Fundamentals of Die Casting Design

Genick BarMeir, Ph. D. 7449 North Washtenaw Ave Chicago, IL 60645 email:barmeir at gmail dot com

Copyright 2009, 2008, 2007, and 1999 by Genick Bar-Meir See the le copying.fdl or copyright.tex for copying conditions. Version (0.1.3 November 9, 2012)

We are like dwarfs sitting on the shoulders of giants

from The Metalogicon by John in 1159

CONTENTS

Nomenclature GNU Free Documentation License . . . . . . . . . . . . . . . . 1. APPLICABILITY AND DEFINITIONS . . . . . . . . . 2. VERBATIM COPYING . . . . . . . . . . . . . . . . . . 3. COPYING IN QUANTITY . . . . . . . . . . . . . . . . 4. MODIFICATIONS . . . . . . . . . . . . . . . . . . . . 5. COMBINING DOCUMENTS . . . . . . . . . . . . . . 6. COLLECTIONS OF DOCUMENTS . . . . . . . . . . . 7. AGGREGATION WITH INDEPENDENT WORKS . . . 8. TRANSLATION . . . . . . . . . . . . . . . . . . . . . 9. TERMINATION . . . . . . . . . . . . . . . . . . . . . 10. FUTURE REVISIONS OF THIS LICENSE . . . . . . . ADDENDUM: How to use this License for your documents CONTRIBUTORS LIST How to contribute to this book . . . . . . . . . . . . . Credits . . . . . . . . . . . . . . . . . . . . . . . . . . Steven from artofproblemsolving.com . . . . . . . Tousher Yang . . . . . . . . . . . . . . . . . . . . Steve Spurgeon . . . . . . . . . . . . . . . . . . . Irene Tan . . . . . . . . . . . . . . . . . . . . . . Your name here . . . . . . . . . . . . . . . . . . . Typo corrections and other minor contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

xv xix xx xxi xxi xxii xxiv xxiv xxv xxv xxv xxv xxvi xxvii xxvii xxvii xxvii xxviii xxviii xxviii xxviii xxviii

Prologue For The POTTO Project xxxi Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxi Why Volunteer? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii

iii

iv

CONTENTS What Has been So Far . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxiii

Prologue For This Book Version 0.1 January 12, 2009 . pages 213 size 1.5M . . . Version 0.0.3 October 9, 1999 pages 178 size 3.2M . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xxxvii . xxxvii . xxxvii . xxxviii . xxxviii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 4 4 5 6 9 10 11 11 12 12 12 14 14 14 21 21 21 23 31 32 33 33 34 36 39 40 41 41 41 45 45 46 46

1 Introduction 1.1 The Importance of Reducing Production Costs . . . 1.2 Designed/Undesigned Scrap/Cost . . . . . . . . . . 1.3 Linking the Production Cost to the Product Design 1.4 Historical Background . . . . . . . . . . . . . . . . 1.5 Numerical Simulations . . . . . . . . . . . . . . . . 1.6 Integral Models . . . . . . . . . . . . . . . . . . 1.7 Summary . . . . . . . . . . . . . . . . . . . . . . . 2 Basic Fluid Mechanics 2.1 Introduction . . . . . . . . . . . . . . . . 2.2 What is uid? Shear stress . . . . . . . . 2.2.1 What is Fluid? . . . . . . . . . . . 2.2.2 What is Shear Stress? . . . . . . . 2.3 Thermodynamics and mechanics concepts 2.3.1 Thermodynamics . . . . . . . . . 2.3.2 Basic Denitions . . . . . . . . . 2.3.3 Momentum Equation . . . . . . . 2.3.4 Compressible ow . . . . . . . . . 2.3.5 Speed of Sound . . . . . . . . . . 2.3.6 Choked Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 Dimensional Analysis 3.0.7 How The Dimensional Analysis Work . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 The Die Casting Process Stages . . . . . . . . . . . . . . 3.2.1 Filling the Shot Sleeve . . . . . . . . . . . . . . . 3.2.2 Plunger Slow Moving Part . . . . . . . . . . . . . 3.2.3 Runner system . . . . . . . . . . . . . . . . . . . . 3.2.4 Die Cavity . . . . . . . . . . . . . . . . . . . . . . 3.2.5 Intensication Period . . . . . . . . . . . . . . . . 3.3 Special Topics . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Is the Flow in Die Casting Turbulent? . . . . . . . 3.3.2 Dissipation eect on the temperature rise . . . . . 3.3.3 Gravity eects . . . . . . . . . . . . . . . . . . . . 3.4 Estimates of the time scales in die casting . . . . . . . . . 3.4.1 Utilizing semi dimensional analysis for characteristic

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . time

CONTENTS 3.5 3.4.2 The ratios of various time scales Similarity applied to Die cavity . . . . . 3.5.1 Governing equations . . . . . . . 3.5.2 Design of Experiments . . . . . Summary of dimensionless numbers . . . Summary . . . . . . . . . . . . . . . . . Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v 54 55 56 59 59 61 61 63 63 63 64 66 66 66 67 67 67 67 68 69 69 71 71 73 73 73 74 75

3.6 3.7 3.8

4 Fundamentals of Pipe Flow 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 4.2 Universality of the loss coecients . . . . . . . . . . 4.3 A simple ow in a straight conduit . . . . . . . . . . 4.3.1 Examples of the calculations . . . . . . . . . 4.4 Typical Components in the Runner andVent Systems 4.4.1 bend . . . . . . . . . . . . . . . . . . . . . . 4.4.2 Y connection . . . . . . . . . . . . . . . . . 4.4.3 Expansion/Contraction . . . . . . . . . . . . 4.5 Putting it all to Together . . . . . . . . . . . . . . . 4.5.1 Series Connection . . . . . . . . . . . . . . . 4.5.2 Parallel Connection . . . . . . . . . . . . . .

5 Flow in Open Channels 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Typical diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Hydraulic Jump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Runner Design 6.1 Introduction . . . . . . . 6.1.1 Backward Design 6.1.2 Connecting runner 6.1.3 Resistance . . . . 7 pQ2 7.1 7.2 7.3 . . . . . . . . . . . . segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7.4

7.5

Diagram Calculations Introduction . . . . . . . . . . . . . . . The common pQ2 diagram . . . . . The validity of the common diagram . 7.3.1 Is the Common Model Valid? 7.3.2 Are the Trends Reasonable? . . 7.3.3 Variations of the Gate area, A3 . The reformed pQ2 diagram . . . . . . 7.4.1 The reform model . . . . . . . . 7.4.2 Examining the solution . . . . . 7.4.3 Poor design eects . . . . . . . 7.4.4 Transient eects . . . . . . . . . Design Process . . . . . . . . . . . . . .

79 . 79 . 80 . 83 . 84 . 86 . 87 . 87 . 88 . 90 . 102 . 102 . 103

vi 7.6 7.7 7.8

CONTENTS The Intensication Consideration . . . . . . . . . . . . . . . . . . . . . 104 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 107 107 108 110 111 111 112 112 112 112 113 113 113 115 116 116 117 117 118 118 118 119 121 123 126 126 127 131 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 133 133 134 135 135 136 137

8 Critical Slow Plunger Velocity 8.1 Introduction . . . . . . . . . . . . . . . 8.2 The common models . . . . . . . . . 8.2.1 Garbers model . . . . . . . . . 8.2.2 Brevicks Model . . . . . . . . . 8.2.3 Brevicks circular model . . . . . 8.2.4 Millers square model . . . . . . 8.3 The validity of the common models . 8.3.1 Garbers model . . . . . . . . . 8.3.2 Brevicks models . . . . . . . . . 8.3.3 Millers model . . . . . . . . . . 8.3.4 EKKs model (numerical model) 8.4 The Reformed Model . . . . . . . . . . 8.4.1 The reformed model . . . . . . . 8.4.2 Design process . . . . . . . . . . 8.5 Summary . . . . . . . . . . . . . . . . . 8.6 Questions . . . . . . . . . . . . . . . . 9 Venting System Design 9.1 Introduction . . . . . . . 9.2 The common models . 9.2.1 Early (etc.) model 9.2.2 Millers model . . 9.3 General Discussion . . . . 9.4 The Analysis . . . . . . . 9.5 Results and Discussion . 9.6 Summary . . . . . . . . . 9.7 Questions . . . . . . . . 10 Density change eects 11 Clamping Force Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

12 Analysis of Die Casting Economy 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . 12.2 The common model, Millers approach . . . . . . 12.3 The validity of Millers price model . . . . . . . . . 12.4 The combined Cost of the Controlled Components 12.5 Die Casting Machine Capital Costs . . . . . . . . . 12.6 Operational Cost of the Die Casting Machine . . . 12.7 Runner Cost (Scrap Cost) . . . . . . . . . . . . . .

CONTENTS 12.8 Startup and Mold Manufacturing Cost 12.9 Personnel Cost . . . . . . . . . . . . . . 12.10Uncontrolled components . . . . . . . . 12.11Minimizing Cost of Single Operation . . 12.12Introduction to Economics . . . . . . . 12.13Summary . . . . . . . . . . . . . . . . . 12.14Question . . . . . . . . . . . . . . . . . A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii 139 140 140 141 144 145 145 147 147 148 149 152 153 156 161 162 162 163 167 171 177 178 180 182 183 184 186 189 190 191 192 195 201 217

Fanno Flow A.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A.2 Fanno Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A.3 NonDimensionalization of the Equations . . . . . . . . . . . . . A.4 The Mechanics and Why the Flow is Choked? . . . . . . . . . . . A.5 The Working Equations . . . . . . . . . . . . . . . . . . . . . . . A.6 Examples of Fanno Flow . . . . . . . . . . . . . . . . . . . . . . A.7 Supersonic Branch . . . . . . . . . . . . . . . . . . . . . . . . . . A.8 Maximum Length for the Supersonic Flow . . . . . . . . . . . . . A.9 Working Conditions . . . . . . . . . . . . . . . . . . . . . . . . . A.9.1 Variations of The Tube Length ( 4f L ) Eects . . . . . . . D A.9.2 The Pressure Ratio, P2 , eects . . . . . . . . . . . . . . . P1 A.9.3 Entrance Mach number, M1 , eects . . . . . . . . . . . . A.10 Practical Examples for Subsonic Flow . . . . . . . . . . . . . . . A.10.1 Subsonic Fanno Flow for Given 4f L and Pressure Ratio . D A.10.2 Subsonic Fanno Flow for a Given M1 and Pressure Ratio . A.11 The Approximation of the Fanno Flow by Isothermal Flow . . . . A.12 More Examples of Fanno Flow . . . . . . . . . . . . . . . . . . . A.13 The Table for Fanno Flow . . . . . . . . . . . . . . . . . . . . . A.14 Appendix Reynolds Number Eects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B What The Establishments Scientists Say B.1 Summary of Referee positions . . . . . . B.2 Referee 1 (from hand written notes) . . B.3 Referee 2 . . . . . . . . . . . . . . . . . B.4 Referee 3 . . . . . . . . . . . . . . . . .

C My Relationship with Die Casting Establishment Bibliography

Index 219 Subjects index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Authors index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

viii

CONTENTS

LIST OF FIGURES

1.1 1.2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15

The prots as a function of the amount of the scrap . . . . . . . . . . Increase of prots as reduction of scrap reduction. . . . . . . . . . . . . The velocity distribution in Couette ow . . . . . . . . . . . . . . . The deformation of uid due to shear stress as progression of time. A very slow moving piston in a still gas. . . . . . . . . . . . . . . . Stationary sound wave and gas moves relative to the pulse. . . . . . Gas ow through a convergingdiverging nozzle. . . . . . . . . . . . The stagnation properties as a function of the Mach number, k=1.4 Various ratios as a function of Mach number for isothermal Nozzle . Rod into the hole example . . . . . . . . . . . . . . . . . . . . . Hydraulic jump in the shot sleeve. . . . . . . . . . . . . . . . . . Filling of the shot sleeve. . . . . . . . . . . . . . . . . . . . . . . Heat transfer processes in the shot sleeve. . . . . . . . . . . . . . Solidication of the shot sleeve time estimates. . . . . . . . . . . Entrance of liquid metal to the runner. . . . . . . . . . . . . . . . Flow in runner when during pressurizing process. . . . . . . . . . . Typical ow pattern in die casting, jet entering into empty cavity. Transition to turbulent ow in instantaneous ow after Wygnanski Flow pattern in the shot sleeve . . . . . . . . . . . . . . . . . . . Two streams of uids into a medium. . . . . . . . . . . . . . . . Schematic of heat transfer processes in the die. . . . . . . . . . . The oscillating manometer for the example 3.1. . . . . . . . . . . Mass Balance on the lest side of the manometer . . . . . . . . . . Rigid body brought into rest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 3 11 13 21 21 23 24 28 33 34 35 36 37 39 39 40 42 42 44 47 50 51 54

ix

x 4.1 4.2 4.3 4.4 4.5 5.1 5.2 6.1 6.2 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13 7.14 7.15 8.1 8.2 8.3 8.4 9.1 9.2 9.3 The results for the ow in a pipe with orice. General simple conduit description. . . . . . . General simple conduit description. . . . . . . A sketch of the bend in die casting. . . . . . A parallel connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 64 65 66 68 69 70 74 75

Equilibrium of Forces in an open channel. . . . . . . . . . . . . . . . . Specic Energy and momentum Curves. . . . . . . . . . . . . . . . . . A geometry of runner connection. . . . . . . . . . . . . . . . . . . . . y connection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Schematic of typical die casting machine. . . . . . . . . . . . A typical trace on a cold chamber machine . . . . . . . . . . pQ2 diagram typical characteristics. . . . . . . . . . . . . . . P as A3 to be relocated . . . . . . . . . . . . . . . . . . . . . Presure of die casting machine. . . . . . . . . . . . . . . . . . P1 as a function of Pmax . . . . . . . . . . . . . . . . . . . . KF as a function of gate area, A3 . . . . . . . . . . . . . . . Die casting characteristics. . . . . . . . . . . . . . . . . . . . Various die casting machine performances . . . . . . . . . . . Reduced pressure performances as a function of Ozer number. Schematic of the plunger and piston balance forces . . . . . . Metal pressure at the plunger tip. . . . . . . . . . . . . . . . . Hydralic piston schematic . . . . . . . . . . . . . . . . . . . . The gate velocity, U3 as a function of the plunger area, A1 . . The reduced power as a function of the normalized ow rate. . A schematic of wave formation in stationary coordinates The two kinds in the sleeve. . . . . . . . . . . . . . . . . A schematic of the wave with moving coordinates . . . . The Froude number as a function of the relative height. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 79 . 80 . 82 . 84 . 85 . 86 . 91 . 93 . 94 . 96 . 96 . 98 . 99 . 101 . 102 . . . . 108 108 109 114

The relative shrinkage porosity as a function of the casting thickness. . 117 A simplied model for the venting system. . . . . . . . . . . . . . . . . 122 The pressure ratios for air and vacuum venting at end. . . . . . . . . . 124

10.1 The control volume of the phase change. . . . . . . . . . . . . . . . . . 127 12.1 Production cost as a function of the runner hydraulic diameter. . . . . . 134 12.2 The reduced power as a function of the normalized ow rate. . . . . . . 137 12.3 Supply and Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 A.1 Control volume of the gas ow in a constant cross section . . . . . . . 147 A.2 Various parameters in Fanno ow as a function of Mach number . . . . 156 A.3 Schematic of Example (A.1) . . . . . . . . . . . . . . . . . . . . . . . 156

LIST OF FIGURES A.4 A.5 A.6 A.7 A.8 A.9 A.10 A.11 A.12 A.13 A.14 A.15 A.16 A.17 A.18 A.19 A.20 A.21 A.22 The schematic of Example (A.2) . . . . . . . . . . . . . . . . The maximum length as a function of specic heat, k . . . . . The eects of increase of 4f L on the Fanno line . . . . . . . D The development properties in of converging nozzle . . . . . . Min and m as a function of the 4f L . . . . . . . . . . . . . . D M1 as a function M2 for various 4f L . . . . . . . . . . . . . D M1 as a function M2 . . . . . . . . . . . . . . . . . . . . . . The pressure distribution as a function of 4f L for a short 4f L D D The pressure distribution as a function of 4f L for a long 4f L D D The eects of pressure variations on Mach number prole . . Mach number as a function of 4f L when the total 4f L = 0.3 D D Schematic of a long tube in supersonic branch . . . . . . . The extra tube length as a function of the shock location . . The maximum entrance Mach number as a function of 4f L . D Unchoked ow showing the hypothetical full tube . . . . . The results of the algorithm showing the conversion rate. . . Solution to a missing diameter . . . . . . . . . . . . . . . . . M1 as a function of 4f L comparison with Isothermal Flow . . D Moody diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi 158 163 164 165 165 167 168 169 170 171 172 173 174 175 178 179 182 183 186

xii

LIST OF FIGURES

LIST OF TABLES

1 1 2.1

Books Under Potto Project . . . . . . . . . . . . . . . . . . . . . . . . xxxiii continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxiv Properties of Various Ideal Gases [300K] . . . . . . . . . . . . . . . . . 19

A.1 Fanno Flow Standard basic Table . . . . . . . . . . . . . . . . . . . . 184 A.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 A.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

xiii

xiv

LIST OF TABLES

NOMENCLATURE

Universal gas constant, see equation (2.36), page 18 Units length., see equation (2.11), page 14

B Bf c Cp Cv EU Eu Ei H h k M n P

Density of the uid, see equation (2.55), page 22 bulk modulus, see equation (2.62), page 22 Body force, see equation (2.19), page 15 Speed of sound, see equation (2.55), page 22 Specic pressure heat, see equation (2.33), page 17 Specic volume heat, see equation (2.32), page 17 Internal energy, see equation (2.13), page 15 Internal Energy per unit mass, see equation (2.16), page 15 System energy at state i, see equation (2.12), page 14 Enthalpy, see equation (2.28), page 17 Specic enthalpy, see equation (2.28), page 17 the ratio of the specic heats, see equation (2.34), page 18 Mach number, see equation (2.64), page 23 The poletropic coecient, see equation (2.60), page 22 Pressure, see equation (2.57), page 22

xv

xvi q Q12 R S U w W12 Energy per unit mass, see equation (2.16), page 15

LIST OF TABLES

The energy transferred to the system between state 1 and state 2, see equation (2.12), page 14 Specic gas constant, see equation (2.37), page 18 Entropy of the system, see equation (2.23), page 16 velocity , see equation (2.14), page 15 Work per unit mass, see equation (2.16), page 15 The work done by the system between state 1 and state 2, see equation (2.12), page 14

The Book Change Log

Version 0.1.3
Nov 8, 2012 (1.9M 265 pages)
Improvements to some of the gures of dimensional analysis chapter (utilizing blender). Add an analysis of the minimum cost ordering supply. The minimum cost ordering refers to the analysis dealing with the minimum cost achieved by nding the optimum number of ordering.

Version 0.1.2
April 1, 2009 (1.9M 263 pages)
Irene Tan provided many English corrections to the dimensional analysis chapter.

Version 0.1.1
Feb 8, 2009 (1.9M 261 pages)
Add Steve Spurgeon (from Dynacast England) corrections to pQ2 diagram. Minor English corrections to pQ2 diagram chapter (unnished). Fix some gures and captions issues. Move to potto style le.

xvii

xviii

LIST OF TABLES

Version 0.1
Jan 6, 2009 (1.6M 213 pages)
Change to modern Potto format. English corrections Finish some examples in Dimensionless Chapter (manometer etc)

Version 0.0.3
Nov 1, 1999 (3.1 M 178 pages)
Initial book of Potto project. Start of economy, dimensional analysis, pQ2 diagram chapters.

Notice of Copyright For This Document:

This document published Modied FDL. The change of the license is to prevent from situations where the author has to buy his own book. The Potto Project License isnt long apply to this document and associated docoments.

GNU Free Documentation License

The modication is that under section 3 copying in quantity should be add in the end. If you print more than 200 copies, you are required to furnish the author with two (2) copies of the printed book. Version 1.2, November 2002 Copyright 2000,2001,2002 Free Software Foundation, Inc. 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.

Preamble
The purpose of this License is to make a manual, textbook, or other functional and useful document free in the sense of freedom: to assure everyone the eective freedom to copy and redistribute it, with or without modifying it, either commercially or noncommercially. Secondarily, this License preserves for the author and

xix

xx

LIST OF TABLES

publisher a way to get credit for their work, while not being considered responsible for modications made by others. This License is a kind of copyleft, which means that derivative works of the document must themselves be free in the same sense. It complements the GNU General Public License, which is a copyleft license designed for free software. We have designed this License in order to use it for manuals for free software, because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does. But this License is not limited to software manuals; it can be used for any textual work, regardless of subject matter or whether it is published as a printed book. We recommend this License principally for works whose purpose is instruction or reference.

1. APPLICABILITY AND DEFINITIONS

This License applies to any manual or other work, in any medium, that contains a notice placed by the copyright holder saying it can be distributed under the terms of this License. Such a notice grants a world-wide, royalty-free license, unlimited in duration, to use that work under the conditions stated herein. The Document, below, refers to any such manual or work. Any member of the public is a licensee, and is addressed as you. You accept the license if you copy, modify or distribute the work in a way requiring permission under copyright law. A Modied Version of the Document means any work containing the Document or a portion of it, either copied verbatim, or with modications and/or translated into another language. A Secondary Section is a named appendix or a front-matter section of the Document that deals exclusively with the relationship of the publishers or authors of the Document to the Documents overall subject (or to related matters) and contains nothing that could fall directly within that overall subject. (Thus, if the Document is in part a textbook of mathematics, a Secondary Section may not explain any mathematics.) The relationship could be a matter of historical connection with the subject or with related matters, or of legal, commercial, philosophical, ethical or political position regarding them. The Invariant Sections are certain Secondary Sections whose titles are designated, as being those of Invariant Sections, in the notice that says that the Document is released under this License. If a section does not t the above denition of Secondary then it is not allowed to be designated as Invariant. The Document may contain zero Invariant Sections. If the Document does not identify any Invariant Sections then there are none. The Cover Texts are certain short passages of text that are listed, as Front-Cover Texts or Back-Cover Texts, in the notice that says that the Document is released under this License. A Front-Cover Text may be at most 5 words, and a Back-Cover Text may be at most 25 words. A Transparent copy of the Document means a machine-readable copy, represented in a format whose specication is available to the general public, that is suitable for revising the document straightforwardly with generic text editors or (for

GNU FREE DOCUMENTATION LICENSE

xxi

images composed of pixels) generic paint programs or (for drawings) some widely available drawing editor, and that is suitable for input to text formatters or for automatic translation to a variety of formats suitable for input to text formatters. A copy made in an otherwise Transparent le format whose markup, or absence of markup, has been arranged to thwart or discourage subsequent modication by readers is not Transparent. An image format is not Transparent if used for any substantial amount of text. A copy that is not Transparent is called Opaque. Examples of suitable formats for Transparent copies include plain ASCII without markup, Texinfo input format, LaTeX input format, SGML or XML using a publicly available DTD, and standard-conforming simple HTML, PostScript or PDF designed for human modication. Examples of transparent image formats include PNG, XCF and JPG. Opaque formats include proprietary formats that can be read and edited only by proprietary word processors, SGML or XML for which the DTD and/or processing tools are not generally available, and the machine-generated HTML, PostScript or PDF produced by some word processors for output purposes only. The Title Page means, for a printed book, the title page itself, plus such following pages as are needed to hold, legibly, the material this License requires to appear in the title page. For works in formats which do not have any title page as such, Title Page means the text near the most prominent appearance of the works title, preceding the beginning of the body of the text. A section Entitled XYZ means a named subunit of the Document whose title either is precisely XYZ or contains XYZ in parentheses following text that translates XYZ in another language. (Here XYZ stands for a specic section name mentioned below, such as Acknowledgements, Dedications, Endorsements, or History.) To Preserve the Title of such a section when you modify the Document means that it remains a section Entitled XYZ according to this denition. The Document may include Warranty Disclaimers next to the notice which states that this License applies to the Document. These Warranty Disclaimers are considered to be included by reference in this License, but only as regards disclaiming warranties: any other implication that these Warranty Disclaimers may have is void and has no eect on the meaning of this License.

2. VERBATIM COPYING
You may copy and distribute the Document in any medium, either commercially or noncommercially, provided that this License, the copyright notices, and the license notice saying this License applies to the Document are reproduced in all copies, and that you add no other conditions whatsoever to those of this License. You may not use technical measures to obstruct or control the reading or further copying of the copies you make or distribute. However, you may accept compensation in exchange for copies. If you distribute a large enough number of copies you must also follow the conditions in section 3. You may also lend copies, under the same conditions stated above, and you may publicly display copies.

3. COPYING IN QUANTITY

xxii

LIST OF TABLES

If you publish printed copies (or copies in media that commonly have printed covers) of the Document, numbering more than 100, and the Documents license notice requires Cover Texts, you must enclose the copies in covers that carry, clearly and legibly, all these Cover Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on the back cover. Both covers must also clearly and legibly identify you as the publisher of these copies. The front cover must present the full title with all words of the title equally prominent and visible. You may add other material on the covers in addition. Copying with changes limited to the covers, as long as they preserve the title of the Document and satisfy these conditions, can be treated as verbatim copying in other respects. If the required texts for either cover are too voluminous to t legibly, you should put the rst ones listed (as many as t reasonably) on the actual cover, and continue the rest onto adjacent pages. If you publish or distribute Opaque copies of the Document numbering more than 100, you must either include a machine-readable Transparent copy along with each Opaque copy, or state in or with each Opaque copy a computer-network location from which the general network-using public has access to download using public-standard network protocols a complete Transparent copy of the Document, free of added material. If you use the latter option, you must take reasonably prudent steps, when you begin distribution of Opaque copies in quantity, to ensure that this Transparent copy will remain thus accessible at the stated location until at least one year after the last time you distribute an Opaque copy (directly or through your agents or retailers) of that edition to the public. It is requested, but not required, that you contact the authors of the Document well before redistributing any large number of copies, to give them a chance to provide you with an updated version of the Document.

4. MODIFICATIONS
You may copy and distribute a Modied Version of the Document under the conditions of sections 2 and 3 above, provided that you release the Modied Version under precisely this License, with the Modied Version lling the role of the Document, thus licensing distribution and modication of the Modied Version to whoever possesses a copy of it. In addition, you must do these things in the Modied Version: A. Use in the Title Page (and on the covers, if any) a title distinct from that of the Document, and from those of previous versions (which should, if there were any, be listed in the History section of the Document). You may use the same title as a previous version if the original publisher of that version gives permission. B. List on the Title Page, as authors, one or more persons or entities responsible for authorship of the modications in the Modied Version, together with at least ve of the principal authors of the Document (all of its principal authors, if it has fewer than ve), unless they release you from this requirement. C. State on the Title page the name of the publisher of the Modied Version, as the publisher.

GNU FREE DOCUMENTATION LICENSE D. Preserve all the copyright notices of the Document.

xxiii

E. Add an appropriate copyright notice for your modications adjacent to the other copyright notices. F. Include, immediately after the copyright notices, a license notice giving the public permission to use the Modied Version under the terms of this License, in the form shown in the Addendum below. G. Preserve in that license notice the full lists of Invariant Sections and required Cover Texts given in the Documents license notice. H. Include an unaltered copy of this License. I. Preserve the section Entitled History, Preserve its Title, and add to it an item stating at least the title, year, new authors, and publisher of the Modied Version as given on the Title Page. If there is no section Entitled History in the Document, create one stating the title, year, authors, and publisher of the Document as given on its Title Page, then add an item describing the Modied Version as stated in the previous sentence. J. Preserve the network location, if any, given in the Document for public access to a Transparent copy of the Document, and likewise the network locations given in the Document for previous versions it was based on. These may be placed in the History section. You may omit a network location for a work that was published at least four years before the Document itself, or if the original publisher of the version it refers to gives permission. K. For any section Entitled Acknowledgements or Dedications, Preserve the Title of the section, and preserve in the section all the substance and tone of each of the contributor acknowledgements and/or dedications given therein. L. Preserve all the Invariant Sections of the Document, unaltered in their text and in their titles. Section numbers or the equivalent are not considered part of the section titles. M. Delete any section Entitled Endorsements. Such a section may not be included in the Modied Version. N. Do not retitle any existing section to be Entitled Endorsements or to conict in title with any Invariant Section. O. Preserve any Warranty Disclaimers. If the Modied Version includes new front-matter sections or appendices that qualify as Secondary Sections and contain no material copied from the Document, you may at your option designate some or all of these sections as invariant. To do this, add their titles to the list of Invariant Sections in the Modied Versions license notice. These titles must be distinct from any other section titles.

xxiv

LIST OF TABLES

You may add a section Entitled Endorsements, provided it contains nothing but endorsements of your Modied Version by various partiesfor example, statements of peer review or that the text has been approved by an organization as the authoritative denition of a standard. You may add a passage of up to ve words as a Front-Cover Text, and a passage of up to 25 words as a Back-Cover Text, to the end of the list of Cover Texts in the Modied Version. Only one passage of Front-Cover Text and one of Back-Cover Text may be added by (or through arrangements made by) any one entity. If the Document already includes a cover text for the same cover, previously added by you or by arrangement made by the same entity you are acting on behalf of, you may not add another; but you may replace the old one, on explicit permission from the previous publisher that added the old one. The author(s) and publisher(s) of the Document do not by this License give permission to use their names for publicity for or to assert or imply endorsement of any Modied Version.

5. COMBINING DOCUMENTS
You may combine the Document with other documents released under this License, under the terms dened in section 4 above for modied versions, provided that you include in the combination all of the Invariant Sections of all of the original documents, unmodied, and list them all as Invariant Sections of your combined work in its license notice, and that you preserve all their Warranty Disclaimers. The combined work need only contain one copy of this License, and multiple identical Invariant Sections may be replaced with a single copy. If there are multiple Invariant Sections with the same name but dierent contents, make the title of each such section unique by adding at the end of it, in parentheses, the name of the original author or publisher of that section if known, or else a unique number. Make the same adjustment to the section titles in the list of Invariant Sections in the license notice of the combined work. In the combination, you must combine any sections Entitled History in the various original documents, forming one section Entitled History; likewise combine any sections Entitled Acknowledgements, and any sections Entitled Dedications. You must delete all sections Entitled Endorsements.

6. COLLECTIONS OF DOCUMENTS
You may make a collection consisting of the Document and other documents released under this License, and replace the individual copies of this License in the various documents with a single copy that is included in the collection, provided that you follow the rules of this License for verbatim copying of each of the documents in all other respects. You may extract a single document from such a collection, and distribute it individually under this License, provided you insert a copy of this License into the extracted document, and follow this License in all other respects regarding verbatim copying of that document.

GNU FREE DOCUMENTATION LICENSE

xxv

7. AGGREGATION WITH INDEPENDENT WORKS

A compilation of the Document or its derivatives with other separate and independent documents or works, in or on a volume of a storage or distribution medium, is called an aggregate if the copyright resulting from the compilation is not used to limit the legal rights of the compilations users beyond what the individual works permit. When the Document is included in an aggregate, this License does not apply to the other works in the aggregate which are not themselves derivative works of the Document. If the Cover Text requirement of section 3 is applicable to these copies of the Document, then if the Document is less than one half of the entire aggregate, the Documents Cover Texts may be placed on covers that bracket the Document within the aggregate, or the electronic equivalent of covers if the Document is in electronic form. Otherwise they must appear on printed covers that bracket the whole aggregate.

8. TRANSLATION
Translation is considered a kind of modication, so you may distribute translations of the Document under the terms of section 4. Replacing Invariant Sections with translations requires special permission from their copyright holders, but you may include translations of some or all Invariant Sections in addition to the original versions of these Invariant Sections. You may include a translation of this License, and all the license notices in the Document, and any Warranty Disclaimers, provided that you also include the original English version of this License and the original versions of those notices and disclaimers. In case of a disagreement between the translation and the original version of this License or a notice or disclaimer, the original version will prevail. If a section in the Document is Entitled Acknowledgements, Dedications, or History, the requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual title.

9. TERMINATION
You may not copy, modify, sublicense, or distribute the Document except as expressly provided for under this License. Any other attempt to copy, modify, sublicense or distribute the Document is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.

10. FUTURE REVISIONS OF THIS LICENSE

The Free Software Foundation may publish new, revised versions of the GNU Free Documentation License from time to time. Such new versions will be similar in spirit to the present version, but may dier in detail to address new problems or concerns. See http://www.gnu.org/copyleft/. Each version of the License is given a distinguishing version number. If the Document species that a particular numbered version of this License or any later

xxvi

LIST OF TABLES

version applies to it, you have the option of following the terms and conditions either of that specied version or of any later version that has been published (not as a draft) by the Free Software Foundation. If the Document does not specify a version number of this License, you may choose any version ever published (not as a draft) by the Free Software Foundation.

ADDENDUM: How to use this License for your documents

To use this License in a document you have written, include a copy of the License in the document and put the following copyright and license notices just after the title page: Copyright YEAR YOUR NAME. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no BackCover Texts. A copy of the license is included in the section entitled GNU Free Documentation License. If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, replace the with...Texts. line with this: with the Invariant Sections being LIST THEIR TITLES, with the FrontCover Texts being LIST, and with the Back-Cover Texts being LIST. If you have Invariant Sections without Cover Texts, or some other combination of the three, merge those two alternatives to suit the situation. If your document contains nontrivial examples of program code, we recommend releasing these examples in parallel under your choice of free software license, such as the GNU General Public License, to permit their use in free software.

CONTRIBUTORS LIST
How to contribute to this book
As a copylefted work, this book is open to revisions and expansions by any interested parties. The only catch is that credit must be given where credit is due. This is a copyrighted work: it is not in the public domain! If you wish to cite portions of this book in a work of your own, you must follow the same guidelines as for any other GDL copyrighted work.

Credits
All entries have been arranged in alphabetical order of surname (hopefully). Major contributions are listed by individual name with some detail on the nature of the contribution(s), date, contact info, etc. Minor contributions (typo corrections, etc.) are listed by name only for reasons of brevity. Please understand that when I classify a contribution as minor, it is in no way inferior to the eort or value of a major contribution, just smaller in the sense of less text changed. Any and all contributions are gratefully accepted. I am indebted to all those who have given freely of their own knowledge, time, and resources to make this a better book!
Date(s) of contribution(s): 1999 to present Nature of contribution: Original author. Contact at: barmeir at gmail.com

Steven from artofproblemsolving.com

Date(s) of contribution(s): June 2005 Nature of contribution: LaTeX formatting, help on building the useful equation and important equation macros.

xxvii

xxviii

CONTRIBUTORS LIST

Tousher Yang
Date(s) of contribution(s): Mat 2008 Nature of contribution: Major review of dimensional analysis and intro chapters.

Steve Spurgeon
Date(s) of contribution(s): November 200x Nature of contribution: Correction to pQ2 diagram derivations.

Irene Tan
Date(s) of contribution(s): January, 2009 Nature of contribution: Repair of dimensional analysis chapter.

Your name here

Date(s) of contribution(s): Month and year of contribution Nature of contribution: Insert text here, describing how you contributed to the book. Contact at: my [email protected]

Typo corrections and other minor contributions

John Joansson English corrections 1999 Adeline Ong English corrections 1999 Robert J. Fermin English corrections 1999 Mary Fran Riley English corrections 1999 Joy Branlund English corrections 1999 Denise Pfeifer English corrections 1999 F. Montery, point to typos in the book 2000. Irene Tan, English correction to Fluid Mechanics chapter 2009.

About This Author

Genick Bar-Meir holds a Ph.D. in Mechanical Engineering from University of Minnesota and a Master in Fluid Mechanics from Tel Aviv University. Dr. Bar-Meir was the last student of the late Dr. R.G.E. Eckert. Much of his time has been spend doing research in the eld of heat and mass transfer (related to renewal energy issues) and this includes uid mechanics related to manufacturing processes and design. Currently, he spends time writing books (there are already three very popular books) and softwares for the POTTO project (see Potto Prologue). The author enjoys to encourage his students to understand the material beyond the basic requirements of exams. Bar-Meirs books are used by hundred of thousands of peoples. His book on compressible is the most popular and preferred by practitioners and students. His books books are used in many universities like Purdue, Caltech, Queens University in Canada, and Singapore. One reason that his books are so popular is that they contain up to date material much of it original work by Bar-Meir. In his early part of his professional life, Bar-Meir was mainly interested in elegant models whether they have or not a practical applicability. Now, this authors views had changed and the virtue of the practical part of any model becomes the essential part of his ideas, books and software. He developed models for Mass Transfer in high concentration that became a building blocks for many other models. These models are based on analytical solution to a family of equations1 . As the change in the view occurred, Bar-Meir developed models that explained several manufacturing processes such the rapid evacuation of gas from containers, the critical piston velocity in a partially lled chamber (related to hydraulic jump), application of supply and demand to rapid change power system and etc. All the models have practical applicability. These models have been extended by several research groups (needless to say with large research grants). For example, the Spanish Comision Interministerial provides grants TAP97-0489 and PB98-0007, and the CICYT and the European Commission provides
1 Where

the mathematicians were able only to prove that the solution exists.

xxix

xxx

CONTRIBUTORS LIST

1FD97-2333 grants for minor aspects of that models. Moreover, the authors models were used in numerical works, in GM, British industry, Spain, and Canada. In the area of compressible ow, it was commonly believed and taught that there is only weak and strong shock and it is continue by PrandtlMeyer function. Bar Meir discovered the analytical solution for oblique shock and showed that there is a quiet buer between the oblique shock and PrandtlMeyer. He also build analytical solution to several moving shock cases. He described and categorized the lling and evacuating of chamber by compressible uid in which he also found analytical solutions to cases where the working uid was ideal gas. The common explanation to PrandtlMeyer function shows that ow can turn in a sharp corner. Engineers have constructed design that based on this conclusion. Bar-Meir demonstrated that common PrandtlMeyer explanation violates the conservation of mass and therefor the turn must be around a nite radius. The authors explanations on missing diameter and other issues in fanno ow and naughty professors question are used in the industry. In his book Basics of Fluid Mechanics, Bar-Meir demonstrated that uids must have wavy surface when two dierent materials ow together. All the previous models for the ooding phenomenon did not have a physical explanation to the dryness. He built a model to explain the ooding problem (two phase ow) based on the physics. He also constructed and explained many new categories for two ow regimes. The author lives with his wife and three children. A past project of his was building a four stories house, practically from scratch. While he writes his programs and does other computer chores, he often feels clueless about computers and programing. While he is known to look like he knows about many things, the author just know to learn quickly. The author spent years working on the sea (ships) as a engine sea ocer but now the author prefers to remain on solid ground.

Prologue For The POTTO Project

Preface
This books series was born out of frustrations in two respects. The rst issue is the enormous price of college textbooks. It is unacceptable that the price of the college books will be over $150 per book (over 10 hours of work for an average student in The United States). The second issue that prompted the writing of this book is the fact that we as the public have to deal with a corrupted judicial system. As individuals we have to obey the law, particularly the copyright law with the innite2 time with the copyright holders. However, when applied to small individuals who are not able to hire a large legal rm, judges simply manufacture facts to make the little guy lose and pay for the defense of his work. On one hand, the corrupted court system defends the big guys and on the other hand, punishes the small entrepreneur who tries to defend his or her work. It has become very clear to the author and founder of the POTTO Project that this situation must be stopped. Hence, the creation of the POTTO Project. As R. Kook, one of this authors sages, said instead of whining about arrogance and incorrectness, one should increase wisdom. This project is to increase wisdom and humility. The POTTO Project has far greater goals than simply correcting an abusive Judicial system or simply exposing abusive judges. It is apparent that writing textbooks especially for college students as a cooperation, like an open source, is a new idea3 . Writing a book in the technical eld is not the same as writing a novel. The writing of a technical book is really a collection of information and practice. There is always someone who can add to the book. The study of technical material isnt only done by having to memorize the material, but also by coming to understand and be able to solve related problems. The author has not found any technique that is more useful for this
2 After the last decision of the Supreme Court in the case of Eldred v. Ashcro (see http://cyber.law.harvard.edu/openlaw/eldredvashcroft for more information) copyrights practically remain indenitely with the holder (not the creator). 3 In some sense one can view the encyclopedia Wikipedia as an open content project (see http://en.wikipedia.org/wiki/Main Page). The wikipedia is an excellent collection of articles which are written by various individuals.

xxxi

xxxii

PROLOGUE FOR THE POTTO PROJECT

purpose than practicing the solving of problems and exercises. One can be successful when one solves as many problems as possible. To reach this possibility the collective book idea was created/adapted. While one can be as creative as possible, there are always others who can see new aspects of or add to the material. The collective material is much richer than any single person can create by himself. The following example explains this point: The army ant is a kind of carnivorous ant that lives and hunts in the tropics, hunting animals that are even up to a hundred kilograms in weight. The secret of the ants power lies in their collective intelligence. While a single ant is not intelligent enough to attack and hunt large prey, the collective power of their networking creates an extremely powerful intelligence to carry out this attack4 . When an insect which is blind can be so powerful by networking, So can we in creating textbooks by this powerful tool.

Why Volunteer?
Why would someone volunteer to be an author or organizer of such a book? This is the rst question the undersigned was asked. The answer varies from individual to individual. It is hoped that because of the open nature of these books, they will become the most popular books and the most read books in their respected eld. For example, the books on compressible ow and die casting became the most popular books in their respective area. In a way, the popularity of the books should be one of the incentives for potential contributors. The desire to be an author of a wellknown book (at least in his/her profession) will convince some to put forth the eort. For some authors, the reason is the pure fun of writing and organizing educational material. Experience has shown that in explaining to others any given subject, one also begins to better understand the material. Thus, contributing to these books will help one to understand the material better. For others, the writing of or contributing to this kind of books will serve as a social function. The social function can have at least two components. One component is to come to know and socialize with many in the profession. For others the social part is as simple as a desire to reduce the price of college textbooks, especially for family members or relatives and those students lacking funds. For some contributors/authors, in the course of their teaching they have found that the textbook they were using contains sections that can be improved or that are not as good as their own notes. In these cases, they now have an opportunity to put their notes to use for others. Whatever the reasons, the undersigned believes that personal intentions are appropriate and are the authors/organizers private aair. If a contributor of a section in such a book can be easily identied, then that contributor will be the copyright holder of that specic section (even within question/answer sections). The books contributors names could be written by their sections. It is not just for experts to contribute, but also students who happened to be doing their homework. The students contributions can be done by adding a question and perhaps the solution. Thus, this method is expected to accelerate the creation of
4 see also in Franks, Nigel R.; Army Ants: A Collective Intelligence, American Scientist, 77:139, 1989 (see for information http://www.ex.ac.uk/bugclub/raiders.html)

WHAT HAS BEEN SO FAR

xxxiii

these high quality books. These books are written in a similar manner to the open source software process. Someone has to write the skeleton and hopefully others will add esh and skin. In this process, chapters or sections can be added after the skeleton has been written. It is also hoped that others will contribute to the question and answer sections in the book. But more than that, other books contain data5 which can be typeset in A LTEX. These data (tables, graphs and etc.) can be redone by anyone who has the time to do it. Thus, the contributions to books can be done by many who are not experts. Additionally, contributions can be made from any part of the world by those who wish to translate the book. It is hoped that the books will be error-free. Nevertheless, some errors are possible and expected. Even if not complete, better discussions or better explanations are all welcome to these books. These books are intended to be continuous in the sense that there will be someone who will maintain and improve the books with time (the organizer(s)). These books should be considered more as a project than to t the traditional denition of plain books. Thus, the traditional role of author will be replaced by an organizer who will be the one to compile the book. The organizer of the book in some instances will be the main author of the work, while in other cases only the gate keeper. This may merely be the person who decides what will go into the book and what will not (gate keeper). Unlike a regular book, these works will have a version number because they are alive and continuously evolving.

What Has been So Far

The undersigned of this document intends to be the organizerauthorcoordinator of the projects in the following areas:
Table -1. Books under development in Potto project.

Compressible Flow Die Casting Dynamics Fluid Mechanics Heat Transfer

Pr og re

Project Name

Remarks

Version

Availability for Public Download     

beta alpha NSY alpha NSY

Based on Eckert

0.4.8.4 0.1 0.0.0 0.1.8 0.0.0

120,000 60,000 15,000 -

Data are not copyrighted.

Number DownLoads

ss

xxxiv

PROLOGUE FOR THE POTTO PROJECT

Table -1. Books under development in Potto project. (continue)

Mechanics Open Channel Flow Statics Strength of Material Thermodynamics Two/Multi ow phases

Pr og

Project Name

Remarks

Version

Availability for Public Download      

NSY NSY early alpha NSY early alpha NSY

rst chapter

0.0.0 0.0.0 0.0.1 0.0.0 0.0.01

Number DownLoads -

re s

TelAvivnotes

0.0.0

NSY = Not Started Yet The meaning of the progress is as:

The Alpha Stage is when some of the chapters are already in a rough draft; in Beta Stage is when all or almost all of the chapters have been written and are at least in a draft stage; in Gamma Stage is when all the chapters are written and some of the chapters are in a mature form; and the Advanced Stage is when all of the basic material is written and all that is left are aspects that are active, advanced topics, and special cases.

The mature stage of a chapter is when all or nearly all the sections are in a mature stage and have a mature bibliography as well as numerous examples for every section. The mature stage of a section is when all of the topics in the section are written, and all of the examples and data (tables, gures, etc.) are already presented. While some terms are dened in a relatively clear fashion, other denitions give merely a hint on the status. But such a thing is hard to dene and should be enough for this stage. The idea that a book can be created as a project has mushroomed from the open source software concept, but it has roots in the way science progresses. However, traditionally books have been improved by the same author(s), a process in which books have a new version every a few years. There are book(s) that have continued after their author passed away, i.e., the Boundary Layer Theory originated6 by Hermann Schlichting but continues to this day. However, projects such as the Linux Documentation project
6 Originally authored by Dr. Schlichting, who passed way some years ago. A new version is created every several years.

WHAT HAS BEEN SO FAR

xxxv

demonstrated that books can be written as the cooperative eort of many individuals, many of whom volunteered to help. Writing a textbook is comprised of many aspects, which include the actual writing of the text, writing examples, creating diagrams and gures, and writing the A LTEX macros7 which will put the text into an attractive format. These chores can be done independently from each other and by more than one individual. Again, because of the open nature of this project, pieces of material and data can be used by dierent books.

7 One can only expect that open source and readable format will be used for this project. But more A than that, only LTEX, and perhaps tro, have the ability to produce the quality that one expects for A these writings. The text processes, especially LTEX, are the only ones which have a cross platform ability to produce macros and a uniform feel and quality. Word processors, such as OpenOce, Abiword, and Microsoft Word software, are not appropriate for these projects. Further, any text that is produced by Microsoft and kept in Microsoft format are against the spirit of this project In that they force spending money on Microsoft software.

xxxvi

PROLOGUE FOR THE POTTO PROJECT

Prologue For This Book

Version 0.1 January 12, 2009

pages 213 size 1.5M
Die casting was my focus of my Ph.D. thesis which admittedly, is not my preferred choice. Dr. Eckert, my adviser, asked me to work on die casting and that is where I developed my knowledge. The rst thing that I have done is a literature review which force me to realize that that there is very little scientic known about how to design the die casting process. I have reviewed works/papers by from of Ohio State University by A. Miller, Brevick, J. Wallace from Case Western, Murry from Australia etc. Scientists are categorized in the following categories, Free thinkers, Cathedral builder, research managers, dust collectors (important work but minor), and thus who should be in science and thus those who are very lucky. This author feel that he, in same sense, very luck that die casting research is infested with thus who should be in science. Like moving from the stone age to modern time, this author is using this book as a tool in his attempt to convert die casting design process to be based on real scientic principles. I have found that the book early version (0.0.3) of the have been downloaded over 50,000. It is strange to me that the fact that many were using the economical part of the book to explain many other the economical problems of large scale manufacturing processes. As I am drifting towards a dierent eld (renewal energy), I still have interest in this material but with dierent aspects will be emphasized. Subjects like Fanno Flow that was as written as appendix will be expanded. Moreover, material like the moving shock issue will be explained and add to process description was omitted in the previous version. While this topic is not directly aecting die casting, the issue of future value will be discussed.

xxxvii

xxxviii

PROLOGUE FOR THIS BOOK

Version 0.0.3 October 9, 1999

pages 178 size 3.2M
This book is the rst and initial book in the series of POTTO project books. This book started as a series of articles to answer both specic questions that I have been asked, as well as questions that I was curious about myself. While addressing these questions, I realized that many commonly held truths about die-casting were scientically incorrect. Because of the importance of these results, I have decided to make them available to the wider community of die-casting engineers. However, there is a powerful group of individuals who want to keep their monopoly over knowledge in the die-casting industry and to prevent the spread of this information.8 Because of this, I have decided that the best way to disseminate this information is to write a book. This book is written in the spirit of my adviser and mentor E.R.G. Eckert. Eckert, aside from his research activity, wrote the book that brought a revolution in the education of the heat transfer. Up to Eckerts book, the study of heat transfer was without any dimensional analysis. He wrote his book because he realized that the dimensional analysis utilized by him and his adviser (for the post doc), Ernst Schmidt, and their colleagues, must be taught in engineering classes. His book met strong criticism in which some called to burn his book. Today, however, there is no known place in world that does not teach according to Eckerts doctrine. It is assumed that the same kind of individual(s) who criticized Eckerts work will criticize this work. As a wise person says dont tell me that it is wrong, show me what is wrong; this is the only reply. With all the above, it must be emphasized that this book is not expected to revolutionize the eld but change some of the way things are taught. The approach adapted in this book is practical, and more handson approach. This statement really meant that the book is intent to be used by students to solve their exams and also used by practitioners when they search for solutions for practical problems. So, issue of proofs so and so are here only either to explain a point or have a solution of exams. Otherwise, this book avoids this kind of issues. This book is divided into two parts. The rst discusses the basic science required by a diecasting engineer; the second is dedicated to die-castingspecic science. The die-casting specic is divided into several chapters. Each chapter is divided into three sections: section 1 describes the commonly believed models; section 2 discusses why this model is wrong or unreasonable; and section 3 shows the correct, or better, way to do the calculations. I have made great eorts to show what existed before science came to die casting. I have done this to show the errors in previous models which make them invalid, and to prove the validity of science. I hope that, in the second edition, none of this will be needed since science will be accepted and will have gained validity in the die casting community. Please read about my battle to get the information out and how the establishment react to it.
8 Please read my correspondence with NADCA editor Paul Bralower and Steve Udvardy. Also, please read the references and my comments on pQ2 .

VERSION 0.0.3 OCTOBER 9, 1999 A Plea for LTEX usage

xxxix

Is it only an accident that both the quality of the typesetting of papers in die casting congress and their technical content quality is so low? I believe there is a connection. A All the major magazines of the the scientic world using TEX or LTEX, why? Because it is very easy to use and transfer (via the Internet) and, more importantly, because it produces high quality documents. NADCA continued to produce text on a low quality word processor. Look for yourself; every transaction is ugly. Linux has liberated the world from the occupation and the control of Microsoft OS. We hope to liberate the NADCA Transaction from such a poor quality word processor. TEX and all (the good ones) supporting programs are free and available every where on the web. There is no reason not to do it. Please join me in improving A NADCAs Transaction by supporting the use of LTEX by NADCA.

Will I Be in Trouble?
Initial part
Many people have said I will be in trouble because I am telling the truth. Those with a vested interest in the status quo (North American Die Casting Association, and thus research that this author exposed there poor and or erroneous work). will try to use their power to destroy me. In response, I challenge my opponents to show that they are right. If they can do that, I will stand wherever they want and say that I am wrong and they are right. However, if they cannot prove their models and practices are based on solid scientic principles, nor nd errors with my models (and I do not mean typos and English mistakes), then they should accept my results and help the diecasting industry prosper. People have also suggested that I get life insurance and/or good lawyer because my opponents are very serious and mean business; the careers of several individuals are in jeopardy because of the truths I have exposed. If something does happen to me, then you, the reader, should punish them by supporting science and engineering and promoting the diecasting industry. By doing so, you prevent them from manipulating the industry and gaining additional wealth. For the sake of my family, I have, in fact, taken out a life insurance policy. If something does happen to me, please send a thank you and work well done card to my family.

The Continued Struggle

It was exposed that second reviewer that appear in this book is Brevick from Ohio. It strange that in a dierent correspondence he say that he cannot wait to get this author futur work. This part is holding for some juicy details.

xl

PROLOGUE FOR THIS BOOK

How This Book Was Written

This book started because I was frustrated with the system that promote erroneous research. Then, I realized that the book cannot be stolen if it under open content. The die casting process is interesting enough to insert my contributions. I have found that works or model in this area are lack of serious scientic principles. I have started to write class notes to my clients and I add my research work to create this book. During the writing I add the material on economy which I felt is missing piece of knowledge in the die casting engineering world. Of course, this book was written on Linux (Micro$oftLess book). This book was written using the vim editor for editing (sorry never was able to be comfortable with emacs). The graphics were done by TGIF, the best graphic program that this author experienced so far. The gures were done by grap but will be modied to gle. The spell checking was done by ispell, and hope to nd a way to use gaspell, a program that currently cannot be used on new Linux systems. The gure in cover page was created by Genick Bar-Meir, and is copylefted by him.

xli

xlii

PROLOGUE FOR THIS BOOK

Abstract

Die-casting engineers have to compete not only with other die-casting companies, but also against other industries such as plastics, and composite materials. Clearly, the black art approach, which has been an inseparable part of the engineers tools, is in need of being replaced by a scientic approach. Excuses that science has not and never will work need to be replaced with science does work. All technologies developed in recent years are described in a clear, simple manner in this book. All the errors of the old models and the violations of physical laws are shown. For example, the common pQ2 diagram violates many physical laws, such as the rst and second laws of thermodynamics. Furthermore, the common pQ2 diagram produces trends that are the opposite of reality, which are described in this book. The die casting engineers job is to produce maximum prots for the company. In order to achieve this aim, the engineer must design high quality products at a minimum cost. Thus, understanding the economics of the die casting design and process are essential. These are described in mathematical form for the rst time in this volume. Many new concepts and ideas are also introduced. For instance, how to minimize the scrap/cost due to the runner system, and what size of die casting machine is appropriate for a specic project. The die-casting industry is undergoing a revolution, and this book is part of it. One reason (if one reason can describe the situation) companies such as Doehler Jorvis (the biggest die caster in the world) and Shelby are going bankrupt is that they do not know how to calculate and reduce their production costs. It is my hope that die-casters will turn such situations around by using the technologies presented in this book. I believe this is the only way to keep the die casting professionals and the industry itself, from being left in the dust.

xliii

xliv

PROLOGUE FOR THIS BOOK

Preface
"In the beginning, the POTTO project was without form, and void; and emptiness was upon the face of the bits and files. And the Fingers of the Author moved upon the face of the keyboard. And the Author said, Let there be words, and there were words." 9 .

This book, Fundamentals of Die Casting Design, describes the fundamentals of die casting process design and economics for engineers and others. This book is designed to ll the gap and the missing book on economy and scientic principles of die casting. It is hoped that the book could be used as a reference book for people who have at least some basics knowledge of science areas such as calculus, physics, etc. It has to realized the some material is very advance and required knowledge of uid mechanics particularly compressible ow and open channel ow. This authors popular book on compressible ow should provide the introductory in that area. The readers reactions to this book and the usage of the book as a textbook suggested that the chapter which deals with economy should be expand. In the following versions this area will strength and expended. The structure of this book is such that many of the chapters could be usable independently. For example, if you need information about, say, economy of the large scale productions, you can read just chapter (12). I hope this makes the book easier to use as a reference manual. However, this manuscript is rst and foremost a textbook, and secondly a reference manual only as a lucky coincidence. I have tried to describe why the theories are the way they are, rather than just listing seven easy steps for each task. This means that a lot of information is presented which is not necessary for everyone. These explanations have been marked as such and can be skipped. Reading everything will, naturally, increase your understanding of the many aspects of uid mechanics.
9 To

the power and glory of the mighty God. This book is only to explain his power.

xlv

xlvi

PROLOGUE FOR THIS BOOK

This book is written and maintained on a volunteer basis. Like all volunteer work, there is a limit on how much eort I was able to put into the book and its organization. Moreover, due to the fact that English is my third language and time limitations, the explanations are not as good as if I had a few years to perfect them. Nevertheless, I believe professionals working in many engineering elds will benet from this information. This book contains many worked examples, which can be very useful for many. I have left some issues which have unsatisfactory explanations in the book, marked with a Mata mark. I hope to improve or to add to these areas in the near future. Furthermore, I hope that many others will participate of this project and will contribute to this book (even small contributions such as providing examples or editing mistakes are needed). I have tried to make this text of the highest quality possible and am interested in your comments and ideas on how to make it better. Incorrect language, errors, ideas for new areas to cover, rewritten sections, more fundamental material, more mathematics (or less mathematics); I am interested in it all. I am particularly interested in the best arrangement of the book. If you want to be involved in the editing, graphic design, or proofreading, please drop me a line. You may contact me via Email at barmeir at gmail dot com. Naturally, this book contains material that never was published before (sorry cannot avoid it). This material never went through a close content review. While close content peer review and publication in a professional publication is excellent idea in theory. In practice, this process leaves a large room to blockage of novel ideas and plagiarism. For example, Brevick from Ohio State is one the individual who attempt to block this author idea on pQ2 diagram. If you would like to critic to my new ideas please send me your comment(s). However, please do not hide your identity, it will cloud your motives. Several people have helped me with this book, directly or indirectly. I would like to especially thank to my adviser, Dr. E. R. G. Eckert, whose work was the inspiration for this book. I also would like to thank to Jannie McRotien (Open Channel Flow chapter) and Tousher Yang for their advices, ideas, and assistance. A I encourage anyone with a penchant for writing, editing, graphic ability, LTEX knowledge, and material knowledge and a desire to provide open content textbooks and to improve them to join me in this project. If you have Internet e-mail access, you can contact me at [email protected].

CHAPTER 1 Introduction
In the recent years, many die casting companies have gone bankrupt (DoehlerJarvis and Shelby to name a few) and many other die casting companies have been sold (St. Paul Metalcraft, Tool Products, OMC etc.). What is/are the reason/s for this situation? Some blame poor management. Others blame bad customers (which is mostly the automobile industry). Perhaps there is something to these claims. Nevertheless, one can see that the underlying reasons are the missing knowledge of how to calculate if there are prots for a production line and how to design, so that costs will be minimized. To demonstrate how the absurd situation is the fact that there is not even one company today that can calculate the actual price of any product that they are producing. Moreover, if a company is able to produce a specic product, no one in that company looks at the redesign (mold or process) in order to reduce the cost systematically. In order to compete with other industries and other companies, the die casting industry must reduce the cost as much as possible (20% to 40%) and lead time signicantly (by 1/2 or more). To achieve these goals, the engineer must learn to connect mold design to the cost of production (charged to the customer) and to use the correct scientic principals involved in the die casting process to reduce/eliminate the guess work. This book is part of the revolution in die casting by which science is replacing the black art of design. For the rst time, a link between the cost and the design is spelled out. Many new concepts, based on scientic principles, are introduced. The old models, which was plagued by the die casting industry for many decades, are analyzed, their errors are explained and the old models are superseded. Science is good, but it is not useful in the oor of our plant!! George Reed, the former president of SDCE, in 1999 announced in a meeting in the local chapter (16) of NADCA. He does not believe that there is A relationship between science and what he does with the die casting machine. He said that because he does not follow NADCA recommendations, he achieves good castings. For instance, he

CHAPTER 1. INTRODUCTION

stated that the common and NADCA supported, recommendation in order to increase the gate velocity, plunger diameter needs to be decreased. He said that because he does not follow this recommendation, and/or others, that is the reason his succeeds in obtaining good castings. He is right and wrong. He is right not to follow the NADCA recommendations since they violate many basic scientic principles. One should expect that models violating scientic principles would produce unrealistic results. When such results occur, this should actually strengthen the idea that science has validity. The fact that models which appear in books today are violating scientic principals and therefore do not work should actually convince him, and others, that science does have validity. Mr. Reed is right (in certain ranges) to increase the diameter in order to increase the gate velocity as will be covered in Chapter 7. The above example is but one Profits of many of models that are errant and % in need of correction. To this date, the 23 author has not found so much as a sin0 gle commonly used model that has been correct in its conclusions, trends, and/or assumptions. The wrong models/methods that have plagued the in~65-75 78 80 dustry are: 1) critical slow plunger velocity, 2) pQ2 diagram, 3) plunger diameter calculations, 4) runner system design, 5) vent system design, etc These incorrect Fig. -1.1. The prots as a function of the amount models are the reasons that science of the scrap. does not work. The models presented in this book are here for the purpose of answering the questions of design in a scientic manner which will result in reduction of costs and increased product quality. Once the reasons to why science does not work are clear, one should learn the correct models for improving quality, reducing lead time and reducing production cost. The main underlying reason people are in the die casting business is to make money. One has to use science to examine what the components of production cost/scrap are and how to minimize or eliminate each of them to increase protability. The underlying purpose of this book is to help the die caster to achieve this target.
23 20 maximum profits invesment cost breakeven point no scrap 0 100 scrap 100 0

1.1 The Importance of Reducing Production Costs

Contrary to popular belief, a reduction of a few percentage points of the production cost/scrap does not translate into the same percentage of increase in prots. The increase is a little bit more complicated function. To study the relationship further, see Figure 1.1 where prots are plotted as a function of the scrap. A linear function describes the relationship, when the secondary operations are neglected. The maximum loss occurs when all the material turned out to be scrap and it is referred to as the investment cost. On the other hand, maximum prots occur when all the material becomes products (no scrap of any kind (see Figure 1.1). The breakeven point (BEP)

1.1. THE IMPORTANCE OF REDUCING PRODUCTION COSTS

has to exist somewhere between these two extremes. Typically, for the die casting industry, the breakeven point lies within the range of 55%75% product (or 25%-35% scrap). Typical prots in the die casting industry are or should be about 20%. When the prots fails below 15% or typical prot in the stock exchange then the production should stop. From Figure 1.1 it can be noticed that relative change in prots% = new product percent BEP 1 100 old product percent BEP (1.1)

Example 1.1: What would be the eect on the prots of a small change (2%) in a amount of scrap for a job with 22% scrap (78% product) and with breakeven point of 65%? Solution 80 65 1 100 = 15.3% 78 65 A reduction of 2% in a amount of the scrap to be 20% (80% product) results in increase of more than 15.3% in the prots.
End Solution

This is a very substantial dierence. Therefore, a much bigger reduction in scrap will result in much, much bigger prots.
Scrap Cost
400
BEP = 50.0

Scrap Cost
BEP = 50.0

140 350
BEP = 55.0 BEP = 55.0

120 300
BEP = 60.0

Incrase of prots

Incrase of prots

100

BEP = 60.0

250

Old Scrap = 10%

BEP = 65.0

Old Scrap = 20%

80
BEP = 65.0

200

BEP = 70.0

150

60
BEP = 70.0

100

40

50

20

0 50 60 70 80 90 100 110 120

0 50 60 70 80 90 100 110 120

Scrap percent
December 8, 2008 December 8, 2008

Scrap percent

Fig a.

For BEP= 10%

Fig b.

For BEP= 10%

Fig. -1.2. The left graph depicts the increase of prots as reduction of the scrap for 10% + BEP. The right graph depicts same for for 20% + BEP.

To analysis this point further Figure intro:g:scrapCostBEP is built for two old scrap values, 10% more than the BEP on the left and 20% more than BEP on the right. The two gures (left and right) in Figure intro:g:scrapCostBEP demonstrate that The higher BEP the change in reduction of scrap is more important. The lower the old scrap point is the more important the reduction is.

CHAPTER 1. INTRODUCTION

1.2 Designed/Undesigned Scrap/Cost

There can be many denitions of scrap. The best denition suited to the die casting industry should be dened as all the metal that did not become a product. There are two kinds of scrap/cost: 1) those that can be eliminated, and 2) those that can only be minimized. The rst kind is referred to here as the undesigned scrap and the second is referred as designed scrap. What is the dierence? It is desired not to have rejection of any part (the rejection should be zero) and of course it is not designed. Therefore, this is the undesigned scrap/cost. However, it is impossible to eliminate the runner completely and it is desirable to minimize its size in such a way that the cost will be minimized. This minimization of cost and this minimum scrap is the designed scrap/cost. The die casting engineer must distinguish between these two scrap components in order to be able to determine what should be done and what cannot be done. Science can make a signicant dierence; for example, it is possible to calculate the critical slow plunger velocity and thereby eliminating (almost) air entrainment in the shot sleeve in order to minimize the air porosity. This means that air porosity will be reduced and marginal products (even poor products in some cases) are converted into good quality products. In this way, the undesigned scrap can be eliminated or minimized. Additional way of minimizing the scrap is changing several parameters. The minimum scrap/cost can be achieved when a combination of the smallest runner volume and the cheapest die casting machine are selected for a single cavity. Similar analysis can be done for multiply cavity molds. This topic will be studied further in Chapter 12. The possibility that a parameter, which reduces the designed scrap/cost will, at the same time, reduce the undesigned scrap/cost. An example of such a parameter is the venting system design. It will be shown that there is a critical design below which air/gas is exhausted easily and above which air is trapped. In the later case, the air/gas pressure builds up and results in a poor casting (large amount of porosity) The meaning of the critical design and above and below critical design will be presented in Chapter 9. The analysis of the vent system demonstrates that a design much above the critical design and design just above the critical design yielding has almost the same results small amount of air entrainment. One can design the vent just above the critical design so the design scrap/cost is reduced to a minimum amount possible. Now both targets have been achieved: less rejections (undesigned scrap) and less vent system volume (designed scrap). It also possible to have an opposite case in which reduction of designed scrap results in poor design. The engineer has to be aware of these points.

1.3 Linking the Production Cost to the Product Design

It is sound accounting practice to tie the cost of every aspect of production to the cost to be charged to the customer. Unfortunately, the practice today is such that the price of the products are determined by some kind of average based on the part weight plus geometry and not on the actual design and production costs. Furthermore, this idea is also perpetuated by researchers who do not have any design factor [14]. Here it is advocated to price according to the actual design and production costs. It

1.4. HISTORICAL BACKGROUND

is believed that better pricing results from such a practice. In todays practice, even after the project is nished, no one calculates the actual cost of production, let alone calculating the actual prots. The consequences of such a practice are clear: it results in no push for better design and with no idea which jobs make prots and which do not. Furthermore, considerable nancial cost is incurred which could easily be eliminated. Several chapters in this book are dedicated to linking the design to the cost (end-price).

1.4 Historical Background

Die casting is, relatively speaking a very forgiving process, in which after tinkering with the several variables one can obtain a medium quality casting. For this reason there has not been any real push toward doing good research. Hence, all the major advances in the understanding of the die casting process were not sponsored by any of die casting institutes/associations. Many of the people in important positions in the die casting industry suer from what is known as the Detroit attitude, which is very dicult to change. We are making a lot of money so why change? and if do not the Government will pay for it.. Moreover, the controlling personnel on the research funds believe that the die casting is a metallurgical manufacturing process and therefore, the research has to be carried out by either Metallurgical Engineers or Industrial Engineers. Furthermore, should come as no surprise that peopleincharge of the research funding fund their own research. One cannot wonder if there is a relationship between so many erroneous models which have been produced and the personnel controlling the research funding. A highlight of the major points of the progress of the understanding is described herein. The vent system design requirements were studied by some researchers, for example Suchs, Veinik, and Draper and others. These models, however, are unrealistic and do not provide no relation to the physics or realistic picture of the real requirements or of the physical situation since they ignore the major point, the air compressibility. However this research extremely poor, it highlights the idea that venting design is a must. One of the secrets of the black art of design was that there is a range of gate velocity which creates good castings depending on the alloy properties being casted. The existence of a minimum velocity hints that a signicant change in the liquid metal ow pattern occurs. Veinik linked the gate velocity to the ow pattern (atomization) and provide a qualitative physical explanation for this occurrence. Experimental work [25] showed that liquid metals, like other liquids, ow in three main patterns: a continuous ow jet, a coarse particle jet, and an atomized particle jet. Other researchers utilized the water analogy method to study ow inside the cavity for example, [6]. At present, the (minimum) required gate velocity is supported by experimental evidence which is related to the ow patterns. However, the numerical value is unknown because the experiments were poorly conducted for example, [30] the dierential equations that have been solved are not typical to die casting. Discussion about this poor research is presented in Chapter 3. At this stage, this question is not understood. In the late 70s, an Australian group [12] suggested adopting the pQ2 diagram for die casting in order to calculate the gate velocity, the gate area, and other parameters.

CHAPTER 1. INTRODUCTION

As with all the previous models they missed the major points of the calculations. As will be shown in Chapter 7, the Australians model produce incorrect results and predict trends opposite to reality. This model took root in die casting industry for the last 25 years. Yet, one can only wonder why this well established method (the supply and demand theory which was build by Fanno (the brother of other famous Fanno from Fanno ow), which was introduced into uid mechanics in the early of this century, reached the die casting only in the late 70s and was then erroneously implemented. This methods now properly build for the rst time for the die casting industry in this book. Until the 1980s there was no model that assisted the understanding air entrapment in the shot sleeve. Garber described the hydraulic jump in the shot sleeve and called it the wave, probably because he was not familiar with this research area. He also developed the erroneous model which took root in the industry in spite the fact that it never works. One can only wonder why any die casting institutes/associations have not published this fact. Moreover, NADCA and other institutes continue to funnel large sums of money to the researchers (for example, Brevick from Ohio State) who used Garbers model even after they knew that Garbers model was totally wrong. The turning point of the understanding was when Prof. Eckert, the father of modern heat transfer, introduced the dimensional analysis applied to the die casting process. This established a scientic approach which provided an uniform schemata for uniting experimental work with the actual situations in the die casting process. Dimensional analysis demonstrates that the uid mechanics processes, such as lling of the cavity with liquid metal and evacuation/extraction of the air from the mold, can be dealt when the heat transfer is assumed to be negligible. However, the uid mechanics has to be taken into account in the calculations of the heat transfer process (the solidication process). This proved an excellent opportunity for simple models to predict the many parameters in the die casting process, which will be discussed later in this book. Here, two examples of new ideas that mushroomed in the inspiration of prof. Eckerts work. It has been shown that [5] the net eect of the reactions is negligible. This fact is contradictory to what was believed at that stage. The development of the critical vent area concept provided the major guidance for 1) the designs to the venting system, and 2) criterion when the vacuum system needs to be used. In this book, many of the new concepts and models, such as economy of the runner design, plunger diameter calculations, minimum runner design, etc, are described for the rst time.

1.5 Numerical Simulations

Numerical simulations have been found to be very useful in many areas which lead many researchers attempting to implement them into die casting process. Considerable research work has been carried out on the problem of solidication including uid ow which is known also as Stefan problems [21]. Minaie et al in one of the pioneered work use this knowledge and simulated the lling and the solidication of the cavity using nite dierence method. Hu et al used the nite element method to improve the grid

1.5. NUMERICAL SIMULATIONS

problem and to account for atomization of the liquid metal. The atomization model in the last model was based on the mass transfer coecient. This model atomization is not appropriate. Clearly, this model is in waiting to be replaced by a realistic model to describe the mass transfer1 . The Enthalpy method was further exploded by Swaminathan and Voller and others to study the lling and solidication problem. While numerical simulation looks very promising, all the methods (nite dierence, nite elements, or boundary elements etc) 2 suer from several major drawbacks which prevents them from yielding reasonable results.
There is no theory (model) that explains the heat transfer between the mold walls and the liquid metal. The lubricant sprayed on the mold change the characteristic of the heat transfer. The dierence in the density between the liquid phase and solid phase creates a gap during the solidication process between the mold and the ingate which depends on the geometry. For example, Osborne et al showed that a commercial software (MAGMA) required ddling with the heat transfer coecient to get the numerical simulation match the experimental results3 . As it was mentioned earlier, it is not clear when the liquid metal ows as a spray and when it ows as continuous liquid. Experimental work has demonstrated that the ow, for a large part of the lling time, is atomized [4]. The pressure in the mold cavity in all the commercial codes are calculated without taking into account the resistance to the air ow out. Thus, builtup pressure in the cavity is poorly estimated, or even not realistic, and therefore the characteristic ow of the liquid metal in the mold cavity is poorly estimated as well. The ow in all the simulations is assumed to be turbulent ow. However, time and space are required to achieved a fully turbulent ow. For example, if the ow at the entrance to a pipe with the typical conditions in die casting is laminar (actually it is a plug ow) it will take a runner with a length of about 10[m] to achieved fully developed ow. With this in mind, clearly some part of the ow is laminar. Additionally, the solidication process is faster compared to the dissipation process in the initial stage, so it is also a factor in changing the ow from a turbulent (in case the ow is turbulent) to a laminar ow. The liquid metal velocity at the entrance to the runner is assumed for the numerical simulation and not calculated. In reality this velocity has to be calculated utilizing the pQ2 diagram. If turbulence exists in the ow eld, what is the model that describes it adequately? Clearly, model such k are based on isentropic homogeneous with mild change in the properties cannot describe situations where the ow changes into two-phase ow (solid-liquid ow) etc.
nds that it is the easiest to critic ones own work or where he/she was involved. or academic versions. 3 Actually, they attempted to prove that the software is working very well. However, the fact that coecient need to eld is excellent proof why this work is meaningless.
2 Commercial 1 One

CHAPTER 1. INTRODUCTION
The heat extracted from the die is done by cooling liquid (oil or water). In most models (all the commercial models) the mechanism is assumed to be by regular cooling. In actuality, some part of the heat is removed by boiling heat transfer. The governing equations in all the numerical models, that I am aware of, neglect the dissipation term in during the solidication. The dissipation term is the most important term in that case.

One wonders how, with unknown ow pattern (or correct ow pattern), unrealistic pressure in the mold, wrong heat removal mechanism (cooling method), erroneous governing equation in the solidication phase, and inappropriate heat transfer coecient, a simulation could produce any realistic results. Clearly, much work is need to be done in these areas before any realistic results should be expected from any numerical simulation. Furthermore, to demonstrate this point, there are numerical studies that assume that the ow is turbulent, continuous, no air exist (or no air leaving the cavity) and proves with their experiments that their model simulate reality [23]. On the other hand, other numerical studies assumed that the ow does not have any eect on the solidication and of course have their experiments to support this claim [11]. Clearly, this contradiction suggest several options:
Both of the them are right and the model itself does not matter. One is right and the other one is wrong. Both of them are wrong.

The third research we mentioned here is an example where the calculations can be shown to be totally wrong and yet the researchers have experimental proofs to back them up. Viswanathan et al studied a noble process in which the liquid metal is poured into the cavity and direct pressure is applied to the cavity. In their calculations the authors assumed that metal enter to the cavity and ll the whole entrance (gate) to the cavity. Based on this assumption their model predict defects in certain geometry. A critical examination of this model present the following. The assumption of no air ow out by the authors (was explained privately that air amount is a small and therefore not important) is very critical as will be shown here. The volumetric air ow rate into the cavity has to be on average equal to liquid metal ow rate (conservation of volume for constant density). Hence, air velocity has to be approximately innite to achieve zero vent area. Conversely, if the assumption that the air ows in the same velocity as the liquid entering the cavity, liquid metal ow area is a half what is assume in the researchers model. In realty, the ow of the liquid metal is in the two phase region and in this case, it is like turning a bottle full of water over and liquid inside ows as blobs 4 . More information can be found on reversible ow in this author book in Potto series of Basics of Fluid Mechanics. In this case the whole calculations do not have much to do with reality since the velocity is not continuous and dierent from what was calculated.
4 Try

it your self! ll a bottle and turn it upside and see what happens.

1.6. INTEGRAL MODELS

Another example of such study is the model of the ow in the shot sleeve by Backer and Sant from EKK [2]5 . The researchers assumed that the ow is turbulent and they justied it because they calculated and found a jet with extreme velocity. Unfortunately, all the experimental evidence demonstrate that there is no such jet [24]. It seems that this jet results from the poor boundary and initial conditions6 . In the presentation, the researchers also stated that results they obtained for laminar and turbulent ow were the same7 while a simple analysis can demonstrate the dierence is very large. Also, one can wonder how liquid with zero velocity to be turbulent. With these results one can wonder if the code is of any value or the implementation is at fault. The bizarre belief that the numerical simulations are a panacea to all the design problem is very popular in the die casting industry. Any model has to describe and account for the physical situation in order to be useful. Experimental evidence which is supporting wrong models as a real evidence is nonsense. Clearly some wrong must be there. For example, see the paper by Murray and colleague in which they use the fact that two unknown companies (somewhere in the outer space maybe?) were using their model to claim that it is correct.. A proper way can be done by numerical calculations based on real physics principles which produce realistic results. Until that point come, the reader should be suspicious about any numerical model and its supporting evidence.8

1.6 Integral Models

Unfortunately, the numerical simulations of the liquid metal ow and solidication process do not yield reasonable results at the present time. This problem has left the die casting engineers with the usage of the integral approach method. In this method the calculations are broken into simplied models. One of the most important tool in this approach is the pQ2 diagram, one of the manifestations of the supply and demand theory. In this diagram, an engineer insures that die casting machine ability can fulll the die mold design requirements; the liquid metal is injected at the right velocity range and the lling time is small enough to prevent premature freezing. One can, with the help of the pQ2 diagram and by utilizing experimental values for desired lling time and gate velocities improve the quality of the casting. The gate velocity has to be above a certain value to assure atomization and below a critical value to prevent erosion of the mold. This two values are experimental and no reliable theory is available today known. The correct model for the pQ2 diagram has been developed and will be presented in Chapter 7. A byproduct of the above model is the plunger diameter calculations and it is discussed in Chapter 7.
5 It was suggested by several people that the paper was commissioned by NADCA to counter BarMeirs equation to shot sleeve. This fact is up to the reader to decide if it is correct. 6 The boundary and initial conditions were not spelled out in the paper!! However they were implicitly stated in the presentation. 7 So why to use the complicate turbulent model? 8 With all these harsh words, I would like to take the opportunity for the record, I do think that work by Daveys group is a good one. They have inserted more physics (for example the boiling heat transfer) into their models which I hope in the future, leads us to have realistic numerical models.

10

CHAPTER 1. INTRODUCTION

It turned out that many of the design parameters in die casting have a critical point above which good castings are produced and below which poor castings are produced. Furthermore, much above and just above the critical point do not change much the quality but costs much more. This fact is where the economical concepts plays a signicant role. Using these concepts, one can increase the protability signicantly, and obtain very good quality casting and reduce the leading time. Additionally, the main cost components like machine cost and other are analyzed which have to be taken into considerations when one chooses to design the process will be discussed in the Chapter 12. Porosity can be divided into two main categories; shrinkage porosity and gas/air entrainment. The porosity due to entrapped gases constitutes a large part of the total porosity. The creation of gas/air entrainment can be attributed to at least four categories: lubricant evaporation (and reaction processes9 ), vent locations (last place to be lled), mixing processes, and vent/gate area. The eects of lubricant evaporation have been found to be insignicant. The vent location(s) can be considered partially solved since only qualitative explanation exist. The mixing mechanisms are divided into two zones: the mold, and the shot sleeve. Some mixing processes have been investigated and can be considered solved. The requirement on the vent/gate areas is discussed in Chapter 9. When the mixing processes are very signicant in the mold, other methods are used and they include: evacuating the cavities (vacuum venting), Pore Free Technique (in zinc and aluminum casting) and squeeze casting. The rst two techniques are used to extract the gases/air from the shot sleeve and die cavity before the gases have the opportunity to mix with the liquid metal. The squeeze casting is used to increase the capillary forces and therefore, to minimize the mixing processes. All these solutions are cumbersome and more expensive and should be avoided if possible. The mixing processes in the runners, where the liquid metal ows vertically against gravity in relatively large conduit, are considered to be insignicant10 . The enhanced air entrainment in the shot sleeve is attributed to operational conditions for which a blockage of the gate by a liquid metal wave occurs before the air is exhausted. Consequently, the residual air is forced to be mixed into the liquid metal in the shot sleeve. With Bar-Meirs formula, one can calculate the correct critical slow plunger velocity and this will be discussed in Chapter 8.

1.7 Summary
It is an exciting time in the die casing industry because for the rst time, an engineer can start using real science in designing the runner/mold and the die casting process. Many new models have been build and many old techniques mistake have been removed. It is the new revolution in the die casting industry.

9 Some researchers view the chemical reactions (e.g. release of nitrogen during solidication process) as category by itself. 10 Some work has been carried out and hopefully will be published soon. And inside, in the book Basic of Fluid Mechanics in the two phase chapter some inside was developed.

CHAPTER 2 Basic Fluid Mechanics

2.1 Introduction
This chapter is presented to ll the void in basic uid mechanics to the die casting community. It was observed that knowledge in this area cannot be avoided. The design of the process as well as the properties of casting (especially magnesium alloys) are determined by the uid mechanics/heat transfer processes. It is hoped that others will join to spread this knowledge. There are numerous books for introductory uid mechanics but the Potto series book Basic of Fluid Mechanics is a good place to start. This chapter is a summary of that book plus some pieces from the Fundamentals of Compressible Flow Mechanics. It is hoped that the reader will nd this chapter interesting and will further continue expanding his knowledge by reading the full Potto books on uid mechanics and compressible ow. First we will introduce the nature of U uids and basic concepts from thermodynamics. Later the integral analysis will be discussed in which it will be divided into introduction of the control volume concept and Continuity equations. The energy equation will be explained in the next section. Later, the momentum equa- Fig. -2.1. The velocity distribution in Couette tion will be discussed. Lastly, the chapter ow will be dealing with the compressible ow gases. Here it will be refrained from dealing with topics such boundary layers, non viscous ow, machinery ow etc which are not essential to understand the rest of this book. Nevertheless, they are important and it is advisable that the reader will read on these topics as well.

11

12

CHAPTER 2. BASIC FLUID MECHANICS

2.2 What is uid? Shear stress

Fluid, in this book, is considered as a substance that moves continuously and permanently when exposed to a shear stress. The liquid metals are an example of such substance. However, the liquid metals do not have to be in the liquidus phase to be considered liquid. Aluminum at approximately 4000 C is continuously deformed when shear stress are applied. The whole semisolid die casting area deals with materials that looks solid but behaves as liquid.

2.2.1

What is Fluid?

The uid is mainly divided into two categories: liquids and gases. The main dierence between the liquids and gases state is that gas will occupy the whole volume while liquids has an almost xed volume. This dierence can be, for most practical purposes considered, sharp even though in reality this dierence isnt sharp. The dierence between a gas phase to a liquid phase above the critical point are practically minor. But below the critical point, the change of water pressure by 1000% only change the volume by less than 1 percent. For example, a change in the volume by more than 5% will require tens of thousands percent change of the pressure. So, if the change of pressure is signicantly less than that, then the change of volume is at best 5%. Hence, the pressure will not aect the volume. In gaseous phase, any change in pressure directly aects the volume. The gas lls the volume and liquid cannot. Gas has no free interface/surface (since it does ll the entire volume).

2.2.2

What is Shear Stress?

The shear stress is part of the pressure tensor. However, here it will be treated as a separate issue. In solid mechanics, the shear stress is considered as the ratio of the force acting on area in the direction of the forces perpendicular to area. Dierent from solid, uid cannot pull directly but through a solid surface. Consider liquid that undergoes a shear stress between a short distance of two plates as shown in Figure (2.2). The upper plate velocity generally will be U = f (A, F, h) (2.1)

Where A is the area, the F denotes the force, h is the distance between the plates. From solid mechanics study, it was shown that when the force per area increases, the velocity of the plate increases also. Experiments show that the increase of height will increase the velocity up to a certain range. Consider moving the plate with a zero lubricant (h 0) (results in large force) or a large amount of lubricant (smaller force). In this discussion, the aim is to develop dierential equation, thus the small distance analysis is applicable. For cases where the dependency is linear, the following can be written U hF A (2.2)

2.2. WHAT IS FLUID? SHEAR STRESS Equations (2.2) can be rearranged to be U F h A Shear stress was dened as xy = F A

13

(2.3)

(2.4)

From equations (2.3) and (2.4) it follows that ratio of the velocity to height is proportional to shear stress. Hence, applying the coecient to obtain a new equality as xy = U h (2.5)

Where is called the absolute viscosity or dynamic viscosity. In steady state, the distance the upper plate moves after small amount of t0 < time, t is d = U t (2.6)

t1 < t2 <

t3

From gure (2.2) it can be noticed that for a small angle, the regular approximation provides
geometry

d = U t =

(2.7)

Fig. -2.2. The deformation of uid due to shear stress as progression of time.

From equation (2.7) it follows that U =h t (2.8)

Combining equation (2.8) with equation (2.5) yields xy = t (2.9)

If the velocity prole is linear between the plate (it will be shown later that it is consistent with derivations of velocity), then it can be written for small angle that dU = t dy (2.10)

Materials which obey equation (2.9) are referred to as Newtonian uid. For liquid metal used in the die casting industry, this property should be considered as Newtonian uid.

14

CHAPTER 2. BASIC FLUID MECHANICS

2.3 Thermodynamics and mechanics concepts

2.3.1 Thermodynamics
In this section, a review of several denitions of common thermodynamics terms is presented. This introduction is provided to bring familiarity of the material back to the student.

2.3.2

Basic Denitions

The following basic denitions are common to thermodynamics and will be used in this book. Work In mechanics, the work was dened as mechanical work = Fd = P dV (2.11)

This denition can be expanded to include two issues. The rst issue that must be addressed, that work done on the surroundings by the system boundaries similarly is positive. Two, there is a transfer of energy so that its eect can cause work. It must be noted that electrical current is a work while heat transfer isnt. System This term will be used in this book and it is dened as a continuous (at least partially) xed quantity of matter (neglecting Einsteins law eects). For almost all engineering purposes this law is reduced to two separate laws: mass conservation and energy conservation. Our system can receive energy, work, etc as long as the mass remains constant the denition is not broken. Thermodynamics First Law This law refers to conservation of energy in a non accelerating system. Since all the systems can be calculated in a non accelerating system, the conservation is applied to all systems. The statement describing the law is the following: Q12 W12 = E2 E1 (2.12)

The system energy is a state property. From the rst law it directly implies that for process without heat transfer (adiabatic process) the following is true W12 = E1 E2 (2.13)

Interesting results of equation (2.13) is that the way the work is done and/or intermediate states are irrelevant to nal results. The internal energy is the energy that depends on the other properties of the system. Example: for pure/homogeneous and

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

15

simple gases it depends on two properties like temperature and pressure. The internal energy is denoted in this book as EU and it will be treated as a state property. The system potential energy is dependent upon the body force. A common body force is gravity. For such body force, the potential energy is mgz where g is the gravity force (acceleration), m is the mass and the z is the vertical height from a datum. The kinetic energy is K.E. = mU 2 2 (2.14)

Thus the energy equation can be written as mU1 2 mU2 2 + m g z1 +EU 1 + Q = + m g z2 +EU 2 + W (2.15) 2 2 where Bf is a body force. For the unit mass of the system equation (2.15) is transformed into U2 2 U1 2 + gz1 + Eu 1 + q = + gz2 + Eu 2 + w (2.16) 2 2 where q is the energy per unit mass and w is the work per unit mass. The new internal energy, Eu , is the internal energy per unit mass. Since the above equations are true between arbitrary points, choosing any point in time will make it correct. Thus, dierentiating the energy equation with respect to time yields the rate of change energy equation. The rate of change of the energy transfer is DQ =Q Dt (2.17)
Bf Bf

In the same manner, the work change rate transferred through the boundaries of the system is DW =W Dt Since the system is with a xed mass, the rate energy equation is D EU DU Dgz QW = + mU +m Dt Dt Dt (2.19) (2.18)

For the case were the body force, Bf = g, is constant with time like in the case of gravity equation (2.19) reduced to D EU DU Dz QW = + mU + mg Dt Dt Dt

(2.20)

16

CHAPTER 2. BASIC FLUID MECHANICS

The time derivative operator, D/Dt is used instead of the common notation because it refers to system property derivative. Thermodynamics Second Law There are several denitions of the second law. No matter which denition is used to describe the second law it will end in a mathematical form. The most common mathematical form is Clausius inequality which state that Q 0 T (2.21)

The integration symbol with the circle represent integral of cycle (therefore circle) of system which returns to the same condition. If there is no lost, it is referred as a reversible process and the inequality change to equality. Q =0 T (2.22)

The last integral can go though several states. These states are independent of the path the system goes through. Hence, the integral is independent of the path. This observation leads to the denition of entropy and designated as S and the derivative of entropy is ds Q T rev
2

(2.23)

Performing integration between two states results in

2

S2 S1 =
1

Q = T rev

dS
1

(2.24)

One of the conclusions that can be drawn from this analysis is for reversible and adiabatic process dS = 0. Thus, the process in which it is reversible and adiabatic, the entropy remains constant and referred to as isentropic process. It can be noted that there is a possibility that a process can be irreversible and the right amount of heat transfer to have zero change entropy change. Thus, the reverse conclusion that zero change of entropy leads to reversible process, isnt correct. For reversible process equation (2.22) can be written as Q = T dS and the work that the system is doing on the surroundings is W = P dV Substituting equations (2.25) (2.26) into (2.20) results in T dS = d EU + P dV (2.27) (2.26) (2.25)

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

17

Even though the derivation of the above equations were done assuming that there is no change of kinetic or potential energy, it still remains valid for all situations. Furthermore, it can be shown that it is valid for reversible and irreversible processes. Enthalpy It is a common practice to dene a new property, which is the combination of already dened properties, the enthalpy of the system. H = EU + P V The specic enthalpy is enthalpy per unit mass and denoted as, h. Or in a dierential form as dH = dEU + dP V + P dV Combining equations (2.28) the (2.27) yields T dS = dH V dP (2.30) (2.29) (2.28)

For isentropic process, equation (2.27) is reduced to dH = V dP . The equation (2.27) in mass unit is T ds = du + P dv = dh dP (2.31)

when the density enters through the relationship of = 1/v. Specic Heats The change of internal energy and enthalpy requires new denitions. The rst change of the internal energy and it is dened as the following Eu (2.32) T And since the change of the enthalpy involve some kind of work, it is dened as Cv

Cp

h T

(2.33)

The ratio between the specic pressure heat and the specic volume heat is called the ratio of the specic heats and it is denoted as, k. k Cp Cv (2.34)

18

CHAPTER 2. BASIC FLUID MECHANICS

For liquid metal used in die casting, the ratio of the specic heats is bite higher than one (1) and therefore the dierence between them is almost zero and therefore referred as C. Equation of state Equation of state is a relation between state variables. Normally the relationship of temperature, pressure, and specic volume dene the equation of state for gases. The simplest equation of state referred to as ideal gas and it is dened as P = RT (2.35)

Application of Avogadros law, that all gases at the same pressures and temperatures have the same number of molecules per unit of volume, allows the calculation of a universal gas constant. This constant to match the standard units results in R = 8.3145 kj kmol K (2.36)

Thus, the specic gas can be calculated as R= R M (2.37)

The specic constants for select gas at 300K is provided in table 2.1. From equation (2.35) of state for perfect gas it follows d(P v) = RdT For perfect gas dh = dEu + d(P v) = dEu + d(RT ) = f (T ) (only) From the denition of enthalpy it follows that d(P v) = dh dEu (2.40) (2.39) (2.38)

Utilizing equation (2.38) and substituting into equation (2.40) and dividing by dT yields Cp Cv = R This relationship is valid only for ideal/perfect gases. The ratio of the specic heats can be expressed in several forms as R k1 (2.41)

Cv =

(2.42)

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

Table -2.1. Properties of Various Ideal Gases [300K]

19

Gas

Chemical Formula Ar C4 H10 CO2 CO C2 H6 C2 H4 He H2 CH4 Ne N2 C8 H18 O2 C3 H8 H2 O

Molecular Weight 28.970 39.948 58.124 44.01 28.01 30.07 28.054 4.003 2.016 16.04 20.183 28.013 114.230 31.999 44.097 18.015

kj KgK

Cv

kj KgK

CP

kj KgK

Air Argon Butane Carbon Dioxide Carbon Monoxide Ethane Ethylene Helium Hydrogen Methane Neon Nitrogen Octane Oxygen Propane Steam

0.28700 0.20813 0.14304 0.18892 0.29683 0.27650 0.29637 2.07703 4.12418 0.51835 0.41195 0.29680 0.07279 0.25983 0.18855 0.48152

1.0035 0.5203 1.7164 0.8418 1.0413 1.7662 1.5482 5.1926 14.2091 2.2537 1.0299 1.0416 1.7113 0.9216 1.6794 1.8723

0.7165 0.3122 1.5734 0.6529 0.7445 1.4897 1.2518 3.1156 10.0849 1.7354 0.6179 0.7448 1.6385 0.6618 1.4909 1.4108

1.400 1.400 1.091 1.289 1.400 1.186 1.237 1.667 1.409 1.299 1.667 1.400 1.044 1.393 1.327 1.327

Cp =

kR k1

(2.43)

The specic heats ratio, k value ranges from unity to about 1.667. These values depend on the molecular degrees of freedom (more explanation can be obtained in Van Wylen F. of Classical thermodynamics.) The values of several gases can be approximated as ideal gas and are provided in Table (2.1). The entropy for ideal gas can be simplied as the following
2

s2 s1 =
1

dh dP T T

(2.44)

20

CHAPTER 2. BASIC FLUID MECHANICS

Using the identities developed so far one can nd that

2

s2 s1 =
1

Cp

dT T

2 1

R dP T2 P2 = Cp ln R ln P T1 P1

(2.45)

Or using specic heats ratio equation (2.45) transformed into s2 s1 k T2 P2 = ln ln R k 1 T1 P1 For isentropic process, s = 0, the following is obtained T2 ln = ln T1 P2 P1
k1 k

(2.46)

(2.47)

There are several famous identities that results from equation (2.47) as

T2 = T1

P2 P1

k1 k

P2 P1

k1

(2.48)

The ideal gas model is a simplied version of the real behavior of real gas. The real gas has a correction factor to account for the deviations from the ideal gas model. This correction factor is referred to as the compressibility factor and dened as Z= PV RT (2.49)

Control Volume The control volume was introduced by L. Euler1 In the control volume (c.v) the focus is on specic volume which mass can enter and leave. The simplest c.v. is when the boundaries are xed and it is referred to as the Nondeformable c.v.. The conservation of mass to such system can be reasonably approximated by d dt dV =
Vc.v. Sc.v.

Vrn dA

(2.50)

This equation states the change in the volume came from the dierence of masses being added through the boundary. put two examples of simple for mass conservation. For deformable c.v. d dt dV =
Vc.v. Vc.v.

d dV + dt

Vrn dA
Sc.v.

(2.51)

1 A blind man known as the master of calculus, made his living by being a tutor, can you imagine he had eleven kids: where he had the time and energy to develop all the great theory and mathematics.

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

21

2.3.3

Momentum Equation

The second Newton law of motion is written mathematically as D mV Dt This explanation, of course, for uid particles can be written as F = F = D Dt V dV
Vsys

(2.52)

(2.53)

or more explicitly it can be written as F = d dt V dV +

Vc.v. Ac.v.

V Vrn dA

(2.54)

2.3.4

Compressible ow

This material is extensive and requires a semester for student to have good undersound wave standing of this complex material. Yet dU velocity=dU c to give very minimal information is seems P+dP P +d to to be essential to the understanding of the venting design. The summary material here is derived from the book Fundamentals of Compressible Flow Mechanics. Fig. -2.3. A very slow moving piston in a still
gas.

2.3.5

Speed of Sound
Control volume around the sound wave

The speed of sound is a very important parameter in the die casting process because c-dU c it eects and explains the choking in the P+dP die casting process. What is the speed of P +d the small disturbance +as it travels in a quiet medium? This velocity is referred to as the speed of sound. To answer this Fig. -2.4. Stationary sound wave and gas moves question, consider a piston moving from relative to the pulse. the left to the right at a relatively small velocity (see Figure 2.3). The information that the piston is moving passes thorough a single pressure pulse. It is assumed that if the velocity of the piston is innitesimally small, the pulse will be innitesimally small. Thus, the pressure and density can be assumed to be continuous. It is convenient to look at a control volume which is attached to a pressure pulse. Applying the mass balance yields c = ( + d)(c dU ) (2.55)

22

CHAPTER 2. BASIC FLUID MECHANICS

or when the higher term dU d is neglected yields dU = cd = dU = cd (2.56)

From the energy equation (Bernoullis equation), assuming isentropic ow and neglecting the gravity results (c dU )2 c2 dP + =0 2 neglecting second term (dU 2 ) yield cdU + dP =0 (2.58) (2.57)

Substituting the expression for dU from equation (2.56) into equation (2.58) yields d dP dP = = c2 = (2.59) d It is shown in the book Fundamentals of Compressible Fluid Mechanics that relationship between n, Z and k is c2
k

Cp n= Cv

z+T z+T

z T z T P

(2.60)

Note that n approaches k when z 1 and when z is constant. The speed of sound for a real gas can be obtained in similar manner as for an ideal gas dP = nzRT d (2.61)

Speed of Sound in Almost Incompressible Liquid Even liquid metal normally is assumed to be incompressible but in reality it has a small and important compressible aspect. The ratio of the change in the fractional volume to pressure or compression is referred to as the bulk modulus of the material. The mathematical denition of bulk modulus is as follows dP B= (2.62) d In physical terms it can be written as c= elastic property = inertial property B (2.63)

In summary, the speed of sound in liquid metals is about 5 times faster than the speed of sound in gases in the chamber.

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

23

2.3.6

Choked Flow

In this section a discussion on a steady state ow through a smooth and continuous area ow rate is presented which include the ow through a converging diverging nozzle. The isentropic ow models are important because of two main reasons: Stagnation State for Ideal Gas Model
Fig. -2.5.

PB = P0 P P0
Subsonic M <1

Supersonic
M >1

distance, x

Flow of a compressible substance

It is assumed that the ow is one (gas) through a convergingdiverging nozzle. dimensional. Figure (2.5) describes a gas ow through a convergingdiverging nozzle. It has been found that a theoretical state known as the stagnation state is very useful in which the ow is brought into a complete motionless condition in isentropic process without other forces (e.g. gravity force). Several properties can be represented by this theoretical process which include temperature, pressure, and density etc and denoted by the subscript 0. A dimensionless velocity and it is referred as Mach number for the ratio of velocity to speed of sound as M The temperature ratio reads T0 k1 2 =1+ M T 2 (2.65) U c (2.64)

The ratio of stagnation pressure to the static pressure can be expressed as the function of the temperature ratio because of the isentropic relationship as P0 = P T0 T
k k1

k1 2 1+ M 2

k k1

(2.66)

In the same manner the relationship for the density ratio is 0 = T0 T

1 k1

k1 2 1+ M 2

1 k1

(2.67)

A new useful denition is introduced for the case when M = 1 and denoted by superscript . The special case of ratio of the star values to stagnation values are dependent only on the heat ratio as the following: c 2 2 T = 2 = T0 c0 k+1 (2.68)

24 and P = P0

CHAPTER 2. BASIC FLUID MECHANICS

2 k+1

k k1

(2.69)

= 0

2 k+1

1 k1

(2.70)

Static Properties As A Function of Mach Number

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 Mon Jun 5 17:39:34 2006 0 4 5 Mach number 6 7 8 9 P/P0 /0 T/T0

Fig. -2.6. The stagnation properties as a function of the Mach number, k=1.4

The denition of the star Mach is ratio of the velocity and star speed of sound at M = 1. The ow in a convergingdiverging nozzle has two models: First is isentropic and adiabatic model. Second is isentropic and isothermal model. Clearly, the stagnation temperature, T0 , is constant through the adiabatic ow because there isnt heat transfer. Therefore, the stagnation pressure is also constant through the ow because of the isentropic ow. Conversely, in mathematical terms, equation (2.65) and equation (2.66) are the same. If the right hand side is constant for one variable, it is constant for the other. In the same argument, the stagnation density is constant through the ow. Thus, knowing the Mach number or the temperature will provide all that is needed to nd the

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

25

other properties. The only properties that need to be connected are the cross section area and the Mach number. Examination of the relation between properties can then be carried out. The Properties in the Adiabatic Nozzle When there is no external work and heat transfer, the energy equation, reads dh + U dU = 0 (2.71)

Dierentiation of continuity equation, AU = m = constant, and dividing by the continuity equation reads d dA dU + + =0 A U The thermodynamic relationship between the properties can be expressed as T ds = dh dP (2.73) (2.72)

For isentropic process ds 0 and combining equations (2.71) with (2.73) yields dP + U dU = 0 (2.74)

Dierentiation of the equation state (perfect gas), P = RT , and dividing the results by the equation of state (RT ) yields dP d dT = + P T (2.75)

Obtaining an expression for dU/U from the mass balance equation (2.72) and using it in equation (2.74) reads
dU U

dP dA d U2 + =0 A

(2.76)

Rearranging equation (2.76) so that the density, , can be replaced by the static pressure, dP/ yields dP = U2 dA d dP + A dP
1 c2

(2.77)

dA d dP = U2 + dP A

26

CHAPTER 2. BASIC FLUID MECHANICS

Recalling that dP/d = c2 and substitute the speed of sound into equation (2.77) to obtain dP Or in a dimensionless form dP dA 1 M2 = U2 A (2.79) 1 U c
2

= U2

dA A

(2.78)

Equation (2.79) is a dierential equation for the pressure as a function of the cross section area. It is convenient to rearrange equation (2.79) to obtain a variables separation form of dP = U 2 dA A 1 M2 (2.80)

Before going further in the mathematical derivation it is worth while to look at the physical meaning of equation (2.80). The term U 2 /A is always positive (because all the three terms can be only positive). Now, it can be observed that dP can be positive or negative depending on the dA and Mach number. The meaning of the sign change for the pressure dierential is that the pressure can increase or decrease. It can be observed that the critical Mach number is one. If the Mach number is larger than one than dP has opposite sign of dA. If Mach number is smaller than one dP and dA have the same sign. For the subsonic branch M < 1 the term 1/(1 M 2 ) is positive hence dA > 0 = dP > 0 dA < 0 = dP < 0 From these observations the trends are similar to those in incompressible uid. An increase in area results in an increase of the static pressure (converting the dynamic pressure to a static pressure). Conversely, if the area decreases (as a function of x) the pressure decreases. Note that the pressure decrease is larger in compressible ow compared to incompressible ow. For the supersonic branch M > 1, the phenomenon is dierent. For M > 1 the term 1/1 M 2 is negative and change the character of the equation. dA > 0 dP < 0 dA < 0 dP > 0 This behavior is opposite to incompressible ow behavior. For the special case of M = 1 (sonic ow) the value of the term 1 M 2 = 0 thus mathematically dP or dA = 0. Since physically dP can increase only in a nite amount it must be that dA = 0.It must also be noted that when M = 1 occurs only when dA = 0. However, the opposite, not necessarily means that when dA = 0 that M = 1. In that case, it is possible that dM = 0 thus the diverging side is in the subsonic branch and the ow isnt choked.

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS Isentropic Isothermal Flow Nozzle

27

In this section, the other extreme case model where the heat transfer to the gas is perfect, (e.g. Eckert number combination is very small) is presented. Again in reality the heat transfer is somewhere in between the two extremes. So, knowing the two limits provides a tool to examine where the reality should be expected. The perfect gas model is again assumed. In isothermal process the perfect gas model reads P = RT dP = dRT (2.81)

Substituting equation (2.81) into the momentum equation2 yields U dU + RT dP =0 P (2.82)

Integration of equation (2.82) yields the Bernoullis equation for ideal gas in isothermal process which reads U2 2 U1 2 P2 + RT ln =0 2 P1 Then the stagnation velocity is U= 2RT ln P P0 (2.84) (2.83)

It can be shown that the pressure ratio is

k(M1 2 M2 2 ) P2 2 = =e P1

eM1 2 eM2

k 2

(2.85)

As opposed to the adiabatic case (T0 = constant) in the isothermal ow the stagnation temperature ratio can be expressed 1 ! T01 T1 1 + = T02 T 2 1+
2 k1 2 M1 2 k1 2 M2

1+ 1+

2 k1 2 M1 2 k1 2 M2

(2.86)

Combining equation mass conservation with equation (2.85) yields A2 M1 = A1 M2

eM2 eM1 2

k 2

(2.87)

2 The one dimensional momentum equation for steady state is U dU/dx = dP/dx+0(other eects) which are neglected here.

28

CHAPTER 2. BASIC FLUID MECHANICS

The change in the stagnation pressure can be expressed as P02 P2 = P01 P1 1+ 1+

2 k1 2 M2 2 k1 2 M1
k k1

eM1 = eM1 2

k 2

(2.88)

The critical point, at this stage, is unknown (at what Mach number the nozzle is choked is unknown) so there are two possibilities: the choking point or M = 1 to normalize the equation. Here the critical point dened as the point whereM = 1 so results can be compared to the adiabatic case and denoted by star. Again it has to be emphasized that this critical point is not really related to physical critical point but it is only an arbitrary denition. The true critical point is when ow is choked and the relationship between two will be presented. The critical pressure ratio can be obtained from (2.85) to read
(1M 2 )k P = =e 2 P

(2.89)

Equation (2.87) is reduced to obtained the critical area ratio writes A 1 = e A M

(1M 2 )k 2

Isothermal Nozzle
k=14 4 3.5 P/P * A/A * P0 / P0 T 0 / T0 2.5 2 T/T
* * *

(2.90)

Similarly the stagnation temperature reads 2 1 + k1 M1 2 T0 2 = T0 k+1

k k1

1.5 1 0.5 0

0.5

1.5

2 M

2.5

3.5

(2.91)

Tue Apr 5 10:20:36 2005

Finally, the critical stagnation pressure Fig. -2.7. Various ratios as a function of Mach number for isothermal Nozzle reads
(1M 2 )k P0 2 =e P0

2 1 + k1 M1 2 2 k+1

k k1

(2.92)

The maximum value of stagnation pressure ratio is obtained when M = 0 at which is P0 P0 =e

M =0
k 2

2 k+1

k k1

(2.93)

For specic heats ratio of k = 1.4, this maximum value is about two. It can be noted that the stagnation pressure is monotonically reduced during this process. Of course in isothermal process T = T . All these equations are plotted in Figure (2.7). From the Figure 2.7 it can be observed that minimum of the curve A/A isnt on M = 1. The minimum of the curve is when area is minimum and at the point where

2.3. THERMODYNAMICS AND MECHANICS CONCEPTS

29

the ow is choked. It should be noted that the stagnation temperature is not constant as in the adiabatic case and the critical point is the only one constant. The mathematical procedure to nd the minimum is simply taking the derivative and equating to zero as the following d
A A

dM Equation (2.94) simplied to

kM 2 e

k(M 2 1) 2

k(M 2 1) 2

M2

=0

(2.94)

kM 2 1 = 0

1 M= k

(2.95)

30

CHAPTER 2. BASIC FLUID MECHANICS

The shear, S, at the ingate is determined by the average velocity, U, of the liquid and by the ingate thickness, t. Dimensional analysis shows that is directly proportional to (U/ ). The constant of proportionality is dicult to determine, . . .1 Murray, CSIRO Australia

CHAPTER 3 Dimensional Analysis

One of the important tools to understand the die casting process is dimensional analysis. Fifty years ago, this method transformed the uid mechanics/heat transfer into a uniform understanding. This book attempts to introduce to the die casting industry this established method2 . Experimental studies will be expanded/generalized as it was done in convective heat transfer. It is hoped that as a result, separate sections for aluminum, zinc, and magnesium will not exist anymore in die casting conferences. This chapter is based partially on Dr. Eckerts book, notes, and the article on dimensional analysis applied to die casting. Several conclusions are derived from this analysis and they will be presented throughout this chapter. This material can bring great benet to researchers who want to built their research on a solid foundation. For those who are dealing with the numerical research/calculation, it is useful to learn when some parameters should be taken into account and why.

1 Citing The Design of feed systems for thin walled zinc high pressure die castings, Metallurgical and materials transactions B Vol. 27B, February 1996, pp. 115118. This excerpt is an excellent example of poor research and poor understanding. This unknown constant is called viscosity (see Basics of Fluid Mechanics in Potto series. Here, a discussion on some specic mistakes were presented in that paper (which are numerous). Dimensional analysis is a tool which can take cluttered and meaningless paper such as the above and turn them into something with real value. As proof of their model, the researchers have mentioned two unknown companies that their model is working. What a nice proof! Are the physics laws really dierent in Australia? 2 Actually, Prof. E.R.G. Eckert introduced the dimensional analysis to the die casting long before. The author is his zealous disciple, all the credit should go to Eckert. Of course, all the mistakes are the authors and none of Dr. Eckerts. All the typos in Eckerts paper were this authors responsibility for which he apologizes.

31

32

CHAPTER 3. DIMENSIONAL ANALYSIS

3.0.7

How The Dimensional Analysis Work

In dimensional analysis, the number of the eecting parameters is reduced to a minimum by replacing the dimensional parameters by dimensionless parameters. Some researchers point out that the chief advantage of this analysis is to obtain experimental results with a minimum amount of labor, results in a form having maximum utility [18, pp. 395]. The dimensional analysis has several other advantages which include; 1)increase of understanding, 2) knowing what is important, and 3) compacting the presentation3 . The advantage of compact of presentation allows one to see the big picture with minimal eort. Dimensionless parameters are parameters which represent a ratio which does not have a physical dimension. The experimental study assists to solve problems when the solution of the governing equation cannot be obtained. To achieve this, experiments are designed to be similar to the situations which need to be solved or simulated. The base for this concept is mathematical. Two dierent sets of phenomena will produce a similar result if the governing dierential equations with boundaries conditions are similar. The actual experiments are dicult to carry out in many cases. Thus, design experiments with the same governing dierential equations as the actual phenomenon is the solution. This similarity does not necessarily mean that the experiments have to be carried exactly as studied phenomena. It is enough that the main dimensionless parameters are similar, since the minor dimensional parameters, in many cases, are insignicant. For example, a change in Reynolds number is insignicant since a change in Reynolds number in a large range does not aect the friction factor. An example of the similarity applied to the die cavity is given in the section 3.5. Researchers in casting in general and die casting in particular do not utilize this method. For example, after the Russians [6] introduced the water analogy method (in casting) in the 40s all the experiments such as Wallace, CSIRO, etc. conducted poorly designed experiments. For example, Wallace record the Reynolds and Froude number without attempting to match the governing equations. Another example is the experimental study of Gravity Tiled Die Casting (low pressure die casting) performed by Nguyens group in 1986 comparing two parameters and We. Flow of free Re falling, the velocity is a function of the height (U gH). Hence, the equation Remodel = Reactual should lead only to Hmodel Hactual and not to any function of Umodel /Uactual . The value of Umodel /Uactual is actually constant for the same height ratio. The Wallace experiments with Reynolds number matching does not lead to matching of similar governing equations. Many other important parameters which control the governing equations are not simulated [26]. The governing equations in these cases include several other important parameters which have not been controlled

3 The importance of compact presentation is attributed to Prof. M. Bentwitch who was mentor to many including the author during his masters studies.

3.1. INTRODUCTION

33

or even measured, monitored, and simulated4 . Moreover, the Re number is controlled by the ow rate and the characteristics of the ladle opening and not as in the pressurized pipe ow as the authors assumed.

3.1 Introduction
Lets take a trivial example of tting a rode into a circular hole (see Figure 3.1). To solve this problem, it is required to know two parameters; 1) the diameter of the rode and 2) the diameter of the hole. Actually, it is required to have only one parameter, the ratio of the rode diameter to the hole diameter. The ratio is a dimensionless number and with this Fig. -3.1. Rod into the hole example number one can say that for a ratio larger than one, the rode will not enter the hole; and ratio smaller than one, the rod is too small. Only when the ratio is equal to one, the rode is said to be t. This allows one to draw the situation by using only one coordinate. Furthermore, if one wants to deal with tolerances, the dimensional analysis can easily be extended to say that when the ratio is equal from 0.99 to 1.0 the rode is tting, and etc. If one were to use the two diameters description, he will need more than this simple sentence to describe it. In the preceding simplistic example, the advantages are minimal. In many real problems, including the die casting process, this approach can remove clattered views and put the problem into focus. It also helps to use information from dierent problems to a similar situation. Throughout this book the reader will notice that the systems/equations are converted to a dimensionless form to augment understanding.

3.2 The Die Casting Process Stages

The die casting process can be broken into many separated processes which are controlled by dierent parameters. The simplest division of the process for a cold chamber is the following: 1) lling the shot sleeve, 2) slow plunger velocity, 3) lling the runner system 4) lling the cavity and overows, and 5) solidication process (also referred as intensication process). This division into such subprocesses results in a clear picture
many conceptual physical mistakes, the authors have a conceptual mathematical mistake. They tried to achieve the same Re and F r numbers in the experiments as in reality for low pressure die casting. They derived an equation for the velocity ratio based on equal Re numbers (model and actual). They have done the same for F r numbers. Then they equate the velocity ratio based on equal Re to velocity ratio based on equal F r numbers. However, velocity ratio based on equal Re is a constant and does vary with the tunnel dimension (as opposed to distance from the starting point). The fact that these ratios have the same symbols do not mean that they are really the same. These two ratios are dierent and cannot be equated.
4 Besides

34

CHAPTER 3. DIMENSIONAL ANALYSIS

on each process. On one hand, in processes 1 to 3, it is desirable to have a minimum heat transfer/solidication to take place for obvious reasons. On the other hand, in the rest of the processes, the solidication is the major concern. In die casting, the information and conditions do not travel upstream. For example, the turbulence does not travel from some point at the cavity to the runner and ofcourse, to the shot sleeve. This kind of relationship is customarily denoted as a parabolic process (because in mathematics the dierential equations describe these kind of cases as parabolic). To a larger extent it is true in die casting. The pressure in the cavity does not aect the ow in the sleeve or the runner if the vent system is well designed. In other words, the design of the pQ2 diagram is not controlled by downstream conditions. Another example, the critical slow plunger velocity is not aected by the air/gas ow/pressure in the cavity. In general, the turbulence generated downstream does not travel upstream in this process. One has to restrict this characterization to some points. One point is particularly mentioned here: The poor design of the vent system aects the pressure in the cavity and therefore the eects do travel down stream. For example, the pQ2 diagram calculations are aected by poor vent system design.

3.2.1

Filling the Shot Sleeve

The ow from the ladle to the shot sleeve did not receive much attention in the die casting research5 because it is believed that it does not play a signicant role. For low pressure die casting, the ow of liquid metal from the ladle through channel(s) to the die cavity plays an important role6 . The importance of the understanding of this process can show us how to minimize the heat transfer, layer created on the sleeve (solidication layer), and sleeve protection from; a) erosion b) plunger problem. The jet itself has no smooth surface and two kinds of Fig. -3.2. Hydraulic jump in the instability occurs. The rst instability is of Bernoullis shot sleeve. eect and second eect is Bar-Meirs eect that boundary conditions cannot be satised for two phase ow. Yet, for die casting process, these two eects (see Figure 3.2 do not change the global ow in the sleeve. At rst, the hydraulic jump is created when the liquid metal enters the sleeve. The typical time scale for hydraulic jump creation is almost instant and extremely short as can be shown by the characteristic methods. As the liquid metal level in the sleeve rises, the location of the jump moves closer to the impinging center. At a certain point, the liquid depth level is over the critical depth level and the hydraulic jump disappears. The critical depends on the liquid properties and the ratio of impinging momentum or velocity to the hydraulic static pressure. The impinging momentum impact is proportional to U 2 r2 and hydraulic pressure is proportional
Bar-Meirs instability Hydraulic Jump

few papers ( 0) can be found dealing with this aspect. elementary estimates of uid mechanics and heat transfer were made by the author and hopefully will be added to this book.
6 Some

5 Very

3.2. THE DIE CASTING PROCESS STAGES

35

to g h 2 r h. Where r is the radius of the impinging jet and h is the height of the liquid metal in the sleeve. The above statement leads to Ucritical g h2 r (3.1)

The critical velocity on the other hand has to be Uc ritical = g hL (3.2)

where hL is the distance of the ladle to the height of the liquid metal in the sleeve. The height where the hydraulic shock will not exist is hcritical r hL (3.3)

This analysis suggests that decreasing the ladle height and/or reducing less mass ow rate (the radius of the jet) result in small critical height. The air entrainment during that time will be discussed in the book Basic of Fluid Mechanics in the MultiPhase ow chapter. At this stage, air bubbles are entrained in the liquid metal which augment the heat transfer. At present, there is an extremely limited knowledge about the heat transfer during this part of the process, and of course less about how to minimize it. However, this analysis suggests that minimizing the ladle height is one of the ways to reduce it. The heat transfer from liquid metal to the surroundings is aected by the velocity and the ow patterns since the initial stage mechanism of heat transfer is changed from a dominated natural convection to a later H <H> stage dominated force convection. In addition, the liquid metal jet surface is also aected by heat transfer to some degree by change in the properties.
hydralic jump

bubles

air entrainment

Heat Transferred to the Jet The estimate on heat transfer requires some information on jet dynamics. Fig. -3.3. Filling of the shot sleeve. There are two eects that must be addressed; one the average radius and the uctuation of the radius. As rst approximation, the average jet radius changes due to the velocity change. For laminar ow, (for simplicity assume plug ow) the veloc ity function is x where x is the distance from the ladle. For constant ow rate, neglecting the change of density, the radius will change as r 1/ 4 x. Note that this relationship is not valid when it is very near the ladle proximity (r/x 0). The heat transfer increases as a function of x for these two reasons. The second eect is jet radius uctuations. Consider this, the jet leaves the ladle in a plug ow. Due to air friction, the shear stress changes the velocity prole to

36

CHAPTER 3. DIMENSIONAL ANALYSIS

parabolic. For simple assumption of steady state(it is not steady state), the momentum equation which governs the liquid metal is assume uz t
0

constant uz = 1 + ur r r r

uz r

(3.4)

Equation (3.4) is in simplied equation form for the gas and liquid phases. Thus, there are two equations that needs to be satised simultaneously; one for the gas side and one for the liquid side. Even neglecting several terms for this discussion, it clear that both equations are second order dierential equations which have dierent boundary conditions. Any second order dierential equation requires two dierent boundary conditions. Requirement to satisfy additional boundary condition can be achieved. Thus from physical point of view, second order dierential equation which needs to satisfy three boundary conditions is not possible, Thus there must be some wrong either with the governing equation or with the boundary conditions. In this case, the two governing equations must satisfy ve (5) dierent boundary conditions. These boundary conditions are as follows: 1) summitry at r = 0, 2) identical liquid metal and air velocities at the interface, 3) identical shear stress at the interface, 4) zero velocity at innity for the air, and 5) zero shear stress for the air at the innity. These requirements cannot be satised if the interface between the liquid metal and the gas is a straight line. The heat transfer to the sleeve in the impinging area is signicant but at present only very limited knowledge is available due to complexity.

3.2.2

Plunger Slow Moving Part

Fluid Mechanics The main point is the estimate for Shot energy dissipation. The dissipaheat transfer Sleeve to the air 2 tion is proportional to < U > process 1 L. Where the strange velocity, < U > is averaged kinetic velocity provided by jet. This kinetic energy is at most the same as potential energy of liquid metal in heat transfer the ladle. The potential energy in to the sleeve Solidification process 2 the ladle is < H > m g where layer < H > is averaged height see Figure 3.3. The averaged velocity Fig. -3.4. Heat transfer processes in the shot sleeve. in the shot sleeve is 2 g < H > 2 The rate energy dissipation can be estimated as <U > R L as L is the length of R shot sleeve. The shear stress is assumed to occur equally in the volume of liquid metal in the sleeve. This assumption of shear stress grossly under estimates the dissipation.
Temperature

3.2. THE DIE CASTING PROCESS STAGES

37

The actual dissipation is larger due to the larger velocity gradients. The estimated time is then Heat Transfer In this section, the solidication eects are examined. One of the assumptions in the analysis of the critical slow plunger velocity is that the solidication process does not play an important role (see Figure 3.4). The typical time for heat to penetrate a typical layer in air/gas phase is in the order of minutes. Moreover, the density of the air/gas is 3 order magnitude smaller than liquid metal. Hence, most of the resistance to heat transfer is in the gas phase. Additionally, it has been shown that the liquid metal surface is continuously replaced by slabs of material below the surface which is known in scientic literature as the renewal surface theory. Thus, the main heat transfer mechanism is through the liquid metal to the sleeve. The heat transfer rate for a very thin solidied layer can be approximated as Q klm T r L Ls r L t r (3.5)

Where Ls is the latent heat, klm is the thermal conductivity of liquid metal and t is the thickness of the solidication layer. Equation (3.6) results in t klm T r Ls r 2 (3.6)

The value for this die casting process in minutes is in the range of 0.01-0.001 after the thickness reaches to 1-2 [mm]. The relative thickness liquid metal further decreases as the inverse of the square solidied layer increases. If the solidication is less than shot sleeve (steel) one percent of the radius, the speed will be very small compared to the speed of the plunger. If the insulation solidication occur as a mushy zone then the heat transfer is reduced further and it is even lower than Fig. -3.5. Solidication of the shot this estimate and R 1). Therefore, the heat sleeve time estimates. transfer from the liquid metal surface to the air, as shown in Figure 3.4 (mark as process 1), acts as an insulator to the liquid metal. The governing equation in the sleeve is
y
(liquid) (solid) Temperature

d cp d

T = kd t

2T y 2

(3.7)

where the subscript d denotes the properties of the sleeve material. Boundary condition between the sleeve and the air/gas is T n =0
y=0

(3.8)

38

CHAPTER 3. DIMENSIONAL ANALYSIS

Where n represents the perpendicular direction to the die. Boundary conditions between the liquid metal (solid) and sleeve ksteel T y = kAL
y=l

T y

(3.9)
y=l

The governing equation for the liquid metal (solid phase) lm cp lm T = klm t 2T y 2 (3.10)

where lm denotes the properties of the liquid metal. The dissipation and the velocity are neglected due to the change of density and natural convection. Boundary condition between the phases of the liquid metal is given by vs s hsf = kl hsf s vn k T y ks
y=l+

T y

k
y=l+

(Tl Ts ) y

(3.11)
y=l+

the heat of solidication liquid metal density at the solid phase velocity of the liquid/solid interface conductivity

Neglecting the natural convection and density change, the governing equation in the liquid phase is l cp l T 2T = kl 2 t y (3.12)

The dissipation function can be assumed to be negligible in this case. There are three dierent periods in heat transfer; 1. lling the shot sleeve 2. during the quieting time, and 3. during the plunger movement. In the rst period, heat transfer is relatively very large (major solidication). At present, there is not much known about the uid mechanics not to say much about the solidication process/heat transfer in uid mechanics. The second period can be simplied and analyzed as known initial velocity prole. A simplied assumption can be made considering the fact that P r number is very small (large thermal boundary layer compared to uid mechanics boundary layer). Additionally, it can be assumed that the natural convection eects are marginal. In the last period, the heat transfer is composed from two zones: 1) behind the jump and 2) ahead of the jump. The heat transfer ahead of the jump is the same as in the second period; while the heat transfer behind the jump is like heat transfer into a plug ow for low P r number. The heat transfer in such cases have been studied in the past7 .
7 The

reader can refer, for example, to the book Heat and Mass Transfer by Eckert and Drake.

3.2. THE DIE CASTING PROCESS STAGES

39

3.2.3

Runner system

The ow in the runner system has to be divided into sections; 1) ow with free surface 2) lling the cavity liquid when the ow is pressurized (see Figmetal ures 3.6 and 3.7). In the rst section the gravity aects the air entrainment. The dominant parameters in this case air streaks are Weber number, W e and Reynolds number, Re. This phenomenon determines how much metal has to be ushed out. It is well known that the Fig. -3.6. Entrance of liquid metal to the runner. liquid interface cannot be a straight line. Above certain velocity (typical to die casting, high Re number) air leaves streaks of air/gas slabs behind the front line as shown in Figure 3.6. These streaks create a low heat transfer zone at the head of the jet and increases its velocity. The air entrainment created in this case is supposed to be ushed out through the vent system in a proper process design. Unfortunately, at present very little is known about this issue especially the geometry typical to die casting. Gravity Limited in Runner system In the second phase, the ow in the runner system is pressurized. The typical velocity is large of the range of 10-15 [m/sec]. The typical runner length is in order of 0.1[m]. The velocity due to gravity is 2.5[m/sec]. The F r number assumes the value 102 for which gravity play a limited role. The converging nozzle such as the transition into runner system (which a good die casting engineer should design) tends to reduce the turbulence, if turbulence exists, and can even eliminate it. In that view, the liquid metal enters the runner system as a laminar ow (actually close to a plug ow). For a duct with a typical dimension of 10 [mm] and a mean Fig. -3.7. Flow in runner when during pressurvelocity, U = 10[m/sec], (during the sec- izing process. ond stage), for aluminum die casting, the Reynolds number is: Ub Re = 5 107 which is a supercritical ow. However, the ow is probably laminar ow due to the short time. Another look at turbulence issue: The boundary layer is a function of the time (during the lling period) is of order
pressure

= 12t

40

CHAPTER 3. DIMENSIONAL ANALYSIS

The boundary layer in this case can be estimated as8 the time of the rst phase. Anyhow, utilizing the time of 0.01[sec] the viscosity of aluminum in the boundary layer is of the thickness of 0.25[mm] which indicates that ow is laminar.

3.2.4

Die Cavity

All the numerical simulations of die lling are done almost exclusively by assuming that the ow is turbulent and continuous (no two phase ow). In the section 3.3.1 a question about the question whether existence of turbulence is discussed and if so what kind of model is appropriate. Thus, the validity of these numerical models is examined. The liquid metal enters the cavity as a noncontinuous ow. According to some researchers, it is preferred that the ow will be atomized (spray). While there is a considerable literature about many geometries none available to typical die casting congurations9 . The ow can be atomized as either in laminar or turbulent region. The experiments by the author and by others, showed that the ow turns into spray in many cases ( See Figures 3.8).

Fig a.

Flow as a jet.

Fig b.

Flow as a spray.

Fig. -3.8. Typical ow pattern in die casting, jet entering into empty cavity.

In the section 3.4.1 it was shown that the time for atomization is very fast compared with any other process (lling time scale and, of course, the conduction heat transfer or solidication time scales). Atomization requires two streams with a signicant velocity dierence; stronger surface tension forces against the maintaining stability forces. Numerous experimental studies have shown that better castings are obtained when the injected velocity is above a certain value. This fact alone is enough to convince researchers that the preferred ow pattern is a spray ow. Yet, only a very small number of numerical models exist assuming spray ow and are used for die casting (for example, the paper by Hu at el [22].). Experimental work commonly cited as a proof
during the ow in the runner system, no lling of the cavity can just wonder who were the opposition to this research? Perhaps one of the referees as in the Appendix B for the all clues that have been received.
9 One 8 only

3.3. SPECIAL TOPICS

41

of turbulence was conducted in the mid 60s [30] utilizing water analogy10 . The white spats they observed in their experiments are atomization of the water. Because these experiments were poorly conducted (no similarity to die casting process) the observation/information from these studies is very limited. Yet with this limitation in mind, one can conclude that the spray ow does exist. Experiments by Fondse et al [16] show that atomization is larger in laminar ow compared to a turbulent ow in a certain range. This fact further creates confusion of what is the critical velocity needed in die casting. Since the experiments which measure the critical velocity were poorly conducted, no reliable information is available on what is the ow pattern and what is the critical velocity11 .

3.2.5

Intensication Period

The two main concerns in this phase is to extract heat from the die and to solidify the liquid metal as aptly as possible to obtain the nal shape. Thus, two operational parameters are important; one the (minimum) time for the intensication and two the pressure of the intensication (the clamping force). These two operational parameters can improve casting design to obtain good product. The main resistance to the heat ow is in the die and the cooling liquid (oil or water based solution). In some parts of the process, the heat is transformed to the cooling liquid via the boiling mechanism. However, the characteristic of boiling heat transfer time to achieve a steady state is larger than the whole process and the typical equations (steady state) for the preferred situation (heat transfer only in the rst mode) are not accurate. When there is very limited understanding of so many aspects of the process, the eects of each process on other processes are also cluttered.

3.3 Special Topics

3.3.1 Is the Flow in Die Casting Turbulent?
It is commonly assumed that the ow in die casting processes is turbulent in the shot sleeve, runner system, and during the cavity lling. Further, it also assumed that the k model can reasonably represent the turbulence structure. These assumptions are examined herein. The ow can be examined in three zones: 1) the shot sleeve, 2) the runner system, and 3) the mold cavity. Note, even if the turbulence exists in some regions, it doesnt necessarily mean that all the ow eld is turbulent.
10 The problems in these experiments were, among other things, no simulation of the dimensional numbers such as Re, Geometry etc. and therefore dierent dierential equations not typical to die casting were solved. ??punctuation inside quotes] The researchers also look at what is known as a poor design for disturbances to ow downstream (this is like putting screen in the ow). However, a good design requires smooth contours. 11 Beside other problems such as dierent ow velocity in dierent gates which were never really measured, the pressure in the cavity and quality of the liquid metal entering the cavity (is it in two phase?) were never recorded.

42 Transition from laminar to turbulent

CHAPTER 3. DIMENSIONAL ANALYSIS

Is the ow in the shot sleeve turbulent as the EKK sale engineers claim? These sale engineers did not present any evidence or analysis for such claims. For a simple anal ysis, the initial part of the shot sleeve lling, the liquid metal goes through a hy draulic jump. The ow after the hydraulic jump is very slow because the increase of the ratio of cross section areas. For example, casting of the 1[kg] from height Time [sec] of 0.2[m] to a shot sleeve of 0.1[m] cre ates a velocity in shot sleeve of 2[m/sec] Fig. -3.9. Transition to turbulent ow in circuwhich results after the hydraulic jump to lar pipe for instantaneous ow after Wygnanski be with velocity about 0.01[m/sec]. The and others by interpolation. Reynolds number for this velocity is 104 and Froude number of about 10. After the jump the Froude number is reduced and the ow is turbulent. However, by the time the hydraulic jump vanishes, the ow turns into laminar ow and no change (waviness) in the surface can be observed. It can be noticed that the time scale for the dissipation is about the same scale as the time for the operation of the next stage. Figure 3.9 exhibits the transition to a turbulent ow for instantaneous starting ow in a circular pipe. The abscissa represents time and the yaxis represents the Re number at which transition to turbulence occurs. The points on the graphs show the transition to a turbulence. This gure demonstrates that a large time is required to turn the ow pattern to turbulent which is measured in several seconds. Fig. -3.10. Flow pattern in the shot sleeve. The gure demonstrates that the transition does not occur below a certain critical Re number (known as the critical Re number for steady state). It also shows that a considerable time has elapsed before transition to turbulence occurs even for a relatively large Reynolds number. The geometry in die casting however is dierent and therefore it is expected that the transition occurs at dierent times. Our present knowledge of this area is very limited. Yet, a similar transition delay is expected to occur after the instantaneous startup which probably will be measured in seconds. The ow in die casting in many situations is very short (in order of milliseconds) and therefore it is expected that the transition to a turbulent ow does not occur. After the liquid metal is poured, it is normally repose for sometime in a range of 10 seconds. This fact is known in the scientic literature as the quieting time for which the existed turbulence (if exist) is reduced and after enough time (measured in seconds)
12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00

Re10^3

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

thin boundary layer plus solidification

almost still flow

Transition zone

3.3. SPECIAL TOPICS

43

is illuminated. Hence, the turbulence, which was created during the lling process of the shot sleeve, disappear due to viscous dissipation. The question is, whether the ow in the duration of the slow plunger velocity turbulent (see Figure 3.10) can be examined. Clearly, the ow in the substrate (a head of the wave) is still (almost zero velocity) and therefore the turbulence does not exist. The Re number behind the wave is above the critical Re number (which is in the range of 20003000). The typical time for the wave to travel to the end of the shot sleeve is in the range of a 100 second. At present there are no experiments on the ow behind the wave12 . The estimation can be done by looking at what is known in the literature about the transition to turbulence in instantaneous starting pipe ow. It has been shown [32] that the ow changes from laminar ow to turbulent ow in an abrupt manner for a ow with supercritical Re number. A typical velocity of the propagating front (transition between laminar to turbulent) is about the same velocity as the mean velocity of the ow. Hence, it is reasonable to assume that the turbulence is conned to a small zone in the wave front since the wave is traveling in a faster velocity than the mean velocity. Note that the thickness of the transition layer is a monotone increase function of time (traveling distance). The Re number in the shot sleeve based on the diameter is in a range of 104 which means that the boundary layer has not developed much. Therefore, the ow can be assumed as almost a plug ow with the exception of the front region. A Note on Numerical Simulations The most common model for turbulence that is used in the die casting industry for simulating the ow in cavity is k . This model is based on several assumptions 1. isentropic homogeneous turbulence, 2. constant material properties (or a mild change of the properties), 3. continuous medium (only liquid (or gas), no mixing of the gas, liquid and solid whatsoever), and 4. the dissipation does not play a signicant role (transition to laminar ow). The k model is considered reasonable for the cases where these assumptions are not far from reality. It has been shown, and should be expected, that in cases where assumptions are far from reality, the k model produces erroneous results. Clearly, if we cannot determine whether the ow is turbulent and in what zone, the assumption of isentropic homogeneous turbulence is very questionable. Furthermore, if the change to turbulence just occurred, one cannot expect the turbulence to have sucient time to become isentropic homogeneous. As if this is not enough complication, consider the eects of properties variations as a result of temperature change. Large variations of the properties such as the viscosity have been observed in many alloys especially in the mushy zone.
12 It has to be said that similar situations are found in two phase ow but they are dierent by the fact the ow in two phase ow is a sinusoidal in some respects.

44

CHAPTER 3. DIMENSIONAL ANALYSIS

While the assumption of the continuous medium is semi reasonable in the shot sleeve and runner, it is far from reality in the die cavity. As discussed previously, the ow is atomized and it is expected to have a large fraction of the air in the liquid metal and conversely some liquid metal drops in the air/gas phase. In such cases, the isentropic homogeneous assumption is very dubious. For these reasons the assumption of k model seems unreasonable unless good experiments can show that the choice of the turbulence model does not matter in the calculation. The question whether the ow in die cavity is turbulent or laminar is secondary. Since the two phase ow eects have to be considered such as atomization, air/gas entrainment etc. to describe the real ow in the cavity. Additional note on numerical simulation The solution of momentum equation for certain situations may lead to unstable solution. Such case is the case of two jets with dierent velocity ow into a medium and they are adjoined (see Figure 3.11). The solution of such ow can show that the velocity eld can be an unstable Fig. -3.11. Two streams of uids into a medium. solution for which the ow moderately changes to become like wave ow. However, in many cases this ow can turn out to be full with vortexes and such. The reason that this happened is the introduction of instabilities. Numerical calculations intrinsically are introducing instabilities because of truncation of the calculations. In many cases, these truncations results in overshooting or undershooting of the nature instability. In cases where the ow is unstable, a careful study is required to make sure that the solution did not produce an unrealistic solution for larger or smaller than reality introduced instabilities. An excellent example of such poor understating is a work made in EKK company [2]. In that work, the ow in the shot sleeve was analyzed. The nature of the ow is two dimensional which can be seen by all the photos taken by numerous people (staring from the 50s). The presenter of that work explained that they have used 3D calculations because they want to study the instabilities perpendicular to the ow direction. The numerical instability in this case is larger than real instabilities and therefore, the numerical results show phenomena does not exist in reality.
U1 U2 U1 U2

Reverse transition from turbulent ow to laminar ow After lling the die cavity, during the solidication process and intensication, the attained turbulence (if exist) is reduced and probably eliminated, i.e. the ow is laminar in a large portion of the solidication process. At present we dont comprehend when the transition point/criteria occurs and we must resort to experiments. It is a hope

3.3. SPECIAL TOPICS

45

that some real good experiments using the similarity technique, outlined in this book, will be performed. So more knowledge can be gained and hopefully will appear in this book.

3.3.2

Dissipation eect on the temperature rise

The large velocities of the liquid metal (particularly at the runner) theoretically can increase the liquid metal temperature. To study this phenomenon, compare the of maximum eect of all the kinetic energy that is transformed into thermal energy. U2 = cp T 2 This equation leads to the denition of Eckert number Ec = U2 cp T (3.14) (3.13)

When Ec number is very large it means that the dissipation plays a signicant role and conversely when Ec number is small the dissipation eects are minimal. In die casting, Eckert number, Ec, is very small therefore the thermal dissipation is very small and can be ignored.

3.3.3

Gravity eects

The gravity has a large eect only when the gravity force is large relatively to other forces. A typical velocity range generated by gravity is the same as for an object falling through the air. The air eects can be neglected since the air density is very small compared with liquid metal density. The momentum is the other dominate force in the lling of the cavity. Thus, the ratio of the momentum force to the gravity force, also known as Froude number, determines if the gravity eects are important. The Froude number is dened here as Fr = U2 g (3.15)

Where U is the velocity, is the characteristic length g is the gravity force. For example, the characteristic pouring length is in order of 0.1[m], in extreme cases the velocity can reach 1.6[m] with characteristic time of 0.1[sec]. The author is not aware of experiments to verify the ow pattern in such cases (low P r number due to solidication eect)13 Yet, it is reasonable to assume that the liquid metal in such a case, ow in laminar regimes even though the Re number is relatively large ( 104 ) because of the short time and the short distance. The Re number is dened by the ow rate and the thickness of the exiting typical dimension. Note, the velocity reached its maximum value just before impinging on the sleeve surface.
13 It

be interesting to nd such experiments.

46

CHAPTER 3. DIMENSIONAL ANALYSIS

The gravity has dominate eects on the ow in the shot sleeve since the typical value of the Froude number in that case (especially during the slow plunger velocity period) is in the range of one(1). Clearly, any analysis of the ow has to take into consideration the gravity (see Chapter 8).

3.4 Estimates of the time scales in die casting

3.4.1 Utilizing semi dimensional analysis for characteristic time
The characteristic time scales determine the complexity of the problem. For example, if the time for heat transfer/solidication process in the die cavity is much larger than the lling time, then the problem can be broken into three separate cases 1) the uid mechanics, the lling process, 2) the heat transfer and solidication, and 3) dissipation (maybe considered with solidication). Conversely, the real problem in die lling is that we would like for the heat transfer process to be slower than the lling process, to ensure a proper lling. The same can be said about the other processes.

lling time
The characteristic time for lling a die cavity is determined by tf L U (3.16)

Where L denotes the characteristic length of the die and U denotes the average lling velocity, determined by the pQ2 diagram, in most practical cases this time typically is in order of 5100 [millisecond]. Note, this time is not the actual lling time but related to it.

Atomization time
The characteristic time for atomization for a low Re number (large viscosity) is given by ta viscosity = (3.17)

where is the kinematic viscosity, is the surface tension, and is the thickness of the gate. The characteristic time for atomization for large Re number is given by ta momentum = 2U (3.18)

The results obtained from these equations are dierent and the actual atomization time in die casting has to be between these two values. Conduction time (die mold)

3.4. ESTIMATES OF THE TIME SCALES IN DIE CASTING The governing equation for the heat transfer for the die reads Td d cp d = kd t 2 Td 2 Td 2 Td + + x2 y 2 z 2
L

47

(3.19)
cooling liquid

To obtain the characteristic time we dimensionlessed the governing equation and present it with a group of constants that determine value of the characteristic time by setting it to unity. Denoting the following variables as
Fig. -3.12. Schematic of heat transfer processes in the die.

td =

t tc d

xd =

x L

yd = L

y L

zd =

z L

d =

T TB TM TB

(3.20)

the characteristic path of the heat transfer from the die inner surface to the cooling channels

subscript B boiling temperature of cooling liquid M liquid metal melting temperature With these denitions, equation (3.19) is transformed to tc d d = d2 t L 2 d 2 d 2 d + + x 2 y 2 z 2 (3.21)

which leads into estimate of the characteristic time as tc d 2 L d (3.22)

Note the characteristic time is not eected by the denition of the d .

Conduction time in the liquid metal (solid)

The governing heat equation in the solid phase of the liquid metal is the same as equation (3.19) with changing properties to liquid metal solid phase. The characteristic time for conduction is derived similarly as done previously by introducing the dimensional parameters t = t x y z T TB ; x = ; y = ; z = ; s = tc s TM TB (3.23)

48

CHAPTER 3. DIMENSIONAL ANALYSIS

where tc s is the characteristic time for conduction process and, , denotes the main path of the heat conduction process die cavity. With these denitions, similarly as was done before the characteristic time is given by
2

tc s

(3.24)

Note again that s has to be taken for properties of the liquid metal in the solid phase. Also note that the solidied length, , changes during the process and discussing the case where the whole die is solidied is not of interest. Initially the thickness, = 0 (or very small). The characteristic time for very thin layers is very small, tc s 0. As the solidied layer increases the characteristic time also increases. However, the temperature prole is almost established (if other processes were to remain in the same conditions). Similar situations can be found when a semi innite slab undergoes solidication with T changes as well as results of increase in the resistance. For the foregoing reasons the characteristic time is very small.

Solidication time
Millers approach Following Eckerts work, Miller and his student [20] altered the calculations14 and based the assumption that the conduction heat transfer characteristic time in die (liquid metal in solid phase) is the same order magnitude as the solidication time. This assumption leads them to conclude that the main resistance to the solidication is in the interface between the die and mold 15 . Hence they conclude that the solidied front moves according to the following hsl vn = hT (3.25)

Where here h is the innovative heat transfer coecient between solid and solid16 and vn is front velocity. Then the lling time is given by the equation ts = hsl hT (3.26)

14 Miller and his student calculate the typical forces required for clamping. The calculations of Miller has shown an interesting phenomenon in which small casting (2[kg]) requires a larger force than heavier casting (20[kg])?! Check it out in their paper, page 43 in NADCA Transaction 1997! If the results extrapolated (not to much) to about 50[kg] casting, no force will be required for clamping. Furthermore, the force for 20 [kg] casting was calculated to be in the range of 4000[N ]. In reality, this kind of casting will be made on 1000 [ton] machine or more (3 order of magnitude larger than Miller calculation suggested). The typical required force should be determined by the plunger force and the machine parts transient characteristics etc. Guess, who sponsored this research and how much it cost! 15 An example how to do poor research. These kind of research works are found abundantly in Dr. Miller and Dr. J. Brevick from Ohio State Univerity. These works when examined show contractions with the logic and the rest of the world of established science. 16 This coecient is commonly used either between solid and liquid, or to represent the resistance between two solids. It is hoped that Miller and coworkers refer that this coecient to represent the resistance between the two solids since it is a minor factor and does not determine the characteristic time.

3.4. ESTIMATES OF THE TIME SCALES IN DIE CASTING

49

where designates the half die thickness. As a corollary conclusion one can arrive from this construction is that the lling time is linearly proportional to the die thickness since hsl /hT is essentially constant (according to Miller). This interesting conclusion contradicts all the previous research about solidication problem (also known as the Stefan problem). That is if h is zero the time is zero also. The author is not aware of any solidication problem to show similar results. Of course, Miller has all the experimental evidence to back it up! Present approach Heat balance at the liquid-solid interface yields (Tl Ts ) (3.27) n where n is the direction perpendicular to the surface and has to be taken at the solid phase see Appendix 10. Additionally note that in many alloys, the density changes during the solidication and is substantial which has a signicant eect on the moving of the liquid/solid front. It can be noticed that at the die interface ks T/n kd T/n = (opposite to Miller) and further it can be assumed that temperature gradient in the liquid side, T/n 0 , is negligible compared to other uxes. Hence, the speed of the solid/liquid front moves s hsf vn = k vn = k Ts Tl kTM B s hsl n s hsl L (3.28)

Notice the dierence to equation (3.26) The main resistance to the heat transfer from the die to the mold (cooling liquid) is in the die mold. Hence, the characteristic heat transfer from the mold is proportional to TM B /L17 . The characteristic temperature dierence is between the melting temperature and the boiling temperature. The time scale for the front can be estimated by ts = = s hsl
2 L

vs

kd TM B

(3.29)

Note that the solidication time isnt a linear function of the die thickness, , but a function of 2 18 .

Dissipation Time
Examples of how dissipation is governing the ow can be found abundantly in nature. l l l l +u +v +w t x y z = l 2 l 2 l 2 l + + 2 2 x y z 2 + (3.30)

17 The estimate can be improved by converting the resistances of the die to be represented by die length and the same for the other resistance into the cooling liquid i.e. 1/ho + L/k + CD ots + 1/hi . 18 L can be represented by for example, see more simplied assumption leads to pure = 2 .

50

CHAPTER 3. DIMENSIONAL ANALYSIS

Where the dissipation function is dened as =2 w y + u x v z

2

+
2

v y w y

+ +

v y v z
2

+ 2 3

v x

u y

+
2

u v w + + x y z

(3.31)

Since the dissipation characteristic time isnt commonly studied in regular uid mechanics, we rst introduce two classical examples of dissipation problems. First problem deals with the oscillating manometer and second problem focuses on the rigid body brought to a rest in a thin cylinder.
D air air H equilibrioum level H

lowest level for the liquid

Fig a.

Fig b.
Mass, spring

Oscillating manometer

Fig. -3.13. The oscillating manometer for the example 3.1.

Example 3.1: A liquid in manometer is disturbed from a rest by a distance of H0 . Assume that the ow is laminar and neglected secondary ows. Describe H(t) as a function of time. Dened 3 cases: 1)under damping, 2) critical damping, and 3) over damping. Discuss the physical signicance of the critical damping. Compute the critical radius to create the critical damping. For simplicity assume that liquid is incompressible and the velocity prole is parabolic. Solution The conservation of the mechanical energy can be written as d dt (total of inflow of) + potential energy
rate of increase of kinetic and potential energy in system total net rate of surroundings work on the system inflow = (total of energy of) + (total of inflow of) + kinetic potential energy total rate mechanical energy dissipated because viscosity

total work due to expansion or compression of fluid

(3.32)

3.4. ESTIMATES OF THE TIME SCALES IN DIE CASTING

51

The chosen system is the liquid in the manometer. There is no ow in or out of the liquid of the manometer, and thus, terms that deal with ow in or out are canceled. It is assumed that the surface at the interface is straight without end eects like surface tension. This system is unsteady and therefore the velocity prole is function of the time and space. In order to demonstrate the way the energy dissipation is calculated it is assumed the velocity is function of the radius and time but separated. This assumption is wrong and cannot be used for real calculations because the real velocity prole is not separated and can have positive and negative velocities. It is common to assume that velocity prole is parabolic which is for the case where steady state is obtained. This assumption can used as a limiting case and the velocity prole is r U (r, t) = U (r) = U0 (t) 1 R
2
U0
R

(3.33)

H(t)

U0 1

r 2 R

V = H R2

where R the radius of the manometer. The velocity at the center is a function of time but independent of the Length. It can be noticed that this equation dim:eq:velocityH Fig. -3.14. Mass Balance to determine the is problematic because it breaks the assump- relationship between the U0 and the Height, H. tion of the straight line of the interface. The relationship between the velocity at the center, U0 to the height, H(t) can be obtained from mass conservation on left side of the manometer (see Figure 3.14) is d H R2 = dt
R 0

U0

r 1 R

dA

2 r dr

(3.34)

Equation (3.34) relates H(t) to the center velocity, U0 , and the integration results in dH U0 = dt 2 (3.35)

Note that H(t) isnt a function of the radius, R. This relationship (3.35) is based on the denition that U0 is positive for the liquid owing to right and therefore the height decreases. The total kinetic energy in the tube is then
L R 0

Kk =
0

U0 2 2

r 1 R

2 2

dA

2 r dr d =

L U0 2 R 2 6

(3.36)

where L is the total length (from one interface to another) and d is a coordinate running along the axis of the manometer neglecting the curvature of the U shape. It can be noticed that L is constant for incompressible ow. It can be observed that the disturbance of the manometer creates a potential energy which can be measured from a datum at the maximum lower point. The maximum potential energy is obtained when H is either maximum or minimum. The maximum kinetic energy is obtained when H

52

CHAPTER 3. DIMENSIONAL ANALYSIS

is zero. Thus, at maximum height, H0 the velocity is zero. The total potential of the system is then left side
H0 H dV

right side
2 H0 +H 0 dV

Kp =
0

( g ) R d +

( g ) R2 d = H0 2 + H 2 g R2 (3.37)

The last term to be evaluated is the viscosity dissipation. Based on the assumptions in the example, the velocity prole is function only of the radius thus the only gradient of the velocity is in the r direction. Hence
R

Ed = = L
0

dU dr

dA

2 r dr

(3.38)

The velocity derivative can be obtained by using equation (3.33) as dU = U0 dr 2 r R2 = dU dr

2

4 r2 U0 2 R4

(3.39)

Substituting equation (3.39) into equation (3.37) reads ( dU ) dr

R
2

Ed = 2 L
0

4 r2 U0 2 r dr = 2 L R2 U0 2 R2 R R

(3.40)

The work done on system is neglected by surroundings via the pressure at the two interfaces because the pressure is assumed to be identical. Equation (3.32) is transformed, in this case, into d (Kk + Kp ) = Ed dt The kinetic energy derivative with respect to time (using equation (3.35)) is d Kk d = dt dt L U0 2 R 2 6 = L R2 U0 4 L R2 d H d2 H 2 U0 = 6 dt 3 dt dt2 (3.42) (3.41)

The potential energy derivative with respect to time is d d Kp = dt dt H0 2 + H 2 g R 2 = 2 H dH g R2 dt (3.43)

Substituting equations (3.43), (3.42) and (3.38) into equation (3.41) results in 4 L R 2 d H d2 H dH + 2H g R 2 + 2 L U0 2 = 0 2 3 dt dt dt (3.44)

3.4. ESTIMATES OF THE TIME SCALES IN DIE CASTING Equation (3.44) can be simplied using the identity of (3.35) to be d2 H 6 dH 3g + + H=0 2 2 dt dt R 2L

53

(3.45)

This equation is similar to the case mass tied to a spring with damping. This equation is similar to RLC circuit19 . The common method is to assume that the solution of the form of A e t where the value of A and will be such determined from the equation. When substituting the guessed function into result that having two possible solution which are
6R2 6 R2 2

= Thus, the solution is

6g L

(3.46)

H = A e1 t + A e2 t H = Ae
t

= =

1 = 2 1 = 2 = (3.47)

+ Ae

The constant A1 and A2 are to be determined from the initial conditions. The value under the square root determine the kind of motion. If the value is positive then the system is overdamped and the liquid height will slowly move the equilibrium point. If the value in square is zero then the system is referred to as critically damped and height will move rapidly to the equilibrium point. If the value is the square root is negative then the solution becomes a combination of sinuous and cosines. In the last case the height will oscillate with decreasing size of the oscillation. The critical radius is then Rc =
4

6 2 L g 2

(3.48)

It can be observed that this analysis is only the lower limit since the velocity prole is much more complex. Thus, the dissipation is much more signicant.
End Solution

Example 3.2: A thin (t/D 1) cylinder full with liquid is rotating in a velocity, . The rigid body is brought to a stop. Assuming no secondary ows (Bernards cell, etc.), describe the ow as a function of time. Utilize the ratio 1 t/D. d2 X + dt2
2

dX +X =0 dt

(3.49)

Discuss the case of rapid damping, and the case of the characteristic damping
19 An electrical circuit consisting of a resistor (R), an inductor (L), and a capacitor (C), connected in series or in parallel.

54

CHAPTER 3. DIMENSIONAL ANALYSIS

Fig. -3.15. Rigid body brought into rest.

Solution

End Solution

These examples illustrate that the characteristic time of dissipation can be assessed by (du/dy )2 thus given by 2 /. Note the analogy between ts and tdiss , for which 2 appears in both of them, the characteristic length, , appears as the typical die thickness.

3.4.2

The ratios of various time scales

The ratio of several time ratios can be examined for typical die casting operations. The ratio of solidication time to the lling time tf Lkd TM B Ste = ts P r Re U s hsl L where Re Ste Reynolds number Stefan number
cp U lm
lm

lm s

kd klm

L L

(3.50)

TM B

hsl

the discussion is augmented on the importance of equation (3.50). The ratio is extremely important since it actually denes the required lling time. tf = C lm s kd klm L L Ste P r Re (3.51)

3.5. SIMILARITY APPLIED TO DIE CAVITY

55

At the moment, the constant, C, is unknown and its value has to come out from experiments. Furthermore, the constant is not really a constant and is a very mild function of the geometry. Note that this equation is also dierent from all the previously proposed lling time equations, since it takes into account solidication and lling process20 . The ratio of liquid metal conduction characteristic time to characteristic lling time is given by tc d U L2 U L2 L2 = = Re P r tf L L L (3.52)

The solidication characteristic time to conduction characteristic time is given by s hsl Ld ts 1 = 2 tc Ste kd TM B L s d cp lm cp d (3.53)

The ratio of the lling time and atomization is ta viscosity U = Ca tf L L 6 108 (3.54)

Note that , in this case, is the thickness of the gate and not of the die cavity. ta momentum 2U 2 = We tf L L 0.184 (3.55)

which means that if atomization occurs, it will be very fast compared to the lling process. The ratio of the dissipation time to solidication time is given by
2 tdiss kd TM B = ts lm s hsl L

Ste Pr

kd klm

lm s

100

(3.56)

this equation yields typical values for many situations in the range of 100 indicating that the solidication process is as fast as the dissipation. It has to be noted that when the solidication progress, the die thickness decreases. The ratio, /L, reduced as well. As a result, the last stage of the solidication can be considered as a pure conduction problem as was done by the English group.

3.5 Similarity applied to Die cavity

This section is useful for those who are dealing with research on die casting and or other casting process.
20 In this book, this equation because of its importance is referred to as EckertBarMeirs equation. If you have good experimental work, your name can be added to this equation.

56

CHAPTER 3. DIMENSIONAL ANALYSIS

3.5.1

Governing equations

The lling of the mold cavity can be divided into two periods. In the rst period (only uid mechanics; minimum heat transfer/solidication) and the second period in which the solidication and dissipation occur. This discussion deals with how to conduct experiments in die casting21 . It has to be stressed that the conditions downstream have to be understood prior to the experiment with the die lling. The liquid metal velocity prole and ow pattern are still poorly understood at this stage. However, in this discussion we will assume that they are known or understood to same degree22 . The governing equations are given in the preceding sections and now the boundary conditions will be discussed. The boundary condition at the solid interface for the gas/air and for the liquid metal are assumed to be noslip condition which reads ug = vg = wg = ulm = vlm = wlm = 0 (3.57)

where the subscript g is used to indicate the gas phase. It is noteworthy to mention that this can be applied to the case where liquid metal is mixed with air/gas and both are touching the surface. At the interface between the liquid metal and gas/air, the pressure jump is expressed as (3.58) p r1 + r2 where r1 and r2 are the principal radii of the free surface curvature, and, , is the surface tension between the gas and the liquid metal. The surface geometry is determined by several factors which include the liquid movement23 instabilities etc. Now on the dicult parts, the velocity at gate has to be determined from the pQ2 diagram or previous studies on the runner and shot sleeve. The diculties arise due to the fact that we cannot assign a specic constant velocity and assume only liquid ow out. It has to be realized that due to the mixing processes in the shot sleeve and the runner (especially in a poor design process and runner system, now commonly used in the industry), some portions at the beginning of the process have a signicant part which contains air/gas. There are several possibilities that the conditions can be prescribed. The rst possibility is to describe the pressure variation at the entrance. The second possibility is to describe the velocity variation (as a function of time). The velocity is reduced during the lling of the cavity and is a function of the cavity geometry. The change in the velocity is sharp in the initial part of the lling due to the change from a free jet to an immersed jet. The pressure varies also at the entrance, however, the variations are more mild. Thus, it is a better possibility24 to consider the pressure prescription. The simplest assumption is constant pressure P = P0 = 1 U0 2 2 (3.59)

21 Only minimal time and eorts was provided how to conduct experiments on the lling of the die. In the future, other zones and dierent processes will be discussed. 22 Again the die casting process is a parabolic process. 23 Note, the liquid surface cannot be straight, for unsteady state, because it results in no pressure gradient and therefore no movement. 24 At this only an intelligent guess is possible.

3.5. SIMILARITY APPLIED TO DIE CAVITY We also assume that the air/gas obeys the ideal gas model. g = P RT

57

(3.60)

where R is the air/gas constant and T is gas/air temperature. The previous assumption of negligible heat transfer must be inserted and further it has to be assumed that the process is polytropic25 . The dimensionless gas density is dened as = = 0 P0 P
1 n

(3.61)

The subscript 0 denotes the atmospheric condition. The air/gas ow rate out the cavity is assumed to behave according to the model in Chapter 9. Thus, the knowledge of the vent relative area and 4f L are important D parameters. For cases where the vent is well designed (vent area is near the critical area or above the density, g can be determined as was done by [5]). To study the controlling parameters, the equations are dimensionlessed. The mass conservation for the liquid metal becomes lm lm u lm lm v + + t x y
lm

lm w z

lm

=0

(3.62)

where x = x , y = y/ , z = z/ , u = u/U0 , v = v/U0 , w = w/U0 and the dimensionless time is dened as t = tU0 , where U0 = 2P0 /. Equation (3.62) can be similar under the assumption of constant density to read u lm v lm w lm + + =0 x y z (3.63)

Please note that this simplication can be used for the gas phase. The momentum equation for the liquid metal in the x-coordinate assuming constant density and no body forces reads lm u lm lm u lm lm u lm lm u lm +u +v +w = t x y z p lm 1 + x Re 2 ulm 2v 2w + + x 2 ylm 2 zlm 2 (3.64)

where Re = U0 /lm and p = p/P0 . The gas phase continuity equation reads g ug g vg g wg g + + + =0 t x y z
25 There

(3.65)

are several possibilities, this option is chosen only to obtain the main controlling parameters.

58

CHAPTER 3. DIMENSIONAL ANALYSIS

The gas/air momentum equation26 is transformed into g u g t +u g u g x +v g u g y +w g u g z = + 2 vg y

2

p g lm g0 1 + x g lm Re

2 ug x 2

2 wg z
2

(3.66)

Note that in this equation, additional terms were added, (lm /g )(g0 /lm ). The no-slip conditions are converted to: ug = vg = wg = ulm = vlm = wlm = 0 The surface between the liquid metal and the air satisfy 1 p (r1 + r2 ) = We where the p , r1 , and r2 are dened as r1 = r1 / r2 = r2 / The solution to equations has the form of u v w p = = = = fu x , y , z , Re, W e, fv fw fp A 4f L g lm , D , n, , Ac lm g g lm A 4f L x , y , z , Re, W e, , D , n, , Ac lm g A 4f L g lm x , y , z , Re, W e, , , n, , Ac D lm g g lm A 4f L , n, x , y , z , Re, W e, , , Ac D lm g (3.67)

(3.68)

(3.69)

If it will be found that equation (3.66) can be approximated27 by p g u g u g u g u g +u +v +w t x y z x then the solution is reduced to u v w p = = = = fu x , y , z , Re, W e, fv fw fp A 4f L , ,n Ac D A 4f L x , y , z , Re, W e, , ,n Ac D A 4f L x , y , z , Re, W e, , ,n Ac D A 4f L x , y , z , Re, W e, , ,n Ac D (3.70)

(3.71)

26 In writing this equation, it is assumed that viscosity of the air is independent of pressure and temperature. 27 This topic is controversial in the area of two phase ow.

3.6. SUMMARY OF DIMENSIONLESS NUMBERS

59

At this stage, it is not known if it is the case and if it has to come out from the experiments. The density ratio can play a role because two phase ow characteristic is a major part of the lling process.

3.5.2

Design of Experiments
28

Under Construction

3.6 Summary of dimensionless numbers

This section summarizes all the major dimensionless parameters and what eects they have on the die casting process. Reynolds number Re = U 2 / internal Forces = 2 U/ viscous forces

Reynolds number represents the ratio of the momentum forces to the viscous forces. In die casting, Reynolds number plays a signicant role which determines the ow pattern in the runner and the vent system. The discharge coecient, CD , is used in the pQ2 diagram is determined largely by the Re number through the value of friction coecient, f, inside the runner. Eckert number Ec = 1/2U 2 inertial energy = 1/2cp T thermal energy

Eckert number determines if the role of the momentum energy transferred to thermal energy is signicant. Brinkman number Br = heat production by viscous dissipation U 2 / 2 = 2 kT / heat transfer transport by conduction

Brinkman number is a measure of the importance of the viscous heating relative the conductive heat transfer. This number is important in cases where large velocity change occurs over short distances such as lubricant ow (perhaps, the ow in the gate). In die casting, this number has small values indicating that practically the viscous heating is not important.
28 See

for time being Eckerts paper

60 Mach number Ma =

CHAPTER 3. DIMENSIONAL ANALYSIS

U
p

For ideal gas (good assumption for the mixture of the gas leaving the cavity). It becomes U characteristic velocity M = = gas sound velocity RT Mach number determines the characteristic of ow in the vent system where the air/gas velocity is reaching to the speed of sound. The air is chocked at the vent exit and in some cases other locations as well for vacuum venting. In atmospheric venting the ow is not chocked for large portion of the process. Moreover, the ow, in well design vent system, is not chocked. Yet the air velocity is large enough so that the Mach number has to be taken into account for reasonable calculation of the CD . Ozer number Oz =
CD 2 Pmax 2 Qmax A3

A3 Qmax

CD 2

eective static pressure energy Pmax = average kinematic energy

One of the most important number in the pQ2 diagram calculation is Ozer number. This number represents how good the runner is designed. Froude number Fr = U 2 / inertial forces = g gravity forces

F r number represent the ratio of the gravity forces to the momentum forces. It is very important in determining the critical slow plunger velocity. This number is determined by the height of the liquid metal in the shot sleeve. The Froude number does not play a signicant role in the lling of the cavity. Capillary number Ca = inertial forces U 2 / = g gravity forces

capillary number (Ca) determine when the ow during the lling of the cavity is atomized or is continuous ow (for relatively low Re number).

3.7. SUMMARY Weber number We = 1/2U 2 inertial forces = 1/2/ surface forces

61

W e number is the other parameter that govern the ow pattern in the die. The ow in die casting is atomized and, therefore, W e with combinations of the gate design also determine the drops sizes and distribution. Critical vent area Ac = V (0) ctmax mmax

The critical area is the area for which the air/gas is well vented.

3.7 Summary
The dimensional analysis demonstrates that the uid mechanics process, such as the lling of the cavity with liquid metal and evacuation/extraction of the air from the mold, can be dealt with when heat transfer is neglected. This provides an excellent opportunity for simple models to predict many parameters in the die casting process. It is recommended for interested readers to read Eckerts book Analysis of Heat and Mass transfer to have better and more general understanding of this topic.

3.8 Questions
Under construction

62

CHAPTER 3. DIMENSIONAL ANALYSIS

CHAPTER 4 Fundamentals of Pipe Flow

Chapter Under heavy construction

4.1 Introduction
The die casting engineer encounters many aspects of network ow. For example, the liquid metal ows in the runner is a network ow. The ow of the air and other gases out of the mold through the vent system is also another example network ow. The pQ2 diagram also requires intimate knowledge of the network ow. However, most die casting engineers/researchers are unfamiliar with uid mechanics and furthermore have a limited knowledge and understanding of the network ow. Therefore, this chapter is dedicated to describe a brief introduction to a ow in a network. It is assumed that the reader does not have extensive background in uid mechanics. However, it is assumed that the reader is familiar with the basic concepts such as pressure and force, work, power. More comprehensive coverage can be found in books dedicated to uid mechanics and pipe ow (network for pipe). First a discussion on the relevancy of the data found for other liquids to the die casting process is presented. Later a simple ow in a straight pipe/conduit is analyzed. Dierent components which can appear in network are discussed. Lastly, connection of the components in series and parallel are presented.

4.2 Universality of the loss coecients

Die casting engineers who are not familiar with uid mechanics ask whether the loss coecients obtained for other liquids should/could be used for the liquid metal. To answer this question, many experiments have been carried out for dierent liquids owing in dierent components in the last 300 years. An example of such exper-

63

64
1 Head Loss [meter] 0.1 0.01 0.001 0.0001 1e05 1e06
0.00 0.00 0.50 1.00 0.02 0.06 0.12 0.01 0.03 0.25

CHAPTER 4. FUNDAMENTALS OF PIPE FLOW

1.2 1.0
Loss Coefficient

0.8 0.6 0.4 0.2 0.0

Air Crude Oil Hydrogen Mercury Water

2.00 4.00 8.00

1e07

100

1000
Reynolds Number

10000

Velocity[m/sec]

Fig a.

Friction of orice as

Fig b.

The collapsed

a function velocity.

results as funciton Reynolds number.

Fig. -4.1. The results for the ow in a pipe with orice.

iments is a ow of dierent liquids in a pipe with an orice (see Figure 4.1). Dierent liquids create signicant head loss for the same velocity. Moreover, the dierences for the dierent liquids are so signicant that the similarity is unclear as shown in Figure ??. As the results of the past geniuses work, it can be shown that when results are normalized by Reynolds number (Re) instated of the velocity and when the head loss is replaced by the loss coecient, UH g one obtains that all the lines are collapsed on to a 2 /2 single line as shown in Figure 4.1b. This result indicates that the experimental results obtained for one liquid can be used for another liquid metal provided the other liquid is a Newtonian liquid1 . Researchers shown that the liquid metal behaves as Newtonian liquid if the temperature is above the mushy zone temperature. This example is not correct only for this spesic geometry but is correct for all the cases where the results are collapsed into a single line. The parameters which control the problem are found when the results are collapsed into a single line. It was found that the resistance to the ow for many components can be calculated (or extracted from experimental data) by knowing the Re number and the geometry of the component. In a way you can think about it as a prof of the dimensional analysis (presented in Chapter 3).

4.3 A simple ow in a straight conduit

A simple and most common component is a straight conduit as shown in Figure 4.2. The simplest conduit is a circular pipe which would be studied here rst. The entrance problem and the unsteady aspects will be discussed later. The parameters that the die casting engineers interested are the liquid metal velocity, the power to Fig. -4.2. General simple conduit description. drive this velocity, and the pressure dierence occur for the required/desired velocity. What determine these parameters? The velocity is determined by the pressure dierence applied on the pipe and the resistance

1 Newtonian

liquid obeys the following stress law = dU dy

4.3. A SIMPLE FLOW IN A STRAIGHT CONDUIT

65

to the ow. The relationship between the pressure dierence, the ow rate and the resistance to the ow is given by the experimental equation (4.1). This equation is used because it works2 . The pressure dierence determined by the geometrical parameters and the experimental data which expressed by f 3 which can be obtained from Moodys diagram. P = f L U2 L U2 ; H = f D 2 D 2g (4.1)

Note, head is energy per unit weight of uid (i.e. Force x Length/Weight = Length) and it has units of length. Thus, the relationship between the Head (loss) and the pressure (loss) is P = H g (4.2)

The resistance coecient for circular conduit can be dened as L KF = f (4.3) Fig. -4.3. General simple conduit description. D This equation is written for a constant density ow and a constant cross section. The ow rate is expressed as Q=UA (4.4) The cross sectional area of circular is A = r2 = D2 /4, using equation (4.4) and substituting it into equation (4.1) yields 16 L 2 Q (4.5) 2 D3 The equation (4.5) shows that the required pressure dierence, P , is a function of 1/D3 which demonstrates the tremendous eect the diameter has on the ow rate. The length, on the other hand, has mush less signicant eect on the ow rate. The power which requires to drive this ow is give by P = f P = QP (4.6)

These equations are very important in the understanding the economy of runner design, and will be studied in Chapter 12 in more details. The power in terms of the geometrical parameters and the ow rate is given P = {
2 Actually

L Q D

(4.7)

there are more reasons but they are out of the scope of this book this stage, we use dierent denition than one used in Chapter A. The dierence is by a factor of 4. Eventially we will adapt one system for the book.
3 At

66

CHAPTER 4. FUNDAMENTALS OF PIPE FLOW

4.3.1

Examples of the calculations

Example 4.1: calculate the pressure loss (dierence) for a circular cross section pipe for driving aluminum liquid metal at velocity of 10[m/sec] for a pipe length of 0.5 [m] (like a medium quality runner) with diameter of 5[mm] 10[mm] and 15[mm] Solution

This is example 4.3.1 Example 4.2: calculate the power required for the above example Solution

4.4 Typical Components in the Runner andVent Systems

In the calculations of the runner the die casting engineer encounter beside the straight pipe which was dealt in the previous section but other kind of components. These components include the bend, Yconnection and tangential gate, regular gate, the extended Y connection and expansion/contraction (including the abrupt expansion/contraction). In this section a general discussion on the good design practice for the dierent component is presented. A separate chapter is dedicated to the tangential runner due to its complication.

4.4.1

bend

bad english, change it please

The resistance in the bend is created because a change in the momentum and the ow pattern. Engineers normally convert the bend to equivalent conduit length. This conversion produces adequate results in same cases while in other it might introduce larger error. The knowledge of this accuracy of this conversion is very limited because limited study have been carry out for the characteristic of ows in die casting. From the limited information the author of this book gadered it seem that it is reasonable to carry this conversion for the calculations of liquid metal ow resistance while in the air/liquid metal mixture it far from adequate. Moreover, hole of our knowledge of the gas ow in vent system are far more large. Nevertheless, for the engineering purpose at this stage it seem that some of the errors will cancel each other and the end result will be much better. The schematic of a bend commonly used in die casting is shown in Figure ??. The resistance of the bend is a function of several parameters: angle, , radius, R and the geometry before and after the
cross section

Fig. -4.4. A sketch of the bend in die casting.

4.5. PUTTING IT ALL TO TOGETHER

67

bend. Commonly, the runner is made with the same geometry before and after the bend. Moreover, we will assume in this discussion that downstream and upstream do not inuence that ow in the ow. This assumption is valid when there is no other bend or other change in the ow nearby. In cases that such a change(s) exists more complicated analysis is required. In the light of the for going discussion, we left with two parameters that control the resistance, the angle, , and the radius, R As larger the angle is larger the resistance will be. In the practice today, probably because the way the North American Die Casting Association teaching, excessive angle can be found through the industry. It is recommended never to exceed the straight angle (900 ). Figure ?? made from a data taken from several sources. From the Figure it is clear that optimum radius should be around 3.

4.4.2

Y connection
picture of Y connection

The Yconnection reprsent a split in the runner system. The resistence

4.4.3

Expansion/Contraction

One of the undisirable element is the runner system is sudden change in the conduict area. In some instance they are inevodeble. We will disscuss how to design and what are the better design options which availble for the engineer.

4.5 Putting it all to Together

There are two main kinds of connections; series and parallel. The resistance in the series connection has to be added in a fashion similar to electrical resistance i.e. every resistance has to be added plainly to the total resistance. There are many things that contribute to the resistance besides the regular length, i.e. bends, expansions, contractions etc. All these connections are of series type.

4.5.1

Series Connection

The ow rate in dierent locations is a function of the temperature. Eckert [13] demonstrated that the heat transfer is insignicant in the duration of the lling of the cavity, and therefore the temperature of the liquid metal can be assumed almost constant during the lling period (which in most cases is much less 100 milliseconds). As such, the solidication is insignicant (the liquid metal density changes less than 0.1% in the runner); therefore, the volumetric ow rate can be assumed constant: Q1 = Q2 = Q3 = Qi (4.8)

68

CHAPTER 4. FUNDAMENTALS OF PIPE FLOW Clearly, the pressure in the points is dierent and P1 = P2 = P3 = (4.9)

However the total pressure loss is composed of from all the small pressure loss P1 Pend = (P1 P2 ) + (P2 P3 ) + Every single pressure loss can be written as Pi1 Pi = Ki U2 2 (4.11) (4.10)

There is also resistance due to parallel connection i.e. y connections, y splits and manifolds etc. First, lets look at the series connection. (see Figure ??). where: Kbend L f the resistance in the bend length of the duct (vent), friction factor, and

4.5.2

Parallel Connection

An example of the resistance of parallel connection (see Figure ??). The pressure at point 1 is the same for two branches however the total ow rate is the combination Qtotal = Qi + Qj between two branches and the loss in the junction is calculated as To add a gure and check if the old one is good
Fig. -4.5. A parallel connection

(4.12)

CHAPTER 5 Flow in Open Channels

5.1 Introduction
Fl ow

One of the branches of the uid mechanics discussed in Chapter 2. Here we expand this issue further because it is give the basic understanding to the wave phenomenon. There are numerous books that dealing with open channel ow and the interested reader can broader his/her knowledge by reading book such as OpenChannel Hydraulics by Ven Te Chow (New York: McGraw-Hill Book Company, Inc. Fig. -5.1. Equilibrium of Forces in an open 1959). Here a basic concepts for the non- channel. Fluid Mechanics Engineers are given. The ow in open channel ow in steady state is balanced by between the gravity forces and mostly by the friction at the channel bed. As one might expect, the friction factor for open channel ow has similar behavior to to one of the pipe ow with transition from laminar ow to the turbulent at about Re 103 . Nevertheless, the open channel ow has several respects the cross section are variable, the surface is at almost constant pressure and the gravity force are important. The ow of a liquid in a channel can be characterized by the specic energy that is associated with it. This specic energy is comprised of two components: the hydrostatic pressure and the liquid velocity1 .
y x d

1 The

velocity is an average velocity

69

70

CHAPTER 5. FLOW IN OPEN CHANNELS The energy at any point of height in a rectangular channel is e= U2 P + +z 2g (5.1)

why? explain

and, since

+ z = y for any point in the cross section (free surface), e=y+ U2 2g (5.2)

where: e y g U specic energy per unit height of the liquid in the channel acceleration of gravity average velocity of the liquid

If the velocity of the liquid is increased, the height, y, has to change to keep the same ow rate Q = q b = b y U . For a specic ow rate and cross section, there are many combinations of velocity and height. Plotting these points on a diagram, with the ycoordinate as the height and the xcoordinate as the specic energy, e, creates a parabola on a graph. This line is known as the specic energy curve. Several conclusions can be drawn from Figure ??. First, there is a minimum Fig. -5.2. Specic Energy and momentum energy at a specic height known as the Curves. critical height. Second, the energy increases with a decrease in the height when the liquid height is below the critical height. In this case, the main contribution to the energy is due to the increase in the velocity. This ow is known as the supercritical ow. Third, when the height is above the critical height, the energy increases again. This ow is known as the subcritical ow, and the energy increase is due to the hydrostatic pressure component. The minimum point of energy curve happens to be at U = g yc (5.3)
Y EL Y2
specific energy line

specific thrust line

Y3

YC

Y1

e, f

where the critical height is dened by yc = Thrust is dened as f= y2 U2 +y 2 g (5.5)

3

q2 g

(5.4)

5.2. TYPICAL DIAGRAMS The minimum thrust also happens to be at the same point U = can dene the dimensionless number as: g yh Fr = 2 U

71 g y. Therefore, one (5.6)

Dividing the velocity by gy provides one with the ability to check if the ow is above or below critical velocity. This quantity is very important, and its signicance can be studied from many books on uid mechanics. The gravity eects are measured by the Froude number which is dened by equation (5.6).

5.2 Typical diagrams 5.3 Hydraulic Jump

The ow can change only from a supercritical ow to a subcritical ow, in which the height increases and the velocity decreases. There is no possibility for the ow to go in the reverse direction because of the Second Law of Thermodynamics (the explanation of which is out of the scope of this discussion). If there is no energy loss, the ow moves from point 1 to point 2 in Figure ??. In actuality, energy loss occurs in any situation, but sometimes it can be neglected in the calculations. In cases where the ow changes rapidly (such as with the hydraulic jump), the energy loss must be taken into account. In these cases, the ow moves from point 1 to point 3 and has energy loss (EL ). In many cases the change in the thrust is negligible, such as the case of the hydraulic jump, and the ow moves from point 1 to point 3 as shown in Figure ??. In 1981, Garber found the hydraulic jump in the shot sleeve which he called a wave. Garber built a model to describe this wave, utilizing mass conservation and Bernoullis equation (energy conservation). This model gives a set of equations relating plunger velocity and wave velocity to other geometrical properties of the shot sleeve. Over 150 years earlier, Blanger [?] demonstrated that the energy is dissipated, and that e energy conservation models cannot be used to solve hydraulic jump. He demonstrated that the dissipation increases with the increase of the liquid velocity before the jump. This conclusion is true for any kind of geometry. A literature review demonstrates that the hydraulic jump in a circular cross section (like in a shot sleeve) appears in other cases, for example a ow in a storm sewer systems. An analytical solution that describes the solution is BarMeirs formula and is shown in Figure 8.4. The energy loss concept manifests itself in several designs, such as in the energy dissipating devices, in which hydraulic jumps are introduced in order to dissipate energy. The energydissipating devices are so common that numerous research works have been performed on them in the last 200 hundreds years. An excellent report by the U.S. Bureau of Reclamation [7] shows the percentage of energy loss. However, Garber, and later other researchers from Ohio State University [8], failed to know/understand/use this information.

72

CHAPTER 5. FLOW IN OPEN CHANNELS

CHAPTER 6 Runner Design

Under construction please ignore the rest of the chapter.

6.1 Introduction
In this chapter the design and the dierent relationship between runner segments are studied herein. The rst step in runner design is to divide the mold into several logical sections. The volume of every section has to be calculated. Then the design has to ensure that the gate velocity and the lling time of every section to be as recommend by experimental results. At this stage there is no known reliable theory/model known to the author to predict these values. The values are based heavily on semi-reliable experiments. The Backward Design is discussed. The reader with knowledge in electrical engineering (electrical circles) will notice in some similarities. However, hydraulic circuits are more complex. Part of the expressions are simplied to have analytical expressions. Yet, in actuality all the terms should be taken into considerations and commercial software such DiePerfectshould be used.

6.1.1

Backward Design

Suppose that we have n sections with n gates. We know that volume to be delivered at gate i and is denoted by vi.The gate velocity has to be in a known range. The lling time has to be in a known function and we recommend to use Eckert/BarMeirs formula. For this discussion it is assumed that the lling has to be in known range and the ow rate can be calculated by Qi = Vi ti 73 (6.1)

74 Thus, gate area for the section Ai = Qi Ugate

CHAPTER 6. RUNNER DESIGN

(6.2)

Armed with this knowledge, one can start design the runner system.

6.1.2

Connecting runner segments

vents vents

Design of connected runner segments have insure that the ow rate at each segment has to be designed one. In Figure ??a branches i and j are connected to branch at point K. The pressure drop (dierence) on branches i and j has to be the same since the pressure in the mold cavity is the same for both segments. The sum of the ow rates for both branch has to be equal to ow rate in branch Q = Qi + Qj = Qj = Q Qi (6.3) The ow rate in every branch is related to the pressure dierence by Qi = P Ri

Mold
i K j

b Biscuit

Fig. -6.1. A geometry of runner connection.

(6.4)

Where the subscript i in this case also means any branch e.g. i, j and so on. For example, one can write for branch j Oj = DeltaP Rj (6.5)

Utilizing the mass conservation for point K in which Qkappa = Oi + Oj and the fact that the pressure dierence, DeltaP , is the same thus we can write Qk = DeltaP DeltaP Ri Rj + = DeltaP Ri Rj Ri + Rj (6.6)

where we can dene equivalent resistance by R= Ri Rj Ri + Rj (6.7)

Lets further manipulate the equations to get some more important relationships. Using equation (??) and equation (??) DeltaP i = DeltaP j = Oi Ri = Oj Rj (6.8)

6.1. INTRODUCTION

75

The ow rate in a branch j can be related to ow rate in branch i and corresponding resistances Oj = Ri Oi Rj (6.9)

Using equation (??) and equation (??) one can obtain Oi Rj = Qk Rj + Ri Solving for the resistance ratio since the ow rate is known Ri Oi = 1 Rj Qk (6.11) (6.10)

6.1.3

Resistance

1 2

What does the resistance include? How to achieve resistance ratio in the previous equation (??) will be discussed herein further. The total resistance reads

Fig. -6.2. y connection.

R = Rii + R + Rgeometry + Rcontraction + Rki + Rexit

(6.12)

The contraction resistance, Rcontraction , is the due the contraction of the gate. The exit resistance, Rexit , is due to residence of the liquid metal in mold cavity. Or in other words, the exit resistance is due the lost of energy of immersed jet. The angle resistance, R is due to the change of direction. The Rki is the resistance due to ow in the branch on branch i. The geometry resistance Rgeometry , is due to who rounded the connection. DeltaP L U2 =f HD 2 since Ui =
Oi A

(6.13)

DeltaP L Oi 2 =f HD 2A2 L Oi 2 DeltaP = (C)f HD 3 2

(6.14)

(6.15)

76 Lets assume further that Li = Lj ,

Oi Oj

CHAPTER 6. RUNNER DESIGN = known (6.16)

fi = fj = f

(C)f

L Oi 2 L Oj 2 = (C)f HD 3 i 2 HD 3 j 2

(6.17)

HD i = HD j

Oi Oj

2 3

(6.18)

Comparison between scrap between (multi-lines) two lines to one line rst nd the diameter equivalent to two lines L Qk 2 L (Oi + Oj )2 3 = (C)f 2 2 HD k HD 3 k

DeltaP = (C)f

(6.19)

Oi 2 = subtitling in to

DeltaP 2 HD i3 f L

(6.20)

HD k =

(HD 3 + HD 3 i j

(6.21)

Now we know the relationship between the hydraulic radius. Let see what is the scrap dierence between them. put drawing of the trapezoid let scrap denoted by converting the equation i constL

HD i = the ratio of the scrap is

3/2

(6.22)

6.1. INTRODUCTION

77

i + j (HD i + 2 + HD j 2 ) = k HD k 2 and now lets write HD k in term of the two other (HD 2 + HD 2 ) i j HD 3 + HD 3 i j
2/3

(6.23)

(6.24)

78

CHAPTER 6. RUNNER DESIGN

CHAPTER 7 PQ Diagram Calculations

2

In conclusion, its just a plain sloppy piece of work.

7.1 Introduction

Referee II, see In the appendix B

The pQ2 diagram is the most common gate mold calculation, if any at all, are used by most 3 die casting engineers. The importance of gate this diagram can be demonstrated by the runner fact that tens of millions of dollars have been invested by NADCA, NSF, and other 2 major institutes here and abroad in the Shot sleeve 1 pQ2 diagram research. The pQ2 diagram is one of the manifestations of supply and demand theory which was developed by Alfred Marshall (18421924) in the turn of Fig. -7.1. Schematic of typical die casting mathe century. It was rst introduced to the chine. die casting industry in the late70s [12]. In this diagram, an engineer insures that die casting machine ability can fulll the die mold design requirements; the liquid metal is injected at the right velocity range and the lling time is small enough to prevent premature freezing. One can, with the help of the pQ2 diagram, and by utilizing experimental values for desired lling time and gate velocities improve the quality of the casting. In the die casting process (see Figure 7.1), a liquid metal is poured into the shot sleeve where it is propelled by the plunger through the runner and the gate into the mold. The gate thickness is very narrow compared with the averaged mold thickness and the runner thickness to insure that breakage point of the scrap occurs at that gate location. A solution of increasing the discharge coecient, CD , (larger conduits) results 79

perhaps put this section in general discussion

80

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

in a larger scrap. A careful design of the runner and the gate is required. First, the common pQ2 diagram1 is introduced. The errors of this model are analyzed. Later, the reformed model is described. Eects of dierent variables is studied and questions for students are given in the end of the chapter.

7.2

The common pQ2 diagram

The injection phase is (normally) separated into three main stages which are: slow part, fast part and the intensication (see Figure 7.2). In the slow part the plunger moves in the critical velocity to prevent wave formation and therefore expels maximum air/gas before the liquid metal enters the cavity. In the fast part the cavity supposed to be lled in such way to prevent premature freezing and to obtain the right lling pattern. The intensication part is to ll the cavity with additional material to compensate for the shrinkage porosity during the solidication process. The pQ2 diagram deals with the second part of the lling phase. In the pQ2 diagram, the solution is determined by nding the intersecting point of the runner/mold characteristic line with the pump (die casting machine) characteristic line. The intersecting point sometime refereed to as the operational point. The machine characteristic line is assumed to be understood to some degree and it requires nding experimentally two Time coecients. The runner/mold characteristic line requires knowledge on the eciency/discharge coecient, CD , thus it Fig. -7.2. A typical trace on a cold chamber is an essential parameter in the calcula- machine tions. Until now, CD has been evaluated either experimentally, to be assigned to specic runner, or by the liquid metal properties (CD ) [9] which is de facto the method used today and refereed herein as the common pQ2 diagram2 . Furthermore, CD is assumed constant regardless to any change in any of the machine/operation parameters during the calculation. The experimental approach is arduous and expensive, requiring the building of the actual mold for each attempt with average cost of $5,000$10,000 and is rarely used in the industry3 . A short discussion about this issue is presented in the Appendix B comments to referee 2. Herein the common model (constant CD ) is constructed. The assumptions made in the construction of the model as following
plunger location liquid metal reaches to the venting system starging filling the cavity liquid metal pressure at the plunger tip or the hydralic pressure cavity filling

this model is described in NADCAs books method has been suggested in the literature in which the CD is evaluated based on the volume to be lled [10]. The author does not know of anyone who use this method and therefore is not discussed in this book. Nevertheless, this method is as good as the common method. 3 if you now of anyone who use this technique please tell me about it.
2 Another

1 as

7.2.

THE COMMON PQ2 DIAGRAM

81

1. CD assumed to be constant and depends only the metal. For example, NADCA recommend dierent values for aluminum, zinc and magnesium alloys. 2. Many terms in Bernoullis equation can be neglected. 3. The liquid metal is reached to gate. 4. No air/gas is present in the liquid metal. 5. No solidication occurs during the lling. 6. The main resistance to the metal ow is in the runner. 7. A linear relationship between the pressure, P1 and ow rate (squared), Q2 . According to the last assumption, the liquid metal pressure at the plunger tip, P1 , can be written as P1 = Pmax 1 Where: P1 Q max Qmax the pressure at the plunger tip the ow rate maximum pressure which can be attained by the die casting machine in the shot sleeve maximum ow rate which can be attained in the shot sleeve Q Qmax
2

(7.1)

The Pmax and Qmax values to be determined for every set of the die casting machine and the shot sleeve. The Pmax value can be calculated using a static force balance. The determination of Qmax value is done by measuring the velocity of the plunger when the shot sleeve is empty. The maximum velocity combined with the shot sleeve crosssectional area yield the maximum ow rate, Qi = A Ui (7.2)

where i represent any possible subscription e.g. i = max Thus, the rst line can be drawn on pQ2 diagram as it shown by the line denoted as 1 in Figure ??. The line starts from a higher pressure (Pmax ) to a maximum ow rate (squared). A new combination of the same die casting machine and a dierent plunger diameter creates a dierent line. A smaller plunger diameter has a larger maximum pressure (Pmax ) and dierent maximum ow rate as shown by the line denoted as 2. The maximum ow rate is a function of the maximum plunger velocity and the plunger diameter (area). The plunger area is a obvious function of the plunger diameter, A = D2 /4. However, the maximum plunger velocity is a farmore complex function. The force that can be extracted from a die casting machine is essentially the same for dierent plunger diameters. The change in the resistance as results of changing the

82

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

plunger (diameter) depends on the conditions of the plunger. The dry friction will be same what dierent due to change plunger weight, even if the plunger conditions where the same. Yet, some researchers claim that plunger velocity is almost invariant in regard to the plunger diameter4 . Nevertheless, this piece of information has no bearing on the derivation in this model or reformed one, since we do not use it. Example 7.1: Prove that the maximum ow rate, Qmax is reduced and that Qmax 1/DP 2 (see Figure ??). if Umax is a constant

Pmax

Pmax

1 DP
2

Q max DP
2

Q=

2 A 3 CD

Pmax

Pmax Q max D1

Qmax

Fig b. Fig a.
The common pQ2 version.

Pmax and Qmax ass a function of model.

the plunger diameter according to common

Fig. -7.3. The left graph depicts the common version. The right graph depicts Pmax and Qmax as a function of the plunger diameter according to common model.

pQ2

A simplied force balance on the rode yields (see more details in section 7.11 page 97) Pmax = PB DB D1
2

PB DB 2 D1 2

(7.3)

where subscript B denotes the actuator. What is the pressure at the plunger tip when the pressure at the actuator is 10 [bars] with diameter of 0.1[m] and with a plunger diameter, D1 , of 0.05[m]? Substituting the data into equation (7.3) yields P1 = 10 0.1 0.05
2

= 4.0[M P a]

4 More research is need on this aspect. ++ read the comment made by referee II to the paper on pQ2 on page 192.

7.3. THE VALIDITY OF THE COMMON DIAGRAM In the common pQ2 diagram CD is dened as CD = 1 = constant 1 + KF

83

(7.4)

Note, therefore KF is also dened as a constant for every metal5 . Utilizing Bernoullis equation6 . U3 = CD 2P1 (7.5)

The ow rate at the gate can be expressed as Q3 = A3 CD 2P1 (7.6)

The ow rate in dierent locations is a function of the temperature. However, Eckert demonstrated that the heat transfer is insignicant in the duration of the lling of the cavity, and therefore the temperature of the liquid metal can be assumed almost constant during the lling period (which in most cases is much less 100 milliseconds). As such, the solidication is insignicant (the liquid metal density changes less than 0.1% in the runner); therefore, the volumetric ow rate can be assumed constant:
7

to make question about mass balance

Q1 = Q2 = Q3 = Q

(7.7)

Hence, we have two equations (7.1) and (7.6) with two unknowns (Q and P1 ) for which the solution is P1 = 1 Pmax 2CD 2 Pmax A3 2 Qmax 2 (7.8)
insert a discussion in regards to the trends

7.3 The validity of the common diagram

In the construction of the common model, two main assumptions were made: one CD is a constant which depends only on the liquid metal material, and two) many terms in the energy equation (Bernoullis equation) can be neglected. Unfortunately, the examination of the validity of these assumptions was missing in all the previous studies. Here, the question when the common model valid or perhaps whether the common model valid at all is examined. Some argue that even if the model is wrong and do not stand on sound scientic principles, it still has a value if it produces reasonable trends. Therefore, this model should produce reasonable results and trends when varying any parameter in order to have any value. Part of the examination is done by varying parameters and checking to see what happen to trends.
5 The 6 for

insert the calculation with redU3 dP1 spect to and odA1 dA1

author would like to learn who cameout with this clever idea. more details see section 7.4 page 87. 7 read more about it in Chapter 3.

84

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

7.3.1

Is the Common Model Valid?

Is the mass balance really satised in the common model? Lets examine this point. Equation (7.7) states that the mass (volume, under constant density) balance is exist. A1 U1 = A3 U3 (7.9)

So, what is the condition on CD to satisfy this condition? Can CD be a constant as stated in assumption 1? To study this point let derive an expression for CD . Utilizing yields equation (7.5)

A1 U1 = A3 CD

2 P1

(7.10)

From the machine characteristic, equation (7.1), it can be shown that U1 = Umax Pmax P1 Pmax (7.11)
A3

Example 7.2: Derive equation 7.11. Start with machine characteristic equation (7.1)

Fig. -7.4. P as A3 to be relocated

Substituting equation (7.11) into equation (7.10) yield, A1 Umax Pmax P1 = A3 CD Pmax P1 (7.12)

It can be shown that equation (7.12) can be transformed into CD A1 Umax = A3 2 P1 Pmax
Pmax P1

(7.13)

Example 7.3: Find the relationship between CD and Ozer number that satisfy equation (7.13) According to the common model Umax , and Pmax are independent of the gate area, A3 . The term A3 other parameters).
P1 Pmax P1

is not a constant and is a function of A3 (possibility

7.3. THE VALIDITY OF THE COMMON DIAGRAM Example 7.4: nd the relationship between A3 Solution under construction
End Solution

85

P1 Pmax P1

and A3

P1
P

Q 1 DP
2

DP

1 DP P1 Q u3 D1

A1

Fig a.

P as A1 to be relocated

Fig b.

P1 , Q, and U3 as a function of plunger diameter , A1 .

Fig. -7.5. Pressure at the plunger tip, P1 , the ow rate, Q, and the gate velocity, U3 as a function of plunger diameter , A1 .

Example 7.5: A3 what other parameters that CD depend on which do not provide the possibility CD = constant? To maintain the mass balance CD must be a function at least of the gate area, A3 . Since the common pQ2 diagram assumes that CD is a constant it diametrically opposite the mass conservation principle. Moreover, in the common model, a major assumption is that the value of CD depends on the metal, therefore, the mass balance is probably never achieved in many cases. This violation demonstrates, once for all, that the common pQ2 diagram is erroneous. Solution under construction.
End Solution

Use the information from example ?? and check what happened to the ow rate at two location ( 1) gate 2) plunger tip) when discharge coecient is varied CD = 0.40.9

86

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

7.3.2

Are the Trends Reasonable?

Now second part, are the trends predicted by the common model are presumable (correct)? To examine that, we vary the plunger diameter, (A1 or D1 ) and the gate area, A3 to see if any violation of the physics laws occurs as results. The comparison between the real trends and the common trends is discussed in the following section. Plunger area/diameter variation First, the eect of plunger diameter size variation is examined. In section 7.2 it was shown that Pmax 1/D1 2 . Equation (7.8) demonstrates that P1 increases with an increase of Pmax . It also demonstrates that the value of P never can exceed P1 Pmax =
max

Qmax 2 CD A3

(7.14)

The value Pmax can attained is an innite value (according to the common model) therefore P1 is innite as well. The gate velocity, U3 , increases as the plunger diameter decreases as shown in Figure ??. Armed with this knowledge now, several cases can be examined if the trends are realistic. Gate area variation Energy conservation (power supply machine characteristic) Lets assume P that mass conservation is fullled, and, hence the plunger velocity can approach Q 2A C innity, U1 when D1 0 (under constant Qmax ). The hydraulic piston also has to move with the same velocity, U1 . Yet, according to the machine charP acteristic the driving pressure, approaches zero (PB 1 PB 2 ) 0. Therefore, the energy supply to the system is approachFig. -7.6. P1 as a function of Pmax . ing zero. Yet, energy obtained from the system is innite since jet is inject in innite velocity and nite ow rate. This cannot exist in our world or perhaps one can proof the opposite.
1 3 2 max 2 2 D max

to make a question in regards to dissipation and velocity

Energy conservation (power supply) Assuming that the mass balance requirement is obtained, the pressure at plunger tip, P1 and gate velocity, U3 , increase (and can reach innity,(when P1 then U3 ) when the plunger diameter is reduced. Therefore, the energy supply to the system has to be innity (assuming a constant energy dissipation, actually the dissipation increases with plunger diameter in most ranges) However, the energy supply to the system (c.v. only the liquid metal) system would be PB 1 AB 1 U1 (nite amount) and the energy the system provide plus would be innity (innite gate velocity) plus dissipation.

7.4. THE REFORMED PQ2 DIAGRAM

87

Energy conservation (dissipation problem) A dierent way to look at this situation is check what happen to physical quantities. For example, the resistance to the liquid metal ow increases when the gate velocity velocity is increased. As smaller the plunger diameter the larger the gate velocity and the larger the resistance. However, the energy supply to the system has a maximum ability. Hence, this trend from this respect is unrealistic. Mass conservation (strike) According to the common model, the gate velocity decreases when the plunger diameter increases. Conversely, the gate velocity increases when the plunger diameter decreases8 . According to equation (7.2) the liquid metal ow rate at the gate increases as well. However, according to the common pQ2 diagram, the plunger can move only in a nite velocity lets say in the extreme case Umax 9 . Therefore, the ow rate at the plunger tip decreases. Clearly, these diametrically opposing trends cannot coexist. Either the common pQ2 diagram wrong or the mass balance concept is wrong, take your pick. Mass conservation (hydraulic pump): The mass balance also has to exist in hydraulic pump (obviously). If the plunger velocity have to be innite to maintain mass balance in the metal side, the mass ow rate at the hydraulic side of the rode also have to be innite. However, the pump has maximum capacity for ow rate. Hence, mass balance can be obtained.

to put table with dierent trends as a function of A3 and may be with a gure.

7.3.3

Variations of the Gate area, A3

under construction

7.4 The reformed pQ2 diagram

The method based on the liquid metal properties is with disagreement with commonly agreed on in uid mechanics [27, pp. 235-299]. It is commonly agreed that CD is a function of Reynolds number and the geometry of the runner design. The author suggested adopting an approach where the CD is calculated by utilizing data of ow resistance of various parts (segments) of the runner. The available data in the literature demonstrates that a typical value of CD can change as much as 100% or more just by changing the gate area (like valve opening). Thus, the assumption of a constant CD , which is used in common pQ2 calculations10 , is not valid. Here a systematic derivation of the pQ2 diagram is given. The approach adapted in this book is that everything (if possible) should be presented in dimensionless form.
again Figure ?? is the velocity attained when the shot sleeve is empty 10 or as it is suggested by the referee II
9 this 8 check

88

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

7.4.1

The reform model

Q = 1P

Equation (7.1) can be transformed into dimensionless from as (7.15)

Where: P Q reduced pressure, P1 /Pmax reduced ow rate, Q1 /Qmax

Eckert also demonstrated that the gravity eects are negligible11 . Assuming steady state12 and utilizing Bernoullis equation between point (1) on plunger tip and point (3) at the gate area (see Figure 7.1) yields P1 U1 2 P3 U3 2 + = + + h1,3 2 2 where: U h1,3 subscript 1 2 3 velocity of the liquid metal the liquid metal density energy loss between plunger tip and gate exit plunger tip entrance to runner system gate (7.16)

It has been shown that the pressure in the cavity can be assumed to be about atmospheric (for air venting or vacuum venting) providing vents are properly designed BarMeir at el 13 . This assumption is not valid when the vents are poorly designed. When they are poorly designed, the ratio of the vent area to critical vent area determines the build up pressure, P3 , which can be calculated as it is done in BarMeir et al However, this is not a desirable situation since a considerable gas/air porosity is created and should be avoided. It also has been shown that the chemical reactions do not play a signicant role during the lling of the cavity and can be neglected [5]. The resistance in the mold to liquid metal ow depends on the geometry of the part to be produced. If this resistance is signicant, it has to be taken into account calculating the total resistance in the runner. In many geometries, the liquid metal path in the mold is short, then the resistance is insignicant compared to the resistance in the runner and can be ignored. Hence, the pressure at the gate, P3 , can be neglected. Thus, equation (7.16) is reduced to U1 2 U3 2 P1 + = + h1,3 2 2
11 see

(7.17)

for more details chapter 3 in the section 7.4.4 on the transition period of the pQ2 13 Read a more detailed discussion in Chapter 9
12 read

7.4. THE REFORMED PQ2 DIAGRAM The energy loss, h1,3 , can be expressed in terms of the gate velocity as h1,3 = KF
2 U3 2

89

(7.18)

where KF is the resistance coecient, representing a specic runner design and specic gate area. Combining equations (7.7), (7.17) and (7.18) and rearranging yields U3 = CD where CD = f (A3 , A1 ) = 1 1
A3 A1 2

2P1

(7.19)

(7.20) + KF

Converting equation (7.19) into a dimensionless form yields Q= When the Ozer Number is dened as Oz =
CD 2 Pmax 2 Qmax A3

2Oz P

(7.21)

A3 Qmax

CD 2

Pmax

(7.22)

The signicance of the Oz number is that this is the ratio of the eective maximum energy of the hydrostatic pressure to the maximum kinetic energy. Note that the Ozer number is not a parameter that can be calculated a priori since the CD is varying with the operation point. 14 For practical reasons the gate area, A3 cannot be extremely large. On the other hand, the gate area can be relatively small A3 0 in this case Ozer number A3 A3 n where is a number larger then 2 (n > 2). Solving equations (7.21) with (7.15) for P , and taking only the possible physical solution, yields P = 1 1 + 2 Oz (7.23)

which is the dimensionless form of equation (7.8).

14 It

should be margin-note and so please ignore this footnote. how Ozer number behaves as a function of the gate area? Oz = A3 Pmax Qmax 2 1 A3 2 + K F A
1

90

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

7.4.2

Examining the solution

This solution provide a powerful tool to examine various parameters and their eects on the design. The important factors that every engineer has to nd from these calculations are: gate area, plunger diameters, the machine size, and machine performance etc15 . These issues are explored further in the following sections. The gate area eects Gate area aects the reduced pressure, P , in two ways: via the Ozer number which include two terms: one, (A3 /Qmax ) and, two, discharge coecient CD . The discharge coecient, CD is also aected by the gate area aects through two dierent terms in the denition (equation 7.20), one, (A3 /A1 )2 and two by KF . Qmax eect is almost invariant with respect to the gate area up to small gate area sizes16 . Hence this part is somewhat clear and no discussion is need. (A3 /A1 )2 eects Lets look at the denition of CD equation (7.20). For illustration purposes let assume that KF is not a function of gate area, KF (A3 ) = constant. A small perturbation of the gate area results in Taylor series, CD = CD (A3 + A3 ) CD (A3 ) 1 A3 A3 = + A3 2 A3 2 1 A 1 2 + KF A1 2 1 A 1 2 + K F
3 A3 2 2 A 2 A1 4 1 A3 2 +KF
1

perhaps to put discussion pending on the readers response.

(7.24)
3 2

1 A 2 A1 2 1 A3 2 +KF
1

A3 2 + O(A3 )
3

A3 2 A1 2

+ KF

In this case equation (7.8) still hold but CD has to be reevaluated. repeat the example ?? with KF = 3.3 First calculate the discharge coecient, CD for various gate area starting from 2.4 106 [m2 ] to 3 104 [m2 ] using equation 7.20. This example demonstrate the very limited importance of the inclusion of the term (A3 /A1 )2 into the calculations. KF eects The change in the gate area increases the resistance to the ow via several contributing factors which include: the change in the ow cross section, change in the direction of the ow, frictional loss due to ow through the gate length, and the loss due to the abrupt expansion after the gate. The loss due to the abrupt expansion is a major contributor and its value changes during the lling process. The liquid metal
15 The machine size also limited by a second parameter known as the clamping forces to be discussed in Chapter 11 16 This is reasonable speculation about this point. More study is well come

7.4. THE REFORMED PQ2 DIAGRAM

91

enters the mold cavity in the initial stage as a free jet and sometime later it turns into an immersed jet which happens in many geometries within 5%-20% of the lling. The change in the ow pattern is believed to be gradual and is a function of the mold geometry. A geometry with many changes in the direction of the ow and/or a narrow mold (relatively thin walls) will have the change to immersed jet earlier. Many sources provide information on KF for various parts of the designs of the runner and gate. Utilizing this information produces the gate velocity as a function of the given geometry. To study further this point consider a case where KF is a simple function of the gate area. When A3 is very large then the eect on KF are relatively small. Conversely, KF when A3 0 the resistance, KF . The simplest function, shown in Figure 7.7, that represent such behavior is KF = C 1 + C2 A3
K0

(7.25)

A3 C1 and C2 are constants and can be calculated (approximated) for a specic geometry. The value of C1 determine the Fig. -7.7. KF as a function of gate area, A3 value of the resistance where A3 eect is minimum and C2 determine the range (point) where A3 plays a signicant eect. In practical, it is found that C2 is in the range where gate area are desired and therefore program such as DiePerfect are important to calculated the actual resistance.

Example 7.1: Under construction create a question with respect to the function 7.25 Solution Under construction The combined eects Consequently, a very small area ratio results in a very large resistance, and when A3 0 therefore the resistance resulting in a zero gate A1 velocity (like a closed valve). Conversely, for a large area ratio, the resistance is insensitive to variations of the gate area and the velocity is reduced with increase the gate area. Therefore a maximum gate velocity must exist, and can be found by dU3 =0 dA3 (7.26)

which can be solved numerically. The solution of equation (7.26) requires full information on the die casting machine. A general complicated runner design conguration can be converted into a straight conduit with trapezoidal crosssection, provided that it was proportionally designed

92

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

leave it for now, better presentation needed

to create equal gate velocity for dierent gate locations17 . The trapezoidal shape is commonly used because of the simplicity, thermal, and for cost reasons. To illustrate only the eects of the gate area change two examples are presented: one, a constant pressure is applied to the runner, two, a constant power is applied to the runner. The resistance to the ow in the shot sleeve is small compared to resistance in the runner, hence, the resistance in the shot sleeve can be neglected. The die casting machine performance characteristics are isolated, and the gate area eects on the the gate velocity can be examined. Typical dimensions of the design are presented in Figure ??. The short conduit of 0.25[m] represents an excellent runner design and the longest conduit of 1.50[m] represents a very poor design. The calculations were carried for aluminum alloy with a density of 2385[kg/m3 ] and a kinematic viscosity of 0.544 106 [m2 /sec] and runner surface roughness of 0.01 [mm]. For the constant pressure case the liquid metal pressure at the runner entrance is assumed to be 1.2[MPa] and for the constant power case the power loss is [1Kw]. lling time that tmax t = V Qmax 2Oz P (7.27)

The gate velocity is exhibited as a function of the ratio of the gate area to the conduit area as shown in Figure ?? for a constant pressure and in Figure ?? for a constant power. General conclusions from example 7.7 For the constant pressure case the common18 assumption yields a constant velocity even for a zero gate area. The solid line in Figure ?? represents the gate velocity calculated based on the common assumption of constant CD while the other lines are based on calculations which take into account the runner geometry and the Re number. The results for constant CD represent averaged of the other results. The calculations of the velocity based on a constant CD value are unrealistic. It overestimates the velocity for large gate area and underestimates for the area ratio below 80% for the short runner and 35% for the long runner. Figure ?? exhibits that there is a clear maximum gate velocity which depends on the runner design (here represented by the conduit length). The maximum indicates that the preferred situation is to be on the right hand side branch because of shorter lling time. The gate velocity is doubled for the excellent design compared with the gate velocity obtained from the poor design. This indicates that the runner design is more important than the specic characteristic of the die casting machine performance. Operating the die casting machine in the right hand side results in smaller requirements on the die casting machine because of a smaller lling time, and therefore will require a smaller die casting machine. For the constant power case, the gate velocity as a function of the area ratio is shown in Figure ??. The common assumption of constant CD yields the gate velocity
17 read 18 As

about poor design eect on pQ2 diagram it is written in NADCAs books

7.4. THE REFORMED PQ2 DIAGRAM

93

U3 A1 /A3 shown by the solid line. Again, the common assumption produces unrealistic results, with the gate velocity approaching innity as the area ratio approaches zero. Obviously, the results with a constant CD over estimates the gate velocity for large area ratios and underestimates it for small area ratios. The other lines describe the calculated gate velocity based on the runner geometry. As before, a clear maximum can also be observed. For large area ratios, the gate velocity with an excellent design is almost doubled compared to the values obtained with a poor design. However, when the area ratio approaches zero, the gate velocity is insensitive to the runner length and attains a maximum value at almost the same point. In conclusion, this part has been shown that the use of the common pQ2 diagram with the assumption of a constant CD may lead to very serious errors. Using the pQ2 diagram, the engineer has to take into account the eects of the variation of the gate area on the discharge coecient, CD , value. The two examples given inhere do not represent the characteristics of the die casting machine. However, more detailed calculations shows that the constant pressure is in control when the plunger is small compared to the other machine dimensions and when the runner system is very poorly designed. Otherwise, the combination of the pressure and power limitations results in the characteristics of the die casting machine which has to be solved.

U3

pressure
efficiency
common model

realistic velocity

power
A3

Fig a.

U3 as a function of gate area, A3

power

Fig b.

General characteristic of a pump.

Fig. -7.8. Velocity, U3 as a function of the gate area, A3 and the general characterstic of a pump

94

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

The die casting machine characteristic eects There are two type of operation of the die casting machine, one) the die machine is operated directly by hydraulic pump (mostly on the old machines). two) utilizing the non continuous demand for the power, the power is stored in a container and released when need (mostly on the newer machines). The container is normally a large tank contain nitrogen and hydraulic liquid19 . The eects of the tank size and gas/liquid ratio on the pressure and ow rate can easily be derived.

Meta

The power supply from the tank with can consider almost as a constant pressure but the line to actuator is with variable resistance which is a function of the liquid velocity. The resistance can be consider, for a certain range, as a linear function of the velocity square, UB 2 . Hence, the famous a assumption of the common die casting machine p Q2 .

Meta End
The characteristic of the various pumps have been studied extensively in the past [15]. The die casting machine is a pump with some improvements which are patented by dierent manufactures. The new congurations, such as double pushing cylinders, change somewhat the characteristics of the die casting machines. First let discuss some general characteristic of a pump (issues like impeller, speed are out of the scope of this discussion). A pump is mechanical devise that transfers and electrical power (mostly) into hydraulic power. A typical characteristic of a pump are described in Figure ??. Two similar pumps can be connect in two way series and parallel. The serious connection increase mostly the pressure as shown in Figure ??. The series connection if normalized is very close to the original pump. However, the parallel connection when normalized show a better performance. To study the eects of the die casting machine performances, the following Fig. -7.9. Various die casting machine perforfunctions are examined (see Figure 7.9): mances
1 0.8 0.6

0.4

P= 1 Q P= 1 Q P= 1 Q

0.2

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

machine : linearQ = machine : sqrtQ = machine : sq Q =

4

1P 1P 1P

(7.28) (7.29) (7.30)

19 This similar to operation of water system in a ship, many of the characteristics are the same. Furthermore, the same dierential equations are governing the situation. The typical questions such as the necessarily container size and the ratio of gas to hydraulic liquid were part of my study in high school (probably the simplied version of the real case). If demand to this material raised, I will insert it here in the future.

7.4. THE REFORMED PQ2 DIAGRAM

95

The functions (??), (??) and (??) represent a die casting machine with a poor performance, the common performance, and a die casting machine with an excellent performance, respectively. Combining equation (7.21) with (7.28) yields 1P = 1P =
4

2Oz P 2Oz P 2Oz P

(7.31) (7.32) (7.33)

1P =

rearranging equation (7.31) yields P 2 2(1 + Oz)P + 1 = 1 P (1 + 2Oz) = 4Oz P 2 + P 1 = 0 0 0 (7.34) (7.35) (7.36)

Solving equations (7.34) for P , and taking only the possible physical solution, yields P = P = P = 1 + Oz (2 + Oz) Oz 1 1 + 2 Oz 1 + 16 Oz 2 1 8Oz 2 (7.37) (7.38) (7.39)

The reduced pressure, P , is plotted as a function of the Oz number for the three die casting machine performances as shown in Figure 7.10. Figure 7.10 demonstrates that P monotonically decreases with an increase in the Oz number for all the machine performances. All the three results convert to the same line which is a plateau after Oz = 20. For large Oz numbers the reduced pressure, P , can be considered to be constant P 0.025. The gate velocity, in this case, is U3 0.22CD Pmax (7.40)

The Ozer number strongly depends on the discharge coecient, CD , and Pmax . The value of Qmax is relatively insensitive to the size of the die casting machine. Thus, this equation is applicable to a well designed runner (large CD ) and/or a large die casting machine (large Pmax ).

96
1.00 .. 0.90 0.80 0.70 0.60

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

. . . . . .. . 4 . . . 0.45 . P = 1Q . .. . . . . .. . . .. 2 . . 0.40 .. P = 1Q . . .. . . .. . . .. . 0.35 . .. .......... . P = 1Q . .. . . .. . .. . . 0.30 .. . . .. . . .. . . . . .. . 0.25 . .. . . .. . . .. . .. . .. . 0.20 . . . .. . .. .. .. .. .. .. 0.15 .. .. ... .. ... ... .. .... .. 0.10 .. .. 1.00 2.00 3.00 4.00 .. .. .. ... .... ..... . . .. .. .. . . . .. .. . . . . .. .. .. .. .. . . . . . . . .. .. .. .. . . . . .. . . .. .. .. .. .. . . .. .. .. .. .. . . . . . . . . . . . . . .. .. .. .. .. .. .. .. .. .. . . . . . . . . . . .. .. .. .. . . .. .. . . . . .

P
0.50 0.40 0.30 0.20 0.10 0.00

0.00

3.00

6.00

9.00

12.00

15.00

18.00

21.00

24.00

27.00

30.00

Oz
Fig. -7.10. Reduced pressure, P , for various machine performances as a function of the Oz number.

The reduced pressure for a very small value of the Oz number equals to one, P 1 or Pmax = P1 , due to the large resistance in the runner (when the resistance in the runner approaches inn ity, KF , then P = 1). Hence, the gate velocity is determined by the approximation of

PB 2 D B PB 1

DR

atmospheric pressure P 1 D1

rode hydraulic piston plunger

Fig. -7.11. Schematic of the plunger and piston balance forces.

U3

CD

2Pmax

(7.41)

The dierence between the various machine performances is more considerable in the middle range of the Oz numbers. A better machine performance produces a higher reduced pressure, P . The preferred situation is when the Oz number is large and thus indicates that the machine performance is less important than the runner design parameters. This observation is further elucidated in view of Figures ?? and ??.

7.4. THE REFORMED PQ2 DIAGRAM Plunger area/diameter eects

97

The pressure at the plunger tip can be evaluated from a balance forces acts on the hydraulic piston and plunger as shown in Figure 7.11. The atmospheric pressure that acting on the left side of the plunger is neglected. Assuming a steady state and neglecting the friction, the forces balance on the rod yields DR 2 D1 2 DB 2 (PB 1 PB 2 ) + PB 2 = P1 4 4 4 In particular, in the stationary case the maximum pressure obtains DB 2 DR 2 D1 2 (PB 1 PB 2 )|max + PB 2 |max = P1 |max 4 4 4 (7.43) (7.42)

(why? perhaps to create a question for the students)

The equation (7.43) is reduced when the rode area is negligible; plus, notice that P1 |max = Pmax to read DB 2 D1 2 (PB 1 PB 2 )|max = Pmax 4 4 Rearranging equation (7.44) yields DB D1
20 2

(7.44)

Pmax = Pmax = (PB 1 PB 2 )|max (PB 1 PB 2 )|max

DB D1

(7.45)

The gate velocity relates to the liquid metal pressure at plunger tip according to the following equation combining equation (7.5) and (??) yields (PB 1 PB 2 )|max 1+
2 CD A3 Qmax 2 DB D1 2 2

U3 = CD

(7.46)

(PB 1 PB 2 )|max

DB D1

Under the assumption that the machine characteristic is P1 Q2 = P = 1 Q2 ,

20 Note that P | 1 max = [P1 ]max . The dierence is that P1 |max represents the maximum pressure of the liquid metal at plunger tip in the stationary case, where as [P1 ]max represents the value of the maximum pressure of the liquid metal at the plunger tip that can be achieved when hydraulic pressure within the piston is varied. The former represents only the die casting machine and the shot sleeve, while the latter represents the combination of the die casting machine (and shot sleeve) and the runner system. Equation (7.14) demonstrates that the value of [P1 ]max is independent of Pmax (for large values of Pmax ) under the assumptions in which this equation was attained (the common die casting machine performance, etc). This suggests that a smaller die casting machine can achieve the same job assuming average performance die casing machine.

98
1.0 ... .. .. .. .. 0.9 .. .. .. .. 0.8 .. .. .. .. 0.7 .. .. . .. .. 0.6 .. .. .. .. 0.5 .. .. .. .. .. 0.4 .. .. .. .. 0.3 .. .. .. .. 0.2 .. ... .. 0.1 0.0 0.0 1.0 2.0 3.0

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

............

...

...

....

. . . . . .. . .. .. .. .. . . . . .. .. . . .. .. .. .. .. .. .. .. .. . . . . . . . . . 4.0 5.0 6.0 7.0

.. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . . . . . . . . . . 8.0 9.0 10.0

Fig. -7.12. Reduced liquid metal pressure at the plunger tip and reduced gate velocity as a function of the reduced plunger diameter.

Meta

the solution for the intersection point is given by equation ? To study equation (7.46), lets dene = (PB 1 PB 2 )|max Qmax C D A3 D1 DB (7.47)

and the reduced gate velocity y= U3 A3 Qmax (7.48)

Using these denitions, equation (7.46) is converted to a simpler form: y= 1 2 + 1 (7.49)

With these denitions, and denoting = P1 2 C D A3 Qmax

2

= 2 Oz P

(7.50)

7.4. THE REFORMED PQ2 DIAGRAM

99

one can obtain from equation (??) that (make a question about how to do it?) = 2 1 +1 (7.51)

The coecients of P1 in equation (7.50) and D1 in equation (7.47) are assumed constant according to the common pQ2 diagram. Thus, the plot of y and as a function of represent the aect of the plunger diameter on the reduced gate velocity and reduced pressure. The gate velocity and the liquid metal pressure at plunger tip decreases with an increase in the plunger diameters, as shown in Figure 7.12 according to equations (7.49) and (7.51).

Meta End
A control volume as it is shown in Figure 7.13 is constructed to study the eect of the plunger diameter, (which includes the plunger with the rode, hydraulic piston, and shot sleeve, but which does not include the hydraulic liquid or the liquid metal jet). The control volume is stationary around the shot sleeve and is moving with the hydraulic piston. Applying the rst law of thermodynamics, when that the atmospheric pressure is assumed negligible and neglecting the dissipation energy, yields Uin Q + min hin + 2
2

why? should be included in the end.

= mout hout +

Uout 2

dm dt

e+
c.v.

Uc.v. 2

(7.52) + Wc.v.

In writing equation (7.52), it should be noticed that the only change in the control volume is in the shot sleeve. The heat transfer can be neglected, since the lling process is very rapid. There is no ow into the control volume (neglecting the air ow into the back side of the plunger and the Fig. -7.13. A general schematic of the control volume change of kinetic energy of the of the hydraulic piston with the plunger and part of the air, why?), and therefore the sec- liquid metal ond term on the right hand side can be omitted. Applying mass conservation on the control volume for the liquid metal yields
the moving part of the control volume UP liquid metal out atmospheric pressure air flow in PB 2 PB 1 friction is neglected

dm dt

= mout
c.v.

(7.53)

The boundary work on the control volume is done by the left hand side of the plunger and can be expressed by Wc.v. = (PB 1 PB 2 )AB U1 (7.54)

100

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

The mass ow rate out can be related to the gate velocity mout = A3 U3 Mass conservation on the liquid metal in the shot sleeve and the runner yields
2 A1 U1 = A3 = U1 2 = U3

(7.55)

A3 A1

(7.56)

Substituting equations (7.53-7.56) into equation (7.52) yields (PB 1 PB 2 )AB U3 A1 = A3 U3 (hout e) + Rearranging equation (7.57) yields (PB 1 PB 2 ) Solving for U3 yields
A 2 (PB 1 PB 2 ) A1B (hout e) 2 U3 2

A3 A1

(7.57)

AB U2 = (hout e) + 3 A1 2

A3 A1

(7.58)

U3 =

A3 A1

(7.59)

Or in term of the maximum values of the hydraulic piston 2

(PB 1 PB 2 )|max AB 1+2 Oz A1

U3 =

(hout e)
2

(7.60)

A3 A1

When the term (hout e) is neglected (Cp Cv for liquid metal) 2

(PB 1 PB 2 )|max AB 1+2 Oz A1 2 1 A3 A1

U3 =

(7.61)

Normalizing the gate velocity equation (7.61) yields y= U3 A3 = Qmax CD 2 [1 + 2 Oz] 1

A3 A1 2

(7.62)

7.4. THE REFORMED PQ2 DIAGRAM

101

The expression (7.62) is a very complicated function of A1 . It can be shown that when the plunger diameter approaches innity, D1 (or when A1 ) then the gate velocity approaches U3 0. Conversely, the gate velocity, U3 0, when the plunger diameter, D1 0. This occurs because mostly K and CD 0. Thus, there is at least one plunger diameter that creates maximum velocity (see gure 7.14). A more detailed study shows that depending on the physics in the situation, more than one local maximum can occur. With a small plunger diameter, the gate velocity approaches zero because CD approaches innity. For a large plunger diameter, the gate velocity approaches zero because the pressure dierence acting on the runner is approaching zero. The mathematical expression for the maximum gate velocity takes several pages, and therefore is not shown here. However, for practical purposes, the maximum velocity can easily (relatively) be calculated by using a computer program such as DiePerfect. Machine size eect
U3 common model

The question how large the die casting machine depends on how ecient it is used. To maximized the utilization of the die casting machine we must understand under what condition it happens. It is important to realize that the injection of the liquid metal into the cavity requires power. The power, we can extract from a ma- Fig. -7.14. The gate velocity, U3 as a function chine, depend on the plunger velocity and of the plunger area, A1 other parameters. We would like to design a process so that power extraction is maximized. Lets dened normalized machine size eect
realistic velocity
po ssib le m ax

A1

pwrm =

QP Pmax Qmax

QP

(7.63)

Every die casting machine has a characteristic curve on the pQ2 diagram as well. Assuming that the die casting machine has the common characteristic, P = 1 Q2 , the normalized power can be expressed pwrm = Q(1 Q2 ) = Q2 Q3 (7.64)

where pwrm is the machine power normalized by Pmax Qmax . The maximum power of this kind of machine is at 2/3 of the normalized ow rate, Q, as shown in Figure ??. It is recommended to design the process so the ow rate occurs at the vicinity of the maximum of the power. For a range of 1/3 of Q that is from 0.5Q to 0.83Q, the average power is 0.1388 Pmax Qmax , as shown in Figure ?? by the shadowed rectangular. One may notice that this value is above the capability of the die casting machine in two ranges of the ow rate. The reason that this number is used is because with some

102
0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 . 0.00 .. .. .. 0.00 ... . ..

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

..

. ... ... ... .. .. .. .. .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

...

.....

. .... . .

...

.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 0.80 0.90 1.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Q
Fig. -7.15. The reduced power of the die casting machine as a function of the normalized ow rate.

improvements of the the runner design the job can be performed on this machine, and there is no need to move the job to a larger machine21 . Precondition eect (wave formation)

Meta

discussion when Q1 = Q3

Meta End
7.4.3 Poor design eects

Meta

discussed the changes when dierent velocities are in dierent gates. Expanded on the sudden change to turbulent ow in one of the branches.

Meta End
7.4.4 Transient eects
that requirements on the clamping force is meet.

Under construction
21 Assuming

7.5. DESIGN PROCESS To put the discussion about the inertia of the system and compressibility. the magnitude analysis before intensication eects

103
insert only general remarks until the paper will submitted for publication insert the notes from the yellow folder

7.5 Design Process

Now with these pieces of information how one design the process/runner system. A design engineer in a local company have told me that he can draw very quickly the design for the mold and start doing the experiments until he gets the products running well. Well, the important part should not be how quickly you get it to try on your machine but rather how quickly you can produce a good quality product and how cheap ( little scrap as possible and smaller die casting machine). Money is the most important factor in the production. This design process is longer than just drawing the runner and it requires some work. However, getting the production going is much more faster in most cases and cheaper (less design and undesign scrap and less experiments/starting cost). Hence, for given die geometry, four conditions (actually there are more) need to satised U3 =0 A1 U3 =0 A3 (7.65)

(7.66)

the clamping force, and satisfy the power requirements. For these criteria the designer has to check the runner design to see if gate velocity are around the recommended range. A possible answer has to come from nancial considerations, since we are in the business of die casting to make money. Hence, the optimum diameter is the one which will cost the least (the minimum cost). How, then, does the plunger size determine cost? It has been shown that plunger diameter has a value where maximum gate velocity is created. A very large diameter requires a very large die casting machine (due to physical size and the weight of the plunger). So, one has to chose as rst approximation the largest plunger on a smallest die casting machine. Another factor has to be taken into consideration is the scrap created in the shot sleeve. Obviously, the liquid metal in the sleeve has to be the last place to solidify. This requires the biscuit to be of at least the same thickness as the runner. Trunner = Tbiscuit Therefore, the scrap volume should be D1 2 D1 2 Tbiscuit = Trunner 4 4 (7.68) (7.67)

General relationship between runner hydraulic diameter and plunger diameter.

104

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

When the scrap in the shot sleeve becomes signicant, compared to scrap of the runner D1 2 Trunner = LTrunner 4 Thus, the plunger diameter has to be in the range of D1 =
To discussed that the plunger diameter should not be use as varying the plunger diameter to determine the gate velocity

(7.69)

4 L

(7.70)

7.6 The Intensication Consideration

Intensication is a process in which pressure is increased making the liquid metal ows during the solidication process to ensure compensation for the solidication shrinkage of the liquid metal (up to 20%). The intensication is applied by two methods: one) by applying additional pump, two) by increasing the area of the actuator (the multiplier method, or the prell method)22 . . The rst method does not increase the intensication force to Pmax by much. However, the second method, commonly used today in the industry, can increase considerably the ratio.

put schematic gure of how it is done from the patent by die casting companies

Meta
why? to put discussion

Analysis of the forces demonstrates that as rst approximation the plunger diameter does not contribute any additional force toward pushing the liquid metal.

Meta End
A very small plunger diameter creates faster solidication, and therefore the actual force is reduced. Conversely, a very large plunger diameter creates a very small pressure for driving the liquid metal.

discuss the the resistance as a function of the diameter

7.7 Summary
In this chapter it has been shown that the common diagram is not valid and produces unrealistic trends therefore has no value what so ever23 . The reformed pQ2 diagram was introduced. The mathematical theory/presentation based on established scientic principles was introduced. The eects of various important parameters was discussed. The method of designing the die casting process was discussed. The plunger diameter
22 A note for the manufactures, if you would like to have your system described here with its advantages, please drop me a line and I will discuss with you about the material that I need. I will not charge you any money. 23 Beside the historical value

7.8. QUESTIONS

105

has a value for which the gate velocity has a maximum. For D1 0 gate velocity, U3 0 when D1 the same happen U3 0. Thus, this maximum gate velocity determines whether an increase in the plunger diameter will result in an increase in the gate velocity or not. An alternative way has been proposed to determine the plunger diameter.

7.8 Questions

106

CHAPTER 7. PQ2 DIAGRAM CALCULATIONS

Garber concluded that his model was not able to predict an acceptable value for critical velocity for ll percentages lower than 50% . . . Brevick, Ohio

CHAPTER 8 Critical Slow Plunger Velocity

8.1 Introduction
This Chapter deals with the rst stage of the injection in a cold chamber machine in which the desire (mostly) is to expel maximum air/gas from the shot sleeve. Porosity is a major production problem in which air/gas porosity constitutes a large portion. Minimization of Air Entrainment in the Shot Sleeve (AESS) is a prerequisite for reducing air/gas porosity. This can be achieved by moving the plunger at a specic speed also known as the critical slow plunger velocity. It happens that this issue is related to the hydraulic jump, which was discussed in the previous Chapters 5 (accidentally? you thought!). The common model, also known as Garbers model, with its extensions made by Brevick1 , Miller2 , and EKKs model are presented rst here. The basic fundamental errors of these models are presented. Later, the reformed and simple model is described. It followed by the transient and poor design eects3 . Afterwords, as usual questions are given at the end of the chapter.

8.2 The common models

In this section the common models are described. Since the popular model also known as Garbers model never work (even by its own creator)4 , several other models have appeared. These models are described here to have a clearer picture of what
1 Industrial and Systems Engineering (ISE) Graduate Studies Chair, ISE department at The Ohio State University 2 The chair of ISE Dept. at OSU 3 It be added in the next addition 4 I wonder if Garber and later Brevick have ever considered that their the models were simply totally false.

107

108

CHAPTER 8. CRITICAL SLOW PLUNGER VELOCITY

was in the pre BarMeirs model. First, a description of Garbers model is given later Brevicks two models along with Millers model5 are described briey. Lastly, the EKKs numerical model is described.

8.2.1

Garbers model
1

The description in this section is based on one of the most cited paper in the die castv v ing research [17]. Garbers model deals h v only with a plug ow in a circular cross P h P section. In this section, we improve the model to include any geometry cross section with any velocity prole6 . Consider a duct (any cross section) Fig. -8.1. A schematic of wave formation in with a liquid at level h2 and a plunger mov- stationary coordinates ing from the left to the right, as shown in Figure 8.1. Assuming a quasi steady ow is established after a very short period of time, a unique height, h1 , and a unique wave velocity, Vw , for a given constant plunger velocity, Vp are created. The liquid in the substrate ahead of the wave is still, its height, h2 , is determined by the initial ll. Once the height, h1 , exceeds the height of the shot sleeve, H, there will be splashing. The splashing occurs because no equilibrium can be achieved (see Figure 8.2a). For h1 smaller than H, a reecting wave from the opposite wall appears resulting in an enhanced air entrainment (see Figure 8.2b). Thus, the preferred situation is when h1 = H (in circular shape H = 2R) in which case no splashing or a reecting wave result.
p w
2 H

L(t)

initial height

Fig a.

A schematic of

Fig b.

A schematic of

built wave.

reecting wave.

Fig. -8.2. The left graph depicts the common pQ2 version. The right graph depicts Pmax and Qmax as a function of the plunger diameter according to common model.

It is easier to model the wave with coordinates that move at the wave velocity, as shown in Figure 8.3. With the moving coordinate, the wave is stationary, the plunger moves back at a velocity (Vw Vp ), and the liquid moves from the right to the left. Dashed line shows the stationary control volume.
model was developed at Ohio State University by Millers Group in the early 1990s. addition to the original Garbers paper is derived here. I assumed that in this case, some mathematics will not hurt the presentation.
6 This 5 This

8.2. THE COMMON MODELS Mass conservation of the liquid in the control volume reads: Vw dA =
A2 A1
1

109

(vw vp) h1 v1 P1 L(t)

v=0 vw

(v1 Vp )dA (8.1)

P2 h2

where v1 is the local velocity. Under quasi-steady conditions, the corresponding Fig. -8.3. A schematic of the wave with moving average velocity equals the plunger veloc- coordinates ity: 1 A1 v1 dA = v1 = Vp
A1

(8.2)

What is justication for equation 8.2? Assuming that heat transfer can be neglected because of the short process duration7 . Therefore, the liquid metal density (which is a function of temperature) can be assumed to be constant. Under the above assumptions, equation (8.1) can be simplied to
hi

build a question about what happens if the temperature changes by a few degrees. How much will it aect equation 8.2 and other parameters?

Vw A(h2 ) = (Vw Vp )A(h1 )

A(hi ) =
0

dA

(8.3)

Where i in this case can take the value of 1 or 2. Thus, Vw = f (h12 ) (Vw Vp ) (8.4)

where f (h12 ) = A(h1 ) is a dimensionless function. Equation (8.4) can be transformed A(h2 ) into a dimensionless form: v ( 1) v f (h12 ) = v = f (h12 ) 1 f (h12 ) = (8.5) (8.6)

where v = Vw . Show that A(h1 ) = 2R2 for h1 = 2R Assuming energy is conserved Vp (the Garbers model assumption), and under conditions of negligible heat transfer, the energy conservation equation for the liquid in the control volume (see Figure 8.3) reads: PB E (Vw Vp )2 + (Vw Vp )dA = 2 P2 Vw 2 + Vw dA 2 (8.7)

A1

A2

where =
7 see

1 A1 (Vw Vp )3

(Vw v1 )3 dA =
A1

1 A1 ( 1)3 v

v
A1

v1 Vp

dA

(8.8)

Chapter 3 for a detailed discussion

110

CHAPTER 8. CRITICAL SLOW PLUNGER VELOCITY

underconstruction The shape factor, E , is introduced to account for possible deviations of the velocity prole at section 1 from a pure plug ow. Note that in die casting, the ow is pushed by the plunger and can be considered as an inlet ow into a duct. The typical Re number is 105 , and for this value the entry length is greater than 50m, which is larger than any shot sleeve by at least two orders of magnitude. The pressure in the gas phase can be assumed to be constant. The hydrostatic y pressure in the liquid can be represent by Rc g [28], where Rc is the center of the y cross section area. For a constant liquid density equation (8.7) can be rewritten as: Rc1 g + E y (Vw Vp )2 Vw 2 (Vw Vp )A(h1 ) = Rc2 g + y Vw A(h2 ) 2 2 (8.9)

Garber (and later Brevick) put this equation plus several geometrical relationships as the solution. Here we continue to obtain an analytical solution. Dening a dimensionless parameter F r as Fr = Rg , Vp 2 (8.10)

Utilizing denition (8.10) and rearranging equation (8.9) yields 2F rE yc1 + E ( 1)2 = 2F rE yc2 + v 2 v Solving equation (8.11) for F rE the latter can be further rearranged to yield: F rE = 2(c1 yc2 ) y (1+E )f (h12 ) E f (h12 )1 (8.12) (8.11)

Given the substrate height, equation (8.12) can be evaluated for the F rE , and the corresponding plunger velocity ,Vp . which is dened by equation (8.10). This solution will be referred herein as the energy solution.

8.2.2

Brevicks Model

The square shot sleeve Since Garbers model never work Brevick and coworkers go on a shing expedition in the uid mechanics literature to nd equations to describe the wave. They found in Lambs book several equations relating the wave velocity to the wave height for a deep liquid (water)8 . Since these equations are for a two dimensional case, Brevick and coworkers built it for a squared shot sleeve. Here are the equations that they used. The instantaneous height dierence (h = h1 h2 ) is given as h = h2
8I

V p + 1 2 gh2

h2

(8.13)

have checked the reference and I still puzzled by the equations they found?

8.2. THE COMMON MODELS This equation (8.13), with little rearranging, obtained a new form Vp = 2 The wave velocity is given by Vw = gh2 3 1+ h 2 h2 gh2 h1 1 h2

111

(8.14)

(8.15)

Brevick introduces the optimal plunger acceleration concept. By plotting the height and position of each incremental wave with time, their model is able to predict the stability of the resulting wave front when the top of the front has traveled the length of the shot sleeve.9 . They then performed experiments on this miracle acceleration 10 .

8.2.3

Brevicks circular model

Probably, because it was clear to the authors that the previous model was only good for a square shot sleeve 11 . They say let reuse Garbers model for every short time steps and with dierent velocity (acceleration).

8.2.4

Millers square model

Miller and his student borrowed a two dimensional model under assumption of turbulent ow. They assumed that the ow is innite turbulence and therefor it is a plug ow12 . Since the solution was for 2D they naturally build model for a square shot sleeve13 . The mass balance for square shot sleeve Vw h2 = (Vw Vp )h1 Momentum balance on the same control volume yield PB (Vw Vp )2 P2 Vw 2 + (Vw Vp )h1 = + Vw h 2 2 2 and the solution of these two equations is F rmiller =
9 What

(8.16)

(8.17)

1 h1 2 h2

h1 +1 h2

(8.18)

an interesting idea?? Any physics? to say this is not good enough a fun idea, they also invented a new acceleration units cm/seccm. 11 It is not clear whether they know that this equations are not applicable even for a square shot sleeve. 12 How they comeout with this conclusion? 13 Why are these two groups from the same university and the same department not familiar with each others work.
10 As

112

CHAPTER 8. CRITICAL SLOW PLUNGER VELOCITY

8.3 The validity of the common models

8.3.1 Garbers model
Energy is known to dissipate in a hydraulic jump in which case the equal sign in equation (8.12) does not apply and the criterion for a nonsplashing operation would read F rE < F roptimal (8.19)

A considerable amount of research work has been carried out on this wave, which is known in the scientic literature as the hydraulic jump. The hydraulic jump phenomenon has been studied for the past 200 years. Unfortunately, Garber, ( and later other researchers in die casting such as Brevick and his students from Ohio State University [8], [31])14 , ignored the previous research. This is the real reason that their model never works. Show the relative error created by Garbers model when the substrate height h2 is the varying parameter.

8.3.2

Brevicks models

square model There are two basic mistakes in this model, rst) the basic equations are not applicable to the shot sleeve situation, second) the square geometry is not found in the industry. To illustrate why the equations Brevick chose are not valid, take the case where 1 > h1 /h2 > 4/9. For that case Vw is positive and yet the hydraulic jump opposite to reality (h1 < h2 ). Improved Garbers model Since Garbers model is scientic erroneous any derivative that is based on it no better than its foundation15 .

8.3.3

Millers model

The ow in the shot sleeve in not turbulent16 . The ow is a plug ow because entry length problem17 . Besides all this, the geometry of the shot sleeve is circular. This mistake is discussed in the comparison in the discussion section of this chapter.
with these major mistakes NADCA under the leadership of Gary Pribyl and Steve Udvardy continues to award Mr. Brevick with additional grands to continue this research until now, Why? 15 I wonder how much NADCA paid Brevick for this research? 16 Unless someone can explain and/or prove otherwise. 17 see Chapter 3.
14 Even

8.4. THE REFORMED MODEL

113

8.3.4

EKKs model (numerical model)

This model based on numerical simulations based on the following assumptions: 1) the ow is turbulent, 2) turbulence was assume to be isentropic homogeneous every where (k model), 3) unspecied boundary conditions at the free interface (how they solve it with this kind of condition?), and 4) unclear how they dealt with the corner point in which plunger perimeter in which smart way is required to deal with zero velocity of the sleeve and known velocity of plunger. Several other assumptions implicitly are in that work18 such as no heat transfer, a constant pressure in the sleeve etc. According to their calculation a jet exist somewhere in the ow eld. They use the k model for a eld with zero velocity! They claim that they found that the critical velocity to be the same as in Garbers model. The researchers have found same results regardless the model used, turbulent and laminar ow!! One can only wonder if the usage of k model (even for zero velocity eld) was enough to produce these erroneous results or perhaps the problem lays within the code itself19 .

8.4 The Reformed Model

The hydraulic jump appears in steadystate and unsteadystate situations. The hydraulic jump also appears when using dierent crosssections, such as square, circular, and trapezoidal shapes. The hydraulic jump can be moving or stationary. The wave in the shot sleeve is a moving hydraulic jump in a circular crosssection. For this analysis, it does not matter if the jump is moving or not. The most important thing to understand is that a large portion of the energy is lost and that this cannot be neglected. All the uid mechanics books20 show that Garbers formulation is not acceptable and a dierent approach has to be employed. Today, the solution is available to die casters in a form of a computer program DiePerfect.

8.4.1

The reformed model

In this section the momentum conservation principle is applied on the control volume in Figure 8.3. For large Re ( 105 ) the wall shear stress can be neglected compared to the inertial terms (the wave is assumed to have a negligible length). The momentum balance reads: P B + M (Vw Vp )2 dA =
A1 A2

P2 + Vw 2 dA

(8.20)

where M =
18 This 19 see

1 A1 (Vw Vp )2

(Vw v1 )2 dA =
A1

1 A1 ( 1)2 v

v
A1

v1 Vp

dA

(8.21)

paper is a good example of poor research related to a poor presentation and text processing. remark on page 44 20 in the last 100 years

114

CHAPTER 8. CRITICAL SLOW PLUNGER VELOCITY

Given the velocity prole v1 , the shape factor M can be obtained in terms of v . The expressions for M for laminar and turbulent velocity proles at section 1 easily can be calculated. Based on the assumptions used in the previous section, equation (8.20) reads: Rc1 g + M (Vw Vp ) y
2

A(h1 ) = Rc2 g + Vw 2 A(h2 ) y

(8.22)

Rearranging equation (8.22) into a dimensionless form yields: f (h12 ) yc1 F r + M ( 1) v Combining equations (8.5) and (8.23) yields
f (h12 ) f (h12 )1 2 2

= yc2 F r + v 2

(8.23)

F rM =

M f (h12 )

f (h12 ) f (h12 )1

[f (h12 )c1 yc2 ] y

(8.24)

where F rM is the F r number which evolves from the momentum conservation equation. Equation (8.24) is the analogue of equation (8.12) and will be referred herein as the BarMeirs solution.
16.00 14.40 12.80 11.20 9.60 .......... momentum energy Miller Karni Garber .. .. .. .. . .. .. .. . .. .. .. . .. .. .. .. .. .. .. . ... .. ... .. ... ...

Fr

8.00 6.40 4.80 3.20 1.60 0.00 0.00

. . . ... ... . . . . . . . . . .... ... ... ... ... ... ... ... ... ... 0.17 0.33 0.49 0.66

.. . .. . . . ... .. . .. . .. . 0.83

.... ..

....

....

0.99

1.16

1.32

1.49

1.65

h2
R

Fig. -8.4. The Froude number as a function of the relative height.

21 and the energy solution can be presented in a simple form. Moreover, these solutions can be applied to any cross section for the transition of the free surface ow to pressurized ow. The discussion here focuses on the circular cross section, since it is the only one used by diecasters. Solutions for other velocity proles, such as laminar
21 This

model was constructed with a cooperation of a another researcher.

8.4. THE REFORMED MODEL

115

ow (Poiseuille paraboloid), are discussed in the Appendix 22 . Note that the Froude number is based on the plunger velocity and not on the upstream velocity commonly used in the twodimensional hydraulic jump. The experimental data obtained by Garber , and Karni and the transition from the free surface ow to pressurized ow represented by equations (8.12) and (8.24) for a circular cross section are presented in Figure 8.4 for a plug ow. The Millers model (two dimensional) of the hydraulic jump is also presented in Figure 8.4. This Figure shows clearly that the BarMeirs solution is in agreement with Karnis experimental results. The agreement between Garbers experimental results and the BarMeirs solution, with the exception of one point (at h2 = R), is good. The experimental results obtained by Karni were taken when the critical velocity was obtained (liquid reached the pipe crown) while the experimental results from Garber are interpretation (kind of average) of subcritical velocities and supercritical velocities with the exception of the one point at h2 /R = 1.3 (which is very closed to the Bar Meirs solution). Hence, it is reasonable to assume that the accuracy of Karnis results is better than Garbers results. However, these data points have to be taken with some caution23 . Non of the experimental data sets were checked if a steady state was achieved and it is not clear how the measurements carried out. It is widely accepted that in the two dimensional hydraulic jump small and large eddies are created which are responsible for the large energy dissipation [19]. Therefore, energy conservation cannot be used to describe the hydraulic jump heights. The same can be said for the hydraulic jump in dierent geometries. Of course, the same has to be said for the circular cross section. Thus, the plunger velocity has to be greater than the one obtained by Garbers model, which can be observed in Figure 8.4. The Froude number for the Garbers model is larger than the Froude number obtained in the experimental results. Froude number inversely proportional to square of the plunger velocity, F r 1/Vp 2 and hence the velocity is smaller. The Garbers model therefore underestimates the plunger velocity.

8.4.2

Design process

To obtain the critical slow plunger velocity, one has to follow this procedure: 1. Calculate/estimate the weight of the liquid metal. 2. Calculate the volume of the liquid metal (make sure that you use the liquid phase property and not the solid phase). 3. Calculate the percentage of lling in the shot sleeve, 4. Find the F r number from Figure 8.4. 5. Use the F r number found to calculate the plunger velocity by using equation (8.10).
22 To 23 Results

height . r

appear in the next addition. of good experiments performed by serious researchers are welcome.

116

CHAPTER 8. CRITICAL SLOW PLUNGER VELOCITY

8.5 Summary
In this Chapter we analyzed the ow in the shot sleeve and developed a explicit expression to calculated the required plunger velocity. It has been shown that Garbers model is totally wrong and therefore Brevicks model is necessarily erroneous as well. The same can be said to all the other models discussed in this Chapter. The connection between the wave and the hydraulic jump has been explained. The method for calculating the critical slow plunger velocity has been provided.

8.6 Questions

CHAPTER 9 Venting System Design

The dierence between the two is expressed by changing standard atmospheric ambient conditions to those existing in the vacuum tank.

9.1 Introduction

Millers student, p. 102

Proper design of the venting system is one of the requirements for reducing air/gas porosity. Porosity due to entrainment of gases constitutes a large portion of the total porosity, especially when the cast walls are very thin (see Figure ??). The main causes of air/gas porosity are insucient vent area, lubricant evaporation (reaction processes), incorrect placement of the vents, and the mixing processes. The present chapter considers the inuence of the vent area (in atmospheric and vacuum venting) on the residual gas (in the die) at the end of the lling process. Atmospheric venting, the most widely used casting method, is one in which the vent is opened to the atmomaximum sphere and is referred herein as air ventshrinkage porosity porosity ing. Only in extreme cases are other solutions required, such as vacuum venting, Pore Free Technique (in zinc and aluminum casting) and squeeze casting. Vacwall thickness uum is applied to extract air/gas from the mold before it has the opportunity to mix Fig. -9.1. The relative shrinkage porosity as a with the liquid metal and it is call vacuum function of the casting thickness. venting. The Pore Free technique is a variation of the vacuum venting in which the oxygen is introduced into the cavity to replace the air and to react with the liquid metal, and therefore creates a vacuum [5]. Squeeze

117

118

CHAPTER 9. VENTING SYSTEM DESIGN

casting is a dierent approach in which the surface tension is increased to reduce the possible mixing processes (smaller Re number as well). The gases in the shot sleeve and cavity are made mostly of air and therefore the term air is used hereafter. These three solutions are cumbersome and create a far more expensive process. In this chapter, a qualitative discussion on when these solutions should be used and when they are not needed is presented. Obviously, the best ventilation is achieved when a relatively large vent area is designed. However, to minimize the secondary machining (such as trimming), to ensure freezing within the venting system, and to ensure breakage outside the cast mold, vents have to be very narrow. A typical size of vent thicknesses range from 12[mm]. These conicting requirements on the vent area suggest an optimum area. As usual the common approach is described the errors are presented and the reformed model is described.

9.2 The common models

9.2.1 Early (etc.) model
The rst model dealing with the extraction of air from the cavity was done by Sachs. In this model, Sachs developed a model for the gas ow from a die cavity based on the following assumptions: 1) the gas undergoes an isentropic process in the die cavity, 2) a quasi steady state exists, 3) the only resistance to the gas ow is at the entrance of the vent, 4) a maximum mass ow rate is present, and 5) the liquid metal has no surface tension, thus the metal pressure is equal to the gas pressure. Sachs also dierentiated between two cases: choked ow and unchoked ow (but this dierentiation did not come into play in his model). Assumption 3 requires that for choked ow the pressure ratio be about two between the cavity and vent exit. Almost the same model was repeat by several researchers1 . All these models, with the exception of Veinik , neglect the friction in the venting system. The vent design in a commercial system includes at least an exit, several ducts, and several abrupt expansions/contractions in which the resistance coecient ( 4f L see [29, page 163]) D can be evaluated to be larger than 3 and a typical value of 4f L is about 7 or more. D In this case, the pressure ratio for the choking condition is at least 3 and the pressure ratio reaches this value only after about 2/3 of the piston stroke is elapsed. It can be shown that when the ow is choked the pressure in the cavity does not remain constant as assumed in the models but increases exponentially.

9.2.2

Millers model

Miller and his student, in the early 90s, constructed a model to account for the friction in the venting system. They based their model on the following assumptions: 1. No heat transfer
1 Apparently,

no literature survey was required/available/needed at that time.

9.3. GENERAL DISCUSSION

119

2. Isothermal ow (constant temperature) in the entrance to vent (according to the authors in the presentation) 3. Fanno ow in the rest vent 4. Air/gas obeys the ideal gas model Miller and his student described the calculation procedures for the two case as choked and unchoked conditions. The calculations for the choked case are standard and can be found in any book about Fanno ow but with an interesting twist. The conditions in the mold and the sleeve are calculated according the ambient condition (see the smart quote of this Chapter)2 . The calculations about unchoked case are very interesting and will be discussed here in a little more details. The calculations procedure for the unchoked as the following:
Assume Min number (entrance Mach number to the vent) lower than Min for choked condition Calculate the corresponding star (choked conditions) the temperature ratio for the assume Min number Calculate the dierence between the calculated
4f L D 4f L D ,

the pressure ratio, and

4f L D .

and the actual

Use the dierence 4f L to calculate the double stars (theoretical exit) conditions D based on the ambient conditions. Calculated the conditions in the die based on the double star conditions.

Now the mass ow out is determined by mass conservation. Of course, these calculations are erroneous. In choked ow, the conditions are determined only and only by upsteam and never by the down steam3 . The calculations for unchoked ow are mathematical wrong. The assumption made in the rst step never was checked. And mathematically speaking, it is equivalent to just guessing solution. These errors are only fraction of the other other in that model which include among other the following: one) assumption of constant temperature in the die is wrong, two) poor assumption of the isothermal ow, three) poor measurements etc. On top of that was is the criterion for required vent area.

9.3 General Discussion

When a incompressible liquid such as water is pushed, the same amount propelled by the plunger will ow out of the system. However, air is a compressible substance and thus the above statement cannot be applied. The ow rate out depends on the resistance to the ow plus the piston velocity (piston area as well). There could be three situations
model results in negative temperature in the shot sleeve in typical range. otherwise, can it be? It is like assuming negative temperature in the die cavity during the injection. Is it realistic?
3 How 2 This

120

CHAPTER 9. VENTING SYSTEM DESIGN

1) the ow rate out is less than the volume pushed by the piston, 2) the ow rate out is more than the volume pushed by the piston, or 3) the ow rate out is equal to the volume pushed by the piston. The last case is called the critical design, and it is associated with the critical area. Air ows in the venting system can reach very large velocities up to about 350 [m/sec]. The air cannot exceed this velocity without going through a specially congured conduit (converging diverging conduit). This phenomena is known by the name of choked ow. This physical phenomenon is the key to understanding the venting design process. In air venting, the venting system has to be designed so that air velocity does not reach the speed of sound: in other words, the ow is not choked. In vacuum venting, the air velocity reaches the speed of sound almost instantaneously, and the design should be such that it ensures that the air pressure does not exceed the atmospheric pressure. Prior models for predicting the optimum vent area did not consider the resistance in the venting system (pressure ratio of less than 2). The vent design in a commercial system includes at least an exit, several ducts, and several abrupt expansions/contractions in which the resistance coecient, 4f L , is of the order of 37 or D more. Thus, the pressure ratio creating choked ow is at least 3. One of the dierences between vacuum venting and atmospheric venting occurs during the startup time. For vacuum venting, a choking condition is established almost instantaneously (it depends on the air volume in the venting duct), while in the atmospheric case the volume of the air has to be reduced to more than half (depending on the pressure ratio) before the choking condition develops - - and this can happen only when more than 2/3 or more of the piston stroke is elapsed. Moreover, the ow is not necessarily choked in atmospheric venting. Once the ow is choked, there is no dierence in calculating the ow between these two cases. It turns out that the mathematics in both cases are similar, and therefore both cases are presented in the present chapter. The role of the chemical reactions was shown to be insignicant. The dierence in the gas solubility (mostly hydrogen) in liquid and solid can be shown to be insignicant [1]. For example, the maximum hydrogen release during solidication of a kilogram of aluminum is about 7cm3 at atmospheric temperature and pressure. This is less than 3% of the volume needed to be displaced, and can be neglected. Some of the oxygen is depleted during the lling time [5]. The last two eects tend to cancel each other out, and the net eect is minimal. The numerical simulations produce unrealistic results and there is no other quantitative tools for nding the vent locations (the last place(s) to be lled) and this issue is still an open question today. There are, however, qualitative explanations and reasonable guesses that can push the accuracy of the last place (the liquid metal reaches) estimate to be within the last 10%30% of the lling process. This information increases the signicance of the understanding of what is the required vent area. Since most of the air has to be vented during the initial stages of the lling process, in which the vent locations do not play a role. Air venting is the cheapest method of operation, and it should be used unless acceptable results cannot be obtained using it. Acceptable results are dicult to obtain

9.4. THE ANALYSIS

121

1) when the resistance to the air ow in the mold is more signicant than the resistance in the venting system, and 2) when the mixing processes are augmented by the specic mold geometry. In these cases, the extraction of the air prior to the lling can reduce the air porosity which require the use of other techniques. An additional objective is to provide a tool to combine the actual vent area with the resistance (in the venting system) to the air ow; thus, eliminating the need for calculations of the gas ow in the vent in order to minimize the numerical calculations. Hu et al. and others have shown that the air pressure is practically uniform in the system. Hence, this analysis can also provide the average air pressure that should be used in numerical simulations.

9.4 The Analysis

The model is presented here with a minimal of mathematical details. However, emphasis is given to all the physical understanding of the phenomena. The interested reader can nd more detailed discussions in several other sources [4]. As before, the integral approach is employed. All the assumptions which are used in this model are stated so that they can be examined and discussed at the conclusion of the present chapter. Here is a list of the assumptions which are used in developing this model: 1. The main resistance to the air ow is assumed to be in the venting system. 2. The air ow in the cylinder is assumed onedimensional. 3. The air in the cylinder undergoes an isentropic process. 4. The air obeys the ideal gas model, P = RT . 5. The geometry of the venting system does not change during the lling process (i.e., the gap between the plates does not increase during the lling process). 6. The plunger moves at a constant velocity during the lling process, and it is determined by the pQ2 diagram calculations. 7. The volume of the venting system is negligible compared to the cylinder volume. 8. The venting system can be represented by one long, straight conduit. 9. The resistance to the liquid metal ow, 4f L , does not change during the lling D process (due to the change in the Re, or Mach numbers). 10. The ow in the venting system is an adiabatic ow (Fanno ow). 11. The resistance to the ow,
4f L D ,

is not aected by the change in the vent area.

With the above assumptions, the following model as shown in Figure ?? is proposed. A plunger pushes the liquid metal, and both of them (now called as the piston) propel the air through a long, straight conduit.

122

CHAPTER 9. VENTING SYSTEM DESIGN

The mass balance of the air in the cylinder yields dm + mout = 0. dt (9.1)
Fig. -9.2. A simplied model for the venting system.

This equation (9.1) is the only equation that needed to be solved. To solve it, the physical properties of the air need to be related to the geometry and the process. According to assumption 4, the air mass can be expressed as m=

PV RT

(9.2)

The volume of the cylinder under assumption 6 can be written as V (t) = V (0) 1 t tmax (9.3)

Thus, the rst term in equation (9.1) is represented by d dm = dt dt P V (0) 1 RT

t tmax

(9.4)

The lling process occurs within a very short period time [milliseconds], and therefore the heat transfer is insignicant 3. This kind of ow is referred to as Fanno ow4 . The instantaneous ow rate has to be expressed in terms of the resistance to the ow, 4f L , the pressure ratio, and the characteristics of Fanno ow [29]. Knowledge of D Fanno ow is required for expressing the second term in equation (9.1). The mass ow rate can be written as mout = P0 (0)AMmax where k1 2 f [Min (t)] = 1 + (Min (t)) 2
(k+1) 2(k1)

Min (t) Mmax

P0 (0) P0 (t)

k+1 2k

k f [Min (t)] RT0 (0)

(9.5)

(9.6)

The Mach number at the entrance to the conduit, Min (t), is calculated by Fanno ow characteristics for the venting system resistance, 4f L , and the pressure ratio. Mmax is D
4 Fanno ow has been studied extensively, and numerous books describing this ow can be found. Nevertheless, a brief summary on Fanno ow is provided in Appendix A.

9.5. RESULTS AND DISCUSSION

123

the maximum value of Min (t). In vacuum venting, the entrance Mach number, Min (t), is constant and equal to Mmax . Substituting equations (9.4) and (9.5) into equation (9.1), and rearranging, yields: k 1 dP = dt
tmax k1 2k tc M f (Min )P

1t

P;

P (0) = 1.

(9.7)

The solution to equation (9.7) can be obtained by numerical integration for P . The residual mass fraction in the cavity as a function of time is then determined using the ideal gas assumption. It is important to point out the signicance of the tmax . This tc parameter represents the ratio between the lling time and the evacuation time. tc is the time which would be required to evacuate the cylinder for a constant mass ow rate at the maximum Mach number when the gas temperature and pressure remain at their initial values, under the condition that the ow is choked, (The pressure dierence between the mold cavity and the outside end of the conduit is large enough to create a choked ow.) and expressed by tc = m(0) AMmax P0 (0)
k RT0 (0)

(9.8)

Critical condition occurs when tc = tmax . In vacuum venting, the volume pushed by the piston is equal to the ow rate, and ensures that the pressure in the cavity does not increase (above the atmospheric pressure). In air venting, the critical condition ensures that the ow is not choked. For this reason, the critical area Ac is dened as the area that makes the time ratio tmax /tc equal to one. This can be done by looking at equation (9.8), in which the value of tc can be varied until it is equal to tmax and so the critical area is Ac = m(0) tmax Mmax P0 (0)
k RT0 (0)

(9.9)

Substituting equation (9.2) into equation (9.9), and using the fact that the sound velocity can be expressed as c = kRT , yields: Ac = V (0) ctmax Mmax (9.10)

where c is the speed of sound at the initial conditions inside the cylinder (ambient conditions). The tmax should be expressed by Eckert/BarMeir equation.

9.5 Results and Discussion

The results of a numerical evaluation of the equations in the proceeding section are presented in Figure ??, which exhibits the nal pressure when 90% of the stroke has A elapsed as a function of Ac .

124

CHAPTER 9. VENTING SYSTEM DESIGN

A Parameters inuencing the process are the area ratio, Ac , and the friction parameter, 4f L . From other detailed calculations [4] it was found that the inuence of the D parameter 4f L on the pressure development in the cylinder is quite small. The inuence D is small on the residual air mass in the cylinder, but larger on the Mach number, Mexit . A The eects of the area ratio, Ac , are studied here since it is the dominant parameter. Note that tc in air venting is slightly dierent from that in vacuum venting [3] by A a factor of f (Mmax ). This factor has signicance for small 4f L and small Ac when the D Mach number is large, as was shown in other detailed calculations [4]. The denition chosen here is based on the fact that for a small Mach number the factor f (Mmax ) can be ignored. In the majority of the cases Mmax is small. A For values of the area ratio greater than 1.2, Ac > 1.2, the pressure increases the volume ow rate of the air until a quasi steadystate is reached. In air venting, this quasi steadystate is achieved when the volumetric air ow rate out is equal to the volume pushed by the piston. The pressure and the mass ow rate are maintained constant A after this state is reached. The pressure in this quasi steadystate is a function of Ac . A A For small values of Ac there is no steadystate stage. When Ac is greater than one A the pressure is concave upwards, and when Ac is less than one the pressure is concave downwards. These results are in direct contrast to previous molds by Sachs , Draper , Veinik and Lindsey and Wallace , where models assumed that the pressure and mass ow rate remain constant and are attained instantaneously for air venting. To refer to the stroke completion (100% of the stroke) is meaningless since 1) no gas mass is left in the cylinder, thus no pressure can be measured, and 2) the vent can be blocked partially or totally at the end of the stroke. Thus, the completion (end of the process) of the lling process is described when 90% of the stroke is elapsed. Figure ?? presents the nal presA sure ratio as a function Ac for 4f L = 5. Fig. -9.3. The pressure ratios for air and vacD The nal pressure (really the pressure ra- uum venting at 90% of the piston stroke. A tio) depends strongly on Ac as described in Figure ??. The pressure in the die cavity increases by about 85% of its initial value A A when Ac = 1 for air venting. The pressure remains almost constant after Ac reaches A the value of 1.2. This implies that the vent area is suciently large when Ac = 1.2 for A air venting and when Ac = 1 for vacuum venting. Similar results can be observed when the residual mass fraction is plotted. This discussion and these results are perfectly correct in a case where all the assumptions are satised. However, the real world is dierent and the assumptions have to examined and some of them are:
20.0 18.0 16.0 14.0 Vacuum venting Air venting ......... 12.0 P (t =0.9) 10.0 P (t =0) 8.0 6.0 4.0 2.0 0.0 4fL =5.0 D .. .. .. .. .. .. .. .. ... .. . . . .. . .. . ... . ... ... ... ... ... . . . . . ... ... ... . . . ... . .... ... .. . . .. .. ... .. . . . . . . . 1.2 1.5 A Ac 1.8 2.1 2.4 2.7 0.0 0.3 0.6 0.9 3.0

1. Assumption 1 is not a restriction to the model, but rather guide in the design. The engineer has to ensure that the resistance in the mold to air ow (and metal ow) has to be as small as possible. This guide dictates that engineer designs the

9.5. RESULTS AND DISCUSSION path for air (and the liquid metal) as as short as possible.

125

2. Assumptions 3, 4, and 10 are very realistic assumptions. For example, the error in using assumption 4 is less than 0.5%. 3. This model is an indication when assumption 5 is good. In the initial stages (of the lling process) the pressure is very small and in this case the pressure (force) to open the plates is small, and therefore the gap is almost zero. As the lling process progresses, the pressure increases, and therefore the gap is increased. A signicant gap requires very signicant pressure which occurs only at the nal A stages of the lling process and only when the area ratio is small, Ac < 1. Thus, this assumption is very reasonable. 4. Assumption 6 is associated with assumption 9, but is more sensitive. The change in the resistance (a change in assumption in 9 creates consequently a change in the plunger velocity. The plunger reaches the constant velocity very fast, however, this velocity decrease during the duration of the lling process. The change again depends on the resistance in the mold. This can be used as a guide by the engineer and enhances the importance of creating a path with a minimum resistance to the ow. 5. Another guide for the venting system design (in vacuum venting) is assumption 7. The engineer has to reduce the vent volume so that less gas has to be evacuated. This restriction has to be design carefully keeping in mind that the resistance also has to be minimized (some what opposite restriction). In air venting, when this assumption is not valid, a dierent model describes the situation. However, not fullling the assumption can improve the casting because larger portion of the liquid metal which undergoes mixing with the air is exhausted to outside the mold. 6. Assumption 8 is one of the bad assumptions in this model. In many cases there is more than one vent, and the entrance Mach number for dierent vents could be a dierent value. Thus, the suggested method of conversion is not valid, and therefore the value of the critical area is not exact. A better, more complicated model is required. This assumption cannot be used as a guide for the design since as better venting can be achieved (and thus enhancing the quality) without ensuring the same Mach number. 7. Assumption 9 is a partially appropriate assumption. The resistance in venting system is a function of Re and Mach numbers. Yet, here the resistance, 4f L , D is calculated based on the assumption that the Mach number is a constant and equal to Mmax . The error due to this assumption is large in the initial stages where Re and Mach numbers are small. As the lling progress progresses, this error is reduced. In vacuum venting the Mach number reaches the maximum instantly and therefore this assumption is exact. The entrance Mach number is very small (the ow is even not choke ow) in air venting when the area ratio, A Ac >> 1 is very large and therefore the assumption is poor. However, regardless

126

CHAPTER 9. VENTING SYSTEM DESIGN the accuracy of the model, the design achieves its aim and the trends of this model are not aected by this error. Moreover, this model can be improved by taking into consideration the change of the resistance.

8. The change of the vent area does aect the resistance. However, a detailed calculation can show that as long as the vent area is above half of the typical cross section, the error is minimal. If the vent area turns out to be below half of the typical vent cross section a improvement is needed.

9.6 Summary
This analysis (even with the errors) indicates there is a critical vent area below which the ventilation is poor and above which the resistance to air ow is minimal. This critical area depends on the geometry and the lling time. The critical area also provides a mean to combine the actual vent area with the vent resistance for numerical simulations of the cavity lling, taking into account the compressibility of the gas ow. Importance of the design also was shown.

9.7 Questions
Under construction

CHAPTER 10 Density change eects

In this appendix we will derive the boundary condition for phase change boundry at t=0 with a signicant density change. Traditionally in die casting the density change is assumed to be insignicant in die casting. The author is not aware of any liquid solid model in die casting that take this phenomenon into account. In materials like boundry steel and water the density change isnt after some time dx large enough or it does not play furthermore important role. While in die castA B ing the density change play a signicant B A role because a large dierence in values for example aluminum is over 10%. Fig. -10.1. The control volume of the phase Furthermore, the creation of shrinkage change. porosity is a direct consequence of the density change. A constant control volume1 is constructed as shown in gure 10.1. Solid phase is on the right side and liquid phase is on the left side. After a small time increment the moved into the the dashed line at a distance dx. The energy conservation of the control volume reads
1 1 2 2

d dt

h dV =
V A

hvi dA +
A

T dA n

(10.1)

1 A discussion on the mathematical aspects are left out. If explanation on this point will be asked by readers I will added it.

127

128

CHAPTER 10.

DENSITY CHANGE EFFECTS

Analogy the mass conservation for the control volume is d dt dV =

V A

vi dA

(10.2)

The equations (10.1) and (10.2) do not have any restrictions of the liquid movement which has to be solved separately. Multiply equation (10.1) by a constant hl results in d hl dV = hl vi dA (10.3) dt V A Subtraction equation (10.3) from equation (10.1) yields d dt (h hl ) dV =
V A

(h hl )vi dA +
A

T dA n

(10.4)

put explanation or question

The rst term on the right hand side composed from two contributions: one) from the liquid side and two) from solid side. At the solid side the contribution is vanished because (h hl )vi is zero due to vi is identically zero (no movement of the solid, it is a good assumption). In the liquid phase the term h hl is zero (why? ) thus the whole term is vanished we can write the identity (h hl )vi dA 0
A

(10.5)

maybe the derivations are too long. shorten them?

where vi is the velocity at the interface. The rst term of equation (10.4 ) can be expressed in the term of the c.v.2 as
solid liquid=0

s A2 (hs hl ) s A1 (hs hl ) + ( (hl hl )) dt V dx = (s (hs hl )) = s (hs hl )vn (10.6) dt liquid side contribution is zero since h hl 0 and the solid contribution appears only in transitional layer due to transformation liquid to solid. The second term on right hand side of equation (10.4) is simply (h hl ) dV = k
A

d dt

T T T dA = ks kl n n n

(10.7)

Thus, equation (10.4) is transformed into s (hs hl )vn = ks T T kl n n (10.8)

It is noteworthy that the front propagation is about 10previously was calculated. Equation (10.7) holds as long as the transition into solid is abrupt (sharp transition).
2 please

note some dimensions will canceled each other out and not enter into equationsssss

129

Meta

For the case of where the transition to solid occurs over temperature range we have create three zones. Mathematically, it is convenient to describe the the mushy zone boundaries by two boundary conditions.

Meta End Meta Meta End

The creation of voids is results of density changes which change the heat transfer mechanism from conduction to radiation. The location of the void depends on the crystallization and surface tension eect, etc. The possibility of the liquid channels and the ow of semi-solid and even solid compensate for this void.

Kleins paper

Meta

Yet, one has to take into consideration the pressure eect The liquidation temperature and the latent heat are aected somewhat by the pressure. At pressure between the atmospheric to typical intensication pressure the temperature and latent heat are eected very mildly. However, for pressure near vacuum the latent heat and the temperature are eected more noticeably.3

Meta End
The velocity of the liquid metal due to the phase change can be related to the front propagation utilizing the equation (10.2). The left hand side can be shown to be (s l )vn . The right hand side is reduced into only liquid ow and easily can be shown to be l vl . (s l ) vn + l vl = 0 vl ( 1) = vn where is the density ratio, s /l .

(10.9)

3 I have used Clapyrons equation to estimate the change in temperature to be over 10 degrees (actually about 400 [C]). However, I am not sure of this calculations and I had not enough time to check it in the literature. If you have any knowledge and want to save me a search in the library, please drop me a line.

130

CHAPTER 10.

DENSITY CHANGE EFFECTS

CHAPTER 11 Clamping Force Calculations

Under construction

131

132

CHAPTER 11. CLAMPING FORCE CALCULATIONS

It doesnt matter on what machine the product is produce, the price is the same Prof. Al Miller, Ohio

CHAPTER 12 Analysis of Die Casting Economy

12.1 Introduction
The underlying reason for the existence of the die casting process is so that people can make money. People will switch to more ecient methods/processes regardless of any claims die casting engineers make1 . To remain competitive, the die casting engineer must totally abandon the Detroit attitude, from which the automotive industry suffered and barely survived during the 70s. The die casting industry cannot aord such a luxury. This topic is emphasized and dwelt upon herein because the die casting engineer cannot remain stagnate, but rather must move forward. It is a hope that the saying We are making a lot of money why should we change? will totally disappear from the die casting engineers jargon. As in the dairy industry, where keeping track of specics created the super cow, keeping track of all the important information plus using scientic principles will create the super die casting economy. This would be true even if a company, for marketing reasons, needed to oer a wide variety of services to their customers. Which costs the engineer can alter, and what he/she can do to increase prots, are the focus of this chapter. First as usual a discussion on the common model is presented, the validity and the usefulness is discussed, and nally a proper model is unveiled.

12.2 The common model, Millers approach

They started with idea that the price is eected by the following parameters: 1)weight, 2)alloy cost, 3)complexity, 4)tolerance, 5)surface roughness, 7)aspect ratio, 8)produc1 The DDC, a sub set of NADCA operations, is now trying to convince die casting companies to advertise through them to potential customers. Is the role of the DCC or NADCA to be come the middle man? I do not think so. The role of these organizations should be to promote the die casting industry and not any particular company/ies.

133

134

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

tion quantity, and 9)secondary machining. After statistical analysis they have done they comeout with the following equation price = 0.485 + 2.20weight 0.505zinc + 0.791mag + 0.292details +0.637tolerance 0.253quantity where mag, zinc, details(<100 dimension), and tolerance are on/o switch. They claim that this formula is good for up to ten pounds (about 4.5[kg]). In summary, if you expect to get equation that does not have much with the actual cost, you got one. (12.1)

12.3 The validity of Millers price model

There is a saying garbage in garbage out. The proponent conclusion from equation 12.2 is that it does not matter how good the design how much scrap the product generates the price is the same. This is exactly what we are preaching against. The question must be asked, how they calculate the average price of the product that statistically they analyzed, if they have no idea how to the calculate the actual price in the rst place. So, how they determine that the product will produce prot if the price have no relationship to the actual production cost? The critical/optimum point is the point above which the quality is good and below which the quality is unacceptable. As it turns out, much above and just above the critical point produces an acceptable quality product for many design parameters in the die casting process. However, the cost is considerably higher2 . The hydraulic diameter of the runner system is one such example (see Figure ??). Fig. -12.1. Production cost as a function of the The price of the runner system (scrap) runner hydraulic diameter. is proportional to the hydraulic diameter squared, HD 2 (a parabola), as shown by the scrap cost curve in Figure ??. The machine cost is constant (as a rst approximation) up to the point below which the machine cannot produce an acceptable quality. The engineer would like to design the runner diameter just above this point. Machine cost as a function of the runner diameter for several dierent machines is shown by the machine curves in Figure ??. The combined cost of the scrap and the machine usage can be drawn, and clearly the combinedcost curve has a minimum point, and is referred to here as the optimum point3 . This is a typical example of how a design parameter (runner hydraulic diameter) eects the cost and quality. The components of the production cost now should be dissected and analyzed, and then a model will be constructed. It has to be realized that there are two kinds of
cost combined cost
m1 m2

scrap cost

machine cost

m3

m1 > m2 > m3

HD

2 The price of a die casting machine increase almost exponentially with the machine size. Thus, nding the smallest die casting machine to run the job is critical importance. 3 The change in the parts numbers per shot will be discussed in section ??.

12.4. THE COMBINED COST OF THE CONTROLLED COMPONENTS

135

cost components: 1) those which the engineer controls, and 2) those which the engineer does not control. The uncontrolled components include overhead, secondary operations, marketing, space4 , etc. This category should be considered as a constant, since the engineers actions/choices do not aect the cost and therefore do not aect the cost of design decisions. However, the costs of die casting machine capital and operations, personnel cost, melting cost, and scrap cost5 are factors which have to be considered, and are discussed in the succeeding sections. In this analysis it is assumed that the die casting company is here to make a buck, and it is also assumed that competitive price wars for a specic project and/or any other personal reasons inuencing decision making are not relevant6 . This issue is formulated in such a way that the engineer will have the needed tools to make appropriate decisions.

12.4 The combined Cost of the Controlled Components

The engineer has to choose the least expensive machines available, yet produce a product of acceptable quality. The least expensive machine has to chosen. The price for production cost each machine is determined from the sums of every component. If the customer is in a rush, the cost should be calculated for the available die casting machine as follows: total =
i

(12.2)

12.5 Die Casting Machine Capital Costs

The capital cost of a die casting machine (like any other industrial equipment) has two components: 1) money cost and 2) depreciation cost. The money cost in many cases is also comprised of two components: 1) loan cost and 2) desired prot7 . The cost of a loan is interest. The value of the interest rate is easy to evaluate just ask a banker. However, the value of the desired prot is harder to estimate. One possible way to estimate this is by checking how much it costs to lease a similar machine. Adding these two numbers yields a good estimate of the money costs. In todays values, the money cost value is about 12%25%. Depreciation is a loss in value of the die casting machine8 .
4 The roomamount cost for the machine is almost insensitive to the engineers choice of the size or brand of the die casting machine 5 See the discussion on this topic in section 12.7 page 137 for the more detail. 6 such as doing a project to keep a customer for another project are not relevant here. Yet, this information can be used to make intelligent decision in regard to the customer. 7 This prot is dierent from the operational prot. For example, if one own a taxi, he should have two kind of prots: 1) those from owning the taxi and 2) those from operating the taxi. He can rent the taxi and have a prot just for owning the vehicle. The owner should earn additional income for the eight hours shift. These refereed herein as operational earnings. 8 The eects of taxes on the depreciation analysis are sometime signicant, but to reduce the complexity of the explanation here, it is ignored.

136

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

In this analysis, it is assumed (or at least hoped) that the other die casting machines have other jobs waiting for them. If a company for a short time is not working to full capacity, the analysis will still be valid with minor modications. However, a longer duration of being below full capacity requires the company to make surgical solutions. The cost of the die casting machine depends on the market and not on the value the accountant has put on the books for that machine. Clearly, if the machine is to be sold/leased, the value obtained will be according to the market as average value. The market value should be used since the machine can be sold and this money can be invested in other possibilities. Amortization is estimated in the same manner. The dierence between the current value and the value at one year older is the depreciation value.9 Having these numbers, the capital cost can be estimated. For example, a one million dollar machine with a 20% money cost and a 5% depreciation cost equals about $250,000.00 a year. To convert this number to an hourly base rate, the number of idle days (on that specic machine) is required, and in many case is about 60 65 days. Thus, hourly capital cost of that specic machine is about $34.70. A change in the capital cost per unit can be via the change in the cycle time. The change in the cycle time is determined mostly by the solidication processes, which are controlled slightly by the runner design. Yet this eect can be diminished by controlling the cooling rate. Hence, the capital price is virtually unaected once the die casting machine has been selected for a specic project. Here, the cost per unit can be expressed as follows: capital = capital cost per hour Nc Np (12.3)

where capital is the capital cost per unit produced, Nc is the number of cycles per hour, and Np the number of parts shot.

12.6 Operational Cost of the Die Casting Machine

Operational costs are divided into two main categories: 1) energy cost, and 2) maintenance cost. The energy cost is almost insensitive to the mold/runner design. The maintenance cost is determined mostly by the amount of time the die casting machine is in operation. This cost is comprised of the personnel cost of doing the work, hydraulic uid maintenance, components (ladle, etc.) and maintenance, etc., which is dierent for each machine and company. However, the value of this cost can be considered invariant for a specic machine in regard to design parameters. The engineers duty is to calculate the operation cost for every die casting machine that is in the company. This can be achieved by keeping records of the maintenance for each machine and adding up all related costs performed on that machine in the last year. The energy costs are the costs of moving the die casting machine and its parts and accessories. The energy needed to move all parts is the electrical energy which can easily be measured. Today, electrical energy costs are far below one dollar for one [kW]hour (0.06-0.07 of a dollar according to NSP prices). Even a large job will require
9 This

also depends on changes in the condition of the machine.

12.7. RUNNER COST (SCRAP COST)

137

less than 10[kWhour]. Thus, the total energy cost is in most cases at most $1.00 per hour. The change in the energy is insensitive to the runner and venting system designs and can vary by only 30% (15 cents for a very very large job), which is insignicant compared to all other components. The operation cost can be expressed as operation = operation cost per hour f(machine size, type etc.) Nc Np (12.4)

12.7 Runner Cost (Scrap Cost)

The main purpose of the runner is to deliver the liquid metal from the shot sleeve to the mold, since the mold cannot be put (hooked) directly on (to) the shot sleeve. The requirements of the runner have conicting demands. Here is a partial list of the requirements for the runner: 1. As small as possible so it will create less scrap. 2. Large enough so that there is less resistance in the runner to the liquid metal ow, so that the job can be performed on a smaller die casting machine. 3. Small enough so that the plunger will need to propel only a minimum amount of liquid metal. In a way this is the same as requirement 1 above but less important. Clearly, a large runner volume creates more scrap and is a linear function of the size of the runner volume, which is
area
0.20 0.18 0.16 0.14 0.12 . ... ... .. .. .. .. .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . .. ... .. . .. . ..... . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 0.80 0.90

Vrunner

HD T 2 = LT 4

length

0.10 0.08

(12.5)

0.06 0.04 0.02 . 0.00 ... ... 0.00 ... . .. ..

..

where HD is the typical size of the hyQ draulic diameter, and LT its length (these values are not the actual values, but they are used to represent the sizes of the run- Fig. -12.2. The reduced power of the die casting ner). From equation (12.5), it is clear that machine as a function of the normalized ow the diameter has one of the greater im- rate. pacts on the scrap cost. The minimum diameter at which a specic machine can produce good quality depends on the required lling time, gate velocity, other runner design characteristics, and the characteristics of the specic machine. Scrap cost is a linear function of the volume10 . The scrap cost per volume/weight consists of three components: 1) the melting cost, 2) the dierence between the buying price and the selling price (assuming that the scrap can be sold), and 3) the handling
0.10 0.20 0.30 0.40 0.50 0.60 0.70 1.00

10 Up

to a point about which it becomes more sensitive to the volume.

138

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

cost. The melting cost includes the cost to raise the metal temperature to the melting point, to melt the metal, and to hold the metal temperature above the melting point. The melting cost can be calculated by measuring energy used (crude oil or natural oil in most cases) plus the maintenance cost of the furnace divided by the amount of metal that has been casted (the parts and design scrap). The buying price is the price paid for the raw material; the selling price is the price for selling the scrap. Sometime it is possible to reuse the scrap and to remelt the metal. In some instances, the results of reusing the scrap will be a lower grade of metal in the end product. If reuse is possible, the dierence in cost should be substituted by the lost metal cost, which is the cost of 1) metal that cannot be recycled and 2) metal lost due to the chemical reactions in the furnace. The handling cost is the cost encountered in selling the metal, and it includes changing the mechanical or chemical properties of the scrap, transportation, cost of personnel, storage, etc. Each handling of the metal costs a dierent amount, and the specics can be recorded for the specic metal. Every job/mold has typical ranges for the lling time and gate velocity. Moreover, a rough design for the runner system can be produced for the mold. With these pieces of information in place, one can calculate the gate area (see pQ2 diagram calculations in Chapter 7 for more details, and this part is repeated in that Chapter. I am looking for the readers input to decided what is the best presentation.). Then the ow rate for the mold can be calculated by Q= Vmold Agate Vgate (12.6)

Additionally, the known design of the runner with ow rate yields the pressure dierence in the runner, and this yields the power required for the runner system, Pr = QP or in normalized form, Pr = QP Pmax Qmax QP (12.8) (12.7)

Every die casting machine has a characteristic curve on the pQ2 diagram as well. 2 Assuming that the die casting machine has the common characteristic, P = 1 Q , the normalized power can be expressed Pm = Q(1 Q ) = Q Q
2 2 3

(12.9)

where Pm is the machine power normalized by Pmax Qmax . The maximum power of this kind of machine is at 2/3 of the normalized ow rate, Q, as shown in Figure ??. It is recommended to design the process so the ow rate occurs at the vicinity of the maximum of the power. For a range of 1/3 of Q that is from 0.5Q to 0.83Q, the average power is 0.1388 Pmax Qmax , as shown in Figure ?? by the shadowed rectangular. One may notice that this value is above the capability of the die casting machine in two

12.8. STARTUP AND MOLD MANUFACTURING COST

139

ranges of the ow rate. The reason that this number is used is because with some improvements of the the runner design the job can be performed on this machine, and there is no need to move the job to a larger machine. If the machine power turns out to be larger than the required power of the runner, Pm > Cs Pr , the job can then be performed on the machine; otherwise, a bigger die casting machine is required. In general, the number of molds castable in a single cycle is given by Np = Pm Cs Pr = Pm Cs Pr (12.10)

The oor symbol being used means that the number is to be rounded down to the nearest integer. Cs denotes the safety factor coecient. In the case that Np is less than one, Np > 1, that specic machine is too small for this specic job. After the number of the parts has been determined (rst approximation) the runner system has to be redesigned so that the required power needed by the runner can be calculated more precisely. Plugging the new numbers into equation (12.10) yields a better estimation of the number of parts. If the number does not change, this is the number of parts that can be produced; otherwise, the procedure must be repeated. In this analysis, the required clamping forces that the die casting machine can produced are not taken into consideration. Analysis of the clamping forces determines the number of possible parts and it is a dierent criterion which required to satised, this will be discussed in more detail in Chapter 11. The actual number of parts that has to be taken into consideration is the smaller of the two criteria. Next, the new volume of the runner system has to be calculated. The cost per cavity is the new volume divided by the number of cavities: scrap = Vrunner (cost per volume) Np (12.11)

12.8 Startup and Mold Manufacturing Cost

The cost of manufacturing of a mold is aected slightly by the the shape of the runner. The only exemption to the above statement is the eect of change of the cross section shape and size on the cost of manufacturing which will be discussed in Chapter 6. A larger part of expense is the startup time cost which is composed of 1) rebuilding the mold, 2) lost time (personnel time, machine time, etc), and 3) lost material. When dealing with calculation of the startup time two things have to be taken into account 1) the ratio of the start up cost to the total cost, and 2) how long it is expected to take to achieve a product of acceptable quality. The cap cost has to be determined from the total cost per unit, and then multiplied by the total number of units. This number is the net production cost. The startup cost cannot (should not) exceed 10%15% of that number. Presently, it is very hard to determined the number of trials that will required per mold. This number is related to the complexity of the shape. The more complex

140

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

the shape is, the more likely it is that the number of attempted shots will increase. If it is assumed that the engineer is experienced, the only factor that will aect the number of shots will be the complexity provided that the job can be performed on the same die casting machine. The complexity of the shape should present a general idea of the number of expected attempts, and should be used in calculating the startup cost, startU p = (Cost per attempt) Na Nr (12.12)

where Na is the number of attempts, and Nr is the number of the total parts to be produced.

12.9 Personnel Cost

The cost of personnel is aected by the cycle time plus the number of parts produced per cycle. With todays automatization, the number of operators is decreasing. In some companies, one operator controls three or more machines. Hence, the personnel cost is: personnel = salary per hour number of machines number of cycle (12.13)

In toadys market, the operator cost is in the range of $10$20 per hour. When automatization is used, the personnel cost is signicantly reduced to the point that it is insignicant.

12.10 Uncontrolled components

The price to be charged to the customer has to include the uncontrolled components as well. There are several methods for adding this fragment to the part cost. First, the total cost of the uncontrolled components has to be calculated. This can be done by adding up the costs from the previous year and estimated for this year This cost includes salaries that were paid in the last year plus the legal expenses, rent, and marketing, etc. Dividing the uncontrolled components of cost has many reasonable options. Here is a selected list according to:
the number of parts the number of parts and their size/weight the number of the parts and their complexity

12.11. MINIMIZING COST OF SINGLE OPERATION

141

12.11 Minimizing Cost of Single Operation

In this section several issues related to cost minimization and/or prot increases are discussed. An example of such problem is when an engineer has to make a decision about supplies. Ordering a supply commonly to has to do with two or more conicting costs. These two conicting costs have a minimum of ordering cost associated with an optimum number of orders. Consider a simple situation where ABC company produces x devices per year, for example x hard drive frames per year. These x devices require y items supporting components. For example, these hard drive frames are made of y mass of aluminum11 . So, the ABC company has to order y items supporting components a year. It is assumed that the number of items ordered and consumed per year is constant12 In a typical situation, these items are ordered several times, the cost is composed of two components. The rst component is the cost per order which includes such things as the delivery cost, the time to order (verication if it is in stock). The second component is associated with the cost of keeping the stock beyond one day supply. This category includes such as the money tied to material, the storage used (cost of space and handling) etc. For simplicity, it is assumed that the cost per delivery is constant and does not change during the year. It also assumed that the storage per day cost is also constant. To illustrate this point, consider two extreme cases of ordering. One possibility is to order everyday and other possibility is to order once per year. The cost for ordering very day is 36513 times the ordering cost. In this case, there is no stocking cost. The other extreme is one order a year for which there is one time order cost with 364 days of stocking cost. Example 12.1: Under conditions below, calculate the ordering cost for ordering every day of the year and once a year. The delivery cost of aluminum is $150 per delivery. The cost of storing a kilogram aluminum is 0.10 dollars. And the cost of a kilogram aluminum is 10 dollars. The daily aluminum consumption is 750 kilogram. The money cost is 0.01 percent a day. Solution The ordering cost is $150. Hence, the cost for the rst case is composed of the ordering cost (in this case only delivery cost) only. The stocking cost is vanished because no stocking is involved. Thus, the total cost is total cost = 365 150 = $54, 750 (12.I.a)

The ordering occurs once a year with 364 days of stocking cost which must be added.
the weight for the total of the hard drives [in kilograms]. the number is not constant or even seasonal (depend on the season) this model can be expanded by numerical analysis. 13 It is assumed that the year is 365 days which about 75% of the occurrence. There are years with 356 days.
12 If 11 Or

142

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

The daily stocking cost one day consumption is

stocked material amount cost per kilogram

stocking cost = per day

750

10.0 0.0001 = $0.75

(12.I.b)

On the rst day has 364 portions (days) to be stocked. On the second day has 363 portions (days) to be stocked. On every sequence day there is one less portions to be stocked. Hence, there is a series of 364 items which starts with 364 and end at zero. This exactly algebraic series which can be calculated as a1 + a364 364 + 0 total = 364 = 364 = 66248 portions 2 2 Thus, the total cost for ordering once is total cost for once = $150 + 66248 0.75 = $49836 ordering (12.I.c)

(12.I.d)

For the purpose of this example, the choice of one time order is better. While the cost of one time ordering looks better in this example, in real life other factors should be considered. In this case case, the erosion of the company credit which was not a considered here. That fact should reversed the decision.
End Solution

Example 12.1 exhibits the eect of the number of orders on the cost of operation. Intuitively, for example, it can be observed that for two (2) orders, as compare to one order, the cost is reduced by half plus $150 (why?). The focus of this discussion is to nd the optimum number of times to order the items per year. Not all years are the same (some years are 366 days) but here it will be assumed that all years are 365 days. In this discussion, months and weeks do not appear and the dissimilarity is not discussed. The number of items ordered per year is assumed to be constant for this discussion and is denoted as N . Hence, the number of order items per day is N/365. The cost per day of store of item is denoted y. The ordering cost is r. The unknown number of periods is denoted as p. The number of days per period is D= 365 p (12.14)

The period cost of storage is the number parts not used which were kept for the following days of the period. The number of items remained for the storage for the rst day of the period are Remainder = (D 1) N 365 (12.15)

12.11. MINIMIZING COST OF SINGLE OPERATION On the second day, the number of items that remain in storage is Remainder = (D 2) N 365

143

(12.16)

There are no items to stored on the last day of the period. The number of days that we had to store is (D 1) (no need to store at the last day). The cost of the storage per period is
rst day second day 3rd last day day before last last day

N N N number of = (D 1) + (D 2) + + 2 + 365 365 365 items

N 365

+0

N 365 (12.17)

Equation (12.17) can be rearranged (because it is a regular algebraic series) is N number of = 365 items or
averaged to be stored days number of days D1

D1i
i=1

(12.18)

number of (D 1) + 1 = 2 items The cost per period is

(D 1)

N D (D 1) N = 365 2 365

(12.19)

storage cost D (D 1) N = y per period 2 365 365 p 365 1 p 2 365 1 p y

(12.20)

Or in terms of the number of period, p (see also equation (12.14)) the storage cost is storage cost = per period The yearly storage cost is N y 365

(12.21)

yearly storage = storage cost p = N per period 2 cost The total cost yearly ordering cost, C, is C=

(12.22)

(12.23)

yearly yearly + storage cost ordering cost

144

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

In term of the number of periods the total ordering cost is yearly cost = rp+ N 2 365 1 p y (12.24)

The minimum cost expressed by the expression (12.24) a derivative with respect to p number of period as C N y 365 =r+ p 2 The solution of equation (12.25) is p= N y 365 2r (12.26) 1 p2 =0 (12.25)

These calculations were made under the assumption that the number of periods is a real number. However, the number of periods and several other parameters must be an integer. It can be argued that the number of orders can be 6.5 on the account that in two years planning to have 13 orders (13/2 = 6.5) It is more common to have solution with a totally irrational number which leads in practicality solution that cannot be used. The real solution (in a yearly integer planning sense) lay either on one of adjoining sides of the continuous solution. It has to be manually calculated. The calculation can yield a number of periods to be below one. The actual meaning is that the ordering cost is signicant (dominate) so the order must be continuous. On the other extreme, when the ordering cost is so insignicant to order can occur several time a day. Example 12.2: In ABC die casting company has 68000 kg of aluminum a year. The cost of storage of 1 kg a day is $0.04 and cost of ordering is $130. What is the optimum order period? Solution The information provides that p= 68000 0.04 365 = 3.23443016 2 130 (12.II.a)

The solution is either 3 or 4 times when additional consideration has to be taken into account.
End Solution

12.12 Introduction to Economics

The main goal of a company is to increase the total prots which will be denoted as T . The total prots, T , is a function of (x) the number of the items sold. There is a cost

12.13. SUMMARY

145

(C) and revenue (R) which associated with these items both which are a function of x. Thus, T (x) = R(x) C(x) (12.27)

The revenue is a function of number items, x, the price, P (x) is also a function of x. To add discussion about the discrete v. approximation of the smooth derivation. The typical cost is
cost per item more complex situation

C(x) = f ixed +

x+

f (x)

(12.28)

Hence, the cost per item (if a is a function of x) is c(x) = 1 f (x) C(x) = a + d(x) + x x x (12.29)

Examining the terms in equation (12.29) reveals that there are terms that vanish as the number of items increase to innity x 0(1/x) = 1/ = 0. However, other terms behaves according the supply and demand diagram (pQ2 diagram). In the supply and demand diagram is diagram and was proposed by Marco Fanno (the other Fanno). The interesting part of the supply diagram show that the price decreases as the number of items increases. This part of the analysis shows indicate that the profits are a strong function item parts. These facts shows that situations where the total prots is reduced when the total income is increased. A typical equation repressing this situation is

TA2

Fig. -12.3. Supply and Demand.

C(x) = 10, 000 + 45 x + 100 x

(12.30)

12.13 Summary
In this chapter the economy of the design and choices of the casting process have been presented. It is advocated that the averaged approach commonly used in the die casting industry be abandoned. Adopt a more elaborate method, in which more precise calculations are made is also advocated. It is believed that the new method will create the super die casting economy.

12.14 Question

146

CHAPTER 12. ANALYSIS OF DIE CASTING ECONOMY

APPENDIX A Fanno Flow

An adiabatic ow with friction is named after Ginno Fanno a Jewish flow direction engineer. This model is the second pipe ow model described here. The main restriction for this model is that heat transfer is negligible and can be c.v. ignored 1 . This model is applicaNo heat transer ble to ow processes which are very fast compared to heat transfer mechFig. -A.1. Control volume of the gas ow in a conanisms with small Eckert number.

This model explains many industrial ow processes which includes emptying of pressured container through a relatively short tube, exhaust system of an internal combustion engine, compressed air systems, etc. As this model raised from need to explain the steam ow in turbines.

stant cross section

A.1 Introduction
Consider a gas owing through a conduit with a friction (see Figure (A.1)). It is advantages to examine the simplest situation and yet without losing the core properties of the process. Later, more general cases will be examined2 .
1 Even 2 Not

the friction does not convert into heat ready yet, discussed on the ideal gas model and the entry length issues.

147

148

APPENDIX A. FANNO FLOW

A.2 Fanno Model

The mass (continuity equation) balance can be written as m = AU = constant 1 U1 = 2 U2 (A.1)

The energy conservation (under the assumption that this model is adiabatic ow and the friction is not transformed into thermal energy) reads T0 1 = T0 2 T1 + U1 U2 = T2 + 2cp 2cp
2 2

(A.2)

(A.3) Or in a derivative from Cp dT + d U2 2 =0 (A.4)

Again for simplicity, the perfect gas model is assumed3 . P = RT P1 P2 = 1 T1 2 T2 (A.5)

It is assumed that the ow can be approximated as onedimensional. The force acting on the gas is the friction at the wall and the momentum conservation reads AdP w dAw = mdU It is convenient to dene a hydraulic diameter as DH = Or in other words A= DH 2 4 (A.8) 4 Cross Section Area wetted perimeter (A.7) (A.6)

3 The equation of state is written again here so that all the relevant equations can be found when this chapter is printed separately.

A.3. NONDIMENSIONALIZATION OF THE EQUATIONS

149

It is convenient to substitute D for DH and yet it still will be referred to the same name as the hydraulic diameter. The innitesimal area that shear stress is acting on is dAw = Ddx (A.9)

Introducing the Fanning friction factor as a dimensionless friction factor which is some times referred to as the friction coecient and reads as the following: f= w 1 U 2 2 (A.10)

By utilizing equation (A.2) and substituting equation (A.10) into momentum equation (A.6) yields
A w 2

D dP Ddx f 4

1 2 U 2

m A

= A U dU

(A.11)

Dividing equation (A.11) by the cross section area, A and rearranging yields dP + 4f dx D 1 2 U 2 = U dU (A.12)

The second law is the last equation to be utilized to determine the ow direction. s2 s1 (A.13)

A.3 NonDimensionalization of the Equations

Before solving the above equation a dimensionless process is applied. By utilizing the denition of the sound speed to produce the following identities for perfect gas M2 = U c
2

U2 k RT
P

(A.14)

Utilizing the denition of the perfect gas results in M2 = U 2 kP (A.15)

Using the identity in equation (A.14) and substituting it into equation (A.11) and after some rearrangement yields 4f dx DH 1 kP M 2 2 U dU dU = kP M 2 U U
2 U 2

dP +

(A.16)

150 By further rearranging equation (A.16) results in dP 4f dx P D kM 2 2

APPENDIX A. FANNO FLOW

= kM 2

dU U

(A.17)

It is convenient to relate expressions of (dP/P ) and dU/U in terms of the Mach number and substituting it into equation (A.17). Derivative of mass conservation ((A.2)) results in
dU U

d 1 dU 2 + =0 2 U2

(A.18)

The derivation of the equation of state (A.5) and dividing the results by equation of state (A.5) results dP d dT = + P dT (A.19)

Derivation of the Mach identity equation (A.14) and dividing by equation (A.14) yields d(M 2 ) d(U 2 ) dT = 2 M U2 T (A.20)

Dividing the energy equation (A.4) by Cp and by utilizing the denition Mach number yields dT + T 1 kR (k 1)
Cp

1 U2 d T U2

U2 2

dT (k 1) U 2 + d T kRT U 2
c2

U2 2

dT k 1 2 dU 2 + M =0 T 2 U2

(A.21)

Equations (A.17), (A.18), (A.19), (A.20), and (A.21) need to be solved. These equations are separable so one variable is a function of only single variable (the chosen as the independent variable). Explicit explanation is provided for only two variables, the rest variables can be done in a similar fashion. The dimensionless friction, 4f L , D is chosen as the independent variable since the change in the dimensionless resistance, 4f L D , causes the change in the other variables. Combining equations (A.19) and (A.21) when eliminating dT /T results dP d (k 1)M 2 dU 2 = P 2 U2 (A.22)

A.3. NONDIMENSIONALIZATION OF THE EQUATIONS

151

The term d can be eliminated by utilizing equation (A.18) and substituting it into equation (A.22) and rearrangement yields dP 1 + (k 1)M 2 dU 2 = P 2 U2 The term dU 2 /U 2 can be eliminated by using (A.23) kM 2 1 + (k 1)M 2 4f dx dP = P 2(1 M 2 ) D (A.24) (A.23)

The second equation for Mach number, M variable is obtained by combining equation (A.20) and (A.21) by eliminating dT /T . Then d/ and U are eliminated by utilizing equation (A.18) and equation (A.22). The only variable that is left is P (or dP/P ) which can be eliminated by utilizing equation (A.24) and results in 1 M 2 dM 2 4f dx = D kM 4 (1 + k1 M 2 ) 2 Rearranging equation (A.25) results in kM 2 1 + k1 M 2 4f dx dM 2 2 = M2 1 M2 D (A.26) (A.25)

After similar mathematical manipulation one can get the relationship for the velocity to read dU kM 2 4f dx = U 2 (1 M 2 ) D and the relationship for the temperature is 1 dc k(k 1)M 4 4f dx dT = = T 2 c 2(1 M 2 ) D density is obtained by utilizing equations (A.27) and (A.18) to obtain d kM 2 4f dx = 2 (1 M 2 ) D The stagnation pressure is similarly obtained as kM 2 4f dx dP0 = P0 2 D The second law reads ds = Cp ln dP dT R ln T P (A.31) (A.30) (A.29) (A.28) (A.27)

152

APPENDIX A. FANNO FLOW

The stagnation temperature expresses as T0 = T (1 + (1 k)/2M 2 ). Taking derivative of this expression when M remains constant yields dT0 = dT (1 + (1 k)/2M 2 ) and thus when these equations are divided they yield dT /T = dT0 /T0 (A.32)

In similar fashion the relationship between the stagnation pressure and the pressure can be substituted into the entropy equation and result in ds = Cp ln dP0 dT0 R ln T0 P0 (A.33)

The rst law requires that the stagnation temperature remains constant, (dT0 = 0). Therefore the entropy change is ds (k 1) dP0 = Cp k P0 Using the equation for stagnation pressure the entropy equation yields ds (k 1)M 2 4f dx = Cp 2 D (A.35) (A.34)

A.4 The Mechanics and Why the Flow is Choked?

The trends of the properties can be examined by looking in equations (A.24) through (A.34). For example, from equation (A.24) it can be observed that the critical point is when M = 1. When M < 1 the pressure decreases downstream as can be seen from equation (A.24) because f dx and M are positive. For the same reasons, in the supersonic branch, M > 1, the pressure increases downstream. This pressure increase is what makes compressible ow so dierent from conventional ow. Thus the discussion will be divided into two cases: One, ow above speed of sound. Two, ow with speed below the speed of sound. Why the ow is choked? Here, the explanation is based on the equations developed earlier and there is no known explanation that is based on the physics. First, it has to be recognized that the critical point is when M = 1. It will be shown that a change in location relative to this point change the trend and it is singular point by itself. For example, dP (@M = 1) = and mathematically it is a singular point (see equation (A.24)). Observing from equation (A.24) that increase or decrease from subsonic just below one M = (1 ) to above just above one M = (1 + ) requires a change in a sign pressure direction. However, the pressure has to be a monotonic function which means that ow cannot crosses over the point of M = 1. This constrain means that because the ow cannot crossover M = 1 the gas has to reach to this speed, M = 1 at the last point. This situation is called choked ow.

A.5. THE WORKING EQUATIONS The Trends

153

The trends or whether the variables are increasing or decreasing can be observed from looking at the equation developed. For example, the pressure can be examined by looking at equation (A.26). It demonstrates that the Mach number increases downstream when the ow is subsonic. On the other hand, when the ow is supersonic, the pressure decreases. The summary of the properties changes on the sides of the branch Subsonic decrease increase increase decrease decrease decrease Supersonic increase decrease decrease increase increase increase

Pressure, P Mach number, M Velocity, U Temperature, T Density, Stagnation Temperature, T0

A.5 The Working Equations

Integration of equation (A.25) yields 4 D
Lmax k+1 2 1 1 M2 k+1 2 M + ln k1 2 k M 2k 1 + 2 M2

f dx =
L

(A.36)

A representative friction factor is dened as f= 1 Lmax

0 Lmax

f dx

(A.37)

In the isothermal ow model it was shown that friction factor is constant through the process if the uid is ideal gas. Here, the Reynolds number dened in equation (??) is not constant because the temperature is not constant. The viscosity even for ideal gas is complex function of the temperature (further reading in Basic of Fluid Mechanics chapter one, Potto Project). However, the temperature variation is very limit. Simple improvement can be done by assuming constant constant viscosity (constant friction factor) and nd the temperature on the two sides of the tube to improve the friction factor for the next iteration. The maximum error can be estimated by looking at the maximum change of the temperature. The temperature can be reduced by less than 20% for most range of the spesic heats ratio. The viscosity change for this change is for many gases about 10%. For these gases the maximum increase of average Reynolds number is only 5%. What this change in Reynolds number does to friction factor? That depend in the range of Reynolds number. For Reynolds number larger than 10,000 the change in friction factor can be considered negligible. For the other extreme, laminar

154

APPENDIX A. FANNO FLOW

ow it can estimated that change of 5% in Reynolds number change about the same amount in friction factor. With the exception the jump from a laminar ow to a turbulent ow, the change is noticeable but very small. In the light of the about discussion the friction factor is assumed to constant. By utilizing the mean average theorem equation (A.36) yields
k+1 2 4f Lmax 1 1 M2 k+1 2 M = + ln D k M2 2k 1 + k1 M 2 2

(A.38)

It is common to replace the f with f which is adopted in this book. Equations (A.24), (A.27), (A.28), (A.29), (A.29), and (A.30) can be solved. For example, the pressure as written in equation (A.23) is represented by 4f L , and D Mach number. Now equation (A.24) can eliminate term 4f L and describe the pressure D on the Mach number. Dividing equation (A.24) in equation (A.26) yields
dP P dM 2 M2

1 + (k 1M 2 dM 2 2M 2 1 + k1 M 2 2

(A.39)

The symbol * denotes the state when the ow is choked and Mach number is equal to 1. Thus, M = 1 when P = P equation (A.39) can be integrated to yield:

P 1 = P M

1+

k+1 2 k1 2 2 M

(A.40)

In the same fashion the variables ratio can be obtained

k+1 T c2 2 = 2 = T c 1 + k1 M 2 2

(A.41)

1 = M

1+

k1 2 2 M k+1 2

(A.42)

U = U

=M

1+

k+1 2 k1 2 2 M

(A.43)

A.5. THE WORKING EQUATIONS The stagnation pressure decreases and can be expressed by (1+ 1k M 2 ) k1 2 P0 = P0 P0 P P0 P
2 ( k+1 ) k1 k k

155

P (A.44) P

Using the pressure ratio in equation (A.40) and substituting it into equation (A.44) yields P0 = P0 1+
k1 2 2 M k+1 2
k k1

1 M

1+

k1 2 2 M k+1 2

(A.45)

And further rearranging equation (A.45) provides

P0 1 1 + k1 M 2 2 = k+1 P0 M 2 The integration of equation (A.34) yields s s = ln M 2 cp

k+1 2(k1)

(A.46)

2M 2

k+1 1 + k1 M 2 2

k+1 k

(A.47)

The results of these equations are plotted in Figure (A.2) The Fanno ow is in many cases shockless and therefore a relationship between two points should be derived. In most times, the star values are imaginary values that represent the value at choking. The real ratio can be obtained by two star ratios as an example T2 = T1
T T M2 T T M1

(A.48)

A special interest is the equation for the dimensionless friction as following

L2 L1

4f L dx = D

Lmax L1

4f L dx D

Lmax L2

4f L dx D

(A.49)

Hence, 4f Lmax D =
2

4f Lmax D

4f L D

(A.50)

156
* *

APPENDIX A. FANNO FLOW

P/P , / and T/T as a function of M 1e+02 4fL D P P * T/T * P0/P0 U/U*
*

Fanno Flow *

1e+01

0.1

0.1 Tue Sep 25 10:57:55 2007

0.01

1 Mach number

10

Fig. -A.2. Various parameters in Fanno ow as a function of Mach number

A.6 Examples of Fanno Flow

Example A.1: Air ows from a reservoir and enters a uni form pipe with a diameter of 0.05 [m] and length of 10 [m]. The air exits to the at mosphere. The following conditions prevail at the exit: P2 = 1[bar] temperature T2 = 27 C M2 = 0.94 . Assume that the average friction factor to be f = 0.004 and that the ow from the reservoir up to the pipe inlet Fig. -A.3. Schematic of Example (A.1) is essentially isentropic. Estimate the total temperature and total pressure in the reservoir under the Fanno ow model.

Solution For isentropic, the ow to the pipe inlet, the temperature and the total pressure at the
4 This

property is given only for academic purposes. There is no Mach meter.

A.6. EXAMPLES OF FANNO FLOW

157

pipe inlet are the same as those in the reservoir. Thus, nding the total pressure and temperature at the pipe inlet is the solution. With the Mach number and temperature known at the exit, the total temperature at the entrance can be obtained by knowing the 4f L . For given Mach number (M = 0.9) the following is obtained. D M
4fL D P P P0 P0 U U T T

0.90000 0.01451 1.1291 So, the total temperature at the exit is T |2 = T T

1.0089

1.0934

0.9146

1.0327

T2 =
2

300 = 290.5[K] 1.0327

4f L D

To move to the other side of the tube the

4f L D

is added as 3.21

4f L D

4f L D

4 0.004 10 + 0.01451 0.05

4f L D

The rest of the parameters can be obtained with the new by interpolations or by utilizing the attached program. M
4fL D P P P0 P0

either from Table (A.1)

U U

T T

0.35886 3.2100

3.0140

1.7405

2.5764

0.38814 1.1699

Note that the subsonic branch is chosen. The stagnation ratios has to be added for M = 0.35886 M
T T0 0 A A P P0 AP A P0 F F

0.35886 0.97489 0.93840 1.7405

0.91484 1.5922

0.78305

The total pressure P01 can be found from the combination of the ratios as follows:
P1 P

P01 = P2

P P

P P

P0 P

1 1 3.014 = 2.91[Bar] =1 1.12913 0.915

158

APPENDIX A. FANNO FLOW

T1 T

T01 = T2

T T0 2 T 1 T 1 1 1 1.17 =300 1.0327 0.975

T T

348K = 75 C

End Solution

Another academic question/example: Example A.2: A system is composed of a convergentdivergent nozzle followed by a tube with length of 2.5 [cm] in diameter and 1.0 [m] long. The system is supplied by a vessel. The vessel conditions are at 29.65 [Bar], 400 K. With these conditions a pipe inlet Mach number is 3.0. A normal shock wave occurs in the tube and the ow discharges to the atmosphere, determine: (a) the mass ow rate through the system; (b) the temperature at the pipe exit; and (c) determine the Mach number when a normal shock wave occurs [Mx ]. Take k = 1.4, R = 287 [J/kgK] and f = 0.005. Solution

shock d-c nozzle

atmosphere conditions

Fig. -A.4. The schematic of Example (A.2)

(a) Assuming that the pressure vessel is very much larger than the pipe, therefore the velocity in the vessel can be assumed to be small enough so it can be neglected. Thus, the stagnation conditions can be approximated for the condition in the tank. It is further assumed that the ow through the nozzle can be approximated as isentropic. Hence, T01 = 400K and P01 = 29.65[P ar]

A.6. EXAMPLES OF FANNO FLOW

159

The mass ow rate through the system is constant and for simplicity point 1 is chosen in which, m = AM c The density and speed of sound are unknowns and need to be computed. With the isentropic relationship the Mach number at point one (1) is known, then the following can be found either from Table (A.1) or the PottoGDC M 3.0000
T T0 0 A A P P0 AP A P0 F F

0.35714 0.07623 4.2346

0.02722 0.11528 0.65326

The temperature is T1 = T1 T01 = 0.357 400 = 142.8K T01

Using the temperature, the speed of sound can be calculated as c1 = kRT = 1.4 287 142.8 239.54[m/sec] The pressure at point 1 can be calculated as P1 = P1 P01 = 0.027 30 P01 0.81[Bar]

The density as a function of other properties at point 1 is 1 = P RT =

1

8.1 104 287 142.8

1.97

kg m3

The mass ow rate can be evaluated from equation (A.2) m = 1.97 0.0252 kg 3 239.54 = 0.69 4 sec

(b) First, check whether the ow is shockless by comparing the ow resistance and the maximum possible resistance. From the Table (A.1) or by using the Potto GDC, to obtain the following M 3.0000
4fL D P P P0 P0 U U T T

0.52216 0.21822 4.2346

0.50918 1.9640

0.42857

160 and the conditions of the tube are

4f L D

APPENDIX A. FANNO FLOW

4 0.005 1.0 = 0.8 0.025

Since 0.8 > 0.52216 the ow is choked and with a shock wave. The exit pressure determines the location of the shock, if a shock exists, by comparing possible Pexit to PB . Two possibilities are needed to be checked; one, the shock at the entrance of the tube, and two, shock at the exit and comparing the pressure ratios. First, the possibility that the shock wave occurs immediately at the entrance for which the ratio for Mx are (shock wave Table (??))

Mx 3.0000

My 0.47519

Ty Tx

y x

Py Px

P0 y P0x

2.6790

3.8571

10.3333

0.32834

After the shock wave the ow is subsonic with M1 = 0.47519. (Fanno ow Table (A.1))

4fL D

P P

P0 P0

U U

T T

0.47519 1.2919

2.2549

1.3904

1.9640

0.50917 1.1481

The stagnation values for M = 0.47519 are M

T T0 0 A A P P0 AP A P0 F F

0.47519 0.95679 0.89545 1.3904

0.85676 1.1912

0.65326

The ratio of exit pressure to the chamber total pressure is

1 1

P2 = P0 = =

P2 P

P0 y P P1 P0 x P1 P0 y P0 x P0 1 1 0.8568 0.32834 1 2.2549 0.12476

A.7. SUPERSONIC BRANCH

161

The actual pressure ratio 1/29.65 = 0.0338 is smaller than the case in which shock occurs at the entrance. Thus, the shock is somewhere downstream. One possible way to nd the exit temperature, T2 is by nding the location of the shock. To nd the location of the shock ratio of the pressure ratio, P2 is needed. P1 With the location of shock, claiming upstream from the exit through shock to the entrance. For example, calculate the parameters for shock location with known 4f L in the y side. Then either by utilizing shock table or the program, D to obtain the upstream Mach number. The procedure for the calculations: Calculate the entrance Mach number assuming the shock occurs at the exit: 1) a) set M2 = 1 assume the ow in the entire tube is supersonic: b) calculated M1 Note this Mach number is the high Value. Calculate the entrance Mach assuming shock at the entrance. a) set M2 = 1 2) b) add 4f L and calculated M1 for subsonic branch D c) calculated Mx for M1 Note this Mach number is the low Value. According your root nding algorithm5 calculate or guess the shock location and then compute as above the new M1 . a) set M2 = 1 3) b) for the new 4f L and compute the new M for the subsonic branch y D c) calculated Mx for the My d) Add the leftover of
4f L D

and calculated the M1

4) guess new location for the shock according to your nding root procedure and according to the result, repeat previous stage until the solution is obtained. M1 3.0000 M2 1.0000
4fL D up 4fL D down

Mx 1.9899

My 0.57910

0.22019

0.57981

(c) The way of the numerical procedure for solving this problem is by nding

4f L D

that will produce M1 = 3. In the process Mx and My must be calculated (see the chapter on the program with its algorithms.).
End Solution

up

A.7 Supersonic Branch

In Chapter (??) it was shown that the isothermal model cannot describe adequately the situation because the thermal entry length is relatively large compared to the pipe length

162

APPENDIX A. FANNO FLOW

and the heat transfer is not sucient to maintain constant temperature. In the Fanno model there is no heat transfer, and, furthermore, because the very limited amount of heat transformed it is closer to an adiabatic ow. The only limitation of the model is its uniform velocity (assuming parabolic ow for laminar and dierent prole for turbulent ow.). The information from the wall to the tube center6 is slower in reality. However, experiments from many starting with 1938 work by Frossel7 has shown that the error is not signicant. Nevertheless, the comparison with reality shows that heat transfer cause changes to the ow and they need/should to be expected. These changes include the choking point at lower Mach number.

A.8 Maximum Length for the Supersonic Flow

It has to be noted and recognized that as opposed to subsonic branch the supersonic branch has a limited length. It also must be recognized that there is a maximum length for which only supersonic ow can exist8 . These results were obtained from the mathematical derivations but were veried by numerous experiments9 . The maximum length of the supersonic can be evaluated when M = as follows:
k+1 2 4f Lmax 1 M2 k+1 2 M = + ln = D kM 2 2k 2 1 + k1 M 2 2 k + 1 (k + 1) 4f L D (M ) k + 2k ln (k 1) 1 k + 1 (k + 1) = + ln k 2k (k 1)

4f L D (M

, k = 1.4) = 0.8215

The maximum length of the supersonic ow is limited by the above number. From the above analysis, it can be observed that no matter how high the entrance Mach number will be the tube length is limited and depends only on specic heat ratio, k as shown in Figure (A.5).

A.9 Working Conditions

It has to be recognized that there are two regimes that can occur in Fanno ow model one of subsonic ow and the other supersonic ow. Even the ow in the tube starts as a supersonic in parts of the tube can be transformed into the subsonic branch. A shock wave can occur and some portions of the tube will be in a subsonic ow pattern.
6 The word information referred to is the shear stress transformed from the wall to the center of the tube. 7 See on the web http://naca.larc.nasa.gov/digidoc/report/tm/44/NACA-TM-844.PDF 8 Many in the industry have diculties in understanding this concept. The author seeks for a nice explanation of this concept for nonuid mechanics engineers. This solicitation is about how to explain this issue to non-engineers or engineer without a proper background. 9 If you have experiments demonstrating this point, please provide to the undersign so they can be added to this book. Many of the pictures in the literature carry copyright statements.

A.9. WORKING CONDITIONS

163

The maximum length in supersonic flow

In Fanno Flow 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1.2 1.25 1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 spesific heat, k Thu Mar 3 16:24:00 2005

Fig. -A.5. The maximum length as a function of specic heat, k

The discussion has to dierentiate between two ways of feeding the tube: converging nozzle or a converging-diverging nozzle. Three parameters, the dimensionless friction, 4f L , the entrance Mach number, M1 , and the pressure ratio, P2 /P1 are conD trolling the ow. Only a combination of these two parameters is truly independent. However, all the three parameters can be varied and they are discussed separately here.

A.9.1

Variations of The Tube Length ( 4f L ) Eects D

In the analysis of this eect, it should be assumed that back pressure is constant and/or low as possible as needed to maintain a choked ow. First, the treatment of the two branches are separated. Fanno Flow Subsonic branch For converging nozzle feeding, increasing the tube length results in increasing the exit Mach number (normally denoted herein as M2 ). Once the Mach number reaches maximum (M = 1), no further increase of the exit Mach number can be achieved. In this process, the mass ow rate decreases. It is worth noting that entrance Mach number is reduced (as some might explain it to reduce the ow rate). The entrance temperature increases as can be seen from Figure (A.7). The velocity therefore must decrease because the loss of the enthalpy (stagnation temperature) is used. The density decrease

4fL maximum length, max D

164

APPENDIX A. FANNO FLOW

Fig. -A.6. The eects of increase of

4f L D

on the Fanno line

P because = RT and when pressure is remains almost constant the density decreases. Thus, the mass ow rate must decrease. These results are applicable to the converging nozzle.

In the case of the convergingdiverging feeding nozzle, increase of the dimensionless friction, 4f L , results in a similar ow pattern as in the converging nozzle. Once D the ow becomes choked a dierent ow pattern emerges.

Fanno Flow Supersonic Branch There are several transitional points that change the pattern of the ow. Point a is the choking point (for the supersonic branch) in which the exit Mach number reaches to one. Point b is the maximum possible ow for supersonic ow and is not dependent on the nozzle. The next point, referred here as the critical point c, is the point in which no supersonic ow is possible in the tube i.e. the shock reaches to the nozzle. There is another point d, in which no supersonic ow is possible in the entire nozzletube system. Between these transitional points the eect parameters such as mass ow rate, entrance and exit Mach number are discussed. At the starting point the ow is choked in the nozzle, to achieve supersonic ow. The following ranges that has to be discussed includes (see Figure (A.8)):

A.9. WORKING CONDITIONS

constant pressure lines

165

1 1

1 2 2 2

Fanno lines

Fig. -A.7. The development properties in of converging nozzle

0
4f L D 4f L D 4f L D choking shockless chokeless

< < < <

4f L D 4f L D 4f L D 4f L D

< < < <

4f L D 4f L D 4f L D

choking shockless chokeless

0a ab bc c

all supersonic flow

mixed supersonic with subsonic flow with a shock between

the nozzle is still choked

Fig. -A.8. The Mach numbers at entrance and exit of tube and mass ow rate for Fanno Flow as a function of the 4f L . D

The 0-a range, the mass ow rate is constant because the ow is choked at the nozzle. The entrance Mach number, M1 is constant because it is a function of the nozzle design only. The exit Mach number, M2 decreases (remember this ow is on the supersonic branch) and starts ( 4f L = 0) as M2 = M1 . At the end of the range a, M2 = 1. In the D

166

APPENDIX A. FANNO FLOW

range of a b the ow is all supersonic. In the next range a b The ow is double choked and make the adjustment for the ow rate at dierent choking points by changing the shock location. The mass ow rate continues to be constant. The entrance Mach continues to be constant and exit Mach number is constant. The total maximum available for supersonic ow b b , 4f L , is only a D max theoretical length in which the supersonic ow can occur if nozzle is provided with a larger Mach number (a change to the nozzle area ratio which also reduces the mass ow rate). In the range b c, it is a more practical point. In semi supersonic ow b c (in which no supersonic is available in the tube but only in the nozzle) the ow is still double choked and the mass ow rate is constant. Notice that exit Mach number, M2 is still one. However, the entrance Mach number, M1 , reduces with the increase of 4f L . D It is worth noticing that in the a c the mass ow rate nozzle entrance velocity and the exit velocity remains constant!10 In the last range c the end is really the pressure limit or the break of the model and the isothermal model is more appropriate to describe the ow. In this range, the ow rate decreases since (m M1 )11 . To summarize the above discussion, Figures (A.8) exhibits the development of M1 , M2 mass ow rate as a function of 4f L . Somewhat dierent then the subsonic D branch the mass ow rate is constant even if the ow in the tube is completely subsonic. This situation is because of the double choked condition in the nozzle. The exit Mach M2 is a continuous monotonic function that decreases with 4f L . The entrance Mach D M1 is a non continuous function with a jump at the point when shock occurs at the entrance moves into the nozzle. Figure (A.9) exhibits the M1 as a function of M2 . The Figure was calculated by utilizing the data from Figure (A.2) by obtaining the 4f L for M2 and subtracting D the given 4f L and nding the corresponding M1 . D The Figure (A.10) exhibits the entrance Mach number as a function of the M2 . Obviously there can be two extreme possibilities for the subsonic exit branch. Subsonic velocity occurs for supersonic entrance velocity, one, when the shock wave occurs at the tube exit and two, at the tube entrance. In Figure (A.10) only for 4f L = 0.1 and D 4f L 4f L D = 0.4 two extremes are shown. For D = 0.2 shown with only shock at the exit only. Obviously, and as can be observed, the larger 4f L creates larger dierences D between exit Mach number for the dierent shock locations. The larger 4f L larger M1 D must occurs even for shock at the entrance. For a given 4f L , below the maximum critical length, the supersonic entrance ow D has three dierent regimes which depends on the back pressure. One, shockless ow,
10 On a personal note, this situation is rather strange to explain. On one hand, the resistance increases and on the other hand, the exit Mach number remains constant and equal to one. Does anyone have an explanation for this strange behavior suitable for nonengineers or engineers without background in uid mechanics? 11 Note that increases with decreases of M but this eect is less signicant. 1 1

max

A.9. WORKING CONDITIONS

167

Fanno Flow
M1 as a function of M2 1 0.9 0.8 0.7 Entrace Mach number 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 4fL = 0.1 D = 1.0 = 10.0 = 100.0

Exit Mach number Tue Oct 19 09:56:15 2004

Fig. -A.9. M1 as a function M2 for various
4f L D

tow, shock at the entrance, and three, shock at the exit. Below, the maximum critical length is mathematically 4f L 1 1+k k+1 > + ln D k 2k k1 For cases of 4f L above the maximum critical length no supersonic ow can be over the D whole tube and at some point a shock will occur and the ow becomes subsonic ow12 .

A.9.2

The Pressure Ratio,

P2 , P1

eects

In this section the studied parameter is the variation of the back pressure and thus, the pressure ratio P2 variations. For very low pressure ratio the ow can be assumed P1 as incompressible with exit Mach number smaller than < 0.3. As the pressure ratio
12 See

more on the discussion about changing the length of the tube.

168

APPENDIX A. FANNO FLOW

Fanno Flow
M1 as a function of M2 for the subsonic brench 5 4.5 4 3.5 3 M1 2.5 2 1.5 1 0.5 0 0 0.2 0.4 0.6 0.8 1 M2 1.2 1.4 1.6 1.8 2 4fL = 0.1 D = 0.2 = 0.4 = 0.1 shock = 0.4

Tue Jan 4 11:26:19 2005

Fig. -A.10. M1 as a function M2 for dierent
4f L D

for supersonic entrance velocity.

increases (smaller back pressure, P2 ), the exit and entrance Mach numbers increase. According to Fanno model the value of 4f L is constant (friction factor, f , is independent D of the parameters such as, Mach number, Reynolds number et cetera) thus the ow remains on the same Fanno line. For cases where the supply come from a reservoir with a constant pressure, the entrance pressure decreases as well because of the increase in the entrance Mach number (velocity). Again a dierentiation of the feeding is important to point out. If the feeding nozzle is converging than the ow will be only subsonic. If the nozzle is converging diverging than in some part supersonic ow is possible. At rst the converging nozzle is presented and later the converging-diverging nozzle is explained. Choking explanation for pressure variation/reduction Decreasing the pressure ratio or in actuality the back pressure, results in increase of the entrance and the exit velocity until a maximum is reached for the exit velocity.

A.9. WORKING CONDITIONS

169

a shock in the nozzle fully subsoinic flow

critical Point a

criticalPoint b

critical Point c

critical Point d

Fig. -A.11. The pressure distribution as a function of

4f L D

for a short

4f L D

The maximum velocity is when exit Mach number equals one. The Mach number, as it was shown in Chapter (??), can increases only if the area increase. In our model the tube area is postulated as a constant therefore the velocity cannot increase any further. However, for the ow to be continuous the pressure must decrease and for that the velocity must increase. Something must break since there are conicting demands and it result in a jump in the ow. This jump is referred to as a choked ow. Any additional reduction in the back pressure will not change the situation in the tube. The only change will be at tube surroundings which are irrelevant to this discussion. If the feeding nozzle is a convergingdiverging then it has to be dierentiated between two cases; One case is where the 4f L is short or equal to the critical length. The D critical length is the maximum Short
4f L D 4f L D max

that associate with entrance Mach number.

Figure (A.12) shows dierent pressure proles for dierent back pressures. Before the ow reaches critical point a (in the Figure) the ow is subsonic. Up to this stage the

170

APPENDIX A. FANNO FLOW

a shock in the nozzle fully subsoinic flow

critical Point a criticalPoint b critical Point c

Fig. -A.12. The pressure distribution as a function of

4f L D

for a long

4f L D

nozzle feeding the tube increases the mass ow rate (with decreasing back pressure). Between point a and point b the shock is in the nozzle. In this range and further reduction of the pressure the mass ow rate is constant no matter how low the back pressure is reduced. Once the back pressure is less than point b the supersonic reaches to the tube. Note however that exit Mach number, M2 < 1 and is not 1. A back pressure that is at the critical point c results in a shock wave that is at the exit. When the back pressure is below point c, the tube is clean of any shock13 . The back pressure below point c has some adjustment as it occurs with exceptions of point d. Long
4f L D 4f L D

In the case of

>

4f L D

max

reduction of the back pressure results in the same

4f L D

process as explained in the short

up to point c. However, point c in this case is

4f L D

dierent from point c at the case of short tube

13 It

<

4f L D

max

. In this point the

is common misconception that the back pressure has to be at point d.

A.9. WORKING CONDITIONS

171

Mach number in Fanno Flow

4fL D 2 1.8 1.6 1.4 Mach Number 1.2 1 0.8 0.6 0.4 0.2 0 0 0.05 0.1 0.15 4fL D 0.2 0.25

shock at
75% 50% 5%

Tue Jan 4 12:11:20 2005

Fig. -A.13. The eects of pressure variations on Mach number prole as a function of when the total resistance 4f L = 0.3 for Fanno Flow D

4f L D

exit Mach number is equal to 1 and the ow is double shock. Further reduction of the back pressure at this stage will not move the shock wave downstream the nozzle. At point c or location of the shock wave, is a function entrance Mach number, M1 and the extra 4f L . The is no analytical solution for the location of this point c. The D procedure is (will be) presented in later stage.

A.9.3

Entrance Mach number, M1 , eects

In this discussion, the eect of changing the throat area on the nozzle eciency is neglected. In reality these eects have signicance and needs to be accounted for some instances. This dissection deals only with the ow when it reaches the supersonic branch reached otherwise the ow is subsonic with regular eects. It is assumed that in this P2 discussion that the pressure ratio P1 is large enough to create a choked ow and 4f L D is small enough to allow it to happen. The entrance Mach number, M1 is a function of the ratio of the nozzles throat area to the nozzle exit area and its eciency. This eect is the third parameter discussed

172

APPENDIX A. FANNO FLOW

P2/P1 Fanno Flow

4fL D 4.8 4.4 4 3.6 3.2 P2/P1 2.8 2.4 2 1.6 1.2 0.8 0.4 0 0 0.05 0.1 0.15 4fL D 0.2 0.25 5% 50 % 75 %

Fri Nov 12 04:07:34 2004

Fig. -A.14. Mach number as a function of
4f L D

when the total

4f L D

= 0.3

here. Practically, the nozzle area ratio is changed by changing the throat area. As was shown before, there are two dierent maximums for 4f L ; rst is the total D maximum 4f L of the supersonic which depends only on the specic heat, k, and second D the maximum depends on the entrance Mach number, M1 . This analysis deals with the case where 4f L is shorter than total 4f L . D D
max

Obviously, in this situation, the critical point is where 4f L is equal to 4f L D D max as a result in the entrance Mach number. The process of decreasing the convergingdiverging nozzles throat increases the entrance14 Mach number. If the tube contains no supersonic ow then reducing the nozzle throat area wouldnt increase the entrance Mach number. This part is for the case where some part of the tube is under supersonic regime and there is shock as a transition to subsonic branch. Decreasing the nozzle throat area
14 The word entrance referred to the tube and not to the nozzle. The reference to the tube is because it is the focus of the study.

A.9. WORKING CONDITIONS

173

shock

Fig. -A.15. Schematic of a long tube in supersonic branch

moves the shock location downstream. The payment for increase in the supersonic length is by reducing the mass ow. Further, decrease of the throat area results in ushing the shock out of the tube. By doing so, the throat area decreases. The mass ow rate is proportionally linear to the throat area and therefore the mass ow rate reduces. The process of decreasing the throat area also results in increasing the pressure drop of the nozzle (larger resistance in the nozzle15 )16 . In the case of large tube 4f L > 4f L the exit Mach number increases with the D D max decrease of the throat area. Once the exit Mach number reaches one no further increases is possible. However, the location of the shock wave approaches to the theoretical location if entrance Mach, M1 = . The maximum location of the shock The main point in this discussion however, is to nd the furthest shock location downstream. Figure (A.16) shows the possible 4f L as function of retreat of the location of the shock wave from the maximum D location. When the entrance Mach number is innity, M1 = , if the shock location is at the maximum length, then shock at Mx = 1 results in My = 1. The proposed procedure is based on Figure (A.16). i) Calculate the extra 4f L and subtract the actual extra D the left side (at the max length). ii) Calculate the extra 4f L and subtract the actual extra D the right side (at the entrance).
4f L D

assuming shock at assuming shock at

4f L D

iii) According to the positive or negative utilizes your root nding procedure.
Frictionless nozzle has a larger resistance when the throat area decreases is one of the strange phenomenon that in one way increasing the resistance (changing the throat area) decreases the ow rate while in a dierent way (increasing the 4f L ) does not aect the ow D rate.
16 It 15 Strange?

174

APPENDIX A. FANNO FLOW

4f L D

Fig. -A.16. The extra tube length as a function of the shock location,

supersonic branch

From numerical point of view, the Mach number equal innity when left side assumes result in innity length of possible extra (the whole ow in the tube is subsonic). To overcome this numerical problem it is suggested to start the calculation from distance from the right hand side. Let denote 4f L 4f L 4f L = (A.51) D D actual D sup Note that
4f L D sup

is smaller than

4f L D

max

. The requirement that has to be satis-

ed is that denote 4f L as dierence between the maximum possible of length D retreat in which the supersonic ow is achieved and the actual length in which the ow is supersonic see Figure (A.15). The retreating length is expressed as subsonic but 4f L D =
retreat

4f L D

max

4f L D

(A.52)
sup

Figure (A.17) shows the entrance Mach number, M1 reduces after the maximum length is exceeded. Example A.3: Calculate the shock location for entrance Mach number M1 = 8 and for assume that k = 1.4 (Mexit = 1).
4f L D

= 0.9

A.9. WORKING CONDITIONS

175

4f L D max

4f L D

Fig. -A.17. The maximum entrance Mach number, M1 to the tube as a function of supersonic branch

Solution The solution is obtained by an iterative process. The maximum

4f L D 4f L D 4f L D max

for k =

exceed the maximum length for this entrance 1.4 is 0.821508116. Hence, 4f L Mach number. The maximum for M1 = 8 is D = 0.76820, thus the extra tube is
4f L D 4f L D

= 0.9 0.76820 = 0.1318. The left side is when the shock occurs at

= 0.76820 (ow is choked and no additional 4f L ). Hence, the value of left side is D 0.1318. The right side is when the shock is at the entrance at which the extra 4f L is D calculated for Mx and My is Mx 8.0000 With (M1 ) M 0.39289 The extra
4fL D P P P0 P0 U U T T

My 0.39289

Ty Tx

y x

Py Px

P0y P0 x

13.3867

5.5652

74.5000

0.00849

2.4417
4f L D

2.7461

1.6136

2.3591

0.42390

1.1641

is 2.442 0.1318 = 2.3102 Now the solution is somewhere

176

APPENDIX A. FANNO FLOW

between the negative of left side to the positive of the right side17 . In a summary of the actions is done by the following algorithm: (a) check if the 4f L exceeds the maximum D cordingly continue.
4f L D up 4f L D 4f L D 4f L D max

for the supersonic ow. Ac-

(b) Guess

max

(c) Calculate the Mach number corresponding to the current guess of

4f L D up ,

(d) Calculate the associate Mach number, Mx with the Mach number, My calculated previously, (e) Calculate
4f L D

for supersonic branch for the Mx

4f L D up

(f) Calculate the new and improved

(g) Compute the new

4f L D down

4f L D

4f L D up

(h) Check the new and improved stop or return to stage (b). Shock location are:

4f L D

down

against the old one. If it is satisfactory

M1 8.0000

M2 1.0000

4fL D up

4fL D down

Mx 1.6706

My 0.64830

0.57068

0.32932

The iteration summary is also shown below

17 What if the right side is also negative? The ow is chocked and shock must occur in the nozzle before entering the tube. Or in a very long tube the whole ow will be subsonic.

A.10.

PRACTICAL EXAMPLES FOR SUBSONIC FLOW i 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

4fL D up 4fL D down

177 My
4fL D

Mx 1.3838 1.5286 1.6021 1.6382 1.6554 1.6635 1.6673 1.6691 1.6699 1.6703 1.6705 1.6706 1.6706 1.6706 1.6706 1.6706 1.6706 1.6706

0.67426 0.62170 0.59506 0.58217 0.57605 0.57318 0.57184 0.57122 0.57093 0.57079 0.57073 0.57070 0.57069 0.57068 0.57068 0.57068 0.57068 0.57068

0.22574 0.27830 0.30494 0.31783 0.32395 0.32682 0.32816 0.32878 0.32907 0.32921 0.32927 0.32930 0.32931 0.32932 0.32932 0.32932 0.32932 0.32932

0.74664 0.69119 0.66779 0.65728 0.65246 0.65023 0.64920 0.64872 0.64850 0.64839 0.64834 0.64832 0.64831 0.64831 0.64830 0.64830 0.64830 0.64830

0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000 0.90000

This procedure rapidly converted to the solution.

End Solution

A.10 The Practical Questions and Examples of Subsonic branch

The Fanno is applicable also when the ow isnt choke18 . In this case, several questions appear for the subsonic branch. This is the area shown in Figure (A.8) in beginning for between points 0 and a. This kind of questions made of pair given information to nd the conditions of the ow, as oppose to only one piece of information given in choked ow. There many combinations that can appear in this situation but there are several more physical and practical that will be discussed here.

18 This questions were raised from many who didnt nd any book that discuss these practical aspects and send questions to this author.

178

APPENDIX A. FANNO FLOW

A.10.1

Subsonic Fanno Flow for Given

4f L D

and Pressure Ratio

This pair of parameters is the most M2 4f L M1 4f L natural to examine because, in most P1 P2 D P = P D cases, this information is the only inM =1 formation that is provided. For a hypothetical section 4f L given pipe , neither the enD trance Mach number nor the exit Fig. -A.18. Unchoked ow calculations showing the Mach number are given (sometimes hypothetical full tube when choked the entrance Mach number is give see the next section). There is no exact analytical solution. There are two possible approaches to solve this problem: one, by building a representative function and nd a root (or roots) of this representative function. Two, the problem can be solved by an iterative procedure. The rst approach require using root nding method and either method of spline method or the half method found to be good. However, this author experience show that these methods in this case were found to be relatively slow. The NewtonRapson method is much faster but not were found to be unstable (at lease in the way that was implemented by this author). The iterative method used to solve constructed on the properties of several physical quantities must be in a certain range. The rst fact is that the pressure ratio P2 /P1 is always between 0 and 1 (see Figure A.18). In the gure, a theoretical extra tube is added in such a length that cause the ow to choke (if it really was there). This length is always positive (at minimum is zero). The procedure for the calculations is as the following: 1) Calculate the entrance Mach number, M1 assuming the (chocked ow);
4f L D

4f L D

max

2) Calculate the minimum pressure ratio (P2 /P1 )min for M1 (look at table (A.1)) 3) Check if the ow is choked: There are two possibilities to check it. a) Check if the given
4f L D

is smaller than

4f L D

obtained from the given P1 /P2 , or

b) check if the (P2 /P1 )min is larger than (P2 /P1 ), continue if the criteria is satised. Or if not satised abort this procedure and continue to calculation for choked ow. 4) Calculate the M2 based on the (P /P2 ) = (P1 /P2 ), 5) calculate 4f L based on M2 , D 6) calculate the new (P2 /P1 ), based on the new f (remember that 4f L = D
4f L D 2 4f L D 1

4f L D

),

A.10.

PRACTICAL EXAMPLES FOR SUBSONIC FLOW

179

7) calculate the corresponding M1 and M2 , 8) calculate the new and improve the 4f L by D 4f L D 4f L D
P2 P1 P2 P1

=
new

old

given old

(A.53)

Note, when the pressure ratios are matching also the

4f L D

will also match.

9) Calculate the improved/new M2 based on the improve 4f L D 10) calculate the improved
4f L D

as

4f L D

4f L D

given

+
4f L D .

4f L D

new

11) calculate the improved M1 based on the improved

12) Compare the abs ((P2 /P1 )new (P2 /P1 )old ) and if not satised returned to stage (6) until the solution is obtained. To demonstrate how this procedure is working consider a typical example of 4f L = D 1.7 and P2 /P1 = 0.5. Using the above algorithm the results are exhibited in the following gure. Figure (A.19) demonstrates that the conversion occur at about 7-8

3.0

Conversion occurs around 7-9 times

2.5

M1 M2
4f L D

P2/P1
2.0

4f L D

1.5

1.0

0.5

10

20

30

40

50

60

70

80

90

100

110

120

130

Number of Iterations, i

October 8, 2007

Fig. -A.19. The results of the algorithm showing the conversion rate for unchoked Fanno ow model with a given 4f L and pressure ratio. D

iterations. With better rst guess this conversion procedure will converts much faster (under construction).

180

APPENDIX A. FANNO FLOW

A.10.2

Subsonic Fanno Flow for a Given M1 and Pressure Ratio

This situation pose a simple mathematical problem while the physical situation occurs in cases where a specic ow rate is required with a given pressure ratio (range) (this problem was considered by some to be somewhat complicated). The specic ow rate can be converted to entrance Mach number and this simplies the problem. Thus, the problem is reduced to nd for given entrance Mach, M1 , and given pressure ratio calculate the ow parameters, like the exit Mach number, M2 . The procedure is based on the fact that the entrance star pressure ratio can be calculated using M1 . Thus, using the pressure ratio to calculate the star exit pressure ratio provide the exit Mach number, M2 . An example of such issue is the following example that combines also the Naughty professor problems. Example A.4: Calculate the exit Mach number for P2 /P 1 = 0.4 and entrance Mach number M1 = 0.25. Solution The star pressure can be obtained from a table or Potto-GDC as M
4fL D P P P0 P0 U U T T

0.25000 8.4834

4.3546

2.4027

3.6742

0.27217 1.1852

And the star pressure ratio can be calculated at the exit as following P2 P2 P1 = = 0.4 4.3546 = 1.74184 P P1 P And the corresponding exit Mach number for this pressure ratio reads M
4fL D P P P0 P0 U U T T

0.60694 0.46408 1.7418

1.1801

1.5585

0.64165 1.1177

A bit show o the PottoGDC can carry these calculations in one click as M1 0.25000 M2 0.60693
End Solution

4fL D

P2 P1

8.0193

0.40000

While the above example show the most simple from of this question, in reality this question is more complicated. One common problem is situation that the diameter is not given but the ow rate and length and pressure (stagnation or static) with some combination of the temperature. The following example deal with one of such example.

A.10.

PRACTICAL EXAMPLES FOR SUBSONIC FLOW

181

Example A.5: A tank lled with air at stagnation pressure, 2[Bar] should be connected to a pipe with a friction factor, f = 0.005, and and length of 5[m]. The ow rate is (should be) kg 0.1 sec and the static temperature at the entrance of the pipe was measured to be 27 C. The pressure ratio P2 /P1 should not fall below 0.9 (P2 /P1 > 0.9). Calculate the exit Mach number, M2 , ow rate, and minimum pipe diameter. You can assume that k = 1.4. Solution The direct mathematical solution isnt possible and some kind of iteration procedure or root nding for a representative function. For the rst part the naughty professor procedure cannot be used because m/A is not provided and the other hand 4f L is D not provided (missing Diameter). One possible solution is to guess the entrance Mach and check whether and the mass ow rate with the naughty professor procedure are satised. For Fanno ow at for several Mach numbers the following is obtained M1 0.10000 0.15000 0.20000 M2 0.11109 0.16658 0.22202
4fL D P2 P1

Diameter 0.00748 0.01716 0.03136

13.3648 5.8260 3.1887

0.90000 0.90000 0.90000

From the last table the diameter can be calculated for example for M1 = 0.2 as D= 4f L
4f L D

= 4 0.005 5/3.1887 = 0.03136[m]

The same was done for all the other Mach number. Now the area can be calculated and therefor the m/A can be calculated. With this information the naughty professor is given and the entrance Mach number can be calculated. For example for M1 = 0.2 one can obtain the following: m/A = 0.1/( 0.031362 /4) 129.4666798 The same order as the above table it shown in naughty professor (isentropic table). M
T T0 0 A A P P0 AP A P0 F F

1.5781 0.66752 0.36404 1.2329 0.36221 0.97443 0.93730 1.7268 0.10979 0.99760 0.99400 5.3092

0.24300 0.29960 0.56009 0.91334 1.5772 0.77785 0.99161 5.2647 2.2306

The rst result are not reasonable and this process can continue until the satisfactory solution is achieved. Here an graphical approximation is shown. From this exhibit it can be estimated that M1 = 0.18. For this Mach number the following can be obtained

182
0.4

APPENDIX A. FANNO FLOW

guessed M1 calculated M1
0.3

Entrace Mach Nubmer

0.2

0.1

Solution
0.1 0.15 0.2 0.25 0.3

Conversion of the guesing the Mach Number

October 18, 2007

Fig. -A.20. Diagram for nding solution when the pressure ratio and entrance properties (T and P0 are given

M1 0.18000

M2 0.19985

4fL D

P2 P1

3.9839

0.90000

Thus, the diameter can be obtained as D 0.0251[m] The ow rate is m/A 202.1[kg/sec m2 ] M
T T0 0 A A P P0 AP A P0 F F

0.17109 0.99418 0.98551 3.4422

0.97978 3.3726

1.4628

The exact solution is between 0.17 to 0.18 if better accuracy is needed.

End Solution

A.11

The Approximation of the Fanno Flow by Isothermal Flow

The isothermal ow model has equations that theoreticians nd easier to use and to compare to the Fanno ow model. One must notice that the maximum temperature at the entrance is T0 1 . When the Mach number decreases the temperature approaches the stagnation temperature (T T0 ). Hence, if one allows certain deviation of temperature, say about 1% that ow can be assumed to be isothermal. This tolerance requires that (T0 T )/T0 = 0.99 which requires that enough for M1 < 0.15 even for large k = 1.67. This requirement provides that somewhere (depend) in the vicinity of 4f L = 25 the ow can be assumed D isothermal. Hence the mass ow rate is a function of 4f L because M1 changes. Looking D at the table or Figure (A.2) or the results from PottoGDC attached to this book shows that reduction of the mass ow is very rapid. As it can be seen for the Figure (A.21) the dominating parameter is 4f L . The results are very similar for isothermal ow. The D only dierence is in small dimensionless friction, 4f L . D

A.12.

MORE EXAMPLES OF FANNO FLOW

183

M1 Fanno flow
with comperison to Isothermal Flow

0.4 P2 / P1 P2 / P1 P2 / P1 P2 / P1 P2 / P1 P2 / P1 = 0.1 iso = 0.8 iso = 0.1 = 0.2 = 0.5 = 0.8

0.3 M1 0.2 0.1 0 0

10

20

30

40

Wed Mar 9 11:38:27 2005

50 4fL D

60

70

80

90

100

Fig. -A.21. The entrance Mach number as a function of dimensionless resistance and comparison with Isothermal Flow

A.12

More Examples of Fanno Flow

Example A.6: To demonstrate the utility in Figure (A.21) consider the following example. Find the mass ow rate for f = 0.05, L = 4[m], D = 0.02[m] and pressure ratio P2 /P1 = 0.1, 0.3, 0.5, 0.8. The stagnation conditions at the entrance are 300K and 3[bar] air. Solution First calculate the dimensionless resistance,
4f L D .

4f L 4 0.05 4 = = 40 D 0.02 From Figure (A.21) for P2 /P1 = 0.1 M1 0.13 etc. or accurately by utilizing the program as in the following table.

184 M1 0.12728 0.12420 0.11392 0.07975 M2 1.0000 0.40790 0.22697 0.09965

4fL D 4fL D 1

APPENDIX A. FANNO FLOW

4fL D 2 P2 P1

40.0000 40.0000 40.0000 40.0000

40.0000 42.1697 50.7569 107.42

0.0 2.1697 10.7569 67.4206

0.11637 0.30000 0.50000 0.80000

Only for the pressure ratio of 0.1 the ow is choked. M 0.12728 0.12420 0.11392 0.07975
T T0 0 A A P P0 AP A P0

0.99677 0.99692 0.99741 0.99873

0.99195 0.99233 0.99354 0.99683

4.5910 4.7027 5.1196 7.2842

0.98874 0.98928 0.99097 0.99556

4.5393 4.6523 5.0733 7.2519

Therefore, T T0 and is the same for the pressure. Hence, the mass rate is a function of the Mach number. The Mach number is indeed a function of the pressure ratio but mass ow rate is a function of pressure ratio only through Mach number. The mass ow rate is m = P AM k 0.022 = 300000 0.127 RT 4 1.4 0.48 287300 kg sec

and for the rest P2 0.1242 = 0.3 0.48 = 0.468 P1 0.1273 P2 0.1139 m = 0.5 0.48 = 0.43 P1 0.1273 P2 0.07975 m = 0.8 0.48 = 0.30 P1 0.1273 m
End Solution

kg sec kg sec kg sec

A.13

The Table for Fanno Flow

Table -A.1. Fanno Flow Standard basic Table

M 0.03 0.04

4fL D

P P

P0 P0

U U

T T

787.08 440.35

36.5116 27.3817

19.3005 14.4815

30.4318 22.8254

0.03286 1.1998 0.04381 1.1996

A.13.

THE TABLE FOR FANNO FLOW

Table -A.1. Fanno Flow Standard basic Table (continue)

185

M 0.05 0.06 0.07 0.08 0.09 0.10 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00

4fL D

P P

P0 P0

U U

T T

280.02 193.03 140.66 106.72 83.4961 66.9216 14.5333 8.4834 5.2993 3.4525 2.3085 1.5664 1.0691 0.72805 0.49082 0.32459 0.20814 0.12728 0.07229 0.03633 0.01451 0.00328 0.0 0.30500 0.52216 0.63306 0.69380 0.72988 0.75280 0.76819 0.77899 0.78683 0.81265 0.81582 0.81755 0.81860 0.81928 0.81975 0.82008

21.9034 11.5914 18.2508 9.6659 15.6416 8.2915 13.6843 7.2616 12.1618 6.4613 10.9435 5.8218 5.4554 2.9635 4.3546 2.4027 3.6191 2.0351 3.0922 1.7780 2.6958 1.5901 2.3865 1.4487 2.1381 1.3398 1.9341 1.2549 1.7634 1.1882 1.6183 1.1356 1.4935 1.0944 1.3848 1.0624 1.2893 1.0382 1.2047 1.0207 1.1291 1.0089 1.061 1.002 1.00000 1.000 0.40825 1.688 0.21822 4.235 0.13363 10.72 0.089443 25.00 0.063758 53.18 0.047619 1.0E+2 0.036860 1.9E+2 0.029348 3.3E+2 0.023905 5.4E+2 0.00609 1.5E+4 0.00390 4.6E+4 0.00271 1.1E+5 0.00200 2.5E+5 0.00153 4.8E+5 0.00121 8.6E+5 0.000979 1.5E+6

18.2620 15.2200 13.0474 11.4182 10.1512 9.1378 4.5826 3.6742 3.0702 2.6400 2.3184 2.0693 1.8708 1.7092 1.5753 1.4626 1.3665 1.2838 1.2119 1.1489 1.0934 1.044 1.000 0.61237 0.50918 0.46771 0.44721 0.43568 0.42857 0.42390 0.42066 0.41833 0.41079 0.40988 0.40938 0.40908 0.40889 0.40875 0.40866

0.05476 0.06570 0.07664 0.08758 0.09851 0.10944 0.21822 0.27217 0.32572 0.37879 0.43133 0.48326 0.53452 0.58506 0.63481 0.68374 0.73179 0.77894 0.82514 0.87037 0.91460 0.95781 1.00 1.633 1.964 2.138 2.236 2.295 2.333 2.359 2.377 2.390 2.434 2.440 2.443 2.445 2.446 2.446 2.447

1.1994 1.1991 1.1988 1.1985 1.1981 1.1976 1.1905 1.1852 1.1788 1.1713 1.1628 1.1533 1.1429 1.1315 1.1194 1.1065 1.0929 1.0787 1.0638 1.0485 1.0327 1.017 1.000 0.66667 0.42857 0.28571 0.20000 0.14634 0.11111 0.086957 0.069767 0.057143 0.014815 0.00952 0.00663 0.00488 0.00374 0.00296 0.00240

186

APPENDIX A. FANNO FLOW

Table -A.1. Fanno Flow Standard basic Table (continue)

M 55.00 60.00 65.00 70.00

4fL D

P P

P0 P0

U U

T T

0.82033 0.82052 0.82066 0.82078

0.000809 0.000680 0.000579 0.000500

2.3E+6 3.6E+6 5.4E+6 7.8E+6

0.40859 0.40853 0.40849 0.40846

2.447 2.448 2.448 2.448

0.00198 0.00166 0.00142 0.00122

A.14

Appendix Reynolds Number Eects

Almost Constant Zone

Constant Zone

Linear Representation Zone

Small Error Due to Linear Assumption

Fig. -A.22. Moody diagram on the name Moody who netscaped H. Rouse work to claim as his own. In this section the turbulent area is divided into 3 zones, constant, semiconstant, and linear After S Beck and R. Collins.

The friction factor in equation (A.25) was assumed constant. In Chapter ?? it was shown that the Reynolds number remains constant for ideal gas uid. However, in Fanno ow the temperature does not remain constant hence, as it was discussed before, the Reynolds number is increasing. Thus, the friction decreases with the exception of the switch in the ow pattern (laminar to turbulent ow). For relatively large relative roughness larger /D > 0.004 of 0.4% the friction factor is constant. For smother pipe /D < 0.001 and Reynolds number between 10,000 to a million the friction factor vary between 0.007 to 0.003 with is about factor of two. Thus, the error of 4f L is D limited by a factor of two (2). For this range, the friction factor can be estimated as a linear function of the log10 (Re). The error in this assumption is probably small of the assumption that involve in fanno ow model construction. Hence, f = A log10 (Re) + B (A.54)

A.14.

APPENDIX REYNOLDS NUMBER EFFECTS

187

Where the constant A and B are function of the relative roughness. For most practical purposes the slop coecient A can be further assumed constant. The slop coecient A = 0.998125 Thus, to carry this calculation relationship between the viscosity and the temperature. If the viscosity expanded as Taylor or Maclaren series then A1 T = A0 + + 1 T0 Where 1 is the viscosity at the entrance temperature T1 . Thus, Reynolds number is Re = DU 1 A0 + AT0T + DU A0 +
A1 T2 T1

(A.55)

(A.56)

Substituting equation (A.56) into equation (A.54) yield f = A log10 + +B (A.57)

Left hand side of equation (A.25) is a function of the Mach number since it contains the temperature. If the temperature functionality will not vary similarly to the case of constant friction factor then the temperature can be expressed using equation (A.41).
constant

4 A log10 D

A0 + A1

DU 1 + k1 M1 2 2 1+
2 k1 2 M2

+ B

(A.58)

Equation (A.58) is only estimate of the functionally however, this estimate is almost as good as the assumptions of Fanno ow. Equation fanno:eq:d2 can be improved by using equation (A.58)
constant

4 Lmax D

A log10

1 1 M2 k+1 2 k+1 DU 2 M + ln + B 2k k M2 1 + k1 M 2 1 + k1 M 2 2 2 A0 + A1 1 + k1 2 (A.59)

In the most complicate case where the ow pattern is change from laminar ow to turbulent ow the whole Fanno ow model is questionable and will produce poor results. In summary, in the literature there are three approaches to this issue of non constant friction factor. The friction potential is recommended by a researcher in Germany and it is complicated. The second method substituting this physical approach with numerical iteration. In the numerical iteration method, the expression of the various relationships are inserted into governing dierential equations. The numerical methods does not allow exibility and is very complicated. The methods described here can be expended (if really really needed) and it will be done in very few iteration as it was shown in the Isothermal Chapter.

188

APPENDIX A. FANNO FLOW

APPENDIX B What The Establishments Scientists Say

What a Chutzpah? to say samething like that!

anonymous In this section exhibits the establishment experts reaction the position that the common pQ2 diagram is improper. Their comments are responses to the authors paper: The mathematical theory of the pQ2 diagram (similar to Chapter 7)1 . The paper was submitted to Journal of Manufacturing Science and Engineering. This part is for the Associate Technical Editor Dr. R. E. Smelser. I am sure that you are proud of the referees that you have chosen and that you do not have any objection whatsoever with publishing this information. Please send a copy of this appendix to the referees. I will be glad to hear from them. This concludes comments to the Editor. I believe that you, the reader should judge if the mathematical theory of the pQ2 diagram is correct or whether the experts position is reasonable. For the reader unfamiliar with the journal review process, the associate editor sent the paper to readers (referees) which are anonymous to the authors. They comment on the paper and according to these experts the paper acceptance is determined. I have chosen the unusual step to publish their comments because I believe that other motivations are involved in their responses. Coupled with the response to the publication of a summary of this
1 The exact paper can be obtain free of charge from Minnesota Suppercomputing Institute, http://www2.msi.umn.edu/publications.html report number 99/40 The mathematical theory of the pQ2 diagram or by writing to the Supercomputing Institute, University of Minnesota, 1200 Washington Avenue South, Minneapolis, MN 55415-1227

189

190

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY

paper in the Die Casting Engineer, bring me to think that the best way to remove the information blockage is to open it to the public. Here, the referees can react to this rebuttal and stay anonymous via correspondence with the associated editor. If the referee/s choose to respond to the rebuttal, their comments will appear in the future additions. I will help them as much as I can to show their opinions. I am sure that they are proud of their criticism and are standing behind it 100%. Furthermore, I am absolutely, positively sure that they are so proud of their criticism they glad that it appears in publication such as this book.

B.1 Summary of Referee positions

The critics attack the article in three dierent ways. All the referees try to deny publication based on grammar!! The rst referee didnt show any English mistakes (though he alleged that he did). The second referee had some hand written notes on the preprint (two dierent hand writing?) but it is not the grammar but the content of the article (the fact that the common pQ2 diagram is wrong) is the problem. Here is an original segment from the submitted paper: The design process is considered an art for the 8billiondollar die casting industry. The pQ2 diagram is the most common calculation, if any that all, are used by most die casting engineers. The importance of this diagram can be demonstrated by the fact that tens of millions of dollars have been invested by NADCA, NSF, and other major institutes here and abroad in pQ2 diagram research. In order to correct grammar, the referee change to: The pQ2 diagram is the most common calculation used by die casting engineers to determine the relationship between the die casting machine and gating design parameters, and the resulting metal ow rate. It seems, the referee would not like some facts to be written/known. Summary of the referees positions: Referee 1 Well, the paper was published before (NADCA die casting engineer) and the errors in the common pQ2 are only in extreme cases. Furthermore, it actually supports the common model. Referee 2 Very angrily!! How dare the authors say that the common model is wrong. When in fact, according to him, it is very useful. Referee 3 The bizzarro approach! Changed the meaning of what the authors wrote (see the ovaled boxed comment for example). This produced a new type logic which is almost absurd. Namely, the discharge coecient, CD , is constant for a runner or can only vary with time. The third possibility, which is the topic of the paper, the fact that CD cannot be assigned a runner system but have to calculated for every set of runner and die casting machine can not exist possibility, and therefore the whole paper is irrelevant.

B.2. REFEREE 1 (FROM HAND WRITTEN NOTES) Genick BarMeirs answer:

191

Let me say what a smart man once said before: I dont need 2000 scientists to tell me that I am wrong. What I need is one scientist to show what is wrong in my theory.
Please read my rebuttal to the points the referees made. The referees version are kept as close as possible to the original. I put some corrections in a square bracket [] to clarify the referees point. Referee comments appear in roman font like this sentence, and rebuttals appear in a courier font as this sentence.

B.2 Referee 1 (from hand written notes)

1. Some awkward grammar See highlighted portions Where? 2. Similarity of the submitted manuscript to the attached Die Casting Engineer Trade journal article (May/June 1998) is Striking. The article in Die Casting Engineer is a summary of the present article. It is mentioned there that it is a summary of the present article. There is nothing secret about it. This article points out that the common model is totally wrong. This is of central importance to die casting engineers. The publication of this information cannot be delayed until the review process is finished. 3. It is not clear to the reader why the constant pressure and constant power situations were specically chosen to demonstrate the authors point. Which situation is most like that found [likely found] in a die casting machine? Does the constant pressure correspond closely to older style machines when intensifyer [intensier] bottle pressure was applied to the injection system unthrottled? Does the constant power situations assume a newer machines, such as Buher Sc, that generates the pressure required to achieve a desired gate velocity? Some explanation of the logic of selecting these two situations would be helpful in the manuscript. As was stated in the article, these situations were chosen because they are building blocks but more importantly to demonstrate that the common model is totally wrong! If it is wrong for two basic cases it should be absolutely wrong in

192

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY any combinations of the two cases. explanation is given in Chapter 7. Nevertheless, an additional

4. The authors approach is useful? Gives perspective to a commonly used process engineering method (pQ2 ) in die casting. Some of the runner lengths chosen (1 meter) might be consider exceptional in die casting yet the author uses this to show how much in error an average value for CD be. The author might also note that the North American Die Casting Association and many practitioners use a A3 /A2 ratio of .65 as a design target for gating. The author analysis reinforces this value as a good target, and that straying far from it may results in poor design part lling problems (Fig. 5) The reviewer refers to several points which are important to address. All the four sizes show large errors (we do not need to take 1[m] to demonstrate that). The one size, the referee referred to as exceptional (1 meter), is not the actual length but the represented length (read the article again). Poor design can be represented by a large length. This situation can be found throughout the die casting industry due to the common model which does not consider runner design. My office is full with runner designs with represent 1 meter length such as one which got NADCAs design award2 . In regards to the area ratio, please compare with the other referee who claim A2 /A2 = 0.8 - 0.95. I am not sure which of you really represent NADCAs position (I didnt find any of NADCAs publication in regards to this point). I do not agree with both referees. This value has to be calculated and cannot be speculated as the referees asserted. Please find an explanation to this point in the paper or in even better in Chapter 7.

B.3 Referee 2
There are several major concerns I have about this paper. The [most] major one [of these] is that [it] is unclear what the paper is attempting to accomplish. Is the paper trying to suggest a new way of designing the rigging for a die casting, or is it trying to add an improvement to the conventional pQ2 solution, or is it trying to somehow suggest a mathematical basis for the pQ2 diagram ? The paper shows that 1) the conventional pQ2 solution is totally wrong, 2) the mathematical analytical solution for the pQ2 provides an excellent tool for studying the effect of various parameters.
2 to

the best of my understanding

B.3. REFEREE 2

193

The other major concern is the poor organization of the ideas which the authors [are] trying to present. For instance, it is unclear how specic results presented in the results section where obtained ([for instance] how were the results in Figures 5 and 6 calculated?). I do not understand how the organization of the paper relates to the fact that the referee does not understand how Fig 5 and 6 were calculated. The organization of the paper does not have anything do with his understanding the concepts presented. In regard to the understanding of how Figure 5 and 6 were obtained, the referee should referred to an elementary fluid mechanics text book and combined it with the explanation presented in the paper. Several specic comments are written on the manuscript itself; most of these were areas where the reviewer was unclear on what the authors meant or areas where further discussion was necessary. One issue that is particularly irksome is the authors tendency in sections 1 and 2 to wonder [wander] o with editorials and other unsupported comments which have no place in a technical article. Please show me even one unsupported comment!! Other comments/concerns include what does the title have to do with the paper? The paper does not dene what a pQ2 diagram is and the results dont really tie in with the use of such diagrams.

The paper presents the exact analytical solution for the pQ2 diagram. The results tie in very well with the correct pQ2 diagram. Unfortunately, the common model is incorrect and so the results cannot be tied in with it. p.4 The relationship Q P is a result of the application of Bernoullis equation system like that shown in Fig 1. What is the rational or basis behind equation 1; e.g. Q (1 P )n with n =1, 1/2, and 1/4? Here I must thank the referee for his comment! If the referee had serious problem understanding this point, I probably should have considered adding a discussion about this point, such as in Chapter 7.
p.5 The relationship between equation 1(a) to 1(c) and a die casting machine as poor, common, and excellent performance is not clear and needs to be developed, or at least dened.

see previous comment

It is well known that CD for a die casting machine and die is not a constant. In fact it is common practice to experimentally determine CD for use on dies

194

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY with similar gating layouts in the future. But because most dies have numerous gates branching o of numerous runners, to determine all of the friction factors as a function of Reynolds number would be quite dicult and virtually untractable for design purposes. Generally die casting engineers nd conventional pQ2 approach works quite well for design purposes. This several points comment give me the opportunity to discusses the following points: I would kindly ask the referee, to please provide the names of any companies whom experimentally determine CD . Perhaps they do it down under (Australia) where the regular physics laws do not apply (please, forgive me about being the cynical about this point. I cannot react to this any other way.). Please, show me a company that uses the common pQ2 diagram and it works. Due to the computer revolution, today it is possible to do the calculations of the CD for a specific design with a specific flow rate (die casting machine). In fact, this is exactly what this paper all about. Moreover, today there is a program that already does these kind of calculations, called DiePerfect. Here the referee introduce a new idea of the family -the improved constant CD . In essence, the idea of family is improve constant CD in which one assigned value to a specific group of runners. Since this idea violate the basic physics laws and the produces the opposite to realty trends it must be abandoned. Actually, the idea of family is rather bizarre, because a change in the design can lead to a significant change in the value of CD . Furthermore, the family concept can lead to a poor design (read about this in the section poor design effects of this book). How one can decided which design is part of what family? Even if there were no mistakes, the authors method (calculating the CD ) is of course cheaper and faster than the referees suggestion about family of runner design. In summary, this idea a very bad idea. What is CD =constant? The referee refers to the case where CD is constant for specific runner design but which is not the case in reality. The CD does not depend only on the runner, but on the combination of the runner system with the die casting machine via the Re number. Thus, a specific runner design cannot have CD assigned to it.

B.4. REFEREE 3 The CD has to be calculated for any combination runner system with die casting machine. I would like to find any case where the common pQ2 diagram does work. Please read the proofs in Chapter 7 showing why it cannot work.
Discussion and results A great deal of discussion focuses on the regions where A3 /A2 0.1; yet in typical die casting dies A3 /A2 0.8 to 0.95.

195

Please read the comments to the previous referee In conclusion, its just a plain sloppy piece of work I hope that referee does not mind that I will use it as the chapter quote. (the Authors even have one of the references to their own publications sited incorrectly!). Perhaps, the referee should learn that magazines change names and, that the name appears in the reference is the magazine name at the time of writing the paper.

B.4 Referee 3
The following comments are not arranged in any particular order. General: The text has a number of errors in grammar, usage and spelling that need to be addressed before publication. p 6 1st paragraph - The rsts sentence says that the ow rate is a function of temperature, yet the rest of the paragraph says that it isnt. The rest of the paragraph say the flow rate is a weak function of the temperature and that it explains why. I hope that everyone agrees with me that it is common to state a common assumption and explain why in that particular case it is not important. I wish that more people would do just that. First, it would eliminate many mistakes that are synonymous with research in die casting, because it forces the smart researchers to check the major assumptions they make. Second, it makes clear to the reader why the assumption was made. p 6 - after Eq 2 - Should indicate immediately that the subscript[s] refer to the sections in Figure 1. I will consider this, Yet, I am not sure this is a good idea. p 6 - after equation 2 - There is a major assumption made here that should not pass without some comment[s]3 Assuming steady state - This assumption goes
3 Is

the referee looking for one or several explanations?

196

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY

to the heart of this approach to the lling calculation and establishes the bounds of its applicability. The authors should discuss this point. Well, I totally disagree with the referee on this point. The major question in die casting is how to ensure the right range of filling time and gate velocity. This papers main concern is how to calculate the CD and determine if the CD be assigned to a specific runner. The unsteady state is only a side effect and has very limited importance due to AESS. Of course the flow is not continuous/steady state and is affected by many parameters such as the piston weight, etc, all of which are related to the transition point and not to the pQ2 diagram per se. The unsteady state exists only in the initial and final stages of the injection. As a general rule, having a well designed pQ2 diagram will produce a significant improvement in the process design. It should be noted that a full paper discussing the unsteady state is being prepared for publication at the present time. In general the organization of the paper is somewhat weak - the introduction especially does not very well set the technical context for the pQ2 method and show how the present work ts into it. The present work does not fit into past work! It shows that the past work is wrong, and builds the mathematical theory behind the pQ2 diagram. The last paragraph of the intro is confused [confusing]. The idea introduced in the last sentence on page 2 is that the CD should vary somehow during the calculation, and subsequently variation with Reynolds number is discussed, but the intervening material about geometry eects is inconsistent with a discussion of things that can vary during the calculation. The last two sentences do not t together well either - the assumption of constant CD is not valid - okay, but is that what you are going to talk about, or are you going to talk about particularly the eects of the gate area? Firstly, CD should not vary during the calculations it is a constant for a specific set of runner system and die casting machine. Secondly, once any parameter is changed, such as gate area, CD has to be recalculated. Now the referees statement CD should vary, isnt right and therefore some of the following discussion is wrong. Now about the fitting question. What do referee means by fit together? Do the paper has to be composed in a rhyming verse? Anyhow, geometrical effects are part of Reynolds number (review fluid mechanics). Hence, the effects of the gate area shows that CD varies as well and has to be recalculated. So what is inconsistent? How do these sentences not fit together? On p 8, after Eq 10 - I think that it would be a good idea to indicate immediately

B.4. REFEREE 3

197

that these equations are plotted in Figure 3, rather than waiting to the next section before referring to Fig 3. Also, making the Oz-axis on this graph logarithmic would help greatly in showing the dierences in the three pump characteristics. Mentioning the figure could be good idea but I dont agree with you about the log scale, I do not see any benefits. On p. 10 after Eq 11 - The solution of Eq 11 requires full information on the die casting machine - According to this model, the machine characterized by Pmax, Qmax and the exponent in Eq 1. The wording of this sentence, however, might be indicating that there is some information to be had on the machine other that these three parameters. I do not think that that is what the authors intend, but this is confusing. This is exactly what the authors intended. The model does not confined to a specific exponent or function, but rather gives limiting cases. Every die casting machine can vary between the two extreme functions, as discussed in the paper. Hence, more information is needed for each individual die casting machine. p 12 - I tend to disagree with the premises of the discussion following Eq 12. I think that Qmax depends more strongly on the machine size than does Pmax. In general, P max is the intensication pressure that one wants to achieve during solidication, and this should not change much with the machine size, whereas the clamping force, the product of this pressure and platten are, goes up. On the other hand, when one has larger area to make larger casting, one wants to increase the volumetric ow rate of metal so that ow rate of metal so that ll times will not go up with the machine size. Commonly, the shot sleeve is larger, while the maximum piston velocity does not change much. Here the referee is confusing so many different concepts that it will take a while to explain it properly. Please find here a attempt to explain it briefly. The intensification pressure has nothing to do with the pQ2 diagram. The pQ2 does not have much to do with the solidification process. It is designed not to have much with the solidification. The intensification pressure is much larger than Pmax . I give up!! It would take a long discussion to teach you the fundamentals of the pQ2 diagram and the die casting process. You confuse so many things that it impossible to unravel it all for you in a short paragraph. Please read Chapter 7 or even better read the whole book. Also, following Eq 13, the authors should indicate what they mean by middle range of the Oz numbers. It is not clear from Fig 3 how close one needs to get to Oz=0 for the three curves to converge again. The mathematical equations are given in the paper. They are very simple that you can use hand calculator to find how much close you

198

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY

need to go to Oz = 0 for your choice of error. A discussion on such issue is below the level from an academic paper. . Besides being illustrative of the results, part of the value of an example calculation comes from it making possible duplication of the results elsewhere. In order to support this, the authors need to include the relationships that used for CD in these calculations. The literature is full of such information. If the referee opens any basic fluid mechanics text book then he can find information about it. The discussion on p 14 of Fig 5 needs a little more consideration. There is a maximum in this curve, but the authors criterion of being on the right hand branch is said to be shorter ll time, which is not a criterion for choosing a location on this curve at all. The ll time is monotone decreasing with increasing A3 at constant A2, since the ow is the product of Vmax and A3. According to this criterion, no calculation is needed - the preferred conguration is no gate whatsoever. Clearly, choosing an operating point requires introduction of other criteria, including those that the authors mention in the intro. And the end of the page 14 discussion that the smaller lling time from using a large gate (or a smaller runner!!??) will lead to a smaller machine just does not follow at all. The machine size is determined by the part size and the required intensication pressure, not by any of this. Once again the referee is confusing many issues; let me interpret again what is the pQ2 diagram is all about. The pQ2 diagram is for having an operational point at the right gate velocity and the right filling time. For any given A2 , there are two possible solutions on the right hand side and one on the left hand side with the same gate velocity. However, the right hand side has smaller filling time. And again, the referee confusing another issue. Like in many engineering situations, we have here a situation in which more than one criterion is needed. The clamping force is one of the criteria that determines what machine should be chosen. The other parameter is the pQ2 diagram. It seems that they authors have obscured some elementary results by doing their calculations.4   For example, the last sentence of the middle paragraph on p 15 illustrates that as CD reaches its limiting value of 1, the discharge velocity reaches its maximum. This is not something we should be publishing in 1998.   CD ? There is no mention of the alleged fact of CD reaches its limiting value of 1. There is no discussion in the whole article
4 If it is so elementary how can it be obscured. I have brokenout this paragraph for purposes of illustration.

B.4. REFEREE 3

199

about CD reaching its maximum (CD = 1). Perhaps the referee was mistakenly commenting on different articles (NADCAs book or an other die casting book) which he has confused with this article. Regarding the concluding paragraph on p 15: 1. The use of the word constant is not consistent throughout this paper. Do they mean constant across geometry or constant across Reynolds number, or both. To the readers: The referee means across geometry as different geometry and across Reynolds number as different Re number5 . I really do not understand the difference between the two cases. Arent actually these cases the same? A change in geometry leads to a change in Reynolds number number. Anyhow, the referee did not consider a completely different possibility. Constant CD means that CD is assigned to a specific runner system, or like the common model in which all the runners in the world have the same value. 2. Assuming that they mean constant across geometry, then obviously, using a xed value for all runner/gate systems will sometimes lead to large errors. They did not need to do a lot calculation to determine this. And yet this method is the most used method in the industry(some even will say the exclusive method). 3. Conversely, if they mean constant across Reynolds number, i.e. CD can vary through the run as the velocity varies, then they have not made their case very well. Since they have assumed steady state and the P3 does not enter into the calculation, then the only reason that mention for the velocity to vary during the ll would be because Kf varies as a function of the ll fraction. They have not developed this argument suciently. Let me stress again the main point of the article. CD varies for different runners and/or die casting machines. It is postulated that the velocity does not vary during run. A discussion about P3 is an entirely different issue related to the good venting design for which P3 remains constant. 4. If the examples given in the paper do not represent the characteristics of a typical die casting machine, why to present them at all? Why are the more detailed calculations not presented, instead of the trivial results that are shown?
5 if

the interpretation is not correct I would like to learn what it really mean.

200

APPENDIX B. WHAT THE ESTABLISHMENTS SCIENTISTS SAY These examples demonstrate that the common method is erroneous and that the authors method should be adopted or other methods based on scientific principles. I believe that this is a very good reason.

APPENDIX C My Relationship with Die Casting Establishment

I cannot believe the situation that I am in. The hostility I am receiving from the establishment is unbelievable, as individual who has spent the last 12 years in research to improve the die casting. At rst I was expecting to receive a welcome to the club. Later when my illusions disappeared, I realized that it revolves around money along with avoiding embarrassment to the establishment due to exposing of the truths and the errors the establishment has sponsored. I believe that the establishment does not want people to know that they had invested in research which produces erroneous models and continues to do so, even though they know these research works/models are scientically rubbish. They dont want people to know about their misuse of money. When I started my research, I naturally called what was then SDCE. My calls were never returned. A short time later SDCE developed into what is now called NADCA. I had hoped that this new creation would prove better. Approximately two years ago I wrote a letter to Steve Udvardy, director of research and education for NADCA ( a letter I never submitted). Now I have decided that it is time to send the letter and to make it open to the public. I have a long correspondence with Paul Bralower, former director of communication for NADCA, which describes my battle to publish important information. An open letter to Mr. Baran, Director of Marketing for NADCA, is also attached. Please read these letters. They reveal a lot of information about many aspects of NADCAs operations. I have submitted ve (5) articles to this conference (20th in Cleveland) and only one was accepted (only 20% acceptance compared to 70% to any body else). Read about it here. During my battle to insert science in die casting, many curious things have taken place and I wonder: are they coincidental? Read about these and please let me know what you think.

201

202

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

Open Letter to Mr. Udvardy

Steve Udvardy NADCA, 9701 West Higgins Road, Suite 880 Rosemont, IL 60018-4721 January 26, 1998 Subject: Questionable ethics Dear Mr. Udvardy: I am writing to express my concerns about possible improprieties in the way that NADCA awards research grants. As a NADCA member, I believe that these possible improprieties could result in making the die casting industry less competitive than the plastics and other related industries. If you want to enhance the competitiveness of the die casting industry, you ought to support die casting industry ethics and answer the questions that are raised herein. Many of the research awards raise serious questions and concerns about the ethics of the process and cast very serious shadows on the integrity those involved in the process. In the following paragraphs I will spell out some of the things I have found. I suggest to you and all those concerned about the die casting industry that you/they should help to clarify these questions, and eliminate other problems if they exist in order to increase the die casting industrys prots and competitiveness with other industries. I also wonder why NADCA demonstrates no desire to participate in the important achievements I have made. On September 26, 1996, I informed NADCA that Garbers model on the critical slow plunger velocity is unfounded, and, therefore so, is all the other research based on Garbers model (done by Dr. Brevick from Ohio State University). To my great surprise I learned from the March/April 1997 issue of Die Casting Engineer that NADCA has once again awarded Dr. Brevick with a grant to continue his research in this area. Also, a year after you stated that a report on the results from Brevicks could be obtained from NADCA, no one that I know of has been able to nd or obtain this report. I and many others have tried to get this report, but in vain. It leaves me wondering whether someone does not want others to know about this research. I will pay $50 to the rst person who will furnish me with this report. I also learned (in NADCAs December 22, 1997 publication) that once more NADCA awarded Dr. Brevick with another grant to do research on this same topic for another budget year (1998). Are Dr. Brevicks results really that impressive? Has he changed his model? What is the current model? Why have we not heard about it? I also learned in the same issue of Die Casting Engineer that Dr. Brevick and his colleagues have been awarded another grant on top of the others to do research on the topic entitled Development and Evaluation of the Sensor System. In the September/October 1997 issue, we learned that Mr. Gary Pribyl, chairman of the NADCA Process Technology Task Group, is part of the research team. This Mr. Pribyl is the chairman of the very committee which funded the research. Of course, I am sure, this could not be. I just would like to hear your explanation. Is it legitimate/ethical to have a man on the committee awarding the chairman a

203 grant? Working on the same research project with this Mr. Pribyl was Dr. Brevick who also received a grant mentioned above. Is there a connection between the fact that Gary Pribyl cooperated with Dr. B. Brevick on the sensor project and you deciding to renew Dr. Brevicks grant on the critical slow plunger velocity project? I would like to learn what the reasoning for continuing to fund Dr. Brevick after you had learned that his research was problematic. Additionally, I learned that Mr. Steve Udvardy was given a large amount of money to study distance communications. I am sure that Mr. Udvardy can enhance NADCAs ability in distance learning and that this is why he was awarded this grant. I am also sure that Mr. Udvardy has all the credentials needed for such research. One can only wonder why his presentation was not added to the NADCA proceedings. One may also wonder why there is a need to do such research when so much research has already been done in this area by the worlds foremost educational experts. Maybe it is because distance communication works dierently for NADCA. Is there a connection between Mr. Steve Udvardy being awarded this grant and his holding a position as NADCAs research director? I would like to learn the reasons you vouchsafe this money to Mr. Udvardy! I also would like to know if Mr. Udvardys duties as director of education include knowledge and research in this area. If so, why is there a need to pay Mr. Udvardy additional monies to do the work that he was hired for in the rst place? We were informed by Mr. Walkington on the behalf of NADCA in the NovDec 1996 issue that around March or April 1997, we would have the software on the critical slow plunger velocity. Is there a connection between this softwares apparent delayed appearance and the fact that the research in Ohio has produced totally incorrect and obase results? I am sure that there are reasons preventing NADCA from completing and publishing this software; I would just like to know what they are. I am also sure that the date this article came out (Nov/Dec 1996) was only coincidentally immediately after I sent you my paper and proposal on the shot sleeve (September 1996). What do you think? Likewise, I learned that Mr. Walkington, one of the governors of NADCA, also received a grant. Is there a connection between this grant being awarded to Mr. Walkington and his position? What about the connection between his receiving the grant and his former position as the director of NADCA research? I am sure that grant was awarded based on merit only. However, I have serious concerns about his research. I am sure that these concerns are unfounded, but I would like to know what Mr. Walkingtons credentials are in this area of research. The three most important areas in die casting are the critical slow plunger velocity, the pQ2 diagram, and the runner system design. The research sponsored by NADCA on the critical slow plunger velocity is absolutely unfounded because it violates the basic physics laws. The implementation of the pQ2 diagram is also absolutely unsound because again, it violates the basic physics laws. One of the absurdities of the previous model is the idea that plunger diameter has to decrease in order to increase the gate velocity. This conclusion (of the previous model)

204

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

violates several physics laws. As a direct consequence, the design of the runner system (as published in NADCA literature) is, at best, extremely wasteful. As you also know, NADCA, NSF, the Department of Energy, and others sponsoring research in these areas exceed the tens of millions, and yet produce erroneous results. I am the one who discovered the correct procedure in both areas. It has been my continuous attempt to make NADCA part of these achievements. Yet, you still have not responded to my repeated requests for a 1 grant. Is there a reason that it has taken you 1 2 years to give me a negative answer? Is there a connection between any of the above information and how long it has taken you? Please see the impressive partiallist of the things that I have achieved. I am the one who found Garbers model to be totally and absolutely wrong. I am also the one responsible for nding the pQ2 diagram implementation to be wrong. I am the one who is responsible for nding the correct pQ2 diagram implementation. I am the one who developed the critical area concept. I am the one who developed the economical runner design concept. In my years of research in the area of die casting I have not come across any research that was sponsored by NADCA which was correct and/or which produced useful results!! Is there any correlation between the fact that all the important discoveries (that I am aware of) have been discovered not inbut outside of NADCA? I would like to hear about anything in my area of expertise supported by NADCA which is useful and correct? Is there a connection between the foregoing issues and the fact that so many of the die casting engineers I have met do not believe in science? More recently, I have learned that your secretary/assistant, Tricia Margel, has now been awarded one of your grants and is doing research on pollution. I am sure the grant was given based on qualication and merit only. I would like to know what Ms. Margels credentials in the pollution research area are? Has she done any research on pollution before? If she has done research in that area, where was it published? Why wasnt her research work published? If it was published, where can I obtain a copy of the research? Is this topic part of Ms. Margels duties at her job? If so, isnt this a double payment? Or perhaps, was this an extra separated payment? Where can I obtain the nancial report on how the money was spent? Together we must promote die casting knowledge. I am doing my utmost to increase the competitiveness of the die casting industry with our arch rivals: the plastics industry, the composite material industry, and other industries. I am calling on everyone to join me to advance the knowledge of the die casting process. Thank you for your consideration. Sincerely, Genick BarMeir, Ph.D. cc: NADCA Board of Governors NADCA members Anyone who care about die casting industry

205

Correspondence with Paul Bralower

Paul Bralower is the former director of communications at NADCA. I have tried to publish articles about critical show shot sleeve and the pQ2 diagram through NADCA magazine. Here is an example of my battle to publish the article regarding pQ2 . You judge whether NADCA has been enthusiastic about publishing this kind of information. Even after Mr. Bralower said that he would publish it I had to continue my struggle.

He agreed to publish the article but

At rst I sent a letter to Mr. Bralower (Aug 21, 1997): Paul M. Bralower NADCA, Editor 9701 West Higgins Road, Suite 880 Rosemont, IL 60018-4721 Dear Mr. Bralower: Please nd enclosed two (2) copies of the paper The mathematical theory of the pQ2 diagram submitted by myself for your review. This paper is intended to be considered for publication in Die Casting Engineer. For your convenience I include a disk DOS format with Microsoft WORD for window format (pq2.wid) of the paper, postscript/pict les of the gures (gures 1 and 2). If there is any thing that I can do to help please do not hesitate let me know. Thank you for your interest in our work. Respectfully submitted, Dr. Genick BarMeir cc: Larry Winkler a short die casting list Documents, encl: Disk He did not responded to this letter, so I sent him an additional one on December 6, 1997. Paul M. Bralower NADCA, Editor 9701 West Higgins Road, Suite 880 Rosemont, IL 60018-4721 Dear Mr. Bralower: I have not received your reply to my certied mail to you dated August 20, 1997 in which I enclosed the paper The mathematical theory of the pQ2 diagram authored by myself for your consideration (a cc was also sent to Larry Winkler from Hartzell). Please consider publishing my paper in the earliest possible issue. I believe that this paper is of extreme importance to the die casting eld. I understand that you have been very busy with the last exhibition and congress. However, I think that this paper deserves a prompt hearing.

206

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

I do not agree with your statement in your December 6, 1996 letter to me stating that This paper is highly technical-too technical without a less-technical background explanation for our general readers . I do not believe that discounting your readers is helpful. I have met some of your readers and have found them to be very intelligent, and furthermore they really care about the die casting industry. I believe that they can judge for themselves. Nevertheless, I have yielded to your demand and have eliminated many of the mathematical derivations from this paper to satisfy your desire to have a simple presentation. This paper, however, still contains the essentials to be understood clearly. Please note that I will withdraw the paper if I do not receive a reply stating your intentions by January 1, 1998, in writing. I do believe this paper will change the way pQ2 diagram calculations are made. The pQ2 diagram, as you know, is the central part of the calculations and design thus the paper itself is of same importance. I hope that you really do see the importance of advancing knowledge in the die casting industry, and, hope that you will cooperate with those who have made the major progress in this area. Thank you for your consideration. Sincerely, Dr. Genick Bar-Meir cc: Boxter, McClimtic, Scott, Wilson, Holland, Behler, Dupre, and some other NADCA members ps: You probably know by now that Garbers model is totally and absolutely wrong including all the other investigations that where based on it, even if they were sponsored by NADCA. (All the researchers agreed with me in the last congress) Well that letter got him going and he managed to get me a letter in which he claim that he sent me his revisions. Well, read about it in my next letter dated January 7, 1998. Paul M. Bralower, NADCA, Editor 9701 West Higgins Road, Suite 880 Rosemont, IL 60018-4721 Dear Mr. Bralower: Thank you for your fax dated December 29, 1997 in which you alleged that you sent me your revisions to my paper The mathematical theory of the pQ2 diagram. I never receive any such thing!! All the parties that got this information and myself nd this paper to of extreme importance. I did not revise my paper according to your comments on this paper, again, since I did not receive any. I decided to revised the paper since I did not received any reply from you for more than 4 months. I revised according to your comments on my previous paper on the critical slow plunger velocity. As I stated in my letter

207 dated December 6, 1997, I sent you the revised version as I send to all the cc list. I resent you the same version on December 29, 1997. Please note that this is the last time I will send you the same paper since I believe that you will claim again that you do not receive any of my submittal. In case that you claim again that you did not receive the paper you can get a copy from anyone who is on the cc list. Please be aware that I changed the title of the paper (December, 6, 1997 version) to be How to calculate the pQ2 diagram correctly. I would appreciate if you respond to my e-Mail by January 14, 1998. Please consider this paper withdrawn if I will not hear from you by the mentioned date in writing (email is ne) whether the paper is accepted. I hope that you really do see the importance of advancing knowledge in the die casting industry, and, hope that you will cooperate with those who have made the major progress in this area. Sincerely, Dr. Genick Bar-Meir ps: You surely know by now that Garbers model is totally and absolutely wrong including all the other investigations that where based on it He responded to this letter and changed his attitude I thought. January 9, 1998. Dear Mr. Bar-Meir: Thank you for your recent article submission and this follow-up e-mail. I am now in possession of your article How to calculate the pQ2 diagram correctly. It is the version dated Jan. 2, 1998. I have read it and am prepared to recommend it for publication in Die Casting Engineer. I did not receive any earlier submissions of this article, I was confusing it with the earlier article that I returned to you. My apologies. However I am very pleased at the way you have approached this article. It appears to provide valuable information in an objective manner, which is all we have ever asked for. As is my policy for highly technical material, I am requesting technical personnel on the NADCA sta to review the paper as well. I certainly think this paper has a much better chance of approval, and as I said, I will recommend it. I will let you know of our decision in 2-3 weeks. Please do not withdraw itgive us a little more time to review it! I would like to publish it and I think technical reviewer will agree this time. Sincerely, Paul Bralower Well I waited for a while and then I sent Mr. Bralower a letter dated Feb 2, 1998. Dear Mr. Bralower, Apparently, you do not have the time to look over my paper as you promise. Even a negative reply will demonstrate that you have some courtesy. But apparently the paper is not important as your experts told you and I am only a small bothering cockroach. Please see this paper withdrawn!!!!

208

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

I am sorry that we do not agree that an open discussion on technical issues should be done in your magazine. You or your technical experts do not have to agree with my research. I believe that you have to let your readers to judge. I am sure that there is no other reasons to your decision. I am absolutely sure that you do not take into your consideration the fact that NADCA will have to stop teaching SEVERAL COURSES which are wrong according to this research. Thank you for your precious time!! Dr. Bar-Meir Please note that this letter and the rest of the correspondence with you in this matter will be circulated in the die casting industry. I am sure that you stand by your decision and you would like other to see this correspondence even if they are NADCA members. Here is the letter I received in return a letter from Paul Bralower Feb 5, 1998. Dear Mr. Bar-Meir: Ill have you know that you have inconvenienced me and others on our sta today with your untoward, unnecessary correspondence. If you had a working telephone or fax this e-mail would not be necessary. As it is I must reply to your letter and take it to someone elses oce and have them e-mail it to you right away. I tried to telephone you last week on Thurs. 1/29 with the news that we have agreed to publish your article, How to Calculate the pQ2 diagram correctly. I wanted to ask you to send the entire paper, with graphics and equations, on a disk. Because of the current status of our e-mail system, I would advise you not to e-mail it. Send it on any of the following: Syquest, Omega ZIP or Omega JAZ. Use Microsoft Oce 97, Word 6.0 or Word Perfect 6.0. The problem is I couldnt reach you by phone. I tried sending you a fax several times Thurs. and last Friday. There was no response. We tried a couple of dierent numbers that we had for you. Having no response, I took the fax and mailed it to you as a letter on Monday 2/2. I sent Priority 2-day Mail to your attention at Innovative Filters, 1107 16th Ave. S.E., Minneapolis, Minn, 55414. You should have received it today at latest if this address is correct for you, which it should be since it was on your manuscript. Now, while Im bending over backwards to inform you of your acceptance, you have the nerve to withdraw the paper and threaten to spread negative gossip about me in the industry! I know you couldnt have known I was trying to contact you, but I must inform you that I cant extend any further courtesies to you. As your paper has been accepted, I expect that you will cancel your withdrawl and send me the paper on disk immediately for publication. If not, please do not submit any further articles. My response to Paul M. Bralower. Feb 9, 1998 Dear Mr. Bralower: Thank you for accepting the paper How to calculate the pQ2 diagram correctly. I strongly believe that this paper will enhance the understanding of your readers

209 on this central topic. Therefore, it will help them to make wiser decisions in this area, and thus increase their productivity. I would be happy to see the paper published in Die Casting Engineer. As you know I am zealous for the die casting industry. I am doing my utmost to promote the knowledge and protability of the die casting industry. I do not apologize for doing so. The history of our correspondence makes it look as if you refuse to publish important information about the critical slow plunger velocity. The history shows that you lost this paper when I rst sent it to you in August, and also lost it when I resubmitted it in early December. This, and the fact that I had not heard from you by February 1, 1998, and other information, prompted me to send the email I sent. I am sure that if you were in my shoes you would have done the same. My purpose was not to insults anyone. My only aim is to promote the die casting industry to the best of my ability. I believe that those who do not agree with promoting knowledge in die casting should not be involved in die casting. I strongly believe that the editor of NADCA magazine (Die Casting Engineer) should be interested in articles to promote knowledge. So, if you nd that my article is a contribution to this knowledge, the article should be published. I do not take personal insult and I will be glad to allow you to publish this paper in Die Casting Engineer. I believe that the magazine is an appropriate place for this article. To achieve this publication, I will help you in any way I can. The A paper was written using LTEX, and the graphics are in postscript les. Shortly, I will send you a disc containing all the les. I will also convert the le to Word 6.0. I am afraid that conversion will require retyping of all the equations. As you know, WORD produces low quality setup and requires some time. Would you prefer to have the graphic les to be in TIFF format? or another format? I have enhanced the calculations resolution and please be advised that I have changed slightly the graphics and text. Thank you for your assistance. Sincerely, Dr. Genick Bar-Meir

Is the battle over?

Well, I had thought in that stage that the paper would nally be published as the editor had promised. Please continue to read to see how the saga continues. 4/24/98 Dear Paul Bralower: To my great surprise you did not publish my article as you promised. You also did not answer my previous letter. I am sure that you have a good reasons for not doing so. I just would like to know what it is. Again, would you be publishing the article in the next issue? any other issue? published at all? In case that you intend to publish the article, can I receive a preprint so I can proof-read the article prior to the publication?

210

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

Thank you for your consideration and assistance!! Genick Then I got a surprise: the person dealing with me was changed. Why? (maybe you, the reader, can guess what the reason is). I cannot imagine if the letter was an oer to buy me out. I just wonder why he was concerned about me not submitting proposals (or this matter of submitting for publication). He always returned a prompt response to my proposals, yah sure. Could he possibly have suddenly found my research to be so important. Please read his letter, and you can decide for yourself. Here is Mr. Steve Udvardy response on Fri, 24 Apr 1998 Genick, I have left voice mail for you. I wish to speak with you about what appears to be non-submittal of your proposal I instructed you to forward to CMC for the 1999 call. I can and should also respond to the questions you are posinjg to Paul. I can be reached by phone at 219.288.7552. Thank you, Steve Udvardy Since the deadline for that proposal had passed long before, I wondered if there was any point in submitting any proposal. Or perhaps there were exceptions to be made in my case? No, it couldnt be; I am sure that he was following the exact procedure. So, I then sent Mr. Udvardy the following letter. April 28, 1998 Dear Mr. Udvardy: Thank you very much for your prompt response on the behalf of Paul Bralower. As you know, I am trying to publish the article on the pQ2 diagram. I am sure that you are aware that this issue is central to die casting engineers. A better design and a signicant reduction of cost would result from implementation of the proper pQ2 diagram calculations. As a person who has dedicated the last 12 years of his life to improve the die casting industry, and as one who has tied his life to the success of the die casting industry, I strongly believe that this article should be published. And what better place to publish it than Die Casting Engineer? I have pleaded with everyone to help me publish this article. I hope that you will agree with me that this article should be published. If you would like, I can explain further why I think that this article is important. I am very glad that there are companies who are adopting this technology. I just wish that the whole industry would do the same. Again, thank you for your kind letter. Genick ps: I will be in my oce Tuesday between 9-11 am central time (612) 3782940

211 I am sure that Mr. Udvardy did not receive the comments of/from the referees (see Appendix B). And if he did, I am sure that they did not do have any eect on him whatsoever. Why should it have any eect on him? Anyhow, I just think that he was very busy with other things so he did not have enough time to respond to my letter. So I had to send him another letter. 5/15/98 Dear Mr. Udvardy: I am astonished that you do not nd time to answer my letter dated Sunday, April, 26 1998 (please see below copy of that letter). I am writing you to let you that there is a serious danger in continue to teach the commonly used pQ2 diagram. As you probably know (if you do not know, please check out IFIs web site www.dieperfect.com), the commonly used pQ2 diagram as it appears in NADCAs books violates the rst and the second laws of thermodynamics, besides numerous other common sense things. If NADCA teaches this material, NADCA could be liable for very large sums of money to the students who have taken these courses. As a NADCA member, I strongly recommend that these classes be suspended until the instructors learn the correct procedures. I, as a NADCA member, will not like to see NADCA knowingly teaching the wrong material and moreover being sued for doing so. I feel that it is strange that NADCA did not publish the information about the critical slow plunger velocity and the pQ2 diagram and how to do them correctly. I am sure that NADCA members will benet from such knowledge. I also nd it beyond bizarre that NADCA does not want to cooperate with those who made the most progress in the understanding die casting process. But if NADCA teaching the wrong models might ends up being suicidal and I would like to change that if I can. Thank you for your attention, time, and understanding! Sincerely, Dr. Genick Bar-Meir ps: Here is my previous letter. Now I got a response. What a dierent tone. Note the formality (Dr Bar-Meir as oppose to Genick). May 19, 1998 Dear Dr Bar-Meir, Yes, I am here. I was on vacation and tried to contact you by phone before I left for vacation. During business travel, I was sorry to not be able to call during the time period you indicated. As Paul may have mentioned, we have approved and will be publishing your article on calculating PQ2. The best t for this is an upcoming issue dedicated to process control. Please rest assured that it will show up in this appropriate issue of DCE magazine.

212

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

Since there has been communications from you to Paul and myself and some of the issues are subsequently presented to our Executive Vice President, Dan Twarog, kindly direct all future communications to him. This will assist in keeping him tied in the loop and assist in getting responses back to you. His e-mail address is [email protected]. Thank you, Steve Udvardy Why does Mr. Udvardy not want to communicate with me and want me to write to Executive Vice President? Why did they change the title of the article and omit the word correctly. I also wonder about the location in the end of the magazine. I have submitted other proposals to NADCA, but really never received a reply. Maybe it isnt expected to be replied to? Or perhaps it just got/was lost?

Open letter to Leo Baran

In this section an open letter to Leo Baran is presented. Mr. Baran gave a presentation in Minneapolis on May 12, 1999, on Future Trends and Current Projects to sell NADCA to its members. At the conclusion of his presentation, I asked him why if the situation is so rosy as he presented, that so many companies are going bankrupt and sold. I proceeded to ask him why NADCA is teaching so many erroneous models. He gave me Mr. Steve Udvardys business card and told me that he has no knowledge of this and that since he cannot judge it, he cannot discuss it. Was he prepared for my questions or was this merely a spontaneous reaction? Dear Mr. Baran, Do you carry Steve Udvardys business card all the time? Why? Why do you not think it important to discuss why so many die casting companies go bankrupt and are sold? Is it not important for us to discuss why there are so many nancial problems in the die casting industry? Dont you want to make die casting companies more protable? And if someone tells you that the research sponsored by NADCA is rubbish, arent you going to check it? Discuss it with others in NADCA? Dont you care whether NADCA teaches wrong things? Or is it that you just dont give a damn? I am sure that it is important for you. You claimed that it is important for you in the presentation. So, perhaps you care to write an explanation in the next NADCA magazine. I would love to read it. Sincerely, Genick BarMeir

Is it all coincidental?
I had convinced Larry Winkler in mid 1997 (when he was still working for Hartzell), to ask Mr. Udvardy why NADCA continued support for the wrong models (teaching the erroneous Garbers model and fueling massive grants to Ohio State University). He

213 went to NADCA and talked to Mr. Udvardy about this. After he came back, he explained that they told him that I didnt approach NADCA in the right way. (what is that?) His enthusiasm then evaporated, and he continues to say that, because NADCA likes evolution and not revolution, they cannot support any of my revolutionary ideas. He suggested that I needed to learn to behave before NADCA would ever cooperate with me. I was surprised and shaken. What happened, Larry? I asked him. But I really didnt get any type of real response. Later (end of 1997) I learned he had received NADCAs design award. You, the reader, can conclude what happened; I am just supplying you with the facts. Several manufacturers of die casting machines, Buler, HPM, Prince, and UBE presented their products in Minneapolis in April 1999. When I asked them why they do not adapt the new technologies, with the exception of the Buler, the response was complete silence. And just Buler said that they were interested; however, they never later called. Perhaps, they lost my phone number. A representative from one of the other companies even told me something on the order of Yeah, we know that the Garber and Brevick models are totally wrong, but we do not care; just go awayyou are bothering us!. I have news for you guys: the new knowledge is here to stay and if you want to make the die casting industry prosper, you should adopt the new technologies. You should make the die casting industry prosper so that you will prosper as well; please do not look at the short terms as important. The next issue of the Die Casting Engineer (May/Jun 1999 issue) was dedicated to machine products. Whether this was coincidental, you be the judge. I submitted a proposal to NADCA (November 5, 1996) about Garber/Brevick work (to which I never received a reply). Two things have happened since: I made the proposal(in the proposal I demonstrate that Brevicks work from Ohio is wrong) 1) publishing of the article by Bill Walkington in NADCA magazine about the wonderful research in Ohio State University and the software to come. 2)a scientic article by EKK. During that time EKK also advertised how good their software was for shot sleeve calculations. Have you seen any EKK advertisements on the great success of shot sleeve calculations lately? Here is another interesting coincidence, After 1996, I sent a proposal to NADCA, the cover page of DCE showing the beta version of software for calculating the critical slow plunger velocity. Yet, no software has ever been published. Why? Is it accidental that the author of the article in the same issue was Bill Walkington. And after all this commotion I was surprised to learn in the (May/June 1999) issue of DCE magazine that one of the Brevick group had received a prize (see picture below if I get NADCA permission). I am sure that Brevicks group has made so much progress in the last year that this is why the award was given. I just want to learn what these accomplishments are. For a long time NADCA described the class on the pQ2 diagram as a A close mathematical description. After I sent the paper and told them about how the pQ2 diagram is erroneous, they change the description. Well it is good, yet they have to say that in the past material was wrong and now they are teaching something else. or

put the picture of Brevick, Udvardy and price guy

214

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

are they? I have submitted ve (5) papers to the conference (20th in Cleveland) and four (4) have been rejected on the grounds well, you can read the letter yourself: Here is the letter from Mr. Robb. 17 Feb 1999 The International Technical Council (ITC) met on January 20th to review all submitted abstracts. It was at that time that they downselected the abstracts to form the core of each of the 12 sessions. The Call for Papers for the 1999 Congress and Exposition produced 140 possible abstracts from which to choose from, of this number aproximately 90 abstracts were selected to be reviewed as nal papers. I did recieve all 5 abstracts and distribute them to the appropriate Congress Chairmen. The one abstract listed in your acceptance letter is in fact the one for which we would like to review the nal paper. The Congress Chairmen will be reviewing the nal papers and we will be corresponding with all authors as to any changes revisions which are felt to be appropriate. The Congress Chairmen are industry experts and it is there sole discretion as to which papers are solicited based on abstract topic and t to a particular session. It is unfortunate that we cannot accept all abstracts or papers which are submitted. Entering an abstract does not constitue an automatic acceptance of the abstract/or nal paper. Thank you for your inquiry, and we look forward to reviewing your nal paper. Regards, Dennis J. Robb NADCA I must have submitted the worst kind of papers otherwise. How can you explain that only 20% of my papers (1 out of 5) accepted. Note that the other researchers ratio of acceptance on their papers is 65%, which means that other papers are three times better than mine. Please nd here the abstracts and decide if youd like to hear such topics or not. Guess which the topic NADCA chose, in what session and on what day (third day).

A Nobel Tangential Runner Design

The tangential gate element is commonly used in runner designs. A novel approach to this runner design has been developed to achieve better control over the needed performance. The new approach is based on scientic principles in which the interrelationship between the metal properties and the geometrical parameters is described.

Vacuum Tank Design Requirements

Gas/air porosity constitutes a large part of the total porosity. To reduce the porosity due to the gas/air entrainment, vacuum can be applied to remove the residual air in the die. In some cases the application of vacuum results in a high quality casting while

215 in other cases the results are not satisfactory. One of the keys to the success is the design of the vacuum system, especially the vacuum tank. The present study deals with what are the design requirements on the vacuum system. Design criteria are presented to achieve an eective vacuum system.

How Cutting Edge technologies can improve your Process Design approach
A proper design of the die casting process can reduce the lead time signicantly. In this paper a discussion on how to achieve a better casting and a shorter lead time utilizing these cutting edge technologies is presented. A particular emphasis is given on the use of the simplied calculations approach.

On the eect of runner design on the reduction of air entrainment: Two Chamber Analysis
Reduction of air entrapment reduces the product rejection rate and always is a major concern by die casting engineers. The eects of runner design on the air entrapment have been disregarded in the past. In present study, eects of the runner design characteristics are studied. Guidelines are presented on how to improve the runner design so that less air/gas are entrapped.

Experimental study of ow into die cavity: Geometry and Pressure eects

The ow pattern in the mold during the initial part of the injection is one of the parameters which determines the success of the casting. This issue has been studied experimentally. Several surprising conclusions can be drawn from the experiments. These results and conclusions are presented and can be used by the design engineers in their daily practice to achieve better casting.

Afterward
At the 1997 NADCA conference I had a long conversation with Mr. Warner Baxter. He told me that I had rued a lot of feathers in NADCA. He suggested that if I wanted to get real results, I should be politically active. He told me how bad the situation had been in the past and how much NADCA had improved. But here is something I cannot understand: isnt there anyone who cares about the die casting industry and who wants it to ourish? If you do care, please join me. I actually have found some individuals who do care and are supporting my eorts to increase scientic knowledge in die casting. Presently, however, they are a minority. I hope that as Linux is liberating the world from Microsoft, so too we can liberate and bring prosperity to the die casting industry.

216

APPENDIX C. MY RELATIONSHIP WITH DIE CASTING ESTABLISHMENT

After better than a year since my rst (and unsent) letter to Steve Udvardy, I feel that there are things that I would like to add to the above letter. After my correspondence with Paul Bralower, I had to continue to press them to publish the article about the pQ2 . This process is also described in the preceding section. You, the reader, must be the judge of what is really happening. Additionally, open questions/discussion topics to the whole die casting community are added. What happened to the Brevicks research? Is there still no report? And does this type of research continue to be funded? Can anyone explain to me how NADCA operates? Is NADCA, the organization, more important than the die casting industry?

BIBLIOGRAPHY

[1] ASM. Metals Handbook , volume 13. ASM, Metals Park, Ohio, 1987. [2] G. Backer and Frank Sant. Using Finite Element Simulation for the Development of Shot Sleeve Velocity Proles . In NADCA 19th Congress and exposition, Minneapolis, Minnesota, November 1997. paper T97-014. [3] G. Bar-Meir, E.R.G. Eckert, and R. J. Goldstein. Pressure die casting: A model of vacuum pumping. Journal of Engineering for Industry , 118:001 007, February 1996. [4] Genick Bar-Meir. On gas/air porosity in pressure die casting . PhD thesis, University of Minnesota, 1995. [5] Genick Bar-Meir. Analysis of mass transfer processes in the pore free technique. Journal of Engineering Materials and Technology , 117:215 219, April 1995. [6] A. Bochvar, A., M. Notkin, E., S. I. Spektorova, and N.M. Sadchikova. The Study of Casting Systems by Means of Models. Izvest. Akad. Nauk U.S.S.R. (Bulletin of the Academy of Sciences of U.S.S.R) , pages 875882, 1946. [7] Hydraulics Laboratory Branch. Hydraulic Design of Stilling Basins and Bucket Energy Dissipatiors . U.S. Bureau of Reclamation, Engineering Monograph 25, Denver, Colorado, 1958. [8] Jerald R. Brevick, Dwaine J. Armentrout, and Yeou-Li Chu. Minimization of entrained gas porosity in aluminum horizontal cold chamber die castings . Transactions of NAMRI/SME , 12:4146, November - December 1994. [9] Derek L. Cocks. DCRF Recommended Procedures: Metal Flow Predictor System. American Die Casting Institute, Inc., Des Plaines, Illinois, 1986.

217

218

BIBLIOGRAPHY

[10] Derek L. Cocks and A. J. Wall. Technology transfer in the united kingdom: Progress and prospects. In Transactions 12th International SDCE, Minneapolis, Minnesota, 1983. paper G-T83-074. [11] K Davey and S. Bounds. Modeling the pressure die casting process using boundary and nite element methods. Journal of material Processing Technology, 63:696700, 1997. [12] A. J. Davis. Eects of the relationship between molten metal ow in feed systems and hydraulic uid ow in die casting machines. In Transactions 8th International SDCE, St. Louis, 1975. paper G-T75-124. [13] E.R.G. Eckert. Similarity analysis applied to the Die Casting Process . Journal of Engineering Materials and Technology , 111:393398, 1989. No. 4 Oct. [14] Mohamed El-Mehalawi, Jihua Liu, and R. A. Miller. A cost estimating model for die cast products. In NADCA 19th Congress and exposition, Minneapolis, Minnesota, NOvember 1997. paper T97-044. [15] Fairbanks. Hydraulic HandBook. Mores and Co., Kansas city, Kansas, 1959. [16] H. Fondse, H. Jeijdens, and G Ooms. On the inuence of the exit conditions on the entrainment rate in the development region of a free, round, turbulent jet. Applied Scientic Research, pages 355375, 1983. [17] L. W. Garber. Theoretical analysis and experimental observation of air entrapment during cold chamber lling. Die Casting Engineer , 26 No. 3:33, May - June 1982. [18] G. Hansen, Arthur. Fluid Mechanics. John Wiley and Sons, Inc., New York, New York, 1967. [19] F. M. Henderson. Open Channel Flow. Macmillan Publishing Co., New York, New York, 1966. [20] A. G. Horacio and R. A. Miller. Die casting die deections: computer simulation of causes and eects. In NADCA 19th Congress and exposition, Minneapolis, Minnesota, NOvember 1997. paper T97-023. [21] Henry Hu and S. A. Argyropoulos. Mathematical modeling of solidication and melting: a review . Modeling Simulation Mater. Sci. Eng. , 4:371396, 1996. [22] J. Hu, S. Ramalingam, G. Meyerson, E.R.G. Eckert, and R. J. Goldstein. Experiment and computer modeling of the ows in pressure die casting casings. In ASME/CIE Design, San Francisco, California, 1992. [23] C. M. Kim and Frank J. Sant. An application of 3D solidication analysis to large complex castings. In 2nd Pacic rim international conference on modeling of casting and solidication, Singapore, January 1995.

BIBLIOGRAPHY [24] P.A. Madsen and Svendsen. Turbulent bores and hydraulic jumps. Journal of Fluid Mechanics, 129:125, 1983.

219

[25] Ralph David Maier. Inuence of liquid metal jet character on heat transfer during die casting. PhD thesis, Case Western Reserve University, 1974. Engineering, metallurgy. [26] T Nguyen and J. Carrig. Water Analogue Studies of Gravity Tilt Casting Copper Alloy components. AFS Trans , pages 519528, 1986. [27] Richard H. F. Pao. Fluid Mechanics. John Wiley and Sons, Inc., New York, New York, 1961. [28] N. Rajaratnam. The hydraulic jump as a wall jet. Journal of Hydraulic Div. ASCE, pages 107131, 1965. 91 (HY5). [29] Ascher H. Shapiro. The Dynamics and thermodynamics of Compressible Fluid Flow , volume I. John Wiley and Sons, New York, 1953. [30] W. F. Stuhrke and J. F. Wallace. Gating of die castings. Transactions of American Foundrymens Society , 73:569597, 1966. [31] Marilyn Thome and Jerald R. Brevick. Optimal slow shot velocity proles for cold chamber die casting. In NADCA Congress and exposition, Indianapolis, Indiana, October 1995. paper T95-024. [32] I Wygnanski and F. H. Champan. The origin of pus and slugs and the ow in a turbulent slug. J. Fluid Mechanics , pages 281335, 1973.

220

BIBLIOGRAPHY

Subjects index
A
absolute viscosity, 13 adiabatic nozzle, 25

B
Bernoullis equation, 22

C
convergingdiverging nozzle, 23

F
fanno second law, 145 fanno ow, 143, 4f L 147 D choking, 148 average friction factor, 149 entrance Mach number calculations, 157, 175 entropy, 148 shockless, 155, 156 star condition, 150

M
Mach number, 23

S
shear stress, 12 speed of sound star, 24 speed of sound, what, 21 stagnation state, 23

AUTHORS INDEX

221

Authors index