Caesar II-4 - 4 Oct02

Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

M

e
c
h
a
n
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
N
e
w
s
FOR THE POWER,
PROCESS AND
RELATED INDUSTRIES
The COADE Mechanical Engineering
News Bulletin is published twice a year
from the COADE offices in Houston,
Texas. The Bulletin is intended to provide
information about software applications
and development for Mechanical
Engineers serving the power, process and
related industries. Additionally, the Bulletin
serves as the official notification
vehicle for software errors discovered in
those Mechanical Engineering programs
offered by COADE.
2002 COADE, Inc. All rights reserved.
V O L U M E 3 3 O C T O B E R 2 0 0 2
Whats New at COADE
CAESAR II Version 4.40 Released................ 1
TANK Version 2.40 Released ........................ 6
Technology You Can Use
Current Approach to Defining Ground
Motion ........................................................ 7
Vessel Stresses due to Seismic, Pressure,
and Weight Loads ................................... 15
The Bend Flexibility Factor ........................... 18
The Pressure Equipment directive (PED)
and CE marking....................................... 20
PC Hardware for the Engineering User
(Part 33) ................................................... 21
Program Specifications
CAESAR II Notices ...................................... 22
TANK Notices .............................................. 23
CodeCalc Notices ........................................ 23
PVElite Notices ............................................ 23
I N T H I S I S S U E :
CAESAR II Version 4.40 Released
CAESAR II Version 4.40 is a major release, incorporating many
enhancements, in addition to the usual plethora of piping code updates. The
major enhancements made to CAESAR II for the 4.40 release are listed in the
table below.
A large number of updates have been made to the 3D input graphics, as
well as partial implementation in the static output processor (including
the Element Viewer).
An alpha-numeric node label option has been added to the piping input
module
Expanded Static Load case options: (1) added load components H, CS,
HP, and WW (hanger loads, cold spring, hydro pressure, and weight
filled with water, respectively), (2) added HYDRO stress type, (3) added
option to set snubber and hanger status for each load case, (4) provided
ability to scale friction factor for each load case.
Added automatic generation of a hydrotest load case (WW+HP, HYD
stress type, and spring hangers locked), triggered by the presence of a
non-zero HP.
The B31.11 piping code (for slurry piping) has been added.
> continued on p.2
COADE Mechanical Engineering News October 2002
2
Updated the spring hanger design algorithm to provide the
option to iterate the Operating for Hanger Travel load case
to include the stiffness of the selected hanger.
Added new configuration options for ambient temperature,
default friction coefficient (if non-zero, automatically applied
to new translational restraints), liberal stress allowable, stress
stiffening, and Bourdon settings, as well as B31.3 handling of
B16.9 welding tee & sweepolet SIFs.
Spring Hanger manufacturers have been added, Pipe Supports
USA and Quality Pipe Supports.
Added the ability to define the flexibility factor on bends.
Included piping and structural files now support long file
names, may be located in any directory path, and the number
of included structural files has been expanded from 10 to 20.
Results of the Hanger Design Cases are now optionally
viewable in the Static Output Processor (set status to KEEP
in the Load Case Options).
Added the ability to filter static Restraint reports by CNODE
status.
Updated API-661 to 4
th
Edition.
In addition to these enhancements, a number of Technical Changes
have been made. Technical Changes are changes necessitated by
either Code changes or implementation changes on our part that
may cause the software to generate results different from prior
versions. The list of these Technical Changes can be found with
the distribution letter, and in Chapter 1 of the Users Guide.
Users are urged to review this list before using Version 4.40.
This article will elaborate on a number of the enhancements made
for Version 4.40. Additional articles will detail some of the more
involved topics.
Node Labels:
Node Labels are an optional alternative to node numbers. These
labels can be used to identify or tag important points in the piping
model. Note however, that CAESAR II still requires the use of
node numbers to uniquely define node points. The node labels
are optional user defined descriptions.
Node Labels can be specified by checking the Name check box
(next to the From / To fields) on the piping input spreadsheet,
as shown in the figure below. When activated, this check box
displays the Node Names Auxiliary dialog, also shown in the
figure below.
In the Node Names Auxiliary dialog, you can define a label for
each node, up to ten characters in length. (This length limit is
dictated by the space available in the output reports.) Use of node
names is as follows:
If defined, node names are used in graphics displays instead
of node numbers.
If defined, node names are used in all text based reports,
instead of node numbers.
In 80 column reports, only the first six characters of the node
name will be used.
In 132 column reports, all ten characters of the node name
will be used.
An example graphics image and output report showing the use of
these node labels are shown below.
October 2002 COADE Mechanical Engineering News
3
Load Case Components H and CS:
Version 4.40 provides a number of new capabilities for constructing
static load cases. One change made here was to break out
implied loads from the F1 vector. Recall that in the input
processor, concentrated loads (forces and/or moments) can be
applied to node points. These loads can be grouped into different
vectors (F1 through F9) as necessary, perhaps to correspond to
temperature vectors T1 through T9. Defining F1 in the input means
it can be incorporated into a load case.
What was not obvious (but explained in the documentation) is the
fact that spring hanger preloads and cold spring loads were also
loads included in the F1 vector. So, you couldnt separately
address F1, hanger loads H, or cold spring, without making
specific adjustments to the input (you would move the applied
forces to F2, define colds spring using the T2 method, and
leave hanger preloads in F1). This is not obvious, and a little bit
of work.
With Version 4.40, two new basic load components have been
defined. H represents spring hanger preloads, and CS represents
cold spring. Now, a load case that in pre-4.40 looked like:
W + T1 + P1 + F1
may in 4.40 and later versions look like:
W + T1 + P1 + F1 + H + CS
The advantage here is that F1, H, and CS can be used
independently.
3D Graphics Improvements:
Quite a lot of work went into the 3D (HOOPS) graphics for Version
4.40. If you have never used this graphics option, you really owe it
to yourself to explore this capability. (Some of these 4.40 graphics
improvements were released in 4.30 builds, as soon as they were
operational.)
Two of the major additions to the 3D graphics are the display of
displacement and force vectors. This display puts all defined
displacements (or forces) in a scrollable list box. The list can be
scrolled vertically to review the data for all nodes, and horizontally
to review the data for all vectors. Additionally, this list box can be
moved around the graphics window, to minimize its effect on the
viewable pipe region. An example of this displacement display is
shown in the figure below.
COADE Mechanical Engineering News October 2002
4
Note in this display the color key at the left. These colors are
correlated to appropriate node points on the model graphic, to assist
in locating the displaced point, even if node numbers are not
displayed. The display of applied forces works identically to the
display of displacements.
These same 3D graphics have begun their migration into the static
output processor in Version 4.40. In the figure below, the new
Element Viewer is shown. This new viewer allows the selection
of various data types (displacements, forces, or stresses), for the
various load cases.
Note that the highlighted element in the Viewer Window corresponds
to the highlighted element on the model graphic. Clicking on a
different element in the Viewer Window highlights the corresponding
element in the graphic. Likewise, clicking on a different element in
the graphic, highlights the corresponding element entry in the Viewer
Window.
Configuration Directive for Friction Coefficient and Load Case
Friction Multiplier
Friction is a concept that cannot be quantified with 100% accuracy.
The value of the friction coefficient changes over time, as the
system ages and is exposed to the environment. Prior to Version
4.40, there was no easy way to specify or alter the value of this
friction coefficient on every restraint. Version 4.40 provides two
new capabilities that address modeling and evaluating friction effects.
The first new friction capability is specification of a default value of
the friction coefficient in the configuration file. (Recall that each
data directory can have its own configuration file, so this default
friction value only affects jobs where the configuration file has been
modified appropriately.) By specifying a non-zero value for the
friction coefficient, as shown in the figure below, all newly defined
translational restraints (X, +X, -X, Y, +Y, -Y, Z, +Z, -Z, Guide, and
Limit) automatically attain this friction coefficient. Existing
(translational) restraints in existing jobs are not affected by this
configuration setting.
As usual, the job can now be analyzed, which will include frictional
effects on all (newly defined) translational restraints for all load
cases. But what if it was necessary to analyze a particular load case
without friction, or what if it was necessary to consider a 20%
increase in the friction coefficient. In Versions prior to 4.40, these
evaluations would mean returning to input and altering the model.
Now with Version 4.40, both of these situations can be addressed at
the load case level, without altering the existing input. For
example, assume a standard set of load cases as shown in the figure
below.
October 2002 COADE Mechanical Engineering News
5
As expected, all basic load cases (1 through 4) are analyzed with the
friction coefficient of 0.35 (on the appropriate restraints). To
consider the Operating case without friction, and again with a 20%
increase in friction, we now can simply add two more load cases,
and change the Friction Multiplier, in the right most column of
the Load Case Options tab. These modifications are shown in
the figure below.
In this figure, cases 5 and 6 are alternate operating cases using
different frictional effects. Using this load case setup, a single run
can evaluate friction effects considering a low value, the expected
value, and a high value of the friction coefficient. The CAESAR II
summary reports can then be used against these load cases to draw
conclusions and make decisions.
Static Restraint Report CNODE Filter
The CNODE is a very powerful modeling construct in CAESAR II.
Essentially, a CNODE is an association of degrees of freedom.
CNODES are most often used to define internal restraints, or put
another way, how one part of the model is connected to another
part. An excellent example of this use is the restraint setup used to
connect an internal core pipe to an external jacket, such as in a riser.
A portion of such a model is shown in the figure below.
This model consists of a 4 inch core pipe (nodes in 4xx series), a 6
inch core pipe (nodes in 6xx series), and an 18 inch jacket pipe
(nodes in 18xx series). Each of these pipes is labeled in the figure
above. The core pipes are positioned/connected to the jacket pipe
using the dummy rigid elements and CNODEs. For example,
node 415 is restrained to node 414 in the X and Z directions
using a CNODE. The dummy (shown as a solid bar in the figure
above) spans from 414 to 1815, with 1815 a jacket node. Similarly,
node 615 is restrained to node 614 414 in the X and Z directions
using a CNODE.
This modeling setup means that as node 1815 moves in X and
Z, nodes 414 / 615 move in an identical manner. Thus the core
and jacket pipes move together in the horizontal plane (at these
locations) but they are independent in the vertical (Y) direction.
This is great, until it is necessary to review the output restraint
report for external loads. In this situation, the loads at the
CNODE restraints are not desired as these are internal loads. Prior
to Version 4.40, the restraint report for this model would have
contained all of the restrained nodes, as shown in the figure below.
COADE Mechanical Engineering News October 2002
6
In this figure, only the nodes with text labels are external restraints.
All other nodes are internal (CNODE) restraints. Because of these
text labels this distinction is easy to view, but what if text labels
were not used. Then the review of external loads would be more
tedious.
Version 4.40 provides a new option on the Restraint Filter, as
shown in the figure below. This option allows restraints to be
filtered based on whether or not a CNODE is associated with the
node.
In this figure, the filter is set to show on restraints without CNODE
associations. This will produce a report containing only external
restraints, as shown in the figure below.
The restraint loads in this figure can be summed to obtain, in this
case, the total applied wave load on the riser system.
TANK Version 2.40 Released
The primary purpose for the release of TANK Version 2.40 is to
address the new API-650 (Edition 10, Addendum 2) and the new
API-653 (Edition 3). Both of these codes were published just as
TANK Version 2.30 was released.
In addition to addressing the changes necessitated by these code
revisions, a number of other enhancements have been made to the
software. Most noticeable of these changes is the addition of a
dynamic sizing tool, on the tank sizing scratchpad. This dynamic
sizing tool, shown in the figure below, allows users to adjust the
height and / or width of the proposed tank by using the slider bars.
As these bars are adjusted, the tank sketch reflects the new size, and
the required shell course thicknesses (in the 5
th
column of the table
above) are recomputed.
October 2002 COADE Mechanical Engineering News
7
Current Approach to Defining
Ground Motion
By: Dave Diehl
Introduction
B31.3 paragraph 301.5.3 (Earthquake) states: Piping shall be
designed for earthquake-induced horizontal forces. The method of
analysis may be as described in ASCE 7 or the Uniform Building
Code. B31.1 paragraph 101.5.3 (Earthquake) is similar but does
not reference ASCE 7 or UBC. However, B31.1 adds that the
effects of earthquakes shall be considered using data for the site as
a guide in assessing the forces involved. This article will review
the current method of defining the nature of ground shaking (the
ground acceleration as a function of frequency or period) as it
would be applied in a modal analysis of the piping systems installed
at a particular site. Anyone who has no interest in seismic analysis
can move on to the next article. If you are interested in seismic
analysis but you dont want to read another article on how they do
things in the US ASCE 7 and UBC are both US building codes
please hang on for a little longer. This article will be useful in
setting useful ground shaking data for sites around the world.
Additionally, youll get a quick overview of the current authorities
on this subject and their documents.
Before defining ground shaking, I want to briefly talk about that
B31.3 quote. ASCE 7 is alive and well but UBC the Uniform
Building Code is no longer an active document. The preface of
the 1997 UBC states The ICBO
1
code development process has
been suspended by the Board of Directors and, because of this
action, changes to the Uniform Building Code will not be processed.
There will be no more UBC. I called the ICBO about this situation
and received a thoughtful e-mail from Roger Mensink. Quoting
Roger, I dont believe there is anything in print that explains the
suspension of the code development process for the UBC in more
detail. As you probably know, ICBO, BOCA, and SBCCI are
integrating with the ICC (International Code Council). Together,
these organizations have already published the International Codes,
of which the International Building Code essentially replaces the 97
Uniform Building Code. The code development process continues
for the IBC. However, many jurisdictions are still under the UBC,
and printings incorporating errata continue to be published. I believe
we are currently in the seventh printing of the 97 UBC. Yes, its a
suspended document but as long as there are laws on the books
referencing UBC, it will remain in print. For more information visit
http://iccint.intlcode.org/.
____________________
1
International Conference of Building Officials
What Ground Shaking Data is in COADE Products Now?
Both CAESAR II and PVElite offer standard shock spectra as
defined in US Nuclear Regulatory Guide 1.60. CAESAR II also
offers canned shock spectra from 1994 UBC. For a complete
definition of ground shaking for Reg. Guide 1.60 and 1994 UBC,
the user need only enter the Zero Period Acceleration or ZPA for
the site. CAESAR II and PVElite provide data from the highly
documented El Centro, California earthquake. El Centro requires
no additional input but has limited application to sites outside
southern California. PVElite also collects ground shaking data for
ASCE 7 and IBC. The definition for ground acceleration is identical
for both of these codes and these will be discussed below. So this
article will review the method for setting an ASCE 7 shock spectrum
and, in doing so, bring you up to date on shock definition.
Let me also add that spectrum analysis is only one approach for
evaluating piping systems in a seismic zone. Most piping systems
and pressure vessels do not require a response spectrum analysis.
Many piping systems do not require even a simpler, static equivalent
load as in many cases, depending on pipe size and service, simple
rules dictate the location of lateral and axial (horizontal) restraints
in order to circumvent any seismic analysis. But if you are going to
use spectrum analysis you might as well be familiar with todays
approach.
Alphabet Soup! Who Sets the Rules?
1994 UBC
The oldest building code
2
referenced in CAESAR II is the 1994
Uniform Building Code published by International Conference of
Building Officials (ICBO). 1994 UBC Figure 16-3 shown here in
Figure 1 illustrates the nature of horizontal ground motion. For any
given period (inverse frequency) of a structure there is a
corresponding acceleration. The response is also affected by ground
composition (rock, soil, or clay). Using this chart, a structure
having a (fundamental) mode of vibration of 0.5 seconds (2 Hz)
would see a load of 2g if constructed on rock and 2.5g if built on
clay (assuming the earth has a maximum acceleration of 1 g).
____________________
2
U.S. Nuclear Regulatory Commission Guide 1.60 (Revision 1)
Design Response Spectra for Seismic Design of Nuclear Power
Plants dates back to 1973 but it is not what you would call a
building code.
COADE Mechanical Engineering News October 2002
8
1994 UBC Normalized Response Spectra Shapes
0
1
2
3
0 0.5 1 1.5 2 2.5 3
Period, T (seconds)
N
o
r
m
a
l
iz
e
d

E
f
f
e
c
ti
v
e

P
e
a
k
G
r
o
u
n
d
A
c
c
e
l
e
r
a
t
io
n

(
g
)
Soft to Medium Clays and Sands
(Soil Type 3)
Deep Cohesionless or Stiff Clay Soils
(Soil Type 2)
Rock & Stiff Soils
(Soil Type 1)
Figure 1
These curves are stored as shock spectra in CAESAR II. You can
select them based on soil type. The only other piece of data
required is the Zero Period Acceleration or ZPA for the site. Note
that these normalized spectra all show 1g acceleration at T = 0
which is a ZPA = 1. The user-specified ZPA will multiply these
accelerations across the entire spectrum. ZPA is a function of the
seismic zone (0, 1, 2A, 2B, 3 & 4) associated with the site. Figure
16-2 in 1994 UBC maps these zones across the US. Appendix
Paragraph 1655.4 states that these normalized spectra can be scaled
by the product ZN where Z is the seismic zone factor (based on
seismic zone)
3
and N is the near-field coefficient
4
(a value between
1.0 & 1.5 based on the magnitude of the ground shaking and
proximity to the active fault). So for a structure constructed in
seismic zone 4 more than 15 km from an active fault (Z=0.40 and
N=1.0), the ZPA is 0.40 and, if constructed on rock, the calculated
response would be 0.80g for that example above (having a
fundamental period of 0.5 seconds).
Unfortunately, Appendix Chapter 16 from which these factors come
is titled Division III Earthquake Regulations for Seismic-Isolated
Structures and there is no similar calculation found in the main body
of the code for all other structures. But there is a way back. In the
____________________
3
Table A-16-C:
Zone Z
0 0
1 0.075
2A 0.15
2B 0.20
3 0.30
4 0.40
4
In Table A-16-D, N is a value between 1.0 & 1.5 based on the
magnitude of the ground shaking and proximity to the active fault.
dynamic procedures (paragraph 1629.5.3 Scaling of Results), it
says dynamic loads should be scaled up or down so that the structures
base shear using static analysis methods essentially match the base
shear using dynamic methods. The static method calculates design
base shear (V) in Equation 28-1: W R ZIC V
w
= ) ( which has a
form of load = (factor) times (weight). This factor, then, serves as
the acceleration or g load on the structure the g load we should see
in the dynamic analysis. The static factor has three parts ZC, I and
R
w
. Guidelines for Seismic Evaluation and Design of Petrochemical
Facilities
5
on page 4-9 states The product ZC represents the spectral
acceleration corresponding to the period of the structure. the
spectral ordinate corresponding to the period of the structure may
be used in lieu of the ZC value, subject to the approval of building
officials. So this ZC, alone, should approximate the response from
our ground shaking response spectrum but there are two other
values I and R
w
. I varies between 1.0 and 1.5 and reflects the
importance of the structure; buildings which must remain operational
through an earthquake or have higher consequences with failure
have the higher value of I. Now since the actual earthquake
magnitude does not change based on what you build, this value
reflects an increased factor of safety instead of an increase in
ground motion. However, since the static and dynamic base shear
must match, it seems sensible to include this importance factor in
the ground shaking spectrum. So now the ZPA in CAESAR II
would be ZNI rather than just ZN. And what of that R
w
? Again to
quote Guidelines
6
, R
w
is a measure of energy absorbing capability
or ductility inherent in a particular type of structural system or
structure. It represents an equivalent reduction in seismic forces by
allowing energy dissipation once the structure begins to respond in
the inelastic range. Local yielding in the structure dampens or
reduces the response to the earthquake; a typical value for tall
towers and piping range between 3.0 and 4.0. There is no clear
indication of how this factor could be fit back into the shock
spectrum but there is one interesting note in Guidelines
7
, When a
dynamic analysis is performed, various building codes generally
require that the base shear (after dividing by R
w
) be compared to
that of an equivalent static analysis. After dividing by Rw
indicates that the dynamic results do not reflect this ductility factor
and it must be incorporated in the dynamic results prior to comparison
with the static base shear. The conclusion, then, is to use ZNI as the
CAESAR II value for ZPA but then compare the total shear (or
total restraint reaction in the appropriate direction) to the calculated
value V.
Another piece of information from 1994 UBC will be noted here for
future reference. Paragraph 1629.2 Ground Motion states The
____________________
5
Published by ASCE in 1997
6
ibid., page 4-10
7
ibid., page 4-22
October 2002 COADE Mechanical Engineering News
9
ground motion representation shall, as a minimum, be one having a
10 percent probability of being exceeded in 50 years. Note 2 of
that paragraph says use damping ratio of 0.05 and Note 5 says the
vertical shock component may be defined as two thirds of this
horizontal shock.
1997 UBC
The 1997 issue of the Uniform Building Code, the last UBC,
dropped the normalized shock spectra and broke away from the
rigid structure of response spectrum timing. The shock spectra still
maintained shape of the 1994 code but the fixed periods starting and
ending the flat peak portion of the spectra are calculated based on
seismic zone and soil characteristics. Paragraph 1631.2 Ground
Motion again refers to a shock with a 10 percent probability of
being exceeded in 50 years; Note 2 says use damping ration of
0.05; and Note 5 says vertical can be 2/3 horizontal. As no COADE
package offers canned data for 1997 UBC we will move on to the
other authorities. To learn more about the ICBO visit http://
www.icbo.org/About_ICBO/.
IBC 2000
The International Building Code is published by the International
Code Council, Inc. (ICC) and is the result of the combined efforts of
three organizations Building Officials and Code Administrators
International, Inc. (BOCA), International Conference of Building
Officials (ICBO), and Southern Building Code Congress
International, Inc. (SBCCI). These International codes are intended
to replace the existing group of Uniform codes. I understand that in
the future IBC will, in turn, delete its seismic provisions and refer,
instead, to ASCE-7. The first thing you notice in the IBC is the
detail and number of seismic maps no more simple set of six
zones. The second is the number of new terms mapped spectral
response accelerations at short period (S
S
) and at 1 second (S
1
), the
maximum considered earthquake (MCE) spectral response
accelerations at short period (S
MS
) and at 1 second (S
M1
), and the
damped design spectral response accelerations at short period (S
DS
)
and at 1 second (S
D1
). Common to each set of terms is the short
period and 1 second period reference. Each component of this
mapped pair is multiplied by a site coefficient (F
a
for S
S
and F
v
for
S
1
) to get the MCE data and that then is scaled down to the design
data.
These maps show MCE accelerations for Site Class B. Site Class B
is assigned to soil profiles consisting of rock. If your structure is to
be build on rock (and there is a definition of hard rock, rock and soft
rock) you can use these values directly. But what if you are
constructing the building, vessel or piping system on a different soil
profile, say, stiff soil or hard rock? Different soil profiles have
different site coefficients to convert mapped accelerations to MCE
accelerations. You must back-calculate the mapped accelerations
from the listed Site Class B MCE accelerations and then re-calculate
the MCE accelerations for the proper Site Class. However, all the
site coefficients (F
a
& F
v
) for Site Class B are 1.0. That means that
the MCE accelerations for Site Class B are equivalent to the mapped
accelerations used with any site. (For Site B, S
MS
=S
S
and S
M1
=S
1
.)
The maps appear in pairs one for short period and one for 1
second period; all are for 5% critical damping; and the MCE maps
reference NEHRP more on that later. Although I could not locate
a direct reference in the maps, these response spectra are now based
on 2% probability of exceedance within a 50 year period rather than
the 10% used in earlier codes. Its this change in the mapped
earthquake magnitude that requires the MCE data be scaled down to
the design data (converting S
MS
to S
DS
and S
M1
to S
D1
). This
reduction in load will match load magnitudes in previous codes.
The shock spectra produced here appears identical to that in ASCE
7 so it will not be developed here. You can visit http://
www.intlcode.org/about/index.htm to learn more about the
International Code Council.
ASCE 7-98
ASCE 7-98 is entitled Minimum Design Loads for Buildings and
Other Structures and is published by American Society of Civil
Engineers. The home page for ASCE is http://www.asce.org. This
document is the other standard listed in B31.3 and the focus of this
article
The tables, maps and equations found here are identical to the IBC
terms. Once again, you go to the maps and collect the mapped
spectral acceleration for an imaginary oscillator tuned to a period of
0.2 seconds (a frequency of 5 Hz) and then an oscillator tuned to 1
sec (1 Hz). Tables provide site multipliers for these accelerations
which then are converted to Maximum Considered Earthquake, or
MCE, accelerations. These site multipliers are based on average
shear wave velocity, average penetration resistance and average
undrained shear strength of the rock or soil. If these data are not
known, the code allows you to use Class D to set the multipliers (see
Exception to 9.4.1.2.1). Site Class D is considered stiff soil while
Site Class B is rock. Well build a shock spectrum in accordance
with ASCE 7 after reviewing a few more documents.
NEHRP
NEHRP is not a document; instead it is a government program. The
National Earthquake Hazards Reduction Program was established
by an act of Congress in 1977 to reduce the risks to life and
property from future earthquakes in the United States. The lead
agency for this program is FEMA, the Federal Emergency
Management Agency. The United States Geological Survey (USGS)
also plays a role and has published seismic data maps for NEHRP.
These are the maps that appear in IBC 2000 and ASCE 7. I
mentioned that these response acceleration maps are based on 5%
critical damping. You may ask Whats that have to do with my
piping system? Modal analysis, by necessity, eliminates damping
from the equations of motion but includes its effect by increasing or
COADE Mechanical Engineering News October 2002
10
decreasing the applied load. So a system having 10% critical
damping would use a lesser response acceleration than a 5% system.
Fortunately for us, piping systems shall be designed assuming a 5%
critical damping. (See ASME Boiler & Pressure Vessel Code Section
III Division 1 Appendix N paragraph N-1231.) Therefore, we can
use the mapped data without modification for damping.
You dont have to have the actual map to find your local mapped
spectral response accelerations S
S
& S
1
. US ground motion data
are available online through USGS. With a specification of latitude
and longitude or even a simple Zip (postal) code, the site will list
local ground motion data. This data is available at http://
eqint.cr.usgs.gov/eq/html/zipcode.shtml. We will use this data source
in the example.
FEMA 302 & 303 and 368 & 369
These NEHRP documents are prepared by the Building Seismic
Safety Council for the Federal Emergency Management Agency.
These books containing NEHRP Recommended Provisions for
Seismic Regulations for New Buildings and Other Structures are
produced in pairs: Part 1 Provisions in one book (about the size of
a phone book) and Part 2 Commentary in the other. The 302/303
pair form the 1997 edition and the 368/369 pair make up the 2000
Edition. These four documents are available online in PDF files at
http://www.bssconline.org/ or they can be sent to you by mail.
There is no charge for these government documents and the order-
takers are quite friendly. The FEMA web site is http://
www.fema.gov/ but I couldnt locate their entire document catalog
there. In North America call 1(800)480-2520 and either ask for
these documents or have them send you a FEMA Publications
Catalog. The catalog will provide instructions for faxing in your
complete order. Its more than just earthquakes; they have
information on fires, floods, hazardous materials and the like.
If you have trouble reading those small US maps in ASCE 7 or
UBC, order a set from FEMA. They are quite large Architectural
E format. You will get response accelerations at both 0.2 seconds
and one second. They are defined as Maximum Considered
Earthquake (MCE) Ground Motion Site Class B but again you can
read these as mapped accelerations since Site Class B multipliers to
convert mapped to MCE accelerations are all 1.0. Maps 27, 28, 31
& 32 are labeled as Probabilistic Earthquake Ground Motion and
appear to match Maps 1,2, 3 & 4 respectively. These maps show
the mapped accelerations and do not reference Site Class B. I guess
they wanted to show it both ways. Use the ones that are based on
2% probability of exceedance. They throw in one set of California
and Nevada maps (Maps 29 & 30) that show accelerations based on
10% probability of exceedance in 50 years the old 1994 UBC
standard. Naturally, an earthquake that has a one in ten chance of
occurring is smaller in magnitude than an earthquake that has only a
one in fifty chance of occurring. And yes, a point with 0.55g
acceleration on the 10% map shows as about 1.20g on the 2% map.
As you would expect, most of the big numbers are along the west
coast. What may be surprising is the very high g loads along the
Mississippi River bordering Kentucky and Tennessee. I couldnt
find anything in California that exceeded the 3g acceleration (at 0.2
seconds) here. South Carolina, too, has impressive numbers of over
1.5g (again at 0.2 seconds) about halfway up the coast.
The Provisions (FEMA 302 & 368) recommend rules for the seismic
design of structures and cover several methods, many of which are
not applicable to our piping systems. Our intent is to use these
documents to define the ground shaking and not the method of
system analysis. Chapter 4 Ground Motion serves that purpose
well. The Commentaries (FEMA 303 & 369) provide the background
and justification for these rules. So if you want to learn more of the
whys and wheres to seismic load specification, these documents
are a great place to start. Theyre certainly priced right.
TI 809-04
A CAESAR II user alerted me to this document. (Thank you Tej.)
Not knowing what it was or who created it, I just went to the internet
and searched on the phrase TI 809-04 and came up with a web
page containing this document authored by the US Army Corps of
Engineers their Technical Instructions, Seismic Design of
Buildings. The Corps of Engineers has the entire document broken
into chapters and downloadable as PDF files. I know of no hard
copy available for this document. The intent is to have the latest
information immediately available for interested parties through the
internet at http://www.hnd.usace.army.mil/techinfo/ti/809-04/
ti80904.htm. There are many references to NEHRP 1997 and
FEMA 302 in TI 809-04.
Paragraph 2-2 Ground Motion in Chapter 2 states the art of
ground motion prediction has progressed in recent years to the point
that nationally approved design criteria have been developed by
consensus groups of geotechnical and building design professionals.
So thats what we are doing here applying those national design
criteria.
One gem in this document is the conversion factors for the old
(1991) UBC data. This gives you a way to relate the old methods to
the new. Additionally, international sites with old UBC data can
now use this conversion to build the new shock spectrum. The
document provides a table showing S
S
& S
1
for many cities around
the world. Download Chapter 3 from the web site to learn more.
TI 809-04 also defines two ground motions Ground Motion A and
Ground Motion B. Ground Motion A matches that found in FEMA
302 and is used for Performance Objectives 1A (Life Safety) and
2A (Safe Egress for Special Occupancy) while Ground Motion B is
used for Performance Objectives 2B (Safe Egress for Hazardous
Occupancy) and 3B (Immediate Occupancy for Essential Facilities).
A is scaled to 2/3 of the mapped spectral acceleration while B is
scaled to of the same acceleration. You apply A (2/3) or B (3/4)
depending on the building use the more important the building,
October 2002 COADE Mechanical Engineering News
11
the higher the design load. Other codes handle this by increasing
the detail of analysis for the more important structures or tightening
design limits (see, for example, Seismic Design Categories in ASCE
7).
Summary
Everybody appears, now, to be saying the same thing when it comes
to ground shaking. If you are interested in such detail, visit the web
site of S. K. Ghosh Associates Inc. at http://
w w w . s k g h o s h a s s o c i a t e s . c o m /
Comparisons%20of%20seismic%20codes.htm. They provide a
detailed comparison of IBC 2000, 1997 NEHRP & ASCE 7-98 and
even throw in 1997 UBC
What terms do we use?
That USGS web site lists more numbers than we need. So what do
we need?
Heres the data for my house in Houston, Texas, zip code 77069
certainly not an exciting place in terms of ground motion but it
shows what data is available.
Figure 2
We will use the last column in the table 2%PE in 50 yr or 2%
probability of exceedance in 50 yrs. The 2% probability of being
exceeded in 50 years equates to an equivalent return period of 2475
years
8
. A return period is an estimate of how long it will be between
events of a given magnitude; here, you can expect an earthquake of
this magnitude to come around once in 2475 years. The 1994 UBC
____________________
8
The relation between return period, T and probability of
exceedance for a Poissonian distribution is given as:
) 1 ln( p t T = where p is the probability of exceedance in t
years. Guidelines for Seismic Evaluation and Design of
Petrochemical Facilities Published by ASCE 1997
code used the first column or 10% in 50 years (once in 475 years)
and called it the maximum capable earthquake, now (2% in 50) is
called maximum considered earthquake or MCE. There are four
rows in this table: PGA or Peak Ground Acceleration, 0.2, 0.3 and
1.0 SA or spectral accelerations for oscillators tuned to 0.2, 0.3 and
1.0 second. We will use the values for 0.2 and 1.0 seconds. The
codes refer to these MCE accelerations as S
MS
and S
M1
. S
MS
is short
period acceleration where short here means 0.2 seconds. If I attach
a simple cantilever to the ground and this cantilever has a first mode
of vibration of 5 Hz. (a frequency of 5 cps is the same as a period of
0.2 seconds), its maximum acceleration would be the value shown
in the table above. S
M1
is the acceleration for a one second
oscillator. Accelerations are listed as percent g so, for my home,
S
MS
= 0.1063 g and S
M1
= 0.0448 g.
For now we need not consider PGA or Peak Ground Acceleration
but it may become a useful term in the future. The B31 codes are
working on a new document B31S for the seismic design of piping
systems. At this time it looks like B31S will use PGA (for my house
its 0.0470 g) in its design rules. You may ask how my site can have
S
MS
and S
M1
values greater than PGA. PGA is the actual ground
acceleration while S
MS
and S
M1
are system responses to that
acceleration. Thats what makes dynamic loads interesting the
response can be greater than the applied load.
Although I could not find an explicit statement, these accelerations
assume a 5% critically damped response. Thats what we should
use for piping systems.
Well, what if you do not have a US Zip Code associated with your
site? Go back to that TI 809-04 web site and open up the Chapter 3
PDF. Table 3-3 (starting on page 13) lists mapped accelerations (S
S
& S
1
) for over 200 sites around the world. If you cant find your city
but know your UBC zone, page 11 starts a discussion on converting
old 1991 UBC data to the needed values. I should also note that
these calculations are approximate and you should always look for
more recent or more applicable data. Even the USGS data carries a
disclaimer stating the data may not match those NEHRP maps.
Localizing your shock spectrum
Another note on the USGS web site states: These ground motion
values are calculated for firm rock sites which correspond to a
shear-wave velocity of 760 m/sec. in the top 30m. Different soil
sites may amplify or de-amplify these values. For a given earthquake
event, the ground response will vary according to the hardness of
the ground either increasing or decreasing the g load to the
structure. This characteristic is accounted for through a term called
Site Class in ASCE 7. Looking at the Site Class definitions in
ASCE 7, we find that the firm rock site referenced here matches
Site Class B. As mentioned earlier in the discussion on the FEMA
maps, Site Class B has multipliers of 1.0 to convert mapped MCE to
MCE or S
S
to S
M
and S
MS
to S
M1
. To keep things straight, the
authorities create mapped earthquake accelerations and you convert
COADE Mechanical Engineering News October 2002
12
these values to MCEs using these site conversions. ASCE 7 Table
9.4.1.2 defines six Site Classes:
Class Definition
A Hard rock
B Rock
C Very dense soil and soft rock
D Stiff soil
E Soil
F Soils requiring site specific evaluation
These classes are sorted by shear wave velocity, penetration
resistance and undrained shear strength. So what sort of soil do I
have at my house? How do I convert the mapped accelerations to
my ground accelerations? Fortunately, ASCE 7 provides this
exception in paragraph 9.4.1.2.1:When the soil properties are not
known in sufficient detail to determine the Site Class, Class D shall
be used. With multipliers greater than 1.0, Class D sites will
increase the mapped accelerations.
How do you build a seismic shock spectrum from S
S
& S
1
?
Eventually, we need a shock spectrum similar to the 1994 UBC
spectra illustrated at the beginning of this article. Referring again to
that figure we see g load plotted versus period. Once we calculate
the piping systems important natural frequencies, we can read the
ground accelerations for these frequencies (the inverse of the period)
from the chart and load the individual system mode shapes
accordingly. We must create something similar to that original
spectrum in Figure 1 for these ASCE 7 numbers. Note, too, that the
figure shows three curves for the three soil types found in that code.
Here are the terms needed to create an ASCE 7 shock spectrum:
S
S
= mapped short period acceleration read from the maps or the
USGS zip code site
S
1
= mapped 1 second acceleration read from the maps or the
USGS zip code site
S
MS
= short period Maximum Considered Earthquake (MCE)
acceleration
S
M1
= one second MCE acceleration
S a MS
S F S =
1 1
S F S
v M
=
Fa = acceleration-based site coefficient from Table 9.4.1.2.4a,
based on Site Class and S
S
, values range between 0.8 to 2.5
Fv = velocity-based site coefficient from Table 9.4.1.2.4b, based
on Site Class and S
1
, values range between 0.8 to 3.5
S
DS
= design spectral acceleration at short periods (0.2 seconds)
S
D1
= design spectral acceleration at a one second period
MS DS
S S =
3
2
1 1
3
2
M D
S S =
(TI 809-04 states in paragraph 3.1c that using 2/3 of MCE will
compare well with the prior codes that were based on 10% probability
of exceedance in 50 years rather than the new 2% probability of
exceedance in 50 years.)
With S
DS
and S
D1
, the response spectrum for ASCE 7 can be built.
Paragraph 9.4.1.2.6 provides the parameters to build the following
curve:
Fig. 9.4.1.2.6 Design Response Spectrum
from ASCE 7-98 (also, 2000 NEHRP / FEMA 368)
Period T
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

A
c
c
e
l
e
r
a
t
i
o
n

S
a
S
DS
T
0
T
S
S
D1
1.0
Figure 3
The S
a
curve is broken into three sections: 1) the straight line S
a
for
periods less than T
0
, 2) the flat S
a
between T
0
and T
S
and 3) the S
a
curve above T
S
.
DS
D
S
S
T
1
0
2 . 0 =
DS
D
S
S
S
T
1
=
|
.
|

\
|
+ =
0
6 . 0 4 . 0
T
T
S S
DS a for T dT
0
DS a
S S = for T
0
< T dT
S
T
S
S
D
a
1
=
for T > T
S
October 2002 COADE Mechanical Engineering News
13
PVElite will collect the values that you enter for S
S
, S
1
, F
a
and F
v
and build this spectrum for you. In CAESAR II you must collect
data from the resulting shock spectrum and enter the table in the
CAESAR II dynamic input or access the table as a text file stored
on the hard disk.
ASCE 7 sample
Lets finish up by building a shock spectrum used to design a piping
system in my back yard. Again, the USGS web site displays S
MS
and
S
M1
(as 0.1063 g and 0.0448 g, respectively) for Site Class B but
since Site Class B has F
a
and F
v
equal to 1.0, these Maximum
Considered Earthquake values are equivalent to the mapped values
S
S
and S
1
. I guess that my back yard is soft soil but since I do not
have sufficient information to determine the soil profile, I may
assume Site Class D. That is unless I have Site Class E or F soils.
1063 . 0 =
S
S g
0448 . 0
1
= S g
6 . 1 =
a
F for Site Class D where S
S
d0.25g (Table 9.4.1.2.4a)
4 . 2 =
v
F for Site Class D where S
1
d0.1g (Table 9.4.1.2.4b)
170 . 0 1063 . 0 6 . 1 = = =
S a MS
S F S g
108 . 0 0448 . 0 4 . 2
1 1
= = = S F S
v M
g
113 . 0 170 . 0
3
2
3
2
= = =
MS DS
S S g
072 . 0 108 . 0
3
2
3
2
1 1
= = =
M D
S S g
127 . 0
113 . 0
072 . 0
2 . 0 2 . 0
1
0
= = =
DS
D
S
S
T
sec
637 . 0
113 . 0
072 . 0 1
= = =
DS
D
S
S
S
T
sec
( ) ( ) T
T
T
T
S S
DS a
+ = + =
|
.
|

\
|
+ = 534 . 0 0452 . 0
127 . 0
6 . 0 4 . 0 113 . 0 6 . 0 4 . 0
0
g
for T dT
0
113 . 0 = =
DS a
S S g for T
0
< T dT
S
T T
S
S
D
a
072 . 0 1
= =
g for T > T
S
Plotting these three S
a
curves we have:
ASCE 7-98 Design Response Spectrum Curve
Unknown Site Class in Dave's Back Yard (Houston, TX)
0.000
0.050
0.100
0.150
0 0.5 1 1.5 2 2.5 3
Period, T (seconds)
A
c
c
e
l
e
r
a
t
i
o
n

(
g
)
Mapped Spectral Response Acceleration
(2%prob. of exceedance in 50 yr.)
short period (T=0.2) Ss = 0.1063g
one second period S1 = 0.0448g
Maximum Considered Earthquake values
(using Site Class D)
Sms = 0.170g
Sm1 = 0.108g
Design Spectral Response Accelerations
SDS = 0.113g
SD1 = 0.072g
T0 = 0.127 sec
TS = 0.637 sec
CAESAR II Data
period (sec) g
0.0001 0.050
0.15 0.113
0.60 0.113
0.80 0.090
1.20 0.060
2.05 0.035
3.00 0.024
Figure 4
Its not much of an earthquake; I would focus on wind loads, here,
instead.
To enter this shock spectrum in PVElite, enter values for S
S
, S
1
, F
a
and F
v
. The CAESAR II input for this spectrum is accomplished in
three components listed below and highlighted in Figure 5.
1. Set the spectrum name and parameters in the Spectrum
Definitions tab
2. Build the spectrum table using the Spectrum Data Points
button (using data similar to the period vs. g table shown in
Figure 4). Here, for example:
Period (sec) g
0.001 0.050
0.15 0.113
0.60 0.113
0.80 0.090
1.20 0.060
2.05 0.035
3.00 0.024
3. Reference the spectrum name in the shock load case(s) in the
Spectrum Load Cases Tab and supply the appropriate factor
to convert g to acceleration in terms of length/time
2
COADE Mechanical Engineering News October 2002
14
Figure 5
Application
Building codes evaluate buildings more on strength methods rather
than stress methods and piping and vessel just dont act like buildings.
For example, the building codes focus on base shear and story drift
terms important in building design but useless in piping design.
Where piping is itemized, the codes reduce the ground shaking to a
static equivalent load on the supports and the relative displacement
of the different supports; again focusing on the building loads and
not the piping response. For the building codes, vessels and piping
are essentially rigid components attached to the building. They
allow a full response spectrum modal analysis but the guidance
given is not as complete as the static Equivalent Lateral Force or
ELF approach.
While the ground motion as defined above can be used directly in
COADEs programs, ASCE 7 and its implementation in PVElite
introduces three additional terms: The Occupancy Importance Factor
(I), the Response Modification Coefficient (R), and the Deflection
Amplification Factor (C
d
). The response Modification Coefficient,
R, essentially represents the ratio of the forces that would develop
under the specific ground motion if the structure had an entirely
linearly elastic response.
9
I consider R analogous to a shape factor
for our pipe. We calculate and evaluate stress at the outer fiber of
the pipe but for the section to form a plastic hinge and yield across
the entire section, a higher bending moment is required. This ratio
of bending moments, M(complete yield)/M(outer fiber yield) is the
shape factor for the cross section and is a function of section
geometry. Likewise, significant yield of one element of a building
may not produce building failure as load will transfer to other
elements creating an increased building capacity. So R reflects the
building ductility. Values for R range from 3 to 8 for most plant
structures with skirt supported towers having a value of 4 and
piping systems at around 3.5. But accounting for ductility also
____________________
9
FEMA 303 (1997 Edition of NEHRP Recommended Provisions,
Part 2 Commentary) page 61.
accepts damage to the building. Since some buildings are too
important to permit such damage, the R value is reduced by the
Occupancy Importance Factor, I. I runs from 1.0 for most buildings
to 1.5 for structures that must function throughout an earthquake.
Usually the modal masses are each multiplied by their corresponding
ground acceleration to produce the modal inertial load but here each
of the modal ground accelerations taken from the response spectrum
is first divided by R/I. So instead of F = m a on the modal level,
you have F = m a / (R/I). In the dynamic analysis, these reduced
loads would produce lower system deflections. While these loads
may be appropriate for ASCE 7 evaluation, the resulting deflections
may not. To correct this, ASCE 7 introduces the Deflection
Amplification Factor, C
d
. Values for C
d
are greater than 1.0 but less
than R. The modal deflections are multiplied by (C
d
/I) in order to
back out the load adjustments and produce a more accurate
deflected shape. One way to calculate these deflections in
CAESAR II would be to run a second dynamic analysis where the
ground accelerations are multiplied by the ratio C
d
/R. ASCE 7
Table 9.5.2.2 lists Values for R and C
d
by building type. Occupancy
Importance Factors are listed in Table 9.1.4 based on Seismic Use
Groups (Table 9.1.3) which is defined according to Occupancy
Category (Table 1-1).
The ratios R/I and C
d
/R can be easily included in the CAESAR II
analysis by specifying the appropriate Factor in the Spectrum Load
Case Definition. A spectrum factor is already implied in the data
above since the shock table shows g load and CAESAR II requires
acceleration in terms of length/time
2
. Converting from gs to inches
per second squared requires a factor of 386. If the ratio R/I is to be
included, simply divide 386 by R/I. So if R = 3.5 and I = 1.0, then
the spectrum factor will be entered as (386)/(3.5/1.0) or 110.
These R, I and C
d
factors are fully incorporated in PVElite analyses
in accordance with the modal analysis guidelines outlines in ASCE
7. Modal accelerations are automatically divided by the R/I ratio.
Additionally, modes occurring at a period greater than 4 seconds
(or at frequencies less than 0.25 cycles per second) use a different
multiplier found in Equation 9.5.4.5-4. Likewise the modal
deflections are multiplied by the C
d
/I ratio.
But would this be appropriate in a piping system analysis where
yield is to be avoided? I think not. Yield plays a large role in
secondary load analysis. Local yielding is acceptable as the piping
system grows to its operating position. This is known as shakedown.
October 2002 COADE Mechanical Engineering News
15
Primary stress evaluation, however, has as its limit the yield stress
of the piping material. Generally, the pipes outer fiber may not be
greater than 88% of the yield stress
10
. If a good design does not
permit local yielding, then yield adjustments such as R and C
d
should not be included in the calculations.
Closing
So here I go through this long narrative on creating a shock spectrum
using building code guidelines and I end up suggesting we pick and
choose some rules and ignore others. The building codes dont look
at piping as distributed flexible structures as we do. They see
piping as a building component that responds to the shock but does
not amplify it. Another issue, so far avoided, is what shock actually
hits the piping system. I have outlined a quick method for generating
a localized ground response spectrum but many piping systems are
not attached to the ground. The structure between the piping and
earth will also modify the pipes shock spectrum. If the structure is
flexible and it carries relatively large piping loads, then the interaction
of piping and structure would require the inclusion of the structural
steel in the seismic analysis of the piping system. Well leave that
for another article. Lets close by saying we now have a robust
means of setting a local ground response spectrum and a good
review of useful internet sites.
____________________
10
B31.3 limits sustained plus occasional stresses to 1.33 times the
hot allowable stress (S
h
). In most cases, the hot allowable stress is
2/3 yield stress (S
y
)at temperature. 1.33 times 2/3 is 0.88 so there is
no yield. Nuclear piping analyzed for earthquakes which may cause
local yielding (a safe shutdown earthquake, SSE) address yield in
two ways 1) the allowable stress runs up to 1.8S
h
or 1.2S
y
and 2)
percent critical damping for SSE was twice as great as the percent
critical damping for an operating basis earthquake (OBE) thereby
reducing seismic load in a manner similar to the ASCE 7 methods
using R. The Nuclear Code now uses the higher percent critical
damping for both the OBE and SSE.
Determination of Stresses
on a Vertical Vessel due to
Earthquake, Pressure and
Weight Loads
By: Scott Mayeux
In a previous article we compared the generation of loads on a tall
vertical vessel using the Response Spectrum Analysis and the static
equivalent method. In this article, we will investigate how to
compute bending stresses on a tower using the UBC 1994 earthquake
code and combine them with pressure and axial weight stresses.
With any engineering analysis a set of design parameters must be
specified, typically in a design specification. Information in the
design specification must then be entered into PVElite. The
earthquake Design data for this analysis is as follows:
Importance Factor [I] - 1.05
Soil Type - Rock
Horizontal Force Factor [R
w
] - 3.0
Seismic Zone - 4
One objective of the analysis is to determine the longitudinal bending
stresses at each section. To determine the stresses there are a
number of steps that must be followed.
These steps are:
1. Compute the Elevations of the beginning and mid points of
each element.
2. Design or Analyze the Vessel for Internal and External
Pressure as required.
3. Determine the weights of the vessel elements as well as the
nozzles, platforms, rings and other attachments to the vessel.
4. Determine the mechanical properties of each element such as
the moment of inertia and cross sectional area.
5. Determine the Fundamental period of Vibration.
6. Use the equations listed in the Earthquake code to calculate
the base shear.
7. Distribute the base shear to each element. This will be a
concentrated lateral force that acts at the center of each
element.
8. From the top down, sum all of the moments at the beginning
of each section.
COADE Mechanical Engineering News October 2002
16
9. Compute the tensile and compressive stresses and compare to
appropriate Code allowable stresses.
10. If the vessel is being re-rated or designed, determine
acceptability. If the geometry is overstressed and the vessel
is being designed, increase the thickness and return to step 2.
A sketch of the vessel is shown in the figure below:
As shown each element is 10 feet long. The major diameter is 96
inches, the minor diameter is 72 inches and all elements are 1 inch
thick. The heads have a 2-inch straight flange. The coordinates of
element are shown in the table below. The values shown in the Y
column are the distances from the bottom of the base to the end of
the element. Because we will be computing the stresses at the
beginning of the element, the vertical dimension of those points will
be computed in addition to the element mid point.
XY Coordinate Calculations
| | | | | |
From| To | X (Horiz.)| Y (Vert.) | DX (Horiz.)| DY (Vert.) |
| | ft. | ft. | ft. | ft. |
10| 20| 0.00000 | 10.0000 | 0.00000 | 10.0000 |
20| 30| 0.00000 | 10.1667 | 0.00000 | 0.16667 |
30| 40| 0.00000 | 20.1667 | 0.00000 | 10.0000 |
40| 50| 0.00000 | 30.1667 | 0.00000 | 10.0000 |
50| 60| 0.00000 | 40.1667 | 0.00000 | 10.0000 |
60| 70| 0.00000 | 50.1667 | 0.00000 | 10.0000 |
70| 80| 0.00000 | 50.3333 | 0.00000 | 0.16667 |
The pressure design for this tower has already been completed so
proceed to step 3. The weights of each element are shown below.
Element and Detail Weights
| | Element | Element | Corroded | Corroded |
From| To | Metal Wgt. | ID Volume | Metal Wgt. | ID Volume |
| | lb. | in | lb. | in |
10| 20| 10348.8 | 0.00000 | 10348.8 | 0.00000 |
20| 30| 3359.40 | 130288. | 3359.40 | 130288. |
30| 40| 10348.8 | 868588. | 10348.8 | 868588. |
40| 50| 9114.26 | 669536. | 9114.26 | 669536. |
50| 60| 7788.26 | 488580. | 7788.26 | 488580. |
60| 70| 7788.26 | 488580. | 7788.26 | 488580. |
70| 80| 1934.79 | 57001.1 | 1934.79 | 57001.1 |

Total 50683 2702574 50682 2702574
Next, calculate the cross sectional properties such as the area and
section modulus.
Longitudinal Stress Constants
| | Metal Area | Metal Area | New & Cold | Corroded |
From| To | New & Cold | Corroded | Sect. Mod. | Sect. Mod. |
| | in | in | in. | in. |
10| 20| 304.734 | 304.734 | 7315.18 | 7315.18 |
20| 30| 304.734 | 304.734 | 7315.18 | 7315.18 |
30| 40| 304.734 | 304.734 | 7315.18 | 7315.18 |
40| 50| 304.734 | 304.734 | 7315.18 | 7315.18 |
50| 60| 229.336 | 229.336 | 4129.60 | 4129.60 |
60| 70| 229.336 | 229.336 | 4129.60 | 4129.60 |
70| 80| 229.336 | 229.336 | 4129.60 | 4129.60 |
Determine the fundamental period of vibration (POV). This is
achieved by using the Freese method or the Eigen solution method.
The POV for this tower turned out to be 10.0067 cycles per second.
The natural frequency is used in many wind and seismic codes.
Altering the fundamental POV can change wind and seismic loads.
Compute the base shear V. This is a single force that acts at the base
of the skirt and is then distributed to all elements. This distribution
depends on the height and mass of each element. From the Uniform
Building Code 1994 Ed. Volume 2, Structural Engineering Design
Provisions section 1628, the base shear is computed as follows:
W
R
ZIC
V
w
=
where
(

= 75 . 2 ,
25 . 1
min
3 / 2
T
S
C
The terms in these equations are:
Z - the seismic zone factor given in UBC 94 table 16-I
I - the specified importance factor
C - numerical Coefficient specified in UBC Section 1628.2.1
W - the total dead weight
S - site coefficient for soil characteristics in table 16-J
T - fundamental period of vibration in seconds
R
w
- numerical coefficient found in tables 16-N and 16-P
For Zone 4, a highly active zone area, the value of Z from the table
is 0.4. The site coefficient S is 1.0 for rock-like soils. I, the
importance factor as specified above is 1.05 and R
w
is 3.0. W, the
total dead load is the sum of all component and element weights.
October 2002 COADE Mechanical Engineering News
17
W is 50683 pounds. In order to determine V, the soil and frequency
factor C must also be computed. C = min( 5.80, 2.75) = 2.75.
Now, the total base shear
lbs V 8 . 19512 50683
0 . 3
75 . 2 * 05 . 1 * 4 . 0
= =
With a known base shear and element masses, a single concentrated
load can be computed for each element in the vessel. The equations
for vertical distribution of force are found in UBC 94 paragraph
1628.4. In addition to computing the force at each level, the
concentrated top force F
t
must also be

computed.
Ft = 0.07TV
There are some restrictions on Ft and it is zero if T is less than 0.7
seconds. In our case T = 1/10.00067 = 0.09999, therefore Ft = 0.0.
Now, compute the force at level x, the midpoint of each element.
This is determined by using the following relation:
( )

=
n
i
i i
x x t
x
h w
h w F V
F
1
The first task in solving this equation is to acquire the mass at each
level and multiply by the distance from the base of the skirt to the
center of the element and add them all together. After the summation
is complete, F
x
can be computed. Evaluation of the expression

=
n
i
i i
h w
1
provides a result of 1194830 ft.-lbs. Next, determine the
element load on each section. Take the skirt element 10-20 and use
the relation above to compute the equivalent lateral force at the skirt
mid point.
( )
. 03 . 845
1194830
0 . 5 * 8 . 10348 0 8 . 19512
lbs F
x
=

=
In a similar manner, compute the element loads for the other tower
elements. The results for the F
x
calculation are shown in the table
below and are displayed as the Element Load.
Earthquake Load Calculation
| | Earthquake | Earthquake | Element |
From| To | Height | Weight | Load |
| | ft. | lb. | lb. |
10| 20| 5.00000 | 10348.8 | 845.030 |
20| 30| 10.0833 | 3359.40 | 553.195 |
30| 40| 15.1667 | 10348.8 | 2563.26 |
40| 50| 25.1667 | 9114.26 | 3745.93 |
50| 60| 35.1667 | 7788.26 | 4472.85 |
60| 70| 45.1667 | 7788.26 | 5744.75 |
70| 80| 50.2500 | 1934.79 | 1587.75 |
Top 51.8333 0.00
With the element loads and centroidal distances known, the bending
moment at the beginning or base of each element can be computed.
The summation of the moments is given in the following report:
Wind/Earthquake Shear, Bending
| | Elevation | Cummulative| Earthquake | Wind | Earthquake |
From| To | of To Node | Wind Shear| Shear | Bending | Bending |
| | ft. | lb. | lb. | ft.lb. | ft.lb. |
10| 20| 5.00000 | 3779.64 | 19512.8 | 97458.3 | 640560. |
20| 30| 10.0833 | 3047.31 | 18667.7 | 63323.5 | 449658. |
30| 40| 15.1667 | 3035.10 | 18114.5 | 62816.7 | 446593. |
40| 50| 25.1667 | 2300.45 | 15551.3 | 36138.9 | 278264. |
50| 60| 35.1667 | 1557.42 | 11805.4 | 16849.5 | 141480. |
60| 70| 45.1667 | 852.013 | 7332.50 | 4802.35 | 45791.1 |
70| 80| 50.2500 | 94.3213 | 1587.75 | 70.6834 | 1189.84 |
At the base of the large cylindrical shell course element 30-40, it
can be seen that the bending moment is 446593 ft-lbs. The axial
weight stress at the element base is the
weight above that section divided by the cross sectional metal area.
That result is:
2
33 . 121
73 . 304
36974
in
lbs
Area
Weight
w
= = =
In a similar manner compute the bending stress. The bending stress
is tensile on one side and compressive on the other. PVElite
computes all of the individual stresses and classifies them in the
Stresses due to report. This is the bending stress:
2
6 . 732
18 . 7315
) 12 ( 446593
in
lbs
ulus SectionMod
Moment
b
= = =
Another important stress to compute is the longitudinal pressure
stress. Rearrange the ASME or appropriate code formula to obtain
this value. It should be about the value of the hoop stress.
( ) ( ) ( ) ( )
2
int
3570
1 * 2
1 * 4 . 0 48 * 150
2
4 . 0
in
lbs
T
T r P
ca
ca ins
lp
=

=
On the tensile side the stress due to pressure adds with the stress due
to bending. The weight stress compresses the section, so that stress
component subtracts. The resulting combination of these stresses
is: 3570 + 732.6 121.33 = 4181.3 psi (tension). On the compressive
side the combined stress is 3570 732.6 121.33 = 2716 psi
(tension). So for this combination of loads it is seen that the
compressive side is not in compression. This is why the pressure
element compressive stresses are blank in the table below. The
allowable tensile stress is 1.2 * Allowable Tensile Stress *
Circumferential Joint Efficiency. After each stress combination has
been computed, divide that result by its allowable. If the result is
greater than 1, the section is overstressed. The tensile ratio is
4181.27/24000, which equals 0.1742.
f
f

COADE Mechanical Engineering News October 2002


18
Analyzing Stresses for Load Case : IP+OW+EQ+FS Stress Units: psi
From Tensile All. Tens. Comp. All. Comp. Tens. Comp.
Node Stress Stress Stress Stress Ratio Ratio

10 884.47 24000.00 -1217.11 -18948.07 0.0369 0.0642


20 4175.27 24000.00 -18948.07 0.1740
30 4181.27 24000.00 -18948.07 0.1742
40 3939.10 24000.00 -18948.07 0.1641
50 3004.76 24000.00 -19825.55 0.1252
60 2760.67 24000.00 -19825.55 0.1150
70 2665.02 24000.00 -19825.55 0.1110
These results are for a typical operating scenario. The components
IP, OW and EQ listed above are the values computed above. IP
designates internal pressure stress, OW designates Operating Weight
Stress and EQ designates Bending Stress due to Earthquake Loads.
It is important to note that it is empirically difficult to determine
which load case will govern the overall analysis. PVElite computes
a broad range of scenarios and lists those results in the Stress due
to Combined Loads report.
In conclusion, we have determined how to compute combinations
of stresses using PVElite and the Uniform Building Code 1994
Edition. Most all of the other earthquake codes work in a similar
manner. As with any Engineering Code Analysis, the rules will
change periodically as the industry's understanding of load generation
on structures improves.
The Bend Flexibility Factor
By: Dave Diehl
Introduction
CAESAR II Version 4.40 now has another input item for bends a
user-specified flexibility factor. To use it properly, one should
know what it is and, then, how it can be used to improve piping
models.
Definition
Simple beam models which work well in the mathematical
representation for straight pipe do not reflect true distortion for
other piping components such as bends and branch connections.
Unlike the simple cylinder of straight pipe, the bends cross section
will distort and ovalize when a bending moment (either in or out of
plane) is applied. This change in cross section will reduce the
moment of inertia of the bend thereby making it more flexible. This
added component flexibility is represented by a flexibility factor.
To quote Markl
1
, the so-called flexibility factor can be defined
as the relation of the rotation per unit length of the part in question
produced by a moment, to the rotation per unit length of a straight
pipe of the same nominal size and schedule or weight produced by
the same moment.
The flexibility factor is defined (by Markl) in Appendix D of B31.3
as 1.65/h where h, the flexibility characteristic, is defined
as
2
2 1
r R T h =
2
. CAESAR II lists this bend flexibility factor in
the Bend SIF scratchpad in the input processor (reproduced below)
and also in the Miscellaneous Data output report.
So a straight pipe of length
l
, having an anchor at one end and a
moment M on the other, will have an end rotation of while a
piping elbow (of the same cross section, material and support) with
a centerline length of
l
with the same in-plane bending moment M,
will have an end rotation of k, where k is the flexibility factor. In
equation form:
Bend end rotation: ( ) I E M k
bend
= l
Model A
____________________
1
Piping-Flexibility Analysis by A.R.C. Markl , from Tube Turns
Piping Engineering sixth edition, 1986. Piping Engineering can be
purchased directly from Tube Turns Technologies, Inc. in Louisville
Kentucky. The book contains many tables, charts and nomographs
of piping properties and several important technical papers in addition
to Markls (a Tube Turns employee) significant works.
2
T is nominal pipe thickness, R
1
is bend radius, and r
2
is mean pipe
diameter
October 2002 COADE Mechanical Engineering News
19
Cantilever end rotation: I E M
cant
= l
Model B
or
cant bend
k =
Since end rotation is linearly related to beam length, another way to
look at this is to say a bend of arc lengthl will exhibit the same
rotation as a cantilever of length kl under the same moment.
Model C
Two ways to illustrate this
Compare the end rotation of a 6 inch standard, A106 Gr. B long
radius bend (R
1
=9 inches) to a straight run of an equivalent length.
Both components will be anchored at one end and carry a moment
of -1200 in-lbf on the other. The length of the straight run will be
equal to the arc length of the elbow or 14.137 inches ((/2)R
1
). The
reported free-end rotation for the bend (Model A) is -0.0772 degrees
and the reported rotation for the straight (Model B) is -0.0117
degrees. The ratio
cant bend
can be calculated as
(-0.0772) / (-0.0117) = 6.59
3
. The flexibility factor for this elbow is
reported as 6.59002. The ratio of the deflections, on the other hand,
does not follow this ratio:
57 . 5 ) 0014 . 0 ( ) 0078 . 0 ( = =
cant bend
.
Or, you can change length of the straight run to k times the elbows
arc length as shown in Model C. Here, the new length is
6.59002 14.137 inches or 7.764 feet and the end rotation is once
again -0.0772 degrees which matches the elbow.
____________________
3
Based on the values and uncertainty of the input (0.00005), the
uncertainty of the quotient is 0.0324. The calculated value for k,
then, is 6.590.0324.
Code-defined values for k
Lets return to the definition of k. The piping codes define k for us
in Appendix D. Markls paper provides the history of the flexibility
factor starting with work by von Krmn in 1911 and then by
Hovgaard in 1926. Even then the parameter known as the flexibility
characteristic (
2
2 1
r R T h = ) was in use. But the von Krmn-
Hovgaard flexibility factor of that day
( ) ( ) 1 12 10 12
2 2
+ + = h h k erred when this h dropped below
0.3. (A 6 inch std long radius elbow has h=0.25.) It was Beskin, in
1945, who suggested a much simpler equation for k as in h k 65 . 1 =
and k 1. And that is the elbow flexibility factor in use today.
Markl went on to discuss adjustments for conditions that prevent
the bend from ovalizing and allowing this higher flexibility. He
described corrections for internal pressure and attached flanges.
Both of these adjustments are defined in Appendix D and both are
available in CAESAR II
4
.
An interesting aside here is that we still work with flexibility factors
in a stiffness package. Computerized pipe flexibility models were
originally based on the flexibility method rather than todays stiffness
method. They used to solve X=AF (where A is the flexibility matrix
of the piping system) while todays programs use F=KX as the basis
equation. Even so, the basic 12 x 12 elbow flexibility matrix as
found in the original Mare Island program (MEC 21) is still in use in
CAESAR II and probably all the other pipe flexibility and stress
analysis programs. The bend flexibility factor, k, adjusts the
coefficients in the flexibility matrix. This matrix is then inverted to
serve as the bend stiffness matrix used with the rest of the stiffness
model.
Application
Now so far, all of these calculations are automatically performed
for bends and miter joints and there is no need for explicitly setting
a bend flexibility factor. But Mr. Markl has another comment in his
paper Fatigue Tests of Piping Components
5
If the inherent flexibility
of a curved tube is to be fully utilized, a tangent length equal to at
least two pipe diameters, preferably greater, must be present between
its ends and adjoining flanges or anchors. The adjustment to k for
____________________
4
Specifying one or two attached flanges in the CAESAR II bend
definition will only change the stiffness and stress intensification
for the bend, CAESAR II will not automatically add flange weight
or a short, rigid section for the flange length itself. Likewise, rigid
elements connected to bends will not automatically adjust this bend
stiffness.
5
Also in Tube Turns Piping Engineering.
COADE Mechanical Engineering News October 2002
20
attached flanges addresses this issue but there are other situations
that may reduce the flexibility of the bend. Three situations come to
mind refractory-lined pipe, FRP pipe and bends with support
attachments. Refractory or brick, as an insulator in high temperature
piping, will certainly modify the bends ability to distort and flex.
In fact, direct entry of the flexibility factor was added to CAESAR II
at the request of users modeling refractory-lined pipe. Bends made
up of fiberglass reinforced plastic pipe also fail to respond like
homogeneous materials. One source states that The results showed
that the bend was up to a factor of 20 stiffer than what is found by
improper but common use of the (B31.3) standards.
6
And, It was
recommended that flexibility and stress intensification factors should
be measured and supplied by the manufacturers.
The third item, change in stiffness due to attachments, is common in
many systems where a dummy leg is welded to the bend extrados.
This attachment will certainly stiffen up the bend and some
CAESAR II users, for want of a better method, have modeled this
added stiffness by indicating a flange attached to the bend. Equations
are available for this added stiffness but, with todays easy access to
finite element methods, programs such as FE/Pipe
7
can also produce
a value for k. One equation in the literature appears in a paper
presented at the Seventh Annual Convention of SPED (Society of
Piping Engineers and Designers) in 1987 entitled Flexibility of
Trunnion Piping Elbows by G. D. Lewis of Duke Power and Yuh-
Jin Chao of the University of South Carolina. The paper produced
an equation for in-plane bend flexibility: ) / ( 81 . 0 D d k k
e
= ,
where k
e
is the standard bend flexibility factor, d is the trunnion pipe
diameter and D is the elbow pipe diameter. This equation is
qualified only where D d D < 2 and
T t T 2
(t is the
trunnion wall thickness and T is the elbow pipe thickness). If a 3.5
inch, 0.3 inch wall pipe is attached to the 6 inch standard elbow, the
flexibility factor reduces from 6.59 to 4.81. Compare this value to
the calculated flexibility of the clean elbow with one flange attached
5.232 and with two flanges attached 4.154.
Comparison
Simply to say that the flexibility factor changes from 6.59 to, say,
4.81 is not significant. However, if that same change to bend
flexibility greatly increases a thermal load on a nozzle, then it would
be important. The simple system shown below illustrates the change
in anchor load with different bend flexibilities calculated or specified.
In this example, the trunnion model which reduced the flexibility
factor by 27%, increased the bending moment by 19%.
____________________
6
See http://www.sintef.no/units/matek/2490/prosj/deep_water.htm
- the web site of SINTEF (The Foundation for Scientific and
Industrial Research at the Norwegian Institute of Technology)
7
FE/Pipe is a trademark of Paulin Research group
Component at 20 k Moment at 10
Short radius elbow 9.885 -3735 ft-lbf
Long radius (LR) elbow 6.590 -3920 ft-lbf
LR elbow with 1 flange 5.232 -4472 ft-lbf
LR elbow with trunnion 4.810 -4677 ft-lbf
LR elbow with 2 flanges 4.154 -5036 ft-lbf
Square corner n/a -7563 ft-lbf
Elbow with k=1 1 -7978 ft-lbf

Conclusion
CAESAR II Version 4.40 now provides direct specification of
bend flexibility. By understanding its application and effect on a
piping model, the user can decide if and when bend flexibility
should be explicitly defined.
The Pressure Equipment
directive (PED) and CE marking
By: Mandeep Singh
What is the PED?
The European Parliament and European Council adopted the Pressure
Equipment Directive (97/23/EC) on May 1997. After the 2 year
transition period, it became mandatory in May 29, 2002. Now, the
pressure equipment that is stated for installation in European
Economic Area (European Union plus Norway, Iceland and
Liechtenstein), must comply with the PED requirements. Equipment
(vessels, piping, assemblies etc) that complies with this directive
will attain a CE marking. Pressure equipment that has a design
pressure that is less than or equal to 7.25 psig (0.5 barg) is excluded
(as long as it is not a generic safety accessory).
This Directive arises from the European Communitys program for
the elimination of technical barriers to trade and is formulated under
October 2002 COADE Mechanical Engineering News
21
the New Approach to Technical Harmonization and Standards.
Its purpose is to harmonize national laws of Member States regarding
the design, manufacture, testing and conformity assessment of
pressure equipment and assemblies of pressure equipment.
Equipment with a CE marking is accepted any where in the European
Economic Area countries. There may be some additional local
restrictions (similar to state level restrictions in the USA).
The PED concerns manufacturers of items such as vessels,
pressurized storage containers, heat exchangers, steam generators,
boilers, industrial piping, safety devices and pressure accessories.
Such pressure equipment is widely used in the process industries
(oil & gas, chemical, pharmaceutical, plastics and rubber and the
food and beverage industry), high temperature process industry
(glass, paper and board), energy production and in the supply of
utilities, heating, air conditioning and gas storage and transportation.
What are the requirements of the PED?
The PED does not contain detailed technical requirements as in
ASME or PD 5500. It consists of Conformity Assessment procedures
that should be followed to demonstrate that the essential safety
requirements are met. These Conformity Assessment procedures
are based on the hazard category of the pressure equipment. Pressure
equipment is classified into hazard category 1,2,3 or 4, (in increasing
order), depending upon:
Type of equipment (vessel, piping, boiler)
Type of fluid (dangerous, non-dangerous)
Nature of fluid (gas, liquid)
Service condition (pressure, volume/nominal pipe size)
For a higher hazard category, the Conformity Assessment procedures
to be followed are more involved and usually require more
involvement of the notified body. These procedures involve
documentary checks, materials, design approval, quality assurance,
manufacturing etc. The European Union member states have
designated some semi-official or private technical organizations,
either for approval and monitoring of the manufacturers quality
assurance system or for direct product inspection. These
organizations are termed as notified bodies. Notified bodies must
verify and certify that the equipment complies with all the
requirements of Conformity Assessment procedure selected. Many
of the notified bodies are also ASME Authorized Inspection agencies
operating in USA and Canada.
How is the PED related to the design Code?
As mentioned earlier, the PED is not a design code. It permits the
use of any established National code and standard (ASME, PD
5500, CODAP, AD-M), provided all the provisions of the PED are
satisfied. If the equipment is designed per Unified European codes
(also referred as the Harmonized Standard) then it is presumed to
conform to the PED. Following are the EN Harmonized standards
related to this directive:
EN 13445 : Unfired Pressure Vessels
EN 13480 : Industrial Piping
EN 12 952 : Water Tube Boilers
EN 12 953 : Shell Boilers
Conclusion:
For US pressure equipment manufacturers, exporting to Europe has
been a major challenge. Traditionally, each European country used
to have independent pressure vessel/piping codes and standards,
making the exports across the borders costly and challenging
especially for smaller manufacturers. With the PED, there is one set
of requirements for exporting to over 18 European countries.
Manufacturers can also keep using their current code, utilize the
existing in-house QA procedures and work with the notified bodies
to upgrade the procedures that are required.
The margin of safety built in the ASME codes is typically higher
compared to most major European codes, but the production/QA
requirements are less stringent. Now for manufacturers exporting to
Europe having to satisfy the PED requirements, it will probably be
more economical to build per the unified European code.
Useful Urls:
1. PEDs official website http://ped.eurodyn.com/ (information
on PED, guidelines on PED, list of notified bodies)
2. Harmonized standards - http://europa.eu.int/comm/enterprise/
newapproach/standardization/harmstds/reflist/equippre.html
3. Some related directives - Simple pressure vessels (87/404/
EC), transportable pressure equipment (99/36/EC) and Aerosol
Dispensers (75/324/EEC).
PC Hardware/Software for the
Engineering User (Part 33)
Windows XP and Java
As a result of the court agreement between Microsoft Corporation
and Sun Microsystems, the latest operating system from Microsoft
(Windows XP) doesnt directly support Java applications. This
results in some applications being unable to run, and some webs
sites being rendered improperly when accessed from a computer
running Windows XP.
COADE Mechanical Engineering News October 2002
22
There is however a solution to this dilemma. On Suns web site at
http://java.sun.com/ there is a link (shown below) to download a
small plug-in. This plug-in will provide the missing pieces of
the operating system, allowing Windows XP to properly handle
Java applications.
CAESAR II Notices
Listed below are those errors & omissions in the CAESAR II
program that have been identified since the last newsletter. These
corrections are available for download from our WEB site, for
Version 4.40.
1) Accounting Module:
Changed the method used to get the date range off of the
dialog box. This avoided a crash when generating accounting
reports.
2) Large Job Printing Module:
Updated versioning information, to comply with new
4.40 file structures.
Corrected the labeling of the ChinaPower spring size.
3) Dynamic Output Module:
Corrected the Input Echo dialog box to include the new
node names option.
4) Static Solution Module:
Corrected stiffness matrix adjustment when using the new
spring hanger control options and switching from the
hanger status from rigid to ignore.
Corrected the stiffness for constant effort supports when
changing the hanger status.
5) Static Stress Computation Module:
Corrected the computation of allowables for DNV piping
code when more than a single thermal vector was specified
in the job.
6) Static Output Module:
Corrected printing of configuration parameters in the
input echo.
Corrected the labeling of the ChinaPower spring size.
Corrected the presentation of nodal displacement data in
the input echo.
Corrected the memory allocation for node names.
7) Static Load Case Setup Module / Dynamic Input:
Corrected the use of the revised number of spring hanger
design cases instead of the old number from existing ._J
files, when verifying load case data..
Corrected the upgrade of load cases when pre-440 jobs
included (XYZ) spring and rod type restraints.
Corrected an error message if reuse Eigen solution was
activated and the Eigenvalue file was missing.
Corrected an error where if the first run of a job was batch
run from input, the friction multipliers were not set, and
zero was used.
Corrected the import of pre-4.40 static load case definitions
when load cases incorrectly included undefined primitive
load components.
8) Piping Error Checker:
Corrected the display of warning 162.
Corrected the storage of the new bend flexibility factor
for included piping jobs.
Modified all error/warning/note messages to include the
letter E following the ID number. This is to avoid
confusion with fatal errors that may occur during other
operations.
9) Piping Input Module:
Corrected printing of configuration parameters in the
input echo.
Corrected the axis labeling on the LIST Duplicate
dialog for mirroring, when the Z axis is up.
Corrected the presentation of nodal displacement data in
the input echo.
Fixed a problem in the standard graphics where clicking
on the Spreadsheet button caused a crash.
Fixed a problem preventing usage when including piping/
structural files and global coordinates were set.
Changed the grid cell style for nodes 1 & 3 in the bend list
to display as integer values.
10) Buried Pipe Module:
Modified the generation of element lengths to avoid
extremely small elements.
October 2002 COADE Mechanical Engineering News
23
TANK Notices
Listed below are those errors & omissions in the TANK program
that have been identified since the last newsletter. These corrections
are available for download from our WEB site.
1) Solution Module:
Corrected the use of St for the test case in API-653
thickness determination.
Corrected the use of fluid height in the seismic routines to
avoid using the pressure increase.
For rafter supported cone roofs, initialized the value for
percent roof weight supported by the shell.
2) Main/Input Module:
Corrected the startup of the Material Database Editor.
Addressed the handling of zero-byte length user material
files.
CodeCalc Notices
Listed below are those errors & omissions in the CodeCalc program
that have been identified since the last newsletter.
1) ASME and TEMA Tubesheets:
Fixed the selection of design optimization option in U-
tubesheet design.
(Re)added input for user specified bolt root area.
For the fixed tubesheet design with the multiple load cases
selected, when the tube corrosion input field was left blank,
the tube corrosion allowance was not being correctly applied.
This has been fixed now.
2) Nozzles:
Corrected the UG-45 computation for a nozzle on a flat
head with user supplied req. thk.
Corrected pad width for hillside nozzle.
Fixed some issues in the large nozzle (App 1-7) bending
calculations.
Corrected the nozzle reinforcement calculations for some
cases where the reinforcement limit cut through welds.
3) Horizontal Vessels:
Added inputs for sliding saddles friction coefficient, user
longitudinal load, user wind pressure and saddle reaction
force factor.
Corrected the moment used in the AISC unity check on the
saddle ribs.
4) Shell:
Corrected the weld size status bar message.
5) WRC 107:
Corrected the mean shell diameter value used.
Printout/Graphics/Nozzle Pro interface Enhancements.
6) WRC 297/Annex G:
Corrected some graphs and modified the stress combination
method for a sphere for Annex G calcs.
7) Leg & Lug:
Corrected the Wind Exposure category selection.
PVElite Notices
Listed below are those errors & omissions in the PVElite program
that have been identified since the last newsletter. These corrections
are available for download from our WEB site.
1) Several changes were made for PD:5500 Closely Spaced Nozzles.
The program was showing that many were close when they were
not.
2) A ft. in. units problem that resulted in the stress due to overturning
by items such as nozzles were off by the conversion factor of 12.
3) The Aspect Ratio for 2:1 Elliptical heads for PD:5500, Nozzle
Analysis was taken based on inside dimensions and not outside
dimensions.
4) The value of Ft, the top load in UBC 97 was not divided by 1.4.
5) For cones with knuckles, the program was using the cone
thickness and not the knuckle thickness to compute moment of
inertia and cross sectional area.
6) Held up packing liquid was not being considered in seismic
calculations.
7) The moments used in the cone calculations may have been one
element off in some cases. Other miscellaneous fixes to the cone
calculations involving moments at the large and small end were
made.
8) For ASME nozzles when the diameter and thickness limits cut
the nozzle/shell weld the weld area calculations may have been
in error.
COADE Mechanical Engineering News October 2002
24
12777 Jones Rd. Suite 480 Tel: 281-890-4566 Web: www.coade.com
Houston, Texas 77070 Fax: 281-890-3301 E-Mail: [email protected]
COADE Engineering Software
COADE Sponsors Autodesk World One Team Conference and Autodesk University
This year, COADE has joined Hewlett Packard, Intel, Microsoft, IBM, and nVidia as sponsors of Autodesks two premier
conferences the World One Team Conference and Autodesk University, held September 11-13 and December 3-6, 2002
respectively, at the MGM Grand Hotel in Las Vegas, Nevada.
Whereas One Team Conference is conducted for the benefit of Autodesks dealers, distributors, and development partners,
Autodesk University is, as its name implies, four days of courses, offering users of Autodesk products the chance to improve
their skills and understanding of both software and the industries they serve.
COADE and its CADWorx product figure prominently in both conferences. In the above picture taken at One Team Conference,
Jim Meyer, Autodesks Director of Strategic Marketing, presents a sponsors recognition award to COADEs John Brinlee, Tom
Van Laan, and Bill Evans (left to right). This event gave COADE a good chance to visit with many of our dealers from throughout
the world, as well as meet new ones interested in representing COADE. At Autodesk University, COADE personnel are
scheduled to lead two seminars dealing with the plant design industry, as well as a tutorial on plant design using CADWorx. This
will give us further opportunity to speak to many of our long-term (and newer) users.
Wed like to see you all come out and visit us at Autodesk University!

You might also like