Ductile-Regime Grinding
Ductile-Regime Grinding
Ductile-Regime Grinding
Bifano
Department of Aerospace and Mechanical
Engineering,
Boston University,
Boston, MA 02215
T. A. Dow
Department of Mechanical and
Aerospace Engineering,
Precision Engineering Center,
North Carolina State University,
Raleigh, NC 27695
R. 0. Scattergood
Materials Engineering Department,
Precision Engineering Center,
North Carolina State University,
Raleigh, NC 27695
Introduction
When machined, a brittle material can deform through a
The transition from brittle to ductile material-removal at
variety of mechanisms. If the critical resolved shear stress at smaller cutting depths can be argued purely from consideraany point within the material exceeds the elastic yield stress, tions of material-removal energy. Specifically, for lower mathe mechanism of deformation will change from one of re- chining depths-of-cut, it can be shown that plastic flow is a
versible energy storage via elastic stretching to one of irrever- more energetically favorable material-removal process than
sible energy dissipation. Examples of irreversible deformation fracture. The material property characterizing resistance to
include macroscopic fracture propagation, microcrack for- plastic flow is the yield stress, ay. The energy (Ep) required to
mation, phase transformation, dislocation motion (in crystals), plastically deform a specified volume of material (Vp) can be
and intermolecular sliding (in amorphous materials). Irrever- written as:
sible material-removal mechanisms can be divided into two
EP = o V
(1)
types: brittle and ductile. In brittle mechanisms, material removal is accomplished through the propagation and intersec- The material property characterizing resistance to fracture is
tion of cracks, while ductile mechanisms produce plastic flow the Griffith crack propagation parameter, G. The energy (Ef)
of material in the form of severely sheared machining chips. required for fracture is a function of the area {Af) of new
Recent improvements in machining tolerances have exposed surface created by crack propagation. Thus:
a new possibility for material-removal from brittle substances.
Ef = GAf
(2)
It has been noted that plastically deformed chips are formed
For
a
machining
depth
d,
it
is
reasonable
to
assume
that
the
in the machining of brittle substances if the scale of the machining operation is small (less than 1 Lim depth of cut) (Toh order of magnitude of both Vp and af are determined by d.
McPherson, 1986). Similar ductile chip-formation has been That is:
observed in fine scale machining debris from a wide range of
VP ~ <?
0)
ceramics, glasses, and crystals (Huerta and Malkin, 1976; Bi2
fano et al, 1987; Yoshioka et al, 1984; Molloy et al, 1987).
Af ~ d
(4)
This suggests that the process of ductile chip formation may The ratio of material-removal energies, then, is given by
be independent of nature of the material (e.g, brittle or ductile,
Plastic Flow Energy
hard, or soft, crystalline or amorphous, etc.). Grinding of
(5)
brittle materials under conditions that allow predominantly
Fracture Energy
jzf
ductile material-removal is a new technology known as ductileConsequently, as the scale of machining decreases, plastic flow
regime grinding, which is the subject of this paper.
becomes an energetically more favorable material-removal
mechanism. The specific depth at which a brittle-ductile tranContributed by the Production Engineering Division for publication in the sition occurs is a function of the intrinsic material properties
JOURNAL OF ENGINEERING FOR INDUSTRY. Manuscript received May 1990.
governing plastic deformation and fracture.
184/Vol. 113, MAY 1991
Consideration of this energy argument leads to a generalization that we named the "Ductile-Regime Grinding Hypothesis." This hypothesis states that for any material, if the
dimensional scale of material-removal is made small enough,
material-removal will proceed by a mechanism of plastic flow
and not fracture. A grinding apparatus capable of achieving
this ductile regime of material-removal can be used to combine
the fine tolerances that are achievable in a deterministic grinding process with the superfinished surfaces that are achievable
in a nondeterministic polishing or lapping process.
In grinding, the "scale" of material-removal is characterized
by the depth-of-cut imposed on a given abrasive grit. For
plunge grinding, this parameter is determined by the grinding
infeed rate. Therefore, the hypothesis implies that in plunge
grinding, there will be a critical-grinding-infeed-rate, below
which no fracture will occur for a given brittle workpiece
material.
One way of viewing the ductile-regime grinding problem is
that there is a challenge, first described by Miyashita, to fill
a gap in the achievable material-removal rate in abrasive machining (Miyashita, 1985). If the currently achievable materialremoval rates for grinding and polishing are compared, there
is a gap in which neither technique has successfully been utilized. This region of material-removal has been termed the
mircogrinding gap. For grinding processes, material removal
is accompanied by localized fracture for virtually all brittle
materials (Molly et al, 1987; Huerta and Malkin, 1976; Yoshoika et al, 1982). Polishing processes, on the other hand, result
in material removal without fracture. The importance of this
material-removal rate gap, then, is that it represents the threshold between ductile and brittle grinding regimes for a wide
range of glasses, ceramics, and semiconductors (Bifano, 1988).
This paper describes a research effort to characterize the
physical parameters that control the brittle-to-ductile transition
in the grinding of brittle materials. The results of this investigation lend strong evidence in support of the ductile-regime
grinding hypothesis. Also, a model is proposed, defining the
brittle-to-ductile transition in terms of the material properties
of the workpiece and the rate of material removal.
In this paper, three aspects of the research are described:
9
Development of a grinding apparatus capable of ductileregime grinding on brittle materials. The relevant design
features include:
Machine configuration
Machine Stiffness
Infeed control system
Wheel trueing techniques
Experimental evaluation of the grinding apparatus, in
terms of its capacity for ductile-regime grinding on various
materials
Formulation and evaluation of a ductile-regime grinding
model.
Fig. 1
through the three stages: plastic grooving, generation of median and lateral cracks, and finally crushing (Swain, 1979). In
this study, it was demonstrated that the progression of material-removal mechanisms was directly related to the force on
the abrasive grain, with lower forces corresponding to a decrease in the observed surface fracture. Other experiments of
single grit abrasion tests on myriad brittle materials including
glasses (Schinker and Doll, 1987; Molloy et al., 1987; Yoshioka
et al, 1982), semiconductors (Danyluk, 1986), and advanced
ceramics (Swain, 1979; Toh and McPherson, 1986) demonstrated similar transitions in the material-removal process as
a function of the grinding force or depth of cut.
The first grinding apparatus specifically designed to take
advantage of the relationship between small grinding infeeds
(0.2 (im per pass on a surface grinder) and improved surface
finish was built by Yoskioka (Yosioka et al., 1982). With this
apparatus, ductile-regime grinding was demonstrated for several brittle, crystalline materials. Later improvements in the
precision of this grinder were shown to translate directly into
improved surface finish on the brittle workpieces (Yoshioka
et al., 1985). It was from this research effort that the concept
of the microgrinding gap evolved. While the correlation between higher grinding precision and enhanced grinding ductility was qualitatively demonstrated by these Japanese efforts,
quantitative relationships between machine parameters, material properties, and grinding ductility have yet to be established.
The Ductile-Regime Grinding Apparatus
Machine Configuration and Actuation Mechanisms. As an
initial design criterion, it was decided that a ductile-regime
grinder must provide a union of high rigidity (to ensure precision) and low infeed rates (to ensure ductile material-removal). From Yoshioka's early attempts to grind hard materials
in the ductile regime, it was established that a grinding infeed
resolution of ~ 50 nm or better is required to prevent significant
surface fracture damage in glass (Yoshioka et al, 1982). This
represents a level of machine precision that is not often associated with the grinding process. Such motion accuracy, in
turn, demands an extremely rigid structural loop so that disturbance forces experienced by either the workpiece or the
grinding wheel will not be translated into significant relative
motion between the two.
A schematic of the device used in this study is illustrated in
Fig. 1. This test bed has been given the acronym PEGASUS
(Precision Engineering Grinding Apparatus for Super-finishing Ultrahard Surfaces). The machines provides a mechanism
for plunge grinding with the 6 mm wide rim of a 100 mm
diameter cup shaped grinding wheel. The workpieces to be
ground are rectangular parallelepipeds measuring 6 mm x 6
mm x 18 mm.
MAY 1991, Vol. 113/185
Fig. 2
GRINDING
MICROGRINDING
..........------------------------------ .........--
..
10
10
10
10
.,
POLISHING
80
::l
c:;
at'"
60
Q)
'":;
.ll!
PEGASUS
40
tJl
10
10
..
10
.,
.,
10
20
0
.0001
.001
.01
.1
Fig. 3
SUS
FIRST PASS
TH IRD PASS
FOURTH PASS
75
150
225
300
375
450
525
600
675
Degrees of Revolution
Summary of Microgrinding Apparatus Design Features. As illustrated in Fig. 3, the microgrinding apparatus
described in this chapter extends well into the "microgrinding
gap" described previously. Due to its piezoelectric infeed system and its plunge-grinding configuration, PEGASUS is capable of grinding with controlled cutting depths as small as 2
nm. This has proven to be the quintessential requirement for
ductile-regime grinding of brittle materials.
(b)
ysis tool, the ground material surfaces were examined for evidence of surface fracture, which would be indicative of the
grinding ductility. A grid counting technique was devised to
quantify the real percentage of surface fracture. By applying
MAY 1991, Vol. 113 J 187
juuin
9 Alumina
10 Toughened zirconia
A9
000
5A
a.
AS
1
2
3
4
5
6
7
8
A j>
100i
k /
3
2
o
10-
i /
"i
10
100
Fused silica
Soda iime glass
Zerodur
Quartz
Zirconia
Germanium
Silicon
Silicon carbide
1ii
i i i i
1000
ER
H2
dr =
0)
Q
a>
ta
A 10
(6)
where dc is the critical indentation depth, E is the elastic modulus, R is the material's fracture energy, and His the hardness.
For materials that exhibit a plastic zone near the crack tip, the
value of R can be evaluated using Griffith's classical crack
propagation analysis. One approach to defining fracture energy at small scales is to replace it with a dimensionally analogous measure of the energy needed to propagate cracks,
namely:
R ~ *
H
(7)
*" I
as a measure of the brittle transition depth-of-cut. Experimental results using this formula for indentation testing have
shown a remarkable degree of consistency, even using bulk
material properties (Marshall and Lawn, 1986). If this type of
analysis were adopted for grinding, dc would represent a measure of the critical-grinding-infeed-rate expected to change the
material-removal mechanism from a ductile regime to a brittle
one. Thus, presumably a series of brittle materials could be
ranked according to their properties to determine the grinding
wheel infeed rate below which fracture would not occur.
The relevant properties were measured for each material
using microindentation techniques (Bifano, 1988). While the
measurement of hardness by indentation is a standard procedure, determination of Kc and E by indentation is a developing area of research (Marshall and Lawn, 1986). The
properties of the material surface vary with the indentation
depth at which they are measured. This surface property variability is especially troubling for the measurement of Kc. Sizescale effects lead to a dependence of Kc on crack size (R-curve
behavior), which can be a large effect in certain materials
(Scattergood et al, 1988). Such material-related property variations complicate the extrapolation of properties from the
scale of indentation (~ 10 jum) to the scale of microgrinding
( < 1 /an). In spite of these problems calculating the material
properties, a reasonable correlation was obtained between the
calculated critical-depth-of-cut and the measured criticalgrinding-infeed-rate (i.e., the grinding infeed that will produce
10 percent surface fracture). This correlation is illustrated by
the graph of Fig. 6. From this correlation, the constant of
proportionality for eq. (8) can be estimated, yielding:
(9)
H
Only materials exhibiting significant variations in Kc with
indentation depth were not well represented by the model. For
these materials, estimates of the Kc behavior at small indentations can be used to modify the calculated critical-grindinginfeed-rate (Bifano, 1988).
0J5
[H
Summary
The following is a summary of the results brought out in
this paper.
Transactions of the ASME
9
9
Ductile-regime grinding is a newly established materialremoval technique. By controlling a stiff, accurate grinding apparatus so that it has an exceptionally small scale
of material removal, brittle materials can be ground in a
ductile manner. As a result, brittle workpieces can be
machined in a deterministic process while producing surface finishes characteristic of those achieved in nondeterministic, inherently ductile processes such as lapping
and polishing.
Ductile-regime grinding can be achieved by ensuring that
the grinding apparatus has a stiff structural loop, realtime control of the grinding infeed, relative isolation from
environmental disturbances, and state-of-the-art wheel
trueing techniques.
All brittle materials will undergo plastic flow rather than
fracture if the depth of machining is small enough.
There is a correlation between the grinding infeed rate
that corresponds to the brittle-to-ductile transition for a
particular brittle material and the properties (Kc, H, and
E) of the material. This correlation is reasonably described
by a simple power-law equation.
Acknowledgments
This work was supported by a University Research Initiative
grant from the Office of Naval Research, and by the Affiliates
of the Precision Engineering Center of North Carolina State
University.
References
1 Bifano, T. G., 1988, "Ductile-Regime Grinding of Brittle Materials,"
Ph.D. Thesis, NC State University, Raleigh, NC.
2 Bifano, T. G., and Dow, T. A., 1985, "Real Time Control of Spindle
Runout," Optical Engineering, Vol. 24, No. 5.
3 Bifano, T. G., Blake, P., Dow, T. A., and Scattergood, R. O., "Precision
Machining of Ceramic Materials," Proc. of the Intersociety Symposium on the
Machining of Ceramic Materials and Components, R. E. Parks, K. Subramanian, K. Ball, eds., Am. Cer. S o c , ASME Abras. Eng. S o c , pp. 99-120, April,
1987 (Updated and Reprinted in American Ceramic Society Bulletin, June, 1988,
Vol. 67, No. 6, pp. 1038-1044.
4 Chandrasekar.S., andSathyanarayanan.G., 1987, "An Investigation into
the Mechanics of Diamond Grinding of Brittle Materials,'' 15th North American
Manufacturing Research Conference Proceedings, Vol. 2, Manufacturing Technology Review, pp. 499-505.
5 Danyluk, S., 1986, "Smoother Scribing of Silicon Wafers," NASA Tech
Briefs, September/October.
6 Dow, T. A., Bifano, T. G., and Cagle, C. M. 1989, "Spindle Error
Compensation," Proceedings of the InternationalMachine ToolResearch Forum,
Chicago.
7 Huerta, M., and Malkin, S., 1976, "Grinding of Glass: The Mechanics
of the Process," ASME JOURNAL OF ENGINEERING FOR INDUSTRY, May, pp.
459-467.
8 King, R. F., and Tabor, D., 1954, "The Strength Properties and Frictional
Behaviour of Brittle Solids," Proc. of the Roy. Soc. London, A223, p . 225.
9 Lawn, B. R., Jensen, T., and Aurora, A., "Brittleness as an Indentation
Size Effect," J. Matl. Sci. Lett., Vol. 11, p. 575, 1976.
10 Marshall, D. B., and Lawn, B. R., 1986, "Indentation of Brittle Materials," Microindentation Techniques in Materials Science and Engineering, ASTM
STP 889, P. J. Blau and B. R. Lawn, eds., ASTM, Philadelphia, pp. 26-46.
11 Miyashita, M., 1985,1st Annual Precision Engineering Conference, North
Carolina State Univ., Raleigh, NC.
12 Molloy, P., Schinker, M. G., and Doll, W., 1987, "Brittle Fracture Mechanisms in Single-Point Glass Abrasion," Intl. Tech. Symp. on Optical and
Electro-Optical Appl. Sci. and Eng., The Hague, NL, (SPIE Vol. 802).
13 Scattergood, R. O., Srinivasan, S., and Bifano, T. G., 1988, "R-Curve
Effects for Machining and Wear of Ceramics," 7th International Symposium
on Ceramics, Bologna, Italy.
14 Schinker, M. G., and Doll, W., 1987, "Turning of Optical Glasses at
Room Temperature," Intl. Tech. Symp. on Optical and Electro-Optical Appl.
Sci. and Eng., The Hague, NL, (SPIE Vol. 802).
15 Swain, M. V., 1979, "Microfracture About Scratches in Brittle Solids,"
Proc. Roy. Soc. London, A366, pp. 575-597.
16 Toh, S. B., and McPherson, R., 1986, "Fine Scale Abrasive Wear of
Ceramics by a Plastic Cutting Process," Science of Hard Materials, Inst. Phys.
Conf. Serf. No. 75, Chap. 9, Adam Hilger, Ltd., Rhodes, pp. 865-871.
17 Yoshioka, J., Koizumi, K., Shimizu, M., Yoshikawa, H., Miyashita, M.,
and Kanai, A., 1982, "Surface Grinding with Newly Developed Ultra Precision
Grinding Machine," SME Technical Paper MR82-930.
18 Yoshioka, J., Miyashita, M., Hashimoto, F., andDaitoh, M., 1984, "High
Precision Centerless Grinding of Glass as a Preceding Operation to Polishing,"
SME Technical Paper MR84-542.
19 Yoshioka, J., Hashimoto, F., Miyashita, M., Kanai, A., Abo, T., and
Daito, M., 1985, "Ultraprecision Grinding Technology for Brittle Materials:
Application to Surface and Centerless Grinding Processes," Milton C. Shaw
Grinding Symposium, R. Komanduri, D. Maas, eds., ASME Production Engineering Division, Vol. 16, pp. 209-227.