You are on page 1of 7

ARTICLE IN PRESS

Tribology International 39 (2006) 1421–1427


www.elsevier.com/locate/triboint

Analysis of the stress field in a threshold-galling test


R.A. Waite, S.R. Hummel, A. Herr, G. Dalton
Department of Mechanical Engineering, Lafayette College, 256 Accopian Engineering Center, Easton, PA 18042, USA
Received 2 June 2005; received in revised form 28 December 2005; accepted 2 January 2006
Available online 29 March 2006

Abstract

The wear phenomena of galling is characterized by macroscopic roughening and deformation of the surfaces of mated materials.
Currently, the ASTM G98 button-on-block test method is the only standardized testing procedure for determining the galling resistance
of material couples. The pattern of surface damage resulting from performing the ASTM test tends to be predicable. In particular, the
galling generally occurs near the outer edge of the button and migrates inwards. In this study, the stress distribution between the mated
specimens leading up to the onset of galling was investigated. A suitable analytical model of the system geometry and loading conditions
was developed and applied. The results of the model show that a stress concentration exists along the edge of the button specimen when
loaded. The stress concentration is believed to contribute to the initiation of galling at this site. The results of the analytical model have
also shown that the magnitude of the stress concentration is dependent on the size of the corner radius. An alternative testing
arrangement is presented and preliminary results are shown.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Stress concentration; Contact stress; Adhesive wear; Galling

1. Introduction aforementioned standard test method to rank material


couples by their ability to resist galling. Therefore, in
Galling is a severe form of adhesive wear that occurs accordance with the current ASTM standard, galling will
when relative motion exists between contacting metal be considered based on the presence or absence of
components. The ASTM terminology standard G40 characteristic surface damage.
defines galling as ‘‘a form of surface damage arising Galling is a concern in applications such as threaded
between sliding solids, distinguished by macroscopic, connections. For example, in oil exploration galling is a
usually localized, roughening and creation of protrusions problem in the drill collars that are prone to galling when
above the original surface’’ [1]. Although galling is the threaded box and pin connection is released after being
recognized as an undesirable severe form of wear, experts assembled at a high torque [5]. Another common situation
in the field of tribology have differing opinions on the where galling effects performance is in inadequately
definition of galling versus other forms of wear. lubricated valves [6]. Valves are often designed with
There have been attempts to quantify galling; that is to surfaces that slide past one another. Another example is
develop a method of measuring the extent of galling that the sheet metal forming process, where metal sheets slide
has occurred and determine the relationships among the over dies [7]. The medical field also has applications where
various factors that contribute to causing galling [2,3]. galling is a potential problem. Stainless steel surgical
Previous theory and experimentation, including the ASTM instruments such as drill bits, pins and plates are
standard G98 the button-on-block test, have classified particularly susceptible due to the lack of adequate
galling as a binary mechanism, meaning that it is either lubrication [8].
present or it is not [4]. Design engineers have been using the Design engineers most useful tool to help reduce the
occurrence of galling is the choice of material [8]. Material
Corresponding author. Tel.: +1 610 330 5587; fax: +1 610 330 5059. selection is often based on the results of the current ASTM
E-mail address: hummels@lafayette.edu (S.R. Hummel). galling resistance test standard G98 [4]. The current

0301-679X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.triboint.2006.01.002
ARTICLE IN PRESS
1422 R.A. Waite et al. / Tribology International 39 (2006) 1421–1427

standard test method requires that the end of a cylindrical the specimens are visually inspected for the presence of
button be placed against a flat plate, as shown in Fig. 1. galling. The load is then increased until one of the
The specimens are then loaded via a universal tensile specimens experiences galling. Couples of materials can
testing machine or by any other constant force apparatus then be ranked based on the maximum applied load.
and the button is slowly turned one revolution with a One of the steps in the engineering design process is
wrench. The specimens are subsequently removed from the developing working models and prototypes to be tested
machine and are visually inspected for surface damage in under operating conditions. These models typically yield
the form of galling. If galling has not occurred, new more accurate galling test results because they are designed
specimens are placed in the machine and a higher load is specifically for the given application. An example of this is
applied. Following this procedure for a variety of different a threaded connection-galling test. A bolted joint design
combinations of materials, material couples can be ranked was tested by inserting the bolts into the desired config-
qualitatively for their resistance to galling. uration, and torquing them to the specified level [15]. The
This test has been used for several decades to determine bolts were then loosened and completely removed. This
relative galling resistance. Test results can be found in the was repeated until either galling had occurred, or the
literature for material couples including aluminum [8], threads were severely damaged. The testing procedure
stainless steel [9,10] and titanium [11], among others [6]. required that the bolts have a life of 900 cycles. The factors
The button-on-block test is useful in design applications in under investigation were material, lubrication and thread
that it can help in the selection of materials, however, type.
prototypical tests that more closely simulate the applica- Although several galling resistance test methods have
tion configuration are necessary to quantitatively deter- been developed, the ASTM G98 button-on-block test
mine galling thresholds [4]. While the ASTM standard G98 arrangement remains the only standardized procedure for
is the only current standard testing procedure, other ranking galling resistance. The results of this standard test
methods have been developed. These tests investigate are useful in determining the relative ranking of material
different aspects of contact wear that can be found to couples; however, as Hummel and Partlow [10] point out,
exist in machinery, structures, and equipment [6]. there can be discrepancies between inter-laboratory results.
The pin-on-flat testing arrangement has been used to Specifically, it was found that when two different labs ran
determine the galling resistance of material couples [12]. In tests on the same material couples, the galling threshold
this method, a spherically tipped pin slides along a straight could vary by as much as 500%. The present ASTM galling
line across a flat block. The testing method requires a resistance test standard G98 states that the ‘‘subjective
predetermined sliding distance, speed, and surface finish of determination of threshold galling stress by visual exam-
the specimens. The surface topography is measured by a ination makes it difficult to have high precision in test
stylus profilometer in parallel sections along the length of results from different laboratories’’ [4].
the line of sliding. The topographical data is used to Determining the underlying cause of the lack of
determine a parameter that describes the average maximum repeatability is the primary focus of the current study.
peak-to-valley distances. This parameter is then used to There are many factors that researchers have found that
correlate galling damage to the surface profile of the affect galling resistance. These factors include surface
specimens. finish, sliding distance, geometry, and sliding speed
The button-on-cylinder test arrangement was developed [3,6,8]. While many of the factors that effect galling are
by Hummel to investigate galling of components in line not fully understood, the lack of repeatable results using
contact [13,14]. This test requires that the flat end of a the current testing method suggests that the test method
stationary button be in tangential contact with a cylinder. needs to be examined. The purpose of this investigation
A load is applied to the button, and the cylinder is rotated was to take the initial step in understanding the specific test
a prescribed number of cycles. After the test is performed, conditions during a button-on-block test. In particular the
stress field of the contact region was examined to determine
if it plays a significant role in the galling phenomena during
a button-on-block test.

2. Current procedure

The galling resistance test uses two specimens, a button


and a block. The testing arrangement is shown in Fig. 1.
The critical dimension of the button is the diameter of the
cylindrical portion, because it is the characteristic dimen-
sion of the contact area. The block is a flat plate, generally
having larger dimensions than the button so that it may be
used for multiple tests. Specimen materials vary so that
Fig. 1. Button-on-block configuration. different combinations of materials can be investigated.
ARTICLE IN PRESS
R.A. Waite et al. / Tribology International 39 (2006) 1421–1427 1423

The ASTM standard test for galling resistance begins


with the cleaning of the specimens. It is important that the
specimens are cleaned directly prior to performing the test,
so that any scale, oil film, or other debris is removed from
the surfaces of interest. ASTM provides suggested methods
of cleaning the specimens that will not be discussed here.
The specimens are then mounted in the testing machine,
are lightly loaded, and the button is carefully rotated to
ensure that the flat surfaces are in contact and the surfaces
are seated.
The next step is to apply the desired load, and rotate one
of the specimens, usually the button, one full revolution.
The standard states that the rotation can be performed
manually using a wrench or other type of tool. The time for
the rotation should range between 3 and 20 s. Upon
completion of the test, the specimens are visually inspected
for the presence of galling without the assistance of optical
magnification. If the specimens remain smooth and
undamaged, then the procedure is repeated at higher loads
with new specimens until galling occurs. Galling is
characterized by the appearance of ‘‘macroscopic rough-
ening and raised protrusions above the original surface on
one or both of the specimens [1]’’. Several photographs of
galled and non-galled specimens are shown in the standard
as a guideline to aide users in determining the presence of
galling.
Upon completion of the testing, the threshold galling
stress can be calculated by averaging the highest stress
where no galling occurred with the lowest stress where
galling did occur. These stress values are based upon a
normal uniform stress distribution across the surface of the Fig. 2. Typical ASTM G98 test results.
button. Specifically, the stress is calculated by simply
dividing the applied force by the apparent contact area.
the galling. The second factor, the stress distribution, seems
One should note that this stress is the average stress across
unimportant when calculating the threshold galling stress
the contact area and does not represent the local stress. The
because this calculation implies a uniform stress distribu-
local stress field is discussed in subsequent sections of this tion; however, the stress distribution is not uniform. In
paper. It is important to note that the ASTM standard G98
fact, the contact conditions in the button-on-block test
inherently assumes that the highest non-galled stress will be
cause a stress concentration to exist along the perimeter of
less than the highest galled stress; however, this is not
the button. The location of the stress concentration occurs
always the case [10].
at the point of maximum sliding distance and the area
A common characteristic of the galled specimens is the
where galling is most often found in G98 testing. In the
pattern that the galling occurs. In virtually all test
following section the non-uniform stress distribution is
specimens that galling has occurred, the damage is along
modeled, using both an analytical model and a finite
the perimeter of the button. This can be seen in the element model. These results are then compared to
photographs included in the ASTM G98 standard [4],
experimental results.
as well as in specimens tested in this study and shown in
Fig. 2. The consistency of this galling pattern suggests that
there is some factor responsible for the location of the 3. Analytical model
galling damage.
Two factors that likely play a central role in causing the In a recent study, Ciavarella et al. [16] mathematically
galling pattern around the perimeter are the sliding modeled the stress field of an elastic punch on an elastic
distance and the stress magnitude. Jarrell and Bejbl found thick plate. The model incorporated rounded edges on the
that increasing the sliding distance generates higher punch. This model closely simulates the contact conditions
frequency of galling [8]. In the button-on-block test, sliding seen in the G98 button-on-block test and was used in the
distance increases radially from the center of the button. current study to estimate the contact stresses.
This means that the maximum sliding distance occurs at The study performed by Ciavarella et al. is based on the
the outer edge of the button, coinciding with the location of geometric configuration shown in Fig. 3, where a is the
ARTICLE IN PRESS
1424 R.A. Waite et al. / Tribology International 39 (2006) 1421–1427

P was assumed to be 12.7 mm and the material properties


were assumed to be that of steel. The applied load was set
to 4500 N and the corner radius of the button was assumed
to be 0.3 mm. The resulting stress concentration results can
be seen graphically in Fig. 4. From this figure it is clear that
the mathematical model does not predict a uniform stress
x distribution.
It is important to note that the corner radius that was
-b -a a b used in developing Fig. 4 is not specified by the ASTM G98
y [4] standard. The standard makes no mention of a corner
radius and it is implied from the specimen drawings in the
Fig. 3. Geometric schematic of flat punch model [16]. standard that corner should be sharp and unbroken.
However, the mathematical model developed by Ciavarella
dimension of the flat region of the punch and b represents et al. [16] requires a non-zero value for the corner radius. In
the diameter of the contact area. fact due to the assumptions made in the development of the
The pressure distribution is given as a non-dimensiona- model, the corner radius must be of significant size in order
lized function, as for the model to yield accurate results. For the purpose of
( modeling the button-on-block test, the corner radius is
bpðfÞ 2=p large enough to yield accurate results from the mathema-
¼  ðp  2f0 Þ cosðfÞ
P p  2f0  sinð2f0 Þ tical model and it is small enough that it does not
"  significantly reduce the apparent contact area between
sinðf þ f0 ÞsinðfÞ
þ ln   the specimens.
sinðf  f0 Þ The graph shown in Fig. 4 shows that the stress
    sinðf0 Þ #) concentration near the perimeter of the button approaches
 f þ f f  f 
 tan 0
tan 0 
 , ð1Þ three for a corner radius of 0.3 mm. One can easily show
2 2
that by changing the corner radius to larger and larger
where f0 is defined as a measure of the contact zone and is values, the stress concentration decreases as one would
given by the transcendental function, expect. Unfortunately, due to the limitations of the
mathematical model one cannot show that the stress at
2PR p  2f0 cotðf0 Þ
 ¼  , (2) the corner goes to infinity as the corner radius vanishes.
2
aE 2 2
4sin ðf0 Þ However, it can be shown that the stress at the corner does
go to infinity using a Hertzian mathematical model
1 1 1 assuming a rigid button [17].
n ¼ ð1  n21 Þ þ ð1  n22 Þ, (3)
E E1 E2 A finite element model (FEM) of the contact conditions
where P is the applied normal load, R the corner radius of of the ASTM test was developed to confirm the non-
the punch (button), and E* the composite stiffness of the uniform stress distribution predicted by the mathematical
two materials under plane strain conditions, where E and n models. The FEM is shown schematically in Fig. 5, where
are the material properties of Young’s modulus and the upper rectangle is the steel button and the lower
Poisson’s ratio of the two mated materials. The position
along the diameter of the punch was quantified using the
Stress Concentration along Diameter of Button
following equations, (1.3mm Corner Radius)
sinðfÞ 3.0
x¼a , (4)
sinðf0 Þ
2.5
Stress Concentration

a
b¼ . (5) 2.0
sinðf0 Þ
The mathematical model of the stress field presented 1.5
above was used to predict the stress concentrations during
1.0
a button-on-block test. The results of the findings are
presented in the following section.
0.5

4. Discussion 0.0
-1 -0.5 0 0.5 1
Using Eqs. (1)–(5), the stress concentration at the Radial Position (Normalized)
specimen interface was determined for what is considered Fig. 4. Plot of stress concentration factors along the diameter of the
to be typical test. Specifically, the diameter of the button punch.
ARTICLE IN PRESS
R.A. Waite et al. / Tribology International 39 (2006) 1421–1427 1425

Pressure F

Button

Block

Fig. 5. Schematic of the finite element model.

rectangle is the steel block. Axisymmetric boundary


conditions were used along the vertical axis. A pressure Fig. 6. Schematic of new test configuration.
of 34.5 MPa, representative of the 4500 N applied load
assumed in the mathematical model, was applied to the top
of the button. The material was modeled using a bilinear Upper specimen
stress–strain curve to model plastic yielding should a
singularity be encountered. Additionally, contact elements
were used to model the contacting surfaces of the speci- Alignment pin
mens. As expected, the FEM predicted that the highest
stress occurs at the outer edge of the button.
As mentioned previously, two of the factors that play a
central role in causing galling to occur are sliding distance
and stress magnitude. From the mathematical models and Contact Region
the finite element model it is clear that a stress concentra-
tion exists near the perimeter of the button during the
ASTM G98 test. This stress concentration will cause
the local stress value to be approximately three times the Lower specimen
nominal stress value or higher. Moreover, the location of Fig. 7. Section view of new test configuration.
this stress concentration coincides with the location on the
specimens where the relative distance traveled is greatest.
These two factors appear to have a significant impact in
causing galling to originate most often along the perimeter is spherically shaped so when the top cylinder is in position,
of the button during galling testing. In test conducted by the sphere guides the specimens ensuring that the outer and
the author as well as in the photographs located in the inner edges of the specimens are concentric. By ensuring
ASTM standard one can see that the galling typically that the edges of both specimens coincide, the lines of
originates at the edge of the button. constant force running through the specimens from the
applied load to the machine base are continuous and
straight from one specimen to the next. Additionally, the
5. New setup spherical end on the alignment pin ensures that the contact
surfaces of the specimens are in area contact throughout
A new testing configuration is currently being investi- the entire cycle of the test. The spherical end of the pin
gated to eliminate the stress concentration problem. The performs a similar function to that of the sphere on the top
new configuration consists of two concentric hollow of the button during a button-on-block test. This new
cylinders that produce a contact area in the shape of an arrangement provides a uniform stress distribution at the
annulus. The new configuration is shown schematically in contact surfaces, thus eliminating the stress concentration
Fig. 6. In order to maintain proper alignment during issue that is currently being encountered with the ASTM
testing, an alignment pin is placed inside the cylinders. This G98 test method. One should note that, the friction
alignment pin can be seen in the section view of the test between the guide pin and the specimens is assumed to
configuration shown in Fig. 7. The top of the alignment pin be negligible. This will be verified in future studies.
ARTICLE IN PRESS
1426 R.A. Waite et al. / Tribology International 39 (2006) 1421–1427

Fig. 8. Initial results from new galling specimen configuration.

Preliminary tests were run in a tensile testing machine on proposed, and while preliminary tests using the proposed
seven sets of type 303 stainless steel. All of the tests were test setup are encouraging, more tests need to be conducted
run at a load of 445 N. Galling occurred in six of the seven before any conclusions can be drawn.
tests, and the relative location of the galling was rather
scattered. A photograph of the six tests where galling Acknowledgements
occurred is shown in Fig. 8. As can be seen in the photo,
galling occurred on the outer two-thirds of the contact The authors wish to thank the Lafayette College EXCEL
region for most of the specimens. This appears to indicate Scholars Program for providing support to Ryan Waite’05.
that the problem with the stress concentration at the edge
of the button has been eliminated by this new testing
References
arrangement. This new test arrangement may lead to more
accurate galling resistance test results although more [1] Standard Terminology Relating to Wear and Erosion, G40, Annual
testing will need to be performed to provide insight as to book of ASTM standards, vol. 03.02, ASTM.
the precision of the method. [2] Ives LK, Peterson MB, Whitenton EP. The mechanism, measure-
ment, and influence of properties on the galling of metals. US
6. Conclusion Department of Commerce, National Institute of Standards and
Technology, Materials Science and Engineering Laboratory, Cera-
mics Division, Gaithersburg, MD, USA, NISTIR 89-4064, 1989.
The current ASTM G98 standard galling resistance test [3] Peterson MB, et al. Galling wear of metals. Wear of materials.
needs to be reevaluated with the intent of increasing the Vancouver, BC: ASME; 1985. p. 293–301.
repeatability of the galling data collected. The geometry of [4] Standard Test Method for Galling Resistance of Materials, G98,
Annual book of ASTM standards, vol. 03.02, ASTM.
the test specimens allows for considerable stress concentra-
[5] Frick DG. Drill Collar Connection Trial. Report DGF2-85,
tions to exist due to the sharp edges of the button. The Carpenter Technology Corporation, 1985.
stress concentrations are not accounted for in the calcula- [6] Magee JH. ASM handbook, vol. 8. ASM International; 2000.
tions of threshold galling stress. The stress field along the [7] Hirasaka M, Hisashi N. J Mater Process Technol 1994;47:153–66.
contact surfaces has been modeled using both a closed [8] Jarrell JD, Bejbl F. Med Device Deagn Ind 1999;21:50.
form analytical model and using a finite element model. [9] Hummel SR, Partlow B. Threshold galling load and frictional
behavior of stainless steel couples in line contact. Wear 2003;255:
Both models support the existence of a non-uniform stress 504–8.
distribution across the contact area of the button-on-block [10] Hummel SR, Partlow B. Comparison of threshold galling results
test. from two testing methods. Tribol Int 2004;37:291–5.
Developing a test with a uniform stress distribution [11] Wiklund U, Hutchings IM. Investigation of surface treatments for
would move one step closer to correcting the galling galling protection of titanium alloys. Wear 2001;251:1034–41.
[12] Ives LK, Peterson MB, Whitenton EP. Galling: mechanism and
threshold calculation and would likely yield more accurate measurement. National Bureau of Standards Report, 1987. p. 33–40.
galling resistance results than those currently being found [13] Hummel SR. New test method and apparatus for measuring galling
with the button-on-block test. A new test has been resistance. Tribol Int 2001;34:593–7.
ARTICLE IN PRESS
R.A. Waite et al. / Tribology International 39 (2006) 1421–1427 1427

[14] Hummel SR. An application of frictional criteria for determining [16] Ciavarella M, Hills DA, Monno G. The influence of rounded
galling thresholds in line contact tests. Tribol Int 2002;35:801–7. edges on indentation by a flat punch. Proc Inst Mech Eng 1998;
[15] Vo DD, Wissing Jr CE. Failures of bolted connections due to wear 12(4, Part C):319–28.
and galling in bolt threads. In: Proceedings of use of new technology [17] Young WC. Roark’s formulas for stress & strain, 6th ed. New York,
to improve mechanical readiness, reliability, and maintainability, NY: McGraw-Hill; 1989.
Cambridge, 1985.

You might also like