Proceedings of the 1st WSEAS International Conference on VISUALIZATION, IMAGING and SIMULATION (VIS'08)

Finite element analysis used for

determination the endurance of surfaces
Delia GARLEANU, Victor POPOVICI, Gabriel GARLEANU, Mihnea COSTOIU
Department Materials Technology and Welding
University POLITEHNICA Bucharest
Address Splaiul Independentei nr. 313
COUNTRY: ROMANIA

Abstract: - The paper shows a mean of prediction of the endurance of surfaces by using the finite element analysis. In
the case of surfaces that have been subjected to repeated strain, wearing, which is a consequence of the mechanical and
chemical behaviour of the material, may become critical to some friction based applications which imply the friction of
carrier surfaces. The model presented, made be means of the finite element analysis using the ANSYS package is
placed at a medium level of prediction of the war based on Archards theory. In order to simulate and compute the wear
by using ANSYS, creep is used. The input data for the modelling have been obtained by measurement. The results
obtained by modelling are very close to those obtained experimentally.

Key-Words: - Endurance, Contact pair

1 Introduction The wearing phenomenon, at a macroscopic scale, is

The finite element analysis of the endurance of put forward by the Archard equation:
surfaces subjected to abrasion strain sets to make a
prediction model of the wear resistance of a surface. The W = K sP (1)
wear, which is a consequence of the mechanical and
chemical behaviour of the material, becomes important where: W is the volume dislodged; s is the gliding
in surfaces that are subjected to repeated strain and may distance; P is the applied strain and K is the wear per
become critical for friction applications which rely on unit of strain per squared distance.
the friction of carrier surfaces. K may be called wear quota; experiments, which implied
The wear produces by dry gliding friction may the same combination of materials, have indicated the
be approached by finite element modelling by including modification of K is connected to the modification of the
the properties regarding surface interactions, surface surfaces status.
roughness included, in the model. In order to simulate The Archard equation takes into consideration
the grinding or the impact, it is necessary for the finite that the wear is independent of the contact area but we
elements to be removed. neglect the micro geometry of the surface. Effects
There are birth and death elements in ANSYS. produces by the roughness are contained in the wearing
When the particles removed are of molecular size, the quota, experimentally determined.
abstraction density, at least in the surface, must assure a In order to simulate and compute wear with
size of the element which is the same as that of the ANSYS, the creep may be used. The wearing strain
molecules. This would need a large number of finite based on the final configuration of the surface at the end
elements in order to generate data which would exceed of the loading step is determined by using explicit creep
the computation capacity. Details regarding the micro computed after that for elastic and plastic creep.
and nano sizes of the project must thus be overlooked. The advantage of using explicit creep for computing
At a macroscopic scale, the size of the elements will be the wear (creep) is of making a correction to the elastic
too big to anticipate changes due to wear. strain and this correction is made after the computation
The model created for the finite element analysis of plasticity which may thus b included in the analysis.
is scaled onto physical reality by comparing the results
obtained by modelling with the experimental results. The equation of creep strain is programmed into ANSYS
The surface subjected to abrasion strain like:
indicates a variation of physical-mechanical properties, decr C4
as it wears down, due to the loading by welding with one = C1 efort C 2 exp (2)
dt T
to three hard layers.

ISSN: 1790-2769 151 ISBN: 978-960-474-022-2

 Proceedings of the 1st WSEAS International Conference on VISUALIZATION, IMAGING and SIMULATION (VIS'08)

where: C1, C2, C3 and C4 are constants input In order to create the geometry, the Preprocessor
by the user. is activated from the main menu, which generates the
For each time step, the creep strain is computed geometry of the volumes of base material and welding
by using the second equation and then the incremental band Fig.2. The base material has the sizes 2x4x0.4
creep strain is multiplied to the time span and added to
inches, and the welding band 0.6x0.12 inches. The
the previous creep strain.
meshing of the volume of base material and the welding
band Fig.3 is made by means of the Solid 5 (3-D 8-
Node) element.
2 Creating the geometry, mesing and The nodes and the elements corresponding to the
definition of the contact pair welding band of hard added material laid onto the base
The model Fig.5 is made up of two bodies: an material are selected and replaced by the Solid 5 with
inferior one a parallelepiped which the properties of Targe170 Fig.4.
the base material are added, onto which the welding
band made up of 1 - 3 hard layers approximately 1
mm thick is laid, with enhanced values of hardness
(layer 1 58 HRC, layer 3 66 HRC and layer 3
68 HRC) and the upper one a hard plate which during
the wearing process comes in contact with the over
heightening of the welding band.
For the modelling, three elements are used: SOLID 5 (3-
D Coupled-Field Solid), CONTA174 (3-D 8-Node
Surface-to-Surface) and TARGE170 (3-D Target
Segment), shown in Fig.1.
Fig. 2. Generating the volumes
of the base material
and that of the welding band

Fig. 3. Meshing the volumes of the

base material
and that of the welding band

Fig. 4. Elements that correspond

to the hard layers onto
Fig, 1. Elements used for the model the surface of the base material

ISSN: 1790-2769 152 ISBN: 978-960-474-022-2

 Proceedings of the 1st WSEAS International Conference on VISUALIZATION, IMAGING and SIMULATION (VIS'08)

Fig. 6. a,b. Generating movement

for making the contact

At the second step, the movement in Z direction is

replaced by the stress applied on the same direction
and also in the negative direction so that during the
wearing process a pressure force would be present
between the over heightening of the welding band
and the hard (abrasive) surface.

The third step (static) is a collection of loading steps,

Fig. 5. Meshing with finite elements each consisting of 106 repetitions of the load. This is
of the volumes of base material, completed by initiating a loading loop of a desired
welding band and hard (abrasive) number of times, which results in the generation of
surface area loading step sets.

At a distance from the over heightening of the 3 Results regarding creep strain
welding band Fig.5, a rectangular contact are is created In the case of this mode, the creep becomes an
which abstracts with just on element type Conta174. instrument for measuring wear. The results of the
analysis are presented in the Postprocessor by selecting
2.1 Performing the analysis the elements of the type Solid 5.

A nonlinear analysis in three loading steps is performed. Figure 7 shows the von Mises creep strain at the end
The first step consists of a Z-wise movement in the of the loading steps in isometric depiction with auto
negative direction Fig.6, in order to perform contact computed scaling and isometric and upfront view with
between the hard (abrasive) surface and the over real scaling. It can be observed that the surface is
heightening of the welding band, with rheological smoother.
properties defines in the Preprocessor. In table 1, the creep strain values with the respective
time of the loading steps are presented.

a
Fig. 7. The von Mises creep strain
at the end of the loading step

ISSN: 1790-2769 153 ISBN: 978-960-474-022-2

 Proceedings of the 1st WSEAS International Conference on VISUALIZATION, IMAGING and SIMULATION (VIS'08)

Table 1
Time / 20 Creep strain 4 Conclusion
[s] [mm]
Modelling by means of the finite element model, in
this case using the ANSYS package, allows making a
0 0
prediction of the wearing behaviour of hardened surfaces
6 0.0355
12 0.0711 by lying down through welding.
18 0.1062 This method relieves the researcher from having to
24 0.1388 perform further mechanical trials.
30 0.1534 The input data for the modelling by finite element
36 0.1659 method have been the values obtained by measurements,
42 0.1769 where most of the times it is necessary for just one input
48 0.1863 value to be taken by measurement in order to scale the
54 0.1945 model onto reality.
60 0.2016 The results obtained by modelling have been very
66 0.2077 close to those resulted experimentally.
72 0.2130 It is possible to analyse the wearing phenomenon at
78 0.2175 a macroscopic scale.
84 0.2215 The model presented is placed at a medium level of
90 0.2248 prediction of the wear based on the Archard theory.
96 0.2278 Details regarding the behaviour to wear of the
surface by introducing effort in the equation of the wear
By using the creep-time value pairs, resulted from can be highlighted through computations.
the analysis with finite elements, the variation curve
presented in Fig.8 is built. The shape of the curve is the
same as that of the wear curve. References:
A relation which approximates the dependency [1]N.P.Suh.http://ocw.mit.edu/NR/rdonlyres/Mechanical
of the creep strain with time, determined by the same -Engineering/2-800Fall-2004/3EB7EE5A-43BF-
spreadsheet program, TC2D, is presented in Fig.9. 4B4B-860B-1FD57DD8207F/0/ch4a_wear_intro.pdf
[2] Archard, J. F. and Hirst, W. Wear of metals under
0.25
unlubricated conditions. Proceedings of the Royal
0.2 Society of London. Series A, Mathematical and
Physical Sciences, Vol.236, No.1206. (Aug. 2, 1956),
Creep strain

0.15

0.1 pp. 397-410.

0.05
[3] Kauzlarich and Willams. Archard wear and
0
component geometry. Proc. Instn. Mech. Engrs. Vol
0 20 40 60 80 100 215 Part J.
Time/20 [s]
[4] Hegadekatte et al. Finite element based simulation
Fig. 8. Variation of the creep strain in time of dry sliding wear. Modelling Simul. Mater. Sci.
Eng. 13 (2005) 57-75
[5] Podra, P. and Andersson, S. Simulating sliding
wear with finite element method. Tribology
International 32 (1999) 71-81.

Fig. 9. The equation of creep strain

ISSN: 1790-2769 154 ISBN: 978-960-474-022-2