Finite element based simulation of dry sliding wear Dr. Vishwanath Hegadekatte University of Karlsruhe

Prof. Norbert Huber GKSS Forschungszentrum Geestacht GmbH Technical University of Hamburg-Harburg

Prof. Oliver Kraft Forschungszentrum Karlsruhe GmbH University of Karlsruhe

Workshop:Finite element and analytical modeling of contact problems TU Chemnitz, 8 Dec 2006 SFB 499 · Entwicklung, Produktion und Qualitätssicherung urgeformter Mikrobauteile aus metallischen und keramischen Werkstoffen

1

Contents 1.

Motivation and Strategy

2. Dry sliding wear 1.

Finite element based wear simulation tool (WearProcessor)

2. Global incremental wear models (GIWM) 3. Wear simulation results 3. Dry rolling / sliding wear 1.

User subroutine "Umeshmotion" in ABAQUS

2. GIWM 4. Simulation of wear in spinning contact and transient contact tribosystems 5. Future work IMF-III

IPEK

2

Motivation

SFB 499 Model System

™ ™ ™ ™

Micromachines are finding a wide variety of applications High surface to volume ratio High operating speeds Wear is a critical factor which can limit their life span

3

Strategy

V. Hegadekatte et al. In: Micro-Engineering in Metals and Ceramics , Part II, 2005, Wiley-VCH Verlag GmbH, Weinheim, Germany, 605 - 623

4

Archard’s Wear Model

FN ∗ VW = k s H = k D FN s h = k D ps w

VW : Worn volume, FN : Applied normal load, H : Hardness, s : Sliding distance, k D : Dimensional wear coefficient , p : Contact pressure, h w : Linear wear,

*J.

F. Archard, J. Appl. Phys. 1953, 24, 981 – 988.

5

Wear-Processor Wear Processor

FEM (ABAQUS) Update Surface Topography (Re-meshing)

Compute Wear

V. Hegadekatte et al., Modelling Simul. Mater. Sci. Eng., 2005, 13, 57 - 75.

6

Wear-Processor Pin Wear Wear Processor

FEM (ABAQUS) Update Surface Topography (Re-meshing) Disc Wear Compute Wear

V. Hegadekatte et al., Modelling Simul. Mater. Sci. Eng., 2005, 13, 57 - 75.

7

Wear-Processor Pin Wear Wear Processor

FEM (ABAQUS) Update Surface Topography (Re-meshing) Disc Wear Compute Wear

V. Hegadekatte et al., Modelling Simul. Mater. Sci. Eng., 2005, 13, 57 - 75.

8

Simulation results

Si3N4 in Air

s yy [N/mm 2 ] -1200

s = 0 mm

-900 -600 -300 0 0 0.01 x [mm]

0.02

-0.01 -0.02

0

0.01

0.02

0.03

z [mm]

s yy [N/mm 2 ] -1200

s = 71.36 mm

-900 -600 -300 0 0 0.01 x [mm]

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

0.02

-0.01 -0.02

0

0.01

0.02

z [mm]

0.03

9

Simulation results

Si3N4 in Air s zz [N/mm 2 ] Sliding Direction

s = 0 mm

s zz [N/mm 2 ] Sliding Direction

s = 71.36 mm

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

10

Global incremental wear model (GIWM)

FN p= 2 πai

V. Hegadekatte et al. In: Micro-Engineering in Metals and Ceramics , Part II, 2005, Wiley-VCH Verlag GmbH, Weinheim, Germany, 605 - 623

Assumptions ™ Archard’s wear law ™ Average contact pressure over circular contact area ™ Assumes axisymmetric wear on pin surface ™ Normal elastic displacement taken into account (Oliver & Pharr, 1992) ™ Linear wear integrated over sliding distance

11

Simulation results – pin wear

Si3N4 in Air

w

h = kD p s Prediction

Fit

™ GIWM can fit pin-on-disc experimental results ™ GIWM can also predict pin-on-disc experimental results when load is doubled V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

12

Comparison between Wear-Processor and GIWM

Si3N4 in Air

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

13

Simulation results – pin wear

Si3N4 in Air

∗ hw 2 = k D p 1 + 3μ s μ : Coefficient of Friction

Prediction Fit

™ Modified Archard’s wear law (Sarkar, 1980) ™ Identified wear coefficient is lower by a factor 1 + 3μ 2 ™ Marginal influence of friction on the wear behavior for silicon nitride *A.

D. Sarkar, Friction and Wear. 1980, Academic Press (London), 411 – 43.

14

Simulation results – pin wear

Fit

Prediction

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

Prediction

15

Simulation results – pin wear

WC-Co in Water

Expt. Data: S. Kurzenhaeuser et al., In: Proc. Tribologie Fachtagung, Goettingen, Germany, 2006.

16

Simulation results – pin wear

WC-Co in Water

Expt. Data: S. Kurzenhaeuser et al., In: Proc. Tribologie Fachtagung, Goettingen, Germany, 2006.

17

Simulation results – pin wear

WC-Co in Water

Expt. Data: S. Kurzenhaeuser et al., In: Proc. Tribologie Fachtagung, Goettingen, Germany, 2006.

18

GIWM for Disc Wear

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

Assumptions ™ Archard’s wear law ™ Average contact pressure over initially circular and later elliptical ™ Normal elastic displacement taken into account (Oliver & Pharr, 1992) ™ The cross section of the disc wear track has the same curvature as the pin ™ Linear wear integrated over sliding distance

19

Results – disc wear (WC pin on DLC coated steel disc)

Fit

Prediction V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

20

Wear mechanism maps

S. M. Hsu, M. C. Shen, Wear 1996, 200, 154 – 175.

Advantage of FE based wear simulations Local effects

21

Sliding direction

z [mm] 0 -0.08

-0.03

0.03

Increasing µ

2

p [N/mm ]

-500

0.08

-1000

-1500

-2000

µ=0 µ=0.5 µ=0.75 µ=1

60 % higher than the frictionless condition

80 % lower than the frictionless condition

V. Hegadekatte et al., In: Proc. 1st Vienna Intl. Conf. on Micro- and Nano-Technology, Vienna, Austria, March 9 - 11, 2005, 181 - 190.

22

Dimensionless Parameters - Derivation

dh w = kD p ds w

FN dh = kD w ds 2πRP h FN h = kD s πRP w

FN h = 2 EC a e

=

FN 2 EC 2 RP h

w

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

RP : Radius of pin , FN : Applied normal load, EC : Contact modulus, s : Sliding distance, p : Contact pressure, k D : Dimensional wear coefficient , h w : Linear wear, a : Contact radius, h e : Elastic displacement

23

Dimensionless Elastic and System Parameters

FN e

2 EC 2 RP h h = w h FN kD s πRP

w

he

FN 2 EC RP k D s

Πe

Πs

1 = π w h RP 2

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

24

Π e vs Π s

™ Wear coefficients for different material pairs determined by fitting experimental results from literature using GIWM V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

25

Π e vs Π s

™ Wear coefficients for different material pairs determined by fitting experimental results from literature using GIWM V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

26

Local Effects

V. Hegadekatte et al., Tribology Letters, 2006, 24, 51 - 60.

27

Twin-disc tribometer

V. Hegadekatte et al. 2006, under preparation

28

Twin-disc tribometer

V. Hegadekatte et al. 2006, under preparation

29

UMESHMOTION for wear simulation

V. Hegadekatte et al. 2006, under preparation

30

Archard’s wear model for rolling / sliding contact

For rolling/sliding contact,

V − Velocity [mm/s]

n

w i +1

h

= h + kD w i

R − Radius [mm]

∫ p Δs

Subscript 1, 2 stand for parameters of

j

j =1

disc 1 and 2 respectively. θ =2π

hi +1 = hi + k D ∫ p j R1i dθ w

w

θ =0

For a given time increment of

Δti

Δti | (V1 −V2 ) | θ =2π hi+1 = hi + kD ∫θ =0pj R1i dθ 2πR1i w

w

V. Hegadekatte et al. 2006, under preparation

31

Simulation results

V. Hegadekatte et al. 2006, under preparation

32

Simulation results

V. Hegadekatte et al. 2006, under preparation

33

Simulation results

V. Hegadekatte et al. 2006, under preparation

34

Simulation results

V. Hegadekatte et al. 2006, under preparation

35

Simulation results

V. Hegadekatte et al. 2006, under preparation

36

Wear simulation in a spinning contact

V. Hegadekatte et al., Modelling Simul. Mater. Sci. Eng., 2005, 13, 57 - 75.

37

Wear simulation in a spinning contact

V. Hegadekatte et al., Modelling Simul. Mater. Sci. Eng., 2005, 13, 57 - 75.

38

Guidelines Experiment Verification

GIWM

Wear-Processor / UMESHMOTION

•

• •

• •

Fast tool for quick identification Wear Model (Global Scale) Limitations – Geometries, wearing surface

• •

Wear Model (Local Scale) Ideal for studies of local effects Different Geometry Computationally Expensive

Verify each other

39

Transient contact wear simulation s=0

C F F

s=smax

B

F F

A

p

Cylinder Zylinder

Slab Platte

s

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

40

Simulation results

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

41

Simulation results

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

42

Simulation results

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

43

Simulation results

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

44

Simulation results

N. Huber et al., In: Proc. Multiscale Materials Modeling Conference, Freiburg, Germany, Sep 18 - 22, 2006, 970-973.

45

Future work

•

Defined contact

•

Quasi-static conditions

™ Wear simulations using FE-models of the planetary gear train from University of Karlsruhe ™ Investigation of pressure and sliding velocities as a function of wear ™ Influence of starting tolerances on wear and evolution of the contact loading N. Huber et al., Microsystem Technologies, 2003, 9, 465 - 469.

46

Future work

http://www.rci.rutgers.edu/~ozel/projects.htm

Goals ™ To study wear evolution and lifetime assessment of multilayer multifunctional ceramic coated cutting tools ™ Local coating wear and intrinsic stresses in the multilayer coating ™ Plastic deformation of the tool substrate