Wear refers to the gradual removal of material from a surface due to mechanical contact. In practical engineering systems like bearings, gears, fasteners, and seals, wear can influence long-term durability and lead to performance degradation. Ansys Mechanical enables users to simulate this phenomenon using empirical wear models. Chief among them is the Archard wear model.
In this blog, we’ll explore how wear is modeled in Ansys, how to implement the Archard model, the use of APDL commands, postprocessing strategies, and meshing guidelines.
In FEA, wear is modeled by shifting nodes on contact surfaces to replicate material erosion. The simulation iteratively updates geometry and recalculates equilibrium after each update, making the process nonlinear.
Ansys handles wear at the continuum scale using empirical laws that relate pressure, relative motion, and material resistance to surface loss.
Wear modeling is only available for the following contact element types:
To activate wear, a material definition with the TB,WEAR model is applied to these contact elements.
There are two types of wear models:
Archard Wear Model (native)
User-Defined Wear Model (custom via USERWEAR)
This article focuses on the Archard method.
The Archard model approximates wear rate using the relationship:
Where:
K = Wear coefficient
H = Hardness of material
P = Contact pressure
v = Relative sliding velocity
m = Pressure exponent
n = Velocity exponent
By default, the wear vector points opposite the contact normal, though this can be changed using directional inputs.
To activate wear modeling in Ansys Mechanical, a dedicated material model must be defined using the TB,WEAR command. This command, combined with TBDATA, allows users to input wear-related material constants that control how wear is calculated and applied across contact surfaces.
| TB,WEAR,1,,,ARCD ! Activates Archard wear model for material ID 1 TBDATA,1,K,H,m,n ! Wear constants (required) TBDATA,6,nx,ny,nz ! Optional direction cosines for custom wear direction |
The ARCD keyword activates the built-in Archard model.
The material ID (1 in this case) must correspond to a contact element set (e.g.,CONTA172, CONTA174 or CONTA175).
Additional Archard wear constants are defined as follows:
| Constant | Parameter Description |
|---|---|
| C1 | Wear Coefficient (K) – Determines the rate of material removal |
| C2 | Hardness (H) – Resistance to wear; can be estimated from Brinell or Vickers hardness (in Pa) or derived from material yield |
| C3 | Pressure Exponent (m) – Nonlinear influence of pressure on wear |
| C4 | Velocity Exponent (n) – Nonlinear influence of sliding velocity on wear |
| C5 | Model Flag – Controls behavior of the wear algorithm (optional, explained below) |
| C6-C8 | Direction Cosines (nx, ny, nz) – Optional override for wear direction |
The Archard equation implemented in Ansys is generalized, meaning that pressure and velocity dependencies are tunable via m and n.
Estimating Material Hardness
In plasticity-based material models like TB,BISO, C2 (hardness) can be estimated automatically from the current yield stress:
| TBDATA,2,-99 ! Activates automatic hardness = Yield Stress / 3 |
Conditions:
Applicable only if underlying elements are: PLANE182, PLANE183, SOLID185, SOLID186, SOLID187 or SOLID285
No wear will occur in the first substep because prior yield data is not yet available.
The fifth constant, C5, alters how the wear increment is calculated and applied. Each option is suited to a different modeling scenario:
| Value | Description |
|---|---|
| 0 | Default: Wear based on contact pressure |
| 1 | Use nodal stress from the underlying solid element instead of contact pressure |
| 10 | Averaged Wear (pressure): Equalizes wear increment across contact area |
| 11 | Averaged Wear (stress): Same as above but uses nodal stress |
| -99 | Postprocessing only: Wear calculated but not applied to geometry |
Use C5 = 1 for smoother wear behavior on asymmetric contact between dissimilar meshes.
Use C5 = 10 or 11 to ensure uniform wear distribution across interfaces.
Use C5 = -99 to debug or study wear trends without affecting convergence or geometry.
If wear must be applied in a direction other than opposite the contact normal, use direction cosines:
| TBDATA,6,nx,ny,nz |
This is useful when:
Wear is induced by rotational contact
Oblique sliding dominates material removal
Experimental calibration justifies a directional bias
Ensure that (nx, ny, nz) represent a unit vector in global coordinates.
Use TBFIELD or TBTEMP to define time- or temperature-varying wear coefficients.
| TBFIELD,TIME,0 TBDATA,1,K,H,m,n TBFIELD,TIME,10 TBDATA,1,K,H,m,n |
Use this for applications like fretting or fatigue where wear evolves slowly over time.
| TBTEMP,25 TBDATA,1,K,H,m,n TBTEMP,100 TBDATA,1,K,H,m,n |
Apply this in cases where sliding friction causes localized heating and temperature affects material resistance.
Scales wear at every substep based on contact penetration depth.
| TB,WEAR,matid,,,AUTS TBDATA,1,0.1,1e5 ! C1 = safety factor, C2 = max scale |
Updates wear only at the end of load cycles. Ideal for periodic loads.
| TB,WEAR,matid,,,CBCS TBDATA,1,0.1,1e5 |
AUTS is generally smoother; use CBCS only when cycles are clearly defined.
At each substep, once convergence is met, wear increments shift surface nodes. This temporarily disrupts equilibrium, which is resolved through additional iterations. Small timesteps are advised to avoid issues like unexpected contact separation.
Wear modeling alters surface geometry and may distort underlying solid elements. To manage mesh integrity:
Use rezoning or nonlinear adaptivity (NLADAPTIVE)
Enable morphing for tet/brick elements
Avoid mixed meshes in contact regions
| nladaptive,contwearel,add,contact,wear,0.75 nladaptive,contwearel,on,,,5,1,7 |
Explanation of Parameters:
contwearel: Component name containing the contact elements undergoing wear
add: Adds a new criterion for triggering adaptive remeshing
contact: Type of criterion to evaluate (based on contact conditions)
wear: Subtype specifying that wear is the criterion of interest
0.75: Threshold; triggers remeshing when 75% of material is worn
on: Activates the adaptive process
5: Frequency; the program checks the wear criterion every 5 substeps
1,7: Start and end time between which this adaptivity rule is active
NMISC: Total wear vector per node
NLHIST: Wear volume tracking and auto-stop criteria
| NLHIST,ALL NLHIST,STOP_VALUE,WearVol,1e-3 |
View wear displacement in .rst file
Plot worn vs. unworn geometries
Analyze trends in contact pressure or surface damage
For highly specific wear models, implement the USERWEAR routine.
Input variables include:
Slip, stress, and temperature
Contact node coordinates
Material constants (from TBDATA)
Define output:
WearInc = Wear increment per substep
WearDir = Optional vector direction
Ideal for applications requiring custom physics or highly nonlinear wear behavior.
Ansys Mechanical offers robust wear modeling capabilities, particularly through the Archard model. With careful consideration of wear direction, scaling, and element distortion, engineers can simulate realistic material degradation over time.
To summarize:
Use Archard for general cases, USERWEAR for custom needs
Apply scaling for efficiency
Protect mesh quality with rezoning or adaptivity
Postprocess wear effects to inform design improvements
By integrating these tools, engineers can confidently predict wear performance and make data-driven design choices.