Abaqus is one of two commercial finite element programs which are regularly used to model fretting.  Nearly all academic papers using finite elements to study fretting use Abaqus; however, another finite element program has the capability to model fretting and is widely used in industry.  Wahab, Talemi, and De Baets (2011) have published a paper comparing the two programs when modeling fretting fatigue.

Information about the finite element mesh and boundary conditions which were used in each of the computer programs are shown in the figures below.  Abaqus was more computationally efficient.  It ran much faster (16.8 vs 70.8 CPU time) even though it had more degrees of freedom (28,434 vs 20, 147).  This point alone gives a large advantage to Abaqus.

Wahab et al. (2011) used the penalty method contact solver with the other program and a Lagrange multiplier contact solver with Abaqus.  The Lagrange multiplier method is more accurate because it enforces a zero overlap condition rather than approximating it as the penalty method does.  In the other program the Lagrange multipliers did not converge, so the investigators were forced to switch to the penalty method.

The other program gave a higher maximum von-Mises stress and tangential stress than Abaqus.  This was expected based upon to the different contact algorithms being used.  However, both codes indicated the same locations for maximums.

Comparison of stress between finite element codes

Overall Abaqus appears to be the better finite element program for modeling fretting fatigue even though both codes can handle the problem.  I expect that Abaqus will become more advantageous if fretting wear is modeled instead of fretting fatigue.

• Wahab, M.A., Talemi, R.H., De Baets, P., 2011, “FEA of fretting fatigue: a comparison between ANSYS and ABAQUS,” Key Engineering Materials, Vol. 488-489, pp. 662-665.