Contact wear modeling

Output variable CWEAR is requested in all these tests for nodal wear distances. In simple cases, CWEAR is verified to match hand calculations of wear distances based on the Archard's wear model. Most of these test cases are complex enough that closed form solutions are not available.

For the test with a flat analytical rigid surface compressing and sliding over a block-shaped C3D8 element without friction in a sequence of seven steps, the stress state in the block remains uniform uniaxial with varying magnitude. Wear is not proportional to the friction coefficient in this test. Wear accumulates during increments in which the contact pressure and incremental slip are both nonzero. Contact pressure and slip increments are uniform across the nodes in contact, so the wear is also uniform. The accumulated wear distance is consistent with hand calculations.

For tests with a block modeled with C3D10 elements pressed into and sliding across a flat surface, wear is proportional to the friction coefficient. Frictional forces lead to spatial variation in the stress state within the block and spatial variation of contact pressure over the surfaces, such that incremental wear accumulation is greatest on specific edge locations of the block. Wear accumulation is consistent with the evolution of frictional stress and incremental sliding. Small differences in stress state and wear across these tests is due to numerical details, such as whether a penalty method or Lagrange multiplier method is used to enforce contact conditions.

For tests with two blocks twisting while pressed together with wear properties assigned to each face, wear accumulation is small near centers and corners of faces in contact because these regions experience relatively small slip or are not in contact for most of the simulation. Contour plots of wear distance on contact faces show that most wear occurs within the ring region at intermediate distances from the center of these faces, where contact pressures and slip rates are often large during the simulation. The test case with linear tetrahedral elements is overly stiff for the bending-related mode that should occur for much of the twisting motion while the respective corners are not aligned. Therefore, this test overestimates contact stresses and wear distances near edges and underestimates contact stresses and wear distances away from edges.

For the test involving two hollow hemispherical parts in frictionless contact with relative spinning between two parts, wear accumulates on the contact surface of the outer part but not on the contact surface of the inner part because the wear model assigned to the outer part is not proportional to the friction coefficient and the wear model assigned to the inner part is proportional to the friction coefficient (and the friction coefficient is zero for this test). The contact pressure is relatively uniform over the surfaces in contact. The sliding distance is proportional to the distance from the axis of relative spinning between the contact surfaces. The wear distance is approximately proportional to the distance from the axis of relative spinning.

For the test involving compression of a gasket, wear develops on some regions of the gasket due to contact with a rigid surface and develops on other regions of the gasket due to self-contact. Wear accumulation is not proportional to the friction coefficient for this test, so the incremental wear distance at a particular location of the gasket surface is related to the contact pressure and the incremental slip at that location during that increment. Wear is assigned as a surface property, so the incremental wear distance at a given surface location does not depend on characteristics of the opposing surface.

For the test involving self-contact of a ring, "rolling" motion in the second step leads to significant sliding and wear where self-contact is active.