Numerical optimization of wear performance – Utilizing a metamodel based friction law

• Wear performance optimization with finite element analysis and genetic algorithm.
• Modeling of wear considering local frictional characteristics in contact interface.
• Equation for friction applying metamodels for local frictional characteristics.
• Finite element wear analysis accounting for geometry alteration due to wear.
• Method applicable to structural design, e.g. to tire design.

In this contribution, a structural design optimization approach focusing on the enhancement of wear performance, which employs the finite element analysis and a genetic optimization algorithm, is presented. The numerical wear analysis between two sliding bodies is based on constitutive wear and friction equations. Since local stress concentrations are identified as the origin of wear, the local characteristics of tangential stresses should be appropriately captured by means of the constitutive equation for friction. Therefore, a new friction model is developed, enabling to account for the nonlinear dependency of the friction coefficient on the local contact pressure and sliding velocity in the contact interface through the utilization of metamodels, especially artificial neural networks. With the help of the proposed constitutive equation for friction, the energy dissipation rate due to friction, which is utilized for the evaluation of the wear rate within the wear model, can be properly calculated. The numerical wear analysis is coupled to an optimization approach based on a binary coded genetic algorithm. Since geometrical configuration changes due to wear can be modeled with the help of the utilized wear analysis, criteria to pursue a reduced wear volume and a regular wear distribution are developed and implemented into the formulation of the optimization task. The application of the presented methods to a pneumatic tire design task is shown as an example.