Laboratory simulation methods have been used for many years in the automotive industry to accelerate durability testing of vehicle components. As a result, component simulation testing has become well established. The tests use a simplified mechanical system, consisting of a physical prototype of the component under investigation and some fixturing that allows appropriate restraint and loading. Due to the remaining complexity and cost of prototype build and physical test, companies increasingly use computer simulation to provide initial design guidance. A common difficulty for this approach is that the load and boundary conditions are hard to determine. This is especially due to the situation that only some components, rather than the whole vehicle, are of interest and the external loads are not applied directly to these components. Therefore, analysts rely on simplified estimates and experience to determine the load and boundary conditions. The results of this approach may not be satisfactory with respect to the speed of development and the cost/weight optimization level of the final design.
This paper proposes a method of virtual testing to simulate the established component test to evaluate the design in the early development stage. There are two main advantages to this approach. First, since the component test usually greatly simplifies the real system, the test is much easier to model than the real vehicle system on the road. Second, the load and boundary conditions of well established component tests usually are relatively easily obtained. Since the virtual test models are direct representations of the component tests, and the component tests have been proven to reproduce the real service life of the component, the virtual test should be a representative evaluation of the component design. This paper presents the results of a study where physical and virtual testing was employed. Failures in both tests corresponded very closely with respect to location and time to failure.