Improving the Suspension Design Process by Integrating Multibody System Analysis and Design of Experiments 930264
Increasingly, product engineers must influence customer satisfaction with robust designs, as opposed to waiting for manufacturing to build quality into a product. To enable and facilitate these efforts, simulation and analysis must be used to guide the system to an improved or even optimal state versus merely a functional one. The integration of Multibody System Analysis (MSA) with Design of Experiments (DOE) creates a powerful combination of tools for thorough investigation of a specified design space, identification of the optimal system configuration, and illustration of the effects of system changes on a given output.
This paper demonstrates this approach for a specific output: the vertical loads into a suspension's front strut/shock tower due to a severe pothole road event. This event is designed to test the energy management function of a suspension for severe impact events. Improper energy management leads to excessive forces transmitted to the body structure. The analysis presented in this report is an introductory example. The power of this tool to impact customer satisfaction manifests itself primarily in the analysis of the ride and handling functions of the suspension. The fundamental lesson, though, is that the design of experiments approach provides much more information than the haphazard one-factor-at-a-time experimentation.
Since implementing this approach, the authors have found that every system examined behaves differently. The interactions are different, and the factors differ in the amount of their impact on each system. This will be apparent to the reader as the two strut suspensions and one SLA (short-long arm) suspension are presented. Also, the chosen component design spaces dictate the levels of each factor chosen, which impacts the contribution of each factor. Thus, it is worthwhile using this approach on all systems since each differs from the others and engineers can not necessarily generalize from one system to the next.