Process Modeling: Use of Uncertainty, Sensitivity and Optimization Techniques for Improved Understanding of Compaction Model Outputs

Stone, T.W., Sanderow, H. I., Grewal, H., Acar, E., Hammi, Y., Allison, P., & Solanki, K.N. (2009). Process Modeling: Use of Uncertainty, Sensitivity and Optimization Techniques for Improved Understanding of Compaction Model Outputs. Advances in Powder Metallurgy & Particulate Materials. Las Vegas, NV: MPIF.

Abstract

Math-based models developed by the MSU/CAVS team have considered both the compaction and sintering processes. Due to the very large number of input variables for the compaction model it is helpful to understand how sensitive the model output is to small changes in these input variables in order to better apply the model to real world processes. Using numerical methods both uncertainty and sensitivity analysis can be applied to the model and the most significant terms identified. Through further numerical analysis the output of the compaction model can be optimized for one or more output parameters, e.g. least density variation, lowest mass, minimal part thickness. These techniques will be illustrated using a well-established automotive PM product, the main bearing cap.