CFD-Based Optimization Methods Applied to Polymer Die Design

The integration of optimization algorithms with computational fluid dynamics (CFD) is becoming more practical as the software becomes more robust and the hardware more efficient. In order to investigate the application of CFD-based optimization methods in the chemical process industry, two polymer die design examples were explored using different slot die geometries, different CFD codes, and different optimization packages and methods. In both cases, various geometric parameters were modified in order to meet the two often competing objectives of uniform flow at the die exit while minimizing or limiting the pressure drop. In some of the trials, CFD was used to create response surface models (RSM) and the optimization algorithms were applied to the RSM. In other trials, the optimization algorithms were directly coupled to the CFD. The advantage of the response surface model approach is that the RSM runs much faster than the CFD model. However, this is offset by the need to run numerous CFD cases up front in order to generate an accurate RSM. Designs with improved performance over the base cases were identified for both die geometries.