PROFILE OPTIMIZATION FOR THE PREDICTION OF INITIAL PARISON DIMENSIONSFROM FINAL BLOW MOLDED PART SPECIFICATIONS

The critical stage of the extrusion blow moulding process involves the inflation of a molten parison into the final part. Modelling of the i nflation stage generally entails predicting the final part thickness d istribution, given the initial parison thickness profile. We refer to this formulation of the process as the ''forward formulation'' or as t he ''forward predictor''. It would be more desirable to model the proc ess using the inverse formulation, that is, given a final part thickne ss distribution, obtain the required initial parison thickness profile . We refer to this formulation of the process as the ''inverse formula tion''. Tooling costs and machine downtimes can be minimized with the information obtained from the simulation. Computational times and man- hours involved with trial-and-error runs with the forward predictor ca n also be minimized. The approach chosen involves placing a forward pr edictor in an iterative optimization loop. The loop searches for the p arison thickness profile that results in the minimum overall differenc e between the specified final part thickness distribution and the indi vidual iteration's output from the forward predictor. The forward pred ictor employs a discretized Newton-Raphson technique in the solution. The profiled optimization methodology utilizes updating techniques ana logous to classical process control equations.