Nonlinear optimization-based design of ram velocity profiles for isothermal forging

The development of good material models, accurate nonlinear finite element codes, and computer-controlled presses make practical the application of control techniques to metal forging. This paper considers the open-loop problem of selecting a nominal ram velocity profile to produce a desired microstructure, given a specified die and preform geometry and forging temperature. Two approaches for doing so are applied to a simple, but representative problem. The first is based on classical numerical optimization techniques. The second is based on inverse neural networks and offers potential savings in critical computations. Simulation studies for the two methods show good results