Rudolf Seefeldt, Jörg C. Sturm, Alexander Pawlowski
The demands on the productivity and robustness on the high pressure die casting process for high quality components are continuously rising. At the same time, financial considerations mean that an exact and reliable planning of the die layout and production process is required. Today, experience provides a basis for a dependable production planning. With the ever-increasing complexity of die castings, however, the application of experience from previous projects to new castings is increasingly difficult and questionable. A new and novel approach using autonomous optimization makes it possible to use information gained from an existing casting process with a good component quality and to use this information as the target to be achieved in the optimization of the layout of a new die. In doing so, a virtual test program for the die layout is carried out using casting process simulation. The program learns on a stand-alone basis, through trial and error, and searches for the best concepts for runner and gate design or shot conditions in order to reproduce previously effective conditions in the new casting.
Please read the complete publication in the linked PDF.