Predicted versus actual compensation in a stamping die

摘要

Traditional methods used to produce a die set (from developing initial machining cutter paths through finalized die tryout to produce a part that meets design intent) begin with draw simulation and development. It is here, traditionally, that scientific evaluation of actual metal stretch and theoretical ideals end. In past programs, a designed part would be simulated for stretch and a development model created to include various die compensations (i.e., springback, overcrown, etc.) based on past experience for area and amount. At this point, the die is cut and undergoes a metamorphosis through die tryout to finally produce a quality part. This is currently an open-loop system. This paper will focus on the differences in the predicted way the die should look and the actual outcome (after part buyoff). We will document the original ""plan'' (compensation areas and amounts), the resulting development model, supporting data of the initial cut die and the differences between the die at original cut and at final part buyoff. This will create a closed-loop system. We will also show how this data can be used to regenerate a new development model and new cutter path for the purpose of creating a second set of dies. This information may also be used as a benchmark or a ""zero point'' to which the die may be restored after wear. This will also provide a database, which can be used to expedite future die development, by knowing the die surface required to produce a quality part at the production source.