Prediction of geometrical variation of forged and stamped parts for assembly variation simulation

Paper in proceeding, 2013

In assembly variation simulations a total sensitivity matrix, implicitly defined in a FEA-based simulation model describing all mating conditions, kinematic relations and non-rigid behavior, is used together with Monte Carlo simulation to predict variation in critical product dimensions. A major drawback with this kind of simulation is that form and shape variation of parts, stemming from their individual manu-facturing processes, are unknown. Therefore, better ways to predict form and size variation are needed. In this paper a five critical material and process parameters are varied in a full factorial computer experiment for two different test cases; one forged part and one stamped part. The geometrical deviation in each factori-al run is registered. These results are then used as input for construction of a meta-model, well suited for Monte Carlo based variation simulation. The contributions from the different material and process parame-ters to the total variation are discussed. The cases show that the proposed method is well suited both for forging and stamping and represents a general way to describe, model and simulate part variation in varia-tion simulation for assemblies.