The hydroforming process involves a number of complex phenomena and presents several types of non-linearities (geometric, behavioral law, etc.). Optimizing a hydroforming operation requires a great deal of testing to determine precisely the optimum load paths and obtain a defect-free part. Advances in numerical tools have enabled manufacturers to simulate and optimize their production resources before production begins, with the aim of minimizing the rate of part failure
[M3185]
. This approach is justified as much by the multitude of parameters to be controlled as by the exorbitant cost of a real test. A number of techniques have been proposed over the last ten years to ensure that a shaping operation is carried out properly
[M3000][M3002]
.
Most of these techniques combine the finite element method with optimization approaches. With these means, manufacturers can virtually simulate their processes, which helps answer certain questions, mainly concerning the feasibility of the part and also the suitability of the loading path for the correct shape
[BM7518]
. This coupling often leads to a clear improvement. However, it does not ensure process stability, nor does it rule out the occurrence of certain types of plastic instability during manufacture, given that there are several sources of uncertainty linked to the material, the loads, the press and also the operator.
The aim of this article is to take account of uncertainties in the analysis and optimization of the hydroforming...