### Inverse Analysis in Hydroforming of a Refrigerator Door Handle Using MOGA

Abstract

An inverse problem of load path design in tube hydroforming of a refrigerator door handle is
presented. The motivation is to inversely analyze the load path for tube hydroforming and
verify the applicability of developed algorithms. Several parameters are given as inputs: the
material properties of the pre-bent tube, the geometry of the tube and the die. The study
also investigates the influence of the coefficient of friction (COF) at two levels: with COP
of 0.1 and 0.05. The process parameters with regard to pressure and end feeding are to be
determined. An optimization strategy using MOGA was implemented to solve the inverse
problem. In order to obtain a quality part without failure, six objectives were calculated to
evaluate the forming severity and the final product geometry, namely, fracture/necking,
wrinkling, severe thinning, die filling, and the maximum thinning ratio of tube thickness. The
MOGA was coupled with finite element analysis (PEA) and used to optimize the load path.
To reduce the calculation cost, a surrogate Kriging model was established to replace the PEA,
and was updated every five generations. A Pareto optimal solution set was generated after 20
steps and the optimal load path was determined with a minimum thinning ratio. The results were
validated by both the FEA simulation results and the comparison of the optimal load path with
the actual load paths. It was noted that the thinning ratio was less than 25% for the lower
coefficient of friction (0.05) and the maximum stress was far below the forming stress limit.
Meanwhile, the tube fully filled the die cavity and no failure occurred.