Effect of input variables uncertainty in free tube hydroforming process

Abstract

Tube hydroforming (THF) is a plastic forming process that uses tubes with an initial circular cross section, in which pressurized fluid and axial feeds are applied for producing parts with various cross-sectional shapes. Despite of the complexity of THF process, a great progress in the automotive and aerospace industry has been made due to its advantages, such as, consolidation and weight reduction over conventional stamped and welded parts. The analysis of THF process is typically based on deterministic approaches, excluding scattering effects that influence the process reliability. Thus, robust design of tube hydroforming aims to vanish noise factors effects on process responses by considering the influence of process parameters variability. If this fluctuation is not monitored, then the fluctuation of the hydroformed parts quality may contribute to high scrap rates. In this work, the influence of variability in the THF material and process parameters (e.g. yield stress, strength coefficient, strain hardening exponent, plastic anisotropy, initial tube thickness and bulged length) on the bursting pressure is analyzed resorting to a response surface model. The statistically significant variables, which mostly influence the free bulge hydroforming process, are identified through an analysis of variance. Assuming that the input parameters variability follows the normal distribution, the probability distribution of the bursting pressure is evaluated by involving random process variables into the built response surface model. It was shown that the initial tube thickness is the most statistically significant variable, whereas the strain hardening exponent is the least statistically significant variable.