Feasible pressure and axial feed path determination for fuel filler tube hydroforming by genetic algorithm

Abstract

Successful fuel filler tube hydroforming largely depends on proper loading paths, that is, application of internal pressure and axial feeding during the forming time duration. Generally, two part quality criteria are considered in selecting the feasible loading paths: (a) minimum part wall thinning and (b) part wrinkle free. Due to the highly nonlinear nature of the tube hydroforming process, iterative finite element analyses with adjustments based on forming experience are typically conducted to design the loading paths. In this research, genetic algorithm was integrated into the finite element analysis–based optimization, resulting in enhanced determination of the feasible loading paths. Genetic algorithm is a heuristic search based on mechanics of natural selection. A pair of pressure and axial feeding profiles was represented by connecting genes making up to be a chromosome. In each search, mutation and crossover operations generated a new generation of chromosomes. Fitness functions were formulated to assess performance of the chromosomes reflecting the part quality. Generations after generations, the optimal chromosomes are found only when the evaluated fitness function value falls within a user-defined tolerance. Unlike the typical iterative finite element analysis approach, it was shown that the iterative finite element analysis augmented with genetic algorithm was able to determine feasible pressure and axial feeding paths autonomously. The current approach still requires a lot of simulation runs, which must be offset by high-performance computing resources.

The multi-objective optimization of the loading paths for T-shape tube hydroforming using adaptive support vector regression

“…Di Lorenzo et al [18,19] proposed a gradient decomposition method, which aimed to reduce the number of evaluations of FE simulation, to optimize the internal pressure and counter punch action in Y-shape THF process. Intarakumthornchai et al [20] applied genetic algorithm (GA) to obtain and determine the loading path for fuel filler THF process. Mirzaali et al [21,22] used simulated annealing algorithm to find the optimal loading parameters in THF process with and without axial force loading conditions.…”

A state of the art review of hydroforming technology

“…In general, the fluid pressure is assumed to be hydrostatic and universal throughout the application and follow the pressure curve. These kinds of simulations enable the hydroforming process by allowing for the optimization of a function (often based upon minimal material thinning and wrinkling) [45] instead of a more traditional empirically derived equation method, e.g. a calculation of a pressure curve that suppresses buckling using the energy method [55].…”

“…A load path is a statement of how much fluid pressure (and axial force in the case of tube hydroforming) is applied at a given time through a forming process. They are created by Biterative finite element analyses with adjustments based on forming experience… due to the highly nonlinear nature of the tube hydroforming process [ 45]. There are numerous studies in the relevant literature which explore how process windows and load paths are created and how they can be expanded or optimized but there are many challenges in doing so.…”

“…Alternately, instead of designing a specific function ahead of time, Xiang et al looked at 9 different theoretical load paths and made a determination about which was best by looking at their effects on the results of an FEA analysis [53]. Intarakumthornichai et al used a genetic algorithm based upon a two part quality criteria (minimization of thinning and elimination of wrinkles) [45]. Ge et al used a similar multiobjective optimization function with differential evolution in a tube hydroforming operation to optimize a loading path [54].…”

Optimization of Process Parameters during Hydroforming of Tank Bottom using NSGA-III Algorithm

“…Huang et al [26] used the Kriging method to build proxy model for loading path design of T-tube hydroforming, and the robustness and reliability of this method were verified. Intarakumthornchai et al [27] integrated genetic algorithms into finite element analysis for determination of feasible loading paths.…”