Multiscale SPH simulations of viscoelastic injection molding processes based on bead-spring chain model
Xiaoyang Xu,Lingyun Tian,Peng Yu
DOI: https://doi.org/10.1016/j.enganabound.2023.01.035
IF: 3.25
2023-02-03
Engineering Analysis with Boundary Elements
Abstract:In this paper we present an extension of a multiscale smoothed particle hydrodynamics (SPH) method to transient viscoelastic injection molding processes. Specifically, we use the method presented by Xu and Yu, J. Non-Newtonian Fluid Mech. 229 (2016) pp. 27–42, in which the bead-spring chain model is employed to describe the viscoelastic behavior of the fluid without resort to a closed-form constitutive equation. Moreover, to resolve the heavy computation and the time-consuming problem, an efficient parallel algorithm, including parallelizations of both SPH particles and Brownian configuration fields, is developed. To validate the algorithm, we first simulate the viscoelastic Couette flow based on Hookean dumbbell, FENE dumbbell, and FENE chain models. The SPH results are compared with the analytical solutions or those obtained by other methods. The parallel performances of the algorithm are analyzed. Then, we extend the method to the challenging injection molding problem. A number of numerical examples including the injection molding of an F-shaped cavity in two dimensions and a door handle in two and three dimensions are investigated. Some molecular information such as the molecular stretch, the orientation angle, and the mean configuration thickness are displayed. It is found that the injection molding processes could be successfully investigated by SPH from a micro perspective. The proposed parallel algorithm is essential for the efficient simulation of injection molding processes. The maximum speedup can reach more than 130 for the case of the door handle in three dimensions.
engineering, multidisciplinary,mathematics, interdisciplinary applications