Improved Artificial Viscosity in Finite Element Method (FEM) for Hypervelocity Impact Calculations. (2015)

• Record Type: Journal Article
• Authors: Stowe, David; Kupchella, Ryan; Cogar, John
• Is Part Of: Procedia engineering. Volume 103(2015)
• Page Start: 593
• Page End: 600
• Publication Date: 2015
• Subjects: artificial viscosity -- numerical method -- numerical stability -- numerical accuracy -- shock propagation; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Conference proceedings; Periodicals; 620.005
• DOI: 10.1016/j.proeng.2015.04.077
• Languages: English
• ISSNs: 1877-7058

Abstract

Abstract: In this paper we develop methods based primarily on the work of Kuropatenko and Wilkins to improve the application of artificial viscosity in 3D finite element method (FEM) codes. The primary goal is to obtain better shock predictions for hypervelocity impacts (HVI) and reduce the need for user calibration. We focus on examining factors such as geometric variability with respect to shock direction, dynamic adaptation to changes in compressibility in the shock front, and anisotropic compression in multi-dimensional formulations. We implement the methods in the Velodyne hydro-structural code and investigate the effects on shock propagation using a series of simple flyer impact test cases which cover a range of system responses including strong and weak shocks. Various initial mesh geometries are utilized to examine mesh effects. Energetic materials using the Ignition and Growth Reactive Burn (IGRB) equation of state (EOS) are also examined due to the rapid change in compressibility and energy density which occurs due to reaction. These rapid changes can lead to insufficient damping in artificial viscosity calculations and thus provide an effective test case. We employ the CTH hydrocode to evaluate baseline shock behavior. The regular, ordered mesh of CTH allows for a consistent and precise application of the artificial viscosity. Direct numerical comparisons are used rather than experimental data to eliminate uncertainty due to factors such as material characterizations, EOS models, and mesh resolution. We compare the CTH results against various FEM artificial viscosity implementations to evaluate performance. It is demonstrated that shock response in FEM codes can be significantly improved by using updated artificial viscosity methods.