2008JOURNAL OF THE JAPAN SOCIETY FOR AERONAUTICAL AND SPACE SCIENCESOpen access

Computational Efficiency Improvements in Cylindrical SPH Analysis of Hypervelocity Impacts

Ryo Ito

Open full text 1 citations

Abstract

Numerical simulations of solid-solid hypervelocity impacts (HVIs) using smoothed particle hydrodynamics (SPH) method in two-dimensional axisymmetric coordinates, are performed. It is known that Gaussian kernel is adequate for the SPH method in axisymmetric coordinates, however, Gaussian kernel is computationally less efficient because of its infinite range. We propose to adopt a simple function having only even-order terms as the kernel for the SPH method in axisymmetric coordintates, to improve the computational efficiency. The equations of conservation laws, the strain rate and rotation rate are formulated by using the quadric kernel and the quartic kernel. The HVIs are numerically simulated by using the quadric kernel, quartic kernel and Gaussian kernel for comparison. Each numerical result shows a good agreement with the experimental result, and the conservation of energy is assured for each calculation. Consequently, it is found that the quadric kernel and the quartic kernel have far better computational efficiency and comparable computational accuracy to Gaussian kernel.

Open-access reader

About this research paper

What this paper is about

Numerical simulations of solid-solid hypervelocity impacts (HVIs) using smoothed particle hydrodynamics (SPH) method in two-dimensional axisymmetric coordinates, are performed. It is known that Gaussian kernel is adequate for the SPH method in axisymmetric coordinates, however, Gaussian kernel is computationally less efficient because of its infinite range. We propose to adopt a simple function having only even-order terms as the kernel for the SPH method in axisymmetric coordintates, to improve the computational efficiency. The equations of conservation laws, the strain rate and rotation rate are formulated by using the quadric kernel and the quartic kernel. The HVIs are numerically simulated by using the quadric kernel, quartic kernel and Gaussian kernel for comparison. Each numerical result shows a good agreement with the experimental result, and the conservation of energy is assured for each calculation. Consequently, it is found that the quadric kernel and the quartic kernel have far better computational efficiency and comparable computational accuracy to Gaussian kernel.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Numerical simulations of solid-solid hypervelocity impacts (HVIs) using smoothed particle hydrodynamics (SPH) method in two-dimensional axisymmetric coordinates, are performed. It is known that Gaussian kernel is adequate for the SPH method in axisymmetric coordinates, however, Gaussian kernel is computationally less efficient because of its infinite range. We propose to adopt a simple function having only even-order terms as the kernel for the SPH method in axisymmetric coordintates, to improve the computational efficiency. The equations of conservation laws, the strain rate and rotation rate are formulated by using the quadric kernel and the quartic kernel. The HVIs are numerically simulated by using the quadric kernel, quartic kernel and Gaussian kernel for comparison. Each numerical result shows a good agreement with the experimental result, and the conservation of energy is assured for each calculation. Consequently, it is found that the quadric kernel and the quartic kernel have far better computational efficiency and comparable computational accuracy to Gaussian kernel.

Key concepts: Kernel (algebra), Quadric, Quartic function, Kernel smoother, Gaussian function, Mathematics, Hypervelocity, Smoothed-particle hydrodynamics

Related papers

Back to paper searchBrowse research topicsOriginal source
Computational Efficiency Improvements in Cylindrical SPH Analysis of Hypervelocity Impacts — Research Paper | ScholarLens