<p>This paper presents an artificial intelligence (AI) assisted revival of FORTRAN 66 code which calculates laminar flow through curved pipes. Dean numbers <i>D</i> for the entire laminar range are covered: <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(10 \le D \le 5000\)</EquationSource> </InlineEquation>. Results from the code were originally presented in Greenspan (Secondary flow in a curved tube. J Fluid Mech 57: 167–176, 1973). The coupled non-linear system of partial differential equations was solved numerically using a finite difference method. We demonstrate a step-by-step AI-assisted code revival process and compare original (coarse grid resolution) results to updated (fine grid resolution) solutions. Both the structure of streamwise (primary) and secondary flows are covered. The purpose of our paper is both to demonstrate the revival of legacy scientific code and to make the code available as modern Fortran for the scientific community. The code runs quickly on modern hardware architectures and enables fast understanding of the physical effects included.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Application of artificial intelligence to revive numerical studies of fluid motion in a curved pipe

  • Nils Tångefjord Basse

摘要

This paper presents an artificial intelligence (AI) assisted revival of FORTRAN 66 code which calculates laminar flow through curved pipes. Dean numbers D for the entire laminar range are covered: \(10 \le D \le 5000\) . Results from the code were originally presented in Greenspan (Secondary flow in a curved tube. J Fluid Mech 57: 167–176, 1973). The coupled non-linear system of partial differential equations was solved numerically using a finite difference method. We demonstrate a step-by-step AI-assisted code revival process and compare original (coarse grid resolution) results to updated (fine grid resolution) solutions. Both the structure of streamwise (primary) and secondary flows are covered. The purpose of our paper is both to demonstrate the revival of legacy scientific code and to make the code available as modern Fortran for the scientific community. The code runs quickly on modern hardware architectures and enables fast understanding of the physical effects included.