These slides were presented at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

AIAA has recently published a recommended practice for code verification for CFD, R-141-2021, and is now working on the publication of a full new international standard. This standard will be the first international standard for code verification (to the authors knowledge) in any field of engineering simulation and, as such, it represents an important step forward in terms of how to achieve confidence in the simulation tools that the NAFEMS membership uses. Of key importance is the concept of the observed order of accuracy. Code verification requires the error in a simulation to be quantitatively evaluated, which by implication requires the solution to the simulation to be known. The error is calculated on a series of progressively refined meshes so that the trend in the error with the spatial discretization can be determined. Specifically, this trend is plotted on a log log graph and the observed trend should conform to a straight line, the gradient of which is the observed order of accuracy. This should then be compared with the formal order of accuracy claimed by the developer of the code. NAFEMS SGM and CFD working groups have been working on some code verification exemplars to demonstrate the processes set out by AIAA: specifically, they are real cases using real commercial software tools where the observed order of accuracy is found not to correspond to the formal order claimed by the software developer. In each case, the reason for the discrepancy is explored and identified, so that suitable modifications to how the software tool is used can be made, and the tool can be used with confidence. The exemplars demonstrate the essence of the new standard and why considering the observed order of accuracy is important. Whilst the focus of the standard is on the performance of the simulation solver, as the exemplars show, considering the observed order of accuracy can also identify potential issues with the pre and post processing steps.