Design and realization of automated testing cloud platform for CFD verification and validation

Based on the idea of cloud computing, this paper proposes an automated testing cloud platform solution for large-scale computational fluid dynamics (CFD) software verification and validation. The main objective is to reduce labor costs, improve software quality and development efficiency, and be adapted to high performance computing for future development as well. This platform uses modularized browser/server (B/S) network architecture and LAMP (Linux+Apache+MySQL+PHP/Python) development tools. A continuous integration professional database is established, and a cloud environment is built, covering reliable cluster computing, complicated operation scheduling and massively parallel computing. This solution is able to automatically load test cases, submit cluster computing, monitor real-time process, post-process automatically and export analysis results containing comparisons with detailed experimental data, error analysis and summary reports. This platform has been applied to a large-scale parallel computing CFD software, and the feasibility and practicability of this solution have been verified.