CFD Benchmark of foamStar: validation of a CFD tool

Abstract

In predicting the resistance of ship hulls, experiments in towing tanks have been mostly performed before the extensive use of CFD applications. Time and cost turn out to be key factors in consideration along with the accuracy of results. High costs of experiments and limitations in testing several scenarios in towing tests led to an increased use of CFD codes for ship resistance and propulsion applications. Bureau Veritas (BV) Solutions has interests in developing in-house CFD codes in order to reduce license costs and to enable the engineers to contribute to the code development. One of these CFD tool is being jointly developed by Bureau Veritas (BV) and École Centrale de Nantes (ECN). It is based on OpenFOAM open-source libraries for obvious reasons of accessibility and cost. Its name is foamStar. The work performed in this internship enabled to verify the robustness and sensitivity of foamStar in view of its industrial use on resistance cases. At first, multiple sensitivity studies both on numerical parameters and grid parameters were conducted. It showed the importance of good practices to optimise its use and highlighted some of the key setup options to be used. Moreover, some unstable cases/configurations were encountered showing the poor stability of foamStar, especially with regard to the mesh. Then, the benchmark cases of KCS and DTMB 5415 hulls have been performed in calm water resistance computations. Good agreement was observed in resistance coefficients, free surface pattern and wave profiles between foamStar and experimental results. It can be said that foamStar can capture flow phenomena with the certain degree of accuracy. These results are encouraging and tend to validate the capability of the software for resistance predictions with a small tendency to over-estimate it. For industrialisation purpose, the setup procedure to run foamStar calculations have also been automatized by modifying the BVS’s existing script. The workability and the reliability is well established by testing it on hull optimisation cases. In addition, the results obtained for this optimisation study showed that while over-estimating the resistance, foamStar results are in line with ISIS-CFD results with regard to relative comparison of the various hull forms tested. It is a positive outcome which tends to validate its use for such applications. However, it needs to be investigated further prior to any industrial use as one case is not enough for validation.