Feeding VPP applications by means of RANS CFD data for modern hull hydrodynamics design on HPC platforms

Raffaele Ponzini

CINECA

Roberto Pieri

SCS Italy

Andrea Penza, Francesco Pasqua

SCAI - SuperComputing Applications and Innovation – CINECA

Paolo Motta

VPP developer

 

Pushed by hardware and software improvements modern Velocity Prediction Program (VPP) applications are getting more and more rich in terms of predictive capabilities and they are becoming more and more the elective design application in marine yachts design. These improvements are driven by larger amount of more accurate data concerning the forces acting on the hull dynamics. SCAI Department at CINECA has started a feasibility study in order to assess the usability of OpenFOAM as numerical engine to feed modern VPPs with CFD RANS data in an automatic fashion. To do so we first selected a well-established industrial test case used to identify a possible automated and cost-effective work-flow able to provide a short time-to-result on the HPC infrastructure. OpenFOAM results are compared over experimental and numerical (obtained using commercial codes) data in order to evaluate either solver accuracy and performances related to computational costs, scalability on HPC platforms and time-to-result as a byproduct. The preliminary outcomes of the analysis are encouraging in terms of both accuracy and computational efficiency on HPC platforms.

Back