Experimental Validation of Numerical Simulations: A Comparison of Computational Fluid Dynamics and the Oil Film Method

Background Today Computational Fluid Dynamics (CFD) is used for simulating flow in many applications. The quality of the results, however, depends on various factors, like grid quality, boundary conditions and the computational model of the fluid. For this reason, it is important to validate the performed computation with experimental results. In this work, a comparison of numerical simulation with the oil film method was performed for two cardiovascular applications.Methods The investigations were conducted at various geometries, such as a bended cannula tubing, an impeller of a magnetically levitated rotary blood pump and tips of inflow cannulas. The oil film for the experimental validation was composed of black oil color and varnish. In the numerical simulation, color abrasion was displayed with a special post-processing tool by means of wall-attached pathlines.Results With the proper choice of numerical parameters, the computer simulations and the oil film method demonstrated good correlation. Improper generation of the simulation grid did lead to divergent results between the numerical simulation and the experiment. For the pump impeller as well as for the inflow cannulas, the calculation and the experiment showed similar flow patterns with backflow and stall zones.Conclusion The oil film method represents a fast and simple approach to help validate numerical simulations of fluid flow. The experimentally generated near wall flow patterns can be easily compared with the solution of the CFD analysis.