CFD is now an essential tool for the design of all types of turbomachinery. However, as engineers are exposed more and more to the results of CFD and less and less to experimental data there is a danger that they may not realise the limitations of CFD and so will view its predictions as more reliable than they really are and not question them sufficiently. This is particularly dangerous when CFD is use as part of an optimisation procedure. The objective of this paper is to try to expose some of the limitations of CFD as used for routine turbomachinery design. CFD is not an exact science. Errors can arise from the following sources: • Numerical errors due to finite difference approximations. • Modeling errors, where the true physics is not known or is too complex to model — e.g. turbulence modeling. • Unknown boundary conditions, such as inlet pressure or temperature profiles. • Unknown geometry such as tip clearances or leading edge shapes. • Assumption of steady flow. Each of these sources of error is discussed and examples of the differences they can cause in the predictions are shown. Despite these limitations CFD remains an extremely valuable tool for turbomachinery design but it should be used on a comparative basis and not trusted to give quantitative predictions of performance.

This content is only available via PDF.
You do not currently have access to this content.