Calculations of the Secondary Flow in a Turbine Cascade

Detailed experimental data has been obtained from a low speed linear turbine cascade at Durham University. This data from the so called ‘Durham Cascade’ has been used one of the test cases for a series of seminar/workshops organised by the Special Interest Group in Turbomachinery of the European Research Community On Flow Turbulence And Combustion (ERCOFTAC). A large number of different computations have been made of the secondary flow in this cascade. This paper reviews the various results, together with some from the open literature. It is shown that there is a wide variation between the codes in the prediction of the strength and movement of the main passage vortex, and of the strength and position of the loss core. The predictions for a given code do not vary much qualitatively with grid refinement. There is little evidence that the more sophisticated turbulence models perform any better than the simpler models. The experimental data indicate that the flow is transitional, and so those codes which allow for some specification of the point of transition could show advantages. Overall the Durham Cascade poses a severe test, and it seems that there is some way to go before consistently good predictions by computer codes are achieved.