THE USE OF INVERSE METHOD OF SOLVING BOUNDARY-LAYER EQUATIONS FOR THETESTING OF TURBULENCE MODELS

Comparison is made of the classical direct method and the so-called in verse method of solving turbulent boundary-layer equations as applied to the problem of testing of turbulence models. In the case of the dir ect method, it is assumed that the velocity distribution along the ext ernal bound of the boundary layer is known from experiment, and, in th e case of the inverse method, it is assumed that it is the longitudina l distribution of the displacement thickness that is preassigned from experiment. It is demonstrated that the inverse method (which, unlike the direct one, does not require the preassignment of the longitudinal gradient of velocity at the external bound of the boundary layer) ena bles one to derive more objective (free of arbitrariness in preassigni ng the input data) information about the capabilities of the turbulenc e models being tested. This method is used to perform a detailed inves tigation of the capabilities of a wide scope of algebraic, semidiffere ntial, and differential turbulence models in calculating boundary laye rs with adverse, accelerating, and alternating-sign pressure gradients , and conclusions are formulated about the advantages and disadvantage s of individual models.