An experimental study of a plane mixing layer.

Abstract

Examination of existing literature is revealing in that although many investigations have been carried out for a plane mixing layer, only a few present turbulence measurements. It is well-known that both mean velocity and turbulence structure in a plane mixing layer are self-preserving, however there appears to be some variations in these measurements. This investigation presents new data on the mean velocity and the turbulence properties of a plane mixing layer. The turbulence measurements differ by about 25% from those obtained by previous investigators and this is attributed to differences in the experimental set up (i.e., presence or absence of a solid wall in the plane x = 0) and errors associated with hot-wire probes (i.e., longitudinal cooling and wake interference). To avoid difficulties experienced by previous investigators in calculating shear stress distribution from a measured nondimensional mean velocity profile a new method is suggested and this provides good agreement between the calculated and the measured shear stress distribution. ECENTLY, considerable effort has been directed towards the study of free turbulent shear flows. In this group of shear flows a plane, turbulent, incompressible mixing layer between a uniform stream and quiescent surroundings is a com- paratively simple flow to investigate both theoretically and experimentally. However, except for a few investigations (as for example Garshore1; Hackett and Cox2) dealing with mean velocity measurements not much renewed attention appears to have been given to the plane turbulent mixing layer since the appearance of the work of Liepmann and Laufer.3 It should be recalled that Liepmann and Laufer did not measure both the lateral turbulent intensity and the shear stress presumably because is expected to be of the same magnitude as and is expected to be zero. Furthermore, techniques of hot-wire anemometry have developed considerably since their investigation and therefore it is of interest to reinvestigate the plane mixing layer. Since the completion of the present investigation Wygnanski and Fiedler4 have reported extensive and sophisticated turbu- lence measurements in this type of flow. An interesting aspect of their investigation was the geometry of the experimental apparatus. They used a trip wire and a solid wall on the zero velocity side in the plane x = 0. They also used conventional DISA x-wire probes to measure the turbulence parameters and it is now known that the conventional DISA x-wire probes are contaminated by the thermal cross talk.5 The measurements of Wygnanski and Fiedler are compared with the results of the present investigation obtained with single hot-wires in a plane mixing layer without a solid wall in the plane x = 0. It is anticipated that such a comparison would show up differences, if any, due to the geometry of the experimental apparatus and the probes. Recently Spencer and Jones6 have investigated a general problem of a mixing layer between two parallel streams and as a special case they have reported some data on a plane mixing layer. This information consists of the spreading rate and the mean velocity profile. Although not directly related to the