Experiments in Fluids 6, 327-334 (1988) Experiments in Fluids 9 Springer-Verlag 1988 Three-component LDA measurements in a turbulent boundary layer O. Ozcan Faculty of Aeronautics and Astronautics, Istanbul Technical University, Istanbul, Turkey Abstract. An experimental study of a three-dimensional, pressure- driven, attached turbulent boundary-layer flow was made at Mach 0.4. Both the mean velocities and the full Reynolds stress tensor were measured simultaneously by a three-component LDA system. Value of the resultant shear stress to turbulent kinetic energy ratio varied between 0.1 and 0.2 and did not remain constant across the boundary-layer. Slopes of the streamwise and azimuthal mixing length distributions in the wall region were around 0.4 and 1.2, respectively. Skew angle of the turbulent shear stress was larger than skew angle of the velocity gradient. 1 Introduction Progress in understanding the momentum transfer in three- dimensional turbulent shear-layers, as well as development of turbulence models and prediction methods rely heavily on detailed experimental data. Measurements should include all components of the mean velocity vector and the Reynolds stress tensor, magnitude and direction of skin-friction and the outer flow conditions. This paper presents such data for the flow developing over a swept circular-arc bump which simulate the highly three-dimensional boundary-layer flow over a swept-wing. The swept-bump geometry provides two planes of symmetry and well-defined boundary conditions and, therefore, is ideally suited for a straightforward compu- tation. Data of the reported flow can provide guidance and validation for numerical solutions of the Navier-Stokes equations which are becoming more common with recent advances in computer technology. The presented data set should also be useful for development, evaluation and com- putational validation of turbulence closure models for three- dimensional shear flows. Oil flow visualization was performed to measure the di- rection of skin-friction. Magnitude of skin-friction was ob- tained from the measured velocity profiles. Mean velocity measurements were made in the inviscid flow region to ob- tain the static pressure distributions. All three components of the mean velocity vector and six components of the Rey- nolds stress tensor were measured simultaneously by a three- component LDA system. The streamwise distance measured from the bump leading edge was observed to be a correla- tion parameter for all mean flow quantities. Several assump- tions used for turbulence modeling of three-dimensional boundary-layers were checked for their validity in this flow. Surveys of experimental studies on three-dimensional turbulent boundary-layers are given by Johnston (l 976) and Cebeci (1985). Turbulent boundary-layer data obtained by three-component laser anemometers are reported by Yanta and Aushermann (1983) and Driver and Hebbar (1987). Fernholz and Vagt (1981) point out that the turbulent shear- stress components involving the velocity fluctuations in the spanwise direction are influenced in different ways by the streamwise and azimuthal pressure gradients. The interest- ing features of the flow of the present study are the attached flow conditions and the existence of both favorable and adverse streamwise and azimuthal pressure gradients. The resultant boundary-layer does not have a uniformly mono- tonic skewness. The unseparated flow structure is expected to be free from any large scale unsteadiness which may con- tribute spuriously to the turbulent stresses. In addition to the pressure gradients, the concave/convex streamwise and spanwise curvature of the model should also affect the struc- ture of the nonequilibrium turbulence field. 2 Experimental setup and conditions The experiments were conducted in the NASA-Ames 6 x 6 ft. Supersonic Wind Tunnel. The free-stream Mach number, Reynolds number, velocity, pressure and tempera- ture were Moo =0.4 _+ 0.005, Re = (6.10 _+ 0.01) x 106 (l/m), U| = 134.5 + 1.5 (m/s), p~ = (6.93 + 0.06) x 10~ (N/m z) and T~ = 16.2 + 2.1 ~ respectively. The model consisted of a swept circular-arc profile bump mounted on a hollow cylin- der aligned with the freestream direction. Figure I gives a schematic of the model and shows the three tangent circles which define the bump profile. Figure 1 also shows the (x, y, z) cartesian and the (x, r, 0) cylindrical coordinate sys- tems whose origins are located on the symmetry axis of the cylinder 590.6 mm downstream of the cylinder leading edge. The sweep angle of the bump is 45 ~ . The cord length of the bump and the maximum bump thickness are c = 260.4 mm