Engineering topics

Mehta, Rabindra D.

Publications and source records attributed to Mehta, Rabindra D..

29 records · Page 2

NASA NTRS1990

Development of a two-stream mixing layer from tripped and untripped boundary layers

The effects of the state of the initial boundary layers on the development of a two-stream, plane mixing layer, with a velocity ratio of 0.6 are experimentally investigated. Spanwise-average profiles are compared for the first time. The results indicate that both the near and far-field growth rates for the untripped case are significantly higher than the tripped case. The maximum Reynolds stresses and higher-order products for the two cases behave very differently in the near-field, but asymptote to approximately the same constant levels far downstream. The mean velocity and turbulence profiles in this region also collapse adequately for the two cases when plotted in similarity coordinates. The distance required to achieve self-similarity is distinctly shorter for the tripped case, in contrast to previous observations. The higher growth rate for the untripped case is attributed to the presence of streamwise vortices which result in additional entrainment by the mixing layer.

Bell, James H.↗

NASA NTRS1989

Asymptotic behavior of a flat plate wake

An experimental study has been conducted to investigate the far-field, self-similar properties of a flat plate wake. A plane turbulent wake was generated at the trailing edge of a smooth splitter plate separating two legs of a Mixing Layer Wind Tunnel, with both initial boundary layers tripped. For the present study, both legs were operated at a free-steam velocity in the test section of 15 m/s, giving a Reynolds number based on wake momentum thickness of about 1750. Single profile measurements were obtained at five streamwise locations using a Pitot probe for the mean velocity measurements and a single cross-wire probe for the turbulence data, which included statistics up to third order. The mean flow data indicated a self-similar behavior beyond a streamwise distance equivalent to about 350 wake momentum thicknesses. However, the turbulence data show better collapse beyond a distance equivalent to about 500 momentum thicknesses, with all the measured peak Reynolds stresses achieving constant, asymptotic levels. The asymptotic mean flow behavior and peak primary stress levels agree well with theoretical predictions based on a constant eddy viscosity model. The present data also agree reasonably well with previous measurements, of which only one set extends into the self-similar region. Detailed comparisons with previous data are presented and discussed in this report.

Weygandt, James H.↗

NASA NTRS1989

Design and calibration of the mixing layer and wind tunnel

A detailed account of the design, assembly and calibration of a wind tunnel specifically designed for free-shear layer research is contained. The construction of this new facility was motivated by a strong interest in the study of plane mixing layers with varying initial and operating conditions. The Mixing Layer Wind tunnel is located in the Fluid Mechanics Laboratory at NASA Ames Research Center. The tunnel consists of two separate legs which are driven independently by centrifugal blowers connected to variable speed motors. The blower/motor combinations are sized such that one is smaller than the other, giving maximum flow speeds of about 20 and 40 m/s, respectively. The blower speeds can either be set manually or via the Microvax II computer. The two streams are allowed to merge in the test section at the sharp trailing edge of a slowly tapering splitter plate. The test section is 36 cm in the cross-stream direction, 91 cm in the spanwise direction and 366 cm in length. One test section side-wall is slotted for probe access and adjustable so that the streamwise pressure gradient may be controlled. The wind tunnel is also equipped with a computer controlled, three-dimensional traversing system which is used to investigate the flow fields with pressure and hot-wire instrumentation. The wind tunnel calibration results show that the mean flow in the test section is uniform to within plus or minus 0.25 pct and the flow angularity is less than 0.25 deg. The total streamwise free-stream turbulence intensity level is approximately 0.15 pct. Currently the wind tunnel is being used in experiments designed to study the three-dimensional structure of plane mixing layers and wakes.

Bell, James H.↗

NASA NTRS1989

Boundary-layer predictions for small low-speed contractions

The present scheme for the prediction of boundary-layer development in small, low-speed wind tunnel contraction sections proceeds by calculating the wall pressure distributions, and hence the wall velocity distributions, by means of a three-dimensional potential-flow method. For the family of contractions presently treated, the assumption of a laminar boundary layer appears to be justified; the measured boundary layer momentum thicknesses at the exit of the four contractions were found to lie within 10 percent of predicted values.

Mehta, Rabindra D.↗

NASA NTRS1989

Three-dimensional structure of a plane mixing layer

An experimental study is reported which shows the spanwise structure of a plane mixing layer originating from laminar boundary layers. Quantitative measurements are obtained for the first time which show the origin and evolution of streamwise vortices within the mixing layer. After the first spanwise vortex rollup occurs, streamwise structures are triggered in clusters within the braid region of the mixing layer. Further downstream, they realign to form pairs of counterrotating vortices. The maximum vorticity diffuses rapidly with increasing downstream distance while the vortices grow, scaling with the mixing layer vorticity thickness. The presence of the streamwise vortices leads to significant spanwise distortions in the mean and turbulence properties of the mixing layer. A consistent peak in the secondary shear stress exists for each streamwise vortex.

Bell, James H.↗

NASA NTRS1989

An experimental study of forced streamwise vortical structures in a plane mixing layer

Streamwise structures have been shown to ride among the primary spanwise vortices in past flow visualization investigations of plane mixing layers. More recently, quantitative measurements were obtained which showed the origin and evolution of streamwise vortices within a mixing layer. In the present study, the effects of perturbing the mixing layer using two different mechanisms are investigated. A serration on the splitter plate trailing edge was found to have a relatively small effect, confined to the near-field development of the streamwise structures. The installation of cylindrical pegs in the high-speed side boundary layer, however, not only generated a regular array of vortex pairs, but also affected the mean development of the mixing layer far downstream. In both cases, the mean streamwise vorticity was found to decay rapidly with increasing downstream distance.

Mehta, Rabindra D.↗

NASA NTRS1988

Contraction design for small low-speed wind tunnels

An iterative design procedure was developed for two- or three-dimensional contractions installed on small, low-speed wind tunnels. The procedure consists of first computing the potential flow field and hence the pressure distributions along the walls of a contraction of given size and shape using a three-dimensional numerical panel method. The pressure or velocity distributions are then fed into two-dimensional boundary layer codes to predict the behavior of the boundary layers along the walls. For small, low-speed contractions it is shown that the assumption of a laminar boundary layer originating from stagnation conditions at the contraction entry and remaining laminar throughout passage through the successful designs if justified. This hypothesis was confirmed by comparing the predicted boundary layer data at the contraction exit with measured data in existing wind tunnels. The measured boundary layer momentum thicknesses at the exit of four existing contractions, two of which were 3-D, were found to lie within 10 percent of the predicted values, with the predicted values generally lower. From the contraction wall shapes investigated, the one based on a fifth-order polynomial was selected for installation on a newly designed mixing layer wind tunnel.

Bell, James H.↗

NASA NTRS1988

Contraction design for small low-speed wind tunnels

An iterative design procedure was developed for 2- or 3-dimensional contractions installed on small, low speed wind tunnels. The procedure consists of first computing the potential flow field and hence the pressure distributions along the walls of a contraction of given size and shape using a 3-dimensional numerical panel method. The pressure or velocity distributions are then fed into 2-dimensional boundary layer codes to predict the behavior of the boundary layers along the walls. For small, low speed contractions, it is shown that the assumption of a laminar boundary layer originating from stagnation conditions at the contraction entry and remaining laminar throughout passage through the successful designs is justified. This hypothesis was confirmed by comparing the predicted boundary layer data at the contraction exit with measured data in existing wind tunnels. The measured boundary layer momentum thicknesses at the exit of four existing contractions, two of which were 3-D, were found to lie within 10 percent of the predicted values, with the predicted values generally lower. From the contraction wall shapes investigated, the one based on a 5th order polynomial was selected for newly designed mixing wind tunnel installation.

Bell, James H.↗

NASA NTRS1988

A streamwise vortex embedded in a plane mixing layer

Experimental results are presented on the interaction of a single streamwise vortex with a plane turbulent mixing layer. In the present setup, the vortex is generated by a half-delta wing mounted in the settling chamber of a blower-driven wind tunnel. Initially, the vortex is shown to ride just below the mixing layer, locally distorting the mean velocity and turbulence Reynolds stress distributions in the lower part of the mixing layer. Once the vortex becomes embedded within the mixing layer, the distortions are found to spread through the whole width of the layer, and the induced extra strain rates in the mixing layer result in the generation of additional, relatively large, Reynolds normal and shear stresses, mainly in the regions above the vortex.

Bell, James H.↗

NASA NTRS1987

Interaction of an oscillating vortex with a turbulent boundary layer

The effects on mean flow and turbulence caused by meander of a vortex embedded in a two-dimensional boundary layer were investigated experimentally by driving a forced lateral oscillation of the vortex generator. Upstream, the vorticity contours without forcing were found to be round. The forced vortex generator oscillation caused a flattening of the time-averaged vorticity contours and changes in some of the Reynolds stresses. The results indicate that the unforced vortex did not meander significantly upstream, and that the effects of meander can be understood qualitatively from production of Reynolds stresses by the forced motion acting with the exiting three-dimensional velocity field. Farther downstream, the observed differences in the mean vorticity and Reynolds stresses caused by forcing were smaller, mainly because the vortex was substantially diffused at this station, resulting in smaller mean velocity gradients.

Westphal, Russell V.↗

NASA NTRS1987

Properties of a half-delta wing vortex

The mean flow and turbulence structure of a single longitudinal vortex generated by a half-delta wing placed at a small angle of attack were investigated. Particular consideration was given to the near-field properties of the generator in order to establish the role of the generator wake in the initial rolling-up of the vortex, as well as to the far-field properties so that the approach to equilibrium could be studied. Measurements were made on fine cross-plane grids at seven streamwise locations using hot cross wires. The results show that the point of maximum vorticity and the generator wake do not merge until a streamwise distance equivalent to three generator heights is reached. Comparison with previous data on vortices produced by double-branched generators confirmed that the present vortex had achieved a fully developed state, and at a relatively short streamwise distance.

Mehta, Rabindra D.↗