Search

A number of aspects of transition from laminar to turbulent flow were investigated on a flat plate with free stream water velocities between .75 feet per second and 1.2 feet per second. The study was primarily visual, using dye techniques, although some measurements were made using a hot film anemometer. Also investigated were the later stages of natural transition. Dye patterns associated with the chain of events leading from laminar to turbulent flow in natural transition were analyzed. (Author).

The bistable, steady, turbulent separation phenomenon observed in a two-dimensional fully-stalled diffuser is studied. Detailed mean flow data including pressure, dissipation and velocity (both magnitude and direction) inside the flow field as well as development of boundary layers on both side walls are reported for cases with thin and thick boundary layers at inlet and with and without fairing curves at the throat. For all cases, most important changes in flow data occur in a region within two inlet widths up and downstream of the throat. Pressure gradients in this region are so large that the boundary layer on the unstalled side changes shape abruptly while the layer on the stalled side separates in the vicinity of the throat. Streamline patterns are obtained from the velocity distributions. The location of the dividing streamline is different for the different cases. Based on the dividing streamline, a displacement line is defined to extend the concept of displacement thickness to the free shear layer which is developed from the separated boundary layer. This displacement line together with the loci of boundary layer displacement thicknesses on the solid walls form an effective channel. The shape and area ratio of this effective channel explain the different trends of pressure distributions in the main flow. (Author).

An investigation of the use of flat vanes in two-dimensional subsonic diffusers was made. Using optimum designs of vane installations, high pressure recoveries and steady flows were obtained for diffuser-wall divergence angles up to 42 degrees. Criteria for optimum configurations were developed which indicated that the vanes should be symmetrically arranged in the vicinity of the diffuser throat, that the vane-passage divergence angle should be approximately 7.0 degrees, and that the vanes should have a certain predictable length dependent upon the diffuser geometry.

A combination of visual and quantitative measurements is presented, providing a physical picture of the turbulent boundary layer flow structure on a flat plate. The flow structure is shown to consist of three zones, each zone has a one to one correspondence to the well known regions of the u+, y+ mean velocity profile. A wall layer region is shown to exist below y+ = 10. An apparently fully turbulent region exists corresponding to the logarithmic ''law of the wall'' and the ''buffer'' region. An intermittent zone appears to agree closely with the ''wake'' deviation region. An entirely new result of the investigation is the delineation of the structure of the wall layer region. This region is shown to contain a relatively regular structure of low and high velocity fluid streaks alternating in the span direction, together with the ejection of low momentum fluid into the outer flow. Correlations are given for the rate of ejection and the streak spacing. A qualitative description of other features of the wall layer region and the character of the remainder of the boundary layer flow structure is presented. (Author).

A unified performance prediction method was developed for class A diffusers. Class A diffusers have specified flow and geometry restrictions. The prediction method is a system of linear, first order differential equations. These equations embody the momentum equation, continuity equation, and two empirical correlation equations. In addition to the prediction method, two parameters were developed to characterize inlet conditions and separation respectively.