Font Size: a A A

STUDY OF BOUNDARY LAYER STRUCTURE USING LASER DOPPLER VELOCIMETRY, CONDITIONAL SAMPLING, AND POLYMER ADDITIVES (HOT-FILM, VITA, WATER TABLE)

Posted on:1987-11-14Degree:Ph.DType:Dissertation
University:Harvard UniversityCandidate:ARIK, BAHA ENGINFull Text:PDF
GTID:1470390017958379Subject:Engineering
Abstract/Summary:
The results of an investigation of the boundary layer structure on a free-surface water table are presented. Two-dimensional laminar flows as well as turbulent flows are achievable up to Reynolds number (based on mean velocity and flow depth) of 8,900. Strain rates and flow depth can be varied independently of the Reynolds number. Flow thicknesses in the range from 100 to 500v/u(,(tau)) (viscous wall units) can be obtained.; A small scattering volume (SSV) LDV system has been constructed, with a probe volume having diameter and length of approximately 1 and 10 viscous units respectively, and with 15 fringes in water. Laminar mean velocity profiles measured using SSV LDV agree well with calculations, hot film, and selective seeding LDV measurements, thus validating the uniform seeding SSV LDV.; Measurements with SSV LDV in turbulent flows agree very well with universal profile data from other flow geometries for fully developed flows. The extent of the log-region does not affect the universal turbulent profile, and the presence of the free-surface has negligible effect on the flow.; Average bursting frequencies (f(,B,avg)) are measured using the VITA (Variable Interval Time Averaging) technique on the u-velocity signal from SSV LDV, and two different length hot film probes. The results are in good qualitative agreement with each other, and with results of other investigations. Their quantitative differences are explainable by spatial averaging and probe orientation effects. For fully developed flows, scaling of f(,B,avg) with wall variables seems appropriate. Non-fully-developed flows show decreased frequency of bursting.; The VITA technique can not be used directly as a diagnostic tool in laminar flows, since it predicts a higher number of "indicated bursts" than in turbulent flows. The cause of this serious procedural weakness has been understood by studying measured laminar and turbulent normalized local variance signals. The results have been confirmed by analysis of computer simulated laminar flow LDV tracker signals, and "synthesized" turbulent flow signals.; Finally, a novel technique for flow diagnostics using the mean of the normalized local variance signals has been developed. This "flow indicator" technique shows "laminarization" (i.e. inhibition of bursting) in the near wall region of drag reducing polymer flows, and can be used in boundary layers unambiguously to indicate whether they are laminar, turbulent, or drag reducing.
Keywords/Search Tags:Boundary, Flows, Laminar, SSV LDV, VITA, Water, Turbulent, Using
Related items