Fluids and Thermal

An investigation has been conducted to determine the applicability of optical matched filtering (OMF) concepts (Vander Lugt, 1960) to interpretation of holographic particle velocimetry (HPV) holograms (Trolinger et al, 1969). The motivation for this work is that the conventional procedure of CCD data readout and digitization before processing is attended by serious difficulties (Meng et al, 1993). However, investigators had not previously determined how OMF could be applied to this problem.

The theory developed in this study shows that not only can a viable optical system (Fig. 1) be designed to employ OMF for readout of the HPV holograms but that such a system embodies the principle of wavelet transformation (Grossman and Morlet, 1984; Mallet, 1989), a subject currently of keen interest relative to voice, data and video communications, computer vision, image compression, and numerous other areas. In the context of HPV the principle involved is one of scaling the OMF through whatever range is required to accommodate the varying particle depths as they existed at the moment of hologram exposure. For image compression (Wallace, 1991), the OMF is designed to extract essential features and their relative scales from the image. Future work will be addressed to image compression possibilities as well as HPV data reduction.

References
Grossman, A. and J. Morlet. "Decomposition of Hardy Functions into Square Integrable Wavelets of Constant Shape." Siam J. Math. Anal.15 (1984): 723-736.
Mallet, S.G. "A Theory for Multiresolution Signal Decomposition: the Wavelet Representation." IEEE Trans. Pattern Analysis and Machine Intelligence 11 (1989): 674-693.
Meng, H., W. L. Anderson, F. Hussain and D. Liu. "Intrinsic Speckle Noise in In-Line Particle Holography." J. Opt. Soc. Am. (to be published).
Trolinger, J.D., R.A. Belz and W.M. Farmer. "Holographic Techniques for the Study of Dynamic Particle Fields." Applied Optics 8 (1969): 957-961.
Turin, G.L. "An Introduction to Matched Filters." IRE Trans. Inform. Theory (1960): 311-329.
Vander Lugt, A.B. "Signal Detection by Complex Spatial Filtering." IEEE Trans. Inform. Theory IT-6 (1960): 311.
Wallace, G. K. "The JPEG Still Picture Compression Standard." Communications of the ACM 34 (1991): 31-44.

Figure 1. Illustration of one method for scaled optical matched filtering The matched filter ('F") is located in the (x2,y2) plane), and the hologram ("H") is illuminated with a diverging wavefront. In practice, the source of the diverging wavefront would be a virtual point source created by a negative lens, which would be shifted along the z-direction to give a range of effective values of d.

An Experimental Investigation of High-Turbulence Heat Transfer

The aerospace industry has a critical need for reliable predictions of the heat transfer rates into gas turbine blades. Data from running gas turbine engines indicate that the heat transfer coefficients on the blades are 1.5 to 3.0 times larger than those predicted by current boundary-layer theory. This discrepancy is due mainly to the effect of a highly turbulent free-stream flow surrounding the blades. Velocity fluctuations as high as 30% of the local average velocity have been measured in the free-stream flow in running engines. The poor understanding of heat transfer in these conditions is a major impediment in the design of new engines employing advanced- technology blade materials and advanced performance levels. A better understanding would allow new engine designs to proceed with less feedback from costly engine tests.

Applications also exist in which a highly turbulent free-stream is purposely produced to enhance heat transfer. One example is the cooling phase of a materials processing operation in which the material is cooled by heat transfer in a turbulent pool of liquid. The production of new-technology alloys requires greater precision in tailoring this cooling process and in predicting the effects of free-stream turbulence. One Houston-area company, Cooper Industry's Cameron Division, worked with this laboratory on a specific high-turbulence quenching application.

A unifying descriptor of past experimental results is limited by the wide variety of characteristics that describe the intensity, scale, and orientation of the velocity fluctuations in a highly turbulent flow. The most basic of these descriptions is "turbulence intensity" which is the ratio of the root-mean-square velocity fluctuation to the average velocity. Experiments have shown that heat transfer can be enhanced by 20% to 50% for moderate levels of free-stream turbulence intensity (10% or less), and by as much as a factor of 5 for turbulence intensities of 50% to 60%.

Free-stream turbulence flows
The method of generating free-stream turbulence in the laboratory is an important problem. Grid-generated turbulence is well described, but it is of relatively low intensity (<10%) and it decays quickly as it convects downstream. Turbulence generated by other means can have a much greater intensity (>20%) and lon-gevity. However, the hydrodyna-mic description is more complex, and problems such as reversing flow and poor repeatability may limit the applicability of the results. This project investigates a solution to the problems encountered in other experiments by using a turbulent two-stream mixing layer to produce the free-stream flow over a heated plate. Free-stream turbulence produced by a mixing layer has three advantages:
(1) Turbulent mixing layers and wakes can form in gas turbine engines downstream of combustor sections, individual blades, and other engine parts. These layers can impact blades directly or contribute to the level of turbulence around the blades.
(2) The classical mixing layer is well-described in the literature. It is known to contain large-scale, coherent motions that may be important in heat transfer enhancement.
(3) The turbulence intensity reaches a value that is uniform in the flow direction. The opportunity to examine heat transfer in the presence of sustained high levels of free- stream turbulence energy is an innovative and intriguing aspect of this project.

A preliminary survey of this flow field without heat transfer was the subject of a 1991 master's of science thesis written by H. A. Bourgogne and supported by the Energy Laboratory supported during the summer of 1991. A streamwise-uniform turbulence intensity of over 15% was obtained for a velocity ratio across the mixing layer of 2.5:1. At that turbulence intensity, the surface shear stress was enhanced by as much as 72% compared to standard predictions. These results show that the UH facility can produce a unique and very aggressive high-turbulence flow field.

Project accomplishments
During the summer of 1992, a heated test surface was designed and partially constructed. This activity is the central focus of the master's thesis of Clarissa Belbas, the student supported by the project and awarded an AMOCO Graduate Fellowship. Her work will be completed when the apparatus is operating and preliminary heat transfer data obtained and analyzed. Specific ISSO project accomplishments are:
(1) design of a state-of-the-art instrumented heat transfer surface.
(2) purchase of the complete instrumentation system to be used with the project (these items were purchased with equipment funds from another source).
(3) development of the various techniques required to build the heated surface.
(4) partially completed construction of the apparatus.
(5) development of the data acquisition system and software.

Design and construction of the instrumented test surface
The heated surface has the same dimensions as the unheated plate, but a 0.002" thick stainless steel foil covers both faces of the plate. The subframe of the plate is constructed of aluminum square-tubing, and fiberglass panels form the upper and lower faces of the plate. Developing the precise technique for applying thin foil sheets to the fiberglass panels was a major task of the project. After unsuccessful attempts with other techniques, a high-temperature double-stick adhesive membrane was chosen as the best system for securing the foil to the sub-surface of the plate. The surfaces of the plate are heated by passing an electric current through the foil. This arrangement produces a good experimental approximation of a uniform-heat-flux boundary condition. A current of about 160 amperes is required to heat the foil to 15&degree;C to 20&degree;C above the ambient wind tunnel temperature. An array of approximately 150 thin-film-resistance temperature sensors attach-ed to the underside of the foil (internal to the plate) measures surface temperature distribution. Temperature sensors are interrogated by an 80-channel multiplexer system connected to a digital voltmeter and controlled by an Intel-type personal computer. Also, plans are being laid to cover a portion of the surface with thermochromic liquid crystals for attempts to image the surface temperature field directly with a color still or video camera.

References
Bourgogne, H.A. "The Development of a Turbulent Boundary Layer Beneath a Two-Stream Mixing Layer." M.S. Thesis, Dept. of Mech. Eng., Houston, 1991.
Hollingsworth, D. K., R. J, Moffat. and J. P Johnstone, .. "Observations of the Effects of High Free-Stream Turbulence Levels on the Heat Transfer to a Concavely Curved Turbulent Boundary Layer." Report HMT-39, Thermosciences Div., Dept. of Mech. Eng., Stanford University, 1989.
Johnson, P. L. and J. P. Johnston. "The Effects of Grid-Generated Turbulence on Flat and Concave Turbulent Boundary Layers." Report MD-53, Thermosciences Div., Dept. of Mech. Eng., Stanford University, 1989.
Maciejewski, P. K. and R. J. Maffat. "Effects of Very High Turbulence on Heat Transfer." Proc. of the Seventh Symp. on Turbulent SHear Flows, Stanford, California. 1989.

On the Control and Detection of Long-Lived Wakes

IT is found that hydrodynamic instability patterns and wakes can be put into correspondence with a globally stabilized subset of surfaces of tangential discontinuities (which consist of topological limit points). This theoretical result creates a basis for a new understanding of hydrodynamic wakes, their creation, their control, their elimination, and their associated noise production and detection.

Counter to the conventional wisdom which presumes that most wakes are artifacts of incompressible viscous diffusion (div V = 0, curl V # 0), the topologically based theory implies that the creation of wakes is a diffraction phenomena associated with the finite compressibility of all fluids (div V # 0, curl V = 0) and independent from viscosity. Both extremes have their domain of applicability, but the older viscous methods have not been very successful in describing, analytically, the details of observable wakes. On the other hand, new topological methods, based on the qualities of hyperbolic open domains, yield an impressive comparison with experiment. In addition, it appears that a new type of "torsional" wave phenomena, different from longitudinal sound, is permissible in the discontinuity layer of a fluid. In such layers long-lived wake signatures are possible.

The point of departure from current methods of hydrodynamic analysis stems from the fact that all real fluids are slightly compressible and that in the topological open (or hyperbolic) domains of such fluids (near sharp edges and corners or abutments) there exists the possibility of non-unique solutions to hydrodynamic flow problems; i.e., real discontinuities can be created and exist for finite times in hyperbolic domains. Conventional wisdom subsumes that such discontinuities are not "physical," a prejudiced point of view built on the dogma that only uniqueness is admissible in the scientific arena. Although such domains of discontinuity are generally unstable (a fact that exacerbates the prejudice described above), there exists a subset of singular solutions of tangential discontinuities, which, though locally unstable, are globally stabilized by synergetic interactions. These globally stabilized surfaces are like soap films, in that they are associated with minimal surfaces.

Investigation indicates that there can exist persistent long-lived wake signatures that admit of measurement. Torsional wave modes can exist that propagate differently from longitudinal sound waves, and these waves can propagate along the diffraction discontinuity surface without significant attenuation. These torsion waves have a transverse component, which implies that it is theoretically possible to conceive of an induction mechanism that would permit tuned antennas to be constructed for the detection and production of such torsion waves. Such tuned antennas could have high noise rejection qualities found in electromagnetic antennas. The control of hydrodynamic wakes would be useful in submarine warfare, on the flight deck of aircraft carriers, in aircraft with long wings and high aspect ratios having the best lift-to-drag ratios, in the dynamic, unsteady operation of butterfly flow valves, such as the main hydrogen fuel lines of the NASA orbiter, and in numeric simulations of orbiter reentry.

Liquid Crystal Imaging of Surface-Tension-Driven Bčnard Convection

D. Keith Hollingsworth, Ph.D., and Jennifer Drapp, Department of Mechanical Engineering

Surface-driven-tension convection (a.k.a. Bénard convection) is a common phenomenon in microgravity environments. Spatial variations in fluid temperature at the free surface cause surface tractions that will induce fluid motion. In delicate microgravity experiments (crystal growth or biological tissue development), such fluid motions are disruptive. At the UH Heat Transfer Laboratory, investigators have been studying the feasibility of using liquid crystal thermography to map the thermal patterns of Bénard convection in thin layers of silicone oil.

To use liquid crystal thermography for effectively mapping temperature distributions in a heated fluid layer, the heater must approximate a uniform-heat-flux boundary condition. Previous Bénard convection experiments have all used heaters that approximate uniform-temperature boundaries. Thus, the first step in establishing feasibility was to ensure that convection cells would occur in response to uniform heating. After thorough research, investigators determined, to their best knowledge, that no one had yet achieved this effect experimentally. Using an electrically heated 0.003-inch. nickel-alloy foil to approximate a uniform-heat-flux heater, they found that faint convection cells were indeed visible.

Above. Jennifer Drapp, B. S. in mechanical engineering from the University of Houston and currently pursuing a master's degree, is engaged in boiling experiments in the Heat Transfer and Phase Change Laboratory to determine turboidal-eddy flow patterns.

The second step required developing a reliable procedure for calibrating liquid crystal paint for its temperature vs. color behavior. In the calibration, liquid crystal paint was applied to a uniform-temperature surface. As the temperature of the surface was gradually increased, digitized NTSC-RGB images of the surface were acquired. The "color" of the surface was quantified using a scalar, hue, a descriptor of the spectral colors present in the image.

Nikhil Dukle examines responses in the boiling tank of a flow-loop for forced convection boiling experiments. He employs this equipment in research toward his masters degree on liquid crystal imaging of boiling incipience of a submerged jet. Dukle combines his research with his responsibilities as a teaching assistant in the Department of Mechanical Engineering. Dukle earned his baccalaureate degree in mechanical engineering at the College of Engineering in Poona, Maharashtra, in western India.

Jim Huang, who earned both his baccalaureate degree and master's degree from Fudan University in Shanghai, China, is shown in the magnetic shield chamber experimenting with superconductive materials.

Effective Thermal Conductivity of Fibrous Materials

Larry C. Witte, Ph.D., Department of Mechanical Engineering, and Michael W. Futschik, McDonnell-Douglas Corp., Houston, Texas

Insulating materials are used abundantly for thermal protection of space hardware. Such materials can be subjected to harsh conditions that vary from the hot vacuum conditions during a lunar day to the frigid vacuum conditions of the lunar night; from the low-pressure CO2 environment of Mars to the enormously high temperatures and pressures encountered on the surface of a spacecraft during re-entry. Many insulating materials, either used now or envisaged for use in future space applications, are fibrous in nature; that is, they have fibers existing in a gas matrix. Nomex is such an insulator, and it has been chosen for scrutiny in this study because of its potential for space-related applications.

Through this study, investigators sought to gain a better understanding of the various mechanisms of heat transfer through fibrous materials, especially insight into how different fill gasses of different pressures affect the effective thermal conductivity of a fibrous insulator. To accomplish this study, the effective conductivity of Nomex was measured with air, carbon dioxide, and nitrogen fill gases at pressures ranging from atmospheric down to -10-4 Pa and for mean insulation temperatures of 20&degree;C, 0&degree;C, -30&degree;C , and -50&degree;C. By using first principles and some empiricism, two mathematical models were constructed to correlate experimental data. The models were evaluated to determine their effectiveness in predicting the effective conductivity of the fibrous material Nomex.

The models predicted the effective conductivity of Nomex extremely well. The influence of gas conduction proved to be \the most influential component in predicting effective conductivity, investigators discovered.

Investigators found that the gas conduction representation had to be altered by a correction factor K for CO2, regardless the model used to correlate experimental data. The exact nature of the correction factor could not be determined, but speculation centers on its being a function of both the structure of the gas molecules and structural parameters of the insulating material.