Publications

Here we illustrate the free search for the optimal geometry of flow channel cross-sections that meet two objectives simultaneously: reduced resistances to heat transfer and fluid flow. The element cross section and the wall material are fixed, while the shape of the fluid flow opening, or the wetted perimeter is free to vary. Two element cross sections are considered, square and equilateral triangular. We find that the two objectives are best met when the solid wall thickness is uniform, i.e., when the wetted perimeters are square and triangular, respectively. We also consider arrays of square elements and triangular elements, on the basis of equal mass flow rate per unit of array cross sectional area. The conclusion is that the array of triangular elements meets the two objectives better than the array of square elements.

: This report describes the friction and heat transfer characteristics of a solid surface covered with flexible fibers. The work consisted of analysis, numerical simulations, and laboratory experiments. Chapter 1 documents the pressure drop and heat transfer through bundles of parallel cylinders in a domain that has been overlooked, namely, low Reynolds numbers, arrays that are long in the direction of flow such that the flow is hydraulically and thermally fully developed, and cylinders inclined relative to the flow direction. The heat transfer, friction and mechanical (elastic) interaction between an external laminar flow and a solid surface covered by a layer of fibers is documented in Chapter 2. The flow is initially perpendicular to the surface, and the fibers can bend. The effect of fiber bending,

In this paper we use constructal theory to determine the configuration of two rows of pin–fins so that the total heat transfer rate is maximized. The heat transfer across the fins is by laminar forced convection bathed by a free-stream that is uniform and isothermal. The optimization is subjected to fixed total volume of fin materials. The dimensions of the optimized configuration are the result of balancing conduction along the fins with convection transversal to the fin. The resulting flow structure has multiple scales that are distributed non-uniformly through the flow structure. Numerical results on the effect of Reynolds number and the thermal conductivity ratio on the optimal configuration are reported. The results predicted based on scale analysis are in good agreement with the numerical results. The results also show that the flow structure performs best when the fin diameters and heights are non-uniform.

This paper summarizes an analytical and numerical study of buoyancy-driven convection in a rectangular cavity filled with fluid. The new feature of this study, suggested by architectural applications of the convection phenomenon, is the presence of constant heat flux heating and cooling along the vertical side walls. It is shown analytically that in the boundary layer regime the boundary layer thickness must be constant (independent of altitude), that the core must be motionless and linearly stratified, and that the vertical walls temperature must vary linearly with the same gradient as the core temperature. The Nusselt number (q″/ΔT)H/k was found to be equal to 0.34 (H/L)1/9 Ra2/9, where H/L is the height/length ratio and Ra = gβq″ H4/ (kαν). The second part of the paper presents a numerical study of the same phenomenon: the numerical results agree very well with the analytical predictions made in the first part of the paper.

No abstract is provided for this article.

This paper reports an analytical solution and a numerical study of natural convection in a rectangular porous layer heated and cooled with uniform heat flux along the vertical side walls. It is shown analytically that in theboundary layer regime the vertical boundary layer thickness isconstant (independent of altitude) and the core region is motionless. The vertical temperature gradient is the same constant everywhere in the porous layer. Numerical results are reported in the range 100 < Ra < 5000, 1 ≤ H/L ≤ 10, where Ra = KgßH2q″/(xvk). The boundary layer analytical solution is shown to agree well with the numerical results.

In this paper we develop flow architectures for “vascularizing” smart materials that have self-healing capabilities. The flow architectures are configured as two trees matched canopy to canopy. A single stream flows through both trees and bathes every subvolume (crack site) of the material. Several types of tree-tree configurations are optimized. Trees that have only one level of branching and bathe a rectangular domain have optimal external shapes that are nearly square. They also have optimal ratios of channel sizes before and after branching. Trees optimized on square domains perform nearly as well as trees on freely morphing rectangular domains. The minimized global flow resistance decreases slowly as the number of subvolumes increases. It is more beneficial to bathe the entire volume with a single (optimized) one-stream architecture than to bathe it with several streams that serve small clusters of volume elements. These conclusions are reinforced by an analytical optimization of the same class of architectures in the limit of a large number of assembled subvolumes. We also show that the freedom to morph the design and to increase its performance can be enhanced by using tree-tree architectures with more than one level of branching.

In this paper we describe a natural language processing system which is able to predict whether or not a patient exhibits a specific phenotype using the information extracted from the narrative reports associated with the patient. Furthermore, the phenotypic annotations from our report dataset were performed at the report level which allows us to perform the prediction of the clinical phenotype at any point in time during the patient hospitalization period. Our experiments indicate that an important factor in achieving better results for this problem is to determine how much information to extract from the patient reports in the time interval between the patient admission time and the current prediction time.

In this article we rely on constructal theory to show that the hierarchy of universities is rigid, and that the explanation lies in the nature of education (science, news, information) as a natural fl ow system that bathes the globe most effectively. The article begins with two observations: (i) the rankings of the best engineering universities in the USA closely mirror the rankings of the universities that have the most names of researchers on the list of the most highly cited authors; and (ii) the log‐log plot of the number of highly cited authors of one school versus the rank of that school is nearly a straight line with slope between ‐1/2 and ‐1. The straight line is the same as the distribution of city sizes versus city rank throughout the history of Europe. From this follows the argument that the hierarchy of universities is tied to geography, to how each nodule of knowledge generation serves the area allocated to it. Education fl ows from point to area. The compounding of areas to cover the landscape is the origin of the hierarchical and stable arrangement of universities. The rank of a university is closely related to the visibility of its producers of ideas. The tapestry of a university on the landscape is predicted. All universities grow and improve in time (like all the river channels during the rain), but their hierarchy remains the same.

This is a study of a fundamental problem in electronic cooling; namely, conduction in a two-dimensional (2-D) domain cooled by one or more streams that make one or more passes. The fundamental objective is to determine the relation between the chosen cooling patterns and the temperature distribution in the domain, with particular emphasis on the maximum and minimum temperatures. The practical objective is to develop a consistent method of evaluating the cooling performance of various flow patterns, so that the trends of performance improvement are visible, and the selection of the best cooling pattern can be made with minimum additional computation. The method begins with a general analytical treatment based on the use of finite cosine Fourier transforms, and ends with an efficient numerical implementation of the analytical formulation. The results show that cooling patterns with more cold inlets maintain lower hot-spot temperatures. Furthermore, patterns where adjacent flow passes are oriented in counterflow guarantee more uniform temperature distributions than patterns with adjacent passes in parallel flow. The distribution of temperature is also illustrated experimentally for 15 cases using three different flow patterns.

This paper is a proposal to embed tree-shaped vasculatures in a wall designed such that the wall withstands without excessive hot spots and peak stresses the intense heating and pressure that impinge on it. The vasculature is a quilt of square-shaped panels, each panel having a tree vasculature that connects the center with the perimeter. The vascular designs for volumetric cooling can be complemented by the shaping and distributing of channels for maximum strength and thermal performance at the same time. Numerical simulations of heat flow and thermal stresses in three directions show that it is possible to determine the optimal geometric features of configurations with radial channels and trees with radial and one level of bifurcations. The global performance is evaluated in terms of the overall thermal resistance and peak von Mises stresses. The dendritic design is superior under the studied thermal condition.

Here we draw attention to the development of smart materials with embedded vasculatures that provide multiple functionality: volumetric cooling, self-healing, mechanical strength, etc. Vascularization is achieved by using tree-shaped (dendritic) and grid-shaped flow architectures. As length scales become smaller, dendritic vascularization provides dramatically superior volumetric bathing and transport properties than the use of bundles of parallel microchannels. Embedded grids of channels provide substantially better volumetric bathing when the channels have multiple diameters that are selected optimally and put in the right places. Two novel dendritic architectures are proposed: trees matched canopy to canopy, and trees that alternate with upside down trees. Both have optimized length scales and layouts. Flow architectures are derived from principle, in accordance with constructal theory, not by mimicking nature.

Here we report the discovery that even the simplest, oldest and most prevalent forms of evolutionary movement--rolling bodies and whirls of turbulence--exhibit the same body-size effect on life time and life travel as the evolutionary movement united by the body-size effect so far: animals, rivers, vehicles, jets and plumes. In short, the bigger should last longer and travel farther. For rolling bodies, the life span (t) and the life travel (L) should increase with the body mass (M) raised to the powers 1/6 and 1/3, respectively. The number of rolls during this movement is constant, independent of body size. For an eddy of turbulence, t should increase with the eddy mass (M) raised to the power 2/3, while L should increase with M(2/3) times the bulk speed of the turbulent stream that carries the eddy. The number of rolls during the eddy life span is a constant independent of eddy size.

No abstract is provided for this article.

It has been proposed to extract energy from the subterranean hot dry rock bed (HDR) by creating one or more narrow fractures in the rock and circulating cold water through the fractures. In time, the temperature of the rock region surrounding the crack drops under the influence of time-dependent conduction. This study presents the most basic thermodynamic aspects (first law and second law) of the HDR energy extraction process. It shows which parameters most influence the amount of useful energy (exergy) extracted from the HDR reservoir over a fixed time interval. For example, the water flow rate can be selected optimally in order to maximize the delivery of energy over the lifetime of the HDR system.

The constructal law accounts for the universal phenomenon of generation and evolution of design (configuration, shape, structure, pattern, rhythm). This phenomenon is observed across the board, in animate, inanimate and human systems. The constructal law states the time direction of the evolutionary design phenomenon. It defines the concept of design evolution in physics. Along with the first and second law, the constructal law elevates thermodynamics to a science of systems with configuration. In this article we review the more recent work of our group, with emphasis on the advances made with the constructal law in the natural sciences. Highlighted are the oneness of animate and inanimate designs, the origin of finite-size organs on animals and vehicles, the flow of stresses as the generator of design in solid structures (skeletons, vegetation), the universality and rigidity of hierarchy in all flow systems, and the global design of human flows. Noteworthy is the tapestry of distributed energy systems, which balances nodes of production with networks of distribution on the landscape, and serves as key to energy sustainability and empowerment. At the global level, the constructal law accounts for the geography and design of human movement, wealth and communications.

Technical Briefs Predicting the Pool Fire Vortex Shedding Frequency A. Bejan A. Bejan Department of Mechanical Engineering and Materials Science, Duke University, Durham, NC 27706 Search for other works by this author on: This Site PubMed Google Scholar Author and Article Information A. Bejan Department of Mechanical Engineering and Materials Science, Duke University, Durham, NC 27706 J. Heat Transfer. Feb 1991, 113(1): 261-263 (3 pages) https://doi.org/10.1115/1.2910540 Published Online: February 1, 1991 Article history Received: January 5, 1990 Revised: June 6, 1990 Online: May 23, 2008