Advanced thermal analysis of microelectronics using spreading resistance models

Thermal analysis of electronic devices is one of the most important steps for designing of modern devices. Precise thermal analysis is essential for designing an effective thermal management system of modern electronic devices such as batteries, LEDs, microelectronics, ICs, circuit boards, semiconductors and heat spreaders. For having a precise thermal analysis, the temperature profile and thermal spreading resistance of the device should be calculated by considering the geometry, property and boundary conditions. Thermal spreading resistance occurs when heat enters through a portion of a surface and flows by conduction. It is the primary source of thermal resistance when heat flows from a tiny heat source to a thin and wide heat spreader. In this thesis, analytical models for modeling the temperature behavior and thermal resistance in some common geometries of microelectronic devices such as heat channels and heat tubes are investigated. Different boundary conditions for the system are considered. Along the source plane, a combination of discretely specified heat flux, specified temperatures and adiabatic condition are studied. Along the walls of the system, adiabatic or convective cooling boundary conditions are assumed. Along the sink plane, convective cooling with constant or variable heat transfer coefficient are considered. Also, the effect of orthotropic properties is discussed. This thesis contains nine chapters. Chapter one is the introduction and shows the concepts of thermal spreading resistance besides the originality and importance of the work. Chapter two reviews the literatures on the thermal spreading resistance in the past fifty years with a focus on the recent advances. In chapters three and four, thermal resistance of a twodimensional flux channel with non-uniform convection coefficient in the heat sink plane is studied. The non-uniform convection is modeled by using two functions than can simulate a wide variety of different heat sink configurations. In chapter five, a non-symmetrical flux channel with different heat transfer coefficient along the right and left edges and sink plane is analytically modeled. Due to the edge cooling and non-symmetry, the eigenvalues of the system are defined using the heat transfer coefficient on both edges and for satisfying the orthogonality condition, a normalized function is calculated. In chapter six, thermal behavior of two-dimensional rectangular flux channel with arbitrary boundary conditions on the source plane is presented. The boundary condition along the source plane can be a combination of the first kind boundary condition (Dirichlet or prescribed temperature) and the second kind boundary condition (Neumann or prescribed heat flux). The proposed solution can be used for modeling the flux channels with numerous different source plane boundary conditions without any limitations in the number and position of heat sources. In chapter seven, temperature profile of a circular flux tube with discretely specified boundary conditions along the source plane is presented. Also, the effect of orthotropic properties are discussed. In chapter 8, a three-dimensional rectangular flux channel with a non-uniform heat convection along the heat sink plane is analytically modeled. In chapter nine, a summary of the achievements is presented and some systems are proposed for the future studies. It is worth mentioning that all the models and case studies in the thesis are compared with the Finite Element Method (FEM).

Design construction and testing of a heat loop for study of fouling and its interrelationship to corrosion in heat exchanger tubes

A heat loop suitable for the study of thermal fouling and its relationship to corrosion processes was designed, constructed and tested. The design adopted was an improvement over those used by such investigators as Hopkins and the Heat Transfer Research Institute in that very low levels of fouling could be detected accurately, the heat transfer surface could be readily removed for examination and the chemistry of the environment could be carefully monitored and controlled. In addition, an indirect method of electrical heating of the heat transfer surface was employed to eliminate magnetic and electric effects which result when direct resistance heating is employed to a test section. The testing of the loop was done using a 316 stainless steel test section and a suspension of ferric oxide and water in an attempt to duplicate the results obtained by Hopkins. Two types of thermal ·fouling resistance versus time curves were obtained . (i) Asymptotic type fouling curve, similar to the fouling behaviour described by Kern and Seaton and other investigators, was the most frequent type of fouling curve obtained. Thermal fouling occurred at a steadily decreasing rate before reaching a final asymptotic value. (ii) If an asymptotically fouled tube was cooled with rapid cir- ·culation for periods up to eight hours at zero heat flux, and heating restarted, fouling recommenced at a high linear rate. The fouling results obtained were observed to be similar and 1n agreement with the fouling behaviour reported previously by Hopkins and it was possible to duplicate quite closely the previous results . This supports the contention of Hopkins that the fouling results obtained were due to a crevice corrosion process and not an artifact of that heat loop which might have caused electrical and magnetic effects influencing the fouling. The effects of Reynolds number and heat flux on the asymptotic fouling resistance have been determined. A single experiment to study the effect of oxygen concentration has been carried out. The ferric oxide concentration for most of the fouling trials was standardized at 2400 ppM and the range of Reynolds number and heat flux for the study was 11000-29500 and 89-121 KW/M², respectively.