A previously reported numerical technique, which uses adjoint method when optimizing 1D space variation of thermal dose in hyperthermia treatment is extended to the case of 2D problems [1]. Using a 2D transient computational model in the Penne’s bio-heat equation, optimization of thermal treatment during hyperthermia is shown to be effective and accurate. Numerical simulations are carried out with the 2D computational method to retrieve the prescribed thermal dose from the solution of an inverse thermal problem. The non-continuity of the temperature variation of thermal dose around the reference temperature is linearized over a small temperature interval. The adjoint method in conjunction with the conjugate gradient algorithm are employed to perform the optimization of the objective function. The results are obtained in terms of the computed thermal dose and the heat source to be applied. The numerical results, obtained for a simple treatment case show a good agreement between the estimated and desired thermal dose. However, the non-uniqueness of the obtained solution (the control function: heat source to be applied) is still under investigation and supports the need for further fundamental research.

Blood perfusion is the local, non-directional blood flow through living tissue. It is measured as the volumetric flow rate of blood per volume of tissue and a large range of perfusion values have been reported for human tissue (i.e. 0.002–0.5 ml/ml/s). This large range is thought to be due to measurement sensitivity, environmental factors, and tissue type and location.

The objective of this work is to quantify the advantages of using double-sided cooling as the thermal management approach for the integrated power electronics modules. To study the potential advantage of the Embedded Power packaging method for the double-sided cooling, experiments were conducted. Three different cases were studied. To eliminate the effect of the heat sink on either side of the module, no heat sink was used in all three cases. The thermal tests were conducted such that the integrated power electronics modules were placed in the middle of flowing air in an insulated wind tunnel. Modules without additional top DBC, with additional top DBC, and with additional top DBC as well as heat spreaders on both sides were tested under the same condition. A common parameter, junction-to-ambient thermal resistance, was used to compare the thermal performance of these three cases. Despite the shortcoming of this parameter in describing the three-dimensional heat flow within the integrated power electronics modules, the concept of the thermal resistance is still worthwhile for evaluating various cooling methods for the module. The results show that increasing the top surface area can help in transferring the heat from the heat source to the ambient through the top side of the module. Consequently, the ability to handle higher power loss can also be increased. In summary, the Embedded Power technology provides an opportunity for implementing double-sided cooling as thermal management approach compared to modules with wire-bonded interconnects for the multichips.

Radio-frequency (RF) ablation is one of the most widely used methods for the treatment of hepatic malignancies. A finite element method (FEM) analysis was employed to determine the thermal dose delivered to the tumor/tissue region. We simulated heating within a RF probe implanted in generic tumor surrounded by healthy tissue using ANSYS. The 3-D model consists of a tumor / tissue region into which the RF probe is embedded inside the tumor. One-quarter symmetry was then invoked. The blood flow was modeled using Penne’s bio-heat transfer equation with differing perfusion rates between the healthy tissue and tumor volume based on literature values. The resulting temperature distribution throughout the region was determined over time. A program was written in Visual Basic to extract the temperature distribution data in the tumor/tissue region and calculate the thermal dose throughout the region. This was done by using a time–temperature Arrhenius relationship for chemical and physical rate process. Tissue necrosis is assumed complete when a thermal dose of one hour has been achieved at 43 °C. In the present study, the geometry of the electrode had a significant effect on the size of the volume of necrosis. It was found that the lower portion of the tumor did not receive the specified thermal dose relative to the upper portion of the tumor in single setting during the RF ablation therapy. This might be due to the Ni-Ti electrode, which protruded only from the top surface of the trocar. The effectiveness of the existing probe can be improved by having one more set of electrodes protruding out from the lower curved surface of the trocar. It was found that the modified probe significantly improved heating in the lower portion of tumor/tissue area, providing more symmetry between the upper and lower portion.

A numerical technique with potential applications in hyperthermia treatment planning is presented. The treatment is simulated using a 2D transient computational model of the Pennes bioheat equation within an optimization algorithm. The algorithm recovers the heating protocol which will lead to a desired damage field. The relationship between temperature, time and thermal damage is expressed as a first order rate process using the Arrhenius equation. The objective function of the control problem is based on this thermal damage model. The adjoint method in conjunction with the conjugate gradient algorithm is used to minimize the objective function. The results from a numerical simulation show good agreement between the optimal damage field and the damage field recovered by the algorithm. A comparison between the recovered damage field and the commonly used thermal dose is also made.

The task of system-level thermal management in electrical systems has never been simple and has only become more difficult in recent years as industry pushes toward devices with higher power densities. A common practice in many power electronic devices is to use an aluminum heat sink along with at least one cooling fan. This paper presents a parametric study that was performed on a DC-DC power converter to determine the thermal response of heat dissipating components when the incoming cool air is diverted. The scenarios that are investigated include diverting the air to flow only over the top of the heat sink and diverting the air to flow only over the top of the heat sink and diverting the air to flow only through the heat sink. The analysis performed in this paper was done so using IDEAS, a three-dimensional finite element analysis (FEA) program that allows the user to simulate the affects of conduction and convection heat transfer in a forced-air cooling environment. To divert the incoming air, a plate was inserted into the model between the heat sink and the two inlet fans such that the fans were partitioned into four sections. The amount of air that each fan section produced was governed, thus diverting the air to flow in pre-specified areas. The study looked at six scenarios, each with its own unique volumetric flow setting. The scenario that involved diverting all of the air to flow through the fins of the heat sink proved to cause the lowest system-wide temperatures. This best-case scenario reduced the maximum system temperature by nearly 50% from the worst-case scenario.

The conjugate gradient method using vectorial descent parameters is used to solve the inverse problem of simultaneously estimating one boundary condition and one thermo-physical property of an heterogeneous material. The advantage of the conjugate gradient method lies in that no a priori information is needed on the variation of the unknown quantities. The heterogeneous material model under consideration consists of a matrix in imperfect contact with embedded separated particles. The first unknown of interest in this inverse problem is the constant coupling parameter Ψ which characterizes the thermal contact (conductance) behavior between the matrix and the particles. The second unknown is the time dependent applied heat flux on one external side of the matrix. As a result, this problem is concerned with a combined parameter and function estimation at the same time. Thus two descent parameters, each one corresponding to each unknown function parameter, are derived and used in the iterative process. The developed inverse analysis is based on the transient temperature measurements taken from some sensors implanted inside the matrix only during the process of heating. Several numerical test cases were performed and show that the developed method provides an accurate estimation of thermo-physical properties and boundary condition in a very short practical time.

This paper describes a numerical procedure conducted to estimate thermo-physical properties of the human tissue during hyperthermia treatment of a cancerous region. The estimation algorithm is based on the solution of an inverse heat conduction problem. The Gauss-Newton method is used to estimate simultaneously the volumetric heat capacity, the thermal conductivity, and the volumetric blood rate (blood perfusion) in the bio-heat transfer equation during a hyperthermia treatment cycle. The treatment quality of hyperthermia is analyzed by the computation of the thermal dose which is obtained from the resulting temperature field in the tissue. The importance of an accurate estimation of the thermo-physical properties of the tissue lies in that they are the most important factors for achieving a high precision heating cycle which results in an optimized treatment. The inverse analysis is based on the temperature measurements taken inside the cancerous tissue region during the transient heating process. An experimental optimization procedure is conducted to make the estimated parameters as accurate as possible. Several numerical tests were performed and show that the developed method provides an accurate estimation of thermo-physical properties in a very short practical time. As the blood perfusion is very sensitive to the temperature variation in the tissue, this estimation tool can be implemented during one cycle treatment which results in an on-line thermo-physical parameter correction while the treatment is performed.

A methodology was developed to optimize the 3D geometrical design layout of an active integrated power electronics module (IPEM) by considering both electrical and thermal performance. This paper is focused on the thermal analysis, which was performed using 3D finite element and computational fluid dynamic (CFD) analyses. A parametric study was conducted to determine the thermal performance of several different design layouts. A sensitivity analysis was performed to determine the overall uncertainty of the predicted simulations. The final design, Gen-II.C, provided a 70% reduction in the common mode current, a 4% reduction in the size of the geometric footprint, and a 3°C reduction in the maximum temperature over Gen-II.A, thus providing an increase in the overall performance.

This work describes steps in the ongoing development of a system to noninvasively estimate blood perfusion using thermal methods. A combination thermocouple/thermopile sensor is used to record heat flux and temperature measurements on the tissue of interest. The resulting data are compared to a biothermal computer model to determine blood perfusion. The system has been tested on a canine medial saphenous fasciocutaneous free tissue flap that allows experimental control of blood flow to the measurement site. These experiments show distinct correlation between blood flow and bioprobe output demonstrating the validity of the measurement method.

A simple and accurate perfused tissue phantom would be a useful tool for biomedical research. The phantom described here is an agar-saline model that can be easily constructed. Space between finely diced pieces of agar is perfused with saline. A rotator Boekel-Orbitron ® is used to perfuse the agarsaline mixture, mimicking the non-directional movement of blood. Dye injected into the phantom records the volumetric movement of saline for a set time period on the Orbitron. These data are used to determine perfusion estimates in units of g/s/mL. Experimental results can be appropriately scaled and additional modifications can be better mimic specific in-vivo conditions.

An in depth parametric study was conducted to optimize the thermal performance of an active system integrated power electronic module (IPEM). Commercial finite element software was utilized in these analyses. Design parameters included material selection, geometry, and placement. Model-based optimization software tool were investigated for use in this procedure and demonstrated promise in automating the design process.

The focus of this paper is on the system-level cooling of integrated circuits and the effect of various cooling scenarios on the temperature response. Specifically, this paper addresses the thermal response of a system of integrated circuits when the heat sink fin arrangement and the fan location are varied in order to optimize the effects of conduction and forced-convection heat transfer to cool the system.

Blood perfusion represents the local, multi-directional blood flow through capillaries and intracellular spaces of living tissue. Blood perfusion is crucial in the assessment of many physiological phemonena, such as skin grafts, burn injuries, and head traumas. Current methodologies are often invasive or highly sensitive to environmental factors [1]. Previous studies have focused on the development of a noninvasive measurement system to accurately measure blood perfusion. The objective of this study is to validate the system performance via experiments using a phantom test stand. Several previous studies hae been conducted on canines; these tests demonstrated that changes and trends in perfusion can be detected using this system [2,3]. The use of a phantom test stand was motivated by the large range of blood perfusion values reported for human tissue (i.e., 0.002–3 mL/mL/s. [4]) This large perfusion range is thought to be due to several factors: measurement device sensitivity, envoronmental factors, and state, type, and location of the tissue of interest. An agar solution phantom test stand was developed to eliminate physiolocigial factors in evaluating the thermal probe’s function and veriability [4]. The overall blood perfusion measurement system consists of a probe with a heat flux sensor and air housing (see Fig. 1), a bioheat transfer model of the probe and tissue, and a parameter estimation program as described below and in [2].

This paper presents thermal design optimization of an insulated gate bipolar transistor (IGBT) integrated power electronic module (IPEM). A commercially available finite element package was used to create a 3D geometric layout of the IGBT module. Thermal simulations were performed under different forced air convection conditions, and for both single and double-sided cooling, to study the effects on the hot-spots and maximum temperature rise of the module. The design optimization for the module was performed by varying parameters (choice of materials and layer thicknesses) and studying their effect on the thermal performance of the module. The results of these studies were several improved designs for the module.