Search

The permeability of mesh wicks is important for various applications, for example two-phase heat transfer. However, our understanding of the permeability of single-layer, free-standing mesh wicks (SLFSMW), with liquid-gas interfaces on both sides, is limited. We present a novel and simpler method to determine the permeability of a SLFSMW and apply it to a representative mesh. Our method involves modifying the capillary pressures elevation and simultaneously measuring the permeability to determine the permeability-capillary pressure relationship. When applied to a copper mesh with plain weave having undergone careful surface cleaning, the permeability is found to decrease as capillary pressure for deionized (DI) water increases. We present dimensional analysis to generalize our data for other mesh sizes with similar weaves, and fluids. We model the behavior of mesh in application based on the integration of Darcy’s law with an analytic function fit to measured data and conduct parametric studies to investigate the superficial velocity of liquids through the mesh under varying driving pressures, transport lengths, and liquid viscosity based on the obtained capillary pressure-permeability relationship. This study provides valuable insights into the transport properties of mesh wicks, with potential applications in fields such as electronics cooling, electrochemical devices, and fluid purification technologies.

This paper presents a magnetic field-based method for vehicle misalignment estimation in Dynamic Wireless Power Transfer (DWPT) systems. Road-embedded Transmitter (Tx) pads are employed to both charge a moving Electric Vehicle (EV) and participate in the estimation of the EV lateral and vertical misalignments. The EV is equipped with a power Receiver (Rx) pad and four sensing coils. The proposed system exploits the existing road-embedded pads to generate a magnetic field for misalignment estimation. The sensing coils measure the magnetic field produced by Tx pads. These measurements are uniquely affected by EV lateral and vertical misalignments. Misalignments are estimated by analyzing the voltage induced in sensing coils. Six Artificial Neural Networks (ANNs) are employed to calculate misalignment values in real-time and report them to the EV control system or the vehicle driver. To demonstrate the system practicality, a 500-W DWPT prototype consisting of two Tx pads is developed and tested. Misalignment estimation is tested in the range -15, 15 cm laterally and 16.5, 22.5 cm vertically, and the estimation error is identified to be less than 1.9% and 3.8% of the full misalignment ranges, respectively.

Dynamic Wireless Power Transfer (DWPT) allows Electric Vehicles (EVs) to be charged while in motion. However, high cost and efficiency concerns limit the widespread adoption of DWPT. A Ground Assembly (GA) accounts for most of the system cost since it is implemented on a significant portion of the road. This paper proposes a cost-efficiency optimization algorithm to determine the optimum design of a DWPT Transmitter (Tx) pad. Elongated rectangular pads are considered a compromise between cost and efficiency. An optimization methodology is put forward to maximize Tx pad efficiency while minimizing GA cost over a selected road. The optimum design allows constant power transfer inside a selected rated power zone while considering EV lateral misalignment as a random variable. Main cost factors are accounted for, including the cost of the pad coil Litz wire and ferrite material and transmitter power electronics and compensation network. Particle swarm optimization allowed the number of finite element analysis simulations to be reduced by intelligently selecting test designs. Statistical analysis is applied to understand the impact of different variables on the final design and the interdependence between variables. Additional analyses are conducted to evaluate the impact of different DWPT aspects, including the wire gauge, probability density of the misalignment variable, multilayer coil design, and capacitor cost modeling. The algorithm is used to design a 3.7 kVA Tx pad with respect to the SAE J2954 receiver test stand VA WPT1/Z1. The optimization Pareto fronts illustrate the family of optimum designs, while the chosen 3.7-kVA pad offers the statistical expected value of pad efficiency as 96%, GA per meter cost of 1,004 USD, and an optimum pad length of 1.75 m. The 3.7 kVA optimized pad is manufactured and tested for several operating conditions verifying the simulation results.

This article is focused on the thermal design and three-dimensional (3-D) package optimization of planar magnetic components (PMCs), including transformers and inductors for application in an electric vehicle composite boost dc–dc converter. Each PMC comprises electrical windings in printed circuit board (PCB) form in combination with a ferrite core. Multiple features of each PMC package are thermally optimized for the proposed device configurations with given core size, core loss distribution, number of turns in the PCB winding, winding copper thickness, and winding loss distribution. These heuristically optimized features include a lower level cold plate structure with a conformal base for enhanced convective heat transfer, an upper level PMC cap structure for doubled-sided cooling through conductive heat flow to the cold plate, the implementation of functionally distributed copper thermal and electrothermal vias in the PCB winding for improved cross-plane thermal conductance, and judicious implementation of select materials at various locations and interfaces within the package. Detailed numerical modeling reveals the combined effect of this 3-D packaging strategy with a 79.3 °C and 48.5 °C maximum temperature reduction in the core and PCB winding, respectively, relative to a baseline device configuration. Select PMC experimental validation confirms the expected thermal performance of an optimized PCB design. The thermal design approach is relevant for a range of high-power-density electronics PMC packaging applications.

Coherent programmable integrated photonics circuits have shown great potential as specialized hardware accelerators for deep learning tasks, which usually involve the use of linear matrix multiplication and nonlinear activation components. We design, simulate and train an optical neural network fully based on microring resonators, which shows advantages in terms of device footprint and energy efficiency. We use tunable coupled double ring structures as the interferometer components for the linear multiplication layers and modulated microring resonators as the reconfigurable nonlinear activation components. We then develop optimization algorithms to train the direct tuning parameters such as applied voltages based on the transfer matrix method and using automatic differentiation for all optical components.

Thermal regulators and switches are nonlinear devices that can greatly aid in the management of transient and/or spatially varying heating events. Various physical mechanisms have been harnessed to achieve nonlinear and switchable thermal behavior, though device characteristics are often evaluated in terms of steady-state performance only. Accurately capturing the transient behavior of these devices is a crucial missing link required for assessing the effectiveness of implementation in real-world scenarios. Here, we explore the physics of binary vapor diffusion through a noncondensable gas cavity as a promising mechanism for high-resistance contrast thermal regulation. Through a parametric steady-state optimization, we present a roadmap for future designs that can potentially reach switching ratios of up to 14.5. In addition, we perform a transient investigation to characterize the thermal response time of the device. The device possesses an extremely desirable attribute under certain transient heat loads, where the effective switching ratio appears significantly greater than the steady-state ratio due to a temporary increase in the OFF-state resistance immediately after switching. This work informs future experimental efforts for design optimization of binary-diffusion-based regulators and establishes simple modeling schemes to approximate device performance in system-level regulation scenarios.

Microchannel reactors are critical in biological plus energy-related applications and require meticulous design of hundreds-to-thousands of fluid flow channels. Such systems commonly comprise intricate space-filling microstructures to control the fluid flow distribution for the reaction process. Traditional flow channel design schemes are intuition-based or utilize analytical rule-based optimization strategies that are oversimplified for large-scale domains of arbitrary geometry. Here, a gradient-based optimization method is proposed, where effective porous media and fluid velocity vector design information is exploited and linked to explicit microchannel parameterizations. Reaction-diffusion equations are then utilized to generate space-filling Turing pattern microchannel flow structures from the porous media field. With this computationally efficient and broadly applicable technique, precise control of fluid flow distribution is demonstrated across large numbers (on the order of hundreds) of microchannels.

This article proposes an unsupervised learning approach for fault prognosis of silicon carbide (SiC) mosfet s. The proposed approach utilizes the changing trend of a device's voltage, current, temperature, and other device characteristics with its degradation. The failure modes of semiconductors are reviewed along with existing methods for fault prognosis. The proposed approach is the first to address prognostics of SiC devices, and it can avoid the effects from system noise and data errors. It is not limited to offline analysis and is targeted at online implementation. It is easy to implement on standard digital platforms, and has fast computational speed. Offline data analysis is performed to verify the effectiveness of the proposed method, and a processor-in-the-loop system is used to verify its ability to perform online fault prognosis.

In recent years, low-temperature polymer electrolyte membrane fuel cells (PEMFCs) have played an increasingly important role in zero emission strategy to halt climate change. The performance improvement is a particular focus of fuel cell research and development, with water management being one of the major areas of interest. Here we report the development of a novel metal foam gas diffusion layer (GDL) material with embedded gas flow channels. Owing to the porous nature of the gas channel walls, the membrane electrode assembly (MEA) performance is significantly improved as a result of enhanced mass transport. Unprecedented high power density of 2.09 W/cm2 was observed at 90% relative humidity, and 2.04 W/cm2 for 149% relative humidity, which are respectively 12% and 37% improvement over current state-of-the-art low temperature PEMFC performance. Furthermore, we demonstrated that with our metal foam GDL, it is possible to simplify the flow field design on the bipolar plate, leading to manufacturing cost reduction of the fuel cell components. The work provides new pathway to achieve both high performance and low cost for fuel cells, water electrolyzers and other related technologies.

This paper presents some practical formulations for heat conduction topology optimization problems. In post-optimization analysis, temperature metrics are often used to compare the performance of optimized structures, yet are not used generally as optimization objectives. In this article, SIMP-based topology optimization is used to explore several objective functions related to electronics applications to demonstrate clearly the impact of improper objective selection. Performance variations over 100% were observed when comparing key metrics between optimized structures. Findings here are extended to problems in electronics domains, where temperature optimization may be used in unconventional ways to capture more realistic design considerations. This includes an investigation in the combinatorial use of objectives and constraints to satisfy electronics requirements. Four case studies are presented where topology optimization methods are used to maximize system performance metrics while satisfying temperature constraints.

A two-layer sintered porous evaporator wick for use in vapor chambers is shown to offer very high performance in passive high-heat-flux dissipation over large areas at a low thermal resistance. The two-layer wick has an upper cap layer dedicated to capillary liquid feeding of a thin base layer below that supports boiling. An array of vertical posts bridges these two layers for liquid feeding, while vents in the cap layer provide an unimpeded pathway for vapor removal from the base wick. The two-layer wick is fabricated using a combination of sintering and laser machining processes. The thermal resistance of the wicks during boiling is characterized in a saturated environment that replicates the capillary-fed working conditions of a vapor chamber evaporator. Thermal characterization tests are first performed using conventional single-layer evaporator wicks to analyze the effect of sintered particle size on capillary-fed boiling of water. Of the particle size ranges tested, wicks sintered from 180 to 212 μm-diameter particles provided the best combination of high dryout heat flux and a low boiling resistance. A two-layer evaporator wick comprising particles of this optimal size and a 15 × 15 array of liquid feeding posts yielded a maximum heat flux dissipation of 485 W/cm2 over a 1 cm2 heat input area while also maintaining a low thermal resistance of only ∼0.052 K/W. The thermal performance of the two-layer wick is compared against various hybrid and biporous evaporator wicks previously investigated in the literature. While previous wick designs are typically restricted to small areas and low power levels or high surface superheats when dissipating such heat fluxes, the unique area-scalability of the two-layer wick design allows it to achieve an unprecedented combination of high total power and low-thermal-resistance heat dissipation over larger areas than were previously possible.

Single-phase thermal-fluidic performance of an embedded silicon microchannel cold-plate (25 parallel channels: 75 μm × 150 μm) with a 3-D liquid distribution manifold (6 inlets: 700 μm × 150 μm) and vapor extraction conduits, is investigated using water as working fluid. A 3D manifold is fabricated from silicon and bonded to a silicon microchannel substrate to form a monolithic microcooler (μ-cooler). A metal serpentine bridge (52 mm2 of footprint) and multiple resistance temperature detectors (RTDs) are used for electrical Joule-heating and thermometry, respectively. The experimental results for maximum and average temperatures of the chip, pressure drop, thermal resistance (as low as 0.68 K/W), average heat transfer coefficient (∼30,000–50,000 W/m2 K) for flow rates of 0.03, 0.06 and 0.1 l/min and heat fluxes of 60, 100 and 250 W/cm2 are reported. The embedded microchannel-3D manifold μ-cooler device is capable of removing 250 W/cm2 at a maximum temperature of 90 °C with less than 3 kPa pressure drop for a flow rate of 0.1 l/min. The results from conjugate thermal-fluidic numerical simulations agree well with the experimental data over the wide range of heat fluxes and flow conditions. The numerical simulation results also hint at the possibility of removing up to ∼850 W/cm2 using single-phase water at a maximum temperature of 166 °C at the same pressure drop and flow rate. This offers a very attractive strategy/option for cooling of high heat flux power electronics using single-phase water.

A MagCap DC–DC converter using GaN semiconductor devices is presented to improve the efficiency and modularity of the vehicular power system. While the topology is as simple as the flyback converter, all the semiconductor switches turn on and off with soft switching. A design guideline to transformer and capacitor is suggested considering the conduction loss and voltage stress of the switches. A quasi-resonant (QR) operation for low output-power condition is also explained. The switches turn on at the valley of the drain-source waveform to secure soft switching and maximize the efficiency. A 180-W prototype hardware was populated on 63 mm × 31 mm printed circuit board using 200-V GaN FET devices. Peak efficiency was 97.1% at 155-W output power and 1.2-MHz switching frequency. Another prototype was built for the QR operation and demonstrated 91.6% peak efficiency by maintaining the soft switching and avoiding excessive switching frequency.

A periodic composite microstructure consisting of permanent magnet and ferromagnetic material can be utilized to realize a controllable permanent magnet. For the microstructure design of the permanent magnet composite, its effective material properties should be estimated accurately and efficiently. For this, an asymptotic homogenization method is developed in this work. Specifically, the formulations for homogenized magnetic permeability and residual flux density are mathematically derived for both scalar and vector potential magnetostatic analysis approaches. Using the asymptotic expansion, cell problem equations are first derived, which will be solved by finite element method in microscopic coordinate. Then, the integral form formulations are derived for the calculation of homogenized effective properties. To show the design possibility of the composite, four numerical examples are investigated in this work. In each example, the accuracy of the obtained properties is numerically validated by comparing the magnetic field and energy produced by a homogeneous composite with the obtained effective properties and those produced by a heterogeneous model with actual periodic composite structures. In addition, the computation time of the homogeneous and heterogeneous models are compared to confirm the computational benefit of the developed homogenization method.

Multiphysics simulation and optimization has emerged as a powerful approach in the design of high power density electronics systems. Such systems involve a broad range of physics including thermal, fluidic, electrical, electromagnetic, and mechanical phenomena at multiple physical length scales. In this article, a brief review is provided on the use of gradient-based topology optimization for simulation-based design of electronics components. As shown herein, heat transfer is of primary interest for modern high-density electronics, and specific examples are provided in different thermal, thermal-mechanical, and thermal-fluid multiphysics contexts. Performance benefits are quantified both numerically and, in select cases, experimentally. Additional research directions are further proposed.

Human Machine Interface (HMI) in the vehicle environment is becoming more complex due to an increased number of desired user functions. These added functions increase the driver’s distraction since the driver must visually identify a particular physical interface that they want to control. A gesture recognition system is one possible method for controlling HMI functionalities without visual cues. Some technical approaches to gesture recognition that exploit deep learning have been proposed. However, these methods require increased computational cost in relation to processing time. Ideally, a gesture recognition system should enable fast and accurate driver action to avoid a distracted driver and maximize safety. In this study, we propose an algorithm that combines geometric calculation and machine learning techniques to estimate gestures with low computational overhead. The algorithm is explained in detail, and results are reported including a gesture recognition accuracy of 85% across three gestures for a moving hand from eight people.

High performance and economically viable cooling solutions must be developed to reduce weight and volume, allowing for a wide-spread utilization of hybrid electric vehicles. The traditional embedded microchannel cooling heat sinks suffer from high pressure drop due to small channel dimensions and long flow paths in two-dimensional (2D) plane. Utilizing direct “embedded cooling” strategy in combination with top access three-dimensional (3D) manifold strategy reduces the pressure drop by nearly an order of magnitude. In addition, it provides more temperature uniformity across large area chips and it is less prone to flow instability in two-phase boiling heat transfer. This study presents the experimental results for single-phase thermofluidic performance of an embedded silicon microchannel cold plate (CP) bonded to a 3D manifold for heat fluxes up to 300 W/cm2 using single-phase R-245fa. The heat exchanger consists of a 5 × 5 mm2 heated area with 25 parallel 75 × 150 μm2 microchannels, where the fluid is distributed by a 3D-manifold with four microconduits of 700 × 250 μm2. Heat is applied to the silicon heat sink using electrical Joule-heating in a metal serpentine bridge and the heated surface temperature is monitored in real-time by infrared (IR) camera and electrical resistance thermometry. The maximum and average temperatures of the chip, pressure drop, thermal resistance, and average heat transfer coefficient (HTC) are reported for flow rates of 0.1, 0.2. 0.3, and 0.37 L/min and heat fluxes from 25 to 300 W/cm2. The proposed embedded microchannels-3D manifold cooler, or EMMC, device is capable of removing 300 W/cm2 at maximum temperature 80 °C with pressure drop of less than 30 kPa, where the flow rate, inlet temperature, and pressures are 0.37 L/min, 25 °C and 350 kPa, respectively. The experimental uncertainties of the test results are estimated, and the uncertainties are the highest for heat fluxes

The cooling performance of a topology optimized and additively-manufactured (AM) multi-pass branching microchannel heat sink is compared with a conventionally machined and diffusion-bonded (DB) aluminum (Al) counterpart. The thermal and fluid performance of each heat sink is evaluated at a single flow rate of 150 ml/min over an input power range of 0 – 200 W. The coolant used for the tests is water. The AM sample is fabricated through a metal powder bed fusion manufacturing process with F357 Al, while the DB sample is conventionally machined out of five layers of 6101-T6 Al. The experimental results show that the cooling performance of the AM multi-pass branching microchannel heat sink is 46% higher than its traditionally machined counterpart. For example, the unit thermal resistance of the DB heat sink at a flow rate of 150 ml/min is 0.471 cm2. °CW-1 compared to 0.323 cm2. °CW-1 for the AM part. On the other hand, the pressure drop of the AM heat sink is approximately 91% higher when compared to the DB heat sink. The increase in both heat transfer and pressure drop is attributed to the local roughness of the microchannel walls arising from the additive process. Relevant differences between the samples and inherent benefits of the additive manufacturing process for such designs is explored.

Silicon carbide (SiC) semiconductors have been identified to have potential to replace silicon devices due to superior electrical and thermal properties for a range of power conversion applications. For electrified vehicle applications, the device configuration leads to high current rates, and this in turn leads to high heat fluxes (~1 kW/cm2) over large bare dies (~1 cm2). SiC devices are capable of operating at higher junction temperatures than Si devices, which warrants revisiting air-cooling solutions that are more simple and reliable than liquid cooling. To enable air cooling for high heat flux dissipation, transformative heat spreading technologies must be developed to increase the heat sink footprint area considering the relatively low heat transfer coefficients available. An advanced vapor chamber technology is being investigated for spreading the high heat fluxes generated by next-generation wide band-gap power devices. For vapor chamber heat spreaders to operate at very high heat fluxes over large areas, the internal wick layer at the evaporator must simultaneously minimize the device temperature rise and the flow resistance to liquid resupply by capillary action during boiling. In this study, a vapor chamber is investigated having an embedded two-layer evaporator wick designed to decouple the functions of liquid resupply (through a cap layer) and capillary-fed boiling heat transfer (within a base layer). The performance of a 50 mm × 50 mm × 5.5 mm vapor chamber with an embedded two-layer evaporator wick is evaluated as the heat spreader under a straight pin fin heat sink cooled via air jet impingement for a 1 cm2 area heat source. The maximum dryout heat flux and thermal resistance are compared with that of a vapor chamber having a traditional monolayer evaporator wick. At a power dissipation of ~500 W, the air-cooled two-layer wick vapor chamber provides a 12% reduction in the thermal resistance compared to the monolayer wick vapor chamber assembly. The results indicate that the design of the evaporator wick of the vapor chamber plays a critical role in determining the overall thermal resistance of the heat sink plus spreader assembly.

The manifold microchannel (MMC) heat sink has been widely studied for liquid-cooling of power-dense electronic components. Conventionally, thermal-fluid performance of an MMC heat sink is analyzed via unit cell simulations and designed by varying the rectangular fin and channel geometries, namely size optimization. To further explore the performance potential of the MMC heat sink, this paper proposes topology optimization (TO) to design the optimal freeform fin/channel geometry to maximize heat transfer performance while minimizing the required pumping power. The heat transfer physics in an MMC heat sink is governed by conjugate heat transfer between an incompressible laminar fluid and a heated conductor. The MMC heat sink fin/channel geometry design is formulated as a material distribution problem in a periodic unit cell. Since TO describes the geometry non-parametrically, it facilitates innovative designs through the exploration of arbitrary shapes. The physics-governed design optimization problem is solved by mathematical programming using design sensitivities and an iterative gradient-based method. The thermal-fluid performance is presented for both conventional size optimization and the proposed TO approach, considering the heat transfer performance versus the required pumping power. It is demonstrated that the TO designed fin/channel geometries outperform those obtained through size optimization. Due to the shape complexity associated with the TO designed fin/channel geometries, they are not readily suitable for conventional manufacturing processes, e.g., machining and metal die-casting. However, such out-of-box designs fully exploit the flexibility offered by the latest advanced manufacturing processes, e.g., additive manufacturing and rapid investment casting.

Thermal metamaterials exhibit thermal properties that do not exist in nature but can be rationally designed to offer unique capabilities of controlling heat transfer. Recent advances have demonstrated successful manipulation of conductive heat transfer and led to novel heat guiding structures such as thermal cloaks, concentrators, etc. These advances imply new opportunities to guide heat transfer in complex systems and new packaging approaches as related to thermal management of electronics. Such aspects are important, as trends of electronics packaging toward higher power, higher density, and 2.5D/3D integration are making thermal management even more challenging. While conventional cooling solutions based on large thermal-conductivity materials as well as heat pipes and heat exchangers may dissipate the heat from a source to a sink in a uniform manner, thermal metamaterials could help dissipate the heat in a deterministic manner and avoid thermal crosstalk and local hot spots. This paper reviews recent advances of thermal metamaterials that are potentially relevant to electronics packaging. While providing an overview of the state-of-the-art and critical 2.5D/3D-integrated packaging challenges, this paper also discusses the implications of thermal metamaterials for the future of electronic packaging thermal management. Thermal metamaterials could provide a solution to nontrivial thermal management challenges. Future research will need to take on the new challenges in implementing the thermal metamaterial designs in high-performance heterogeneous packages to continue to advance the state-of-the-art in electronics packaging.

This paper presents three-dimensional topology optimization for the inverse design of unidirectional fiber reinforced composite (FRC) structures. Specifically, a multiscale topology optimization scheme is proposed for the co-design of the composite macrostructure , spatially-varying fiber size and fiber orientation. A FRC with spatially tailored fiber distribution may perform better than a conventional FRC with a fixed fiber structure. A composite macrostructure is designed using the well-established three-field density approach based on Helmholtz filtering and regularized Heaviside function . For the design of spatially-varying fiber size and orientation, a homogenization-based multiscale approach using an orientation tensor variable is proposed. As a post-processing procedure, optimized fiber microstructures are restored at a macroscopic scale . For this, a projection based de-homogenization scheme is proposed for the restoration of a circular fiber structure. The effectiveness of the proposed design scheme is validated through three design examples for compliance minimization and compliant mechanism problems.

The layout of a low-permittivity material is designed in a power semiconductor module to reduce parasitic common-mode (CM) capacitance and hence attenuate the CM current. Without sacrificing the thermal performance between the semiconductor devices and heat sink, a portion of the bottom copper of a direct-bond-copper substrate is replaced by air to decrease capacitance. The proposed power module is verified in a 1-kW 100-kHz buck converter. The CM current is decreased by 8 dB in the frequency domain without any additional components or filters.

This paper presents a topology optimization method for the sequential design of material layout and fiber orientation in functionally graded fiber-reinforced composite structures. Specifically, the proposed method can find the optimal structural layout of matrix and fiber materials together with optimal discrete fiber orientations. In this method, an orientation design variable in the Cartesian coordinate system is employed with a conventional density design variable. The orientation design variable controls not only the fiber orientation, but also fiber volume fraction. The fiber volume fraction control can be used to relax the orientation design problem and simultaneously design a functionally graded structural layout of fiber material. To avoid intermediate fiber orientations and achieve discrete fiber orientation design, a penalization scheme is applied to the orientation design variable. For solving the optimization problem which involves multiple design variables such as the density variable, fiber orientation variable, and target discrete orientation set, a three-step sequential optimization procedure is proposed. In this procedure, the result for each step provides the isotropic design, continuous fiber orientation design, and functionally graded discrete orientation design, respectively. To validate the effectiveness of the proposed approach, numerical examples for structural compliance minimization and compliant mechanism design are provided.

This paper presents multi-material topology optimization for the co-design of permanent magnet segments and iron material. Specifically, a co-design methodology is proposed to find an optimal border of permanent magnet segments, a pattern of magnetization directions, and an iron shape. A material interpolation scheme is proposed for material property representation among air, permanent magnet, and iron materials. In this scheme, the permanent magnet strength and permeability are controlled by density design variables, and permanent magnet magnetization directions are controlled by angle design variables. In addition, a scheme to penalize intermediate magnetization direction is proposed to achieve segmented permanent magnet arrays with discrete magnetization directions. In this scheme, permanent magnet strength is controlled depending on magnetization direction, and consequently the final permanent magnet design converges into permanent magnet segments having target discrete directions. To validate the effectiveness of the proposed approach, three design examples are provided. The examples include the design of a dipole Halbach cylinder, magnetic system with arbitrarily-shaped cavity, and multi-objective problem resembling a magnetic refrigeration device.

Electrothermal design of a synchronous buck converter with heat flow control is presented. A commercial reference circuit design for a synchronous buck converter is selected for the study. A functional printed circuit board (PCB) of the converter is fabricated and tested to establish baseline performance. The baseline PCB layout is modified by incorporating a design of electrothermal traces from a heat flow control optimization study. Electrical plus thermal characteristics of each design are experimentally evaluated. It is found that the modified design manipulates the heat flow across the PCB. At low output current (

This paper proposes a new bidirectional dc–dc converter capable of improving the convenience and modularity of the dc power grid system. By the unique operation of the $LC$ resonant transformer, the proposed converter achieves bidirectional power control with only two switches. Also, the resonant transformer drives the soft switching turn-on and turn-off over a wide power range. Hence, the circuit has high potential to downsize the converter by exploiting high frequency and simplicity. This paper provides the operating principle and the experimental verification of the proposed circuit. A prototype was fabricated by using a simple and high-density printed circuit board ( $\geq$ 10 W/cc). It is verified that the circuit resonant operation successfully drives bidirectional power control with high efficiency ( $\geq$ 92%). An application example is also provided for dc grid system integration.

Rankine Cycle (RC) is a thermodynamic cycle that converts thermal energy into mechanical work, which is commonly found in thermal power generation plants. Recently, there have been many studies focusing on applying Rankine Cycle to recover low-grade waste heat. On-road vehicles, which convert around one third of the fuel energy into useful mechanical energy for propulsion, are moving energy conversion systems that generate considerable waste heat. It is found from prior research that the Rankine Cycle has great potential in automobile waste heat harvesting applications. However, in contrast with other low-grade waste heat applications, vehicles have limited space for the RC system integration, and the waste heat is relatively unsteady. In this work, the efforts in the past few decades to apply RC to on-road vehicles, specifically passenger cars, are reviewed. Characteristics of the waste heat sources found in vehicles and the constraints put on the automotive RC application are identified. Rankine Cycle architectures, system components, and working fluids suitable to different applications are summarized, which provides a guideline for future RC system design in automobiles. Lastly, a new concept and case study into the future application of Rankine Cycle to vehicle waste heat recovery (WHR) is provided.