With the rapid growth of electric vehicle production, the disposal of retired lithium-ion batteries has emerged as a critical challenge. This has accelerated research interest in repurposing second-life batteries (SLBs) for various applications. This vision paper reviews recent advances, identifies key challenges and outlines future directions for SLB applications in the power sector. Following an overview of current SLB applications, the paper examines unresolved challenges from three key perspectives: i) accurate prediction of SLB state of health (SoH), ii) development of effective thermal management technologies, and iii) policy and regulatory gaps hindering widespread SLB adoption. For each challenge, the current state of knowledge is reviewed, and the remaining limitations are identified. The paper further highlights that the thermal safety of SLBs cannot be inferred solely from SoH but depends on the battery aging pathway and the selected safety metric, with mildly aged cells potentially exhibiting higher thermal runaway susceptibility than fresh or deeply aged cells under certain aging and abuse conditions. Based on this insight, the paper proposes an aging-pathway-aware thermal-safety grading framework integrated with safety-adaptive thermal management, supported by advanced predictive methodologies and regulatory standardization, to enable the safe, reliable and economically viable large-scale deployment of SLBs.

During the operation of new energy vehicles, the pressures of the air springs and the reservoir dynamically change in response to road conditions and vehicle loads. Consequently, the air suspension compressor must operate under variable back-pressures to accommodate pressure variations within the air suspension system. This study focuses on a miniature oil-free piston compressor used in the air suspension system of new energy vehicles, with particular emphasis on its operating states and thermodynamic characteristics under unsteady, variable back-pressure conditions (0.1–1.5 MPa). A combined air suspension compressor-reservoir model was established to analyze the thermodynamic performance and inflation characteristics of the compressor. Furthermore, a test platform was constructed for evaluating the variable back-pressure characteristics of the air suspension compressor, and thermal performance tests were conducted on the prototype. These results showed that under extremely high back pressure, the combined effect of expansion within the clearance volume and external leakage was the primary cause of the reduction in the air suspension compressor inflation. Additionally, at high pressure ratios, expansion and external leakage shifted the peak resistance torque of the compressor motor, causing fluctuations in motor speed and subsequently degrading the air suspension compressor inflation performance. Compared to conventional inflation conditions (0.6 MPa), the volumetric efficiency of the air suspension compressor decreased by 69%, and the isothermal indicated efficiency decreased by 48.1% when the maximum pressure of the air suspension system (1.5 MPa) was reached, leading to a sharp deterioration in air suspension compressor performance. Therefore, better oil-free sealing materials, a more compact clearance volume design, and multi-stage configuration are prospects for adapting to higher operating pressures. This study lays a theoretical foundation for subsequent performance optimization and air suspension compressor design.

Offshore industries depend solely on diesel-based power generation systems or mainland grids, which are expensive and carbon-intensive. The demand for renewable energy-based offshore DC microgrids (MGs) has significantly increased due to rising fuel prices, high costs of fuel transportation and storage, extreme operation and maintenance expenses, and associated carbon emissions. This research study optimises the size of an offshore DC MG that integrates wave, solar, energy storage, and diesel, utilising real-world data from a specific geographical location (latitude −33.525587 and longitude 114.772211), thereby accurately representing the availability of renewable energy sources. An algorithm is designed to optimise the utilisation of highly variable renewable sources via battery-based energy management, resulting in optimal energy dispatch. Utilising economic performance metrics, such as levelised cost of energy (LCoE) and net present value (NPV), this research aims to minimise the energy, operating, and greenhouse gas emission costs while maximising the economic feasibility of the system. A sensitivity analysis is performed to determine the impact of fuel prices, discount rates, and system lifespans on the feasibility of the system. The findings demonstrate that the proposed renewable-based offshore DC MG can substantially reduce fuel consumption (93%), operational expenses (77.56%), and carbon emissions (89.50%) compared with a diesel-only system for offshore platforms, while improving the sustainability and reliability of power supply for aquaculture and marine activities. In addition, the proposed renewable-energy-based offshore DC MG achieves a lower LCoE (0.5649 $/kWh) and a higher NPV (2.987 × 104 $) than a conventional diesel-based power generation system for offshore industries. The results provide a decision-making framework for the design and implementation of renewable energy-based offshore DC MGs.

Freezing desalination (FD) has gained growing attention in recent years due to its low scaling and corrosion potential and low theoretical energy demand. However, its development is constrained by limited salt removal efficiency and the substantial consumption of high-grade electrical energy for refrigeration. While LNG-driven Multi-Effect Freezing Desalination (MEFD) is widely recognized as a promising solution to these challenges, existing literature remains predominantly confined to conceptual or qualitative analyses, lacking comprehensive system-level design. To address this gap, this study proposes an engineering-oriented, system-level design for an LNG-MEFD system, grounded in experimental investigation. A series of single-effect FD experiments were conducted to quantify the influence of process variables on salt rejection, recovery rate, and stirring energy consumption. Based on the experimental data, quantitative regression models linking process variables to system performance were established. These models were innovatively integrated into a holistic MEFD system design and optimization framework. Utilizing the NSGA-II algorithm, the optimal specific energy consumption (SEC) and process parameters were predicted across various application scenarios. Results indicate that a three-effect configuration is optimal for potable water production. Under typical feedwater salinity conditions of 35 ppt, the optimized thermal and electrical SEC are 465.46 kWht/m3 and 1.83 kWhe/m3, respectively. These metrics are achieved under optimal operating conditions: an effect temperature of approximately -8.2 degrees C, and stirring speeds of 150, 112, and 248 rpm. Overall, this work provides a valuable theoretical foundation and design basis for the practical engineering of LNG-driven MEFD systems.

The incorporation of coal gangue and fly ash (FA) into cemented backfill composites provides an effective pathway for mining solid-waste valorization and ensuring mining operational safety. However, the intrinsic relationship between pore structure characteristics and the mechanical strength of cemented gangue backfill bodies (CGBB), particularly the regulatory role of aggregate gradation in pore evolution and mechanical performance optimization, remains inadequately explored. In this study, CGBB specimens were prepared using coal gangue as aggregate and FA-cement as the binary binder system. Four distinct aggregate gradation groups were designed with coarse particle contents ranging from 20% to 50%. Nuclear magnetic resonance (NMR) spectroscopy and uniaxial compression tests were integrated to systematically investigate the synergistic effects of curing age (7, 14, 28 days) and aggregate gradation on pore structure (porosity, pore size distribution, connectivity) and mechanical properties of CGBB. The results showed that CGBB pore sizes were predominantly distributed within three ranges (e.g., 0-0.01 mu m, 0.01-0.025 mu m, 0.025-0.1 mu m), accounting for 90% of the total pore volume, with non-damaging pores (<20 nm) predominating. With increasing curing age, total pore volume decreased significantly, and the pore size distribution shifted toward microporosity. A strong negative exponential correlation was identified between uniaxial compressive strength (UCS) and porosity, with harmful pores (>100 nm) acted as the primary strength-limiting factor. Reducing coarse particle content from 50% to 20% enhanced the micro-aggregate effect of medium and fine particles (0-5 mm), improving the initial skeleton compactness. Concurrently, the FA's pozzolanic secondary hydration generated additional C-S-H/C-A-S-H gels, which densified the matrix and filled residual pores. These synergistic effects resulted in a 2.5-fold increase in 28 days UCS. Furthermore, T-1-T-2 relaxation spectra and fractal dimension analysis demonstrated that optimized continuous gradation reduced pore connectivity by 71.6% and increased pore structure fractal dimension, indicating enhanced compactness and structural complexity. The resulting dense internal structure mitigated the interfacial weakening effect between slurry and aggregate, thereby improving the overall structural integrity and mechanical reliability of CGBB.

The severe performance degradation of transcritical CO2 air conditioning in electric vehicles under high ambient temperatures remains a critical barrier to their widespread adoption. Conventional solutions including ejectors, expanders, mechanical subcooling and others offer certain improvements, yet they inevitably increase system complexity by directly intervening in the cycle, introduce additional refrigerants with safety or environmental concerns, or suffer from limited effectiveness under diverse conditions. Integrating a magnetocaloric device based on an active magnetic regenerator offers a new remedy. But the dynamic coupling between the periodic magnetic cycle and the CO2 flow introduces substantial complexity that has not been systematically addressed in previous studies, along with the strong interdependence of operating parameters, such as frequency, flow rate and material mass. Moreover, the material’s magnetocaloric response is highly temperature‐sensitive, which complicates the establishment of an effective subcooling span at extreme heat. To investigate the dynamic coupling between the cyclic magnetic field and CO2 flow, along with the interdependence of frequency, flow rate, and material mass, a simulation model was developed coupling Mean‐Field Theory and a phenomenological material description with a validated 1D transcritical CO2 cycle, capturing both steady‐state performance and transient temperature pulsations from cyclic AMR operation. This integrated framework enables system‐level assessment of magnetocaloric-CO₂ interactions under realistic EV operating conditions. Systematic parametric analyses over a wide range of frequencies (0.1-1.0 Hz), coolant flow rates (0.1-1.0 kg·s-1), and material masses (1.12-3.37 kg) quantify the trade‐offs between cooling enhancement and parasitic power consumption, revealing optimal operational windows for each ambient condition. Results demonstrate that the MRSC integration yields the most pronounced COP improvement of 26.8 % at 40 °C, effectively countering the high‐temperature penalty, whereas at the extreme 45 °C condition the benefit diminishes to only 5.5 % due to the sharp decay of magnetic entropy change at elevated temperatures. Dynamic disturbances are effectively damped by the internal heat exchanger, yet optimal parameter combinations remain highly dependent on ambient conditions. The findings confirms that an optimally designed MRSC provides a viable solution for boosting the efficiency of transcritical CO2 thermal management systems, especially under high-temperature cooling demands. This work highlights inherent difficulties and provides guidance for designing such hybrid thermal management systems.

The increasing demand for power system flexibility, driven by the high penetration of renewable energy sources, has highlighted the importance of fast-response energy storage technologies. Among these, Carnot battery systems have attracted growing attention, in which ultra-high-temperature CO2 heat pumps serve as key power-to-heat conversion units. However, it remains unclear how thermal inertia governs the ramp-rate capability and thermal-output response of ultra-high-temperature CO2 heat pumps during load-following operation. Therefore, a system-level, high-fidelity dynamic model of an 873.15 K-class high-temperature CO2 heat pump is developed for energy storage applications. Two regulation strategies, namely compressor-speed modulation and inventory-tank regulation, were investigated to identify the load-following characteristics of the system within their respective regulation limits of 16.45% and 49.41%. Owing to thermal inertia, the average downward ramp rates reached −5.39 and − 11.19% min−1, respectively, while higher upward ramp rates were observed because of the reduced thermal inertia under low-load conditions. Compressor-speed regulation is primarily limited by the delayed thermal response of the gas cooler and gas heater, while inventory-tank regulation is governed by the coupled effects of CO2 inventory migration and recuperator re-equilibration. Under an AGC-like power command, compressor-speed regulation is better suited for small-amplitude, high-accuracy tracking, with an upward ramp rate of 7.70%·min−1, whereas inventory-tank regulation favors deeper and faster modulation, achieving 29.01%·min−1 during load recovery. At the Carnot-battery level, the proposed heat pump enables estimated round-trip efficiencies of 64.40–64.75% for Greenfield sCO2 deployment and 60.90–61.15% for Brownfield steam-Rankine repurposing, indicating its potential as a flexible charging interface for renewable power absorption.

This article presents a uniform structure of a sustainable energy-based offshore DC microgrid (MG) integrating wave, wind, and solar energy sources. The inherent intermittency of these sustainable energy sources (SESs) and the highly variable offshore loads pose significant challenges in maintaining the dynamic stability of offshore DC MGs. Integrating a battery storage system is identified as a suitable solution to buffer these fluctuations by balancing the power generation and load demand, and enhancing the system's reliability. A common maximum power point tracking (MPPT) algorithm is developed for all the sustainable sources to optimize power extraction and enhance system efficiency. A novel combined control strategy, incorporating average current mode (ACM) and PI controllers tuned via the root locus method, is proposed for the battery converter. This strategy allows to maintain dynamic stability of the system under different operating conditions. A detailed small-signal equivalent model of the proposed offshore DC MG is developed to analyze the system's dynamics under load and generation changes. The stability of the system is analyzed through both root locus and Bode plot methods. The entire offshore DC MG system is modeled and simulated in the MATLAB/Simulink environment, providing a comprehensive analysis of the controller performance across various scenarios. The proposed control method is experimentally validated using a B-Box RCP digital controller for a laboratory prototype of DC-DC converters. Both simulation and experimental results not only demonstrate the feasibility of integrating different sustainable sources into a DC MG in an offshore setting but also underline the effectiveness of the proposed control strategy in enhancing operational efficiency and system reliability.

Scroll compressors are widely utilized in various applications as a type of efficient positive displacement compressor. The improvement of scroll compressor profiles can further improve the performance and efficiency of scroll compressors. In this article, the geometric model of the variable-wall-thickness scroll (VWTS) based on profiles generated by the involute of variable-diameter base circle is derived, with refined calculation methods for average leakage line length and wall thickness of the VWTS. A comprehensive thermodynamic model considering leakage, valve motion, and intermediate discharge port are established for the scroll compressor with the VWTS. Additionally, a dynamic model is developed to calculate the gas forces on the VWTS. The above models are solved using CO2 as the refrigerant. Comparative analysis with the constant-wall-thickness scroll (CWTS) compressor under the same suction volume reveals that the VWTS can be optimized to a certain extent in terms of size, length of leakage line, and gas force. In addition, owing to its variable wall thickness characteristics, the leakage in the initial compression stage increases slightly, leading to a marginal reduction in volumetric efficiency. But the leakage in the later stage is reduced, resulting in a slight improvement in isentropic efficiency. Furthermore, compared with CWTS, VWTS has a 2-3% optimization in indicated power.

The deviation of real pump performance from ideal curves critically reduces energy efficiency at pumping stations. This deviation, caused by differences between laboratory and field conditions and long-term wear, results in unavoidable performance decline. Existing studies using CFD or neural networks trained on lab data often face limitations in physical interpretability and cross-condition applicability, restricting their use in large-scale water transfer projects. To solve this, we propose a hybrid physics-informed neural network that integrates axial-flow pump hydraulic equations and demonstrate its effectiveness for accurately predicting performance along the Eastern Route of the South-to-North Water Diversion Project. Results show that the hybrid PINN produced highly accurate predictions for all four units under normal conditions, with R² values over 0.95 for head and 0.92 for efficiency. The median MAE decreased by about 68% and 67% compared to traditional data models. During off-design load tests, the RMSE and MAE for head and efficiency further dropped by approximately 36%–41%, showing good generalization across different units and conditions. Also, the two-stage incremental training strategy used improved training stability and robustness, increasing effective iterations by 26% and lowering early oscillations by 10.4%. Overall, the proposed hybrid PINN provides an efficient and accurate way to predict axial-flow pump performance and offers valuable guidance for similar pump station assessments.

This paper presents a dynamic analysis of a wave-driven liquid-piston compressed air energy storage (W-CAES) system. A comprehensive dynamic model is developed and validated with experimental data. Buoy motion is simulated via ANSYS-AQWA, while dynamic thermal behaviours are modelled using FORTRAN. Heat transfer between liquid and air during compression/expansion and its effect on system performance are evaluated. The effects of key parameters (e.g. liquid-piston cylinder shape parameter, pre-set pressure, pressure ratio, turbine flow rate, and wave conditions) are investigated. Results showed that heat transfer in the liquid-piston cylinder significantly improved the compression/expansion process, reducing the polytropic index from 1.4 (adiabatic) to similar to 1.1 (nearly isothermal). Compression cylinder shape analysis revealed that small (flat) or large (narrow) height-to-diameter ratios enhanced heat transfer and efficiency compared to moderate shapes. Pre-set pressure had little effect on round-trip efficiency and capture factor, though it raised energy storage density reaching 1.1 MJ/m(3) at 8 bar. The turbine flow rate had minimal impact on efficiency or capture factor. Case studies at five representative wave sites revealed that wave conditions strongly affected the capture factor, peaking at 30% near the East China Sea, and ranging 7-12% at other sites. These findings indicate this W-CAES design performs better in lower-wave-energy regions, but designs should be tailored to specific wave conditions.

A multiscale numerical framework integrating molecular dynamics and finite element methods is established to unravel the causal chain from nanoparticle interfacial structures to agglomeration morphology and macroscopic heat transfer in molten salt-based nanofluids. Microscopic simulations reveal that the dense interfacial layer around SiO2 nanoparticles enhances thermal conductivity in a size-dependent manner, but excessive agglomeration (>4 wt.%) disrupts interface integrity, leading to non-monotonic conductivity variation. Mesoscopically, low-fractal-dimension and chain-like agglomerates form continuous thermal pathways, maximizing effective conductivity. Convective analysis further uncovers a dual effect: moderate agglomeration (e.g., 8-particle chains) achieves a 34% Nusselt number increase over uniform dispersion by balancing thermal enhancement and flow resistance, whereas densely packed clusters (16 particles) elevate viscosity and degrade convection despite higher intrinsic conductivity. This work establishes a cross-scale correlation from interfacial nanostructure to agglomerate topology and ultimately to system-level performance, revealing that optimal heat transfer arises not from uniform dispersion nor dense aggregation, but from a well-tailored chain-like morphology—providing a physical guideline for designing high-performance nanofluids.

With the rapid adoption of electric vehicles, electric vehicle thermal management systems (EVTMS) face pressing challenges in dynamic battery temperature uniformity control and coupling with the passenger compartment. Based on the vehicle thermal demands, a transient model of the EVTMS that uses refrigerant direct cooling for the battery was developed and validated. The dynamic cooling performance and cell-to-cell temperature uniformity of two system architectures with unequal evaporation pressure (UEP) and equal evaporation pressure (EEP) were compared under different control schemes. The suitability of each architecture and control approach was then analyzed. The performance degradation caused by branch mixing under fixed constraints and the corresponding mechanisms are elucidated. Furthermore, the trade-off between battery cooling and cabin thermal management was characterized, and the extent of the mutual influence was quantified. To address branch mixing-induced performance degradation and the trade-off between battery uniformity and cabin comfort, a novel semi-series direct cooling architecture with a temperature following control strategy is proposed. Simulation results indicate that this approach can reduce the battery’s maximum temperature difference by 60%, lower average power consumption by 3%, and maintain cabin comfort. The findings offer new theoretical and methodological guidance for co-optimizing architecture and control in CO2 direct cooling EVTMS for engineering applications.

A multi-energy complementary system integrates multiple uncertain renewable energy sources and storage systems to maximize profit and stability in modern power systems. However, the challenge of dealing with uncertainties in renewable energy generation and fairly distributing profit among different energy entities in a complementary alliance is necessarily addressed. This paper proposes a coordinated operation and profit distribution model for a multi-energy complementary system comprising cascade hydropower plants, wind turbines, photovoltaic plants, and energy storage. The system's operation is optimized using a fuzzy chance-constrained method to account for uncertainties in renewable energy generation to achieve the maximum profit for the joint power output. A modified Shapley value method is introduced for fair profit allocation among the entities, considering stability, reliability, and risk factors. Results show that considering uncertainties leads to a 4.64 % decrease in total exported electricity but improves the utilization of renewable energy by reducing abandonment rates. The modified contribution ratios of the wind turbines, photovoltaic, hydropower plants, and energy storage on the increased profits due to the cooperative alliance are 5.71 %, 39.90 %, 6.36 %, and 48.03 %, respectively. The study demonstrates the effectiveness of the proposed approach in maximizing profit and ensuring system stability in multi-energy complementary systems.

Carbon dioxide (CO2) thermal management systems offer potential advantages for electric vehicles, though their cooling performance requires further optimization. A dual-evaporation-temperature (DET) system, which independently regulates battery and cabin cooling, demonstrates superior energy efficiency compared to conventional single-temperature systems. However, operational constraints arise from uneven energy distribution and coupled thermodynamic effects, complicating system design and control. To well address this, a simulation model is developed and verified using experimental data. Key operational characteristics are examined, with emphasis on compressor displacement design and its influence on performance limitations. Additionally, the coupled effects of coolant temperature and battery heat generation are investigated, alongside system boundary constraints. Furthermore, the DET system is compared with that of the traditional CO2 system, and the energy-saving potential was analyzed. Critical findings indicate that compressor displacement, coolant conditions, and thermal load variations predominantly govern battery cooling capacity by modulating chiller heat transfer dynamics. Excessive cooling demand beyond system capacity induces instability and efficiency degradation. The DET system achieves a 10-21% efficiency improvement over traditional CO2 systems under ambient temperatures of 30-40 degrees C, contingent upon optimal design and control strategies. The results provide a new guideline for the performance enhancement of the automobile CO2 air conditioning.

Self-excited oscillations can occur when a flow passes over a louvered cavity, leading to engineering problems such as vibration and noise. Although studies on single-louver cavities have achieved notable progress, the flow physics in multi-louver cavities commonly encountered in engineering applications remains insufficiently understood. In this work, numerical simulations are performed to investigate self-excited oscillations in the flow over a cavity equipped with two opposing louvers. The results show that each louver exhibits self-excited oscillations similar to those in single-louver configurations, and their interaction induces low-amplitude oscillations within the cavity. A stage-switching phenomenon is identified when the louver position is varied. This stage switching is caused by a change in the vorticity accumulation rate around the louver, which is governed by the inner-side flow direction. When the inner-side flow is directed downstream, vorticity accumulates more rapidly around the louver (i.e., spanwise vortices form faster), resulting in a pronounced increase in the oscillation frequency. In addition, the downstream louver is found to exhibit larger oscillation amplitudes.

The abundant yet intermittent nature of ocean wave energy necessitates efficient storage to ensure a reliable power supply. Wave-driven compressed air energy storage (W-CAES) is a promising solution; however, conventional single-stage W-CAES designs exhibit limited energy capture and low efficiency. To address these limitations, this study proposes a multistage W-CAES system and develops an integrated model for comprehensive performance evaluation. Thermodynamic and energy flow analyses are systematically conducted to optimize the design. Additionally, a comparative evaluation is performed to examine the influence of key parameters, such as storage pressure and number of stages, on overall performance. Results indicate that increasing the number of compression stages significantly improves both capture factor and round-trip efficiency under equivalent storage conditions. In particular, a three-stage W-CAES achieves a capture factor of 9.1 % (versus 1.46 % for a single-stage) and a round-trip efficiency of 70.01 % (up from 64.53 %) at the same wave condition and storage pressure, demonstrating the benefits of the multistage approach. A Tasmanian case study demonstrates feasibility: 500 two-stage units could offset up to 20 % of Tasmania's imported energy, depending on seasons. These findings highlight the enhanced performance of multistage W-CAES and its potential to make significant contributions to renewable energy integration.

Accurate prediction of valve pressure drop in diaphragm compressors is essential for evaluating flow rate, energy consumption, and operational safety. However, empirical-correlation (EC) models and steady CFD models with mean-velocity boundary conditions (S-CFD-MEAN) cannot accurately reproduce the valve pressure-drop characteristics, because pressure loss in diaphragm-compressor valves is dominated by strong transient effects and local losses. In contrast, high-accuracy transient CFD models are limited by complex modelling procedures and high computational cost. To address these limitations, a valve pressure-drop prediction model based on an equivalent-energy boundary condition is proposed. First, a high-accuracy transient CFD model (HA-T-CFD) was established under representative operating conditions to characterize the flow fields upstream and downstream of the valve and quantify the contributions of the acceleration term, friction term, and local-loss term during the operating cycle. Based on this analysis, a steady CFD model using the RMS inlet velocity (S-CFD-RMS) was developed. In this model, the RMS inlet velocity is used as an equivalent-energy boundary condition to approximate the cycle-averaged effect of transient pressure losses. The results show that S-CFD-RMS predicts the cycle-averaged valve pressure drop with a maximum deviation within 5% compared with HA-T-CFD. In comparison, the maximum deviations of S-CFD-MEAN and the EC method are approximately 22–24% and 47–53%, respectively. In addition, S-CFD-RMS reduces the computational time by >96.7% compared with HA-T-CFD. Its high accuracy and low computational cost make the model suitable for rapid pressure-drop evaluation, flow-passage optimization, and multi-condition design screening in diaphragm compressors.

The goal of this research is to improve the cooling performance of IGBT power modules by combining microchannel heat sinks (MCHS) with high-conductivity thermal interface materials (TIMs). Two easy-to-manufacture MCHS designs, parallel rectangular channels and circular pin-fins, are examined to emphasize practical feasibility, while newly developed TIMs, specifically graphite sheet and liquid metal, are incorporated to further improve heat dissipation. Additionally, a double-layer channel structure is proposed to reduce pressure drop with minimal impact on cooling efficiency. The cooling performance of the proposed MCHS and TIM configurations is evaluated using a three-stage coupled CFD simulation under a wide range of electrical loading conditions to ensure realistic operating scenarios. It is found that the double-layer pin-fin design provides the best thermal performance among the heat-sink configurations, and both graphite sheet and liquid metal TIMs significantly enhance cooling; however, considering feasibility, the graphite sheet is preferred over liquid metal. The combined heat-sink and TIM solution reduces the maximum device temperature and temperature non-uniformity by more than 60 °C and 5 °C, enabling a 25 % increase in maximum power-handling capability. Consequently, up to 50 % higher switching frequency, together with a 17 % increase in acceleration factor, demonstrates the achieved enhancement in both reliability and AC output quality. Overall, this study presents a practical and high-performance cooling system design for IGBT power modules that surpasses current solutions while maintaining manufacturing feasibility.