Abstract The increasing power densities of processors and server racks necessitate advanced thermal management solutions for data centers. Pumped two-phase (2P) direct-to-chip (DTC) cooling leverages boiling of a dielectric refrigerant and offers high efficiency, reliability, and scalability. However, a comprehensive framework to analyze thermal contributions in 2P DTC systems is lacking due to fundamental differences between the heat transfer characteristics of 2P and single-phase (1P) DTC cooling. In this work, a systematic thermal analysis is conducted for 2P DTC systems coupled with chilled facility water (FW). The end-to-end temperature difference between processor case and FW inlet is adopted as the system performance metric. Refrigerant temperatures at the cold plate outlet and condenser inlet are introduced as intermediate nodes to break down the end-to-end temperature difference into three contributions from cold plate, vapor line pressure drop, and condenser. Using this framework, system-level comparisons between representative 1P and 2P DTC systems are performed. A correlation-based microchannel model is used to analyze cold plate performance with PG25 in 1P mode and R515B in 2P mode. The two systems exhibit comparable performance at a processor heat flux of 100 W/cm2. Under the representative conditions considered, the 2P system enables a 5 °C higher FW inlet temperature at 200 W/cm2, suggesting potential for reduced facility-side cooling energy consumption. This work establishes a quantitative framework for analyzing 2P DTC systems and supports their large-scale adoption in data centers.
Capillary-fed thin-film evaporation is a promising route for passive thermal management, enabling high-performance cooling devices like heat pipes and vapor chambers. The wick design is crucial, and high performance is defined by sustaining large heat fluxes at low overall thermal resistance. Previous modeling efforts on evaporator wicks have predominantly focused on silicon pillar arrays, which provide limited guidance for microporous wicks. Microporous wicks have been widely used in passive thermal management owing to their capability of dissipating high heat flux by increasing the effective evaporation area and sustaining the liquid supply. In this work, we take copper inverse opal (CIO) as a representative microporous wick and develop a predictive numerical pore-scale evaporation model together with an optimization framework. The model captures pore-scale complexity by computing local contact angles from surface normals to set a dry-out criterion and by evaluating thermal resistance with a spherical cap-based interfacial area model. We then couple the model with a multi-objective genetic algorithm to optimize axially graded CIO wicks at a fixed overall thickness, where the pore geometry is varied section by section along the flow direction, to increase the critical heat flux (CHF) while lowering the overall thermal resistance. For designs with fixed thickness, pore-size graded CIO wicks outperform uniform-pore designs, achieving up to a 33.8% increase in CHF and a 62% reduction in thermal resistance. This framework can be extended to different microporous wicks and offers guidelines for capillary-fed evaporative cooling systems.
Ensuring the vibration reliability of electronic packages is crucial for their long-term operation, especially in high-reliability environments. The research investigates the effects of boundary conditions, component placement, and the number of components on the fatigue lifetime of printed circuit board assemblies (PCBA). While traditional approaches relied on numerical models for estimating fatigue life, the current research utilized advanced machine learning models, extreme gradient boosting (XGBoost), and stochastic gradient boosting (SGB), to enhance accuracy and efficiency. The finite element analysis (FEA) and experimental results were rigorously validated by matching natural frequencies and strain values, confirming the robustness of the models in simulating real-world conditions. The machine learning models demonstrated exceptional performance, with XGBoost and SGB achieving high R2 values of 0.957 and 0.978 and low mean absolute percentage error (MAPE) scores of 0.1587 and 0.1322, respectively. The machine learning predicted and numerically estimated fatigue lifetime was further validated through experimentally resonance-based fatigue analysis. The research reveals that strategic component placement and careful consideration of boundary conditions can significantly improve the fatigue lifetime of PCBA. By integrating Shapley additive explanations (SHAP), the research enhances the interpretability of machine learning models, providing insights that contribute to optimizing the design and performance of electronic packages under vibration loading.
The growing demand for higher computational efficiency in data centers and AI hardware is accelerating the adoption of three dimensional (3D) heterogeneous integration and immersion-cooling technologies. Oil immersion offers significantly enhanced heat dissipation compared to conventional air cooling, improving thermal stability and enabling higher power densities. However, the thermo-mechanical environment created by oil immersion also introduces new reliability challenges. In 3D packages, mismatches in material properties like Coefficient of Thermal Expansion (CTE) and Elastic Modulus (E) can induce interfacial stresses that lead to through-silicon via (TSV) cracking, printed circuit board (PCB) warpage, and solder-joint fatigue failures. This study investigates the influence of oil-immersion cooling on the thermomechanical reliability of solder joints and TSV structures in advanced 3D packages. Plastic strain accumulation in solder joints under extreme thermal cycling is analyzed to evaluate the likelihood of crack initiation and propagation. Additionally, J-integral computations are performed to quantify crack-tip driving forces in TSVs across different PCB configurations, providing a comprehensive assessment of package lifetime reliability. To further enhance TSV resilience, this work examines alternative via-fill materials-including tungsten, silver, nickel, carbon nanotubes, aluminum, molybdenum, gold, graphene, and annealed copper which are possible alternatives-and compares their crack-propagation tendencies using J-integral-based metrics. Overall, this research not only identifies reliability risks associated with oil-immersion environments but also proposes material-level mitigation strategies.
The rapid development of artificial intelligence (AI) technology has placed higher demands on AI computing power, and the power of GPUs is increasing rapidly, which in turn imposes greater requirements on thermal management. Due to the non-uniform component arrangement, the heat distribution within the chip is also uneven, making temperature control of hotspots extremely important. This study focuses on the thermal performance of jet manifold microchannel (JMMC) heat sinks employing a high-thermal-conductivity diamond-copper composite material to mitigate hotspots on large-area, high-power AI chips containing six heat sources. Combined with CFD numerical simulation, regression analysis, and nonlinear programming, the present work establishes a high-precision mathematical model to quickly design the structural parameters of JMMC for high-power AI chips. This study analyzed the impacts of individual structural variables, including jet diameter, jet height, inclination angle, mass flux, and number of outlets, and identified the key factors affecting the maximum chip temperature, temperature uniformity, and pressure drop. A dimensionless constant, a (jet height-diameter ratio), is constructed to illustrate the combined effects of jet height, diameter, and substrate thickness. Multi-objective algorithms combined with the Pareto front are used for structural optimization, achieving optimal structural performance. The optimized JMMC structure demonstrated significant performance improvements compared with the pre-optimized design. Under conditions of a peak heat flux of up to 1200 W/cm2 and a flow rate of 15.73 g/s, the maximum hotspot temperature was reduced by 10.98 degrees C, temperature uniformity was improved by 76.6%, and the thermal resistance of the JMMC was decreased by 35.47%.
This study analyzes the feasibility of introducing laser assisted soldering (LAS) into solder ball mounting (SBM) process. The vertical cavity surface emitting laser (VcSEL) adopted in this LAS system radiates an infrared (IR) laser beam with a wavelength of 980 nm over a 240 mm & times; 95 mm area. The workable range of LAS applicable to SBM process has been determined by varying laser power and radiating time. Among the tested conditions, the most favorable condition is 12,000 W for 1.5 s. Before SBM, the average warpage of packages used in this experiment was 240.3 mu m. The average warpage of them has increased to 271.9 mu m after SBM using LAS, but it has increased to 350.4 mu m after SBM using massive reflow (MR). For LAS, X-section result of solder ball showed that the intermetallic compound (IMC) layer between the ENEPIG pad and the solder was undetectable by scanning electron microscopy (SEM) inspection, but for MR, the IMC layer thickness was thicker than 2 mu m. The reason why the solder balls mounted with LAS have 11.9% stronger ball shear strength than that with MR is expected to be because IMC layer was not detected. Temperature cycle (TC) test results also showed that solder balls mounted by using LAS had thinner IMC than MR even after 1000 cycles of TC test. Therefore, LAS can be considered as a good alternative to MR in SBM process because it has advantages such as improved warpages, bond properties, and shorter process time.
Holistic design of a modular energy-efficient edge datacenter with a compute density of 20 kW/m3 is presented. The datacenter uses a single chiller-less liquid loop to extract heat from server components and reject it to the ambient using a mixture of 25% propylene glycol and water. Heat is extracted from the chips using a multiple unit-cell shedding enhanced (MUSE) cold plate and rejected to the ambient air using a counterflow microchannel polymer heat exchanger (MPHX) dry cooler. The MUSE cold plate consists of a manifold to split flow into parallel paths and a large aspect ratio (7.3:1) microscale pin fin heat sink fabricated using electrochemical additive manufacturing. The counterflow MPHX dry cooler design can provide higher effectiveness than metallic finned tube heat exchangers for identical air side pressure drop. Experiments of the cold plate and MPHX components are presented, and a system model of the cooling loop is developed. At an outdoor temperature of 40 degrees C, baseline results indicate that optimized liquid and fan flow rates reduce cooling power to 2.1% of compute power for a graphical processing unit (GPU) case temperature of 75 degrees C and heat flux of 74 W/cm2. Sensitivity studies are performed to assess the impact of improvements in thermal interface material, cold plate convective resistance, and dry cooler air-side heat transfer coefficient on system performance. Results show that, for this heat flux, the thermal interface material (TIM) and dry cooler have the largest impact on the system, with the best-case cooling power being under 1% of compute power.
Abstract This study investigated void formation during capillary underfill (CUF) encapsulation in chip-to-wafer (C2W) assemblies using multiphase finite volume method (FVM) simulations and machine learning (ML) analysis. A three-dimensional volume-of-fluid (VOF)–based multiphase model was developed to capture the transient, capillary-driven flow of a shear-thinning epoxy underfill within the C2W ball grid array (BGA) architecture. The numerical framework was validated against transparent-model flow visualization experiments, showing close agreement in flow morphology and filling time, with a maximum deviation of 6%, thereby confirming the predictive reliability of the model. Parametric investigations revealed that dispensing pattern was the dominant factor governing void formation, with I-type dispensing achieving approximately 71% reduction in average void content compared to L-type dispensing. Double-dispensing strategies further improved filling uniformity, with the 50–50 split ratio yielding the lowest void percentage, while BGA array density exhibited a secondary influence, as the denser 17 × 17 array produced slightly higher void formation than the 10 × 10 configuration due to increased flow resistance. ML-based prediction demonstrated that a response surface methodology (RSM)-style linear regression model achieved the lowest root mean squared error (RMSE) and highest R2, providing robust and interpretable prediction of void formation. The findings offer practical, data-driven guidelines for optimizing C2W underfill dispensing strategies to minimize void formation and enhance packaging reliability.
Abstract Ball grid array (BGA) inspection is critical for ensuring the reliability of semiconductor assemblies. Conventional Red-Green-Blue (RGB) imaging under a single illumination often yields high classification metrics with pretrained convolutional neural networks (CNNs), yet explainability analyses reveal that models may attend to spurious artifacts rather than solder ball defects. This paper introduces a png4ch multi-illumination dataset by stacking grayscale captures under four illumination intensities into a multichannel representation. A 1 × 1 convolutional projection maps the four channels into three, enabling compatibility with pretrained CNN backbones. Comparative experiments with imgShiny (RGB single-illumination) and png4ch show that while RGB yields more consistent high F1 scores, multi-illumination significantly improves attention alignment with defect regions (gradient-weighted class activation mapping (Grad-CAM)), underscoring the importance of dataset design for trustworthy industrial deployment. Keywords: BGA inspection, defect detection, multi-illumination, convolutional neural networks, explainable AI
Electronics cooling systems depend on compact motor-driven pumps to maintain thermal stability in high-power electronic applications. In such systems, improving motor torque density alone is insufficient, as electromagnetic enhancement must be achieved without compromising thermal safety or mechanical integrity under continuous thermal loading. This study proposes a multiphysics-constrained optimal design procedure for a 1.5 kW axial-flux consequent-pole motor (AFCP) intended for electronics cooling systems. Unlike conventional electromagnetic-only optimization approaches, the proposed procedure integrates topology selection, surrogate-assisted optimization, and multiphysics verification in a systematic manner. Four consequent-pole (CP) configurations were comparatively analyzed to identify a baseline topology exhibiting favorable torque characteristics and magnetic loading distribution. A regression-based surrogate model combined with a genetic algorithm was employed to optimize ten geometric design variables while enforcing explicit constraints on torque ripple, cogging torque, and electromagnetic operating limits. The optimized design was validated through three-dimensional finite element analysis (3D FEA) and subsequently assessed for structural and thermal feasibility under rated operating conditions. The optimized motor achieved a 5.6% increase in average torque, a 24.89% reduction in torque ripple, and a 29.73% decrease in cogging torque while reducing magnet volume by 21.16%. Multiphysics verification confirmed that these improvements were achieved without exceeding structural or thermal limits. These results demonstrate that the proposed optimal design procedure enables simultaneous enhancement of electromagnetic performance and thermal robustness for motor-driven electronics cooling applications.
Pad cratering is a major concern in high I/O ball grid array packaging architectures. Assessment of propensity for pad cratering is presently beyond the state of art. Failure in pad cratering is manifested through the fracture of the bulk resin below the interconnect pad, followed by the fracture at the resin-glass weave interface in the organic laminate. A new metric has been proposed and evaluated for assessing the propensity for pad cratering. A model has been developed using experimental measurements to predict pad cratering resistance, identifying the resin-glass-reflow process condition combinations that have the lowest propensity for pad cratering. Experimental data have been measured by analyzing the bulk behavior of the resins using a tensile specimen and by measuring the fracture toughness at the resin-glass interface through measurement of a specifically fabricated resin-glass bimaterial interface specimen. A total of 12 different resin formulations were studied to quantify the material properties under various reflow exposure conditions, ranging from no-reflow to six-reflow conditions. A four-point bend interface specimen was used for the measurement of the fracture toughness of the resin-glass interface. The interface behavior was measured for 30 glass-resin interfaces consisting of 15 resin types and three glass styles as a function of mode-I stress intensity factor to pristine, two reflow, and six reflow conditions. Multiple replicates have been tested for each combination to create a massive dataset. A predictive model has been developed to assess the pad cratering resistance for resin-glass-reflow condition combinations using principal component regression on experimental data, with the newly developed pad cratering resistance metric as the response variable. The model has been assessed for accuracy versus test data.
Closed loop thermosyphons (CLTs) are passive heat transfer devices, that when operating in a buoyancy-driven two-phase regime, can transfer heat nearly isothermally and thereby achieve excellent thermal performance. However, CLTs are susceptible to flow instabilities, particularly geysering, which can lead to transient fluctuations in pressure and temperature. This work presents an experimental investigation of geysering instabilities during the powering up of a small-scale CLT. Experiments are conducted for heat loads from 45 W to 260 W and water fill ratios from 48% to 81%. The experimental facility uses transparent tubing, enabling flow visualization. Loop performance is characterized through loop thermal resistance, dominant oscillation frequency of the evaporator outlet temperature, and standard deviation of the transient temperature data. Flow visualization revealed that the geysering instability follows a repeating cycle of subcooled boiling in the evaporator, followed by a buoyancy-driven increase in fluid column height in the riser, which draws subcooled liquid back into the evaporator causing bubble collapse and subsequent flow reversal, before the cycle begins again with subcooled boiling in the evaporator. Two distinct geysering oscillation behavior were observed: low-frequency, high amplitude and high-frequency, low amplitude modes. Lower fill ratios transition from low- to high-frequency oscillations with power, while higher fill ratios maintain low-frequency oscillations. Results show that increasing heat load decreases thermal resistance for all fill ratios, with the 62% fill ratio achieving the lowest values. These findings provide insight into the underlying mechanism of geyser boiling and establish trends in loop behavior with fill ratio and heat load.
Abstract This paper reports, to the authors' knowledge, the first use of low-temperature cofired ceramics (LTCC) to construct a monolithic device with all the architectural elements of a gridded RF ion thruster and its complex internal subsystem structures. Electric thrusters have become a common solution for in-space propulsion. Electric propulsion (EP) has continued to gain popularity due to its advantages over conventional chemical thrusters, such as few moving parts and higher efficiencies (specific impulses >1000 s). This allows manufacturing with novel construction materials like LTCC to deliver operational goals in a smaller, more integrated form factor. LTCC is leveraged in high-power and high-frequency electronic packaging because of its low dielectric losses and reliability in extreme conditions. Its manufacturing process involves punching and printing conductors on ceramic “green-tape” sheets that are then laminated and fired into a monolithic product. This enables parallel and scalable fabrication and integration of required propulsion components, such as propellant feeds, an ionization cavity, and embedded electrodes, thus simplifying the complexity, fitment, and assembly processes compared to current state-of-the-art EP systems. This work reports the advantages of LTCC as applied to the manufacturability of an electric thruster architecture through the manufacturability and plasma generation demonstration. Plasma was repeatably generated under vacuum with argon flowing at 95 sccm and RF input power of 22 W at 882 MHz. This effort reflects a NASA technology readiness level (TRL) 2 of a functional EP system and TRL 3 as a plasma generator.
Abstract This paper presents the study of thermodynamics, diffusion, and the activation energy of Sn–Cu solder on copper substrate. The present research presents composition of 99.3Sn–0.7Cu (SC) solder flux with model: SC G227 (provided by Rosin Core Solder Flux for Electronic Co., Ltd.), United State selected as a solder flux for the experimental procedures. Copper plate was used and modified for dynamics surface roughness (Ra) values: (0.26, 0.17, 0.62, 0.37, 0.12, and 0.12 μm) created by respective abrasive grit papers particle size classification reference standard (P: 400, 600, 800, 1000, 1200, and 2000) paper size. Solder joints of samples were tested at extreme temperature of 150 °C at intervals of reaction times (100, 200, 300, 400, 500, 600, 700, 800, 900, and 1000 h). Effect of surface roughness on thermodynamics potentials in intermetallic compounds (IMCs) formation was analyzed, how the diffusion process influences the IMCs formation and growth were also studied. The study focused on the kinetic equations that follow the diffusion-controlled method under liquid and solid-state conditions using first principle of thermodynamics. Activation energy and diffusion coefficients were determined using a modified Arrhenius growth equations with dynamics surface roughness (Ra) values: (0.26, 0.17, 0.62, 0.37, 0.12, and 0.12 μm). Findings reveal that (1000 h) reaction time has the highest activation energy Q (871.43 KJ/mol) with diffusion coefficient Do (3.58 m2/s), and (800 h) reaction time produced the lowest activation energy Q (64.29 KJ/mol) with diffusion coefficient Do (1.97 m2/s). Metallurgical engineering and electronics industry are the beneficiaries of this research.
Abstract In recent years, flexible electronic circuits, particularly flexible printed circuits (FPCs), are becoming widely used in wearable devices, flexible sensors, and liquid crystal displays. Inkjet printing technology has attracted attention for manufacturing FPCs, enabling low-cost and small-scale production of conductive lines using nanoparticle metal inks. The circuits' flexibility and durability are typically evaluated through bending and tensile tests. As electrical devices become smaller, increased current densities and Joule heating can lead to electromigration (EM) damage. EM is a phenomenon in which metal atoms are transported by high-density electron wind, causing damage to electronic lines. Ag nanoparticle lines, due to their porous structure, exhibit distinct EM damage. Therefore, understanding the damage characteristics of EM in actual situation is important for reliability evaluation of these lines. However, the EM damage characteristics of these lines under both mechanical stress such as bending and electrical stress have not been fully studied. This study investigates the effect of mechanical stress on the lifetime of Ag nanoparticle lines under electrical loading. High-density current loading tests were conducted on Ag nanoparticle lines drawn on the flexible substrates. After the tests, the line morphology was analyzed using a laser microscope to assess damage. As the results of current loading, applying tensile stress shortened the line lifetime compared to no stress, while applying compressive stress prolonged the time until voltage instability occurs by approximately 1.54 times. This research provides valuable insights into the reliability of flexible electronic circuits under both mechanical and electrical stresses.
Electronic packages are transitioning from cloud computing (controlled environment) to edge computing (uncontrolled environment), presenting new challenges, especially in high-vibration environments like transportation applications. The durability of critical components, such as solder joints, is crucial for the endurance of these packages. Therefore, there is a growing emphasis on modeling solder joints exposed to vibration environments. This study investigates the performance of board-level ball grid array (BGA) packages under harmonic and random vibration excitations, considering different boundaries and excitation conditions. We modified the boundary conditions from a four-corner screws mounting platform to a six-screws mounting platform and the excitation condition from very low harmonic excitation (linear behavior) to high excitation. For the first time in technical literature, we have demonstrated the geometric nonlinearity in the BGA package and its impact on the fatigue parameters of the solder joint. The results indicate that as the excitation increases, the package behaves nonlinearly, and the source of nonlinearity comes from the dynamics of the printed circuit board (PCB), resulting in a complete change in the fatigue parameters of the solder joint. These results can be instrumental in developing test strategies for the semiconductor design industry and facilitating the research community in devising a nonlinear dynamic modeling approach for electronic packages.
Polyethylene terephthalate (PET) and polymide (PI) are used for the additive fabrication of flexible hybrid electronics (FHE) due to their excellent thermal and dimensional stability and mechanical strength. However, their environmental impact is high due to minimal degradation even after decades in landfills. There is a growing global push for sustainable solutions for printed electronics. Recent studies indicate that alternatives, such as biobased PET, polylactic acid, and regenerated cellulose film, all derived from renewable resources, may not only match but, in some instances, surpass the performance of traditional PET. While these materials show promise in certain environments, comprehensive research on their adaptability across various conditions remains inadequate. With the escalating demand for electronics in diverse areas, ranging from packaging to healthcare, it becomes crucial to thoroughly evaluate the efficiency, longevity, and versatility of these renewable substrates. This research addresses these concerns, focusing on assessing the performance of biobased PET substrates in FHE applications and comparing their performance with that of traditional PET substrates. Research Objectives include: (a) To understand the significance of Flexible Hybrid Electronics in the context of sustainable and environmentally friendly electronics. (b) To determine the compatibility and performance of the biodegradable PET substrate during the screen-printing process, especially while printing with water-based inks. (c) To test the performance of printed circuits and calculate the error percentage between the actual and simulated performance.
Thermal management is crucial in power module design nowadays, significantly influencing the cost, performance, and reliability of the traction inverters. As we strive for greater power density and smaller form factors, innovative thermal solutions become imperative. In this study, we explored a two-phase cooling method for power electronics. The heatsink of a hybrid single- and two-phase cooler is integrated into a power module. We utilize surface enhancement features at the substrate-embedded evaporator to enable capillary-driven flow for passive coolant circulation and optimized phase separation. Additionally, a local reservoir is included in the cooler package that serves as a buffer to optimize the liquid supply to the evaporator. The two-phase cooling in our study leads to a high heat flux removal rate of 370 W/cm2 at a low superheat of 10 degrees C, and the coolant flowrate is less than 1 g/min, nearly 50 times lower than the pumped flow-based two-phase microcoolers. The feasibility of the completely passive operation is also studied. The proposed two-phase embedded cooler provides a promising solution for effective cooling and heat spreading of the next-generation power modules.
