International Journal of Heat and Mass Transfer最新文献

{"title":"Hydrophilic SiO2 nanoparticle deposition on boiling surface: Molecular dynamics insights into deposition behavior and heat transfer performance","authors":"Jielin Luo ,&nbsp;Shihao Wei ,&nbsp;Zhuohang Zhang ,&nbsp;Kaiyin Yang ,&nbsp;Gongran Ye ,&nbsp;Hongxing Yang ,&nbsp;Trevor Hocksun Kwan","doi":"10.1016/j.ijheatmasstransfer.2025.127872","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127872","url":null,"abstract":"<div><div>Because of its outstanding heat transfer performance, boiling phenomenon is widely used in heat dissipation of electronic devices. Despite its boiling enhancement effect, the inclusion of nanoparticle in boiling process leads to severe deposition, showing nonnegligible impact on heat transfer coefficient. However, existing studies lack the analysis on its inherent mechanism from microscale deposition behavior to macroscale boiling performance. In this study, a hydrophilic SiO₂ nanoparticle is investigated in typical boiling conditions, while its deposition process and accompanying effect on boiling performance are analyzed. The boiling process on copper substrates with heating temperature from 460 K to 520 K is simulated via molecular dynamics. At temperature above 500 K, over 96 % of water evaporates within 15 ns, leading to the formation of a dense deposited SiO₂ layer, and thus hindering further water evaporation. Cross-sectional density profiles demonstrate the water evaporation hindering by this deposited layer. Radial distribution functions and O<img>Si<img>O bond‑angle distributions peaking at 108.5° confirm short‑range structural ordering at the interface, without any abnormal bond breakage observed. This deposition layer reduces the interfacial thermal resistance by 70.91 % at 520 K, while correspondingly increasing heat transfer coefficient by 243.86 % for constant heating-temperature condition. The circumstance under constant heat flux is also discussed, with quantitative comparison with existing data. These molecular-scale findings offer new insights into the discovery of nano-altered heat transfer mechanism, promote thermal management for high-power microprocessors, and contribute to the advancement of emerging energy technologies.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127872"},"PeriodicalIF":5.8,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improving ground heat transfer simulations in cold climates through field-validated boundary conditions","authors":"Siim Lomp,&nbsp;Jaanus Hallik,&nbsp;Targo Kalamees","doi":"10.1016/j.ijheatmasstransfer.2025.127864","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127864","url":null,"abstract":"<div><div>Highly insulated slab-on-ground construction with sub-slab insulation is widely used in cold climates. Accurate assessment of heat loss, hygrothermal behaviour, and frost heave risk requires understanding ground temperature dynamics around and beneath buildings. This study combines long-term field measurements with hygrothermal simulations to evaluate ground heat transfer modelling accuracy.</div><div>Temperature data were collected at three sites in Estonia using measurement piles on undisturbed ground and self-installed sensors near a building. One-dimensional simulations were conducted over nine years using a verified finite element software for transient heat, air and moisture transfer in building materials, while three-dimensional building-coupled simulations were performed over three years using a verified finite element software for transient 3D heat transfer. Boundary condition configurations were varied to assess their impact on accuracy.</div><div>Simulations using ground-temperature boundary condition below the surface buffer layer on undisturbed ground achieved the best accuracy, root mean square error (RMSE) ranging from 0.2 to 0.5 K across all depths and sites. Daily deviations up to ±1.4 K occurred at shallow depths during seasonal extremes. In contrast, using air temperature, snow cover, and its constant thermal conductivity increased RMSE up to threefold and caused deviations up to ±3 K. Heavy rainfall events produced transient temperature differences, especially in upper soil layers.</div><div>Near the slab perimeter, model–measurement discrepancies were larger, likely due to simplified treatment of moisture and thermal inertia. Results indicate that accounting for dynamic soil moisture and snow properties, along with improved representation of thermal capacity near building edges, could enhance simulation performance.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127864"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhanced prediction of condensation pressure drop in mini and microchannels using physics-assisted machine learning","authors":"Farshad Barghi Golezani ,&nbsp;Jiayuan Li ,&nbsp;Forouzan Naderi ,&nbsp;Alexander Meylikhov ,&nbsp;Logan M Pirnstill ,&nbsp;Sung-Min Kim ,&nbsp;Issam Mudawar ,&nbsp;Chirag Kharangate","doi":"10.1016/j.ijheatmasstransfer.2025.127838","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127838","url":null,"abstract":"<div><div>Reliable prediction of frictional pressure drop during condensation in mini‑ and microchannels underpins both thermal management effectiveness and overall heat transfer performance in compact two‑phase heat exchangers, cold‑plates, and on‑chip cooling loops. Excess pressure loss burdens pumps, raises electrical consumption, and can destabilize flow, whereas under‑prediction risks temperature overshoot and premature dryout. Conventional empirical correlations and flexible machine-learning models can lose accuracy once channel size, working fluid, or operating conditions stray beyond their testing range. This study uses a physics‑assisted machine‑learning framework that overlays an XGBoost residual learner on the Kim–Mudawar separated‑flow correlation to achieve high fidelity and robustness in pressure drop prediction. A curated database of 6566 condensation data points (40 studies; 0.07 ≤ <em>D</em>_h ≤ 6.22 mm; 32.7 ≤ <em>G</em> ≤ 1926 kg m⁻² s⁻¹; 22 fluids) was assembled. Four feature sets (physical, dimensionless, statistically selected, full) were evaluated, and Bayesian hyper‑parameter optimization combined with five‑fold cross‑validation plus fluid‑ and mass‑velocity holdouts quantified both interpolation and extrapolation. Across the full dataset, physics‑assisted machine‑learning lowered the mean absolute percentage error from 24 % with Kim–Mudawar and 9.5–10.3 % with pure machine learning to 7.4–8.3 %, achieving R² &gt; 0.985. For the benchmark refrigerant R134a, interpolation mean absolute percentage error dropped from 22 % (Kim–Mudawar) to ≈14 %. For dielectric fluids HFE7000/HFE7100 (unseen during training) extrapolation error fell from &gt;150 % with pure machine learning to ≈40 %. Mass‑velocity holdouts confirmed ≤15 % error at high mass velocity and ≤42 % at the most challenging low mass velocity conditions. These advances enable more reliable pump sizing, manifold design, and thermal‑resistance budgeting, directly supporting the development of energy‑efficient, high‑heat‑flux thermal management hardware for electronics, electrified vehicles, and aerospace platforms.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127838"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155119","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mechanism analysis of multi-peak wall temperature of regasification heat transfer deterioration for supercritical methane in a horizontal tube","authors":"Changliang Han ,&nbsp;Zhipeng Chen ,&nbsp;Yuhang Chen ,&nbsp;Haokang Deng","doi":"10.1016/j.ijheatmasstransfer.2025.127876","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127876","url":null,"abstract":"<div><div>Supercritical heat transfer deterioration (HTD) phenomenon in tubes is frequently manifested as the distribution of multi-peak wall temperature along the tube length. Quantifying the HTD characteristics of supercritical methane (S-CH₄) in a horizontal tube is critical for improving thermal control of liquefied natural gas (LNG) vaporizers. In this paper, four different regasification heat transfer modes of S-CH₄ are firstly clarified, and the pseudo-phase distribution and vapor-like film (VLF) variations under different modes are also identified. Moreover, the mechanism of multi-peak HTD is deeply unveiled based on the progressive pseudo-boiling theory and virtual orifice contraction effect. Results demonstrate that one normal heat transfer and three HTD (single-peak, double-peak and triple-peak) modes are observed. The thickness of VLF (<em>δ</em>_VLF) near the top generatrix is much greater than that at the bottom generatrix, the former (0.505 mm) is 14 times that of the latter (0.035 mm) on the typical cross-section under the single-peak HTD mode. When HTD occurs, the expanding VLF will exert a virtual orifice on the core liquid-like flow. The alternating dominance of evaporation momentum force and inertia force mainly causes periodic <em>δ</em>_VLF. The average <em>δ</em>_VLF,bot decreases by 62.19% when mass flux increases from 370 kg/m²·s to 650 kg/m²·s, which attributes to the variations of thermal conductivity and specific heat of VLF. Rising pressure can yield smaller supercritical <em>K</em> number and thinner quantitative <em>δ</em>_VLF. The average <em>δ</em>_VLF,bot decreases 73.91% when pressure increases from 6.93 MPa to 12.5 MPa. Finally, novel assessment system and pattern maps of multi-peak HTD for S-CH₄ are established. Two new dimensionless correlations are proposed to predict the multi-peak wall temperature position and magnitude of S-CH₄ with the mean absolute relative deviations of 16.98% and 6.56%. The present findings not only benefit the understanding of supercritical heat transfer, but also provide crucial reference insights for thermal design and operational safety standards of LNG vaporizers or other related engineering equipment.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127876"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155114","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unlocking the Potential of Integrated Cooling and Power Delivery in Multi-chip Power Electronics Packages with Triply Periodic Minimal Surfaces (TPMS)","authors":"Ahmet Mete Muslu,&nbsp;Yogendra Joshi","doi":"10.1016/j.ijheatmasstransfer.2025.127866","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127866","url":null,"abstract":"<div><div>As power density demands in power electronics escalate, integrated power delivery with embedded fluidic cooling has gained interest to leverage the exceptional characteristics of wide bandgap semiconductor devices. However, application-specific constraints—such as dielectric coolants, fewer package layers, and lack of top-sided access for fluid ports—significantly limit the potential of this technology, leading to excessive fluid heating and reduced cooling effectiveness. These challenges become even more pronounced with homogeneous heat transfer enhancement structures, which lack the ability to provide spatial control of flow patterns. To unlock the potential of integrated cooling and power delivery, we proposed a hybrid architecture combining triply periodic minimal surfaces (TPMS) with micropillars to target local hot spots and enhance flow mixing where most needed. The cold plates were additively fabricated using binder jetting technique, and computational models were experimentally validated to unveil underlying physics, examining the unique flow structures generated by the combination of TPMS and pillars under diagonal flow. The swirling motion through a network of high-curvature channels intensified local vorticity, leading to disrupted thermal boundary layers and enhanced convective heat transfer. Unlike conventional micropillar-based cooling, which suffers from a downstream decline in local Nusselt number (<em>Nu</em>), the hybrid design expands the effective heat removal area by sustaining a positive downstream <em>Nu</em> gradient in the hot spot region. Overall, the hybrid architecture achieves a 17.9–57.0% increase in the average <em>Nu</em> compared to the traditional solution, offering promising potential to meet increasing power density demands with reduced cooling power overhead.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127866"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Thermal analysis of baffle effects in cryogenic hydrogen tanks","authors":"Charles Abdol-Hamid Owens ,&nbsp;Hoyeon Park ,&nbsp;Robert Joseph Flores ,&nbsp;Luke Wentlent ,&nbsp;Jacob Brouwer ,&nbsp;Jaeho Lee","doi":"10.1016/j.ijheatmasstransfer.2025.127863","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127863","url":null,"abstract":"<div><div>Minimizing heat ingress into cryogenic storage tanks is a key challenge in thermal system design. While prior studies have examined insulation materials or baffle effects in isolation, the coupled influence of these mechanisms on overall thermal resistance remains underexplored. This study introduces a novel approach that parameterizes baffle geometry and size and couples it with insulation performance by evaluating how baffle configurations influence both internal convection and insulation resistance in a two-phase cryogenic hydrogen tank. A 2D axisymmetric model at 50 % fill level is simulated under transient conditions using commercial multiphysics simulation software, with the working fluid initially at saturation temperature and standard atmospheric pressure. The model is validated against experimental LH₂ tank studies, with strong agreement in both liquid and vapor phase temperature trends. Increasing circular ring baffle diameter from 1 meter to 2 meters reduces the average Nusselt number by 16 %, velocity magnitude by 44 %, and vorticity magnitude by 17 % compared to the no-baffle case. The optimal 2 m baffle is used to assess insulation-baffle coupling, revealing that baffle inclusion increases the convective-to-conductive resistance ratio by up to 30 % at low insulation thermal conductivities but diminishes to 4 % as conductivity rises. Geometric analysis shows that curved baffles outperform rectangular designs through smoother boundary layer development, stabilizing flow, and suppressing convective transport. These findings establish that thermal performance in cryogenic tanks depends on coupled design variables, underscoring the need to jointly optimize baffle structure and insulation properties rather than treating them as isolated components.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127863"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Novel complementary cathode flow field for enhancing hydrothermal performance of PEMFC: Optimization and analysis","authors":"Huo Lin,&nbsp;Zijun Zheng,&nbsp;Zhou Wang,&nbsp;Zhihui Zhang,&nbsp;Changhong Wang","doi":"10.1016/j.ijheatmasstransfer.2025.127857","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127857","url":null,"abstract":"<div><div>A rational cathode flow field optimization design plays a key role in enhancing the overall performance of proton exchange membrane fuel cell (PEMFC). In this study, a complementary cathode flow field is proposed, which achieves structure-function dual complementarity by ingeniously integrating recesses and blocks. The complementary stepped flow field is further derived based on the flow field characteristics. A three-dimensional numerical PEMFC model is developed to comparatively investigate the effects of flow field geometry parameters on mass transfer and drainage performance. The complementary effects of recesses and blocks are thoroughly analyzed. The results demonstrate that recesses effectively mitigate the high pressure drop caused by blocks, while blocks promote more gas diffusion into recesses. Recesses address the defect of disordered water migration induced by blocks, while blocks compensate for the shortcoming of difficult water discharge in recesses after water collection. Compared with the conventional straight channel (SC), the net power densities of the recess-only and block-only channels increase by 12.56% and 3.92%, respectively, while that of the novel channel increases by 17.25%, even exceeding the sum of the previous two increases. RB2-6 is the best-performing complementary cathode flow field. At the center lines of channel and rib, compared with SC, RB2-6 improves the average O₂ concentration by 17.27% and 430.58%, respectively, and reduces the average water saturation by 16.92% and 17.92%, respectively. ST3 based on RB2-6 is the best-performing complementary stepped flow field, achieving a 38.78% increase in net power density compared with SC and more uniform species distribution.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127857"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synergistic design of bi-level heat sink combining topology-optimized microchannels with jet impingement for high-heat-flux applications","authors":"Qidong Sun ,&nbsp;Junzhe Guo ,&nbsp;Sheng Zhou ,&nbsp;Mingji Chen ,&nbsp;Da Geng ,&nbsp;Ran Tao","doi":"10.1016/j.ijheatmasstransfer.2025.127875","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127875","url":null,"abstract":"<div><div>This study presents a bi-level heat sink designed through a manufacturing-aware framework that integrates macro-scale jet impingement manifolds with micro-scale topology-optimized channels. Prototypes were fabricated via laser powder bed fusion (LPBF) and examined using computed tomography (CT) to capture manufacturing-induced deviations. Coupled simulations and experiments evaluated three inlet–outlet configurations (2IN1OUT, 3IN2OUT, and 4IN5OUT, corresponding to two/one, three/two, and four/five inlet–outlet channel combinations, respectively) under high heat fluxes ranging from 83.3 to 200 W/cm². All designs achieved effective cooling, with the 4IN5OUT configuration delivering the best thermal performance in simulation, maintaining an average temperature of 51.3 °C, a peak of 54.0 °C, and a pressure drop of 31.0 kPa at 200 W/cm². CT analysis revealed that LPBF-induced imperfections, such as channel shrinkage and surface roughness, increased pressure drop by 8.8–30.9 % and explained the experimental–numerical temperature discrepancies of 7.2–16.8 %. The optimized 2IN1OUT design exhibited the smallest deviations, consistent with CT findings showing closer geometric fidelity. These results establish a robust design–fabrication–validation workflow that bridges digital optimization with realizable performance, offering a practical and scalable thermal management solution for next-generation high-power electronics.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127875"},"PeriodicalIF":5.8,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145155186","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Analytical model of thermal contact resistance generated at tool-chip contact interface during cutting H13 hardened steel with TiAlN coated tools","authors":"Guangchao Hao ,&nbsp;Xiaoliang Liang ,&nbsp;Yukui Cai ,&nbsp;Laixiao Lu ,&nbsp;Zhenzhong Zhang","doi":"10.1016/j.ijheatmasstransfer.2025.127858","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127858","url":null,"abstract":"<div><div>Thermal contact resistance (TCR) at tool-chip contact interface is caused by the incomplete tool-chip contact and the oxide layer formed on tool rake face. Value of the TCR characterizes the difficulty of transferring cutting heat from chip to cutting tool. This paper aimed at establishing an analytical model for quantifying the TCR. Firstly, two models were developed to calculate values of the TCR caused by incomplete contact and that caused by oxide layer, respectively. A total TCR analytical model was established to describe the relationship between the two models. Secondly, orthogonal cutting experiments of H13 hardened steel machined by TiAlN coated tool were conducted. Value of the TCR was quantified by the established model with cutting experimental results. At last, heat transfer experiments were carried out to verify correctness and applicability of the model. Different cutting speeds and workpiece materials verified and broadened the applicability of the model. Therefore, the analytical model was applicable to TiAlN coated tool cutting experiments. It provides a possibility for predicting cutting tool temperature accurately.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127858"},"PeriodicalIF":5.8,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145118520","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Direct numerical simulation of Rayleigh–Bénard convection based on physics-informed neural networks with transfer learning","authors":"Shuran YE ,&nbsp;Jianlin Huang ,&nbsp;Yiwei Wang ,&nbsp;Chenguang Huang","doi":"10.1016/j.ijheatmasstransfer.2025.127823","DOIUrl":"10.1016/j.ijheatmasstransfer.2025.127823","url":null,"abstract":"<div><div>Rayleigh–Bénard (RB) convection, characterized by a fluid layer with bottom heating and top cooling, serves as a fundamental model system in fluid dynamics research, serves as an essential paradigm for studying thermally driven flows, offering fundamental understanding of heat transfer, fluid mixing, and turbulent transition processes that occur widely in nature and industrial systems. This study introduces the application of Physics-Informed Neural Networks (PINNs) augmented with transfer learning techniques. Using transfer learning, our aim is to take advantage of the knowledge gained from training PINNs on a Ra condition to improve predictions for other Ra values. Preliminary results show that transfer learning-enhanced PINNs successfully capture the convective regime while avoiding convergence to steady-state solutions, enabling efficient prediction across varying Rayleigh (Ra) numbers without requiring full retraining. Furthermore, different ways of transferring models are also proposed to explore the feasibility of knowledge transfer across different natural convection configurations, including cases with varying inclination angles and Prandtl (Pr) numbers. The effective incorporation of transfer learning into PINNs have demonstrated promising capabilities for RB convection modeling, suggesting several key areas for future investigation. Further advanced transfer strategies suited to particular physical systems and conditions can be investigated as PINNs develop.</div></div>","PeriodicalId":336,"journal":{"name":"International Journal of Heat and Mass Transfer","volume":"255 ","pages":"Article 127823"},"PeriodicalIF":5.8,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145118627","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}