Optimal Heat Transfer Performance Research Articles

Experimental investigation of the heat transfer characteristics, operating limits, and temperature distribution of a prototypically 3 m long two-phase closed thermosyphon for spent fuel pool passive cooling

AbstractThe current study investigated the heat transfer performance, operation limits, and temperature distribution of a 3 m prototype model for a large-scale straight two-phase closed thermosyphon with deionized water designed for passive cooling of nuclear spent fuel pools with a filling ratio ranging from 20% to 100% and a heat source temperature ranging from 45 to 80 °C. After the accident at the Fukushima Daiichi nuclear power plant, the need for a reliable cooling system for these pools increased, with thermosyphons emerging as a promising solution for passive cooling. The experimental procedure implemented in this study yielded a comprehensive understanding of the operation and phenomena inside a thermosyphon, providing crucial data for the validation and enhancement of numerical models that simulate relevant phenomena within nuclear power plants, including passive residual heat removal with thermosyphons. The results indicated that the optimal heat transfer performance was achieved at a filling ratio of 30%, where the thermosyphon begins operation at a heat source temperature of 45 °C. Additionally, the temperature distribution along the thermosyphon confirmed the operation limits, including partial dryout at a 20% filling ratio and geyser boiling and flooding (entrainment) limits at 75% and 100% filling ratios, respectively.

Impact of distance on heat transfer at the leading edge of blades in a rotating state

To enhance the high-temperature capability of turbine blades and meet the requirements of advanced design, this research examines the intricate flow characteristics and the heat transfer occurring on the internal walls effects of leading-edge impingement cooling structures under different impingement distances. In this study, copper block method is employed for heat transfer testing, extending the testing range of rotating impingement cooling channels from Redj = 1 × 104, Ro = 0.2 to Redj = 5 × 104, Ro = 2.0. The findings reveal that the Nusselt number along the channel is significantly influenced by flow distribution and turbulence intensity. At lower Re (Re ≤ 2 × 104), the flow velocity increases along the channel due to the decrease in pressure drop, resulting in enhanced heat transfer. However, at higher Re (Re ＞ 2 × 104), increased flow rate leads to intensified turbulence and thickening of the boundary layer, consuming fluid energy and reducing the temperature difference, thereby adversely affecting heat exchange around the impingement holes. Moreover, Nu decreases with increasing rotation speed, with the highest average decrease of 15.3 % when Ro reaches 0.25. For cases where Redj ＜ 4.3 × 104, an impingement distance-to-diameter ratio (Z/D) = 2 exhibits optimal heat transfer performance. Conversely, when Redj＞4.3 × 104, the influence of impingement distance will diminish. Larger impingement distances increase the HTC but are accompanied by increased non-uniformity. Therefore, selecting an appropriate impingement distance can reduce boundary layer formation and flow separation, thereby forming an optimal flow structure, which contributes to promoting heat exchange and improving heat transfer uniformity. A correlation based on geometrical parameters has been proposed to predict the average Nusselt number for jet impingement arrays with varying geometrical configurations.

Prediction and Optimization of Heat Transfer Performance of Premixed Methane Impinging Flame Jet Using the Kriging Model and Genetic Algorithm

In practical applications, rapid prediction and optimization of heat transfer performance are essential for premixed methane impinging flame jets (PMIFJs). This study uses computational fluid dynamics (CFD) combined with a methane detailed chemical reaction mechanism (GRI–Mech 3.0) to study the equivalence ratio (ϕ), Reynolds number (Re) of the mixture, and the normalized nozzle–to–plate distance (H/d) on the heat transfer performance of PMIFJs. Moreover, the Kriging model (KM) was used to construct a prediction model of PMIFJ heat transfer performance. A genetic algorithm (GA) was used to determine the maximum likelihood function (MLE) of the model parameters for constructing KM and identify the points with the maximum root mean square error (RMSE) as the new infilled points for surrogate–based optimization (SBO). Combining these methods to analyze the simulation results, the results show that the global heat transfer performance of PMIFJs is enhanced with the increase in ϕ, the increase in Re, and the decrease in H/d. Sensitivity analysis points out that Re and ϕ significantly affect enhanced heat transfer, while H/d has a relatively small effect. In addition, GA was also used to search for the optimal heat transfer performance, and the global heat transfer performance at specific conditions was significantly enhanced. This study deepens the understanding of the heat transfer mechanism of impinging flame jets and provides an efficient method framework for practical applications.

Study on coupled heat transfer of pyrolytic kerosene and supercritical CO2 in zigzag-type PCHE used for hypersonic vehicle power generation system

Supercritical CO2 closed cycle is an effective solution to the power generation challenge of hypersonic vehicles. The coupled heat transfer between aviation kerosene and supercritical CO2 within the zigzag-type printed circuit heat exchanger (PCHE) is the core thermal process of the whole cycle. Starting from the enhancement of the cycle's power generation, this study analyzes the influence mechanism of structural parameters on the heat transfer process through numerical simulation. The research indicates that a deep pyrolysis reaction of aviation kerosene, with the chemical heat absorption rate accounting for over 65 % of the total heat transfer rate. Due to the significant nonlinearity introduced by pyrolysis, traditional enhanced heat transfer design methods may even weaken its heat exchange. Unlike most studies, this type of PCEH achieves optimal heat transfer performance with a larger channel diameter and a smaller bend angle. Additionally, the increase in diameter and the decrease in bend angle enhance the power capacity of kerosene and its pyrolysis products, significantly improving the efficiency of thermal-electric conversion for the cycle. Finally, the study has determined a diameter of 2.4 mm and a bend angle of 30° constitute a set of structural parameters most suitable for the PCHE.

Effects of geometry on the thermal-fluid dynamics of a transition water flow in a square heated asymmetric cavity channel

Cavity channels candidate as a promising passive cooling configuration in thermal management. The superior performance and the limited scientific investigation regarding cavity channels dictate the need for better understanding of the thermal-fluid dynamics. To the best knowledge of the authors, there are few works investigating the effect of the geometry on the thermal-fluid dynamics of cavity channels with no focus on thermal data and quantification of the recirculation effect which can depend on geometrical parameters. In this work, numerical RANS simulations were performed in ANSYS Fluent with the kT-kL-ω turbulence model for a transitional water flow in an asymmetric square cavity channel with different attachment angles in the range of 20°–90°. The optimal heat transfer performance in terms of uniform and high Nusselt numbers is provided by the configuration with an angle in the range of approximately 30°–40°. Recirculation plays an important role in enhancing heat transfer and its strength is associated with the penetration capacity in the boundary layer. Two regions of strong and weak recirculation are identified along the channel; when the recirculation is weak, other mechanisms such as the thinning of the thermal boundary layer play a more important role. By increasing the attachment angle it is shown the local thinning of the thermal boundary layer, an augmentation of the magnitude of the negative turbulent heat flux peak, and a similarity of the conductive sublayer thickness with the recirculation strength.

Multi-Output Regression Algorithm-Based Non-Dominated Sorting Genetic Algorithm II Optimization for L-Shaped Twisted Tape Insertions in Circular Heat Exchange Tubes

In this study, an optimization method using various multi-output regression models as model proxies within the NSGA-II framework was applied to determine the geometric parameters (P, W, D) of L-shaped twisted tape inserts for achieving the optimal overall heat transfer performance in a circular heat exchange tube. Herein, 4 multi-output regression models, namely, MOLR, MOSVR, MOGPR, and BPNN, were selected as proxy models and trained on a dataset containing 74 groups of data. The training results indicated that the MOGPR model, balancing high accuracy and low error conditions, exhibited moderate training times among the four algorithms. BPNN showed a comparatively lower comprehensive training effect, obtaining training accuracy close to that of the MOGPR algorithm but with approximately twice the training time. The worst fitting performance was gained with the MOSVR algorithm. Due to its fitting performance, the MOSVR algorithm was excluded from the subsequent NSGA-II model proxy. Through multi-objective optimization with NSGA-II, the optimal structural dimensions for three sets of L-shaped twisted tape inserts were obtained to achieve the best overall heat transfer efficiency within the tube.

Heat transfer enhancement of power-law fluid in a solar-driven pretreatment system for microalgae hydrothermal liquefaction

In this article, the flow and heat transfer performance of microalgae slurry in a solar-driven rectangular absorption tube with embedded vortex generators were studied by data optimization analysis and numerical simulation. The orthogonal test design method was used to establish an orthogonal test table by changing the shape, height, spacing, and fluid flow rate of the vortex generator to determine the optimal combination of various factors. Then the response surface analysis method was used for further optimization analysis. Eventually, the flow and heat transfer performance of microalgae slurry were studied numerically, and the optimal data were verified. The results show that the optimal comprehensive heat transfer performance (PEC) is delivered by the cylindrical vortex generator with height of 9 mm and spacing of 28 mm when the inlet velocity of microalgae slurry is 0.02 m/s. It can be seen that the data analysis method combined with numerical simulation can optimize the experimental process parameters and save the cost of the engineering design process, effectively and practically.

Numerical simulation study on heat transfer and fluid performance of novel calcium carbide heat collection system

This article proposes a novel recovery system for the substantial waste heat from the hot calcium carbide (CA). The study employs numerical simulation to investigate the heat transfer performance of the heat collection system (HCS). The research explores the effects of outlet position, inlet velocity, and inlet temperature on the operational characteristics and heat transfer performance of the HCS. The results indicate that optimal heat transfer performance is achieved when the outlet position is at the center of the collector. Instantaneous efficiency (IE), exergy loss (EL), pressure drop (PD), and heat transfer coefficient (HTC) increase with the velocity, while effectiveness decreases with the velocity. The peak values for IE, effectiveness, and HTC are 28.22%, 29.69%, and 1212.45 W/(m2*K), respectively. The heat transfer performance of the HCS is highest when the inlet temperature is 280K. The proposed CA HCS holds significant importance for recovering waste heat from electric stone production, providing a theoretical basis for subsequent experimental work.

Effect of flow maldistribution on heat transfer performance and temperature field of plate-fin heat exchangers

Plate-fin heat exchangers (PFHEs) are widely utilized in aviation precoolers, emphasizing not only heat transfer performance but also the outlet temperature distribution, with flow maldistribution impacting both parameters. This study developed a numerical model of crossflow PFHE and a two-dimensional flow distribution model under the condition of flow maldistribution. Experimental validation shows a standard deviation of 6.9%. This investigation delves into the effects of flow distribution offset (ξ) and non-uniformity (ε) on heat transfer performance and outlet temperature distribution. With large temperature differences, variations in fluid properties due to temperature lead to changes in the surface heat transfer coefficient (SHTC). The combined effects of SHTC and temperature difference determine the heat transfer performance. The results indicate optimal heat transfer performance and outlet cross-sectional temperature distribution at ξ= − 0.67 and 1, respectively. For a given ξ, increased non-uniformity deteriorates heat transfer efficiency and outlet cross-sectional temperature distribution. At ε=0.66, relative to the uniform flow, the heat transfer performance and standard deviation of outlet temperature distribution decrease by 10.3% and 45.1%, respectively.

An experimental investigation on the nanofluids in a cavity under natural convection with and without the rotary magnetic field

This experimental study involves the use of two distinct categories of nanofluids, namely ferromagnetic and non-magnetic, within a square cavity that facilitates natural convection. There are five distinct concentrations associated with each nanofluid. Natural convection arises as a consequence of the thermal gradient between the opposing surfaces of the copper cavity, which has a thickness of 18 gauge. The purpose of utilizing the constructed electro-magnet assembly is to investigate the impact of the rotational magnetic field on the process of heat transfer. The manipulation of magnetic strength can be achieved by regulating the magnetic power and direct current (DC) power. The manipulation of the electromagnet’s spin can also be regulated. In the context of a rotational magnetic field, it is seen that the magnetic flux undergoes a transition from a positive value to an almost identical negative value throughout a full rotation. The optimal heat transfer performance is observed at a nanoparticle concentration of 0.1% by volume (φ) for both nanofluids. In the absence of a rotating magnetic field, the ferromagnetic nanofluid exhibited superior performance. When the Rayleigh number (Ra) is one order smaller than the critical Rayleigh number value, the heat transfer performance is often superior with nanofluid compared to demineralized water.

Correlation of Steam Velocity and Pipe Diameter with Heat Transfer Performance on 120° Half-Pipe Jacket

Double jacket mixing tanks offer temperature control and versatility for industrial processes where temperature-sensitive materials need to be mixed, stored, or processed. The selection of the appropriate jacket type in a double-jacket mixing tank is based on the structural strength and the optimal heat transfer performance. The type of jackets can be conventional, dimple, and half-pipe. The conventional jacket is easier to manufacture, but it is not resistant to high temperature and pressure. The dimple jacket has good heat transfer, but it is susceptible to damage. The half-pipe jacket has good structural strength, but its heat transfer is affected by the central angle and pipe diameter. 120° central angle has greater heat transfer and pressure drop than 180° central angle. In order to determine the effects of steam velocity on heat transfer performance, heating time, and pressure drop of the 120° half-pipe jacket, research occurred on pipes with 2, 2.5, and 3 inches of diameter. The calculating method of heat transfer in agitated jacketed vessels is applied to visualize the relationships. The effects of steam velocity on heat transfer, heating duration, and pressure drop respectively are polynomials of order 2, power, and polynomials of order 2 with an average R2 close to 1. The greater R2, the better the relationship between two variables, according to the equation. The 120° half-pipe jacket performance will be highly effective, with 1774 W, 2.1 minutes heating duration, 8.94 kPa pressure drop if the steam velocity is 10.50 m/s with 2.5 inches pipe diameter.

Influence of smooth heater size on critical heat flux and heat transfer coefficient of saturated pool boiling heat transfer

In order to study the influence of smooth heater size on the performance of pool boiling heat transfer, saturated boiling experiments of FC-72 were conducted on eight sized smooth silicon chips (the test surfaces are denoted as S5 ∼ S30, and their dimensionless dimensions are in the range of Lh/Lc = 7 ∼ 42). The bubble dynamics during boiling were observed with a high-speed camera. The results indicated that in natural convection regime, the heat transfer coefficient decreases with increasing heater size at a given heat flux. Besides, the larger surfaces enter into the nucleate boiling regime earlier. In a fully developed nucleate boiling regime, S5 surface displays the optimal heat transfer performance. The reality is that no vapor columns are produced even in the high heat flux range due to its small bubble departure diameter and high bubble departure frequency. For S8 ∼ S15 surfaces, the wall temperatures increase continuously with the increase of heat flux, while S20 ∼ S30 surfaces present a trend of first keeping constant and then increasing. The active nucleation site densities on S15 ∼ S30 surfaces follow the law of Na = 0.45q2 in the whole nucleate boiling regime. By contrast, for S5 ∼ S12 surfaces, the nucleation site densities abide by this law only in a fully developed nucleate boiling regime. In addition, the departure diameter and departure frequency of bubbles increase with the increase of heat flux for all surfaces. It is observed that with the increase of smooth heater size, CHF presents a changing trend of first increasing (in range of Lh ≤ λD, λD is the most dangerous Taylor wavelength), then decreasing (in range of λD < Lh ≤ 3λD) and finally leveling off (in range of Lh > 3λD). Accordingly, the CHF value maximizes at the heater size Lh = λD. In the segment of decreased CHF, the decline rate of CHF in range of λD < Lh ≤ 2λD is larger than that in range of 2λD < Lh ≤ 3λD. Meanwhile, the influencing mechanisms of heater size on CHF were explained by consideration into bubble patterns and behavior.

Natural convection heat transfer analysis of a nano-encapsulated phase change material (NEPCM) confined in a porous square chamber with two heat sources

This study focuses on the thermal behavior and performance of an insulated porous cavity containing square tubes, one as the hot block and the other as the cold block. The cavity is filled with a Nano-Encapsulated Phase Change Material (NEPCM) that combines the advantages of phase change materials (PCM) and nanofluid. The main purpose of this research is to examine the heat transfer mechanisms, phase change process, and overall thermal performance of the system. The governing equations of the NEPCM suspension are transformed into a dimensionless form and solved using the finite element approach. Various parameters, including nanoparticles concentration, fusion temperature, Darcy number, Rayleigh number, and Stefan number, are investigated to assess their impact on the system's thermal and dynamic performance. The results reveal that the optimal heat transfer performance is achieved when the NEPCM fusion temperature is set to 0.5. Moreover, a smaller Stefan number significantly improves the system's performance. Additionally, including NEPCM particles generally improves heat transfer, with a 33% increase in mean heat transfer observed when the volume fraction was raised from 0% to 2%.

A direct numerical simulation study to elucidate the enhancement of heat transfer for nucleate boiling on surfaces with micro-pillars

Recent experimental studies have demonstrated great potential of surface engineering in enhancing nucleate boiling heat transfer performance. However, the underlying mechanism remains unclear, especially the role of microlayer evaporation underneath bubbles. In this work, we investigate the heat transfer from microlayer evaporation underneath a growing bubble on micro-pillar arrayed surfaces using Direct Numerical Simulations (DNS). The evolution of the microlayer is reproduced in the DNS by considering a bubble growth driven by the local temperature gradient. The effects of micro-pillar structures on the microlayer profile and the heat transfer performance are systematically studied and analyzed. Our simulation results reveal three distinct microlayer morphologies related to micro-pillar structures: the undisturbed microlayer, the disturbed microlayer, and the disrupted microlayer. It can be further generalized as the greater the spacing and height of the micro-pillars, the more disrupted the microlayer becomes. Due to the reduction of microlayer thickness, more disruption means higher microlayer heat transfer coefficient. However, this accelerates microlayer depletion and thus reduces the overall heat transfer potential from microlayer evaporation during its life cycle in nucleate boiling. Based on these findings, a strategy is revealed for the design of micro-pillar arrayed surfaces to achieve optimal heat transfer performance of the microlayer.

Favorable trade-off between heat transfer and pressure drop in 3D printed baffled logpile catalyst structures

Additive manufacturing has the potential to unlock a large degree of geometric freedom in the shaping of catalytic material, thereby providing new possibilities to optimize catalyst holdup, pressure drop and heat and mass transfer characteristics. In this modelling study, baffled logpile structures are proposed as a promising candidate to exploit this potential, by shaping the catalytic material as a static mixer, generating cross-flow. An OpenFOAM Computational Fluid Dynamics study was performed on various 2D structure designs to map the trade-off between heat transfer, pressure drop and residence time distribution as a function of the design, length and gap spacing of the baffle. It is observed that structures with the longest baffles provide optimal heat transfer performance, and that the baffle gap spacing can be used to tailor the trade-off between heat transfer and pressure drop. In comparison to a packed bed filled with spherical particles, the novel structures offer a heat transfer rate four to six times as high at the same pressure drop. Whilst full 3D simulations, validated by experiments, remain to complete the analysis, the current work illustrates the potential of this novel class of structured catalyst materials for intensified chemical reactors.

Multi-objective optimization design of a plate–fin water evaporator in a low-pressure environment based on the MO-SHERPA algorithm

The aim of this study was to investigate the optimal heat transfer performance and structure of airborne water evaporators in low-pressure environments. To achieve this, a multi-objective optimization design was performed on a plate-fin water evaporator with offset strip fins, operating in a low-pressure environment at an absolute pressure of 7 kPa. The multi-objective Simultaneous Hybrid Exploration that is Robust, Progressive and Adaptive (SHERPA) (MO-SHERPA) algorithm was used to optimize the structure of the plate–fin water evaporator under a low pressure of 7 kPa. There are three conflicting optimization objectives, including comprehensive performance factor (JF), entropy generation (Sgen), and weight. The optimal Pareto frontier design points were compared and analyzed based on the three evaluation indicators. The optimization results showed that compared with the initial design point, the JF increased by 32.14% and the weight decreased by 48%, which indicated that the heat exchange performance has been significantly improved, and the weight of the heat exchanger was also reduced. The internal and external flow fields of the water evaporator, including temperature, pressure, and volume fraction of vapor, were qualitatively compared to prove the optimization effect. In addition, the influence of structural parameters on the heat transfer performance of the water evaporator were analyzed. The method proposed in this study can effectively optimize the fin structural design parameters, providing valuable guidance for the structural design of water evaporators operating in low-pressure environments.

Multi-objective optimisation of thermal and hydraulic performance with various concentrations of hybrid Fe3O4/graphene nanofluids in a microchannel heat sink

An accurate thermal resistance network and an objective function of pressure drop were established for a microchannel heat sink (MCHS) to evaluate the heat transfer and hydraulic performance of hybrid Fe3O4/graphene nanofluids with different concentrations in the MCHS. An optimised geometry of the MCHS with nanofluids at 40 °C under mass concentrations of 0.1, 0.3, and 0.5 wt% was obtained through multi-objective particle swarm optimisation (MOPSO). By using the K-means clustering method, a Pareto front of an optimised solution set was divided into three clustering zones with the optimal heat transfer performance, balanced performance, and optimal hydraulic performance. Combined with a comprehensive enhanced heat transfer factor, suggestions for trade-off in these three zones were proposed. The results indicate that the hybrid Fe3O4/graphene nanofluids is an attractive heat transfer medium in the MCHS and the optimised dimensions obtained by the MOPSO can simultaneously reduce the thermal resistance and pressure drop across the MCHS. Nanofluids with higher concentrations can amplify the effects of reducing the thermal resistance without increasing the loss of pumping power when the working temperature of the MCHS increases. The increase in the concentration of nanofluids can constantly decrease the dimensions of the MCHS in the Pareto front. Therefore, ensuring the stability of nanofluids can prevent the resulting pseudo-high concentration from influencing the accuracy of optimisation data.

Bubble Coalescence During Pool Boiling with Different Surface Characteristics

This work investigates the bubble dynamics during pool boiling from isolated cavities on a horizontal silicon substrate with three surface coatings, silicon oxide, Perfluorodecyltrichlorosilane, and silica nanoparticles. The experiments were conducted with FC-72 under various superheat degrees. In the case of vertical coalescence, coalescences at the boundary between the departed bubble and the subsequent nucleated bubble were observed and analyzed. Vertical coalescence appears with lower superheat degrees on Perfluorodecyltrichlorosilane coated surfaces due to the faster initial bubble growth than on the other surfaces. However, the vertical coalescence has a limited effect on the bubble departure diameter, which is nearly kept constant at around 0.6 mm. In the case of horizontal coalescence, bubbles growing on the silica nanoparticles coated surface are able to coalesce for small cavity spacing with lower superheat degree than other surfaces. Coalescing bubbles from cavities spaced 0.50 mm apart can detach just after coalescence, while for a shorter spacing of 0.25 mm, bubbles remain attached to the surface and further growth is needed for their detachment. Based on the analysis of energy removal per unit of area, the optimal heat transfer performance may be achieved when the cavities spacing is approximate to the single bubble departure diameter.

Investigation of impingement heat transfer in double-wall cooling structures with corrugated impingement plate at small Reynolds numbers

This study investigated an impingement/effusion cooling scheme with a corrugated jet plate applicable to a high-temperature exhaust cooling nozzle. The primary objective was to evaluate the effect of the corrugated wall jet on the impingement heat transfer and flow loss in a narrow target chamber under a low jet Reynolds number ranging from 450 to 2700. Sinusoidal curves with different wavelengths were employed to model the corrugated jet plate, where impingement holes were located at the trough of the plate. The location of the impingement holes creates a staggering pattern relative to the effusion holes. The corrugated jet plates with relative ripple amplitudes (A/d) considered were 0.2, 0.5 and 0.8, and the results were compared with those of a flat jet plate. Three relative average jet-to-plate spacings, Have/d, were selected for use: 1.6, 2.6, and 3.6. The heat transferred over a smooth target surface by multiple jets was measured using the transient liquid crystal (TLC) method. Results show that the amplitude of the corrugated plate was found to profoundly influence the heat transfer of the target surface. The use of corrugated jet plates yielded a better heat transfer performance at a relatively high Re and low impinging distance. The optimal heat transfer performance was achieved at A/d of 0.5, which was 28 % higher than the flat plate. The discharge coefficient for the corrugated jet plate was 2–9.4 % lower than that of the flat jet plate at all Have/d and Re values, and the minimum discharge coefficient was reached at A/d of 0.2. Furthermore, internal impingement heat transfer and flow resistance characteristics of the corrugated impingement/effusion were correlated according to the experimental data.

Multi-objective optimization of elliptical tube fin heat exchangers based on neural networks and genetic algorithm

The application of machine learning based on neural networks (NNs) and genetic algorithm (GA) in multi-objective optimization of heat exchangers is studied. Taking the tube fin heat exchanger (TFHE) as the research object, the inlet air velocity and the ellipticity of tubes are taken as the optimization variables. In order to obtain the optimal heat transfer performance and pressure drop performance, Computational Fluid Dynamics (CFD) simulation is carried out for different Reynolds based on the hydraulic diameter numbers (150–750) and tube ellipticity (0.2–1). Then use simulation data to train the Back-Propagation neural networks and establish the prediction model of heat transfer coefficient and pressure drop. The non-dominated multi-objective genetic algorithm with elitist retention strategy (NSGA-II) is used to optimize two prediction results of NNs. Finally, the optimal heat transfer coefficient and pressure drop are given in the form of Pareto front. The optimization results show that when the Reynolds number is 541 and the ellipticity is 0.34, the pressure drop of the TFHE decreases 20%, and the heat transfer coefficient is basically unchanged, whose j/f is 1.28 times as much as that of the original heat exchanger.