Design method of staggered serpentine anti-blocking heat exchange plate

Abstract

The present application relates to heat exchange plate technical field, especially interlaced serpentine anti-blocking heat exchange plate design method is provided.The design method provided by the present application includes the following steps: determining the geometric parameter set of the interlaced serpentine anti-blocking heat exchange plate, boundary constraint is carried out, and the feasible parameter space is constructed;Orthogonal test sampling is used to generate multiple groups of initial geometric parameter combinations;A CFD simulation model of phase change microcapsule suspension liquid-solid two-phase flow is established, the flow field and heat exchange coupling simulation are carried out for each group of geometric parameter combinations, the average flow velocity of the corner, the inlet and outlet pressure drop and the average comprehensive heat exchange coefficient of the heat exchange plate are extracted, and the initial sample data set is constructed;The Gaussian process regression agent model is trained, the multi-objective genetic algorithm is used to carry out global optimization in the parameter space, and the optimal candidate parameter combination is selected;The sampling, modeling, optimization and simulation verification steps are executed in a loop iteration until any one of the iteration conditions is met, and after the iteration is completed, the parameter combination with the optimal comprehensive performance is selected.

Description

Technical Field

[0001] This invention relates to the field of heat exchanger technology, and in particular to a design method for an interlaced serpentine anti-clogging heat exchanger. Background Technology

[0002] Phase change microcapsule suspensions, as a highly efficient latent heat transfer fluid, are increasingly widely used in energy-saving heat exchange fields such as power battery thermal management, high-density electronic device heat dissipation, central air conditioning energy storage cooling, and industrial low-temperature waste heat recovery. This type of fluid uniformly disperses phase change energy storage microcapsule particles in a base liquid, relying on the solid-liquid phase change process of the particles to absorb or release latent heat. It possesses advantages such as small temperature fluctuations, large equivalent specific heat capacity, and strong heat carrying capacity. During heat exchange, it can maintain near-constant temperature heat transfer characteristics, and its overall heat exchange efficiency is far superior to traditional single-phase water cooling media such as pure water and ethylene glycol, making it a core heat exchange medium in high-end thermal control systems.

[0003] Applying phase change microcapsule suspensions to conventional serpentine heat exchanger plates still presents the following inherent drawbacks: 1) Traditional serpentine heat exchanger channels often employ a straight, uniform width structure with right-angle corners. When the fluid flows through the right-angle corners, the flow direction changes abruptly, the flow velocity inside the channel decreases rapidly, and a large-scale low-velocity stagnation zone is easily formed. The outside of the channel is accompanied by obvious backflow vortices and secondary circulation structures, resulting in poor flow field uniformity; 2) The phase change microcapsule particles and the liquid base have a certain density difference, particle size dispersion, and surface adhesion characteristics. In the low-velocity stagnation zone, they are subjected to the coupling effects of gravity, drag, and inertial forces. Particle sedimentation, wall adsorption, and local accumulation at corners are very likely to occur. Under long-term continuous operation, the deposit layer gradually thickens, causing the effective flow cross section of the flow channel to continuously narrow. 3) Once particle accumulation and blockage occur at the corner of the flow channel, it will directly cause the system flow rate to decrease, the inlet and outlet pressure drop to increase, and the energy consumption of the circulating pump to increase. At the same time, the effective heat exchange area of ​​the heat exchange plate will be covered and the thermal resistance will increase, resulting in a significant decrease in the overall heat exchange performance. This can easily induce hidden dangers such as local thermal runaway of the battery and high temperature frequency reduction of electronic devices. In severe cases, it can also disturb the flow state of the pipeline, impact the operating stability of the pump body, and significantly shorten the service life of the entire heat exchange equipment.

[0004] Currently, the industry has developed some variable cross-section flow channel optimization solutions to address corner blockage issues. However, existing improvement methods mostly employ passive structural designs such as overall diameter reduction, local throttling protrusions, and segmented widening, which generally have significant limitations: First, the addition of additional shrinkage components leads to complex flow channel structures, difficult mold forming, and high machining costs, which is not conducive to large-scale mass production. Second, artificially abrupt changes in local cross-sections can easily cause additional local resistance losses, resulting in a surge in system pressure drop and increased cycle energy consumption, leading to poor economic efficiency. Third, existing variable cross-section solutions only make local size adjustments without addressing the overall flow channel layout logic. They cannot form a naturally tapering and accelerating flow channel through staggered straight sections, nor can they weaken the low-speed dead zone at the corner from the flow field mechanism level. This makes it difficult to specifically address the industry pain points of phase change microcapsule suspensions, such as easy sedimentation, easy adhesion, and easy corner blockage.

[0005] Therefore, it is very important to provide a method for designing a novel serpentine heat exchanger structure that is simple in structure, requires no additional throttling components, achieves natural narrowing acceleration by relying on the staggered layout of the flow channels, can suppress particle deposition from the source, and is suitable for engineering processing and manufacturing, in order to address the solid-liquid two-phase flow characteristics of phase change microcapsule suspensions. Summary of the Invention

[0006] This invention provides a design method for an interlaced serpentine anti-clogging heat exchange plate to solve the problems of existing traditional serpentine heat exchange plates, such as the easy generation of flow dead zones at right-angle corners, the easy sedimentation and accumulation of microcapsule particles causing flow channel blockage, and the shortcomings of traditional design relying on empirical trial and error, lack of quantitative standards for parameters, low simulation iteration efficiency, and difficulty in synergistically considering multiple performance objectives.

[0007] In a first aspect, the present invention provides a design method for an interlaced serpentine anti-clogging heat exchange plate, comprising the following steps:

[0008] S101, the set of geometric parameters of the staggered serpentine anti-clogging heat exchange plate is determined according to the structural design elements of the heat exchange plate. Based on the flow channel processing technology, flow section limitation and system pressure drop limitation, boundary constraints are applied to each geometric parameter to construct a feasible parameter space.

[0009] S102, within the feasible parameter space, multiple sets of initial geometric parameter combinations are generated by orthogonal experimental sampling;

[0010] S103, establish a CFD simulation model of solid-liquid two-phase flow of phase change microcapsule suspension, perform flow field and heat transfer coupling simulation for each set of geometric parameter combinations, extract the average flow velocity at corners, inlet and outlet pressure drop and average comprehensive heat transfer coefficient performance index of heat exchange plate, and construct the initial sample dataset;

[0011] S104: A Gaussian process regression surrogate model is trained based on the initial sample dataset. A multi-objective genetic algorithm is used to perform global optimization in the parameter space and screen the optimal candidate parameter combination.

[0012] S105, iteratively execute the sampling, modeling, optimization and simulation verification steps until any one of the iteration conditions is met;

[0013] S106 After the iteration is completed, the parameter combination with the best overall performance is selected from the final sample set. This parameter combination is the structural design parameter of the heat exchange plate.

[0014] In some embodiments, the set of geometric parameters is {B, h, R, W, α}, where B is the width of the straight section of the flow channel, h is the internal height of the flow channel, R is the radius of the rounded corner of the turning section, W is the lateral offset width of adjacent straight sections, and α is the staggered tilt angle of the flow channels.

[0015] In some embodiments, the boundary constraints specifically include: 0.5B≤R≤B, 0.5B≤W≤0.75B, 5°≤α≤25°, minimum wall thickness between adjacent channels not less than 1.5mm, minimum equivalent flow width at the tapered constriction not less than 0.25B, and the total inlet and outlet pressure drop increase of the optimized structure relative to the traditional serpentine channel of the same size not greater than 15%.

[0016] In some embodiments, the Gaussian process regression surrogate model is constructed using a covariance kernel function, and the input geometric parameters and output performance indicators are preprocessed by normalization. Cross-validation is used to verify the model accuracy, thereby meeting the engineering prediction requirements for flow and heat transfer performance in the flow channel.

[0017] In some embodiments, in step S104, the Gaussian process regression surrogate model takes x=[α,W,R] as input and y=[v,Δp,K] as output, where v is the average flow velocity at the corner, Δp is the total pressure drop at the inlet and outlet of the flow channel, and K is the average comprehensive heat transfer coefficient of the heat exchanger plate; it satisfies f(x)~GP(m(x),k(x,x′)), where m(x) is the mean function, i.e., the prior average level of the function, x represents one set of samples, x′ represents another set of samples, and the observation relationship is y=f(x)+ε, where ε~N (0,σn²), where σn² is the signal variance, and a GPR model is constructed using the squared exponential kernel function k(x,x′)=σf²exp[-1 / 2·(xx′)ᵀΛ⁻¹(xx′)], where σf² is the signal variance and Λ is the length scaling matrix. The predicted mean and prediction uncertainty are output by μ(x*)=k*ᵀ(K+σn²I)⁻¹y and σ²(x*)=k(x*,x*)-k*ᵀ(K+σn²I)⁻¹k*, respectively, where I is the identity matrix and x* is the prediction parameter. The candidate parameter combination is substituted into the CFD simulation model for simulation verification, and the verification results are incorporated into the sample set to dynamically update the surrogate model.

[0018] In some embodiments, the specific process of using a multi-objective genetic algorithm to perform global optimization in the parameter space is as follows: A multi-objective comprehensive evaluation function is constructed, with the sub-optimization objectives being maximizing the average flow velocity at the corner, maximizing the average comprehensive heat transfer coefficient of the heat exchanger plate, and minimizing the total pressure drop at the inlet and outlet of the flow channel, respectively. The expression of the evaluation function is F(x)=w1·v(x) / v0+w2·K(x) / K0-w3·Δp(x) / Δp0, where x=[α,W,R], v (x) represents the average flow velocity at the corner, K(x) represents the average comprehensive heat transfer coefficient of the heat exchange plate, Δp(x) represents the total pressure drop at the inlet and outlet of the flow channel, v0, K0, and Δp0 are the baseline values ​​of the traditional straight right-angle serpentine flow channel, w1 is the anti-clogging weight coefficient, w2 is the heat transfer weight coefficient, and w3 is the low pressure drop weight coefficient, and w1+w2+w3=1; the weights of each sub-objective can be adaptively adjusted according to different heat exchange application scenarios; by using the evaluation function to perform global optimization within the preset constraint range, a reasonable range of values ​​for the staggered inclination angle of the flow channel suitable for the anti-clogging and low pressure drop conditions of the phase change microcapsule suspension is obtained.

[0019] In some embodiments, the iteration conditions include any one of the following conditions: (1) the heat exchange anti-blocking and flow pressure drop performance reaches the preset design threshold, the preset design threshold includes: the average flow velocity at the corner is increased by no less than 15% compared with the traditional straight right-angle serpentine flow channel, the average comprehensive heat transfer coefficient of the heat exchange plate is increased by no less than 10%, and the total pressure drop at the inlet and outlet of the flow channel is increased by no more than 15%; (2) in three consecutive iterations, the relative improvement of the multi-objective comprehensive evaluation function F(x) is less than 1%; (3) the number of simulation calculations reaches the preset iteration limit, the preset iteration limit is that the number of CFD verification simulations does not exceed 60 times, or the number of iterations of the multi-objective genetic algorithm does not exceed 120 generations.

[0020] In some embodiments, when the number of iterations reaches a preset upper limit, if the prediction variance of the surrogate model for the optimal parameter combination exceeds a set threshold, incremental sample simulation is automatically started to supplement the verification, and the optimal parameter combination is re-selected after updating the sample set and the surrogate model.

[0021] Secondly, the present invention provides an interlaced serpentine anti-clogging heat exchange plate designed using the above-mentioned design method. The heat exchange plate includes a substrate and interlaced serpentine flow channels formed on the surface of the substrate. The straight sections of the flow channels are arranged according to the optimized interlacing angle and misalignment width, and the turning sections adopt a smooth transition with matching rounded corner radii. The flow channel cross-sections are of equal width and height and the inner wall is smooth. By relying on the interlacing and misalignment to form a natural gradual narrowing, the flow velocity at the corner is increased, and the sedimentation of phase change microcapsule particles and flow channel blockage are suppressed.

[0022] Thirdly, the present invention provides a heat exchanger equipped with at least one of the above-mentioned staggered serpentine anti-clogging heat exchange plates. The heat exchanger is applied to heat exchange scenarios of phase change microcapsule suspensions for power battery thermal management, electronic device heat dissipation, air conditioning energy storage, and industrial waste heat recovery.

[0023] This invention defines key geometric parameters of the heat exchange plate and constructs a feasible parameter space by combining processing technology and flow pressure drop requirements. It establishes a sample set using orthogonal sampling and extracts multi-dimensional performance indicators based on CFD coupled simulation of solid-liquid two-phase flow. A Gaussian process regression model is introduced to construct a surrogate model, coupled with a multi-objective genetic algorithm for global parameter optimization. The optimal geometric parameters for suitable operating conditions are solved through multiple iterative termination conditions and incremental supplementation verification mechanisms. By relying on a naturally gradual narrowing structure with rounded corners formed by staggered positioning, the low-velocity dead zone is weakened and the wall shear scouring capacity is improved from the flow field mechanism. This achieves anti-clogging, low flow resistance, and high heat transfer synergistic optimization under controllable pressure drop increase, providing a standardized and replicable design scheme for serpentine heat exchange plates specifically for phase change microcapsule suspensions.

[0024] The beneficial effects of the technical solution provided by this invention include:

[0025] 1. This invention introduces a Gaussian process regression surrogate model to replace massive CFD repetitive simulations, and combines it with a multi-objective genetic algorithm for global optimization. It gets rid of the limitations of traditional empirical trial and error and single-point simulation iteration, greatly reduces the design cycle, and provides rigorous quantitative basis for parameter matching. It achieves multi-objective synergistic optimization of anti-blocking, pressure drop and heat exchange, and overcomes the drawbacks of traditional design.

[0026] 2. In the early design stage, the present invention is subject to dual constraints of processing technology and flow pressure drop. The optimized geometric parameters are all within the manufacturable and low-energy consumption range. The structure does not require complex irregular processing. It can be formed by conventional sheet metal stamping and mold grooving. It has strong process adaptability and is conducive to industrial mass production.

[0027] 3. The design method of this invention has strong versatility. It only requires fine-tuning of the parameter constraint range, simulation boundary conditions and target weights to adapt to phase change microcapsule suspensions with different particle sizes and mass fractions. It is compatible with various scenarios such as power batteries, electronic heat dissipation and waste heat recovery, and can form an industry-standard design process. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1a This is an exploded view of the heat exchange plate provided in an embodiment of this application;

[0030] Figure 1b An overall view of the heat exchange plate provided in an embodiment of this application;

[0031] Figure 2a This is a top view of the bottom plate structure of the optimized staggered serpentine anti-clogging heat exchange plate of the present invention.

[0032] Figure 2b A top view of the bottom plate structure of a traditional flat, right-angled serpentine heat exchange plate;

[0033] Figure 3a This is a cloud map showing the corner velocity distribution of the optimized heat exchange plate according to the present invention.

[0034] Figure 3b This is a corner velocity distribution cloud map for a traditional heat exchanger plate.

[0035] Figure 4a This is a vector distribution diagram of the angular velocity of the optimized heat exchange plate according to the present invention;

[0036] Figure 4b This is a vector distribution diagram of the angular velocity of a traditional heat exchanger. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] This invention provides a design method for an interlaced serpentine anti-clogging heat exchange plate, which can solve the problems of existing traditional serpentine heat exchange plates, such as the easy generation of flow dead zones at right-angle corners, the easy sedimentation and accumulation of microcapsule particles causing flow channel blockage, and the shortcomings of traditional design relying on empirical trial and error, lack of quantitative standard parameters, low simulation iteration efficiency, and difficulty in synergistically considering multiple performance objectives.

[0039] The design method of the staggered serpentine anti-clogging heat exchange plate provided by the present invention includes the following steps:

[0040] Step S101, Geometric parameter definition and feasible parameter space construction:

[0041] Based on the structural design elements of the heat exchange plate, the core geometric parameter set {B, h, R, W, α} of the staggered serpentine anti-clogging heat exchange plate is determined, where B is the width of the straight section of the flow channel, h is the internal height of the flow channel, R is the radius of the rounded corner of the turning section, W is the lateral misalignment width of adjacent straight sections, and α is the staggered tilt angle of the flow channel.

[0042] The structural topology is designed according to fixed constraint logic: adjacent straight sections of the heat exchange plate are staggered along the width direction, and the staggered width is always less than the width of the flow channel itself to avoid flow blockage caused by excessive narrowing; the flow channel adopts a rectangular equal cross-section design, and the width and height of the straight sections are consistent throughout the entire process. The natural gradual narrowing is formed only by the staggered angle of the straight sections and the lateral offset, without the need to add an additional irregular throttling structure.

[0043] By considering the processing limits of aluminum alloy sheet stamping and die grooving, as well as the energy consumption of the circulating water pump and the system's allowable pressure drop threshold, upper and lower boundary constraints are set for each geometric parameter to construct a closed and controllable feasible parameter space. The width B of the straight section of the flow channel is limited to 4mm≤B≤12mm, the internal height h of the flow channel is limited to 2mm≤h≤8mm, the radius R of the corner section is limited to 0.5B≤R≤B, the lateral misalignment width W of adjacent straight sections is limited to 0.5B≤W≤0.75B, and the staggered inclination angle α of the flow channels is limited to 5°≤α≤25°. The minimum wall thickness between adjacent flow channels is not less than 1.5mm, the minimum equivalent flow width at the tapered end is not less than 0.25B, and the total inlet and outlet pressure drop increase of the optimized structure compared to the traditional serpentine flow channel of the same external dimensions is not greater than 15%. Through the above boundary constraints, invalid parameter combinations that exceed processing capacity, cause excessive pressure drop, or result in flow failure can be eliminated from the source.

[0044] Step S102, Orthogonal experimental sampling and multiphysics sample set construction

[0045] Within the defined feasible parameter space, orthogonal experimental design is used to conduct multi-factor, multi-level uniform sampling to obtain multiple sets of initial geometric parameter combinations. This approach uses the minimum number of samples to achieve high coverage and characterize the high-dimensional parameter space distribution characteristics, avoiding the drawbacks of random sampling redundancy and local omissions.

[0046] Step S103: Establish a simulation model

[0047] For each set of geometric parameter combinations obtained from sampling, a parameterized simulation model of solid-liquid two-phase flow of phase change microcapsule suspension is established. The Euler two-phase flow model is selected to adapt to solid-liquid two-phase flow, and the RNG k-ε turbulence model is used to adapt to strong swirling flow and backflow conditions at corners. The velocity inlet, pressure outlet and wall thermal boundary are set to match the actual working fluid physical property parameters.

[0048] After the simulation converges, three core performance indicators are extracted: average flow velocity at corners, total pressure drop at the inlet and outlet of the flow channel, and average comprehensive heat transfer coefficient of the heat exchange plate. A one-to-one mapping relationship between geometric input parameters and flow, anti-clogging, and heat transfer output indicators is established to form a standardized initial sample dataset, which provides reliable data support for subsequent surrogate model training.

[0049] Step S104: Construction and accuracy verification of the Gaussian process regression surrogate model

[0050] This invention selects Gaussian process regression algorithm as the core of the surrogate model, which is suitable for fitting the parameter performance characteristics of heat exchanger plates with small samples and strong nonlinearity. Let the input variable be x=[α,W,R], and the output variable be y=[v,Δp,K], where v is the average flow velocity at the corner, Δp is the total pressure drop at the inlet and outlet of the flow channel, and K is the average comprehensive heat transfer coefficient of the heat exchanger plate; the surrogate model satisfies f(x)~GP(m(x),k(x,x′)), where m(x) is the mean function, i.e., the prior average level of the function, x represents one set of samples, x′ represents another set of samples, and the observation relationship is y=f(x)+ε, ε~N(0,σn²), where σn² is the signal variance. The GPR model is constructed using the squared exponential kernel function k(x,x′)=σf²exp[-1 / 2·(xx′)ᵀΛ⁻¹(xx′)], where σf² is the signal variance and Λ is the length scale matrix. For the prediction parameter x*, maximum likelihood estimation is used to optimize the hyperparameters, and the predicted mean and prediction uncertainty are output by μ(x*)=k*ᵀ(K+σn²I)⁻¹y and σ²(x*)=k(x*,x*)-k*ᵀ(K+σn²I)⁻¹k*, respectively, where I is the identity matrix. Normalization preprocessing is performed on the input geometric parameters and output performance indicators of the sample set to eliminate fitting bias caused by different dimensions. Cross-validation is performed by dividing the training and test sets to optimize the model hyperparameters and complete the accuracy verification. The successfully trained surrogate model can quickly predict the comprehensive performance corresponding to any combination of parameters, replacing the traditional time-consuming and labor-intensive repetitive CFD simulations, significantly shortening the optimization iteration cycle and reducing computational resource consumption.

[0051] A multi-objective comprehensive evaluation function is established, with three core sub-objectives: maximizing the average flow velocity at the corner, maximizing the average comprehensive heat transfer coefficient of the heat exchanger plate, and minimizing the total pressure drop at the inlet and outlet of the flow channel. The comprehensive objective function is expressed as F(x)=w1·v(x) / v0+w2·K(x) / K0-w3·Δp(x) / Δp0, where x=[α,W,R], v(x) is the average flow velocity at the corner, K(x) is the average comprehensive heat transfer coefficient of the heat exchanger plate, Δp(x) is the total pressure drop at the inlet and outlet of the flow channel, v0, K0, and Δp0 are the baseline values ​​of the traditional straight right-angle serpentine flow channel under the same operating conditions, and w1, w2, and w3 are the anti-clogging weight coefficient, heat transfer weight coefficient, and low pressure drop weight coefficient, respectively, and satisfy w1+w2+w3=1.

[0052] The trained Gaussian process regression surrogate model is used as the fitness evaluation carrier. A multi-objective genetic algorithm is coupled to carry out global iterative optimization in the feasible parameter space. Through selection, crossover, and mutation operations, the model continuously iterates and evolves to screen the candidate parameter combination with the best comprehensive benefit. The candidate parameters are substituted into the CFD simulation model to carry out real-world verification. The verified parameter-performance samples are incorporated into the dataset to dynamically update the surrogate model, forming a closed-loop optimization process of sampling, modeling, optimization, and verification.

[0053] Step S105: Multiple Iteration Termination Conditions and Incremental Verification

[0054] The sampling, modeling, optimization, and simulation verification steps are executed iteratively until any one of the iteration conditions is met. This invention sets three complementary iteration termination criteria:

[0055] (1) Performance Target Achievement Termination: The iteration stops when the heat exchange anti-clogging index and the flow pressure drop index reach the preset design threshold. The preset design threshold is that the average flow velocity at the corner is increased by no less than 15% compared with the traditional straight right-angle serpentine flow channel, the average comprehensive heat transfer coefficient of the heat exchange plate is increased by no less than 10%, and the total pressure drop at the inlet and outlet of the flow channel is increased by no more than 15%. After meeting the above indicators, there is no need to continue the iteration, and the optimal parameters are directly output.

[0056] (2) Iteration convergence termination: When the relative improvement of the multi-objective comprehensive evaluation function F(x) is less than 1% in three consecutive iterations, it is determined that the parameters have approached the global optimum and the invalid iteration is stopped;

[0057] (3) Simulation upper limit termination: When the number of CFD verification simulations reaches 60, or the number of iterations of the multi-objective genetic algorithm reaches 120, it is determined that the preset iteration upper limit has been reached; if the prediction variance of the surrogate model for the optimal parameter combination is within a reasonable threshold range, the result is directly output; if the prediction variance exceeds the threshold, the incremental sample supplementation simulation is automatically started to expand the sample set, retrain the model, and re-select the optimal parameters to ensure that the optimization results are stable and reliable.

[0058] Step S106: Optimal parameter matching and heat exchanger plate structure finalization

[0059] After the iteration, the parameter combination with the best multi-objective comprehensive performance is selected from the final sample set. The overall layout design of the flow channel is carried out based on the optimal flow channel staggered tilt angle, the lateral misalignment width of adjacent straight sections, and the radius of the turning corner. The straight sections of the flow channel are arranged in an staggered manner with the optimal tilt angle, and the turning sections adopt a smooth transition with a matching radius of the turning corner, with no abrupt changes in cross-section or step protrusions throughout the entire process. The inner wall of the flow channel is polished to reduce the particle adhesion coefficient and enhance the wall shear scouring effect.

[0060] The staggered serpentine heat exchange plate designed in this invention consists of a base plate and a single continuous serpentine flow channel. The flow channel is composed of multiple sets of straight sections and rounded corner sections that are alternately connected. Adjacent straight sections are staggered along the width direction, naturally forming a gradually narrowing structure before entering the corner section. This passively forces the fluid to accelerate before entering the corner, weakening the low-speed stagnation zone and backflow vortex. From the flow mechanism, it inhibits the sedimentation, adhesion, and corner accumulation of microcapsule particles. Without increasing the structural complexity and flow pressure drop, it simultaneously achieves anti-clogging, low flow resistance, and high heat transfer performance.

[0061] This invention employs an integrated design process encompassing parameter constraint definition, orthogonal sample sampling, multiphysics coupled simulation, machine learning surrogate modeling, multi-objective iterative optimization, and convergence verification. First, the core geometric parameters of the heat exchanger are identified, and feasible parameter ranges are defined based on molding process, flow dimensions, and pump pressure drop constraints, eliminating invalid parameter combinations. Initial parameter samples are generated through uniform sampling via orthogonal experiments, and a CFD simulation model of the phase change microcapsule suspension solid-liquid two-phase flow is built. Simultaneously, flow, anti-clogging, and heat transfer performance indicators are extracted, constructing a parameter-performance mapping dataset. A Gaussian process regression algorithm is used to train the surrogate model, replacing costly repetitive CFD simulations. A coupled multi-objective genetic algorithm is used for global optimization within the constraint space, screening high-yield candidate parameters and verifying them through simulation, dynamically updating the sample library and model. Multiple termination conditions are set, including performance attainment, iterative convergence, and simulation upper limit; iteration stops when any condition is met. Finally, the optimal parameters are selected to complete the heat exchanger channel layout and structural design.

[0062] In one embodiment, the phase change microcapsule suspension base liquid is an aqueous solution of ethylene glycol, the microcapsule particles have an average particle size of 20 μm and a mass fraction of 20%; the heat exchange plate substrate is made of aluminum alloy, the standard straight section width of the flow channel is B=8 mm, and the flow channel height is h=5 mm; the design variable constraint range is set as follows:

[0063] The staggered tilt angles α are 5°, 10°, 15°, 20°, and 25°.

[0064] Lateral misalignment width W: 4 mm, 4.5 mm, 5 mm, 5.5 mm, 6 mm;

[0065] Corner radius R: 4 mm, 5 mm, 6 mm, 7 mm, 8 mm.

[0066] Orthogonal experimental sample construction:

[0067] A three-factor, five-level orthogonal array L25 (5³) was used to generate 25 sets of uniformly distributed geometric parameter combinations, covering the staggered tilt angle of the flow channel, the lateral misalignment width of adjacent straight sections, and the radius of the rounded corners of the turning sections, to ensure the representativeness and uniformity of the samples.

[0068] CFD numerical simulation dataset creation:

[0069] Three-dimensional flow channel geometric models were established for each of the 25 sample groups, using structured hexahedral meshes with a mesh quality greater than 0.92. The Eulerian two-phase flow model and the RNG k-ε turbulence model were selected, with the inlet velocity set to 0.5 m / s, the outlet at atmospheric pressure, and the wall at constant heat flow. After the simulation converged, the corner average velocity, the total pressure drop at the inlet and outlet of the flow channel, and the average comprehensive heat transfer coefficient of the heat exchanger plate were extracted for each sample group to construct a complete dataset of 25 input and output samples.

[0070] Training a Gaussian process regression machine learning model:

[0071] The 25 datasets were divided into 18 training sets and 7 test sets in a 7:3 ratio; the input variables α, W, and R, and the output performance metrics v, Δp, and K were normalized; and the quadratic exponential kernel function k(x,x′)=σf was selected. 2 The GPR model is constructed using exp[-1 / 2·(xx′)ᵀΛ⁻¹(xx′)], and the hyperparameters are optimized using maximum likelihood estimation. The predicted mean and prediction uncertainty are output by μ(x*)=k*ᵀ(K+σn²I)⁻¹y and σ²(x*)=k(x*,x*)-k*ᵀ(K+σn²I)⁻¹k*, respectively. After training, the coefficient of determination R² on the test set is 0.986, and the average relative error is 1.7%. The model fitting accuracy meets the engineering prediction requirements and can replace CFD simulation for fast prediction performance.

[0072] NSGA-II Multi-Objective Algorithm Optimization:

[0073] The NSGA-II algorithm is set with a population size of 80, a crossover probability of 0.8, a mutation probability of 0.15, and a maximum number of iterations of 120 generations. The comprehensive evaluation function is F(x) = w1·v(x) / v0 + w2·K(x) / K0 - w3·Δp(x) / Δp0. Multiple objectives are considered, including anti-clogging high shear force, low flow pressure drop, and high heat transfer coefficient. Global iterative optimization is performed within the constrained interval. The optimal weights for the power battery thermal management operating conditions are w1 = 0.40, w2 = 0.35, and w3 = 0.25. The Pareto optimal solution set is output, and the compromise optimal parameters are selected based on the actual operating conditions of the power battery thermal management.

[0074] Optimal stagger angle α opt =24°, optimal lateral misalignment width W opt =5.8mm, optimal corner radius R opt =7.2mm.

[0075] Heat exchanger plate structure finalization and simulation verification:

[0076] The straight sections of the serpentine flow channel are arranged at the optimal staggered angle of 24°, with a lateral offset width of 5.8mm, a turning radius of 7.2mm, a rectangular flow channel width of 8mm and a height of 5mm, and the inner wall of the flow channel is polished smooth to complete the heat exchange plate structure.

[0077] The exploded view and overall view of the staggered serpentine anti-clogging heat exchange plate designed in this invention are shown below. Figure 1a and Figure 1b , Figure 1a In the diagram, 1 is the upper cover plate, 2 is the sealing gasket, and 3 is the lower base plate. Figure 1bIn the diagram, 101 is the flow channel inlet, and 102 is the flow channel outlet.

[0078] A top view of the bottom plate of the optimized staggered serpentine anti-clogging heat exchange plate of this invention is shown below. Figure 2a A top view of the bottom plate structure of a traditional flat, right-angled serpentine heat exchanger is shown below. Figure 2b .

[0079] Under the same operating conditions, the corner velocity distribution cloud map of the optimized heat exchanger plate designed in this invention is shown below. Figure 3a The corner velocity distribution cloud map of a traditional heat exchanger plate is shown below. Figure 3b The optimized angular velocity vector distribution diagram of the heat exchange plate is shown in the figure. Figure 4a The angular velocity vector distribution diagram of a traditional heat exchanger plate is shown below. Figure 4b Simulation results show that, compared with traditional heat exchange plates, the staggered serpentine anti-clogging heat exchange plate designed in this invention increases the average flow velocity at the corners by 22.3%; the total pressure drop at the inlet and outlet of the flow channel increases by only 11.2%, which is a controllable increase; and the average comprehensive heat transfer coefficient increases by 15.7%.

[0080] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment / mode or example is included in at least one embodiment / mode or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0081] It should be noted that in this invention, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise expressly specified.

[0082] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A design method of staggered serpentine anti-blocking heat exchange plate, characterized in that, Includes the following steps: S101, determine the set of geometric parameters of the staggered serpentine anti-clogging heat exchange plate according to the structural design elements of the heat exchange plate, and impose boundary constraints on each geometric parameter to construct a feasible parameter space; S102, orthogonal experimental sampling is used to generate multiple sets of initial geometric parameter combinations in the feasible parameter space; S103, establish a CFD simulation model of solid-liquid two-phase flow of phase change microcapsule suspension, perform coupled simulation of flow field and heat transfer for each set of geometric parameter combinations, extract the average flow velocity at corners, pressure drop at inlet and outlet of flow channel and average comprehensive heat transfer coefficient performance index of heat exchange plate, and construct the initial sample dataset; S104: A Gaussian process regression surrogate model is trained based on the initial sample dataset. A multi-objective genetic algorithm is used to perform global optimization in the parameter space and select the optimal candidate parameter combination. S105, iteratively execute the sampling, modeling, optimization and simulation verification steps until the iteration conditions are met; S106 After the iteration is completed, the parameter combination with the best overall performance is selected from the final sample set. This parameter combination is the structural design parameter of the heat exchange plate.

2. The design method of staggered serpentine anti-blocking heat transfer plate according to claim 1, characterized in that, The set of geometric parameters is {B, h, R, W, α}, where B is the width of the straight section of the flow channel, h is the internal height of the flow channel, R is the radius of the rounded corner of the turning section, W is the lateral offset width of adjacent straight sections, and α is the staggered tilt angle of the flow channel.

3. The method of designing staggered serpentine anti-blocking heat transfer plates according to claim 2, characterized in that, Based on the flow channel processing technology, flow cross-section limitations, and system pressure drop limits, boundary constraints are imposed on each geometric parameter. The specific boundary constraints include: 0.5B≤R≤B, 0.5B≤W≤0.75B, 5°≤α≤25°, minimum wall thickness between adjacent flow channels not less than 1.5mm, minimum equivalent flow width at the tapered constriction not less than 0.25B, and the total inlet and outlet pressure drop increase of the optimized structure compared to the traditional serpentine flow channel of the same size not exceeding 15%.

4. The design method of the staggered serpentine anti-clogging heat exchange plate according to claim 1, characterized in that, In step S104, the Gaussian process regression surrogate model takes x=[α,W,R] as input and y=[v,Δp,K] as output, where v is the average flow velocity at the corner, Δp is the total pressure drop at the inlet and outlet of the flow channel, and K is the average comprehensive heat transfer coefficient of the heat exchanger plate; the surrogate model satisfies f(x)~GP(m(x),k(x,x′)), where m(x) is the mean function. x represents one set of samples, x' represents another set of samples, and the observation relationship is y=f(x)+ε, ε~N(0,σn²), where σn² is the signal variance. The GPR model is constructed using the squared exponential kernel function k(x,x′)=σf²exp[-1 / 2·(xx′)ᵀΛ⁻¹(xx′)], where σf² is the signal variance, Λ is the length scale matrix, and the prediction mean and prediction uncertainty are output by μ(x*)=k*ᵀ(K+σn²I)⁻¹y and σ²(x*)=k(x*,x*)-k*ᵀ(K+σn²I)⁻¹k* respectively, where I is the identity matrix.

5. The design method of the staggered serpentine anti-clogging heat exchange plate according to claim 1, characterized in that, The specific process of global optimization in the parameter space using a multi-objective genetic algorithm is as follows: A multi-objective comprehensive evaluation function is constructed, with the sub-optimization objectives being maximizing the average flow velocity at the corner, maximizing the average comprehensive heat transfer coefficient of the heat exchanger plate, and minimizing the total pressure drop at the inlet and outlet of the flow channel. The expression of the evaluation function is F(x) = w1·v(x) / v0 + w2·K(x) / K0 - w3·Δp(x) / Δp0, where x = [α, W, R], v (x) represents the average flow velocity at the corner, K(x) represents the average comprehensive heat transfer coefficient of the heat exchange plate, Δp(x) represents the total pressure drop at the inlet and outlet of the flow channel, v0, K0, and Δp0 are the reference values ​​of the traditional straight right-angle serpentine flow channel, w1 is the anti-clogging weight coefficient, w2 is the heat transfer weight coefficient, w3 is the low pressure drop weight coefficient, and w1+w2+w3=1; by using the evaluation function to perform global optimization within the preset constraint range, the range of values ​​for the staggered inclination angle of the flow channel suitable for the anti-clogging and low pressure drop conditions of the phase change microcapsule suspension is obtained.

6. The design method of the staggered serpentine anti-clogging heat exchange plate according to claim 5, characterized in that, The iteration conditions include any one of the following conditions: (1) The heat exchange anti-blocking and flow pressure drop performance reaches the preset design threshold, the preset design threshold includes: the average flow velocity at the corner is increased by no less than 15% compared with the traditional straight right-angle serpentine flow channel, the average comprehensive heat transfer coefficient of the heat exchange plate is increased by no less than 10%, and the total pressure drop at the inlet and outlet of the flow channel is increased by no more than 15%; (2) In three consecutive iterations, the relative improvement of the multi-objective comprehensive evaluation function F(x) is less than 1%; (3) The number of simulation calculations reaches the preset iteration limit, the preset iteration limit is that the number of CFD verification simulations does not exceed 60 times, or the number of iterations of the multi-objective genetic algorithm does not exceed 120 generations.

7. A heat exchanger, characterized in that, The heat exchanger includes at least one staggered serpentine anti-clogging heat exchange plate designed using the design method described in any one of claims 1-6.

Images