Liquid cooling plate flow resistance characteristic evaluation method and device, medium and equipment

Abstract

The invention relates to the technical field of thermal management, in particular to a liquid cooling plate flow resistance characteristic evaluation method and device, a medium and equipment. According to the method, the calculation load is remarkably reduced by converting the flow data into the reduced-order model by utilizing the intrinsic orthogonal decomposition. This can achieve faster analysis without sacrificing accuracy, so that the design can be evaluated and optimized in real time. Moreover, different from the traditional method which may depend on a simplified model or empirical correlation, the method based on the intrinsic orthogonal decomposition can capture a main flow mode, namely a dominant mode, which causes resistance. These modes reveal critical flow characteristics, such as high shear stress, vortex formation and flow separation, which are crucial for understanding the root cause of flow resistance. The capability of quantifying and isolating the important flow structure can deeper understand how the specific flow behavior affects the resistance, so that guidance is provided for the optimization design of the geometrical shape of the liquid cooling plate.

Description

Technical Field

[0001] This invention relates to the field of thermal management technology, specifically to a method, apparatus, medium, and equipment for evaluating the flow resistance characteristics of a liquid cooling plate. Background Technology

[0002] Liquid cooling systems are widely used in a variety of applications to manage heat loads in high-performance equipment such as electronic devices, power electronic devices, automotive engines, and heat exchangers. These systems typically use liquid cooling plates to facilitate heat transfer from the source (e.g., electronic components or engines) to the coolant, which then circulates to dissipate the heat. The performance of these systems depends heavily on the flow resistance of the liquid within the cooling plates, which in turn depends on the design of the channels, the properties of the coolant, and the flow characteristics within the plates.

[0003] Flow resistance is typically associated with frictional losses in cooling plate channels, which impede coolant flow, leading to higher pressure drops and reduced cooling efficiency. Optimizing flow resistance is crucial for improving the overall thermal performance of cooling systems, reducing energy consumption, and minimizing pump power requirements. However, current flow resistance analysis methods often fail to accurately capture the impact of complex flow characteristics on flow resistance. Summary of the Invention

[0004] In view of this, the present invention provides a method, apparatus, medium and equipment for evaluating the flow resistance characteristics of a liquid cooling plate, in order to solve the problem that the flow resistance analysis in the prior art cannot accurately capture the influence of complex flow characteristics on flow resistance.

[0005] In a first aspect, the present invention provides a method for evaluating the flow resistance characteristics of a liquid cooling plate. The method includes: acquiring fluid flow data of the liquid cooling plate, constructing a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions; extracting the dominant mode of fluid motion within the liquid cooling plate by intrinsic orthogonal decomposition of the fluid data matrix, and determining the region where the dominant mode is located; determining a high-resistance region in the liquid cooling plate based on the influence of the dominant mode on flow resistance and the region; and adjusting the shape of the high-resistance region by a feedback loop.

[0006] In this invention, intrinsic orthogonal decomposition (POD) is employed to identify and capture the main flow patterns causing resistance. The impact of these patterns on resistance is then analyzed to further determine high-resistance regions, and the shape of the liquid cooling plate is adjusted based on these high-resistance regions. The use of POD significantly reduces computational load by converting flow data into a reduced-order model. This allows for faster analysis without sacrificing accuracy, enabling real-time evaluation and optimization of the design. Furthermore, unlike traditional methods that may rely on simplified models or empirical correlations, the POD-based approach captures the main flow patterns causing resistance. These patterns reveal key flow characteristics, such as high shear stress, vortex formation, and flow separation, which are crucial for understanding the root causes of flow resistance. This ability to quantify and isolate important flow structures provides a deeper understanding of how specific flow behaviors affect resistance, helping engineers make informed design decisions.

[0007] In one optional implementation, the dominant modes of fluid motion within the liquid cooling plate are extracted by intrinsic orthogonal decomposition of the fluid data matrix, and the regions where the dominant modes are located are determined. This includes: calculating the covariance matrix of the fluid data matrix; performing singular value decomposition on the fluid data matrix to obtain a spatial mode matrix, a singular value diagonal matrix, and a temporal mode matrix; and determining the dominant modes of fluid motion within the liquid cooling plate based on the singular value decomposition results, and determining the regions where the dominant modes are located.

[0008] In one optional implementation, determining the dominant mode of fluid motion within the liquid cooling plate based on the singular value decomposition results includes: selecting multiple spatial modes associated with the maximum singular value using a rank truncation method; reconstructing the flow field using the multiple spatial modes, the corresponding singular values, and coefficients associated with the time mode to obtain a reconstructed flow field; and determining whether the multiple spatial modes are the dominant modes of fluid motion within the liquid cooling plate based on the reconstructed flow field.

[0009] In this invention, by performing covariance matrix calculation and singular value decomposition on the data matrix, the obtained spatial pattern can accurately represent the complex flow phenomena in the liquid cooling plate; at the same time, flow field reconstruction is performed based on the selected spatial pattern, which verifies the selected spatial pattern and finally determines the more influential flow pattern, namely the dominant pattern.

[0010] In one optional implementation, determining a high-resistance region in the liquid cooling plate based on the influence of the dominant mode on flow resistance and the region includes: associating the spatial characteristics of the region where the dominant mode is located with the flow structure of the liquid cooling plate; quantifying the flow resistance based on the fluid flow data of the dominant mode and the region; and determining the high-resistance region in the liquid cooling plate based on the association results and the quantification results.

[0011] In one alternative implementation, the flow structure includes: a wall region, a flow separation and recirculation region, and a geometric change region, the geometric change region including channel corners and narrow sections.

[0012] In one alternative implementation, the flow resistance is quantified based on the dominant mode and fluid flow data of the region, including: calculating pressure drop, shear stress, or velocity gradient based on the flow velocity in the region where each dominant mode is located; calculating the impact of each dominant mode on the flow resistance based on the pressure drop, shear stress, or velocity gradient; and obtaining a resistance evaluation index based on the impact of each dominant mode on the flow resistance and singular values.

[0013] This invention quantifies the impact of each mode on resistance by correlating the spatial characteristics of the dominant modes with the flow structure, thus accurately locating high-resistance regions. This provides precise data for optimizing liquid cooling plates, helping to reduce resistance, improve heat dissipation efficiency, and drive iterative design of liquid cooling systems.

[0014] In one optional implementation, acquiring fluid flow data of a liquid cooling plate and constructing a fluid data matrix includes: acquiring fluid flow data of the liquid cooling plate through simulation and measurement; cleaning, standardizing, and sampling the fluid flow data; performing data conversion and visualization verification on the processed fluid flow data; and constructing a fluid data matrix based on the verified fluid flow data.

[0015] In this invention, by processing the acquired fluid flow data, the data quality can be significantly improved, thereby ensuring that the intrinsic orthogonal decomposition can extract meaningful and accurate flow features, thus providing a reference for the subsequent optimization of the cooling plate design.

[0016] Secondly, the present invention provides a device for evaluating the flow resistance characteristics of a liquid cooling plate. The device includes: a data acquisition module for acquiring fluid flow data of the liquid cooling plate and constructing a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions; a pattern recognition module for extracting the dominant pattern of fluid motion within the liquid cooling plate by performing intrinsic orthogonal decomposition on the fluid data matrix and determining the region where the dominant pattern is located; a region determination module for determining a high-resistance region in the liquid cooling plate based on the influence of the dominant pattern on flow resistance and the region; and an optimization module for adjusting the shape of the high-resistance region using a feedback loop.

[0017] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the liquid cooling plate flow resistance characteristic evaluation method of the first aspect or any corresponding embodiment described above.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the liquid cooling plate flow resistance characteristic evaluation method of the first aspect or any corresponding embodiment thereof.

[0019] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the liquid cooling plate flow resistance characteristic evaluation method of the first aspect or any corresponding embodiment described above. Attached Figure Description

[0020] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the method for evaluating the flow resistance characteristics of a liquid cooling plate according to an embodiment of the present invention.

[0022] Figure 2 This is a flowchart illustrating another method for evaluating the flow resistance characteristics of a liquid cooling plate according to an embodiment of the present invention;

[0023] Figure 3 This is a structural block diagram of a liquid cooling plate flow resistance characteristic evaluation device according to an embodiment of the present invention;

[0024] Figure 4 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0025] In related technologies, the following methods can be used to analyze flow resistance:

[0026] The first is Computational Fluid Dynamics (CFD). CFD simulations are widely used to simulate flow fields in cooling plates. CFD helps visualize flow patterns, velocity profiles, and temperature distributions, thus aiding in understanding drag points and optimizing channel design. However, CFD simulations are computationally expensive and can be very time-consuming, especially when dealing with large and complex models.

[0027] Second, there are empirical correlations and analytical models. Some simpler analytical methods use empirical correlations based on experimental data to estimate flow resistance in cooling plates. These methods are generally faster, but they tend to oversimplify the problem and may fail to capture complex flow behaviors, such as turbulence or vortex formation, resulting in less accurate resistance predictions.

[0028] Thirdly, there are flow visualization and experimental measurements. Experimental methods, such as particle image velocimetry (PIV) or flow visualization in transparent models, can directly capture flow patterns. However, these methods are typically limited to small-scale studies and cannot be easily scaled up to evaluate the performance of real-world cooling plate designs in operating environments.

[0029] Fourthly, optimization techniques. Design optimization tools that combine CFD with genetic algorithms, particle swarm optimization, or other optimization techniques have been used to minimize flow resistance in cooling plate designs. These tools are computationally intensive and require multiple iterations to converge to the optimal solution.

[0030] However, despite advancements in CFD simulations and other techniques, current methods for evaluating and optimizing flow resistance in liquid cooling plates still face some limitations and challenges:

[0031] First, the computational cost is high. While CFD-based simulations are accurate, they require significant computational resources, especially when simulating flow on large or complex cooling plates. Simulations can take a long time to complete, thus hindering rapid evaluation or real-time optimization during the design process.

[0032] Second, there is a lack of reduced-order models. Most existing methods either rely on full-scale simulations (which are computationally expensive) or simplified models that cannot capture complex flow phenomena such as recirculation zones, eddy formation, and localized turbulence. The lack of reduced-order models for flow resistance makes it difficult to effectively evaluate and optimize the performance of cooling plates without extensive simulations or physical experiments.

[0033] Third, their ability to capture flow characteristics is limited. Traditional methods, such as empirical correlations or simplified analytical models, often fail to accurately capture the impact of complex flow characteristics (e.g., flow separation, eddies, turbulent eddies) on flow resistance. These characteristics are crucial for understanding how different design or operating conditions affect resistance, and ignoring them can lead to suboptimal designs.

[0034] Fourth, identifying the critical drag region is difficult. Although CFD can simulate detailed flow fields, extracting the most important flow features that directly cause drag (such as high shear stress, recirculation, or turbulent regions) remains challenging. These key features are often hidden in complex data and require sophisticated methods for effective identification and evaluation.

[0035] Fifth, there are optimization bottlenecks. Current optimization methods are usually computationally expensive and require a large number of repeated trials and simulations to converge to the optimal design. Therefore, they may not be suitable for real-time design adjustments or fast-paced engineering development cycles.

[0036] In view of this, this invention proposes a novel method for evaluating and analyzing the flow resistance characteristics of liquid cooling plates using Proper Orthogonal Decomposition (POD). This method integrates computational fluid dynamics (CFD) simulations or experimental measurements to generate flow field data, which is then processed and constructed into a data matrix. By applying the POD method, the fluid data is decomposed into a set of spatial and temporal patterns, revealing the main flow structures contributing to drag. These main patterns are associated with key flow features such as velocity gradients, shear stresses, and eddy formation, which are directly related to flow resistance in the cooling plate. This method can accurately identify high-resistance regions within the cooling plate and provides valuable insights for design optimization. This invention quantifies flow resistance through a reduced-order model, making it highly efficient and practical for optimizing cooling plate design in thermal management applications. This method provides a powerful tool for enhancing cooling performance, reducing energy consumption, and improving overall system efficiency.

[0037] The method of this invention can be used to evaluate and analyze the flow resistance characteristics in liquid cooling plates used in applications such as electronic cooling, heat exchangers, and automotive or industrial cooling systems. This method is particularly relevant to the development of high-performance thermal management systems that require precise flow control and heat dissipation.

[0038] 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.

[0039] According to an embodiment of the present invention, an embodiment of a method for evaluating the flow resistance characteristics of a liquid cooling plate is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0040] This embodiment provides a method for evaluating the flow resistance characteristics of a liquid cooling plate, which can be used in electronic devices such as computers, mobile phones, and tablets. Figure 1 This is a flowchart of a method for evaluating the flow resistance characteristics of a liquid cooling plate according to an embodiment of the present invention, as shown below. Figure 1As shown, the process includes the following steps:

[0041] Step S101: Obtain fluid flow data of the liquid cooling plate and construct a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions. Specifically, the fluid flow data obtained in this embodiment includes parameters such as inlet velocity, temperature distribution, and pressure gradient. These data can be obtained through fluid dynamics (CFD) simulation or through physical measurement methods (such as particle image velocimetry (PIV)).

[0042] In addition, after acquiring the fluid flow data, it can be processed, such as converting it into a form suitable for intrinsic orthogonal decomposition analysis. The working principle of POD is to decompose data (e.g., velocity field, pressure field) into a series of spatial or temporal patterns to capture the most important changes in the system. In this embodiment, the acquired fluid flow data is converted into a fluid data matrix, where each row corresponds to a spatial point (or spatial element), and each column corresponds to a snapshot of the flow field, i.e., the time step. It should be noted that this time step can also be set to other operating conditions according to actual needs.

[0043] Step S102 involves using intrinsic orthogonal decomposition (IOD) to extract the dominant patterns of fluid motion within the liquid cooling plate from the fluid data matrix and determining the regions where these dominant patterns are located. Specifically, IOD is a data dimensionality reduction and feature extraction method based on statistical analysis. This method primarily achieves dimensionality reduction through covariance matrix calculation, singular value decomposition (SVD), and dimensionality reconstruction. Therefore, IOD can be used to determine the dominant patterns corresponding to larger singular values ​​in the fluid data matrix. Furthermore, since each row in the fluid data matrix corresponds to a spatial point, the corresponding spatial location can be determined based on the spatial patterns after singular value decomposition, thus obtaining the regions where the dominant patterns are located. These dominant patterns capture the largest changes in the flow field and are most relevant to flow resistance analysis, characterizing the main flow structures in the flow field. In other words, through IOD, complex flow data is organized into a set of manageable dominant patterns.

[0044] Step S103: Based on the influence of the dominant mode on flow resistance and the defined region, determine the high-resistance region in the liquid cooling plate. Specifically, the flow resistance in a liquid cooling plate is generally related to the resistance of the coolant flowing through channels, pathways, or the cooling plate. It is affected by factors such as geometry, flow velocity, and flow properties (laminar or turbulent). The dominant mode, on the other hand, is the flow pattern that causes resistance, extracted from abstract flow field data. Therefore, by analyzing the spatial characteristics of the dominant mode and linking it to flow resistance, the high-resistance region is determined.

[0045] Step S104 involves adjusting the shape of the high-resistance region using a feedback loop. Specifically, the shape of the liquid cooling plate can be adjusted to address the high-resistance region. For example, modifying the geometry to minimize sharp corners, simplifying the flow path, or adjusting the channel size and direction can help reduce resistance. Furthermore, after shape adjustment, the process can be iterated to verify the impact of these adjustments on flow resistance, ultimately ensuring that the liquid cooling plate design achieves the optimal balance between minimum flow resistance and effective heat transfer performance.

[0046] The liquid cooling plate flow resistance characteristic evaluation method provided in this invention identifies and captures the main flow patterns causing resistance by employing intrinsic orthogonal decomposition (POD). The influence of these patterns on resistance is analyzed to further determine high-resistance regions, and the shape of the liquid cooling plate is adjusted based on these high-resistance regions. Specifically, by utilizing POD and converting flow data into a reduced-order model, the computational load is significantly reduced. This allows for faster analysis without sacrificing accuracy, enabling real-time evaluation and design optimization. Furthermore, unlike traditional methods that may rely on simplified models or empirical correlations, the POD-based method captures the main flow patterns causing resistance. These patterns reveal key flow characteristics, such as high shear stress, vortex formation, and flow separation, which are crucial for understanding the root causes of flow resistance. This ability to quantify and isolate important flow structures provides a deeper understanding of how specific flow behaviors affect resistance, thus helping engineers make informed design decisions.

[0047] This embodiment provides a method for evaluating the flow resistance characteristics of a liquid cooling plate, the process of which includes the following steps:

[0048] Step S201: Obtain fluid flow data of the liquid cooling plate and construct a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions.

[0049] Specifically, step S201 includes:

[0050] Step S2011 involves acquiring fluid flow data for the liquid cooling plate using simulation and measurement. Specifically, while CFD simulation can be used to obtain fluid flow data, the complex geometry of the liquid cooling plate makes accurate modeling in CFD difficult. Therefore, a combination of simulation and measurement is used. Particle image velocimetry (PEV) can be employed for measurement. PEV is one of the most common and accurate techniques for real-time visualization of fluid flow. Tiny tracer particles are injected into the fluid, and a high-speed camera captures their motion within the flow field. By analyzing the particle motion, the fluid's velocity field can be reconstructed.

[0051] The specific implementation steps of the particle image velocimetry method are as follows:

[0052] 1. Tracer particle injection: Small particles are injected into the fluid flow. The density and size of these particles should be similar to the coolant to ensure that they move with the fluid flow.

[0053] 2. Laser Illumination: A laser sheet is used to illuminate the flow field on a specific plane within a cooling plate. This light sheet helps visualize the motion of particles.

[0054] 3. Image acquisition: A high-speed camera continuously captures images as illuminated particles pass through the flow field.

[0055] 4. Velocity calculation: The velocity field of the flow is determined by using the displacement of particles between frames and cross-correlation techniques.

[0056] Step S2012 involves cleaning, standardizing, and sampling the fluid flow data. The processing of the fluid flow data can be divided into data cleaning, standardization, and sampling.

[0057] Specifically, cleaning data mainly includes eliminating noise and artifacts, as well as processing missing data.

[0058] Raw fluid flow data (whether from CFD simulations or experimental measurements) often contain noise or artifacts that can distort the results of POD analysis. These can be caused by numerical errors in the simulation or measurement inaccuracies (e.g., sensor drift, turbulent fluctuations in physical measurements). For CFD simulations, noise is often generated in the numerical solution, especially in regions with steep gradients (e.g., near walls or in the recirculation zone). Filters such as low-pass filters or Fourier filters can be applied to smooth high-frequency noise. For experimental measurements, it is important to detect and eliminate any anomalous readings caused by sensor malfunctions or unexpected environmental conditions. Statistical methods (e.g., Z-scores or IQR (interquartile range) methods) can be used to identify and remove outliers.

[0059] Furthermore, in some cases, data may be missing due to sensor malfunction, simulation errors, or insufficient measurement sampling. If data points are missing within a regular grid (in the case of CFD simulations), interpolation techniques (e.g., linear, cubic splines) can be used to estimate the missing values ​​based on surrounding data points. For data gaps at the edges of the measurement domain or simulation grid, extrapolation methods can be used to estimate values ​​based on the trends of adjacent data points.

[0060] Data standardization ensures that all flow variables (such as velocity, pressure, and temperature) are scaled to a comparable range, making them more suitable for analysis and comparison. This is especially important when combining different types of data or processing data from different conditions. Specifically, data standardization can include both scaling to reference values ​​and normalization.

[0061] Flow variables such as velocity and temperature can span multiple orders of magnitude depending on flow conditions. Therefore, these variables can be standardized using reference values. For velocity: flow velocity can be standardized using the average inlet velocity (i.e., the velocity at the inlet of the cooling plate) or a characteristic velocity scale derived from the flow conditions, expressed by the following formula:

[0062]

[0063] In the formula, v ref It can be the maximum speed or a reference speed value (such as the inlet speed).

[0064] For pressure: Pressure is normalized using the maximum pressure drop or certain characteristic pressure difference observed on the cooling plate. It is expressed by the following formula:

[0065]

[0066] In the formula, p ref This can be the maximum pressure or a reference pressure value (such as the inlet pressure).

[0067] When combining multiple datasets or when different quantities (such as speed and temperature) need to be processed equally, data normalization can be performed. This involves converting each variable to data with a mean of zero and a standard deviation of one, specifically calculated using the following formula:

[0068]

[0069] In the formula, x is the original data point; μ is the mean of the data; and σ is the standard deviation.

[0070] For POD analysis to be effective, flow data should be organized in a uniform grid, especially in the case of CFD simulations. Data points representing the flow field (velocity, pressure, temperature, etc.) need to be arranged in a regular, structured manner to ensure the POD algorithm can correctly decompose the data. Therefore, data sampling or uniform grid processing is necessary. Specifically, if the data is sparse or irregular (e.g., in experimental measurements), it can be interpolated to a uniform grid (spatially or temporally), i.e., uniform resampling, to ensure effective POD analysis. For example, grid interpolation methods can be used to sample velocity or pressure data at fixed grid points in a cooling plate channel. If the flow data is time-dependent (e.g., transient or pulsating flow), temporal resampling may be necessary to make the data temporally uniform. This ensures that flow snapshots from different time instances are consistent and comparable.

[0071] Furthermore, when combining data from multiple simulations or measurements, aggregation methods such as averaging or time averaging can be used to reduce the amount of data and focus on the main flow behaviors. This is particularly useful under turbulent or unsteady flow conditions, as short-term fluctuations can distract from the main flow structure. When using time averaging, the velocity or pressure field can be averaged over multiple time steps to eliminate transient effects and highlight steady-state flow characteristics.

[0072] Step S2013: Perform data conversion and visualization verification on the processed fluid flow data.

[0073] Specifically, data transformation mainly includes spatial resolution of data format and data centering. For the POD to function effectively, the flow data should be constructed as a matrix, where each row corresponds to a spatial point (or spatial element) and each column corresponds to a snapshot of the flow field. If there is a 3D flow field, it may be necessary to flatten the data into vectors at each time step to create this matrix. For example, if there are N spatial grid points in a liquid cooling plate and M time steps (or snapshots), the size of the data matrix X is N×M.

[0074] Additionally, before applying POD, it is common practice to subtract the average value of each variable (e.g., velocity component) at all spatial or temporal points to center the data. This removes any constant background components and ensures that POD identifies dynamic characteristics of the flow rather than static patterns. This is achieved using the following formula:

[0075]

[0076] Where x i is the data at point i, and M is the number of snapshots (or time step).

[0077] Before performing a Point of View (POD), visualizing the prepared data to check for consistency and identify any unexpected patterns can further ensure data quality. Visualizations can include: flow field visualization, i.e., plotting velocity or pressure profiles on a liquid cooling plate to visually examine flow patterns; and time analysis, i.e., creating data plots over time or across space to ensure the data is functioning as expected and without significant anomalies. These visual checks can also help identify areas where the data may still require further cleaning or resampling.

[0078] It is important to note that the quality and structure of the data directly affect the results of POD analysis. If the data is noisy, unscaled, or incorrectly sampled, POD decomposition may fail to identify the true dominant flow patterns and could lead to misleading conclusions regarding flow resistance characteristics. Processing the flow field data in the manner described above ensures that POD can extract meaningful and accurate flow characteristics, thus providing a reference for subsequent optimization of liquid cooling plate design.

[0079] Step S2013: Construct a fluid data matrix based on the validated fluid flow data. Specifically, the validated fluid flow data is constructed into a matrix X, which represents the flow characteristics (e.g., velocity, pressure, temperature) on the liquid cooling plate.

[0080] This data matrix is ​​specifically represented by the following formula:

[0081] X = [x1x2…x] M ]

[0082] Where, x i M represents the flow field (e.g., velocity field) at the i-th snapshot or flow configuration; M represents the number of snapshots (e.g., the number of time steps in a transient flow or the number of different operating conditions being considered).

[0083] Step S202: Extract the dominant mode of fluid motion within the liquid cooling plate by intrinsic orthogonal decomposition of the fluid data matrix, and determine the region where the dominant mode is located.

[0084] Specifically, step S202 includes:

[0085] Step S2021: Calculate the covariance matrix of the fluid data matrix; specifically, in order to decompose the flow field data into meaningful patterns, the covariance matrix of the fluid data is first calculated. The covariance matrix C provides insight into how different spatial locations within the liquid cooling plate are correlated across all snapshots.

[0086] The covariance matrix C is calculated using the following formula:

[0087]

[0088] in, It is a covariance matrix, where each entry C ij This indicates the correlation between spatial locations i and j. It is a data matrix; It is the transpose of the data matrix. Specifically, the covariance matrix can capture the spatial correlation between different flow field values ​​at different locations on the cooling plate. For example, if the covariance values ​​of two adjacent locations are high, it indicates that their flow states (such as velocity changes) have strong synchronicity and may belong to the same flow structure (such as the same vortex or shear region); conversely, it indicates that the flow characteristics of the two are weakly correlated. This correlation analysis provides a basis for the subsequent extraction of dominant modes. That is, the covariance matrix reveals from a statistical perspective which spatial locations in the flow field have "co-changing" flow characteristics, thus providing a reference for the intrinsic structure of the data for mode decomposition.

[0089] Step S2022: Perform singular value decomposition on the fluid data matrix to obtain a spatial pattern matrix, a singular value diagonal matrix, and a temporal pattern matrix; specifically, apply Singular Value Decomposition (SVD) to the data matrix X to extract the main spatial patterns of the flow field. These patterns correspond to the main flow structures that control the drag characteristics of the cooling plate.

[0090] The SVD operation on X is expressed by the following formula:

[0091] X=U∑V T

[0092] In the formula: U contains the spatial patterns of the flow field (each column of U corresponds to a spatial pattern U). k ∑ is a diagonal matrix λ containing singular values. k This indicates the importance of each spatial pattern. T Includes a time pattern (describing the evolution of fluid data over time or different snapshots). Matrix U includes a spatial pattern U. k These patterns represent the most significant flow structures that cause flow resistance, while Σ provides the magnitude or importance of each pattern.

[0093] Step S2023: Based on the singular value decomposition results, determine the dominant mode of fluid motion within the liquid cooling plate and identify the region where the dominant mode is located. This process specifically includes: selecting multiple spatial modes related to the maximum singular value using the rank truncation method; reconstructing the flow field using the multiple spatial modes, their corresponding singular values, and coefficients related to the time mode to obtain a reconstructed flow field; and determining whether the multiple spatial modes are the dominant modes of fluid motion within the liquid cooling plate based on the reconstructed flow field.

[0094] In this process, singular value decomposition (SVD) identifies the spatial patterns of the flow field and their corresponding importance (singular values). Therefore, a rank truncation step can be used to select the top r spatial patterns that contribute the most to the overall flow field, i.e., the multiple spatial patterns associated with the largest singular values. The selected spatial patterns can explain the maximum variance in the flow field (the proportion of the total change in the flow field explained by the top r spatial patterns).

[0095] Specifically, after selecting the first r spatial patterns, the selected patterns are used to reconstruct the flow field.

[0096] Wherein, the flow field x at each i i The following formula can be used to approximate the reconstruction:

[0097]

[0098] Where λk is the maximum singular value; u kIt is a spatial pattern; v k These are time-mode related coefficients, representing the changes in the flow field over time or operating conditions. These time-mode related coefficients can be determined by the time-mode matrix determined through singular value decomposition.

[0099] When reconstructing the flow field at each snapshot, the spatial pattern used is the same, but the coefficients related to the time pattern are different. Therefore, the reconstructed flow field at each snapshot is also different. The reconstructed flow field can be used to verify whether the selected spatial pattern is appropriate.

[0100] Specifically, the errors between the reconstructed flow field and the original flow field at each snapshot can be compared. Smaller errors indicate that the first r models accurately capture key flow field features, allowing for adjustments to the appropriate value of r. Once the value of r is determined, the dominant model can be identified. The values ​​of the dominant model can then be plotted on a 2D or 3D mesh of the liquid cooling plate. Regions with concentrated values ​​indicate the location of the dominant model, which can be considered key drag regions. Thus, by reconstructing the flow field, it is determined which models contribute most to flow drag and which regions or structures in the flow field have the greatest impact. The value of the dominant model can be determined using a spatial model matrix derived from singular value decomposition. Furthermore, this reconstructed flow field can be used for subsequent drag quantification, thereby reducing computational complexity.

[0101] Step S203: Based on the influence of the dominant mode on flow resistance and the region, determine the high resistance region in the liquid cooling plate;

[0102] Specifically, step S203 includes:

[0103] Step S2031: Associate the spatial characteristics of the region where the dominant mode is located with the flow structure of the liquid cooling plate; the flow structure includes: a wall region, a flow separation and recirculation region, and a geometric change region, wherein the geometric change region includes channel corners and narrow sections.

[0104] Specifically, flow resistance in fluid dynamics is generally defined as the resistance to flow caused by friction and pressure losses within a given flow path. It is influenced by factors such as geometry, flow velocity, and flow properties (laminar or turbulent). Regarding geometry, narrow channels, sharp angles, and changes in direction increase resistance. Regarding flow properties, characteristics such as fluid viscosity and density affect the ease with which the fluid passes through the channel. Regarding flow velocity, laminar, turbulent, and transitional flow all significantly affect resistance.

[0105] In liquid cooling plates, flow resistance is typically related to the resistance of the coolant as it flows through channels, pathways, or the cooling plate itself. Specifically, in liquid cooling plates, wall regions, flow separation and recirculation areas, and regions of geometric abrupt changes are areas prone to generating resistance. High velocity gradients or shear stresses near the walls contribute to resistance; flow separation or the formation of eddies leads to higher pressure drops and additional resistance, while recirculation or stagnant flow regions result in even higher resistance. Geometries that produce sudden expansion, contraction, or sharp angles can cause localized turbulence or flow separation, thereby increasing resistance.

[0106] Intrinsic orthogonal decomposition identifies dominant flow modes, which represent the most important flow characteristics (such as recirculation, eddies, and high-shear regions). These modes provide a reduced-order description of the flow field, and by analyzing them, specific flow characteristics can be correlated with flow resistance in liquid cooling plates. Specifically, dominant modes can be used for velocity and flow field analysis, analyzing the relationship between dominant modes and geometric features, and their relationship with turbulence effects, thereby identifying high-drag regions.

[0107] The dominant pattern provides the spatial distribution of velocity or flow characteristics across the entire cooling plate. Patterns corresponding to high velocity gradients, abrupt changes in direction, or large vortex regions are key indicators of high flow resistance. For example, in vortex regions, flow separation and vortex formation lead to increased drag. Patterns capturing these characteristics are crucial for understanding the locations where drag arises due to flow instability. Regions with steep velocity gradients (e.g., near walls or narrow areas) tend to have higher flow resistance. By analyzing patterns associated with large velocity gradients, the regions contributing most to pressure drop can be identified. Based on this, the regions where the identified dominant patterns are located are correlated with these characteristics to identify high flow resistance. For example, if a dominant pattern has particularly large values ​​near the wall or in narrow channels, indicating a steep velocity gradient, that location is identified as a high flow resistance region.

[0108] Secondly, by analyzing the dominant modes and comparing them with the geometry of the cooling plate (e.g., channel shape, plate geometry, etc.), the drag caused by the geometry can be determined. For example, sharp turns or sudden expansion / contraction in the cooling plate geometry often lead to turbulence, thus increasing drag. The dominant modes corresponding to these flow structures can indicate where the flow becomes turbulent or unstable. In narrow channels, the flow encounters higher drag because a higher velocity is required to maintain the same flow rate. The corresponding dominant modes will show velocity concentration and higher shear stress in these areas.

[0109] Furthermore, in turbulent flow, drag is higher due to the chaotic fluid motion. If the flow is turbulent in certain regions, it will be captured by higher-order dominant modes, which represent more complex and irregular flow patterns. Turbulence increases energy dissipation in the fluid, resulting in a higher pressure drop on the cooling plate.

[0110] For example, the dominant mode u1 might show a region of high shear stress near the cooling channel wall, indicating significant flow resistance. Similarly, the dominant mode u2 might correspond to a large vortex or recirculation region that causes additional drag.

[0111] Step S2032: Quantify the flow resistance based on the dominant mode and the fluid flow data of the region. Specifically, in addition to identifying high-resistance regions using the aforementioned correlation method, the resistance can be further quantified based on the results of intrinsic orthogonal decomposition, i.e., the determined dominant mode. During quantization, a flow resistance index related to the mode can be introduced to quantify the flow resistance.

[0112] In one optional implementation, step S2032 includes:

[0113] Step a1: Calculate pressure drop, shear stress, and velocity gradient based on fluid flow data in the region where each dominant mode is located.

[0114] Step a2: Calculate the effect of each dominant mode on flow resistance based on the pressure drop, shear stress, and velocity gradient.

[0115] Step a3: Obtain the resistance assessment index based on the impact of each dominant mode on flow resistance and singular values.

[0116] Specifically, the indicators introduced in this embodiment when quantifying resistance include pressure drop, shear stress, and velocity gradient.

[0117] The pressure drop across a liquid cooling plate is a fundamental measure of flow resistance. This pressure drop can be derived from flow field data and is commonly used as an indicator of the degree of obstruction encountered by the flow as it passes through the cooling plate. For flow through a channel or pipe, the pressure drop ΔP is related to the flow velocity and the channel resistance. This relationship is typically given by the Darcy-Wiesbach equations for laminar or turbulent flow, as shown in the following formula:

[0118]

[0119] Where f is the coefficient of friction (which depends on the flow state), L is the characteristic length of the channel, D is the diameter of the channel or hydraulic diameter, ρ is the fluid density, and u is the fluid velocity.

[0120] Once the pressure drop is determined, the relationship between the pressure drop and the resistance is expressed by the following formula:

[0121]

[0122] Where R is the flow resistance. Let μ be the pressure gradient, μ be the dynamic viscosity of the fluid, and A be the surface area of ​​the cooling plate. Using fluid flow data in the region where the dominant mode is located, the effect of the mode on the local pressure gradient can be calculated, and the flow resistance in different regions of the cooling plate can be determined. For example, the high-pressure gradient associated with a specific mode can point to a high-resistance region.

[0123] Wall shear stress, caused by fluid viscosity and velocity gradient, is a significant factor in flow resistance. Wall shear stress τ w It is related to the velocity gradient near the wall and can be expressed as:

[0124]

[0125] Where μ is the dynamic viscosity of the fluid; u is the flow velocity; and y is the distance from the wall. The dominant mode reveals the locations where these gradients are maximized, which helps to pinpoint the location of the highest drag. The relationship between shear stress and drag can be determined by referring to relevant techniques.

[0126] Specifically, based on the above indicators, a comprehensive analysis of flow resistance can be conducted. This embodiment uses the following formula to calculate the indicators for evaluating flow resistance:

[0127]

[0128] Where, λ k The singular value associated with the k-th dominant mode indicates its importance in explaining flow behavior. The Resistance Contribution of Modek can be calculated based on the pressure drop, shear stress, or velocity gradient associated with that mode, representing the "drag contribution" of each mode. It quantifies the contribution of each mode to the total drag. This metric helps prioritize the most influential flow structures in drag analysis and can guide design optimization to minimize drag by adjusting the geometry of cooling plates or operating conditions.

[0129] Step S2033: Based on the correlation and quantification results, determine the high-resistance region in the liquid cooling plate. Specifically, after identifying the high-resistance region by correlating the spatial characteristics of the region where the dominant mode is located with the flow structure of the liquid cooling plate, the flow resistance can be quantified to verify the high resistance of the high-resistance region, thereby clarifying the final high-resistance region.

[0130] Step S204 involves adjusting the shape of the high-resistance region using a feedback loop. For details, please refer to [link to relevant documentation]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0131] Based on the above steps, the liquid cooling plate flow resistance characteristic evaluation method provided by the present invention has the following effects:

[0132] First, it reduces computational complexity. Traditional methods, such as full-scale CFD simulations, are computationally expensive and time-consuming, especially for complex cooling plate geometries. This invention utilizes intrinsic orthogonal decomposition to transform flow data into a reduced-order model, significantly reducing the computational load. This enables faster analysis without sacrificing accuracy, allowing for real-time evaluation and design optimization.

[0133] Second, an improved flow characteristic identification method is employed. Unlike traditional methods that may rely on simplified models or empirical correlations, the POD-based approach can capture the main flow patterns that contribute to drag. These patterns reveal key flow characteristics, such as high shear stress, vortex formation, and flow separation, which are crucial for understanding the root causes of flow drag. This ability to quantify and isolate important flow structures allows for a deeper understanding of how specific flow behaviors affect drag, thus helping engineers make informed design decisions.

[0134] Third, it offers higher accuracy in assessing flow resistance. The POD decomposition employed in this invention can more accurately represent complex flow phenomena in liquid cooling plates, including turbulent eddies, recirculation zones, and velocity gradients. Traditional methods may fail to capture these complex details, leading to less accurate resistance predictions. By focusing on the most influential flow patterns, this invention ensures a more precise assessment of flow resistance and its relationship with cooling performance.

[0135] Fourth, it enables enhanced design optimization. This invention facilitates more targeted and efficient cooling plate design optimization by identifying high-resistance areas and allowing engineers to focus on areas requiring improvement. This targeted approach reduces the need for extensive trial-and-error simulations common in traditional optimization methods. By improving cooling efficiency and minimizing flow resistance, this invention helps reduce energy consumption and pump power requirements.

[0136] Fifth, it is scalable and practically applicable. While experimental techniques and CFD simulations may be limited by scale or complexity, the POD-based approach is highly scalable and can be applied to various cooling plate geometries and operating conditions. This approach is suitable not only for small-scale studies but also for large-scale or industrial applications, making it applicable to various fields such as electronics, automotive, and renewable energy systems.

[0137] Sixth, it enables real-time design feedback. By significantly reducing the computation time and resources required for flow resistance analysis, this invention enables real-time design feedback. Engineers can quickly assess the impact of design changes on flow resistance, simplify the optimization iteration process, and ensure that the cooling plate more effectively meets performance standards.

[0138] Seventh, it enables comprehensive and systematic analysis. The POD method allows for a more systematic analysis of flow resistance because it organizes complex flow data into a set of manageable dominant patterns. This systematic approach contrasts with traditional methods that may ignore or underestimate certain flow characteristics, providing a clearer path to improving cooling plate design.

[0139] Based on this, the present invention represents a significant advancement in efficiency, accuracy, and scalability, overcoming key limitations of current technology. It provides a faster, more accurate, and computationally more efficient method for evaluating and optimizing the flow resistance of liquid cooling plates, ultimately contributing to improved cooling system performance and reduced energy consumption.

[0140] As a specific application embodiment of the present invention, such as Figure 2 As shown, the method for evaluating the flow resistance characteristics of the liquid cooling plate is implemented using the following process:

[0141] Step 1: Fluid flow simulation or measurement.

[0142] This method first acquires fluid flow data for a given liquid cooling plate. This data can be obtained through computational fluid dynamics simulations or through physical measurement methods. The flow data may include various parameters, such as inlet velocity, temperature distribution, and pressure gradient.

[0143] Step 2, Data Preparation.

[0144] Once the flow field data is obtained, it is processed into a form suitable for analysis. This may involve resampling, normalizing, or discretizing the data to ensure consistency under different flow conditions.

[0145] Step 3, intrinsic orthogonal decomposition (proper orthogonal decomposition).

[0146] The processed data undergoes intrinsic orthogonal decomposition (IOD) to decompose the complex flow field into orthogonal modes. IOD identifies the most important flow modes, or dominant modes, which correspond to the main modes of fluid motion within the cooling plate channels. Each mode represents a unique characteristic of the flow, such as a recirculation zone, eddies, or a high shear stress region, which affects the overall flow resistance.

[0147] Step 4, flow resistance analysis.

[0148] The dominant modes identified were analyzed to assess their impact on overall flow resistance. The relationship between flow resistance and these modes can be further quantified using established fluid dynamics principles (such as the Darcy-Weisbach or Colebrook-White equations).

[0149] Step 5, Optimize and design feedback.

[0150] Based on the analysis of flow resistance patterns, design modifications to the cooling plate structure can be proposed. This may involve changing the channel geometry, optimizing flow distribution, or implementing design features that minimize resistance.

[0151] This method allows for iterative feedback loops, where design variations can be re-simulated and analyzed through an intrinsic orthogonal decomposition process for further optimization.

[0152] This invention enables engineers to understand the impact of fluid dynamics on cooling performance and system pressure drop. By utilizing intrinsic orthogonal decomposition (POD), reduced-order modeling is achieved, significantly lowering computational costs compared to traditional CFD simulations. This allows for faster analysis and real-time optimization during the design process. Furthermore, POD identifies and quantifies the most influential flow characteristics (e.g., velocity gradients, shear stress, eddy formation, and flow separation) that directly contribute to resistance in the cooling plate. This facilitates precise identification of high-resistance areas for design improvements. Additionally, by combining flow resistance analysis with design feedback, the geometry of the liquid cooling plate is optimized, thereby improving cooling efficiency, reducing energy consumption, and enhancing thermal performance. Therefore, this method can be seamlessly integrated into the design and development of liquid cooling systems, benefiting a wide range of applications from electronics and automotive to industrial heat exchangers and power generation systems.

[0153] This embodiment also provides a liquid cooling plate flow resistance characteristic evaluation device, which is used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0154] This embodiment provides a device for evaluating the flow resistance characteristics of a liquid cooling plate, such as... Figure 3 As shown, it includes:

[0155] Data acquisition module 31 is used to acquire fluid flow data of liquid cooling plate and construct fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions;

[0156] The pattern recognition module 32 is used to extract the dominant pattern of fluid motion within the liquid cooling plate by intrinsic orthogonal decomposition of the fluid data matrix, and to determine the region where the dominant pattern is located.

[0157] The region determination module 33 is used to determine the high resistance region in the liquid cooling plate based on the influence of the dominant mode on the flow resistance and the region.

[0158] The optimization module 34 is used to adjust the shape of the high-resistance region using a feedback loop.

[0159] Further functional descriptions of the above modules are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0160] This invention also provides a computer device having the above-described features. Figure 3 The device shown is for evaluating the flow resistance characteristics of a liquid cooling plate.

[0161] Please see Figure 4 , Figure 4 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 4 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 4 Take a processor 10 as an example.

[0162] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0163] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0164] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device as shown by a landing page for an app. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0165] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0166] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0167] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0168] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0169] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for evaluating the flow resistance characteristics of a liquid cooling plate, characterized in that, The method includes: Obtain fluid flow data from the liquid cooling plate and construct a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions; The dominant modes of fluid motion within the liquid cooling plate are extracted by intrinsic orthogonal decomposition of the fluid data matrix, and the regions where the dominant modes are located are determined. Based on the influence of the dominant mode on flow resistance and the region, the high resistance region in the liquid cooling plate is determined. The shape of the high-resistance region is adjusted using a feedback loop.

2. The method according to claim 1, characterized in that, The dominant modes of fluid motion within the liquid cooling plate are extracted by intrinsic orthogonal decomposition of the fluid data matrix, and the regions where the dominant modes are located are determined, including: Calculate the covariance matrix of the fluid data matrix; Perform singular value decomposition on the fluid data matrix to obtain the spatial pattern matrix, the singular value diagonal matrix, and the temporal pattern matrix; Based on the results of singular value decomposition, the dominant mode of fluid motion within the liquid cooling plate is determined, and the region where the dominant mode is located is identified.

3. The method according to claim 2, characterized in that, The dominant modes of fluid motion within the liquid cooling plate were determined based on the results of singular value decomposition, including: Multiple spatial patterns related to the maximum singular value are selected using the rank truncation method. The flow field is reconstructed by using the multiple spatial models, corresponding singular values, and coefficients related to the time model; Based on the reconstructed flow field, it is determined whether multiple spatial patterns are the dominant patterns of fluid motion within the liquid cooling plate.

4. The method according to claim 1, characterized in that, Based on the influence of the dominant mode on flow resistance and the region, high-resistance regions in the liquid cooling plate are determined, including: The spatial characteristics of the region where the dominant mode is located are correlated with the flow structure of the liquid cooling plate; The flow resistance is quantified based on the dominant mode and the fluid flow data of the region. High-resistance regions in the liquid cooling plate were identified based on correlation and quantification results.

5. The method according to claim 4, characterized in that, The flow structure includes: a wall region, a flow separation and recirculation region, and a geometric change region, wherein the geometric change region includes channel corners and narrow sections.

6. The method according to claim 4, characterized in that, The flow resistance is quantified based on the dominant mode and fluid flow data of the region, including: Calculate pressure drop, shear stress, or velocity gradient based on the flow velocity in the region where each dominant mode is located. The effect of each dominant mode on flow resistance is calculated based on the pressure drop, shear stress, or velocity gradient. Resistance assessment metrics are derived based on the impact of each dominant mode on flow resistance and singular values.

7. The method according to claim 1, characterized in that, Acquire fluid flow data from the liquid cooling plate and construct a fluid data matrix, including: Fluid flow data of the liquid cooling plate is obtained through simulation and measurement. The fluid flow data is cleaned, standardized, and sampled. The processed fluid flow data is transformed and visualized for verification. A fluid data matrix is ​​constructed based on the validated fluid flow data.

8. A device for evaluating the flow resistance characteristics of a liquid cooling plate, characterized in that, The device includes: The data acquisition module is used to acquire fluid flow data of the liquid cooling plate and construct a fluid data matrix, wherein the row elements of the fluid data matrix correspond to spatial positions. The pattern recognition module is used to extract the dominant pattern of fluid motion within the liquid cooling plate by intrinsic orthogonal decomposition of the fluid data matrix, and to determine the region where the dominant pattern is located. A region determination module is used to determine high-resistance regions in a liquid cooling plate based on the influence of the dominant mode on flow resistance and the region. An optimization module is used to adjust the shape of the high-resistance region using a feedback loop.

9. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the liquid cooling plate flow resistance characteristic evaluation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the liquid cooling plate flow resistance characteristic evaluation method according to any one of claims 1 to 7.

Images