Interface optimization method for hollow micron-sized silica spheres-aluminum nitride whisker composites

Abstract

This invention relates to the field of composite material interface optimization technology, and discloses a method for interface optimization of hollow micron-sized silica spheres-aluminum nitride whisker composite materials. This method is applied to a material composite system including a dispersion sensor and an interface control system in a hot-pressing mold. The method includes generating a dataset by collecting spherulite dispersion data and interface stress data through sensors; constructing a dispersion matrix using the control system; determining the material interface bonding characteristics, including matrix processing, thermal field distribution modeling, calculating the thermal field distribution gradient and bonding prediction information, constructing a coupled control model to output topological distribution characteristics and update the dataset; and finally, dynamically adjusting composite parameters based on the bonding characteristics. This method achieves real-time monitoring and precise control of the composite material interface, improving interface bonding quality and material properties, and is suitable for interface optimization in the composite material preparation process.

Description

Technical Field

[0001] This invention relates to the field of composite material interface optimization technology, specifically a method for optimizing the interface of hollow micron-sized silica spheres-aluminum nitride whisker composite materials. Background Technology

[0002] In the research and application of advanced composite materials, hollow micron-sized silica spheres-aluminum nitride whisker composites have shown broad application prospects in fields such as electronic packaging and high-temperature structural components due to their combination of the high temperature resistance and insulation properties of silica and the high thermal conductivity and mechanical properties of aluminum nitride. However, these composite materials face severe interface problems during preparation, which seriously restrict the improvement of the overall performance of the materials and their practical application.

[0003] From the perspective of the material properties themselves, hollow micron-sized silica spheres and aluminum nitride whiskers exhibit significant differences in their physicochemical properties. Their coefficients of thermal expansion are mismatched, and during the preparation and use of the composite material, temperature changes can lead to significant thermal stress at the interface, easily causing problems such as interface debonding and crack propagation. Simultaneously, their different surface energies make uniform dispersion difficult during the composite process, easily leading to agglomeration, which in turn affects the uniformity of the material's mechanical and thermal conductivity properties.

[0004] Existing methods for optimizing composite material interfaces have several limitations in addressing such problems. Traditional interface modification methods, such as surface coatings, while improving interfacial bonding to some extent, often struggle to precisely control coating thickness and uniformity, and their effectiveness is less than ideal for complex composite structures. In composite material fabrication processes, hot pressing is a commonly used method, but traditional hot pressing processes make it difficult to monitor and control the material state within the composite region in real time. The lack of effective real-time monitoring methods makes it impossible to accurately obtain spherulite dispersion and interfacial stress data, hindering dynamic adjustments to process parameters based on actual conditions. This results in unstable interfacial bonding strength and significant dispersion in material properties.

[0005] Furthermore, there are shortcomings in the construction of interface analysis and optimization models for composite materials. Most existing models are based on simplified assumptions and fail to fully consider the complex physicochemical changes in materials during the composite process, such as the influence of thermal field distribution gradients and whisker orientation on interfacial bonding. This results in low predictive accuracy and an inability to provide accurate guidance for interface optimization. Moreover, traditional methods often rely on post-construction detection and repair when dealing with interface defects. This passive approach not only increases production costs but also makes it difficult to guarantee the reliability and consistency of materials.

[0006] With the increasing demand for high-performance composite materials in fields such as electronics and aerospace, there is an urgent need for an optimization method that can monitor and precisely control the interfacial bonding state of composite materials in real time. This method needs to address issues such as insufficient interfacial bonding strength, uneven material properties, and lack of real-time process control in existing technologies, in order to improve the overall performance and practical application value of hollow micron-sized silica sphere-aluminum nitride whisker composite materials. Summary of the Invention

[0007] The purpose of this invention is to provide an interface optimization method for hollow micron-sized silica sphere-aluminum nitride whisker composite materials to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: an interface optimization method for hollow micron-sized silica spheres-aluminum nitride whisker composite materials, the method comprising:

[0009] The dispersion sensor is used to collect spherulite dispersion data and interfacial stress data within the composite region to generate a material state dataset.

[0010] The interface control system receives real-time material state data from multiple dispersion sensors to construct a regional dispersion matrix.

[0011] Based on the material state dataset and real-time data from each dispersion sensor, the material interface bonding characteristics are determined. The determination of the material interface bonding characteristics includes: processing the dispersion matrix, extracting interface features from the material state dataset, predicting interface bonding based on the thermal field distribution gradient and whisker orientation information, outputting the topological distribution characteristics of the material interface bonding through a coupling control model, and updating the material state dataset based on the topological distribution characteristics.

[0012] Based on the aforementioned bonding characteristics, the composite parameters of the regional materials are dynamically adjusted.

[0013] Preferably, determining the material interface bonding characteristics includes:

[0014] The dispersion matrix is ​​processed to extract spherulite dispersion topology, interfacial stress characteristics, and bonding strength trends;

[0015] Based on the spherulite dispersion topology and interfacial stress characteristics, the dispersion matrix is ​​modeled for thermal field distribution. The composite region is divided into multiple sub-units and the unit identifiers are marked. The spherulite dispersion of the sub-units is correlated and matched with the material state dataset, and the unit identifiers are marked in the material state dataset.

[0016] The thermal field distribution gradient is calculated based on the location of the dispersion sensor. The interface bonding distribution is predicted based on the thermal field distribution gradient and the bonding strength trend. The bonding prediction information for each sub-unit is calculated.

[0017] A coupling control model is constructed, and the predicted information is used as the input parameter of the coupling control model. The coupling control model is used to perform spatial correlation modeling on the predicted information and output the topological distribution characteristics of the material interface bonding.

[0018] Update the material state dataset based on the topological distribution characteristics to obtain the material interface binding characteristics.

[0019] Preferably, the processing of the dispersion matrix includes:

[0020] The dispersion matrix is ​​normalized, and the stress concentration region in the matrix is ​​truncated by a sliding window. The noise in the concentration region is suppressed, and the combined strength trend is calculated by tensor decomposition algorithm.

[0021] The spatial correlation characteristics of the dispersion matrix are calculated. Based on the spatial correlation characteristics, the stress interference coefficient, whisker stability index and interface defect degree between units are calculated. A feature fusion network is constructed, and the interface stress characteristics are calculated through the feature fusion network.

[0022] Extract the time-domain and frequency-domain features collected by each dispersion sensor, calculate the combined feature vector of the sensor based on the phase difference between the time-domain and frequency-domain features, and perform feature matching on dispersion sensors at different locations based on the combined feature vector to calculate the trend of the combined strength.

[0023] Preferably, the thermal field distribution modeling of the dispersion matrix includes:

[0024] Based on the spherulite dispersion topology, spherulite dispersion sampling points are extracted from each frame of data. The sampling points are correlated and mapped with the interface stress characteristics to generate a thermal field distribution map. The thermal field distribution maps collected by multiple sensors are spatially registered to calculate the thermal field distribution intensity of the region.

[0025] Set a defect threshold, locate the defect source based on the spherulite dispersion value of the multi-frame dispersion matrix, calculate the defect intensity difference, if the defect intensity difference is greater than or equal to the defect threshold, it indicates that there is an interface defect in the unit, perform hot-pressing parameter constraint compensation on the current unit, iteratively correct the thermal field distribution intensity of the current unit according to the heat conduction model corresponding to the current unit, and calculate the thermal field compensation value of the defect area based on the correction result.

[0026] The thermal field distribution is modeled based on the dispersion matrix according to the thermal field distribution intensity, and the thermal field model of the region is stress-labeled by the interface stress characteristics.

[0027] Preferably, the step of calculating the thermal field distribution gradient based on the location of the dispersion sensor includes:

[0028] Based on multiple sets of composite material data, the points of change in thermal field intensity are extracted, and the points of change are mapped to a unified thermodynamic coordinate system according to the deployment location of the sensors. The points of change are then fitted using a heat flux vector interpolation algorithm to generate a thermal field distribution model for the region.

[0029] Equal-interval sampling is performed along the heat transfer path of the thermal field distribution model. The thermal attenuation rate, stress fluctuation index, and thermal field change slope of the path are calculated based on the sampling results. The thermal field change parameters are calculated based on the thermal attenuation rate, stress fluctuation index, and thermal field change slope.

[0030] Based on the deployment parameters and acquisition accuracy of the dispersion sensor, the distribution characteristics of the interface bonding strength in each frame of data are projected onto the thermal field distribution model. The model is divided into zones according to the number of sensors along the heat transfer direction of the thermal field distribution model. The variation law of the interface bonding strength in the zone is analyzed, and the bonding distribution characteristics are calculated based on the variation law.

[0031] Preferably, based on the thermodynamic coordinate range from the first dispersion sensor to the last dispersion sensor, thermal field coordinate points are selected in the sensor deployment direction, and the product of the thermal field intensity characteristic weight value and the combined distribution characteristic weight value within the spatial resolution range is calculated cumulatively, and the influence value of the sensor acquisition frequency on the rate of change of thermal field intensity is superimposed.

[0032] Preferably, the calculation of the combined prediction information for each sub-unit includes:

[0033] Using the heat transfer main diameter of the aforementioned thermal field distribution model as the baseline, and taking the peak position of the interface bonding strength in each frame of data as the reference point, the bonding offset is calculated, and the bonding distribution curve is plotted according to thermodynamic coordinates.

[0034] Based on the thermal field distribution gradient, the growth rate and direction in the bonding strength trend are corrected;

[0035] Starting from the most recent combined distribution point, the distribution curve is continued to be plotted based on the correction results of the growth rate and direction, generating the next period's combined distribution point, until the distribution points cover the entire target area, generating combined prediction information.

[0036] Preferably, the construction of the coupled regulation model includes:

[0037] The input layer is used to organize the combined prediction information into spatially distributed data and perform normalization processing.

[0038] The feature fusion layer is used to extract regional correlation features of interface bonding by processing spatially distributed data, and to construct the dependency relationship between material units.

[0039] The parameter control layer is used to integrate the relationship between the interface and the spatial unit to generate a material composite parameter control strategy.

[0040] Preferably, the acquisition of material interface bonding features includes:

[0041] Based on the topological distribution characteristics of the material interface combination output by the coupling control model, the identifiers of the sub-units are mapped to the topological distribution characteristics.

[0042] The element data in the material state dataset is reorganized according to topological features to generate an element distribution map sorted by interface bonding strength;

[0043] Based on the reorganized unit distribution diagram, the optimized interface combination distribution characteristics are output.

[0044] Preferably, the dynamic control of the composite parameters of the regional materials includes:

[0045] When the interface reaches a preset intensity threshold in the target area, a hot pressing power increase command is triggered in the adjacent area.

[0046] Based on the whisker orientation strategy, hot pressing parameters are dynamically combined to generate a heat preservation time vector;

[0047] Adjust the mold parameters of the hot-pressing node in the target area based on the heat preservation time vector.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] By collecting spherulite dispersion data and interfacial stress data within the composite region using a dispersion sensor and generating a material state dataset, the interface control system receives real-time data to construct a regional dispersion matrix. This enables comprehensive and real-time monitoring of the composite material interface state, providing accurate data support for interface optimization. This real-time monitoring mechanism changes the passive situation of post-processing detection in traditional methods, allowing for timely detection of interface problems during composite material preparation and avoiding material waste and performance defects caused by untimely problem detection.

[0050] In determining the interfacial bonding characteristics of materials, the dispersion matrix is ​​processed to extract spherulite dispersion topology, interfacial stress characteristics, and bonding strength trends. Interfacial bonding is then predicted by combining thermal field distribution gradient and whisker orientation information. A coupled control model outputs the topological distribution characteristics of the material interfacial bonding. This series of operations fully considers the various complex factors affecting interfacial bonding, enabling a precise grasp of the actual interfacial bonding situation. Unlike traditional models based on simplified assumptions, this method constructs a model that better reflects the physicochemical changes in the actual composite process, significantly improving the accuracy of interfacial bonding state prediction and providing a reliable basis for subsequent interfacial optimization and control.

[0051] Based on the bonding characteristics, the composite parameters of the regional materials are dynamically adjusted. When the interface bonding reaches a preset strength threshold in the target area, a hot pressing power increase command is triggered in the adjacent area. The hot pressing parameters are dynamically combined according to the whisker orientation strategy to generate a holding time vector, and the mold parameters of the hot pressing nodes in the target area are adjusted accordingly. This dynamic control method can precisely adjust the process parameters for the interface state of different regions and at different times during the composite process, achieving refined control of the interface bonding of composite materials. Compared with the limitations of traditional hot pressing processes that are difficult to adjust parameters in real time, this method effectively solves the problems of unstable interface bonding strength and large material property dispersion, significantly improving the uniformity and consistency of composite material properties.

[0052] When processing the dispersion matrix, operations such as normalization, sliding window extraction of stress concentration regions, noise suppression, and tensor decomposition algorithms are used to calculate and combine these with intensity trends. Simultaneously, spatial correlation features are calculated to construct a feature fusion network for calculating interface stress features. Time-domain and frequency-domain features of the sensors are extracted and combined with feature vectors for feature matching. These processing methods effectively remove noise interference from the data, extract feature information that better reflects the true state of the interface, improve the accuracy and reliability of data processing, and provide high-quality data support for subsequent interface analysis and optimization.

[0053] When modeling the thermal field distribution of the dispersion matrix, the thermal field distribution map is generated by extracting the correlation mapping between the spherulite dispersion sampling points and the interface stress characteristics. Spatial registration is then performed to calculate the intensity of the regional thermal field distribution. A defect threshold is set to locate the defect source and compensate for the thermal pressure parameter constraints. Finally, stress is labeled on the thermal field model using the interface stress characteristics. This series of thermal field distribution modeling operations achieves an accurate description of the thermal field distribution within the composite region and effective location and compensation of defects. It solves the problems of unclear thermal field distribution and passive defect handling in traditional methods, further optimizes the interfacial bonding state of the composite material, and improves the reliability and stability of the material.

[0054] When calculating the thermal field distribution gradient based on the location of the dispersion sensor, the points of thermal field intensity change are extracted and mapped to a unified thermodynamic coordinate system. A thermal field distribution model is generated by fitting the model using a heat flux vector interpolation algorithm. Thermal field change parameters are calculated by sampling along the heat transfer path, and the bonding intensity variation law at the interface is analyzed to calculate the bonding distribution characteristics. Finally, the thermal field distribution gradient is calculated. This process fully considers factors such as sensor deployment location and acquisition accuracy, accurately describes the thermal field distribution gradient and bonding distribution characteristics, provides more accurate parameters for bonding prediction, and makes the prediction of the interface bonding state more precise, thus enabling more targeted interface optimization and control.

[0055] When calculating the bonding prediction information for each sub-unit, the heat transfer principal diameter of the thermal field distribution model is used as the baseline. The bonding offset is calculated, and a bonding distribution curve is plotted. The growth rate and direction of the bonding strength trend are corrected according to the thermal field distribution gradient to generate the bonding distribution points for the next time period until the target area is covered. This bonding prediction method can accurately predict the bonding distribution in future time periods based on the current interface state and thermal field distribution, providing forward-looking guidance for dynamic control. This makes interface optimization and control more scientific and reasonable, further improving the quality and performance of composite material interface bonding.

[0056] When constructing the coupled control model, the predicted information is organized into spatially distributed data and normalized through an input layer, a feature fusion layer, and a parameter control layer. Regional correlation features of the interface bonding are extracted to construct the material unit dependencies, generating a material composite parameter control strategy. This model effectively integrates and analyzes the spatial distribution characteristics of interface bonding, generating precise control strategies based on the actual needs of interface bonding. This improves the efficiency and effectiveness of interface optimization control, resulting in more optimized interface bonding and significantly enhanced overall performance of the composite material.

[0057] When acquiring material interface bonding characteristics, sub-unit identifiers are mapped to topological distribution features. The unit data in the material state dataset is then reorganized according to topological features to generate a unit distribution map, and the optimized interface bonding distribution features are output. This process achieves a clear presentation and orderly management of interface bonding characteristics, facilitating an intuitive understanding of the overall state and local details of composite material interface bonding, and providing strong support for further optimization of composite material design and fabrication processes. Attached Figure Description

[0058] Figure 1 This is a schematic diagram illustrating the working principle of the interface optimization method for hollow micron-sized silica spheres-aluminum nitride whisker composite materials described in this invention.

[0059] Figure 2 A schematic diagram illustrating the working principle of scatter matrix processing;

[0060] Figure 3 A schematic diagram illustrating the working principle of modeling the thermal field distribution of the dispersion matrix;

[0061] Figure 4 This is a diagram illustrating the working principle of calculating the thermal field distribution gradient. Detailed Implementation

[0062] 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, and 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.

[0063] Please see Figures 1-4 This invention relates to an interface optimization method for hollow micron-sized silica spheres-aluminum nitride whisker composite materials. The method includes multiple dispersion sensors deployed in a hot-pressing mold and an interface control system connecting the dispersion sensors, with adjacent dispersion sensors spaced at a predetermined distance. The specific implementation steps are as follows:

[0064] Data on spherulite dispersion and interfacial stress within the composite region are collected using dispersion sensors to generate a material state dataset. These sensors are uniformly distributed within the hot-pressing mold, monitoring changes in various parameters during the material composite process in real time to ensure the collected data accurately reflects the actual state of the composite region. An interface control system receives real-time material state data from multiple dispersion sensors and constructs a regional dispersion matrix. The interface control system processes and integrates the received data, converting the data collected by each sensor into a matrix format for subsequent analysis.

[0065] Based on the material state dataset and real-time data from each dispersion sensor, the material interface bonding characteristics are determined. Specifically, the dispersion matrix is ​​processed, interface features are extracted using the material state dataset, and interface bonding is predicted based on the thermal field distribution gradient and whisker orientation information. The topological distribution characteristics of the material interface bonding are output through a coupled control model, and the material state dataset is updated based on these topological distribution characteristics.

[0066] Based on the bonding characteristics, the composite parameters of the regional materials are dynamically adjusted. Based on the determined interfacial bonding characteristics, the suitability of the current composite parameters is analyzed; if deficiencies are found, adjustments are made to optimize the interfacial bonding performance of the materials.

[0067] Example 1:

[0068] In the material composite system, the system includes multiple dispersion sensors deployed in a hot press mold, with a set distance between adjacent dispersion sensors, as well as an interface control system connecting these dispersion sensors.

[0069] The dispersion matrix is ​​processed to extract spherulite dispersion topology, interfacial stress characteristics, and bonding strength trends. The first step in processing the dispersion matrix is ​​normalization, which ensures data comparability and consistency, laying the foundation for subsequent analysis. After normalization, a sliding window is used to extract stress concentration regions from the matrix. The size and position of the sliding window can be adjusted according to specific data characteristics and analytical needs to accurately capture stress concentration areas. After extracting the stress concentration regions, noise suppression is applied to these areas. Noise is inevitably introduced during data acquisition and transmission, affecting data analysis and understanding; noise suppression improves data quality and reliability. After noise suppression, a tensor decomposition algorithm is used to calculate the bonding strength trend. Tensor decomposition is a powerful data analysis tool that decomposes complex tensor data into a combination of multiple low-dimensional tensors, thereby extracting latent features and trends. This algorithm can accurately calculate the changing trend of bonding strength.

[0070] The spatial correlation characteristics of the dispersion matrix are calculated. These characteristics reflect the spatial distribution and association of data, and calculating them allows for a better understanding of the interactions between different locations during the material composite process. Based on the calculated spatial correlation characteristics, the stress interference coefficient, whisker stability index, and interface defect degree between elements are further calculated. The stress interference coefficient measures the degree of stress interference between different elements, the whisker stability index reflects the stability of whiskers during the composite process, and the interface defect degree indicates the extent of defects at the interface. After obtaining these parameters, a feature fusion network is constructed. This network integrates and fuses multiple features, fully utilizing the complementarity between features. The interface stress characteristics are calculated through the feature fusion network, thus providing a more comprehensive and accurate description of the interface stress state.

[0071] The time-domain and frequency-domain features acquired by each dispersion sensor are extracted. Time-domain features describe the signal's variation over time, while frequency-domain features reflect the signal's distribution over frequency. The combined feature vector of the sensors is calculated based on the phase difference between the time-domain and frequency-domain features. The phase difference is a crucial parameter between the time-domain and frequency-domain features, containing information about the signal's structure and properties. The combined feature vector calculated using the phase difference comprehensively reflects the sensor's integration at different locations. Then, feature matching is performed on dispersion sensors at different locations based on the combined feature vector. Feature matching determines the correlation and correspondence between different sensors, thereby calculating the trend of integration strength.

[0072] In processing the dispersion matrix, each step is closely interconnected and mutually influential. Normalization provides a solid data foundation for subsequent stress concentration region extraction, noise suppression improves data quality, and the tensor decomposition algorithm accurately extracts the bonding strength trend. The calculation of spatial correlation features provides a basis for calculating parameters between elements, and the feature fusion network fully integrates various features to obtain accurate interface stress characteristics. The extraction of time-domain and frequency-domain features, along with the calculation of combined feature vectors, provides crucial information for feature matching from different sensors and the calculation of bonding strength trends.

[0073] Example 2:

[0074] In determining the interfacial bonding characteristics of materials, it is necessary to model the thermal field distribution of the dispersion matrix based on the spherulite dispersion topology and interfacial stress characteristics. The specific implementation method is as follows:

[0075] Based on the spherulite dispersion topology, spherulite dispersion sampling points are extracted from each frame of data. The spherulite dispersion topology reflects the distribution morphology and structure of spherulites within the composite region. By analyzing each frame of data, the specific location of the spherulite dispersion can be accurately located, thus extracting the corresponding sampling points. These sampling points contain key information about the spherulite distribution and form the basis for subsequent analysis. Next, these sampling points are correlated with interface stress characteristics. Interface stress characteristics describe the distribution and variation of stress at the interface. Correlating the sampling points with interface stress characteristics establishes a connection between spherulite distribution and stress distribution, thereby generating a thermal field distribution map. The thermal field distribution map visually displays the distribution state of the thermal field within the composite region. Then, the thermal field distribution maps collected by multiple sensors are spatially registered. Due to the different positions and acquisition angles of different sensors, the collected thermal field distribution maps may have spatial differences. Spatial registration unifies these maps to the same spatial coordinate system, ensuring data consistency and comparability. Finally, the thermal field distribution intensity of the region is calculated. This intensity value reflects the overall distribution of the thermal field within the entire composite region.

[0076] A defect threshold is set, which needs to be determined based on the material's performance requirements and the actual application scenario. It is an important criterion for determining the existence of interface defects. Defect sources are located based on the spherulite dispersion values ​​of a multi-frame dispersion matrix. Multi-frame data provides more comprehensive information, and analysis of this data allows for accurate determination of the defect source location. The defect intensity difference is calculated by comparing the defect intensity of the current frame with the set defect threshold. If the defect intensity difference is greater than or equal to the threshold, it indicates the presence of interface defects in that unit. When an interface defect is confirmed, thermocompression parameter constraints are applied to the current unit. These thermocompression parameters include pressure, temperature, and time. Adjusting these parameters can improve the interface defect situation. Based on the heat conduction model corresponding to the current unit, which describes the heat conduction process and laws within the unit, the thermal field distribution intensity of the current unit is iteratively corrected. Through continuous adjustment and optimization, the thermal field distribution intensity gradually approaches the ideal state. Finally, the thermal field compensation value for the defect region is calculated based on the correction results. This compensation value is used for subsequent thermal field compensation in the defect region to improve the interfacial bonding performance of the material.

[0077] Thermal field distribution modeling is performed on the dispersion matrix based on the intensity of the thermal field distribution. The intensity of the thermal field distribution is a crucial basis for modeling, reflecting the influence of the thermal field on the dispersion matrix. By modeling the thermal field distribution, the distribution of the thermal field can be incorporated into the dispersion matrix, allowing it to more comprehensively reflect the composite state of the material. Furthermore, stress annotation is performed on the regional thermal field model using interfacial stress characteristics. Interfacial stress characteristics are closely related to the thermal field distribution; annotating them on the thermal field model provides a more intuitive understanding of the relationship between the thermal field distribution and interfacial stress, offering more accurate information for subsequent analysis and control.

[0078] In the entire process of thermal field distribution modeling, each step is interconnected and mutually influential. Extracting spherulite dispersion sampling points and associating them with interface stress characteristics are the foundation for generating thermal field distribution maps. Spatial registration and calculation of thermal field distribution intensity ensure the accuracy and reliability of the thermal field distribution maps. Setting defect thresholds and locating defect sources are key to judging interface defects. Hot-pressing parameter constraint compensation and iterative correction of thermal field distribution intensity are important means to improve interface defects. Modeling the thermal field distribution based on thermal field distribution intensity and stress annotation through interface stress characteristics integrate thermal field and stress information into the model, enabling the model to more comprehensively reflect the interfacial bonding state of the material.

[0079] Example 3:

[0080] When calculating the thermal field distribution gradient, a series of operations need to be performed based on the location of the dispersion sensor. The specific implementation method is as follows:

[0081] Based on multiple sets of material composite data, points of change in thermal field intensity are extracted. These data sets contain thermal field information when materials are composited under different conditions. Analysis and processing of this data accurately identify points where significant changes in thermal field intensity occur. Then, these change points are mapped to a unified thermodynamic coordinate system based on the sensor deployment locations. Since the sensors are deployed at different locations on the hot-pressing mold, the change points at different locations need to be transformed into the same thermodynamic coordinate system to ensure comparability and uniformity. Next, the change points are fitted using a heat flux vector interpolation algorithm. This algorithm can infer the thermal field intensity value at unknown locations based on known change point data, thereby generating a thermal field distribution model for the region. This model visually displays the distribution of the thermal field throughout the entire region.

[0082] Equal-interval sampling is performed along the heat transfer path of the thermal field distribution model. The heat transfer path represents the main direction of heat transfer in the thermal field, and equal-interval sampling along this path allows for the acquisition of detailed information about the thermal field at different locations. Based on the sampling results, the thermal attenuation rate, stress fluctuation index, and thermal field change slope of the path are calculated. The thermal attenuation rate reflects the degree of heat attenuation during transfer, the stress fluctuation index represents the stress fluctuation along the heat transfer path, and the thermal field change slope reflects the rate of change of thermal field intensity with location. Then, thermal field variation parameters are calculated based on the thermal attenuation rate, stress fluctuation index, and thermal field change slope. These parameters comprehensively consider various variation characteristics of the thermal field along the heat transfer path, providing a more comprehensive description of the thermal field's changes.

[0083] Based on the deployment parameters and acquisition accuracy of the dispersion sensors, the distribution characteristics of interfacial bonding strength in each frame of data are projected onto the thermal field distribution model. The deployment parameters of the dispersion sensors determine the location and range of their data acquisition, while the acquisition accuracy affects the accuracy of the data. Projecting the distribution characteristics of interfacial bonding strength onto the thermal field distribution model establishes a relationship between interfacial bonding strength and thermal field distribution. The thermal field distribution model is divided into zones along the heat transfer direction based on the number of sensors. The number of sensors determines the number and range of zones, allowing for more detailed analysis of the changes in interfacial bonding strength in different regions. Analyzing the variation patterns of interfacial bonding strength within each zone reveals that the bonding strength in each zone may exhibit different trends. By analyzing these trends, the variation patterns can be summarized. Finally, the bonding distribution characteristics are calculated based on the variation patterns, reflecting the distribution of interfacial bonding strength throughout the entire thermal field distribution model.

[0084] The thermal field distribution gradient is calculated based on the thermal field change parameters and combined distribution characteristics. The calculation process is as follows: Based on the thermodynamic coordinate range from the first to the last dispersion sensor, thermal field coordinate points are selected along the sensor deployment direction. These coordinate points represent different positions along the sensor deployment direction. The product of the thermal field intensity characteristic weight value and the combined distribution characteristic weight value within the spatial resolution range is accumulated. The spatial resolution determines the accuracy and range of the calculation. The thermal field intensity characteristic weight value and the combined distribution characteristic weight value represent the importance of the thermal field intensity characteristic and the combined distribution characteristic in the calculation, respectively, and their product reflects the combined influence of these two factors. The influence of the sensor acquisition frequency on the rate of change of thermal field intensity is superimposed. The higher the sensor acquisition frequency, the more accurate the monitoring of the rate of change of thermal field intensity; therefore, its influence needs to be included in the calculation.

[0085] When calculating the gradient of the thermal field distribution, a formula is involved:

[0086]

[0087] in, This represents the gradient of the thermal field distribution. It is the final calculated result and is used to describe the spatial distribution and variation of the thermal field. This indicates the number of thermal field coordinate points selected in the sensor deployment direction; the number of coordinate points determines the level of detail in the calculation. Indicates the first The weight value of the thermal field intensity characteristics at each coordinate point is determined based on the importance of the thermal field intensity characteristics, reflecting the degree of influence of the thermal field intensity at that coordinate point. Indicates the first The thermal field intensity value at each coordinate point is the actual thermal field intensity at that coordinate point obtained through the thermal field distribution model. Indicates the first The weight value of the combined distribution characteristics at each coordinate point is determined based on the importance of the combined distribution characteristics, reflecting the magnitude of the combined distribution characteristics' effect at that coordinate point. Indicates the first The bonding distribution characteristic value at each coordinate point is obtained by analyzing the bonding strength distribution characteristics of the interface at that coordinate point. This indicates the sensor's data acquisition frequency, which reflects how quickly the sensor acquires data. This represents the rate of change of thermal field intensity, which describes how the thermal field intensity changes over time.

[0088] Example 4:

[0089] When calculating the bonding prediction information for each sub-unit, it is necessary to use the thermal field distribution model as a basis and combine it with the thermal field distribution gradient to predict the development trend of the interface bonding strength. The specific implementation method is as follows:

[0090] The heat transfer principal diameter of the thermal field distribution model is used as the baseline. The heat transfer principal diameter is the main path for heat transfer in the thermal field, and its direction and distribution reflect the core characteristics of the thermal field. The peak position of the interfacial bonding strength in each frame of data is used as a reference point. The peak position represents the maximum interfacial bonding strength in that frame of data and has important reference significance. The spatial distribution difference of the interfacial bonding strength is measured by calculating the bonding offset, i.e., the deviation between the peak position of the current frame and the heat transfer principal diameter. For example, assuming the heat transfer principal diameter of the thermal field distribution model is a straight line along the mold length, if the peak of the interfacial bonding strength in a certain frame of data appears 5 mm away from the principal diameter, then the bonding offset is 5 mm. Subsequently, a bonding distribution curve is plotted using thermodynamic coordinates. Thermodynamic coordinates comprehensively consider the factors of temperature and spatial location, accurately locating the position of each data point in the thermal field, thus converting the peak position and bonding offset of each frame into a curve, intuitively showing the distribution trend of the interfacial bonding strength in the thermal field.

[0091] The growth rate and direction of the bonding strength trend are corrected based on the thermal field distribution gradient. The thermal field distribution gradient reflects the rate of change of thermal field intensity in space and has a significant impact on the development of interfacial bonding strength. For example, a large thermal field distribution gradient means that the thermal field intensity changes significantly over short distances, which may accelerate the growth of interfacial bonding strength or change its growth direction. Suppose the original predicted growth rate of the bonding strength trend is 10% per unit time, and the direction is positive along the main heat transfer path. However, due to the influence of the thermal field distribution gradient, the actual growth rate may be adjusted to 12%, and the direction may be slightly biased towards one side of the main heat transfer path. This correction requires comprehensive consideration of the magnitude and direction of the thermal field distribution gradient, as well as the characteristics of the bonding strength itself, to ensure that the predicted results are more consistent with reality.

[0092] Starting from the most recent bonding distribution point, the distribution curve is continued to be plotted based on the correction results for the growth rate and direction. The most recent bonding distribution point is obtained based on the latest detection data and can reflect the actual state of the current interface bonding strength. For example, if the most recent bonding distribution point is located at thermodynamic coordinates (10, 20, 30), the bonding strength is 80 MPa, the corrected growth rate is 12% per unit time, and the direction is 20 degrees positively offset from the main heat transfer diameter, then the bonding distribution point at the next time point can be calculated as follows: based on the original coordinates, move a certain distance along the corrected direction, the distance of which is determined by the growth rate and the time interval. Continue this process to generate the bonding distribution point for the next time period until the distribution point covers the entire target area, thereby generating complete bonding prediction information. The target area may be the entire composite area of ​​the hot pressing mold, or it may be a specific sub-unit, depending on the actual application requirements.

[0093] Throughout the calculation process, the accuracy of the thermal field distribution model is crucial. It needs to be generated based on multiple sets of composite material data and real-time data collected by sensors to ensure a true reflection of the thermal field distribution. For example, when fitting the thermal field intensity variation points using a heat flux vector interpolation algorithm, the sensor deployment location and acquisition accuracy must be fully considered to avoid model errors. Simultaneously, the correction of the intensity trend requires comprehensive consideration of factors such as the thermal field distribution gradient and whisker orientation. Whisker orientation affects the mechanical properties and interfacial bonding characteristics of the material. When the thermal field distribution gradient interacts with whisker orientation, it may have a complex impact on the growth rate and direction of the bonding strength, thus requiring full consideration during the correction process.

[0094] The acquisition frequency and accuracy of each frame of data also affect the accuracy of the combined prediction information. A higher acquisition frequency can capture rapid changes in the thermal field and interface bonding strength, while high-precision acquisition data ensures the reliability of the calculation results. For example, if the sensor's acquisition frequency is 10 times per second, it can promptly detect sudden changes in the thermal field distribution gradient and adjust the bonding strength trend accordingly, thereby improving the real-time performance and accuracy of the prediction.

[0095] Example 5:

[0096] When constructing a coupled control model, obtaining material interface bonding characteristics, and dynamically controlling regional material composite parameters, it is necessary to achieve optimized control of the composite material interface through multiple steps. The specific implementation method is as follows:

[0097] When constructing the coupling control model, the input layer is used to organize the binding prediction information into spatially distributed data and perform normalization processing. For example, assuming that the binding prediction information of a certain sub-unit at different locations is obtained through the previous steps, such as the binding strength prediction value of 75 MPa at coordinates (10, 20, 30) and 80 MPa at (15, 25, 35), the input layer will organize these coordinates and corresponding strength values ​​into spatially distributed data, and then perform normalization processing to convert the strength values ​​of different orders of magnitude into values ​​between 0 and 1 for subsequent processing.

[0098] The feature fusion layer is used to process spatially distributed data, extract regional correlation features of interface bonding, and construct dependencies between material units. For example, in a hot press mold, the interface bonding strength of adjacent sub-units may influence each other. The feature fusion layer analyzes the spatial distribution data of these sub-units to find the pattern of strength changes in adjacent sub-units when the strength of a certain sub-unit increases, thereby constructing a dependency model between units. For instance, when the bonding strength of sub-unit A increases, the strength of its adjacent sub-unit B to its right may increase by 0.8 times accordingly; the feature fusion layer captures this correlation feature.

[0099] The parameter control layer is used to integrate the correlation between the interface bonding on the spatial units and generate a material composite parameter control strategy. For example, if the feature fusion layer finds a positive correlation between the bonding strength of sub-unit C and sub-unit D, and the current bonding strength of sub-unit C is not as expected, the parameter control layer will generate a strategy to adjust the hot-pressing temperature and pressure of the regions where sub-units C and D are located, such as increasing the hot-pressing temperature of sub-unit C by 5°C and increasing the pressure of sub-unit D by 10 MPa to promote the interface bonding between the two.

[0100] When acquiring material interface bonding characteristics, the identifiers of sub-units are mapped to the topological distribution characteristics of the material interface bonding output by the coupling control model. For example, if the topological distribution characteristics of a certain region output by the coupling control model are displayed as a grid-like high-intensity distribution, this feature is mapped to the sub-unit identified as "Region A" to clarify the interface bonding state of that sub-unit.

[0101] The element data in the material state dataset is reorganized according to topological features to generate an element distribution map sorted by interface bonding strength. For example, if the material state dataset contains bonding strength data for multiple sub-elements, such as sub-element 1 with a strength of 60 MPa, sub-element 2 with 70 MPa, and sub-element 3 with 85 MPa, after reorganization according to topological features, these sub-elements are arranged from high to low strength to generate an element distribution map that visually displays the differences in bonding strength among the sub-elements.

[0102] Based on the reconstructed unit distribution diagram, the optimized interface bonding distribution characteristics are output. For example, it can be seen from the unit distribution diagram that the sub-unit bonding strength in the central region of the hot pressing mold is generally higher, while that in the edge region is lower. Therefore, the optimized characteristics output are that the interface bonding in the central region is good, while the edge region needs further adjustment, providing a basis for subsequent execution actions.

[0103] When dynamically controlling the composite parameters of regional materials, a hot-pressing power increase command is triggered in the adjacent region when the interfacial bonding in the target region reaches a preset strength threshold. For example, if the interfacial bonding strength threshold of the target region is set to 80 MPa, when the strength of a sub-unit in that region is detected to reach 80 MPa, the hot-pressing power of its adjacent region is increased by 15% to promote the synchronous improvement of interfacial bonding in the adjacent region.

[0104] Based on the whisker orientation strategy, hot-pressing parameters are dynamically combined to generate a holding time vector. For example, if the whisker orientation strategy requires the whiskers to be more closely packed in a certain direction, the hot-pressing temperature is dynamically combined to be 180℃ and the pressure to be 15MPa. A holding time vector is generated based on these parameters, such as holding at 180℃ for 20 minutes or at 15MPa pressure for 15 minutes, to ensure that the whiskers are aligned in the expected orientation.

[0105] The mold parameters of the hot-pressing node in the target area are adjusted based on the heat preservation time vector. For example, if the heat preservation time vector indicates that heat preservation at 180℃ for 20 minutes is required, then the mold temperature parameter of the hot-pressing node in the target area is adjusted to 180℃, the heat preservation time is set to 20 minutes, and the pressure parameter is adjusted to 15MPa. This ensures that the mold operates according to the set parameters during the composite process, achieving precise control of the material interface bonding.

[0106] It should be noted that, in this document, relational terms such as "first" and "second" are used only 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 process, method, article, or apparatus.

[0107] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An interface optimization method for hollow micron-sized silica spheres-aluminum nitride whisker composite materials, applied to a material composite system, the system comprising multiple dispersion sensors deployed in a hot-pressing mold and an interface control system connecting the dispersion sensors, wherein adjacent dispersion sensors are spaced apart by a predetermined distance, characterized in that... The method includes: The dispersion sensor is used to collect spherulite dispersion data and interfacial stress data within the composite region to generate a material state dataset. The interface control system receives real-time material state data from multiple dispersion sensors to construct a regional dispersion matrix. Based on the material state dataset and real-time data from each dispersion sensor, the material interface bonding characteristics are determined. The determination of the material interface bonding characteristics includes: processing the dispersion matrix, extracting interface features from the material state dataset, predicting interface bonding based on the thermal field distribution gradient and whisker orientation information, outputting the topological distribution characteristics of the material interface bonding through a coupling control model, and updating the material state dataset based on the topological distribution characteristics. Based on the aforementioned bonding characteristics, the composite parameters of the regional materials are dynamically adjusted. The characteristic is that determining the material interface bonding features includes: The dispersion matrix is ​​processed to extract spherulite dispersion topology, interfacial stress characteristics, and bonding strength trends; Based on the spherulite dispersion topology and interfacial stress characteristics, the dispersion matrix is ​​modeled for thermal field distribution. The composite region is divided into multiple sub-units and the unit identifiers are marked. The spherulite dispersion of the sub-units is correlated and matched with the material state dataset, and the unit identifiers are marked in the material state dataset. The thermal field distribution gradient is calculated based on the location of the dispersion sensor. The interface bonding distribution is predicted based on the thermal field distribution gradient and the bonding strength trend. The bonding prediction information for each sub-unit is calculated. A coupling control model is constructed, and the predicted information is used as the input parameter of the coupling control model. The coupling control model is used to perform spatial correlation modeling on the predicted information and output the topological distribution characteristics of the material interface bonding. Update the material state dataset based on the topological distribution characteristics to obtain the material interface binding characteristics.

2. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 1, characterized in that, The processing of the dispersion matrix includes: The dispersion matrix is ​​normalized, and the stress concentration region in the matrix is ​​truncated by a sliding window. The noise in the concentration region is suppressed, and the combined strength trend is calculated by tensor decomposition algorithm. The spatial correlation characteristics of the dispersion matrix are calculated. Based on the spatial correlation characteristics, the stress interference coefficient, whisker stability index and interface defect degree between units are calculated. A feature fusion network is constructed, and the interface stress characteristics are calculated through the feature fusion network. Extract the time-domain and frequency-domain features collected by each dispersion sensor, calculate the combined feature vector of the sensor based on the phase difference between the time-domain and frequency-domain features, and perform feature matching on dispersion sensors at different locations based on the combined feature vector to calculate the trend of the combined strength.

3. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 2, characterized in that, The thermal field distribution modeling of the dispersion matrix includes: Based on the spherulite dispersion topology, spherulite dispersion sampling points are extracted from each frame of data. The sampling points are correlated and mapped with the interface stress characteristics to generate a thermal field distribution map. The thermal field distribution maps collected by multiple sensors are spatially registered to calculate the thermal field distribution intensity of the region. Set a defect threshold, locate the defect source based on the spherulite dispersion value of the multi-frame dispersion matrix, calculate the defect intensity difference, if the defect intensity difference is greater than or equal to the defect threshold, it indicates that there is an interface defect in the unit, perform hot-pressing parameter constraint compensation on the current unit, iteratively correct the thermal field distribution intensity of the current unit according to the heat conduction model corresponding to the current unit, and calculate the thermal field compensation value of the defect area based on the correction result. The thermal field distribution is modeled based on the dispersion matrix according to the thermal field distribution intensity, and the thermal field model of the region is stress-labeled by the interface stress characteristics.

4. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 3, characterized in that, The calculation of the thermal field distribution gradient based on the location of the dispersion sensor includes: Based on multiple sets of composite material data, the points of change in thermal field intensity are extracted, and the points of change are mapped to a unified thermodynamic coordinate system according to the deployment location of the sensors. The points of change are then fitted using a heat flux vector interpolation algorithm to generate a thermal field distribution model for the region. Equal-interval sampling is performed along the heat transfer path of the thermal field distribution model. The thermal attenuation rate, stress fluctuation index, and thermal field change slope of the path are calculated based on the sampling results. The thermal field change parameters are calculated based on the thermal attenuation rate, stress fluctuation index, and thermal field change slope. Based on the deployment parameters and acquisition accuracy of the dispersion sensor, the distribution characteristics of the interface bonding strength in each frame of data are projected onto the thermal field distribution model. The model is divided into zones according to the number of sensors along the heat transfer direction. The variation law of the interface bonding strength within the zone is analyzed, and the bonding distribution characteristics are calculated based on the variation law.

5. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 4, characterized in that, The step of calculating the thermal field distribution gradient based on the location of the dispersion sensor further includes: selecting thermal field coordinate points in the sensor deployment direction based on the thermodynamic coordinate range from the first dispersion sensor to the last dispersion sensor, accumulating the product of the thermal field intensity characteristic weight value and the combined distribution characteristic weight value within the spatial resolution range, and superimposing the influence value of the sensor acquisition frequency on the rate of change of thermal field intensity.

6. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 4, characterized in that, The calculation of the combined prediction information for each sub-unit includes: Using the heat transfer main diameter of the aforementioned thermal field distribution model as the baseline, and taking the peak position of the interface bonding strength in each frame of data as the reference point, the bonding offset is calculated, and the bonding distribution curve is plotted according to thermodynamic coordinates. Based on the thermal field distribution gradient, the growth rate and direction in the bonding strength trend are corrected; Starting from the most recent combined distribution point, the distribution curve is continued to be plotted based on the correction results of the growth rate and direction, generating the next period's combined distribution point, until the distribution points cover the entire target area, generating combined prediction information.

7. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 1, characterized in that, The construction of the coupled regulation model includes: The input layer is used to organize the combined prediction information into spatially distributed data and perform normalization processing. The feature fusion layer is used to extract regional correlation features of interface bonding by processing spatially distributed data, and to construct the dependency relationship between material units. The parameter control layer is used to integrate the relationship between the interface and the spatial unit to generate a material composite parameter control strategy.

8. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 1, characterized in that, The acquired material interface bonding features include: Based on the topological distribution characteristics of the material interface combination output by the coupling control model, the identifiers of the sub-units are mapped to the topological distribution characteristics. The element data in the material state dataset is reorganized according to topological features to generate an element distribution map sorted by interface bonding strength; Based on the reorganized unit distribution diagram, the optimized interface combination distribution characteristics are output.

9. The interface optimization method for a hollow micron-sized silica sphere-alumina whisker composite material according to claim 1, characterized in that, The dynamic control of regional material composite parameters includes: When the interface reaches a preset intensity threshold in the target area, a hot pressing power increase command is triggered in the adjacent area. Based on the whisker orientation strategy, hot pressing parameters are dynamically combined to generate a heat preservation time vector; Adjust the mold parameters of the hot-pressing node in the target area based on the heat preservation time vector.

Images