A computer-aided design optimization method and system for injection mold structure

Abstract

The application relates to a computer-aided design optimization method and system for an injection mold structure. The method comprises: obtaining material flow information of an injection mold in an actual production process; comparing the material flow information with preset material flow prediction information to identify differences; in response to the differences, calculating actual flow resistance of the material in a specific area according to the material flow information; adjusting parameters of a plastic flow model according to the actual flow resistance; and using the adjusted plastic flow model to design and optimize the injection mold structure. Through a closed-loop mechanism of model self-adaptive correction driven by actual production feedback and design iterative optimization, the application realizes continuous learning and design improvement of the injection mold design when facing new engineering plastics.

Description

Technical Field

[0001] This application relates to the field of computer-aided design technology for injection molds, and in particular to a method and system for optimizing the computer-aided design of injection mold structures. Background Technology

[0002] In modern industrial production, the design and optimization of injection molds are crucial for manufacturing high-precision, complex-structured plastic products. Traditional computer-aided design systems perform well when handling conventional materials and structures, but their limitations become increasingly apparent when faced with emerging high-performance engineering plastics and increasingly complex mold structures. These new materials often possess unique flow characteristics, and the material databases within existing design systems often lack accurate records of this cutting-edge data. This leads to significant discrepancies between the predictions made by the design system when simulating the plastic filling process and actual production conditions, ultimately resulting in product defects.

[0003] Due to the lack of data on the flow characteristics of novel plastics in the database, fluid motion calculation programs used to simulate the plastic filling process cannot accurately capture the real flow behavior of molten plastic in complex channels and thin-walled regions during simulation. When the melt experiences high shear within the mold, its flow resistance changes may deviate significantly from the system's preset model, resulting in simulation results that cannot accurately reflect the melt front's propulsion speed, pressure distribution, and the generation of local shear heat. This inaccurate input data leads the system to generate channel designs that appear optimal in the simulation environment but may result in incomplete filling or excessive shear heat in actual filling processes.

[0004] After manufacturing the mold based on the simulation results and putting it into trial production, quality defects were frequently found in certain specific areas of the product during the actual injection molding process. To solve these product defects, the production department usually tries to optimize by adjusting the injection molding process parameters. However, in problems caused by defects in the mold structure design, the effective range of process adjustments is often extremely narrow. After a long period of repeated debugging, the production team often finds that they cannot find a set of process parameters that can stably produce qualified products. This forces the engineering team to use costly and time-consuming manual trial and error methods to modify the mold, seriously delaying the project cycle and significantly increasing the production budget.

[0005] In summary, existing technologies lack a mechanism to adaptively correct plastic flow models and drive mold structure design optimization by utilizing material flow feedback information from actual production processes. This results in inaccurate simulation predictions of novel plastics in complex mold structures, long mold development cycles, and high costs, which are urgent problems to be solved in this technical field. Summary of the Invention

[0006] This application provides a computer-aided design optimization method and system for injection mold structure, which at least solves the problems of large prediction deviations and long development cycles in mold design caused by inaccurate plastic flow models in the prior art.

[0007] In a first aspect, this application provides a computer-aided design optimization method for injection mold structure, comprising the following steps: Obtain material flow information during the actual production process of injection molds; The material flow information is compared with preset material flow prediction information to identify the differences between the material flow information and the material flow prediction information; In response to the difference, the actual flow resistance of the material in a specific region is calculated based on the material flow information; Adjust the parameters of the plastic flow model based on the actual flow resistance. The adjusted plastic flow model was used to optimize the design of the injection mold structure.

[0008] Optionally, the step of estimating the actual flow resistance of the material in a specific region based on the material flow information in response to the difference includes: The local shear rate and local temperature of the specific area are continuously monitored, and when the local shear rate and the local temperature simultaneously exceed the preset material degradation critical threshold, an early warning is issued that the specific area is a potentially high-risk area for degradation. Calculate the actual effective viscosity of the specific region; Based on the local shear rate and local temperature of the specific region, predict the expected viscosity of the specific region; By comparing the actual effective viscosity with the expected viscosity, when the actual effective viscosity is lower than the expected viscosity and the decrease exceeds a preset abnormal drop threshold, and an early warning has been triggered in the specific area, it is confirmed that the material has degraded. When the material is confirmed to have degraded, the actual effective viscosity data of the degradation is isolated.

[0009] Optionally, the step of calculating the actual effective viscosity of the specific region includes: Identify the mold geometry in the specific region, and detect wall slippage in the specific region when it is an extremely thin-walled region; Based on the wall slip phenomenon, the boundary conditions in the simplified fluid dynamics formula are adjusted to introduce an equivalent slip length or slip velocity term, thereby correcting the deviation in the flow resistance calculation. Monitor the local temperature change rate and shear rate change rate in the specific region, and evaluate the viscoelastic effect of the material based on the local temperature change rate and shear rate change rate. Based on the viscoelastic effect, a time-dependent or shear history-dependent correction factor is introduced to dynamically adjust the viscosity term in the simplified fluid dynamics formula, thereby compensating for the influence of material viscoelasticity on flow resistance. The actual effective viscosity of the specific region is calculated based on the adjusted simplified fluid dynamics formula.

[0010] Optionally, the step of adjusting the parameters of the plastic flow model based on the actual flow resistance includes: Identify whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise. The identification includes distinguishing the types of deviations that cause the sparsity or measurement noise. The types of deviations include deviations caused by inaccurate initial parameters of the plastic flow model and deviations caused by the dynamic evolution of the internal structure of the material. When the actual flow information is identified as sparse or contains measurement noise, data augmentation is performed based on the actual flow information to generate a set of virtual data points; Based on the actual flow information, the intensity of the measurement noise is evaluated, and the actual flow resistance data is weighted according to the intensity of the measurement noise. A regularization-based optimization strategy is adopted to adjust the parameters of the plastic flow model based on the weighted actual flow resistance data and the virtual data points. The learning rate of the plastic flow model parameters is dynamically adjusted based on whether the actual flow information is sparse or contains measurement noise.

[0011] Optionally, the step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise includes: A sensor array is deployed to collect local pressure, melt front velocity, and temperature data in specific areas of the injection mold; Spatiotemporal correlation analysis was performed on the collected data to extract the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micro-path, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature as flow characteristics. The actual measured flow characteristics are compared with the flow characteristics predicted by the plastic flow inside the system to identify whether the differences between the two in multiple dimensions continue to exceed the preset small difference threshold, thereby identifying potential deviations in rheological behavior. For the identified potential deviations in rheological behavior, statistical analysis was performed over multiple consecutive injection molding cycles to calculate the mean deviation and standard deviation. Based on the average deviation and the standard deviation, identify the deviation caused by inaccurate initial parameters of the plastic flow model; Based on the correlation between the average deviation, the standard deviation, the local cooling rate of the specific region, and the shear rate of the specific region, the deviation caused by the dynamic evolution of the internal structure of the material is identified; Based on the attribution results of the deviation, the sparsity or measurement noise of the actual flow information is identified.

[0012] Optionally, in the step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise, the step of performing spatiotemporal correlation analysis on the collected local pressure, melt front velocity, and temperature data to extract flow characteristics includes: Based on the spatial distribution range and time span of the collected local pressure, melt front velocity, and temperature data, a multi-scale spatiotemporal grid is established. The collected local pressure, melt front velocity, and temperature data are mapped onto the multi-scale spatiotemporal grid. The interpolation process within the multi-scale spatiotemporal grid is used to perform spatial smoothing and temporal synchronization on the mapped data. Within a unified spatiotemporal framework, the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature are calculated.

[0013] Optionally, the step of calculating the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature within a unified spatiotemporal framework includes: Analyze the internal geometry of the mold to identify complex regions, including microchannels, sharp bends, and flow dividers. Based on the geometric characteristics of the complex region, the sampling density and time window length of local pressure, melt front velocity and temperature data are adjusted, and the acquisition frequency of sensor data is increased and the time window of data analysis is shortened within the complex region; For the microchannel region, the average velocity of the melt front at different cross sections is calculated, and the distribution of local shear velocity is inferred by combining the changes in the channel cross section; for the sharp bend region, the pressure difference between the inside and outside of the bend of the pressure sensor array and its change over time is analyzed; for the splitter region, the arrival time of the melt front on different split paths and the pressure drop on each path are compared. Based on the local flow information of the microchannel region, the sharp bend region, and the splitter region, the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature are calculated.

[0014] Optionally, in the step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise, the step of identifying the deviation caused by the dynamic evolution of the material's internal structure based on the fluctuation correlation of the average deviation, the standard deviation, the local cooling rate of the specific region, and the shear rate of the specific region includes: In the specific area, miniature heat flow sensors and miniature shear force sensors are deployed to collect instantaneous fluctuation data of local cooling rate and shear force rate, and to perform spectral analysis on the collected instantaneous fluctuation data to extract its energy distribution characteristics in different frequency ranges. In the specific region, miniature ultrasonic sensors are deployed to collect data on the ultrasonic attenuation coefficient and sound velocity variation inside the material, and time-domain and frequency-domain analyses are performed on the collected ultrasonic attenuation coefficient and sound velocity variation data to extract features related to the internal microstructure of the material. A correlation mapping is established based on the material type and temperature range of the specific region. The correlation mapping associates the spectral energy distribution characteristics of the local cooling rate and the shear rate with the ultrasonic attenuation coefficient and the sound velocity variation characteristics. When the spectral energy distribution characteristics of the local cooling rate and the shearing rate change in a specific pattern, and the change corresponds to the specific pattern change of the ultrasonic attenuation coefficient and the sound velocity change characteristics, it is identified as a deviation caused by the dynamic evolution of the internal structure of the material.

[0015] Optionally, the step of establishing the association mapping includes: Obtain the current injection molding process parameters, and based on the process parameters, retrieve the association patterns that match the process parameters; Based on the correlation mode, the mapping relationship between the spectral energy distribution characteristics of the local cooling rate and the shearing velocity and the ultrasonic attenuation coefficient and the sound velocity change characteristics is dynamically adjusted.

[0016] Secondly, this application also provides a computer-aided design optimization system for injection mold structures, the system comprising: The information acquisition module is used to acquire material flow information of the injection mold during the actual production process; The difference recognition module is used to compare the material flow information with preset material flow prediction information to identify the difference between the material flow information and the material flow prediction information; A resistance calculation module is used to calculate the actual flow resistance of the material in a specific region based on the material flow information in response to the difference. The parameter adjustment module is used to adjust the parameters of the plastic flow model according to the actual flow resistance. The design optimization module is used to optimize the design of the injection mold structure using the adjusted plastic flow model.

[0017] Compared with related technologies, the computer-aided design optimization method and system for injection mold structure provided in this application have at least the following technical advantages: By acquiring material flow information from the actual production process of injection molds and comparing it with preset prediction information, the mold design is upgraded from a one-time simulation relying solely on prior material models to a closed-loop optimization based on actual production feedback, overcoming prediction bias caused by the lack of data for novel plastic materials. Subsequently, by calculating the actual flow resistance of the material in specific areas, the flow differences observed in actual production are quantified into driveable model correction parameters, achieving automated diagnosis from defect discovery to root cause localization. Furthermore, through adaptive correction and regularization optimization strategies for the plastic flow model based on actual flow resistance, the model can continuously learn from production practice and improve the accuracy of predicting the flow characteristics of novel materials. Finally, by re-optimizing the mold structure design using the corrected model, a complete closed loop from production feedback to design improvement is formed, significantly shortening the cycle from mold design to stable production.

[0018] In summary, this application achieves continuous learning and adaptive improvement in injection mold design through a multi-layered closed loop, which includes feedback from actual production material flow, reverse calculation of local effective flow resistance, adaptive correction of plastic flow model, and iterative optimization of mold structure. This improves the accuracy of design prediction and reduces trial and error costs.

[0019] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating a computer-aided design optimization method for injection mold structure according to an exemplary embodiment.

[0021] Figure 2 This is a flowchart illustrating step S3 according to an exemplary embodiment.

[0022] Figure 3 This is a flowchart illustrating step S32 according to an exemplary embodiment.

[0023] Figure 4 This is a flowchart illustrating step S4 according to an exemplary embodiment.

[0024] Figure 5 This is a flowchart illustrating step S412 according to an exemplary embodiment.

[0025] Figure 6 This is a block diagram illustrating a computer-aided design optimization system for injection mold structure according to another exemplary embodiment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0027] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any creative effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of this application.

[0028] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0029] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0030] Example 1 This application provides a computer-aided design optimization method for injection mold structure. Figure 1 This is a flowchart illustrating a computer-aided design optimization method for injection mold structures according to an exemplary embodiment. Figure 1 As shown, the method includes the following steps: S1. Obtain material flow information of the injection mold during the actual production process; In this embodiment, a micro-sensor array is deployed in key areas of the mold cavity to acquire material flow information during the actual production process. The sensors include micro-pressure sensors and melt front velocity sensors located at the gate, runner branches, and the cavity filling end. The pressure sensors are piezoelectric micro-sensors that record local pressure changes in real time as the melt passes through the sensor positions at a sampling frequency of at least 1 kHz. The melt front velocity sensors calculate the local flow velocity by detecting the time difference between the arrival times of the melt at each sensor. Simultaneously, the system also collects process parameters such as screw position, injection speed, and holding pressure provided by the injection molding machine, and correlates these with the sensor data to form a complete material flow information dataset.

[0031] S2. Compare the material flow information with the preset material flow prediction information to identify the differences between the material flow information and the material flow prediction information; In this embodiment, the system pre-generates material flow prediction information for the mold filling process based on currently used plastic flow models, such as the Cross-WLF viscosity model, and the mold's geometric design files, through computational fluid dynamics simulation. This includes the expected time for the melt front to reach each sensor location, the expected pressure distribution curve, and the expected temperature change trajectory. The system then compares the material flow information collected by sensors during actual production with this prediction information point-by-point, calculating differences in multiple dimensions such as time deviation, pressure deviation, and temperature deviation. When the difference in any dimension exceeds a preset threshold, the system marks it as a significant difference region requiring further analysis.

[0032] S3. In response to the difference, calculate the actual flow resistance of the material in a specific region based on the material flow information; In this embodiment, for specific regions where significant differences are identified, the system uses simplified fluid dynamics relationships to inversely calculate the actual flow resistance based on the mold geometry (such as channel length and cross-sectional area) of that region, as well as the local pressure difference and melt front velocity actually measured by the sensor. For a channel with an approximate rectangular cross-section, its flow resistance can be approximately expressed as being directly proportional to the effective viscosity, channel length, and flow velocity, and inversely proportional to the square of the cross-sectional height. By inversely solving using known geometric parameters and measured flow data, the actual effective viscosity value of that region at a specific shear rate and temperature is obtained. This value reflects the flow characteristics of the novel plastic under real production conditions, providing a comparison with potentially incomplete or biased predicted values ​​in the original material information database.

[0033] S4. Adjust the parameters of the plastic flow model according to the actual flow resistance; In this embodiment, the actual effective viscosity data points obtained from multiple specific regions in step S3, each corresponding to a specific shear rate and temperature, are used as the basis for correction. The system employs a regularization-based optimization strategy to minimize the difference between the model-predicted viscosity and the actual effective viscosity, adjusting the nonlinear parameters in the Cross-WLF model. The regularization term prevents the model from overfitting under sparse data conditions. Simultaneously, the system dynamically adjusts the learning rate of parameter updates based on the quality of actual flow information: when actual data is sparse or noisy, the learning rate is reduced for conservative updates; when data is sufficient and reliable, the learning rate is increased for rapid convergence.

[0034] S5. Using the adjusted plastic flow model, optimize the design of the injection mold structure.

[0035] In this embodiment, the revised plastic flow model is reloaded into the mold structure optimization module. The optimization module re-evaluates the filling performance of the current mold design based on the updated model, including re-predicting the melt front advancement process, pressure distribution, and temperature field, and calculating potential defect indices such as incomplete filling probability and weld line location. Based on the new evaluation results, the optimization algorithm automatically adjusts mold structure parameters, such as runner diameter, gate location, and cooling system layout, to eliminate predicted defects and improve overall design performance. The optimized design is output as a new mold manufacturing or modification scheme.

[0036] The technical solution described above constructs a complete closed loop from production data acquisition to model adaptive correction and design iterative optimization. By deploying sensors in key areas of the mold to acquire actual material flow information, comparing it with prediction information based on existing models in multiple dimensions, and then using fluid dynamics to back-calculate the differences into local actual flow resistance, the parameters of the plastic flow model are adaptively adjusted. Finally, the corrected model is used to re-optimize the mold structure design, achieving a continuous improvement in the design prediction accuracy and development efficiency of injection molds when facing new engineering plastics.

[0037] In one possible design, Figure 2 This is a flowchart illustrating step S3 according to an exemplary embodiment. (Refer to...) Figure 2 Step S3, the step of calculating the actual flow resistance of the material in a specific region based on the material flow information in response to the difference, includes: S31. Continuously monitor the local shear rate and local temperature of the specific area, and when the local shear rate and the local temperature simultaneously exceed the preset material degradation critical threshold, warn that the specific area is a potential high-risk area for degradation. In this embodiment, an embedded controller deployed at the front end of a specific area of ​​the mold collects shear rate and temperature data in real time. When the local shear rate exceeds 10^5 per second and the local temperature exceeds 280 degrees Celsius, the system triggers an early warning, marks the area as a high-risk degradation zone, and records the warning timestamp and parameter values.

[0038] S32. Calculate the actual effective viscosity of the specific region; In this embodiment, simplified hydrodynamic relationships are used to inversely calculate the local effective viscosity. For thin-walled regions of an approximate flat-plate flow, the actual effective viscosity can be solved using formulas.

[0039] S33. Based on the local shear rate and local temperature of the specific region, predict the expected viscosity of the specific region; In this embodiment, the system uses the currently uncorrected Cross-WLF model parameters to calculate the expected viscosity value of the undegraded material under the shear rate and temperature conditions, which serves as a benchmark for comparison with the actual viscosity.

[0040] S34. Compare the actual effective viscosity with the expected viscosity. When the actual effective viscosity is lower than the expected viscosity and the decrease exceeds a preset abnormal drop threshold, and the specific area has triggered an early warning, confirm that the material has degraded. In this embodiment, the percentage decrease in actual viscosity relative to expected viscosity is calculated. When the decrease exceeds 30% and an extreme condition warning has been triggered in step S31 for that region, the system determines that the material in that region has degraded. Combining the extreme condition warning and the abnormal viscosity drop as dual conditions forms a reliable degradation confirmation mechanism.

[0041] S35. When it is confirmed that the material has degraded, isolate the actual effective viscosity data of the degradation.

[0042] In this embodiment, actual effective viscosity data points that confirm degradation are marked as anomalous data and excluded from the dataset used to drive the correction of plastic flow model parameters. Only actual effective viscosity data points that are not marked as degradation data will be adopted in the subsequent adaptive correction process of model parameters.

[0043] The technical solution of the above embodiments, by establishing a material degradation monitoring and data isolation mechanism, ensures that the data used for plastic flow model correction represents the inherent rheological properties of undegraded materials, avoids incorrect parameter adjustments caused by mistakenly incorporating degradation data into the model correction process, and ensures the accuracy and reliability of model correction.

[0044] In one example, during the injection molding of a complex part with thin-walled areas, a specific region of the mold experienced a sustained increase in local shear rate and temperature due to narrow runners and high-speed filling. Sensors deployed in this region detected local shear rates reaching 1000 s. -1 The local temperature reached 280℃, exceeding the preset material degradation critical threshold (shear rate threshold of 800s). -1 The temperature threshold is 270℃, therefore this area was flagged as a high-risk zone for potential degradation. The system also calculated the actual effective viscosity in this area to be 50 Pa·s, while the expected viscosity predicted by the material rheology model under current conditions was 80 Pa·s. The comparison revealed that the actual effective viscosity decreased by 37.5%, exceeding the preset abnormal drop threshold of 20%, and the system confirmed that degradation had occurred in this area. The system isolated the actual effective viscosity data collected during the degradation period in this area and excluded it from the plastic flow model parameter adjustment process.

[0045] In one possible design, Figure 3 This is a flowchart illustrating step S32 according to an exemplary embodiment. (Refer to...) Figure 3 In step S32, the step of calculating the actual effective viscosity of the specific region includes: S321. Identify the mold geometry of the specific region, and when the specific region is an extremely thin-walled region, detect the wall slippage phenomenon of the specific region; In this embodiment, the wall thickness data of the region in the mold design file is analyzed to determine whether it is an extremely thin-walled region. For regions with a thickness of less than 0.5 mm, before calculating the effective viscosity, the wall slip is detected by measuring the pressure-flow relationship to see if wall slip occurs, i.e., the flow state in which the velocity of the melt at the wall is not zero.

[0046] S322. Based on the wall slip phenomenon, adjust the boundary conditions in the simplified fluid dynamics formula to introduce an equivalent slip length or slip velocity term, thereby correcting the flow resistance calculation deviation. In this embodiment, when wall slip is detected, the conventional no-slip boundary condition is replaced with an equivalent slip boundary condition in the hydrodynamic formula for inversely solving the effective viscosity. An equivalent slip length parameter is introduced, which is calibrated based on the measured pressure-flow rate relationship.

[0047] S323. Monitor the local temperature change rate and shear rate change rate of the specific region, and evaluate the viscoelastic effect of the material based on the local temperature change rate and shear rate change rate. In this embodiment, local temperature and shear rate data are continuously collected, and their rates of change over time are calculated. When there is a significant phase lag between the rate of temperature change and the rate of change of shear rate, the material is determined to exhibit a viscoelastic effect, meaning that the material flow behavior depends not only on the current conditions but also on the shear history and temperature history.

[0048] S324. Based on the viscoelastic effect, a time-dependent or shear history-dependent correction factor is introduced to dynamically adjust the viscosity term in the simplified fluid dynamics formula, thereby compensating for the influence of material viscoelasticity on flow resistance. In this embodiment, based on the intensity of the viscoelastic effect obtained from the evaluation, a correction factor related to shear history and relaxation time is introduced into the viscosity term, so that the calculated effective viscosity can more accurately reflect the true flow characteristics of the material under unsteady conditions.

[0049] S325. Calculate the actual effective viscosity of the specific region based on the adjusted simplified fluid dynamics formula.

[0050] In this embodiment, the final effective viscosity is calculated by applying the complete formula after wall slip compensation and viscoelastic correction, and a high-precision viscosity value with multiple corrections is output.

[0051] The technical solutions of the above embodiments, through wall slip detection and boundary condition compensation, as well as the introduction of viscoelastic effect evaluation and correction factors, make the reverse calculation of effective viscosity more accurate under complex geometric conditions such as extremely thin walls, and avoid the calculation deviation caused by simplified fluid dynamics assumptions in high shear and microscale flow.

[0052] In one example, the design optimization of a precision medical device injection mold with microchannels (wall thickness less than 0.5 mm) is required. After the melt is injected into the microchannel region, the system identifies this region as an extremely thin-walled area and deploys micro-MEMS sensors to detect wall slippage. After confirming the existence of slippage by analyzing the anomaly in the melt velocity gradient near the wall, the system introduces an equivalent slip length term into the simplified fluid dynamics formula to correct the flow resistance calculation deviation. Simultaneously, the system continuously monitors the local temperature change rate and shear rate change rate in the microchannel region: temperature and shear rate rise rapidly during the rapid filling stage and decrease rapidly during the cooling stage. Based on these dynamic change rates, after evaluating the viscoelastic effect of the material, the system introduces time-dependent and shear history-dependent correction factors to dynamically adjust the viscosity term.

[0053] In one possible design, Figure 4 This is a flowchart illustrating step S4 according to an exemplary embodiment. (Refer to...) Figure 4 Step S4, the step of adjusting the parameters of the plastic flow model according to the actual flow resistance, includes: S41. Identify whether the actual flow information under the current injection molding process parameter combination is sparse or has measurement noise, wherein the identification includes distinguishing the type of deviation that causes the sparsity or measurement noise, the type of deviation includes deviation caused by inaccurate initial parameters of the plastic flow model, and deviation caused by the dynamic evolution of the internal structure of the material; In this embodiment, the number and distribution uniformity of the collected actual flow data points are evaluated using statistical analysis methods. When the number of data points is lower than a preset threshold or there is an abnormally large variance among the data points, it is determined that the data is sparse or that measurement noise exists.

[0054] S42. When it is identified that the actual flow information is sparse or there is measurement noise, data augmentation is performed based on the actual flow information to generate a set of virtual data points. In this embodiment, based on the distribution characteristics of existing actual data points, synthetic data points are generated by adding small random perturbations that conform to the noise statistical model, thereby increasing the amount of data used for model correction. The degree of data augmentation is proportional to the sparsity of the original data.

[0055] S43. Based on the actual flow information, assess the intensity of the measurement noise, and weight the actual flow resistance data according to the intensity of the measurement noise; In this embodiment, the signal-to-noise ratio estimate for each data point is calculated, and a weight proportional to the measurement reliability is assigned to each actual effective viscosity data point. High-noise data points are given lower weights to reduce their impact on model parameter adjustment.

[0056] S44. Using a regularization-based optimization strategy, adjust the parameters of the plastic flow model based on the weighted actual flow resistance data and the virtual data points. In this embodiment, an L2-regularized least-squares optimization strategy is applied, using the weighted actual data points and augmented data points as the fitting targets to adjust the parameters of the Cross-WLF model. The regularization parameters are automatically adjusted according to the data sparsity.

[0057] S45. Based on whether the actual flow information is sparse or contains measurement noise, dynamically adjust the learning rate of the plastic flow model parameters.

[0058] In this embodiment, under conditions of sparse or high noise, a smaller learning rate is used to update the model parameters slowly and conservatively; under conditions of sufficient data and low noise, a larger learning rate is used to quickly converge to the optimal parameters. The switching of the learning rate is based on a quantitative assessment of data sparsity and noise intensity.

[0059] The technical solution of the above embodiments, through the combination of data augmentation, noise-aware weighting and adaptive regularization optimization strategies, enables the plastic flow model to robustly perform parameter correction even when faced with sparse data or measurement noise in actual production, thus avoiding incorrect adjustment of model parameters due to insufficient data quality.

[0060] In one example, when designing an injection mold with a complex flow channel (containing multiple thin-walled regions and sharp turns), the mold structure limits the deployment of only a few sensors in a certain thin-walled region, resulting in sparse actual flow information. Simultaneously, electromagnetic interference from the production environment introduces measurement noise into some sensor data. After identifying the sparsity and noise, the system supplements the sparse data by generating virtual pressure / temperature data points using nearby data points and radial basis function interpolation. For noisy data, wavelet denoising is used to assess noise intensity, followed by weighted processing. L2 regularization is employed to prevent overfitting during model parameter tuning, and the learning rate is dynamically adjusted based on data quality.

[0061] In one possible design, step S41, which involves identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise, includes: S411. Deploy a sensor array to collect local pressure, melt front velocity, and temperature data in a specific area of ​​the injection mold; In this embodiment, an array of multiple types of sensors is deployed at key locations such as the gate, runner branch, and cavity filling end of the mold cavity to simultaneously collect pressure change curves, melt front arrival timestamps, and local temperature trajectories, forming a multi-dimensional, multi-location flow dataset.

[0062] S412. Perform spatiotemporal correlation analysis on the collected data, and extract the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity on the micro-path, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature as flow characteristics. In this embodiment, a spatiotemporal correlation analysis method is employed to analyze the pressure gradient and frontal velocity differences between adjacent sensors in the spatial dimension, and to track the changing trends and fluctuation patterns of various parameters in the temporal dimension. The extracted flow features constitute a multidimensional vector, describing the flow characteristics of each region within a specific time window.

[0063] S413. Compare the actual measured flow characteristics with the predicted flow characteristics of plastic flow inside the system to identify whether the differences between the two in multiple dimensions continue to exceed the preset small difference threshold, thereby identifying potential deviations in rheological behavior. In this embodiment, the actual flow feature vector is compared dimension-by-dimensionally with the predicted feature vector generated based on the current plastic flow model to calculate the Euclidean distance deviation. When the deviation continuously exceeds a preset threshold in multiple consecutive injection molding cycles, potential rheological deviations are identified.

[0064] S414. For the identified potential deviations in rheological behavior, perform statistical analysis over multiple consecutive injection molding cycles to calculate the mean deviation and standard deviation. In this embodiment, deviation data for each key region is collected from at least 100 consecutive injection molding cycles, and the average deviation and standard deviation of each region are calculated. A stable and consistent average deviation points to systematic model bias, while a high variance deviation points to random measurement noise.

[0065] S415. Identify the deviation caused by inaccurate initial parameters of the plastic flow model based on the average deviation and the standard deviation; In this embodiment, when the average deviation is relatively stable and not significantly correlated with fluctuations in injection molding process parameters, the deviation is attributed to inaccurate initial parameters of the plastic flow model. This type of deviation is a systematic deviation and is suitable for model parameter correction.

[0066] S416. Based on the fluctuation correlation between the average deviation, the standard deviation, the local cooling rate of the specific region, and the shear rate of the specific region, identify the deviation caused by the dynamic evolution of the internal structure of the material; In this embodiment, when the deviation mode exhibits high instability and is significantly correlated with fluctuations in local cooling rate or shear rate, the deviation is attributed to the dynamic evolution of the material's internal structure, such as changes in polymer chain orientation, alterations in crystallinity, or microdegradation. Signals of this type of deviation need to be excluded from the model correction data.

[0067] S417. Based on the attribution results of the deviation, identify the sparsity or measurement noise of the actual flow information.

[0068] In this embodiment, the results of the comprehensive deviation attribution analysis are used to determine the type and degree of sparsity (insufficient effective data points) or measurement noise (high variance random error) in the current actual flow information, providing parameters for subsequent data augmentation and weighting processing.

[0069] The technical solution of the above embodiments, through multi-dimensional flow feature extraction, multi-period statistical attribution analysis and deviation pattern classification, upgrades the quality assessment of actual flow information from simple data volume statistics to intelligent discrimination based on physical causal relationships, accurately distinguishes model parameter deviations from deviations, and provides a high-quality data foundation that has been filtered for subsequent model correction.

[0070] In one possible design, Figure 5 This is a flowchart illustrating step S412 according to an exemplary embodiment. (Refer to...) Figure 5 In step S412, the step of performing spatiotemporal correlation analysis on the collected data and extracting flow characteristics includes: S4121. Based on the spatial distribution range and time span of the collected local pressure, melt front velocity and temperature data, establish a multi-scale spatiotemporal grid; In this embodiment, a hierarchical spatiotemporal grid with different spatial granularity and temporal resolution is established based on the actual spatial coverage of the sensor data and the total time span of data acquisition. The coarse grid captures global trends, while the fine grid preserves local details.

[0071] S4122. Map the collected local pressure, melt front velocity and temperature data to the multi-scale spatiotemporal grid; In this embodiment, each data point is assigned to the corresponding cell of the corresponding level grid according to the spatial coordinates and timestamps of each sensor data, forming a structured spatiotemporal data matrix.

[0072] S4123. Using the interpolation processing within the multi-scale spatiotemporal grid, spatial smoothing and temporal synchronization are performed on the mapped data; In this embodiment, spatial interpolation smoothing is performed on the data of adjacent grid cells at the same time layer to eliminate spatial discontinuities caused by discrete sensor arrangement, and temporal interpolation is performed on the grid data of different time layers at the same spatial location to unify the time coordinate system of all measurement points.

[0073] S4124. Under a unified spatiotemporal framework, calculate the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the small path, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature.

[0074] In this embodiment, under a unified Cartesian coordinate system that has completed data mapping and smoothing, the instantaneous values ​​and gradient values ​​of various flow characteristics are calculated by finite difference or fitting differentiation, and a structured multidimensional feature matrix is ​​output.

[0075] The technical solution of the above embodiments unifies heterogeneous data from unevenly distributed sensor arrays into a consistent spatiotemporal coordinate system through multi-scale spatiotemporal grid data mapping and smoothing processing, providing a standardized data foundation for subsequent multi-dimensional flow feature calculation and improving the accuracy and robustness of feature extraction.

[0076] In one possible design, step S416, the step of identifying deviations caused by the dynamic evolution of the internal structure of the material, includes: S4161. In the specific area, a miniature heat flow sensor and a miniature shear force sensor are deployed to collect instantaneous fluctuation data of local cooling rate and shear force rate, and to perform spectrum analysis on the collected instantaneous fluctuation data to extract its energy distribution characteristics in different frequency ranges. In this embodiment, the miniature heat flux sensor employs a thin-film thermopile structure with a response time of less than 10 milliseconds, capable of capturing instantaneous fluctuations in the cooling rate within the range of 0.1 to 10 Hz. The miniature shear force sensor is based on the piezoresistive principle, with a sampling frequency of no less than 1 kHz. Fast Fourier transform spectral analysis is performed on the output data of both types of sensors to extract the energy density distribution characteristics within each characteristic frequency range, forming a hydrodynamic feature fingerprint for that region.

[0077] S4162. In the specific area, a miniature ultrasonic sensor is deployed to collect data on the ultrasonic attenuation coefficient and sound velocity variation inside the material, and the collected ultrasonic attenuation coefficient and sound velocity variation data are analyzed in the time domain and frequency domain to extract features related to the internal microstructure of the material. In this embodiment, a miniature ultrasonic sensor emits pulsed ultrasonic waves at a frequency of 1 to 10 MHz and receives echo signals that penetrate the material. The ultrasonic attenuation coefficient is calculated by comparing the amplitude difference between the emitted and received waves, and the sound velocity is calculated by measuring the propagation time of the ultrasonic waves. Time-domain and frequency-domain analyses are performed on the time series of the two parameters to track their changing trends over time and their periodic fluctuation patterns synchronized with the injection molding cycle.

[0078] S4163. Establish a correlation mapping based on the material type and temperature range of the specific region, wherein the correlation mapping correlates the spectral energy distribution characteristics of the local cooling rate and the shear action velocity with the ultrasonic attenuation coefficient and the sound velocity change characteristics; In this embodiment, the correlation mapping is dynamically constructed based on material type and the current temperature range of the region. A pre-calibrated material-temperature correlation matrix maps the spectral energy distribution characteristics to the corresponding ultrasonic feature space. The correlation mapping employs a multivariate regression model to establish a quantitative correspondence between the two physical measurements, outputting a correlation strength coefficient.

[0079] S4164. When the spectral energy distribution characteristics of the local cooling rate and the shearing rate change in a specific pattern, and the change corresponds to the specific pattern change of the ultrasonic attenuation coefficient and the sound velocity change characteristics, it is identified as a deviation caused by the dynamic evolution of the internal structure of the material.

[0080] In this embodiment, the specific pattern is learned from historical data using a machine learning classifier. During normal cycles, the material flow characteristic spectrum and ultrasonic features exhibit a consistent, stable pattern. When polymer chain breakage, orientation rearrangement, or microcrystallization occurs in the material, the flow characteristic spectrum shows energy enhancement in a specific frequency band, while the ultrasonic attenuation coefficient increases and the sound velocity decreases. When a synchronous pattern change across physical fields is detected and the correlation strength coefficient exceeds a preset threshold, the determination bias originates from the dynamic evolution of the material's internal structure.

[0081] The technical solutions of the above embodiments, through multi-physics field collaborative monitoring of heat flow-shear force sensors and ultrasonic sensors, and the correlation mapping method based on material type and temperature range, upgrade the identification of the dynamic evolution of the internal structure of materials from speculation of a single physical quantity to mutual verification and precise correlation between two independent physical measurements, which significantly improves the reliability and accuracy of material condition diagnosis.

[0082] In one example, during injection molding, it is necessary to identify deviations caused by the dynamic evolution of the internal structure of a semi-crystalline polymer (such as polypropylene) during the production of thin-walled parts. The system deploys miniature heat flux sensors and miniature shear force sensors in this region to collect instantaneous fluctuation data of local cooling rates and shear rates, and performs spectral analysis. Simultaneously, miniature ultrasonic sensors are deployed to collect data on ultrasonic attenuation coefficients and sound velocity changes. By establishing a correlation mapping based on material type and temperature range, the spectral energy distribution characteristics are correlated with the ultrasonic characteristics. When an energy enhancement is detected in the cooling rate and shear rate spectra at specific frequency bands, along with a corresponding pattern change of increasing ultrasonic attenuation coefficient and decreasing sound velocity, it is identified as a dynamic evolution of the material's internal structure in that region.

[0083] In summary, the computer-aided design optimization method for injection mold structure provided in this application constructs a complete closed-loop optimization system from production feedback to design improvement by collecting actual material flow information through sensor arrays, identifying multi-dimensional differences and back-calculating flow resistance, isolating degradation data and assessing data quality, adaptively correcting regularized model parameters, and coordinating the work of each link in the iterative optimization of mold structure.

[0084] Example 2 Embodiment 2 of this application provides a computer-aided design optimization system for injection mold structure. Figure 6 This is a block diagram illustrating a computer-aided design optimization system for injection mold structures according to an exemplary embodiment. (Refer to...) Figure 6 The system includes: The information acquisition module 01 is used to acquire material flow information of the injection mold during the actual production process. This module corresponds to step S1 in Example 1, and collects pressure, speed and temperature data during the actual injection molding process by deploying pressure sensors and melt front velocity sensor arrays in key areas of the mold cavity.

[0085] The difference identification module 02 is used to compare the material flow information with preset material flow prediction information to identify the differences between the material flow information and the material flow prediction information. This module corresponds to step S2 in Embodiment 1, which outputs the difference quantification result by performing multi-dimensional alignment comparison between actual sensor data and simulation prediction data based on the current plastic flow model.

[0086] The resistance calculation module 03 is used to calculate the actual flow resistance of the material in a specific region in response to the difference, based on the material flow information. This module corresponds to step S3 in Example 1, which calculates the actual effective viscosity by back-calculating simplified fluid dynamics relationships based on actual pressure and velocity data and mold geometry parameters.

[0087] The parameter adjustment module 04 is used to adjust the parameters of the plastic flow model according to the actual flow resistance. This module corresponds to step S4 in Example 1, and uses a regularization optimization strategy and adaptive learning rate to drive the correction of the Cross-WLF model parameters using the actual effective viscosity data points obtained by back-calculation.

[0088] Design optimization module 05 is used to optimize the design of the injection mold structure using the adjusted plastic flow model. This module corresponds to step S5 in Example 1, and re-performs mold filling simulation and structural parameter optimization based on the updated plastic flow model.

[0089] In summary, the computer-aided design optimization system for injection mold structure provided in this application embodiment achieves closed-loop automation from actual production data to model updates and then to design improvements through the collaborative work of five modules in accordance with the data flow direction of material flow information acquisition, difference identification and comparison, reverse calculation of local flow resistance, adaptive correction of model parameters, and mold structure design optimization.

[0090] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0091] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A computer-aided design optimization method for injection mold structures, characterized in that, Includes the following steps: Obtain material flow information during the actual production process of injection molds; The material flow information is compared with preset material flow prediction information to identify the differences between the material flow information and the material flow prediction information; In response to the difference, the actual flow resistance of the material in a specific region is calculated based on the material flow information; Adjust the parameters of the plastic flow model based on the actual flow resistance. The adjusted plastic flow model was used to optimize the design of the injection mold structure.

2. The method of claim 1, wherein, The step of calculating the actual flow resistance of the material in a specific region based on the material flow information in response to the difference includes: The local shear rate and local temperature of the specific area are continuously monitored, and when the local shear rate and the local temperature simultaneously exceed the preset material degradation critical threshold, an early warning is issued that the specific area is a potentially high-risk area for degradation. Calculate the actual effective viscosity of the specific region; Based on the local shear rate and local temperature of the specific region, predict the expected viscosity of the specific region; By comparing the actual effective viscosity with the expected viscosity, when the actual effective viscosity is lower than the expected viscosity and the decrease exceeds a preset abnormal drop threshold, and an early warning has been triggered in the specific area, it is confirmed that the material has degraded. When the material is confirmed to have degraded, the actual effective viscosity data of the degradation is isolated.

3. The method of claim 2, wherein, The step of calculating the actual effective viscosity of the specific region includes: Identify the mold geometry in the specific region, and detect wall slippage in the specific region when it is an extremely thin-walled region; Based on the wall slip phenomenon, the boundary conditions in the simplified fluid dynamics formula are adjusted to introduce an equivalent slip length or slip velocity term, thereby correcting the deviation in the flow resistance calculation. Monitor the local temperature change rate and shear rate change rate in the specific region, and evaluate the viscoelastic effect of the material based on the local temperature change rate and shear rate change rate. Based on the viscoelastic effect, a time-dependent or shear history-dependent correction factor is introduced to dynamically adjust the viscosity term in the simplified fluid dynamics formula, thereby compensating for the influence of material viscoelasticity on flow resistance. The actual effective viscosity of the specific region is calculated based on the adjusted simplified fluid dynamics formula.

4. The method of claim 1, wherein, The step of adjusting the parameters of the plastic flow model based on the actual flow resistance includes: Identify whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise. The identification includes distinguishing the types of deviations that cause the sparsity or measurement noise. The types of deviations include deviations caused by inaccurate initial parameters of the plastic flow model and deviations caused by the dynamic evolution of the internal structure of the material. When the actual flow information is identified as sparse or contains measurement noise, data augmentation is performed based on the actual flow information to generate a set of virtual data points; Based on the actual flow information, the intensity of the measurement noise is evaluated, and the actual flow resistance data is weighted according to the intensity of the measurement noise. A regularization-based optimization strategy is adopted to adjust the parameters of the plastic flow model based on the weighted actual flow resistance data and the virtual data points. The learning rate of the plastic flow model parameters is dynamically adjusted based on whether the actual flow information is sparse or contains measurement noise.

5. The method of claim 4, wherein, The step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise includes: A sensor array is deployed to collect local pressure, melt front velocity, and temperature data in specific areas of the injection mold; Spatiotemporal correlation analysis was performed on the collected data to extract the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micro-path, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature as flow characteristics. The actual measured flow characteristics are compared with the flow characteristics predicted by the plastic flow inside the system to identify whether the differences between the two in multiple dimensions continue to exceed the preset small difference threshold, thereby identifying potential deviations in rheological behavior. For the identified potential deviations in rheological behavior, statistical analysis was performed over multiple consecutive injection molding cycles to calculate the mean deviation and standard deviation. Based on the average deviation and the standard deviation, identify the deviation caused by inaccurate initial parameters of the plastic flow model; Based on the correlation between the average deviation, the standard deviation, the local cooling rate of the specific region, and the shear rate of the specific region, the deviation caused by the dynamic evolution of the internal structure of the material is identified; Based on the attribution results of the deviation, the sparsity or measurement noise of the actual flow information is identified.

6. The method of claim 4, wherein, The step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise includes performing spatiotemporal correlation analysis on the collected local pressure, melt front velocity, and temperature data to extract flow characteristics. Based on the spatial distribution range and time span of the collected local pressure, melt front velocity, and temperature data, a multi-scale spatiotemporal grid is established. The collected local pressure, melt front velocity, and temperature data are mapped onto the multi-scale spatiotemporal grid. The interpolation process within the multi-scale spatiotemporal grid is used to perform spatial smoothing and temporal synchronization on the mapped data. Within a unified spatiotemporal framework, the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature are calculated.

7. The method of claim 6, wherein, The steps for calculating the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature within a unified spatiotemporal framework include: Analyze the internal geometry of the mold to identify complex regions, including microchannels, sharp bends, and flow dividers. Based on the geometric characteristics of the complex region, the sampling density and time window length of local pressure, melt front velocity and temperature data are adjusted, and the acquisition frequency of sensor data is increased and the time window of data analysis is shortened within the complex region; For the microchannel region, the average velocity of the melt front at different cross sections is calculated, and the distribution of local shear velocity is inferred by combining the changes in the channel cross section; for the sharp bend region, the pressure difference between the inside and outside of the bend of the pressure sensor array and its change over time is analyzed; for the splitter region, the arrival time of the melt front on different split paths and the pressure drop on each path are compared. Based on the local flow information of the microchannel region, the sharp bend region, and the splitter region, the instantaneous slope of the local pressure curve, the rate of change of the melt front velocity along the micropath, the pressure gradient between adjacent pressure sensors and its change over time, and the rate of change of local temperature are calculated.

8. The method of claim 4, wherein, In the step of identifying whether the actual flow information under the current injection molding process parameter combination is sparse or contains measurement noise, the step of identifying the deviation caused by the dynamic evolution of the material's internal structure based on the correlation between the average deviation, the standard deviation, the local cooling rate of the specific region, and the shear rate of the specific region includes: In the specific area, miniature heat flow sensors and miniature shear force sensors are deployed to collect instantaneous fluctuation data of local cooling rate and shear force rate, and to perform spectral analysis on the collected instantaneous fluctuation data to extract its energy distribution characteristics in different frequency ranges. In the specific region, miniature ultrasonic sensors are deployed to collect data on the ultrasonic attenuation coefficient and sound velocity variation inside the material, and time-domain and frequency-domain analyses are performed on the collected ultrasonic attenuation coefficient and sound velocity variation data to extract features related to the internal microstructure of the material. A correlation mapping is established based on the material type and temperature range of the specific region. The correlation mapping associates the spectral energy distribution characteristics of the local cooling rate and the shear rate with the ultrasonic attenuation coefficient and the sound velocity variation characteristics. When the spectral energy distribution characteristics of the local cooling rate and the shearing rate change in a specific pattern, and the change corresponds to the specific pattern change of the ultrasonic attenuation coefficient and the sound velocity change characteristics, it is identified as a deviation caused by the dynamic evolution of the internal structure of the material.

9. The method of claim 8, wherein, The steps for establishing the association mapping include: Obtain the current injection molding process parameters, and based on the process parameters, retrieve the association patterns that match the process parameters; Based on the correlation mode, the mapping relationship between the spectral energy distribution characteristics of the local cooling rate and the shearing velocity and the ultrasonic attenuation coefficient and the sound velocity change characteristics is dynamically adjusted.

10. An injection mold structure computer-aided design optimization system, comprising: The system includes: The information acquisition module is used to acquire material flow information of the injection mold during the actual production process; The difference recognition module is used to compare the material flow information with preset material flow prediction information to identify the difference between the material flow information and the material flow prediction information; A resistance calculation module is used to calculate the actual flow resistance of the material in a specific region based on the material flow information in response to the difference. The parameter adjustment module is used to adjust the parameters of the plastic flow model according to the actual flow resistance. The design optimization module is used to optimize the design of the injection mold structure using the adjusted plastic flow model.

Images