Method and system for flow stability control of aluminum alloy structural parts in high velocity filling

By optimizing the injection curve and mold temperature control through a zoned gating system and a multiphysics simulation model, the flow instability problem in the high-speed filling process of high-density aluminum alloy structural parts was solved, achieving high-quality and consistent forming of the castings.

CN123060083APending Publication Date: 2026-11-10NINGBO LONGYUE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611579323.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-13
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In the high-speed filling process of high-density aluminum alloy structural parts, existing technologies struggle to control flow stability, especially the flow imbalance and instability caused by uneven wall thickness, which affects the internal quality and mechanical properties of the castings.

Method used

By acquiring a three-dimensional geometric model, a zoned gating system is designed. Combined with a multiphysics simulation model and iterative optimization algorithm, the injection curve and mold temperature control strategy are optimized, and the injection speed and mold temperature are adjusted in real time to achieve online control of flow stability.

Benefits of technology

It effectively reduces casting defects, improves the filling quality and forming consistency of aluminum alloy structural parts, and ensures the internal quality and sealing performance of power supply box components with high density requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123060083A_ABST
    Figure CN123060083A_ABST
Patent Text Reader

Abstract

This invention provides a method and system for controlling the flow stability of aluminum alloy structural parts during high-speed filling, relating to the field of die casting molding control technology. The control method involves acquiring a three-dimensional geometric model to extract cavity feature parameters and determine a zoned gating system accordingly. A multi-physics simulation model is established to obtain flow field and temperature field distribution data during the filling process. Based on the flow field data, a quantitative evaluation index for flow stability is calculated, and this index is used as the optimization target to perform multi-segment collaborative optimization of the injection curve, outputting the optimal injection curve and the corresponding theoretical quantitative index for flow stability. Temperature non-uniformity regions and mold temperature control strategies are determined based on temperature field data. Filling is executed according to the optimal injection curve, and pressure, flow rate, and temperature data are collected in real time during the filling process to calculate the actual quantitative index for flow stability. When the deviation from the theoretical index exceeds a threshold, the injection speed and / or mold temperature control parameters are adjusted in real time to achieve online control of flow stability during high-speed filling.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of die casting molding control technology, and more specifically, to a method and system for controlling the flow stability of aluminum alloy structural parts during high-speed filling. Background Technology

[0002] High-density aluminum alloy structural components, due to their lightweight and high strength characteristics, have wide applications in fields such as integrated power supply enclosures for new energy vehicles, automotive lightweighting, and consumer electronics structural components. However, such castings face severe challenges in controlling flow stability during high-speed die casting: on the one hand, due to uneven casting wall thickness, the melt experiences high flow resistance and rapid cooling in thin-walled regions, making it prone to cold shut defects at the end of filling due to flow stagnation; while thick-walled regions are prone to turbulent flow and trapped gas, severely affecting the density of the power supply enclosure components. On the other hand, when thin-walled and thick-walled regions coexist, the significant difference in flow resistance between regions causes the melt to preferentially fill rapidly along the thick-walled regions with lower resistance, while the thin-walled regions fill slowly. This leads to asynchronous filling of different areas of the cavity, causing flow instability phenomena such as gas entrapment and trapped gas, which seriously affects the internal quality and mechanical properties of the casting.

[0003] To address the flow stability issues during the high-speed filling process, existing technologies mainly focus on the following aspects: In terms of gating system design, existing methods optimize the flow direction and flow rate of the melt into the mold cavity by adjusting the position and size of the ingate; in terms of setting injection process parameters, existing solutions mostly adopt a three-stage injection curve (low speed-high speed-increase), and achieve process control of the filling process by empirically setting the speed values ​​of each stage; in terms of mold temperature control, existing solutions use cooling channels or heating rods to perform overall or zoned temperature control of the mold.

[0004] However, the aforementioned existing technologies still have the following shortcomings: First, the design of the gating system is mostly experience-based, lacking quantitative analysis of the filling requirements of different wall thickness areas in the cavity, making it difficult to configure differentiated material supply channels for thin-walled areas to achieve synchronous filling of each area; Second, the parameters of each stage of the injection curve are mostly fixed empirical values, lacking quantitative correlation with the flow stability of the filling process, making it difficult to adaptively adjust according to the specific filling state; Third, the flow stability in the existing technologies mostly remains at the level of qualitative empirical judgment, lacking quantifiable evaluation indicators, resulting in a lack of objective basis for process optimization in the filling process; Fourth, mold temperature control is mostly based on the overall mold temperature setting, without linkage control with the flow state during the filling process, making it difficult to dynamically respond to local flow instability problems that occur during the filling process.

[0005] In summary, how to achieve quantitative evaluation and online adaptive control of the flow stability during high-speed filling of high-density aluminum alloy structural components has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The problem solved by this invention is one or more of the aforementioned related technical problems.

[0007] To address the aforementioned problems, this invention provides a method and system for controlling the flow stability of aluminum alloy structural components during high-speed molding.

[0008] In a first aspect, the present invention provides a method for controlling the flow stability of aluminum alloy structural components during high-speed filling, comprising: A three-dimensional geometric model of a high-density aluminum alloy structural component is obtained and the characteristic parameters of the mold cavity are extracted. Based on the characteristic parameters of the mold cavity, a partitioned gating system is determined, and the structural parameters of the partitioned gating system are obtained. Based on the structural parameters of the partitioned casting system and the preset injection process parameters, a multiphysics simulation model of the filling process is established. The flow field distribution data and temperature field distribution data at each moment during the filling process are obtained through the multiphysics simulation model. The flow stability quantitative evaluation index is calculated based on the flow field distribution data. Using the aforementioned flow stability quantitative evaluation index as the optimization objective, and with a preset filling time window and a preset solidification fraction threshold as constraints, an iterative optimization algorithm is used to perform multi-segment collaborative optimization of the injection curve, outputting the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve. Based on the temperature field distribution data, determine the regions with uneven temperature distribution on the surface of the mold cavity, and determine the mold temperature control strategy corresponding to each of the uneven temperature distribution regions. The filling process is performed according to the optimal injection curve, and actual pressure data, actual flow rate data, and actual temperature data are collected in real time during the filling process. The actual flow stability quantification index at the current moment is calculated based on the actual pressure data, actual flow rate data, and actual temperature data. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy are adjusted in real time to achieve online control of flow stability during high-speed filling.

[0009] Optionally, the mold cavity characteristic parameters include wall thickness distribution data for each region of the mold cavity; the step of determining the zoned gating system based on the mold cavity characteristic parameters to obtain the zoned gating system structural parameters includes: Based on the wall thickness distribution data, the mold cavity is divided into thin-walled and thick-walled regions; Set the cross-sectional area of ​​each ingate in the partitioned gating system so that the cross-sectional area of ​​the ingate corresponding to the thin-walled region is greater than the cross-sectional area of ​​the ingate corresponding to the thick-walled region. Based on the cross-sectional area of ​​each ingate and the corresponding cavity region location, the structural parameters of the partitioned gating system are generated.

[0010] Optionally, the quantitative evaluation indexes for flow stability include the root mean square error of flow velocity and the volatility of the flow front advance velocity; The method for calculating the mean square error of the flow velocity is as follows: during the filling process, the mold cavity is divided into multiple detection areas according to a preset spatial position. For each detection area, the mean square error of the melt flow velocity of each spatial node in the same time window is calculated to obtain the mean square error of the flow velocity. The method for calculating the flow front advance speed fluctuation rate is as follows: during the filling process, based on the change of the position of the flow front of the aluminum alloy melt over time, the advance speed of the flow front at each moment is calculated, and the ratio of the standard deviation to the average value of the advance speed over the time series is used as the flow front advance speed fluctuation rate.

[0011] Optionally, the multiphysics simulation model includes at least a flow field model, a temperature field model, and a solidification field model. The multiphysics simulation model is used to simulate the flow behavior of aluminum alloy melt during the filling process and to obtain the flow field distribution data, the temperature field distribution data, and the solidification front advancement data at each spatial node at each time.

[0012] Optionally, the injection curve includes at least a low-speed phase, a high-speed phase, and a pressurization phase; When the flow front of the aluminum alloy melt has not reached the ingate position, switch to the low-speed stage, and the low-speed stage controls the injection speed within the preset low-speed range. When the flow front of the aluminum alloy melt reaches the ingate position, the process switches to the high-speed stage. The high-speed stage controls the injection speed within a preset high-speed range until the cavity filling rate reaches a preset filling rate threshold. When the cavity filling rate reaches the preset filling rate threshold, the process switches to the pressurization stage. In the pressurization stage, the injection speed is reduced to a preset pressurization range, and pressurization pressure is applied until the solidification fraction reaches the preset threshold. The switching time between the low-speed stage and the high-speed stage is determined based on the time when the flow front of the aluminum alloy melt reaches the ingate in the multiphysics simulation model.

[0013] Optionally, the high-speed phase includes a first high-speed sub-phase and a second high-speed sub-phase; Wherein, the injection speed of the first high-speed sub-stage is greater than that of the second high-speed sub-stage. The first high-speed sub-stage is used to enable the melt to quickly pass through the partitioned gating system and enter the cavity. The second high-speed sub-stage is used to enable the melt to smoothly fill the thin-walled area of ​​the cavity. The switching time between the first high-speed sub-stage and the second high-speed sub-stage is determined based on the moment when the aluminum alloy melt flow front enters the thin-walled region of the cavity in the multiphysics simulation model.

[0014] Optionally, the mold temperature control strategy includes: When the temperature of any region in the thin-walled region or the thick-walled region exceeds the preset upper temperature limit, the flow rate of the cooling medium in the corresponding region is increased or the temperature of the cooling medium is decreased. When the temperature of any region in the thin-walled region or the thick-walled region is lower than the preset lower limit, the flow rate of the cooling medium in the corresponding region is reduced or heating is started. The preset upper temperature limit and preset lower temperature limit are determined based on the solidification characteristics of the melt in this region in the multiphysics simulation model.

[0015] Optionally, the step of calculating the quantitative index of actual flow stability at the current moment based on the actual pressure data, the actual flow rate data, and the actual temperature data includes: The actual pressure data, the actual flow rate data, and the actual temperature data are input into a pre-built filling state evaluation model, and the filling state evaluation model outputs the quantitative index of the actual flow stability at the current moment. The filling state evaluation model uses pressure data sequences, flow data sequences, and temperature data sequences collected during the historical filling process as input features, and uses the measured flow stability evaluation index at the corresponding time as training labels. It is trained using a long short-term memory network or a gated recurrent unit network architecture.

[0016] Optionally, the method for controlling the flow stability of the aluminum alloy structural component during high-speed filling further includes: The actual pressure data, actual flow rate data, actual temperature data, and corresponding quantitative indicators of actual flow stability collected during each filling process are used as new samples to perform online incremental learning on the filling state evaluation model, update the model parameters, and enable the updated model to be used in the flow stability evaluation of subsequent filling processes.

[0017] In a second aspect, the present invention provides a flow stability control system for aluminum alloy structural components during high-speed filling, for implementing the flow stability control method for aluminum alloy structural components during high-speed filling as described in the first aspect, the system comprising: The acquisition unit is used to acquire the three-dimensional geometric model of the high-density aluminum alloy structural parts and extract the mold cavity feature parameters. Based on the mold cavity feature parameters, the partitioned gating system is determined, and the structural parameters of the partitioned gating system are obtained. The simulation unit is used to establish a multiphysics simulation model of the filling process based on the structural parameters of the partitioned casting system and the preset injection process parameters. The multiphysics simulation model is used to obtain the flow field distribution data and temperature field distribution data at each moment during the filling process. The flow stability quantitative evaluation index is calculated based on the flow field distribution data. The optimization unit is used to optimize the injection curve in multiple segments by using the flow stability quantitative evaluation index as the optimization target, and using a preset filling time window and a preset solidification fraction threshold as constraints, and to output the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve. The processing unit is used to determine the uneven temperature distribution areas on the surface of the mold cavity based on the temperature field distribution data, and to determine the mold temperature control strategy corresponding to each uneven temperature distribution area. The control unit is used to execute the filling of each injection stage according to the optimal injection curve, and to collect actual pressure data, actual flow data and actual temperature data in real time during the filling process. Based on the actual pressure data, actual flow data and actual temperature data, it calculates the actual flow stability quantification index at the current moment. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, it adjusts the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy in real time to achieve online control of flow stability during high-speed filling.

[0018] The beneficial effects of the method and system for controlling the flow stability of aluminum alloy structural components during high-speed filling of the present invention are as follows: By acquiring a three-dimensional geometric model and extracting cavity feature parameters, a zoned gating system was designed. This ensured that the configuration of the gating system's ingates matched the filling requirements of different areas of the cavity, balancing the flow distribution of melt entering each area of ​​the cavity from the source and effectively reducing the filling imbalance problem caused by differences in wall thickness. A multi-physics simulation model was established to obtain flow field and temperature field distribution data during the filling process. Based on the flow field distribution data, a quantitative evaluation index of flow stability was calculated, transforming the previously experience-based flow stability into a quantifiable and optimizable numerical index, providing an objective basis for the precise optimization of the subsequent injection curve. By using the quantitative evaluation index of flow stability as the optimization target and employing an iterative optimization algorithm to perform multi-segment collaborative optimization of the injection curve, the injection curve that optimizes the flow stability during the filling process and its corresponding theoretical quantitative flow stability index were obtained, achieving scientific and systematic optimization of the filling process parameters. Finally, the temperature distribution on the mold cavity surface was determined based on the temperature field distribution data. By identifying uneven areas and developing corresponding mold temperature control strategies, the mold temperature control can specifically compensate for temperature deviations in different areas, avoiding differences in melt flowability caused by uneven temperature, and providing reliable thermal boundary conditions for flow stability control. Finally, by performing mold filling according to the optimal injection curve and collecting pressure, flow rate, and temperature data in real time during the filling process, the actual flow stability quantitative indicators are dynamically compared with theoretical indicators. When the deviation exceeds the preset threshold, the injection speed or mold temperature control parameters are adjusted in real time, realizing real-time perception and active control of the flow instability state during high-speed filling. This forms a complete closed-loop control system of simulation pre-optimization, online monitoring, and real-time control, which effectively reduces casting defects such as porosity and cold shuts caused by flow instability during high-density aluminum alloy structural parts filling, and significantly improves the filling quality and forming consistency of aluminum alloy structural parts. Especially for power supply box components with high density requirements, it can effectively ensure their internal quality and sealing performance. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a method for controlling the flow stability of an aluminum alloy structural component during high-speed filling, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the flow stability control system for an aluminum alloy structural component during high-speed filling, according to an embodiment of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0021] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0022] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0023] It should be noted that the terms "one" and "more" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0024] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0025] Existing molding process control systems are mostly open-loop. That is, once the process parameters are set, they are executed according to a predetermined program. They lack the ability to monitor state parameters such as pressure, flow rate, and temperature in real time during the molding process and the ability to adjust online based on flow stability. They are unable to detect and adaptively adjust for flow instability that occurs during the molding process.

[0026] To solve the above problems, such as Figure 1 As shown in the figure, an embodiment of the present invention provides a method for controlling the flow stability of aluminum alloy structural components during high-speed filling, comprising: Step S100: Obtain the three-dimensional geometric model of the high-density aluminum alloy structural component and extract the mold cavity feature parameters. Determine the partitioned gating system based on the mold cavity feature parameters to obtain the partitioned gating system structural parameters.

[0027] Specifically, high-density aluminum alloy structural components refer to a type of aluminum alloy casting with strict requirements for internal quality and sealing performance. A typical example is the integrated power supply box assembly for new energy vehicles. These structural components must meet high density, high sealing performance, and good mechanical properties during die casting. Any defects such as porosity, cold shuts, or shrinkage caused by unstable flow during filling can lead to part scrapping or leakage failure during service. Therefore, precise control of flow stability during high-speed filling is crucial to ensuring the quality of the final product. During die casting, molten aluminum alloy flows from the pressure chamber through the gating system into the mold cavity, eventually solidifying to form the casting. The gating system connects the pressure chamber and the mold cavity, and its design directly affects the flow pattern and filling sequence of the molten metal after entering the cavity. A significant characteristic of high-density aluminum alloy structural components is the uneven wall thickness distribution in different areas of the mold cavity; thin-walled areas have high flow resistance and fast cooling rates, while thick-walled areas have low flow resistance and slow cooling rates. If the gating system uses a single runner to supply material to all areas uniformly, the melt will preferentially fill the thick-walled areas with lower resistance quickly, while the thin-walled areas will fill slowly, resulting in asynchronous filling of different areas of the cavity. Therefore, the gating system needs to be designed according to the filling requirements of different areas of the cavity, so that each ingate can provide differentiated material flow rates for its corresponding area, achieving synchronous filling of all areas.

[0028] A three-dimensional geometric model is a 3D image data file of a casting that is determined during the product design phase. It contains complete geometric information about the casting, such as its outline, wall thickness distribution, rib layout, hole locations, fillet transitions, and all other structural features. This 3D geometric model can be obtained directly from the product design department and serves as the geometric basis for all subsequent data extraction and analysis.

[0029] Extracting mold cavity feature parameters from a 3D geometric model refers to identifying geometric features and extracting data from the aforementioned 3D geometric model to obtain key parameters characterizing the filling characteristics of each region of the cavity. The core parameter is the wall thickness distribution data of each region of the cavity, i.e., the specific values ​​of the wall thickness of each part of the casting and its distribution pattern in the cavity space. Wall thickness distribution data can be obtained through the cross-sectional analysis function or wall thickness analysis tools of 3D drawing software: multiple cross-sections are cut along different directions on the 3D geometric model, the specific values ​​of the casting wall thickness at each cross-section are measured, and these values ​​are labeled and recorded according to their spatial location, thus obtaining complete distribution data of the wall thickness in each region of the cavity. Mold cavity feature parameters also include the surface area, volume, and flow path length of each region; these parameters are used together to assess the filling difficulty and flow rate requirements of each region.

[0030] After obtaining the characteristic parameters of the mold cavity, a zoned gating system is determined based on these parameters. The core design concept of the zoned gating system is to set multiple ingates in the gating system, each corresponding to a different area of ​​the cavity. The cross-sectional area of ​​each ingate is independently set according to the filling flow requirements of its corresponding area, ensuring that the melt from each area reaches the end of the cavity synchronously during the filling process, achieving synchronous filling of each area. The zoned gating system also includes the layout design of the main runner and branch runners to ensure that the melt smoothly enters each area of ​​the cavity from the pressure chamber through each ingate.

[0031] Finally, based on the cross-sectional area parameters of each ingate and its corresponding location on the cavity, a complete set of structural parameters for the zoned gating system is generated. These parameters typically include the number of ingates, the cross-sectional area of ​​each ingate, the spatial location and orientation of each ingate, and the cross-sectional dimensions and orientation of the main runner and branch runners. These data constitute the geometric input for the subsequent multiphysics simulation model.

[0032] By extracting mold cavity feature parameters such as wall thickness distribution from the three-dimensional geometric model and determining the zoned gating system accordingly, the configuration of each ingate in the gating system can be differentiated according to the filling requirements of different areas of the cavity. This balances the flow distribution of melt into each area of ​​the cavity from the source, effectively avoiding the problem of uneven filling caused by differences in wall thickness. It also enables the melt flow front of each area to advance synchronously during the filling process, providing a reasonable gating system structural basis for the subsequent steps of injection curve optimization and flow stability control.

[0033] Step S200: Based on the structural parameters of the partitioned casting system and the preset injection process parameters, establish a multiphysics simulation model of the filling process, and obtain the flow field distribution data and temperature field distribution data at each moment during the filling process through the multiphysics simulation model; calculate the quantitative evaluation index of flow stability based on the flow field distribution data.

[0034] Specifically, multiphysics simulation models are numerical calculation models used to simulate the melt flow behavior, temperature changes, and solidification process during the filling process of high-density aluminum alloy structural components, such as integrated power supply box components for new energy vehicles. During die casting, the molten aluminum alloy enters the mold cavity through the gating system at high speed and high pressure. The flow state of the melt is simultaneously affected by fluid dynamics (flow velocity, pressure, inertial force, viscous force, etc.), heat transfer (heat transfer, temperature change, thermal expansion and contraction, etc.), and phase transformation (liquid to solid state transition, latent heat release, etc.). These three factors are coupled and mutually influential; therefore, a multiphysics simulation model capable of simultaneously describing the above physical fields is needed. The model is established using specialized finite element or finite volume method simulation software, such as some existing die casting simulation software.

[0035] The specific modeling process is as follows: The structural parameters of the zoned gating system and the preset injection process parameters are used as inputs to the model. First, a three-dimensional geometric model containing the pressure chamber, gating system, and mold cavity is established, and this geometric model is meshed to generate a discretized computational mesh. Then, material properties (including density, viscosity, thermal conductivity, specific heat capacity, latent heat, liquidus temperature, solidus temperature, etc. of aluminum alloy) and boundary conditions (including initial mold temperature, ambient temperature, heat transfer coefficient at the casting-mold interface, etc.) are set on the mesh model. Finally, injection process parameters (including injection speed curve, pressurization pressure, pouring temperature, etc.) are set as driving conditions for the filling process, and the numerical calculation of the filling process begins.

[0036] In the numerical calculation of the filling process, as the time step progresses, the simulation software solves the fluid dynamics control equations (i.e., the Navier-Stokes equations, which describe the velocity and pressure distribution of the melt flow), the heat transfer control equations (i.e., the Fourier heat conduction equations, which describe the temperature changes within the cavity), and the solidification model (which describes the process of the melt transitioning from liquid to solid), thereby obtaining the physical field distribution data of each time point and each spatial node during the filling process.

[0037] The flow field distribution data refers to the melt flow velocity and pressure data at each spatial node within the mold cavity at different times. Specifically, it includes the velocity components of the melt in the X, Y, and Z directions at each grid node, the magnitude and direction of the combined velocity, and the pressure value at that node. The flow field distribution data reflects the flow behavior of the melt in different regions of the mold cavity, such as whether the velocity distribution is uniform, whether eddies or backflows exist, and the advancing pattern of the flow front. The temperature field distribution data refers to the temperature values ​​at each spatial node within the mold cavity at different times, including the melt temperature and the mold cavity wall temperature. The temperature field distribution data reflects the cooling and solidification behavior of the melt during the filling process, such as the difference in the rate of temperature decrease in different regions and whether there is local overheating or undercooling. Both types of data are stored in time series form, meaning that a complete set of full-field data corresponds to each time step.

[0038] After obtaining the flow field distribution data, a quantitative evaluation index for flow stability is calculated based on this data. This quantitative evaluation index is one or more numerical indicators used to quantitatively characterize the uniformity and stability of the melt flow during the filling process. Its basic calculation logic is as follows: based on key information reflecting the uniformity and stability of the flow in the flow field distribution data, including the consistency of melt flow velocity in each region (reflecting flow uniformity) and the smoothness of the flow front advancement (reflecting filling stability), this information is transformed into quantifiable numerical indicators through mathematical statistical methods (such as calculating standard deviation, root mean square error, and volatility). Specifically, the dispersion of velocity can be calculated based on the velocity distribution at each spatial node within the cavity, and the volatility of the advancement velocity can be calculated based on the change in the flow front position over time, thereby obtaining a quantitative value that characterizes flow stability. The smaller the value of this quantitative index, the more uniform and stable the melt flow during the filling process; the larger the value, the more non-uniform and unstable the flow. This quantitative evaluation index serves as the objective function for optimizing the injection curve in subsequent steps, making flow stability an optimizable and controllable quantitative object.

[0039] By establishing a multiphysics simulation model to obtain flow field and temperature field distribution data during the filling process, a comprehensive and quantitative description of the melt flow behavior and temperature changes during the high-speed filling of high-density aluminum alloy structural parts is achieved. Specifically, the flow field distribution data reflects the flow velocity and pressure state of the melt in various regions of the cavity, while the temperature field distribution data reflects the cooling and solidification process of the melt. Based on this, by calculating the quantitative evaluation index of flow stability according to the flow field distribution data, the "flow stability" that originally relied on empirical qualitative judgment is transformed into a quantifiable, comparable, and optimizable numerical index. This provides a clear objective function and quantitative basis for the scientific and systematic optimization of the injection curve in subsequent steps, thereby fundamentally ensuring the optimization accuracy and control effect of the entire online compensation control method.

[0040] Step S300: Taking the flow stability quantitative evaluation index as the optimization target, and using the preset filling time window and preset solidification fraction threshold as constraints, the iterative optimization algorithm is used to perform multi-segment collaborative optimization on the injection curve, and outputs the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve.

[0041] Specifically, the injection profile is the curve showing the change in injection speed over time during die casting. It determines the speed variation of the molten aluminum alloy from the pressure chamber through the gating system into the mold cavity, and is the most critical process parameter affecting the flow stability during the filling process. However, injection profiles set based solely on experience often fail to achieve optimal flow stability during the filling process. For example, excessive speed may lead to melt ejection and air entrapment, while excessively slow speed may cause the temperature at the flow front of the molten aluminum alloy to drop too quickly, resulting in cold shuts. Improper speed transitions between stages can also cause flow instability. Therefore, before processing begins, it is necessary to determine a set of optimal injection profile parameters using numerical optimization methods, so that the quantitative evaluation index of flow stability reaches its optimal level when filling is performed under this injection profile.

[0042] First, the optimization objective and constraints are clearly defined. The optimization objective is a quantitative evaluation index for flow stability, meaning the core objective is to minimize the value of this index (a smaller index value indicates more uniform and stable melt flow during the filling process), thereby achieving a filling process with optimal flow stability. Constraints refer to the limitations that must be met during the optimization process, including a preset filling time window and a preset solidification fraction threshold. The filling time window refers to the allowed time range for completing the entire filling process, including the shortest filling time (ensuring the melt does not over-cool due to slow filling) and the longest filling time (ensuring production efficiency). The solidification fraction threshold refers to the upper limit of the proportion of solidified melt volume to the total volume within the mold cavity at a certain point in the filling process. If the solidification fraction exceeds this threshold, it means that the melt has begun to solidify significantly before completely filling the mold cavity, potentially leading to defects such as cold shuts. The filling time window and solidification fraction threshold together constitute the basic boundary conditions for evaluating filling quality. Any candidate injection curve must satisfy these two constraints; otherwise, even with an excellent flow stability index, the process will not be practically feasible.

[0043] The multi-segment coordinated optimization of the injection curve is achieved using an iterative optimization algorithm. The basic logic is as follows: the injection curve is typically divided into multiple stages, and the injection velocities of each stage are not independent. The velocity state of the previous stage determines the position and velocity of the melt in the pressure chamber and gating system, directly affecting the initial flow rate into the mold cavity, and thus influencing the filling state of subsequent stages. The next stage needs to be adjusted based on the velocity state of the previous stage. Therefore, the injection velocity parameters of each stage need to be considered and adjusted collaboratively during the optimization process. In each iteration, the iterative optimization algorithm generates a set of candidate injection velocity parameter combinations covering each stage. This candidate injection curve is input into the multiphysics simulation model established in step S200. The simulation calculates whether the filling process under the given injection curve conditions meets the constraints of the filling time window and solidification fraction threshold, as well as the specific values ​​of the corresponding flow stability quantitative evaluation index.

[0044] Based on the simulation results, the optimization algorithm adjusts the injection velocity parameters at each stage through its internal search strategies (such as gradient descent, selection crossover mutation in genetic algorithms, and velocity and position updates in particle swarm optimization), generating new candidate injection curves and evaluating them through simulation again. This process is iterated repeatedly, that is, the cycle of generating candidate curves, simulation evaluation, and parameter adjustment is repeated until a set of injection velocity parameter combinations that satisfies all constraints and minimizes the quantitative evaluation index of flow stability is found.

[0045] After multiple iterations, the optimal injection curve and the corresponding theoretical flow stability quantification index are finally output. The optimal injection curve is the curve of injection velocity changing over time that achieves optimal flow stability while satisfying the constraints; the theoretical flow stability quantification index is the numerical value of the flow stability quantification index corresponding to the optimal injection curve, representing the best flow stability level achievable under ideal simulation conditions, and serving as the theoretical optimal benchmark in subsequent online control.

[0046] By using the quantitative evaluation index of flow stability as the optimization target, the parameters of the injection curve are determined with a clear optimization direction, that is, directly optimizing the core quality index of flow stability, rather than relying on trial and error based on experience. By using the filling time window and solidification fraction threshold as constraints, it is ensured that the optimized injection curve not only has optimal flow stability, but also meets the basic process feasibility requirements of filling time and solidification boundary, avoiding extreme solutions that simply pursue flow stability while ignoring the actual process boundary. By using an iterative optimization algorithm for multi-segment collaborative optimization, the coupling relationship between the velocity parameters of each stage of the injection curve is fully considered. That is, the velocity state of the previous stage affects the initial position and velocity of the melt, and thus affects the filling conditions of the next stage, so that the parameters of each stage reach the optimal combination in a global sense rather than a simple superposition of local optima of each stage. The final output of the optimal injection curve provides the best process execution basis for the actual filling process, and the output of the theoretical flow stability quantitative index provides an accurate quantitative benchmark for judging whether the actual filling state deviates from the expectation in the subsequent online control. Thus, it provides a key link in the complete chain of the entire online compensation control method from optimization decision to execution benchmark to deviation judgment.

[0047] Step S400: Based on the temperature field distribution data, determine the regions with uneven temperature distribution on the surface of the mold cavity, and determine the mold temperature control strategy corresponding to each of the uneven temperature distribution regions.

[0048] Specifically, the temperature field distribution data was obtained from multiphysics simulations, recording the temperature values ​​at various points in the cavity at different times during the filling process. This includes both the temperature distribution of the melt inside the cavity and the temperature distribution on the surface of the mold cavity (i.e., the cavity wall). During high-speed filling, after the high-temperature aluminum alloy melt enters the relatively cool mold cavity, a strong heat exchange occurs between the melt and the mold wall. That is, the melt near the wall cools down rapidly, while the melt temperature in the central region of the cavity is relatively high, forming a temperature gradient. Due to the different wall thicknesses in different areas of the casting (thin-walled areas have smaller melt volume and less heat, resulting in faster cooling; thick-walled areas have larger melt volume and more heat, resulting in slower cooling), the temperature field distribution data clearly shows the temperature differences in different areas of the cavity surface: the mold cavity surface temperature corresponding to thin-walled areas drops faster and is lower; the mold cavity surface temperature corresponding to thick-walled areas drops slower and is relatively higher. If this temperature difference is too large, it will lead to significant differences in the fluidity of the melt in different areas. The melt viscosity increases and the flow resistance increases in the low-temperature area, further exacerbating the non-uniformity of the filling process.

[0049] Based on the wall thickness distribution of each region within the mold cavity, the temperature change history of the cavity surface during the filling process is examined for each type of region. That is, at each moment during the filling process, each node on the cavity surface corresponds to a temperature value. Statistical analysis of these temperature values ​​according to spatial location identifies which regions significantly deviate from the expected temperature range under ideal filling conditions. Regions with uneven temperature distribution mainly manifest as: excessively low temperatures, where the cavity surface temperature is significantly lower than normal, indicating that the melt cools too quickly, resulting in poor fluidity and a risk of cold shuts; and excessively high temperatures, where the cavity surface temperature is significantly higher than normal, indicating insufficient heat dissipation and overheating of the melt, potentially affecting the solidification structure. Since the temperature decrease rate in thin-walled regions is usually faster than in thick-walled regions, low-temperature regions tend to appear in thin-walled regions, while high-temperature regions are more likely to occur in thick-walled regions.

[0050] After identifying the regions with uneven temperature distribution, the corresponding mold temperature control strategy for each region is determined. The mold temperature control strategy refers to a pre-control plan formulated to control or improve the uneven temperature state. That is, based on the simulation temperature field analysis results before filling begins, the control direction and control target for each uneven region in the subsequent online control stage are predetermined. For example, for regions with excessively low temperatures, the pre-set heating control direction (such as reducing the cooling intensity of that region) is predetermined; for regions with excessively high temperatures, the pre-set cooling control direction (such as increasing the cooling intensity of that region) is predetermined. The core of this strategy is to provide the control direction for each uneven region, that is, when a temperature deviation occurs in that region during the actual filling process, in which direction (heating or cooling) should be adjusted, and by what relative magnitude? This strategy is a pre-planned scheme based on the simulation temperature field before filling begins, rather than a real-time decision during the filling process. Its function is to establish the correspondence between uneven temperature regions and control directions, that is, different cavity regions should respond with different control directions and intensities in subsequent control.

[0051] By using temperature field distribution data obtained from multiphysics simulation, the temperature differences in various regions of the mold cavity surface are identified, and uneven areas with low or high temperatures are accurately located. This transforms the target of mold temperature control from vague overall temperature control to clear regional precise control. For example, problems such as low temperature in thin-walled areas due to excessively rapid cooling and high temperature in thick-walled areas due to insufficient heat dissipation can be accurately located. Based on this, by pre-determining mold temperature control strategies for each uneven region, the control direction and target of each region are determined, providing a clear operational basis for the subsequent online control stage. This enables targeted and precise mold temperature control during the actual filling process, effectively ensuring that the melt in each region of the cavity maintains a reasonable temperature state and fluidity during the filling process, and providing reliable thermal boundary conditions for flow stability control.

[0052] Step S500: Perform filling at each injection stage according to the optimal injection curve, and collect actual pressure data, actual flow rate data, and actual temperature data in real time during the filling process. Calculate the actual flow stability quantification index at the current moment based on the actual pressure data, actual flow rate data, and actual temperature data. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, adjust the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy in real time to achieve online control of flow stability during high-speed filling.

[0053] Specifically, after optimizing the injection curve, the resulting optimal injection curve is used as the basis for executing the filling process and is sent to the die-casting machine's control system. The die-casting machine drives the injection piston according to the speed pattern set by the optimal injection curve, pushing the molten aluminum alloy sequentially through the pressure chamber and gating system into the mold cavity, thus beginning the actual filling process. Injection speed is the most critical control variable in the filling process; changes in speed directly determine the flow pattern and filling behavior of the melt. Therefore, executing according to the optimal injection curve is a prerequisite for ensuring the flow stability of the filling process.

[0054] During the actual filling process, displacement and pressure sensors installed on the injection cylinder or pressure chamber, pressure sensors installed in the gating system or cavity, flow sensors installed at the main runner inlet or cavity, and temperature sensors installed on the mold cavity surface are used to collect real-time injection position and velocity data (which are converted to the current injection velocity), actual pressure data in the cavity or pressure chamber, actual melt flow data through the gating system, and actual temperature data at various detection points on the mold cavity surface at a set sampling frequency (typically in the millisecond to microsecond range). The above data are continuously collected and recorded in time series format. The actual pressure data reflects the magnitude of resistance and compaction state of the melt during filling; the actual flow data reflects the rate at which the melt enters the cavity and the filling progress; and the actual temperature data reflects the thermal state changes of the melt and the mold cavity surface.

[0055] After obtaining the aforementioned real-time data, it is input into a mathematical model for online calculation. Integrating multi-dimensional information such as pressure, flow rate, and temperature, the actual flow stability quantification index for the current moment is calculated. The basic logic of this calculation is consistent with the calculation method of the flow stability quantification evaluation index in step S200. The difference lies in that step S200 calculates theoretical values ​​based on simulated data under simulation conditions, while here the actual value is calculated based on real-time sensor data under actual filling conditions. This actual flow stability quantification index reflects the actual uniformity and stability of the melt flow during the current filling process.

[0056] After obtaining the actual flow stability quantification index, it is compared with the theoretical flow stability quantification index, and the deviation between the two is calculated. The theoretical flow stability quantification index is the optimal flow stability level achievable under the simulation conditions corresponding to the optimal injection curve, representing the ideal state theoretically achievable under the current filling conditions. If the deviation between the actual and theoretical values ​​is small, it indicates that the actual filling process is basically consistent with the simulation optimization state, and the flow stability is good, requiring no intervention. If the deviation exceeds the preset threshold, it indicates that an unstable flow state deviating from the expected state has occurred during the actual filling process, which may be due to the difference between the actual working conditions and the simulation conditions (such as material batch differences, mold temperature fluctuations, uneven spraying of release agent, etc.), in which case real-time intervention is required.

[0057] When the deviation exceeds a preset threshold, an online control mechanism is triggered. The control targets include two types: one is the adjustment of the injection speed, which uses the current stage's set value in the optimal injection curve as a benchmark and applies a reverse speed correction based on the direction and magnitude of the deviation, bringing the actual injection speed closer to the optimal curve; the other is the adjustment of relevant parameters in the mold temperature control strategy, which adjusts parameters such as the cooling medium flow rate or temperature in the corresponding area according to the direction and magnitude of the current temperature deviation and the control strategy. These two control methods can be used individually or simultaneously depending on the actual operating conditions to synergistically eliminate flow stability deviations.

[0058] By executing the filling process at each injection stage according to the optimal injection curve, the actual processing is based on the best process scheme obtained from simulation optimization, ensuring the flow stability of the filling process from the source. By collecting multi-source sensor data such as pressure, flow rate, and temperature in real time during the filling process, continuous and accurate perception of the filling state is achieved. By calculating the quantitative index of actual flow stability based on real-time sensor data and dynamically comparing it with the theoretical optimal index, and using whether the deviation exceeds the threshold as the criterion for control triggering, the control decision has a clear quantitative basis, avoiding the subjectivity and lag of relying on human experience judgment. When the deviation exceeds the threshold, by adjusting the injection speed and / or mold temperature control parameters in real time, the feedforward optimal process setting (optimal injection curve) is combined with the feedback online control to form a complete online closed-loop control of execution-monitoring-judgment-adjustment. This achieves real-time perception and active suppression of the unstable flow state during high-speed filling, effectively ensuring that the flow stability of the actual filling process is always maintained at a level close to the theoretical optimal level, thus meeting the stringent requirements of high density and high sealing performance for integrated power supply box components in new energy vehicles.

[0059] In this embodiment, by acquiring a three-dimensional geometric model and extracting cavity feature parameters, a zoned gating system is designed accordingly. This ensures that the configuration of the gating system's ingates matches the filling requirements of different areas of the cavity, balancing the flow distribution of melt entering each area of ​​the cavity from the source and effectively reducing the problem of uneven filling caused by differences in wall thickness. A multi-physics simulation model is established to obtain flow field and temperature field distribution data during the filling process. Based on the flow field distribution data, a quantitative evaluation index for flow stability is calculated, transforming the previously experience-based judgment of flow stability into a quantifiable and optimizable numerical index, providing an objective basis for the precise optimization of the subsequent injection curve. By using the quantitative evaluation index for flow stability as the optimization target and employing an iterative optimization algorithm to perform multi-segment collaborative optimization of the injection curve, the injection curve that optimizes the flow stability during the filling process and its corresponding theoretical quantitative flow stability index can be obtained, achieving scientific and systematic optimization of filling process parameters. Finally, the temperature of the mold cavity surface is determined based on the temperature field distribution data. By identifying areas with uneven temperature distribution and developing corresponding mold temperature control strategies, the mold temperature control can specifically compensate for temperature deviations in different areas, avoiding differences in melt flowability caused by uneven temperature, and providing reliable thermal boundary conditions for flow stability control. Finally, by performing mold filling according to the optimal injection curve and collecting pressure, flow rate, and temperature data in real time during the filling process, the actual flow stability quantitative indicators are dynamically compared with theoretical indicators. When the deviation exceeds the preset threshold, the injection speed or mold temperature control parameters are adjusted in real time, realizing real-time perception and active control of the unstable flow state during high-speed filling. This forms a complete closed-loop control system of "simulation pre-optimization - online monitoring - real-time control", which effectively reduces casting defects such as porosity and cold shuts caused by flow instability during high-density aluminum alloy structural parts filling, and significantly improves the filling quality and forming consistency of aluminum alloy structural parts. Especially for power supply box components with high density requirements, it can effectively ensure their internal quality and sealing performance.

[0060] Optionally, the mold cavity characteristic parameters include wall thickness distribution data for each region of the mold cavity; the step of determining the zoned gating system based on the mold cavity characteristic parameters to obtain the zoned gating system structural parameters includes: Based on the wall thickness distribution data, the mold cavity is divided into a thin-walled region and a thick-walled region; Set the cross-sectional area of ​​each ingate in the partitioned gating system so that the cross-sectional area of ​​the ingate corresponding to the thin-walled region is greater than the cross-sectional area of ​​the ingate corresponding to the thick-walled region. Based on the cross-sectional area of ​​each ingate and the corresponding cavity region location, the structural parameters of the partitioned gating system are generated.

[0061] Specifically, geometric feature recognition and data extraction are performed on the three-dimensional geometric model to obtain key parameters characterizing the filling characteristics of each region of the cavity. Among these, the wall thickness distribution data of each region of the cavity is the core parameter, which is obtained through the cross-sectional analysis function or wall thickness analysis tool of the three-dimensional drawing software: multiple cross-sections are cut along different directions on the three-dimensional geometric model, the specific values ​​of the casting wall thickness at each cross-section are measured, and these values ​​are labeled and recorded according to their spatial location, thereby obtaining complete distribution data of the wall thickness of each region of the cavity.

[0062] After obtaining the wall thickness distribution data, the mold cavity is divided into thin-walled and thick-walled regions based on the differences in wall thickness across different areas. The division is based on the following criteria: regions with thinner walls experience higher flow resistance and faster cooling during filling, making them difficult to fill; regions with thicker walls experience lower flow resistance and slower cooling, making them relatively easy to fill. Through statistical analysis of the wall thickness distribution data, a wall thickness threshold is established, classifying regions with wall thicknesses less than or equal to this threshold as thin-walled regions and regions with wall thicknesses greater than this threshold as thick-walled regions. This classification aims to allow the gating system design to be tailored to the filling requirements of different wall thickness regions; thin-walled regions require greater filling flow compensation, while thick-walled regions do not require excessive flow supply.

[0063] After completing the zoning, the cross-sectional area of ​​each genus in the zoned gating system is set. The zoned gating system includes a main runner, branch runners, and multiple genus gates. Each genus gate corresponds to a different area of ​​the mold cavity; that is, each genus gate is specifically responsible for supplying material to a particular area. The cross-sectional area of ​​each genus gate is independently set according to the wall thickness and filling flow requirements of the corresponding area. The core principle is: the cross-sectional area of ​​the genus gate corresponding to thin-walled areas is larger than that corresponding to thick-walled areas. Thin-walled areas, due to higher flow resistance, require larger genus gates to provide higher filling flow and more sufficient filling pressure; thick-walled areas, due to lower flow resistance, can be filled with smaller genus gates. This differentiated configuration ensures that the melt flow fronts of each area advance synchronously during the filling process, avoiding asynchronous filling problems caused by slow filling in thin-walled areas and excessively fast filling in thick-walled areas.

[0064] Finally, based on the cross-sectional area parameters of each gengate and its corresponding location in the cavity, a complete set of structural parameters for the zoned gating system is generated. This includes the number of gengates, the cross-sectional area of ​​each gengate, its corresponding location in the cavity, its spatial orientation, and the cross-sectional dimensions and orientation of the main runner and branch runners. These structural parameters constitute a complete geometric description of the zoned gating system, directly serving as the geometric input for subsequent multiphysics simulation models to verify and optimize the filling effect of the gating system.

[0065] In some embodiments, the cavity can be divided into three regions based on wall thickness: thin-walled, medium-walled, and thick-walled. The cross-sectional area of ​​the ingate corresponding to the medium-walled region is between that of the thin-walled and thick-walled regions. In one embodiment, for thin-walled regions with a wall thickness ≤ 1.7 mm, the corresponding ingate cross-sectional area is set to a first area value; for medium-thickness regions with a wall thickness > 1.7 mm and ≤ 3.0 mm, the corresponding ingate cross-sectional area is set to a second area value; for thick-walled regions with a wall thickness > 3.0 mm, the corresponding ingate cross-sectional area is set to a third area value; the first area value > the second area value > the third area value.

[0066] By extracting wall thickness distribution data from the 3D geometric model and dividing the cavity into thin-walled and thick-walled regions based on wall thickness differences, the design of the partitioned gating system can accurately correspond to the differences in filling difficulty in each region of the cavity. By setting the ingate cross-sectional area corresponding to the thin-walled region to be larger than that corresponding to the thick-walled region, flow compensation is achieved for the difficult-to-fill regions, balancing the flow distribution of melt entering each region of the cavity at the gating system level, effectively avoiding the problem of uneven filling caused by wall thickness differences. The final generated partitioned gating system structural parameters provide accurate geometric boundary conditions for subsequent multiphysics simulation and pressure injection curve optimization, laying the foundation for the gating system of the entire flow stability control method.

[0067] Optionally, the quantitative evaluation indexes for flow stability include the root mean square error of flow velocity and the volatility of the flow front advance velocity; The method for calculating the mean square error of the flow velocity is as follows: during the filling process, the mold cavity is divided into multiple detection areas according to a preset spatial position. For each detection area, the mean square error of the melt flow velocity of each spatial node in the same time window is calculated to obtain the mean square error of the flow velocity. The method for calculating the flow front advance speed fluctuation rate is as follows: during the filling process, based on the change of the position of the flow front of the aluminum alloy melt over time, the advance speed of the flow front at each moment is calculated, and the ratio of the standard deviation to the average value of the advance speed over the time series is used as the flow front advance speed fluctuation rate.

[0068] Specifically, the quantitative evaluation index for flow stability is a numerical indicator used to quantitatively characterize the uniformity and stability of melt flow during the filling process. It consists of two sub-indices: the root mean square error of the flow velocity and the fluctuation rate of the flow front advance velocity. These two sub-indices characterize the flow stability of the filling process from spatial and temporal dimensions, respectively. The former reflects the dispersion of melt flow velocity at different spatial locations within the cavity at the same moment (spatial uniformity), while the latter reflects the fluctuation of the melt flow front advance velocity at the same spatial location over time (temporal stationarity). The two sub-indices complement each other, together forming a complete quantitative description of the flow stability of the filling process.

[0069] The calculation process for the root mean square error of flow velocity is as follows: The root mean square error of flow velocity reflects the dispersion of melt flow velocity in different regions of the mold cavity during the filling process. The more uniform the velocity distribution, the smaller the root mean square error, indicating better flow stability.

[0070] First, the mold cavity needs to be divided into multiple detection zones according to a preset spatial location. Due to the complex structure of the cavity and the significant differences in flow behavior between different regions (slow flow velocity in thin-walled regions and fast flow velocity in thick-walled regions), it is necessary to divide the cavity into several detection zones to ensure relatively consistent flow behavior within each zone and facilitate statistical analysis. The detection zones can be divided using a uniform grid, dividing the cavity into several regular unit regions at equal intervals in three-dimensional space; alternatively, they can be partitioned based on the geometric characteristics of the cavity, such as dividing the cavity into different regions according to wall thickness distribution, segmenting according to flow path, or dividing according to structural features (ribs, corners, hole perimeters, etc.). For each detection zone, it typically contains multiple spatial nodes (i.e., discretized grid nodes in the simulation model), each node having a corresponding melt flow velocity value (including magnitude and direction) at each time point.

[0071] After dividing the detection area, the melt flow velocity data of each spatial node within each detection area is obtained within the same time window. The same time window means that for all detection areas, the velocity data within the same time period during the filling process is selected for statistical analysis. For example, the same time window is selected from the 10% to 90% filling stage to ensure that the velocity statistics of each area are comparable in time.

[0072] After obtaining the aforementioned velocity data, the root mean square deviation (RMS) of the melt flow velocity at each spatial node within each detection area is calculated over the same time window. The SMS is calculated as follows: first, the arithmetic mean of the melt flow velocities at all spatial nodes within the detection area at the same time is calculated; then, the deviation of each node's velocity value from the average is calculated; the squares of each deviation are summed and divided by the total number of nodes; finally, the square root is taken to obtain the SMS of the flow velocity at that time within the detection area. A smaller value indicates that the velocity at each node within the detection area is closer to the average, and the flow is more uniform; a larger value indicates that the velocity differences between nodes within the detection area are greater, and the flow is less uniform. By traversing all detection areas and all times, the SMS of the flow velocity at each time point within each area can be obtained, serving as a quantitative indicator of the uniformity of the flow space.

[0073] The calculation process for the velocity volatility of the flow front is as follows: The flow front advance velocity fluctuation rate reflects the degree of fluctuation of the melt flow front advance velocity over time during the filling process. The smaller the advance velocity fluctuation, the smoother the filling process and the better the flow stability.

[0074] The flow front refers to the interface between the molten aluminum alloy and the gas (or vacuum) inside the mold cavity during the filling process; it is the boundary between the filled and unfilled areas of the melt. During the filling process, the flow front starts from the ingate and gradually advances until it fills the entire mold cavity. At any given moment, the flow front is located in a specific spatial position within the mold cavity, and its advancement speed reflects how quickly the melt fills the cavity.

[0075] After obtaining the spatial position of the flow front at each moment, the advancing velocity of the flow front at each moment is calculated. The advancing velocity refers to the distance the flow front advances in the cavity per unit time, which can be obtained by dividing the distance between the positions of the flow front at adjacent moments by the time interval. For example, if the flow front is at position P1 at time t1 and at position P2 at time t2, then the average advancing velocity during the time interval from time t1 to time t2 is the distance between positions P1 and P2 divided by the time difference between time t2 and time t1. Through the above calculation, the advancing velocity sequence of the flow front at each moment during the filling process can be obtained.

[0076] After obtaining the flow front propulsion velocity sequence at each time point, the ratio of the standard deviation to the mean of this sequence is calculated, which is the flow front propulsion velocity volatility. Specifically: first, calculate the arithmetic mean of the flow front propulsion velocity at each time point throughout the entire filling process (or within a specified time window); then calculate the deviation of the propulsion velocity at each time point from the mean, square each deviation, sum them, divide by the total number of data points, and take the square root to obtain the standard deviation; finally, divide the standard deviation by the mean to obtain the volatility. This ratio is dimensionless, eliminating the influence of the absolute value of the velocity on the judgment of the degree of volatility. The smaller the value, the more stable the flow front propulsion velocity is over time and the smaller the volatility; the larger the value, the more violent the propulsion velocity fluctuations and the more unstable the filling process.

[0077] The two sub-indicators mentioned above, the mean square error of flow velocity (spatial dimension) and the volatility of the flow front propulsion velocity (time dimension), together constitute the quantitative evaluation index of flow stability. In the subsequent optimization of the injection curve, the optimization objective can be the weighted sum of the two sub-indicators or other comprehensive forms to achieve a comprehensive optimization of the flow stability during the filling process.

[0078] By defining the mean square error of flow velocity and the fluctuation rate of flow front advance velocity as sub-indices of the quantitative evaluation index of flow stability from both spatial and temporal dimensions, a comprehensive and accurate quantification of the flow stability of the filling process is achieved. That is, the mean square error of flow velocity reflects the uniformity of melt flow velocity in each region of the cavity from a spatial dimension, while the fluctuation rate of flow front advance velocity reflects the stability of the filling process from a temporal dimension. The two complement each other and are indispensable. By transforming the abstract concept of flow stability into calculable and comparable numerical indicators, the process optimization of the filling process has a clear quantitative target, providing an objective and accurate optimization basis for the subsequent multi-segment collaborative optimization of the injection curve. This fundamentally solves the technical problem in the existing technology that the flow stability relies on empirical qualitative judgment and lacks quantitative evaluation methods.

[0079] Optionally, the multiphysics simulation model includes at least a flow field model, a temperature field model, and a solidification field model. The multiphysics simulation model is used to simulate the flow behavior of aluminum alloy melt during the filling process and to obtain the flow field distribution data, the temperature field distribution data, and the solidification front advancement data at each spatial node at each time.

[0080] Specifically, a multiphysics simulation model is used to simulate the flow behavior, temperature changes, and solidification process of the melt during the filling process of high-density aluminum alloy structural parts. Since the die-casting filling process involves the simultaneous occurrence and coupling of multiple physical phenomena such as fluid flow, heat transfer, and phase transformation, a single physics model cannot fully describe the filling process. Therefore, a multiphysics simulation model is needed, which includes a flow field model, a temperature field model, and a solidification field model. This multiphysics simulation model is used to simulate the flow behavior of the aluminum alloy melt during the filling process, acquiring flow field distribution data, temperature field distribution data, and solidification front advancement data at each spatial node at each time point.

[0081] The flow field model is used to simulate the flow behavior of molten aluminum alloy after entering the mold cavity through the gating system under injection pressure. The flow field model solves the governing equations of fluid dynamics (i.e., the Navier-Stokes equations), which describe the distribution and changes in velocity and pressure of the melt during the filling process. In the flow field model, the molten aluminum alloy is treated as an incompressible or weakly compressible fluid, and its flow behavior is influenced by inertial forces, viscous forces, gravity, and pressure gradients. The main inputs to the flow field model include: the velocity or flow rate conditions at the melt inlet (determined by the injection curve), the geometric boundaries of the mold cavity and gating system (determined by the structural parameters of the zoned gating system), and material properties such as the density and viscosity of the melt. By solving the fluid dynamics equations, the flow field model outputs flow field distribution data for each spatial node at each time point. This includes the velocity components of the melt in the X, Y, and Z directions at each grid node, the magnitude and direction of the combined velocity, and the pressure value at that node. The flow field distribution data reflects the flow behavior of the melt in different regions of the mold cavity, such as whether the velocity distribution is uniform, the presence of eddies or backflow, and the propulsion pattern of the flow front.

[0082] The temperature field model is used to simulate the heat exchange between the melt and the mold cavity during the filling process, as well as the temperature change of the melt itself. During the filling process, high-temperature aluminum alloy melt (typically with a pouring temperature of 650℃-720℃) enters the relatively cool mold cavity (the mold preheating temperature is typically 150℃-250℃). There is a strong heat exchange between the two; the heat from the melt is transferred to the mold through the cavity walls, causing the melt temperature to drop and the mold temperature to rise. The temperature field model solves for the heat transfer governing equation (i.e., the Fourier heat conduction equation), which describes the heat transfer process in the melt and mold, including heat conduction, convection, and radiation. The main inputs to the temperature field model include: the initial temperature of the melt (pouring temperature), the initial temperature of the mold (preheating temperature), the thermal properties of the melt and mold materials such as thermal conductivity, specific heat capacity, and density, and the interfacial heat transfer coefficient between the melt and the mold. By solving the heat transfer equation, the temperature field model outputs temperature field distribution data for each spatial node at each moment, including the temperature values ​​of each node of the melt inside the cavity and the temperature values ​​of each node on the surface of the mold cavity. The temperature field distribution data reflects the cooling and solidification behavior of the melt during the filling process, i.e., the difference in the rate of temperature decrease in different regions, and whether there is local overheating or undercooling.

[0083] The solidification field model is used to simulate the solidification process of a melt transitioning from a liquid to a solid state. When the temperature of the aluminum alloy melt drops below the liquidus temperature, it begins to crystallize and solidify, releasing latent heat of crystallization and transitioning from a liquid to a solid state. Based on temperature field distribution data, the solidification field model compares the temperature of each spatial node with the liquidus and solidus temperatures of the aluminum alloy. Nodes with temperatures above the liquidus temperature are in a liquid state, nodes with temperatures below the solidus temperature are in a solid state, and nodes with temperatures between the two are in a state of coexistence of solid and liquid phases (a pasty region). The solidification front is located at the interface between the liquid region and the pasty or solid region. The solidification field model outputs solidification front advancement data, specifically the spatial position and advancement velocity of the solidification front within the mold cavity at each moment. This includes the three-dimensional coordinates of the solidification front within the cavity, the advancement velocity of the solidification front along the normal direction, and the solidification time at each location. Solidification front advance data reflects the solidification behavior of the melt during the filling process. If the solidification front advances too quickly, the melt may solidify before it completely fills the cavity (risk of cold shut), while if it advances too slowly, it will affect production efficiency.

[0084] The three physical field models mentioned above are coupled: the velocity and pressure distribution calculated by the flow field model affects the heat transfer intensity of the melt to the cavity wall, thus influencing the temperature field; the temperature distribution calculated by the temperature field determines the viscosity and solidification state of the melt, which in turn affects the flow behavior of the flow field; the solidification field model makes solidification determination based on the temperature field data, and the formation of the solidified shell changes the effective flow cross-sectional area of ​​the cavity, which in turn affects the flow field. During the simulation, the three physical field models achieve bidirectional coupling through data exchange, jointly completing a comprehensive description of the filling process.

[0085] By establishing a multiphysics simulation model that includes a flow field model, a temperature field model, and a solidification field model, a comprehensive and accurate numerical simulation of the high-speed filling process of high-density aluminum alloy structural parts was achieved. The flow field model captures the velocity and pressure distribution of the melt (reflecting flow behavior), the temperature field model captures the temperature changes of the melt and the mold (reflecting thermal state), and the solidification field model captures the solidification process of the melt (reflecting phase transformation behavior). The three models, through bidirectional coupling, together constitute a complete physical description of the filling process. The flow field distribution data, temperature field distribution data, and solidification front advancement data obtained through this multiphysics simulation model provide a reliable quantitative data foundation for the calculation of flow stability quantitative evaluation index (flow field distribution data), the formulation of mold temperature control strategies (temperature field distribution data), and the verification of filling process constraints (solidification front advancement data), respectively.

[0086] Optionally, the injection curve includes at least a low-speed phase, a high-speed phase, and a pressurization phase; When the flow front of the aluminum alloy melt has not reached the ingate position, switch to the low-speed stage, and the low-speed stage controls the injection speed within the preset low-speed range. When the flow front of the aluminum alloy melt reaches the ingate position, the process switches to the high-speed stage. The high-speed stage controls the injection speed within a preset high-speed range until the cavity filling rate reaches a preset filling rate threshold. When the cavity filling rate reaches the preset filling rate threshold, the process switches to the pressurization stage. In the pressurization stage, the injection speed is reduced to a preset pressurization range, and pressurization pressure is applied until the solidification fraction reaches the preset threshold. The switching time between the low-speed stage and the high-speed stage is determined based on the time when the flow front of the aluminum alloy melt reaches the ingate in the multiphysics simulation model.

[0087] Optionally, the high-speed phase includes a first high-speed sub-phase and a second high-speed sub-phase; Wherein, the injection speed of the first high-speed sub-stage is greater than that of the second high-speed sub-stage. The first high-speed sub-stage is used to enable the melt to quickly pass through the partitioned gating system and enter the cavity. The second high-speed sub-stage is used to enable the melt to smoothly fill the thin-walled area of ​​the cavity. The switching time between the first high-speed sub-stage and the second high-speed sub-stage is determined based on the moment when the aluminum alloy melt flow front enters the thin-walled region of the cavity in the multiphysics simulation model.

[0088] Specifically, the injection profile is the curve showing the change in injection speed over time during the die casting process. In actual production, the injection profile is typically divided into a low-speed stage, a high-speed stage, and a pressurization stage. The transitions between these stages are based on the position of the aluminum alloy molten flow front and the cavity filling state.

[0089] In some embodiments, the injection speed is low when the molten aluminum alloy flow front has not yet reached the ingate. The injection speed during this low-speed phase is typically controlled within the range of 0.1 m / s to 0.5 m / s. At this speed, the injection punch smoothly propels the melt forward through the pressure chamber and gating system, aiming to avoid splashing or air entrapment within the gating system. Taking a typical mold where the ingate is approximately 350 mm from the end of the pressure chamber as an example, the punch needs to advance approximately 350 mm during the low-speed phase for the melt to reach the ingate from the pressure chamber.

[0090] When the molten aluminum alloy flow front reaches the ingate, the injection speed is switched from the low-speed stage to the high-speed stage. The injection speed in the high-speed stage is typically controlled within the range of 1.5 m / s to 8.0 m / s. For high-density aluminum alloy structural parts, thin-walled areas cool rapidly (the thinner the wall, the faster the injection speed is required). The purpose of high-speed filling is to rapidly fill the entire cavity before the temperature drops. The high-speed stage continues until the cavity filling rate reaches a preset filling rate threshold. The filling rate threshold is typically set to 60% to 90% of the total cavity volume; for example, switching occurs when the filling rate reaches 60%, or when it reaches 80% to 90%.

[0091] When the cavity filling rate reaches the preset filling rate threshold, the process switches to the pressurization stage. During the pressurization stage, the injection speed is reduced to the range of 0.5 m / s to 1.0 m / s, while pressurization pressure is applied. The pressurization pressure is typically 30 MPa to 35 MPa, and the pressurization stage continues until the solidification fraction reaches the preset threshold. The solidification fraction threshold is usually set to switch when the solidification fraction reaches 40% to 80%; for example, some designs recommend starting pressurization when the solidification fraction reaches 80%.

[0092] In some embodiments, the high-speed stage is further divided into a first high-speed sub-stage and a second high-speed sub-stage. The injection velocity of the first high-speed sub-stage (e.g., 3.0 m / s to 5.0 m / s) is greater than that of the second high-speed sub-stage (e.g., 1.5 m / s to 3.0 m / s). The first high-speed sub-stage is used to enable the melt to quickly pass through the zoned gating system into the cavity, overcoming the resistance of the ingate; the second high-speed sub-stage is used to enable the melt to smoothly fill the thin-walled region of the cavity, avoiding jetting or turbulence due to excessive velocity. The switching time between the two sub-stages is determined based on the moment when the flow front of the aluminum alloy melt enters the thin-walled region of the cavity in the multiphysics simulation model.

[0093] The switching times for each stage are not fixed empirical values, but are determined based on the simulation results of the multiphysics simulation model. The simulation calculates the times when the aluminum alloy melt flow front reaches the ingate, the times when the cavity filling rate reaches a preset threshold, and the times when the solidification fraction reaches a preset threshold. These times are used as the basis for the switching times of each stage during the actual filling process. For example, the simulation results for a high-density aluminum alloy structural component show that the aluminum alloy melt flow front reaches the ingate at 0.05 s (switching between low and high speed), the filling rate reaches 85% at 0.12 s (switching between high speed and pressurization), and the solidification fraction reaches 75% at 0.20 s (end of pressurization).

[0094] By dividing the injection curve into low-speed, high-speed, and pressurization stages, and switching between stages based on the position of the aluminum alloy melt flow front and the cavity filling rate, precise matching between injection speed and filling process is achieved. Furthermore, by subdividing the high-speed stage into a first high-speed sub-stage and a second high-speed sub-stage, the melt can smoothly fill the thin-walled area at a lower speed after rapidly entering the cavity through the gating system. This ensures filling efficiency in the thin-walled area while avoiding flow instability problems such as jetting and turbulence caused by excessive speed. The switching time of each stage is based on the moment when the aluminum alloy melt flow front reaches the critical position, as determined by the multiphysics simulation model. This provides a precise quantitative basis for determining the switching timing rather than relying on empirical estimation, thereby ensuring precise matching and dynamic adaptation between the speed state and filling process at each stage.

[0095] Optionally, the mold temperature control strategy includes: When the temperature of any region in the thin-walled region or the thick-walled region exceeds the preset upper temperature limit, the flow rate of the cooling medium in the corresponding region is increased or the temperature of the cooling medium is decreased. When the temperature of any region in the thin-walled region or the thick-walled region is lower than the preset lower limit, the flow rate of the cooling medium in the corresponding region is reduced or heating is started. The preset upper temperature limit and preset lower temperature limit are determined based on the solidification characteristics of the melt in this region in the multiphysics simulation model.

[0096] Specifically, during the die-casting filling process, the temperature distribution on the mold cavity surface directly affects the flow behavior of the melt. When the temperature in a certain area of ​​the cavity deviates from the preset reasonable range, corresponding control measures need to be taken to restore the temperature to the reasonable range. This process sets temperature control logic separately for thin-walled and thick-walled regions.

[0097] Because the thin-walled and thick-walled regions have different wall thicknesses, their melt solidification characteristics differ. The thin-walled region has a smaller melt volume, less heat, and a faster cooling rate, making it more difficult to maintain fluidity. Therefore, the upper and lower preset temperature limits for the thin-walled and thick-walled regions are different. These upper and lower preset temperature limits are not fixed empirical values ​​but are determined based on the solidification characteristics of the melt in that region within a multiphysics simulation model.

[0098] Specifically, from the temperature field distribution data and solidification field data of the multiphysics simulation model, the temperature change curve and solidification initiation temperature (i.e., liquidus temperature) of the melt during the filling process in this region are extracted. Based on this, a reasonable working temperature range for this region is set: the upper limit of the preset temperature is usually set at a certain margin below the liquidus temperature to ensure that the melt always maintains sufficient fluidity during the filling process and does not solidify prematurely; the lower limit of the preset temperature is set above the minimum temperature that ensures the melt can fill the region smoothly. Below this temperature, the melt viscosity is too high and there is a risk of cold shut-off. For example, the liquidus temperature of the aluminum alloy used in a high-density aluminum alloy structural component is 615℃. Simulation shows that the temperature drop rate of the melt in the thin-walled region during the filling process is approximately 50℃ / s. To ensure that the temperature of the melt in the thin-walled region does not fall below 570℃ before the end of filling (still about 30℃ above the solidus), the preset lower limit of the temperature for this region is set to 580℃, and the preset upper limit is set to 600℃. In the thick-walled region, due to the large melt volume and high heat content, the temperature drop rate is slower (approximately 20℃ / s), so the preset lower limit of the temperature for this region is set to 590℃, and the preset upper limit is set to 620℃. The above values ​​are for illustrative purposes only; actual values ​​will vary depending on the alloy composition, casting wall thickness, and process conditions.

[0099] The specific execution logic of the control strategy is as follows: When the temperature of any region in the thin-walled or thick-walled area exceeds the preset upper limit of that region's temperature, it indicates insufficient heat dissipation and excessively high temperature in that region, potentially posing a risk of overheating of the melt. In this case, increasing the flow rate of the cooling medium in that region or decreasing its temperature is crucial. Increasing the flow rate removes more heat, while decreasing the medium temperature increases the heat exchange temperature difference; both enhance the cooling effect and bring the region's temperature back to a reasonable range. When the temperature of any region in the thin-walled or thick-walled area falls below the preset lower limit of that region's temperature, it indicates that the region is cooling too quickly and its temperature is too low, potentially worsening the melt's fluidity. In this case, reducing the flow rate of the cooling medium in that region or initiating heating is essential. Reducing the flow rate decreases heat loss, while initiating heating actively replenishes heat; both weaken the cooling effect or raise the temperature, bringing the region's temperature back to a reasonable range. For example, for integrated power supply enclosure components in new energy vehicles, precise mold temperature control directly affects the density and airtightness of the casting. Uneven temperature can lead to localized shrinkage or microcracks, thereby affecting the sealing performance and long-term reliability of the power supply enclosure.

[0100] By setting preset upper and lower temperature limits based on simulated solidification characteristics for thin-walled and thick-walled regions respectively, differentiated and precise control of mold temperature is achieved. The lower temperature limit for thin-walled regions is set higher to ensure their fluidity, while the upper temperature limit for thick-walled regions is set higher to fully utilize their heat dissipation conditions. This avoids the problem of overcooling in thin-walled regions and overheating in thick-walled regions caused by using a uniform temperature standard to control regions with different wall thicknesses. When the temperature of a certain region deviates from the preset range, specific control methods such as increasing or decreasing the flow rate of the cooling medium, decreasing the temperature of the cooling medium, or starting heating are used to achieve precise adjustment of the temperature of each region in the mold cavity. This ensures that the temperature of each region is always maintained within a reasonable range that guarantees good melt fluidity, thus providing reliable thermal boundary conditions to ensure the flow stability of the melt during the filling process. This is particularly important for integrated power supply box components in new energy vehicles, which need to withstand large heat loads and sealing requirements during operation. A stable temperature field is a prerequisite for ensuring their internal quality and long-term reliability.

[0101] Optionally, the step of calculating the quantitative index of actual flow stability at the current moment based on the actual pressure data, the actual flow rate data, and the actual temperature data includes: The actual pressure data, the actual flow rate data, and the actual temperature data are input into a pre-built filling state evaluation model, and the filling state evaluation model outputs the quantitative index of the actual flow stability at the current moment. The filling state evaluation model uses pressure data sequences, flow data sequences, and temperature data sequences collected during the historical filling process as input features, and uses the measured flow stability evaluation index at the corresponding time as training labels. It is trained using a long short-term memory network or a gated recurrent unit network architecture.

[0102] Optionally, the method for controlling the flow stability of the aluminum alloy structural component during high-speed filling further includes: The actual pressure data, actual flow rate data, actual temperature data, and corresponding quantitative indicators of actual flow stability collected during each filling process are used as new samples to perform online incremental learning on the filling state evaluation model, update the model parameters, and enable the updated model to be used in the flow stability evaluation of subsequent filling processes.

[0103] Specifically, during the actual filling process, although physical quantities such as pressure, flow rate, and temperature can be collected in real time, these raw sensor data cannot be directly equated with flow stability quantification indicators. Flow stability quantification indicators are comprehensive evaluation results obtained after performing spatial and temporal statistical analysis on flow field distribution data. However, in actual processing, it is impossible to obtain complete flow field distribution data inside the cavity as in simulation. Therefore, it is necessary to establish a filling state evaluation model to indirectly deduce the actual flow stability quantification indicators at the current moment using real-time collected pressure, flow rate, and temperature data.

[0104] The filling state evaluation model is a pre-trained neural network model, and its training process is as follows: The actual pressure data sequence, actual flow rate data sequence, and actual temperature data sequence collected during the historical filling process are used as input features. These data sequences are continuously changing sensor signals over time, reflecting the dynamic changes during the filling process. The pressure data reflects the changes in resistance experienced by the melt, the flow rate data reflects the changes in the rate at which the melt enters the cavity, and the temperature data reflects the changes in the thermal state of the melt and the mold.

[0105] The model's training labels are the measured flow stability evaluation indicators at corresponding time points. These indicators can be obtained through offline analysis of flow field data from historical filling processes (e.g., comprehensive evaluation of high-frequency pressure, flow rate, and temperature data collected during historical filling processes, or back-calibration based on quality inspection results of the same batch of castings). The model is trained using a Long Short-Term Memory (LSTM) network or a Gated Recurrent Unit (GRU) network architecture. LSTM and GRU are deep learning network architectures specifically designed for processing time-series data, possessing the ability to remember long-term dependencies. This means they can capture the dynamic patterns of pressure, flow rate, and temperature evolution over time during the filling process and encode the mapping relationship between these dynamic patterns and the flow stability evaluation indicators into the network parameters. After training, the model can quickly output the predicted value of the actual flow stability quantification indicator at the current time point based on real-time sensor data input during actual filling processes.

[0106] The online application of the filling state assessment model also includes an online incremental learning process. Online incremental learning refers to the continuous updating of the model during the actual filling process. After each filling is completed, the actual pressure data, actual flow rate data, actual temperature data, and corresponding quantitative indicators of actual flow stability (obtained through model output or offline analysis) collected during this filling process are used as new samples to incrementally train the LSTM or GRU network and update the model parameters. Incremental learning adopts the elastic weight consolidation algorithm. The core mechanism of this algorithm is to apply additional regularization constraints to the key network parameters that affect the model performance when training with new samples.

[0107] Specifically, EWC (Elastic Weight Consolidation) calculates the importance weights of each network parameter to historical tasks (based on the Fisher information matrix). During incremental training, it imposes greater constraints on highly important parameters (limiting their variation) and less constraint on less important parameters (allowing them to adjust freely to adapt to new data). This allows the model to absorb new knowledge while retaining the memory of old knowledge, effectively preventing the model's predictive ability for old samples from degrading due to training on new samples. The updated model parameters replace the original parameters and are used in the flow stability evaluation of the subsequent filling process.

[0108] By constructing a filling state evaluation model using LSTM or GRU networks and taking real-time collected pressure, flow, and temperature data as input, the model indirectly calculates the quantitative index of actual flow stability at the current moment. This solves the technical problem of difficulty in quantifying and evaluating flow stability in actual processing due to the inability to directly measure the flow field distribution inside the cavity, enabling online control to have real-time and accurate flow stability state perception capabilities. By employing the time series processing capabilities of LSTM or GRU, the model fully explores the dynamic evolution of pressure, flow, and temperature data during the filling process, giving the model prediction temporal accuracy and stability. By using actual data from each filling process as new samples for online incremental learning, the model can continuously absorb new working condition experience and adapt to changes such as material differences between batches and mold wear in actual production, avoiding the decline in prediction accuracy over time due to the model's adherence to historical experience. By using the EWC algorithm in incremental learning and applying regularization constraints to key network parameters during training with new samples, the catastrophic forgetting problem in the incremental learning process is effectively mitigated, ensuring that the model maintains stable prediction accuracy for various working conditions during continuous updates. The aforementioned mechanism enables the filling state assessment model to continuously accumulate knowledge and improve performance as production batches progress, providing continuous and reliable state perception and assessment support for the online control of flow stability during the high-speed filling process of high-density aluminum alloy structural parts in long-term, large-volume production.

[0109] like Figure 2 As shown in the figure, an embodiment of the present invention provides a flow stability control system for aluminum alloy structural components during high-speed filling, used to implement the above-described method for controlling the flow stability of aluminum alloy structural components during high-speed filling. The system includes: The acquisition unit is used to acquire the three-dimensional geometric model of the high-density aluminum alloy structural parts and extract the mold cavity feature parameters. Based on the mold cavity feature parameters, the partitioned gating system is determined, and the structural parameters of the partitioned gating system are obtained. The simulation unit is used to establish a multiphysics simulation model of the filling process based on the structural parameters of the partitioned casting system and the preset injection process parameters. The multiphysics simulation model is used to obtain the flow field distribution data and temperature field distribution data at each moment during the filling process. The flow stability quantitative evaluation index is calculated based on the flow field distribution data. The optimization unit is used to optimize the injection curve in multiple segments by using the flow stability quantitative evaluation index as the optimization target, and using a preset filling time window and a preset solidification fraction threshold as constraints, and to output the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve. The processing unit is used to determine the uneven temperature distribution areas on the surface of the mold cavity based on the temperature field distribution data, and to determine the mold temperature control strategy corresponding to each uneven temperature distribution area. The control unit is used to execute the filling of each injection stage according to the optimal injection curve, and to collect actual pressure data, actual flow data and actual temperature data in real time during the filling process. Based on the actual pressure data, actual flow data and actual temperature data, it calculates the actual flow stability quantification index at the current moment. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, it adjusts the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy in real time to achieve online control of flow stability during high-speed filling.

[0110] This invention provides a flow stability control device for aluminum alloy structural components during high-speed filling, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the flow stability control method for aluminum alloy structural components during high-speed filling as described above when the computer program is executed.

[0111] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the flow stability control method for aluminum alloy structural components during high-speed filling, as described above.

[0112] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for controlling the flow stability of aluminum alloy structural components during high-speed filling, characterized in that, include: A three-dimensional geometric model of a high-density aluminum alloy structural component is obtained and the characteristic parameters of the mold cavity are extracted. Based on the characteristic parameters of the mold cavity, a partitioned gating system is determined, and the structural parameters of the partitioned gating system are obtained. Based on the structural parameters of the partitioned casting system and the preset injection process parameters, a multiphysics simulation model of the filling process is established, and the flow field distribution data and temperature field distribution data at each moment during the filling process are obtained through the multiphysics simulation model. Calculate the quantitative evaluation index of flow stability based on the flow field distribution data; Using the aforementioned flow stability quantitative evaluation index as the optimization objective, and with a preset filling time window and a preset solidification fraction threshold as constraints, an iterative optimization algorithm is used to perform multi-segment collaborative optimization of the injection curve, outputting the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve. Based on the temperature field distribution data, determine the regions with uneven temperature distribution on the surface of the mold cavity, and determine the mold temperature control strategy corresponding to each of the uneven temperature distribution regions. The filling process is performed according to the optimal injection curve, and actual pressure data, actual flow rate data, and actual temperature data are collected in real time during the filling process. The actual flow stability quantification index at the current moment is calculated based on the actual pressure data, actual flow rate data, and actual temperature data. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy are adjusted in real time to achieve online control of flow stability during high-speed filling.

2. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 1, characterized in that, The mold cavity characteristic parameters include wall thickness distribution data for each region of the mold cavity; the step of determining the zoned gating system based on the mold cavity characteristic parameters to obtain the zoned gating system structural parameters includes: Based on the wall thickness distribution data, the mold cavity is divided into thin-walled and thick-walled regions; Set the cross-sectional area of ​​each ingate in the partitioned gating system so that the cross-sectional area of ​​the ingate corresponding to the thin-walled region is greater than the cross-sectional area of ​​the ingate corresponding to the thick-walled region. Based on the cross-sectional area of ​​each ingate and the corresponding cavity region location, the structural parameters of the partitioned gating system are generated.

3. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 1, characterized in that, The quantitative evaluation indicators for flow stability include the mean square error of flow velocity and the volatility of the flow front advance velocity. The method for calculating the mean square error of the flow velocity is as follows: during the filling process, the mold cavity is divided into multiple detection areas according to a preset spatial position. For each detection area, the mean square error of the melt flow velocity of each spatial node in the same time window is calculated to obtain the mean square error of the flow velocity. The method for calculating the flow front advance speed fluctuation rate is as follows: during the filling process, based on the change of the position of the flow front of the aluminum alloy melt over time, the advance speed of the flow front at each moment is calculated, and the ratio of the standard deviation to the average value of the advance speed over the time series is used as the flow front advance speed fluctuation rate.

4. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 1, characterized in that, The multiphysics simulation model includes at least a flow field model, a temperature field model, and a solidification field model. The multiphysics simulation model is used to simulate the flow behavior of aluminum alloy melt during the filling process and to obtain the flow field distribution data, the temperature field distribution data, and the solidification front advancement data at each spatial node at each time.

5. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 1, characterized in that, The injection curve includes at least a low-speed phase, a high-speed phase, and a pressurization phase; When the flow front of the aluminum alloy melt has not reached the ingate position, switch to the low-speed stage, and the low-speed stage controls the injection speed within the preset low-speed range. When the flow front of the aluminum alloy melt reaches the ingate position, the process switches to the high-speed stage. The high-speed stage controls the injection speed within a preset high-speed range until the cavity filling rate reaches a preset filling rate threshold. When the cavity filling rate reaches the preset filling rate threshold, the process switches to the pressurization stage. In the pressurization stage, the injection speed is reduced to a preset pressurization range, and pressurization pressure is applied until the solidification fraction reaches the preset threshold. The switching time between the low-speed stage and the high-speed stage is determined based on the time when the flow front of the aluminum alloy melt reaches the ingate in the multiphysics simulation model.

6. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 5, characterized in that, The high-speed phase includes a first high-speed sub-phase and a second high-speed sub-phase; Wherein, the injection speed of the first high-speed sub-stage is greater than that of the second high-speed sub-stage. The first high-speed sub-stage is used to enable the melt to quickly pass through the partitioned gating system and enter the cavity. The second high-speed sub-stage is used to enable the melt to smoothly fill the thin-walled area of ​​the cavity. The switching time between the first high-speed sub-stage and the second high-speed sub-stage is determined based on the moment when the aluminum alloy melt flow front enters the thin-walled region of the cavity in the multiphysics simulation model.

7. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 2, characterized in that, The mold temperature control strategy includes: When the temperature of any region in the thin-walled region or the thick-walled region exceeds the preset upper temperature limit, the flow rate of the cooling medium in the corresponding region is increased or the temperature of the cooling medium is decreased. When the temperature of any region in the thin-walled region or the thick-walled region is lower than the preset lower limit, the flow rate of the cooling medium in the corresponding region is reduced or heating is started. The preset upper temperature limit and preset lower temperature limit are determined based on the solidification characteristics of the melt in this region in the multiphysics simulation model.

8. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 1, characterized in that, The step of calculating the quantitative index of actual flow stability at the current moment based on the actual pressure data, the actual flow rate data, and the actual temperature data includes: The actual pressure data, the actual flow rate data, and the actual temperature data are input into a pre-built filling state evaluation model, and the filling state evaluation model outputs the quantitative index of the actual flow stability at the current moment. The filling state evaluation model uses pressure data sequences, flow data sequences, and temperature data sequences collected during the historical filling process as input features, and uses the measured flow stability evaluation index at the corresponding time as training labels. It is trained using a long short-term memory network or a gated recurrent unit network architecture.

9. The method for controlling the flow stability of aluminum alloy structural components during high-speed filling according to claim 8, characterized in that, The method for controlling the flow stability of the aluminum alloy structural component during high-speed filling also includes: The actual pressure data, actual flow rate data, actual temperature data, and corresponding quantitative indicators of actual flow stability collected during each filling process are used as new samples to perform online incremental learning on the filling state evaluation model, update the model parameters, and enable the updated model to be used in the flow stability evaluation of subsequent filling processes.

10. A flow stability control system for aluminum alloy structural components during high-speed filling, characterized in that, A method for controlling the flow stability of aluminum alloy structural components during high-speed filling as described in any one of claims 1 to 9, the system comprising: The acquisition unit is used to acquire the three-dimensional geometric model of the high-density aluminum alloy structural parts and extract the mold cavity feature parameters. Based on the mold cavity feature parameters, the partitioned gating system is determined, and the structural parameters of the partitioned gating system are obtained. The simulation unit is used to establish a multiphysics simulation model of the filling process based on the structural parameters of the partitioned casting system and the preset injection process parameters. The multiphysics simulation model is used to obtain the flow field distribution data and temperature field distribution data at each moment during the filling process. The flow stability quantitative evaluation index is calculated based on the flow field distribution data. The optimization unit is used to optimize the injection curve in multiple segments by using the flow stability quantitative evaluation index as the optimization target, and using a preset filling time window and a preset solidification fraction threshold as constraints, and to output the optimal injection curve and the theoretical flow stability quantitative index corresponding to the optimal injection curve. The processing unit is used to determine the uneven temperature distribution areas on the surface of the mold cavity based on the temperature field distribution data, and to determine the mold temperature control strategy corresponding to each uneven temperature distribution area. The control unit is used to execute the filling of each injection stage according to the optimal injection curve, and to collect actual pressure data, actual flow data and actual temperature data in real time during the filling process. Based on the actual pressure data, actual flow data and actual temperature data, it calculates the actual flow stability quantification index at the current moment. When the deviation between the actual flow stability quantification index and the theoretical flow stability quantification index exceeds a preset threshold, it adjusts the injection speed deviation of the current injection stage relative to the optimal injection curve and / or the control parameters in the mold temperature control strategy in real time to achieve online control of flow stability during high-speed filling.