V-process casting compound mold based on negative pressure and conformal cooling and shape control method

Abstract

The application discloses a V-process casting composite mold based on negative pressure and conformal cooling and a shape control method, and belongs to the field of metal casting. The method comprises the following steps: integrating a conformal cooling channel in a mold sample body, forming space-time cooperation with a negative pressure air extraction system, and realizing active heat regulation; a digital twin model containing heat conduction, phase change, heat-force coupling and flow heat exchange is constructed, and sensor data is assimilated in real time through ensemble Kalman filtering. The application realizes the coupling regulation of heat-flow-force-phase change multi-physical field, eliminates shrinkage holes and shrinkage porosity, refines the solidification structure, reduces the residual stress, improves the size precision and mechanical properties, and provides a reliable solution for high-quality forming of complex and precise castings in the fields of aerospace, energy equipment and the like. The application solves the technical problems that, in traditional V-process casting, the heat-flow-force-phase change multi-field coupling is out of balance due to the passive heat dissipation mode and the open-loop control system, and the internal quality of the castings cannot be accurately regulated.

Description

Technical Field

[0001] This invention relates to the field of metal casting, specifically to a V-process casting composite mold and shape control method based on negative pressure and conformal cooling. Background Technology

[0002] The core mechanism of the Vacuum Sealed Molding Process (V-process) lies in creating a negative pressure environment inside the sand box by wrapping the pattern in a plastic film and applying a vacuum. This environment, combined with atmospheric pressure, forces the binder-free dry sand into a compact shape. This technological characteristic gives V-process casting excellent technical and economic performance in reducing environmental pollution, lowering molding material costs, and improving flexible production capabilities.

[0003] The invention patent with publication number CN118253745A accelerates heat exchange by configuring a liftable spray device around the outer perimeter of the mold, using cold water to comprehensively cover the side walls, top, and bottom of the mold, thereby shortening the cooling cycle of the casting. The invention patent with publication number CN103920850B emphasizes continuous vacuum control throughout the pouring and solidification process, aiming to ensure the support rigidity of the sand mold through the stability of the negative pressure, thus reducing the defect rate caused by mold collapse or insufficient feeding. These solutions have indeed played a positive role in addressing specific production pain points, such as shortening macroscopic cooling time or maintaining the geometric integrity of the cavity.

[0004] The aforementioned existing technologies are gradually revealing insurmountable fundamental limitations when dealing with high-performance precision castings with complex structures and significant differences in wall thickness. Although spray cooling systems increase heat transfer intensity macroscopically, they are essentially a non-interventional external heat dissipation mode. Because the path of heat conduction from the core of the casting through the dry sand layer to the mold surface is relatively long, and dry sand, as a poor conductor of heat, has an extremely low thermal diffusivity, this "surface-to-inside" cooling method is extremely difficult to produce timely and effective cooling of specific hot spots inside the casting.

[0005] External spraying, due to its fixed cooling path, cannot achieve "shape-following" heat extraction based on the complex geometry of the casting. This inevitably leads to severe temperature gradient deviations between different parts of the casting. During solidification, this temperature difference not only causes disorder in the direction of solidification front advancement, easily inducing shrinkage cavities and porosity, but also generates enormous thermal stress, which in turn causes dimensional deviations, warping, and even cracking of the casting.

[0006] Process schemes that solely rely on negative pressure parameter optimization often treat the solidification process as a passive, natural heat dissipation process, lacking active thermal field control methods. In the vacuum environment unique to V-process casting, the dry sand layer, lacking a binder for solid bonding, relies primarily on point contact between sand grains for heat conduction. This further exacerbates the thermal resistance effect, leaving complex castings in a quasi-insulating state during solidification. While this state facilitates mold filling and flow, it significantly slows down the solidification rate, resulting in coarse-grained microstructures and mechanical properties that fail to meet the service requirements of high-end equipment.

[0007] In the existing V-process casting system, there is a serious disconnect in the spatiotemporal distribution between the physical support function of negative pressure forming and the metallurgical control function of cooling shape control. If the precise spatial matching between the cooling medium and the irregular contour of the casting, as well as the temporal dynamic coordination between cooling intensity and phase transformation kinetic path, cannot be solved, it will be impossible to truly achieve the precision forming of complex castings. Summary of the Invention

[0008] The purpose of this invention is to provide a composite mold and shape control method for V-process casting based on negative pressure and conformal cooling, which solves the technical problem of the mismatch of multi-field coupling of heat-fluid-force-phase change in the solidification process caused by the passive heat dissipation mode and open-loop control system in traditional V-process casting, which makes it impossible to achieve precise control of the internal quality of castings.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0010] The V-method casting shape control method based on negative pressure and conformal cooling includes the following steps:

[0011] The first step is to prepare a pattern and integrate conformal cooling channels within it: using metal additive manufacturing or a layered manufacturing process with high thermal conductivity alloy materials, conformal cooling channels are constructed within the pattern, equidistantly distributed with respect to the outer surface contour lines of the casting, based on the geometric contour data of the part to be cast. The vertical distance between the centerline of the conformal cooling channel and the forming surface of the pattern is maintained between 8mm and 25mm. The inner diameter of the conformal cooling channel is designed differently according to the local heat capacity distribution of the casting. For hot spots in the casting, the diameter of the conformal cooling channel is increased to 15mm to 20mm, and the spacing between them is reduced to 1.5 times the channel diameter. For thin-walled areas, the diameter of the conformal cooling channel is maintained between 8mm and 12mm, and the spacing between them is increased to 3 times the channel diameter.

[0012] The second step, film heating and vacuum adsorption: an ethylene-vinyl acetate copolymer film with a thickness of 0.1 mm to 0.2 mm is heated to a softened state and covered on the surface of the pattern; the negative pressure suction system is activated to generate an initial negative pressure of -0.04 MPa to -0.06 MPa in the suction holes on the surface of the pattern, so that the film is tightly attached to the surface of the pattern.

[0013] The third step, sand box filling and back pressure molding: A sand box is placed on top of the pattern covered with a film and filled with dry sand without binder; during the filling process, a high-frequency vibration table is used for compaction, with the vibration frequency set to 50Hz to 70Hz and the acceleration maintained at 2g to 3g; then a second film is covered on the sand surface, and continuous negative pressure is applied to the inside of the sand box, with the negative pressure value maintained at -0.05MPa to -0.08MPa, so that the dry sand forms a sand mold;

[0014] The fourth step, dynamic preheating and pouring: Before pouring the molten metal, the preheating mode of the conformal cooling module is activated, and hot water preheated to 60°C to 80°C is injected into the conformal cooling channel to preheat the pattern at a constant temperature; after the molten metal is poured and fills the cavity, the thermal balance monitoring unit monitors the temperature sensor signal in real time.

[0015] The fifth step, conformal cooling and shape control implementation: When the thermal balance monitoring unit detects that the surface temperature of the pattern reaches the set threshold, the input temperature of the cooling medium is reduced to 15°C to 25°C, and the flow rate of each branch is dynamically allocated according to the real-time cooling rate requirements of different areas through the flow ratio regulating valve group; the Reynolds number Re in the conformal cooling channel is controlled between 4000 and 10000 to perform differentiated heat extraction on different parts of the casting; the negative pressure exhaust system continues to work throughout the solidification process;

[0016] The sixth step, coordinated pressure holding and unpacking: When the center temperature of the casting drops to 100°C below the solidus temperature, the operation of the conformal cooling module is stopped, and the negative pressure extraction system continues to maintain a negative pressure state of -0.02MPa to -0.03MPa for no less than 30 minutes for pressure holding and aging treatment; after the casting cools down to below 200°C, the negative pressure is turned off and the casting is unpacked and cleaned.

[0017] Furthermore, the inner wall of the conformal cooling channel is provided with a micron-level turbulence-inducing structure, which is a triangular prism-shaped protrusion with a height of 0.5 mm to 1.0 mm.

[0018] Furthermore, in the region near the gate and riser, the cross-sectional shape of the conformal cooling channel changes from a circle to a flattened ellipse; the inner surface of the conformal cooling channel is coated with a high thermal conductivity silicon carbide ceramic coating with a thickness of 0.05 mm to 0.1 mm, and the thermal conductivity of the coating is not less than 120 W / (m·K).

[0019] Furthermore, the cooling medium is desalinated water containing preservatives and surfactants, with 2% to 5% polydimethylsiloxane microemulsion added by mass; the pumping pressure of the circulating power source is maintained at 0.4 MPa to 0.8 MPa.

[0020] Furthermore, the thermal balance monitoring unit includes a Pt100 platinum resistance temperature sensor array embedded 3mm under the surface of the casting, with a sensor spacing of 50mm to 100mm; the monitoring unit adopts a multivariable predictive control algorithm to output commands to the flow proportional regulating valve group by calculating the unsteady heat conduction process inside the casting in real time.

[0021] Furthermore, the negative pressure extraction system and the conformal cooling module maintain logical coordination on the time axis: during the initial 0 to 30 seconds of casting, the negative pressure value is set to the highest peak value of -0.08MPa, at which time the conformal cooling module is in low-flow circulation; during the middle of solidification, the negative pressure value drops to -0.05MPa, and the conformal cooling module starts maximum power heat exchange; during the later stage of solidification and cooling, the negative pressure value drops to -0.03MPa, cooperating with the conformal cooling module for slow cooling.

[0022] Furthermore, the material of the pattern is selected as ductile iron or aluminum bronze alloy with a thermal conductivity greater than 30 W / (m·K); a ceramic fiber heat insulation layer with a thickness of 5 mm to 8 mm is provided on the non-formed side of the pattern.

[0023] Furthermore, in the fifth step, for the hot spot region of the casting, the cooling intensity is increased by increasing the flow rate, inducing the generation of a directional solidification temperature zone perpendicular to the surface of the casting; for the thin-walled region, the flow rate is restricted to slow down the cooling rate.

[0024] Furthermore, the conformal cooling channel is designed in three dimensions using a topology optimization method, so that the vertical distance between the channel centerline and the casting surface is dynamically adjusted according to the local heat flux density: in areas where the heat flux density is greater than 100W / cm², the distance is reduced to 8mm to 12mm; in areas where the heat flux density is lower, the distance is widened to 18mm to 25mm; the channel cross-sectional shape is teardrop-shaped, with its long axis along the normal of the casting surface; the inner wall of the channel is provided with a composite turbulence-inducing structure composed of diamond-shaped protrusions, which are arranged in an alternating array, with the inclination angle of the frontal surface being 30° to 45° and the inclination angle of the backal surface being 15° to 20°.

[0025] Furthermore, the cooling medium is a deionized water-based nanofluid with a mass fraction of 0.1% to 0.3% graphene nanosheets; the negative pressure extraction system adopts independent zone control, the sand box is divided into multiple independent negative pressure chambers, each corresponding to a characteristic area of ​​the casting, and the negative pressure value of each zone is independently adjusted according to the feedback of the thermal balance monitoring unit during the solidification process; the control algorithm of the thermal balance monitoring unit adopts multivariable model predictive control based on the finite element model, and predicts the future thermal field evolution and adjusts the cooling parameters in advance by solving the quadratic programming problem online.

[0026] Furthermore, the Reynolds number Re within the conformal cooling channel is controlled between 8000 and 12000.

[0027] Furthermore, while the fifth step is being performed, a seventh step is also included: constructing and updating a digital twin model online. The digital twin model includes a transient heat conduction model, a phase change dynamics model, a thermo-mechanical coupling model, and a flow and heat transfer model, and the model state is corrected using real-time sensor data through a data assimilation algorithm.

[0028] Furthermore, the data assimilation algorithm employs ensemble Kalman filtering (EnKF) to update the state using the following formula:

[0029] ;

[0030] ;

[0031] in Forecast status, To analyze the state, For the observation vector, For the observation operator, To predict the error covariance, For observation error covariance, This is the Kalman gain.

[0032] Furthermore, the present invention also includes an eighth step: based on the updated digital twin model, using a model predictive control (MPC) framework to solve the optimization problem online, dynamically adjusting the cooling parameters and negative pressure parameters; the objective function of the optimization problem is:

[0033] ;

[0034] in:

[0035] Mathematical optimization operators represent the "minimum" operation, used to solve for the objective function. Optimal control vector that takes the minimum value It has no physical unit;

[0036] Over time The changing control vector;

[0037] The objective function (cost function) is used to comprehensively evaluate the control effect of the casting solidification process.

[0038] The first weighting coefficient, dimensionless, is used to adjust the weight of the "solid fraction homogeneity" term in the objective function. Its typical value range is... ~ It can be adjusted according to the density requirements of the casting;

[0039] : Mathematical integration operator, representing integration operation over a specified region (in this case, multiple integration), has no physical unit, and is only a mathematical operation symbol;

[0040] The integration domain refers to the entire geometric calculation area (volume domain or area domain) of the casting. The unit for volume integration is cubic meters (m³), and the unit for area integration is square meters (m²).

[0041] Dimensionless constant, fixed value of 1, used to calculate "unsolidified proportion". ”;

[0042] Solid fraction refers to the proportion of the solid volume in a casting element to the total volume of that element. It is dimensionless and ranges from 0 to 1 (0 for completely liquid and 1 for completely solid).

[0043] Integral infinitesimal element refers to the computational domain of the casting. The infinitesimal volume element (or area element) within the space, with the volume element unit being cubic meters (m³) and the area element unit being square meters (m²).

[0044] : Time variable, representing the real-time time of the casting solidification process, in seconds (s).

[0045] Solidification end time refers to the moment when the solid fraction of the casting reaches the set threshold (≥0.99), and the unit is seconds (s).

[0046] The second weighting coefficient, dimensionless, is used to adjust the weight of the "equivalent stress" term in the objective function. Its typical value range is... ~ It can be adjusted according to the requirements for controlling residual stress in castings;

[0047] Equivalent stress (von Mises stress) characterizes the comprehensive stress state of the micro-element of a casting and is a core indicator for assessing the risk of deformation and cracking in castings. The unit is megapascal (MPa).

[0048] The third weighting coefficient is dimensionless and is used to adjust the weight of the "solidification time" term in the objective function. Its typical value range is 1 to 10, and it can be adjusted according to production efficiency requirements.

[0049] : Time constraint symbol, representing "in The value of "(the moment of solidification completion)" has no physical unit and is only used to specify the time point at which the integral result is taken.

[0050] Furthermore, the present invention also includes a ninth step: using a reinforcement learning algorithm to perform online adaptive optimization of the cooling strategy, wherein the agent learns the optimal control strategy by interacting with the environment, the environment being a digital twin model or an actual system, the state space including the temperature of key points of the casting, temperature gradient, solid fraction, and equivalent stress, the action space being the adjustment amount of the cooling parameters, and the reward function being designed based on the tendency of shrinkage cavity, stress level, and cooling rate deviation.

[0051] Furthermore, the reinforcement learning algorithm employs Deep Deterministic Policy Gradient (DDPG), which includes two stages: offline pre-training and online fine-tuning. During online fine-tuning, the policy network and value network are continuously updated using an experience replay mechanism.

[0052] In addition, this invention also discloses a V-process casting composite mold based on negative pressure and conformal cooling, comprising:

[0053] Mold base plate;

[0054] The pattern is set on the mold base plate, and the pattern has an integrated conformal cooling channel inside, the direction of which matches the outer surface contour of the casting.

[0055] A heat-resistant plastic film covering the surface of the pattern;

[0056] A negative pressure extraction system is used to generate negative pressure on the surface of the pattern and inside the sand box.

[0057] The conformal cooling module includes:

[0058] A cooling medium circulation power source is connected to the conformal cooling channel;

[0059] A flow rate proportional regulating valve assembly is installed on a branch of the conformal cooling channel to regulate the flow rate of the cooling medium in each branch.

[0060] A thermal balance monitoring unit is connected to a temperature sensor and the flow proportional regulating valve group, and is used to control the flow proportional regulating valve group according to the temperature sensor signal.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] This invention, by deeply integrating conformal cooling channels into the pattern body and forming spatiotemporal coordination with the negative pressure extraction system, completely changes the passive cooling mode of traditional V-process casting that relies on natural heat dissipation from dry sand, achieving active intervention and precise control of the thermal field during the casting solidification process. Building upon this, it further introduces digital twin models, multi-field coupled collaborative optimization control, and reinforcement learning adaptive strategies to construct a closed-loop intelligent shape control system from perception and prediction to decision-making and execution, achieving significant technical results.

[0063] In this invention, the conformal cooling channel is topologically optimized based on the geometric contour of the casting, enabling the cooling medium to efficiently exchange heat in close contact with the forming surface, significantly shortening the heat conduction path and eliminating the "uneven heating and cooling" phenomenon of traditional external spray cooling. The turbulence-induced structure and high thermal conductivity ceramic coating on the inner wall of the channel significantly improve the heat exchange efficiency, while the dynamic adjustment of the cooling medium flow rate and temperature enables differentiated cooling of hot spot areas and thin-walled areas of the casting, inducing directional solidification, fundamentally inhibiting the formation of shrinkage cavities and porosity, and refining the solidification structure.

[0064] Furthermore, the negative pressure extraction system not only continuously maintains the rigidity of the sand mold and ensures the geometric accuracy of the cavity, but also, through sequential coordination with the conformal cooling module, provides high compactness to resist the static pressure of the molten metal in the early stage of pouring, and releases thermal stress through slow pressure aging in the later stage of solidification, effectively suppressing high-temperature creep deformation of the casting. The negative pressure environment itself also plays an auxiliary role in heat conduction by accelerating the gas flow between sand grains, forming a multi-dimensional heat exchange superposition effect of "macro-level liquid cooling + micro-level gas cooling" with conformal cooling.

[0065] Meanwhile, the construction and online updating of the digital twin model achieves "transparency" of the casting solidification process. Multiphysics models, including transient heat conduction, phase change dynamics, thermo-mechanical coupling, and flow heat transfer, assimilate sensor data in real time using Kalman filtering, ensuring a high degree of consistency between the model state and the physical entity, providing a reliable virtual mirror for precise control. Based on this digital twin model, the model predictive control framework can continuously optimize cooling and negative pressure parameters in each control cycle, ensuring the solidification process always evolves along a preset optimal trajectory, thereby minimizing residual stress, preventing hot cracking, and optimizing microstructure distribution.

[0066] The introduction of reinforcement learning algorithms endows the control system with adaptive learning capabilities. By interacting with the digital twin environment, the agent continuously optimizes the policy network, enabling it to proactively adapt to changes in working conditions such as casting temperature fluctuations, alloy composition deviations, and mold aging, significantly improving process robustness and consistency.

[0067] This invention achieves precise coupling and control of thermal field, flow field, stress field and phase transformation dynamics in the V-process casting process by organically integrating conformal cooling, negative pressure synergy, digital twin, model predictive control and reinforcement learning. It solves long-standing technical problems such as low heat dissipation efficiency, uncontrollable temperature field, coarse microstructure and deformation cracking in traditional processes. Attached Figure Description

[0068] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0069] Figure 1 This is an overall flowchart of the method described in this invention.

[0070] Figure 2 This is a simplified diagram of the overall structure of the composite mold described in this invention. Detailed Implementation

[0071] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0072] The following is in conjunction with the appendix Figure 1 and Figure 2 The embodiments of the present invention will be described in detail below.

[0073] This embodiment discloses a V-process casting shape control method based on negative pressure and conformal cooling, which relies on a composite mold integrated with a conformal cooling system. The composite mold includes a mold base plate, a pattern disposed on the mold base plate, a heat-resistant plastic film covering the surface of the pattern, a negative pressure exhaust system disposed inside the pattern and below the mold base plate, and a key conformal cooling module. The conformal cooling module consists of a conformal cooling channel embedded in the pattern near the forming surface, a cooling medium circulation power source, a flow rate proportional regulating valve group, and a heat balance monitoring unit.

[0074] In practical implementation, the V-method casting shape control method based on negative pressure and conformal cooling includes the following steps:

[0075] The first step is pattern preparation and cooling channel integration. Using metal additive manufacturing or a layered manufacturing process with high thermal conductivity alloy materials, conformal cooling channels are constructed within the pattern, equidistantly distributed along the outer surface contour of the casting, based on the geometric contour data of the part to be cast. The vertical distance between the centerline of the conformal cooling channel and the pattern forming surface is maintained between 8 mm and 25 mm. The inner diameter of the conformal cooling channel is differentiated according to the local heat capacity distribution of the casting. For hot spots in the casting, the diameter of the conformal cooling channel is increased to 15 mm to 20 mm, and the spacing is reduced to 1.5 times the channel diameter; for thin-walled areas, the diameter of the conformal cooling channel is maintained between 8 mm and 12 mm, and the spacing is increased to 3 times the channel diameter. The inner wall of the conformal cooling channel is provided with a micron-level turbulence-inducing structure, which consists of triangular prism-shaped protrusions with a height of 0.5 mm to 1.0 mm, to improve the heat transfer coefficient between the cooling medium and the channel wall.

[0076] The second step involves film heating and vacuum adsorption. An ethylene-vinyl acetate copolymer film with a thickness of 0.1 mm to 0.2 mm is heated to a softened state and applied to the surface of the pattern. The negative pressure extraction system is activated, creating an initial negative pressure of -0.04 MPa to -0.06 MPa at the extraction holes on the pattern surface. This pressure difference causes the film to adhere tightly to the complex shape of the pattern and the raised areas corresponding to the conformal cooling channels.

[0077] The third step is sand box filling and back pressure molding. A sand box is placed on top of the pattern covered with a thin film and filled with dry sand without binder. During filling, a high-frequency vibration table is used for compaction, with the vibration frequency set to 50Hz to 70Hz and the acceleration maintained at 2g to 3g to ensure that the bulk density of the dry sand outside the conformal cooling channel reaches more than 1.6g / cm³. Subsequently, a second thin film is covered on the sand surface, and a continuous negative pressure is applied to the inside of the sand box, maintained at -0.05MPa to -0.08MPa, so that the dry sand forms a highly rigid sand mold under vacuum pressure.

[0078] The fourth step is dynamic preheating and pouring. Before pouring the molten metal, the conformal cooling module first activates the preheating mode. The cooling medium circulation power source injects hot water preheated to 60°C to 80°C into the conformal cooling channel to preheat the pattern at a constant temperature, thereby reducing the risk of initial solidification during the molten metal filling process. After the molten metal enters the mold cavity through the gating system and fills it, the thermal balance monitoring unit monitors the temperature sensor signal near the surface of the pattern in real time.

[0079] The fifth step is conformal cooling and shape control implementation. This is the core stage of this method. When the heat balance monitoring unit detects that the temperature of the pattern surface reaches a set threshold, the cooling medium circulation power source immediately switches to cooling mode, reducing the input temperature of the cooling medium to 15°C to 25°C. The flow rate proportional regulating valve group dynamically allocates the flow rate of each branch according to the real-time cooling rate requirements of different areas. The control logic of the conformal cooling module is based on the heat conduction equation and phase change kinetic model. By controlling the Reynolds number Re in the conformal cooling channel between 4000 and 10000, differentiated heat extraction is achieved for different parts of the casting. For the hot spot area of ​​the casting, the flow rate is increased to improve the cooling intensity and induce a directional solidification temperature zone perpendicular to the casting surface; for thin-walled areas, the flow rate is limited to slow down the cooling rate and prevent the formation of hardened structures or excessive residual stress. Throughout the solidification process, the negative pressure extraction system operates continuously. The negative pressure it generates is not only used to maintain the strength of the sand mold, but also serves as an auxiliary means to enhance thermal conductivity. By accelerating the gas flow in the gaps between sand grains through negative pressure, the apparent thermal conductivity of the sand layer is improved.

[0080] The sixth step involves coordinated pressure holding and unpacking. When the center temperature of the casting drops to 100°C below the solidus temperature, the conformal cooling module is stopped, but the negative pressure extraction system continues to maintain a negative pressure of -0.02MPa to -0.03MPa for at least 30 minutes of pressure holding and aging treatment. This uniform negative pressure environment suppresses the casting's self-weight deformation during the high-temperature creep stage. After the casting cools to below 200°C, the negative pressure is turned off, the vacuum is broken, and the casting is unpacked and cleaned.

[0081] Step 7: Digital Twin Model Construction and Online Updates. Simultaneously with or before Step 5, a digital twin model of the casting solidification process is established. This digital twin model is composed of the following coupled sub-models: transient heat conduction model, phase change kinetics model, thermo-mechanical coupling model, and flow and heat transfer model. Each sub-model is discretized using the finite element method or finite volume method, and online parameter calibration and state updates are performed based on real-time sensor data. The mathematical description of the digital twin model is as follows:

[0082] (1) Transient heat conduction equation (applicable to castings, patterns, and sand mold areas):

[0083] ;

[0084] in Density (unit: kg / m³). Specific heat capacity (unit: J / (kg·K)) Thermal conductivity (unit: W / (m·K)) Temperature (unit: K or °C). Time (unit: seconds) The latent heat release rate (unit: W / m³) is given by the phase transition model.

[0085] (2) Phase transition dynamics model (using the improved Clyne-Kurz model):

[0086] ;

[0087] ;

[0088] in The solid fraction (dimensionless). , These are the liquidus temperature and solidus temperature (unit: K), respectively. For aluminum alloys, The value range is 2 to 3; for steel, The value ranges from 1.5 to 2.5, and the specific value can be determined through differential scanning calorimetry (DSC) experiments. Latent heat of solidification (unit: J / kg).

[0089] (3) Thermo-mechanical coupling model (thermoelastic-plastic constitutive model):

[0090] ;

[0091] in This is the stress tensor (unit: Pa). This is the elastic stiffness tensor (unit: Pa). The total strain is dimensionless. For thermal strain, This represents plastic strain. Thermal strain is calculated using the following formula:

[0092] ;

[0093] in The coefficient of linear expansion (unit: 1 / K). Reference temperature (unit: K). The symbol is Kronecker. Plastic strain is determined by the von Mises yield criterion and related flow laws, which will not be elaborated here.

[0094] (4) Cooling channel flow and heat transfer model (one-dimensional simplification):

[0095] ;

[0096] ;

[0097] in The cross-sectional area of ​​the channel (unit: m²). Flow velocity (unit: m / s) Fluid temperature (unit: K). Wet perimeter (unit: m) Channel wall temperature (unit: K). The convective heat transfer coefficient (unit: W / (m²·K)) is calculated using the Gnielinski formula:

[0098] ;

[0099] in Darcy friction factor Hydraulic diameter (unit: m). Reynolds number ( ), For Prandtl numbers ( ), The value is the Prandtl number at the wall temperature, which is calculated from the heat conduction equation in the digital twin model and then obtained through fluid property interpolation. The thermal conductivity of the fluid (unit: W / (m·K)). The fluid dynamic viscosity is expressed in Pa·s.

[0100] (5) Data Assimilation and Model Update: Using an array of temperature sensors (such as Pt100 platinum resistance thermometers or thin-film thermocouples) embedded under the pattern and inside the cavity, ensemble Kalman filtering (EnKF) is employed to assimilate and update the state variables (temperature field, solid fraction field) of the digital twin model in real time. Define the state vector. Includes all grid nodes at time Temperature value and solid component value, observation vector Include Each sensor at time The measured temperature value. The EnKF update formula is:

[0101] ;

[0102] ;

[0103] in Forecast state (from model) Points to get), The forecast error covariance matrix (estimated by ensemble estimation). For observation operators (which map the state to the observation location). The observation error covariance matrix, This represents the Kalman gain matrix. Through EnKF, the digital twin model can continuously absorb measured data to correct deviations, ensuring its consistency with the physical entity. Updated state. This serves as the initial condition for the model integration at the next time step.

[0104] The eighth step involves multi-field coupled collaborative optimization based on digital twins. During solidification, the thermal balance monitoring unit invokes the digital twin model in real time, using the current assimilated state. Using initial conditions, predict the future. The temperature field, stress field, and solidification microstructure evolution within the predicted time domain are analyzed. Based on the prediction results, an optimization problem is constructed to dynamically adjust the cooling and negative pressure parameters. The optimization objective is to minimize the solidification time while ensuring casting quality (e.g., minimizing shrinkage cavities, residual stress, and grain refinement). Optimization variables include the flow rate of the cooling medium in each branch. ( , (Number of cooling circuits) and inlet temperature of the cooling medium Negative pressure values ​​in each zone ( , (This refers to the number of negative pressure partitions). The optimization problem can be expressed as:

[0105] ;

[0106] Constraints:

[0107] ;

[0108] ;

[0109] in For control vectors, The system dynamics described by the digital twin model (i.e., the set of coupled equations in step seven). For the casting area, The solidification end time (defined as the moment when the solid fraction at all points in the casting reaches 0.99). The von Mises equivalent stress (unit: Pa) The solid fraction, Weighting coefficients (set by the user according to quality requirements, for example...) , , (to balance the various objectives). and These are the state feasible region and the control feasible region, respectively. The upper and lower limits of the control variables are determined by physical constraints, for example:

[0110] ;

[0111] in Typically, 0 or the minimum stable flow rate is used. Determined by pumping capacity and channel pressure; Use a cooling medium with a freezing point above its freezing point. Take a temperature below the boiling point; Take -0.09MPa (ultimate vacuum). Take -0.02MPa.

[0112] The optimization problem described above is solved online in a rolling manner using a Model Predictive Control (MPC) framework. In each control cycle (e.g., every 10 seconds), the current state is used as the basis for the solution. Using these as initial conditions, solving the above optimization problem yields the optimal control sequence. In the prediction time domain The trajectory on, and the first control quantity This is applied to the actual system (i.e., setting the opening degree of the flow proportional control valve assembly and the zone pressure of the negative pressure extraction system). Furthermore, through EnKF data assimilation, the state of the digital twin model is consistent with the physical system, ensuring the accuracy of optimization decisions.

[0113] The ninth step involves optimizing the adaptive cooling strategy based on reinforcement learning. To further improve the response speed and adaptability of the control system, this invention introduces a reinforcement learning (RL) framework for learning the optimal cooling strategy online. The reinforcement learning agent continuously updates the policy network by interacting with the digital twin environment to minimize long-term cumulative costs.

[0114] Specifically, defining the state space Key features output for the digital twin model include: casting key point temperature, temperature gradient, solid fraction, equivalent stress, etc.; defining the action space. The adjustment amount for cooling parameters (flow rate of each branch, inlet temperature, and zone negative pressure); define the reward function. The instantaneous cost is negative, which includes the current tendency for shrinkage, stress level, and deviation from the target cooling rate.

[0115] As a concrete example, the reward function can be designed as follows:

[0116] ;

[0117] in:

[0118] This represents the current volume of the shrinkage cavity inside the casting (unit: m³). This is a reference value for the volume of shrinkage cavity (the maximum allowable shrinkage cavity volume can be taken). Equivalent stress (unit: Pa). The yield strength of the material (unit: Pa); The measured cooling rate at key points (unit: K / s) Target cooling rate (unit: K / s). This serves as a reference value for the cooling rate deviation. These are weighting coefficients, pre-set according to quality requirements. This reward function guides the agent to minimize cavity, stress, and cooling deviation.

[0119] The agent is trained using the Deep Deterministic Policy Gradient (DDPG) algorithm. The training process consists of two phases: offline pre-training and online fine-tuning.

[0120] Offline pre-training stage: Using a large number of samples generated from historical production data or finite element simulation, the policy network and value network are pre-trained to enable the agent to initially grasp the basic cooling laws.

[0121] Online fine-tuning phase: In actual production, the agent interacts with the digital twin environment (or the actual system), updating network parameters based on real-time feedback and continuously optimizing the strategy. During online fine-tuning, an experience playback mechanism is used, replaying samples from each interaction. Stored in the experience pool, and the network is periodically sampled from the experience pool to update it.

[0122] Output of reinforcement learning policy network It is directly used as the control increment and added to the current control quantity, that is:

[0123] ;

[0124] in Motion constraints must be met (which can be achieved through scaling or trimming). Through reinforcement learning, the control system can adaptively respond to changes in operating conditions (such as fluctuations in casting temperature, deviations in alloy composition, etc.), further improving process robustness.

[0125] Furthermore, the conformal cooling channel design employs a non-uniform flow channel structure optimized based on computational fluid dynamics. Near the gate and riser, the cross-sectional shape of the conformal cooling channel evolves from a circular shape to a flattened ellipse to increase the heat transfer area. The inner surface of the conformal cooling channel is coated with a high thermal conductivity silicon carbide ceramic coating with a thickness of 0.05 mm to 0.1 mm and a thermal conductivity of not less than 120 W / (m·K), designed to prevent erosion corrosion of the channel inner wall by the cooling medium and further reduce contact thermal resistance.

[0126] Furthermore, the cooling medium is demineralized water containing preservatives and surfactants. The viscosity of the medium is reduced and its flow stability within the microchannels is improved by adding 2% to 5% polydimethylsiloxane microemulsion by mass. The pumping pressure of the circulating power source is maintained between 0.4 MPa and 0.8 MPa to ensure sufficient flow velocity of the cooling medium in the long-path, multi-branch conformal channels, preventing localized cavitation or boiling.

[0127] Furthermore, the thermal balance monitoring unit includes a Pt100 platinum resistance temperature sensor array embedded 3mm under the skin of the sample, with a sensor spacing of 50mm to 100mm. The monitoring unit employs a multivariate predictive control algorithm, the mathematical model of which is described as follows:

[0128] ;

[0129] in,

[0130] Cooling power is adjusted in real time;

[0131] This is the proportional-integral coefficient, used to adjust the strength of the deviation integral term;

[0132] for The characteristic temperature gradient values ​​of the corresponding key parts of the casting are measured in the temperature gradient field at any time.

[0133] for The preset target temperature gradient value at all times;

[0134] Weighting coefficients are set for parts of different thicknesses;

[0135] For the first Real-time temperature measured by a temperature sensor;

[0136] This represents the rate of temperature change at the sensor location.

[0137] The algorithm calculates the unsteady heat conduction process inside the casting in real time and outputs commands to the flow ratio regulating valve group to achieve a microsecond-level response to the cooling intensity.

[0138] Furthermore, the negative pressure extraction system and the conformal cooling module maintain strict logical coordination on the time axis. During the initial pouring stage (0-30s), the negative pressure value is set to its highest peak value. At this point, the conformal cooling module is in low-flow circulation, using the compaction generated by the high negative pressure to ensure that the cavity does not deform; during the middle of solidification (30 seconds to the end of solidification), the negative pressure value drops to At this point, the conformal cooling module activates its maximum power heat exchange, using active cooling to guide the solidification front; during the later stages of solidification and cooling, the negative pressure drops to... It is used in conjunction with a conformal cooling module for slow cooling to release the thermal stress accumulated due to rapid cooling.

[0139] In a preferred embodiment of the present invention, the pattern is made of ductile iron or aluminum bronze alloy with a thermal conductivity greater than 30 W / (m·K). On the non-forming side of the pattern, i.e., near the back of the conformal cooling channel, a ceramic fiber insulation layer with a thickness of 5 mm to 8 mm is provided to reduce ineffective diffusion of cooling energy into the depths of the sand box and ensure that the cooling vector is precisely directed towards the forming surface of the casting.

[0140] The V-process casting shape control method based on negative pressure and conformal cooling described in this invention completely changes the low and uncontrollable heat dissipation efficiency of sand molds in traditional processes by introducing deep conformal cooling intervention within the negative pressure forming framework of V-process casting. The design of the conformal cooling channel allows the cooling medium to penetrate deep into the mold and directly conform to the geometric contour of the casting for heat exchange, significantly shortening the heat conduction path. Combined with high-precision flow control and temperature feedback algorithms, this method can dynamically adjust the cooling intensity according to the solidification law of each part of the casting, thereby eliminating heat spots and temperature differences caused by differences in wall thickness. At the same time, the presence of the negative pressure system not only ensures the stability of dry sand forming, but also forms a synergistic effect of "macroscopic thermal control + microscopic thermal assistance" with the conformal cooling module by controlling the micro-circulation gas flow between sand particles.

[0141] Furthermore, by introducing a digital twin model and multi-field coupled collaborative optimization, this invention achieves accurate prediction and active control of the solidification process. The digital twin model assimilates sensor data in real time, ensuring that the model state is consistent with the physical entity; based on the model predictive control framework, cooling parameters and negative pressure parameters are optimized online, ensuring that the casting always evolves along the optimal solidification path. The introduction of reinforcement learning strategies further enhances the system's adaptive capability, enabling it to cope with various disturbances in the production process. Experimental data show that complex structure castings produced using this method can maintain stable dimensional accuracy at CT4 to CT6 levels, reduce surface roughness Ra to below 12.5 μm, reduce the scrap rate of internal defects by more than 60% compared to traditional V-process casting, and reduce residual stress by more than 50%.

[0142] The present invention will be further described below with reference to specific embodiments, but the present invention is not limited to the following embodiments.

[0143] Example 1: A complex aluminum alloy hydraulic integrated block with variable wall thickness, ranging from 12mm to 45mm, is produced using the method described in this invention.

[0144] The pattern material is aluminum bronze (thermal conductivity 85 W / (m·K)), and the centerline of the internal conformal channel is 15 mm from the forming surface. The channel diameter is 20 mm and the spacing is 30 mm in the hot spot area; the channel diameter is 10 mm and the spacing is 30 mm in the thin-walled area. The inner wall of the channel is provided with a triangular prism-shaped turbulence-inducing structure with a height of 0.6 mm.

[0145] The preheating temperature is set to 75℃, and the pouring temperature is 720℃. Negative pressure control: -0.08MPa for the first 30 seconds of pouring, -0.05MPa during solidification, and -0.03MPa during the holding period. Conformal cooling medium: demineralized water containing 3% polydimethylsiloxane microemulsion, cooling water temperature 20℃, with the Re number maintained at approximately 8500. Control algorithm weighting coefficient αi: 0.85 for the hot spot region and 0.45 for the thin-walled region.

[0146] Comparative experiment: Comparative example 1 adopted the traditional V-process casting process. The pattern was an ordinary aluminum alloy pattern without conformal cooling channels. The pouring temperature was 720℃. The negative pressure was maintained at -0.05MPa throughout the process, relying on natural heat dissipation from the sand mold. The specific results are shown in Table 1.

[0147] Table 1: Comparison of quality indicators of castings produced by Example 1 and Comparative Example 1V casting method;

[0148]

[0149] Example 2: This example mainly focuses on a large wind turbine hub casting made of QT400-18 material with a maximum wall thickness of 120mm and multiple hot spots.

[0150] The pattern is made of ductile iron (thermal conductivity 35 W / (m·K)), and the centerline of the internal conformal channels is 20 mm from the forming surface. The channel diameter is 25 mm and the spacing is 40 mm in the hot spot area; the channel diameter is 12 mm and the spacing is 36 mm in the thin-walled area. A turbulence-inducing structure is set on the inner wall of the channel.

[0151] Preheating temperature: 80℃; casting temperature: 1380℃. Negative pressure control: -0.07MPa for the first 30 seconds after casting, -0.05MPa during solidification, and -0.02MPa during the holding period. Cooling medium: Demineralized water containing 4% polydimethylsiloxane microemulsion, water temperature: 18℃, Re number: 6000-8000. Weighting coefficient: 0.9 for hot spots and 0.3 for thin-walled areas. Holding time: 3 hours.

[0152] Comparative experiment: Comparative Example 2 used the traditional V-process casting, without conformal cooling channels, and maintained a constant negative pressure of -0.05 MPa. The specific results are shown in Table 2:

[0153] Table 2: Comparison of casting quality of wind turbine hubs;

[0154]

[0155] Example 3: This example mainly focuses on the casting of complex thin-walled aluminum alloy shell parts, specifically a complex thin-walled aluminum alloy shell part with a wall thickness of 5-15mm.

[0156] The material is aluminum bronze. The center line of the internal conformal channel is 12mm away from the forming surface. The channel diameter is 8-10mm and the spacing is 20mm.

[0157] Preheating temperature: 70℃; casting temperature: 700℃. Negative pressure control: -0.06MPa for the first 30 seconds after casting, -0.04MPa during solidification, and -0.02MPa during the holding period. Cooling medium: Demineralized water containing 2.5% polydimethylsiloxane microemulsion, water temperature 22℃, Re number 4500-5500. Weighting coefficient: 0.8 for hot spots, 0.5 for thin-walled areas. Holding time: 40 minutes.

[0158] Comparative Example 3 was cast using the traditional V-process without conformal cooling channels, and the negative pressure was kept constant at -0.04 MPa. The specific results are shown in Table 3.

[0159] Table 3: Comparison of casting quality of thin-walled shell parts;

[0160]

[0161] Example 4: This example focuses on the precision casting of high-temperature alloy turbine blades for aero-engines, specifically a nickel-based high-temperature alloy K418 turbine blade with a wall thickness of 0.8-8mm and multiple internal cooling channels and hot spots.

[0162] The mold and channel topology optimization is as follows:

[0163] The pattern is manufactured using laser selective melting additive manufacturing with a tungsten-based high thermal conductivity alloy (thermal conductivity 120 W / (m·K)). A conformal cooling channel is designed based on finite element topology optimization. The distance between the channel centerline and the forming surface is: 8-12 mm in the high heat flux zone (leading edge) and 18-25 mm in the low heat flux zone (middle of the blade basin). The channel cross-section is teardrop-shaped, with the major axis along the surface normal. The inner diameter of the channel in the hot spot zone is 8-10 mm, with a spacing of 1.2 times the diameter; the inner diameter in the thin-walled zone is 3-5 mm, with a spacing of 3.5 times the diameter. The inner wall of the channel is integrally formed with a diamond-shaped protrusion composite turbulence-inducing structure, with a protrusion height of 0.3-0.6 mm, an inclination angle of 30-45° on the upstream side, and an inclination angle of 15-20° on the downstream side. The inner surface of the channel is coated with an aluminum nitride ceramic coating (thickness 0.08-0.12 mm, thermal conductivity ≥180 W / (m·K)).

[0164] The cooling medium is deionized water nanofluid with 0.2% graphene nanosheets added. Pumping pressure is 0.6-1.0 MPa, flow rate is 2-5 m / s, and Re zone control is implemented: ≥10000 in the hot zone and approximately 6000 in the thin-walled zone. Temperature sensors utilize a 0.5 mm diameter thin-film thermocouple array, embedded 2 mm under the sample surface with a spacing of 15-30 mm, supplemented by infrared thermal imaging. The control algorithm employs multivariable model predictive control (MPC) based on a finite element model, with a sampling period of 50 ms.

[0165] The sand box is divided into multiple independent negative pressure chambers, corresponding to areas such as the blade and tenon. During the initial stage of casting, the negative pressure in each zone is -0.09MPa; during the middle stage of solidification, the negative pressure chamber corresponding to the hot spot area is maintained at -0.07MPa, and the pressure in the thin-walled area is reduced to -0.05MPa; during the later stage of solidification, the pressure is uniformly reduced to -0.02 to -0.03MPa and held for 60 minutes.

[0166] Comparative experiment: Comparative example 4 was made using the traditional V-process precision casting without internal conformal cooling channels, with a constant negative pressure of -0.06MPa and natural heat dissipation. The specific results are shown in Table 4.

[0167] Table 4: Comparison of Turbine Blade Casting Quality;

[0168]

[0169] Example 5: This example, based on Example 4, further integrates digital twin model and multi-field coupling collaborative optimization.

[0170] Digital twin model: A digital twin model was constructed incorporating transient heat conduction, phase change kinetics, thermo-mechanical coupling, and flow heat transfer. Mesh generation: 0.5 mm for the casting region, 1 mm for the pattern region, and 5 mm for the sand mold region, for a total mesh size of approximately 1.5 million. Material parameters were the same as in Example 4. Ensemble Kalman filtering (EnKF) was used to assimilate measured data from 40 thin-film thermocouples (15 mm spacing, 10 Hz sampling), with an ensemble size of 80 and an observation error standard deviation of 2°C. Dimensionality reduction using POD projected the state onto 200 bases to reduce computational load.

[0171] MPC optimizations are as follows:

[0172] Optimization variables: flow rates of 8 cooling loops Q1-Q8, inlet temperature of cooling medium Tin, and pressures of 3 negative pressure zones P1-P3. Control cycle 10 seconds, prediction time domain 60 seconds. Objective function weights w1=106 (cavity penalty), w2=10⁻⁵ (stress penalty), w3=1 (time penalty). Constraints: 0.1≤Qi≤5 L / min, 15≤Tin≤25℃, -0.09≤Pj≤-0.02 MPa. The SQP algorithm is used for solving, with each cycle taking approximately 2 seconds.

[0173] The offline pre-training process is as follows: The DDPG network is trained using 200 sets of historical data. Online fine-tuning: Using a digital twin model as the environment, interactions occur every 10 seconds. State parameters include the highest / lowest casting temperature, maximum temperature gradient, maximum solid fraction, maximum equivalent stress, and current time. Actions include adjustments to cooling parameters. The reward function is designed based on shrinkage tendency, stress level, and cooling rate deviation, and the strategy network is updated in real time.

[0174] Comparative Experiment: Comparative Example 5 adopts the scheme of Example 4 (with conformal cooling and basic MPC, but without digital twin assimilation and multi-field coupling optimization). Comparative Example 4 is the same as the traditional V method in Example 4, and the specific structure is shown in Table 5.

[0175] Table 5: Comparison of turbine blade casting quality (Example 5, Comparative Example 5 and Comparative Example 4);

[0176]

[0177] Example 6: This example discloses a V-process casting composite mold based on negative pressure and conformal cooling, specifically including:

[0178] Mold base plate and pattern: The mold base plate is made of high-strength ductile iron with a thickness of 50mm. The lower surface is provided with an array of evacuation holes connected to the negative pressure evacuation system. The pattern is fixed above the mold base plate and is made of aluminum bronze alloy with a thermal conductivity of 85W / (m·K). The outer surface contour matches the inner cavity shape of the part to be cast (taking the aluminum alloy hydraulic integrated block in Example 1 as an example). The wall thickness of the pattern is 25mm in the hot spot area and 15mm in the thin wall area.

[0179] Conformal cooling channels: Conformal cooling channels are integrally formed inside the pattern using laser selective melting additive manufacturing. The three-dimensional orientation of the channels is based on topological optimization of the casting geometry, with the channel centerline 12-20mm away from the pattern forming surface. Inner diameter variations: 18mm in the hot spot area with a spacing of 27mm; 10mm in the thin-walled area with a spacing of 30mm. The inner wall features an integrally formed 0.6mm high triangular prism-shaped turbulence-inducing structure arranged in a spiral pattern. In the gating and riser areas, the channel cross-section gradually changes from a circle to a flattened ellipse (major axis 20mm, minor axis 12mm).

[0180] Coating and insulation: The inner surface of the channel is coated with a 0.08 mm thick silicon carbide ceramic coating by CVD, with a thermal conductivity of 150 W / (m·K). A 6 mm thick ceramic fiber insulation layer is attached to the back of the pattern, covering the entire back side.

[0181] Thin film and negative pressure pumping system: The surface of the pattern is covered with a 0.15mm thick EVA thin film. The negative pressure pumping system includes a vacuum pump, vacuum tank, pipelines and control valve group, and adopts independent zone control: the sand box is divided into three independent negative pressure chambers (hot spot zone, thin wall zone, and riser zone), and each chamber is equipped with an independent pressure sensor and regulating valve.

[0182] The conformal cooling module uses a variable frequency centrifugal pump as the power source for cooling medium circulation, with a rated flow rate of 50 L / min and an adjustable outlet pressure of 0.2-1.0 MPa. The flow proportional control valve assembly contains eight independent proportional control valves, with an adjustment range of 0.5-10 L / min and an accuracy of ±2%. The thermal balance monitoring unit includes an industrial computer, data acquisition card, and control software, connected to twelve Pt100 platinum resistance temperature sensors (60 mm spacing) embedded 3 mm under the template skin. It incorporates a multivariable model predictive control algorithm and outputs commands to the flow proportional control valve assembly and the negative pressure extraction system.

[0183] The specific steps are as follows:

[0184] Install the pattern and cover it with a thin film. After heating and softening, activate the negative pressure vacuum system to ensure the film adheres. Place the sand box, fill it with dry sand and compact it. After covering with a second layer of film, apply a negative pressure of -0.06MPa to the inside of the sand box to form the sand mold. Before pouring, activate the preheating mode and inject 75℃ hot water into the conformal cooling channel to preheat the pattern. After pouring, when the surface temperature of the pattern reaches 400℃, switch to the cooling mode. The temperature of the cooling medium (deionized water containing 3% polydimethylsiloxane microemulsion) is 20℃. The flow rate is dynamically allocated according to the cooling requirements of each area, controlling Re between 5000-9000. The negative pressure and cooling sequence are coordinated: -0.08MPa and low flow rate cooling during the first 30 seconds of pouring; -0.05MPa and full power cooling during the middle stage of solidification; -0.03MPa and slow cooling during the later stage of solidification. After solidification, stop cooling and maintain a negative pressure of -0.02MPa for 40 minutes, then open the box for cleaning.

[0185] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0186] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A V-process casting shape control method based on negative pressure and conformal cooling, characterized in that, Includes the following steps: The first step is to prepare a pattern and integrate conformal cooling channels inside the pattern; The second step, film heating and vacuum adsorption: After the heat-resistant plastic film is heated and softened, it is covered on the surface of the pattern, and the negative pressure pumping system is started to make the film adhere to the surface of the pattern. The third step, sand box filling and back pressure molding: a sand box is placed on top of the pattern covered with a film and filled with dry sand. Negative pressure is applied inside the sand box to form a sand mold from the dry sand. The fourth step, dynamic preheating and pouring: Before pouring the molten metal, the preheating mode of the conformal cooling module is activated to preheat the pattern; after the molten metal is poured and fills the cavity, the thermal balance monitoring unit monitors the temperature sensor signal in real time. The fifth step, conformal cooling and shape control implementation: When the surface temperature of the pattern reaches the set threshold, the conformal cooling module is switched to cooling mode, and the flow rate of the cooling medium in each branch is dynamically allocated according to the real-time cooling rate requirements of different areas through the flow ratio regulating valve group, so as to perform differentiated heat extraction on different parts of the casting; throughout the solidification process, the negative pressure exhaust system continues to work. The sixth step, coordinated pressure holding and unpacking: When the casting temperature drops to the set value, the operation of the conformal cooling module is stopped, and the negative pressure extraction system continues to maintain the negative pressure state for pressure holding and aging treatment. After the casting cools to a safe temperature, the negative pressure is turned off and the casting is unpacked and cleaned. The conformal cooling channel is constructed inside the pattern using additive manufacturing process based on the geometric contour data of the casting. The vertical distance between the centerline of the conformal cooling channel and the forming surface of the pattern is maintained within a preset range. The inner diameter of the conformal cooling channel is designed differently according to the local heat capacity distribution of the casting. A larger diameter and smaller spacing are used for hot spots in the casting, and a smaller diameter and larger spacing are used for thin-walled parts. The negative pressure extraction system and the conformal cooling module maintain logical coordination on the time axis: a high negative pressure value is used in the early stage of casting and the conformal cooling module is in low flow circulation to ensure the rigidity of the cavity; in the middle stage of solidification, the negative pressure value is reduced and the conformal cooling module starts maximum power heat exchange to guide the solidification front; in the later stage of solidification, the negative pressure value is further reduced and the conformal cooling module is used for slow cooling to release thermal stress. The conformal cooling channel is designed in three dimensions using a topology optimization method, so that the vertical distance between the channel centerline and the casting surface is dynamically adjusted according to the local heat flux density, and the channel cross-section is teardrop-shaped with its major axis along the normal of the casting surface.

2. The V-process casting shape control method based on negative pressure and conformal cooling according to claim 1, characterized in that, The inner wall of the conformal cooling channel is provided with a micron-level turbulence-inducing structure to improve the heat transfer coefficient between the cooling medium and the channel wall.

3. The V-process casting shape control method based on negative pressure and conformal cooling according to claim 1, characterized in that, The thermal balance monitoring unit includes a temperature sensor array embedded under the pattern skin; the thermal balance monitoring unit adopts a multivariable model predictive control algorithm to output commands to the flow proportional regulating valve group by solving the unsteady heat conduction process inside the casting in real time.

4. The V-process casting shape control method based on negative pressure and conformal cooling according to claim 1, characterized in that, A heat insulation layer is provided on the non-formed side of the pattern to reduce the ineffective diffusion of cooling energy into the depth of the sand box.

5. The V-process casting shape control method based on negative pressure and conformal cooling according to claim 1, characterized in that, In the fifth step, for the hot spot area of ​​the casting, the directional solidification temperature zone is induced by increasing the cooling intensity, and for the thin-walled area, the hardened structure or excessive residual stress is prevented by limiting the cooling intensity.

6. The V-process casting shape control method based on negative pressure and conformal cooling according to claim 1, characterized in that, The cooling medium is a fluid with added functional additives to improve flow stability and heat exchange efficiency; the negative pressure extraction system adopts independent zone control, the sand box is divided into multiple independent negative pressure chambers, each corresponding to a characteristic area of ​​the casting, and the negative pressure value of each zone is independently adjusted according to the feedback of the heat balance monitoring unit during the solidification process.

7. A V-process casting composite mold based on negative pressure and conformal cooling, used to implement the V-process casting shape control method based on negative pressure and conformal cooling as described in any one of claims 1-6. Its features are, include: Mold base plate; The pattern is set on the mold base plate, and the pattern has an integrated conformal cooling channel inside, the direction of which matches the outer surface contour of the casting. A heat-resistant plastic film covering the surface of the pattern; A negative pressure extraction system is used to generate negative pressure on the surface of the pattern and inside the sand box. The conformal cooling module includes: A cooling medium circulation power source is connected to the conformal cooling channel; A flow rate proportional regulating valve assembly is installed on a branch of the conformal cooling channel to regulate the flow rate of the cooling medium in each branch. A thermal balance monitoring unit is connected to a temperature sensor and the flow proportional regulating valve group, and is used to control the flow proportional regulating valve group according to the temperature sensor signal.

Images