Numerical simulation based optimization method for filling parameters of wind power box body casting

Abstract

This invention relates to the fields of metal casting and computer-aided engineering technology, specifically to a method for optimizing filling parameters in wind turbine housing casting based on numerical simulation. The method includes: acquiring initial filling parameters, including spatial topology parameters, initial kinetic parameters, and initial thermodynamic parameters; generating a mesh model composed of multiple spatial nodes; acquiring flow field evolution data and thermal field evolution data; calculating the local kinetic energy dissipation rate of each spatial node in the mesh model and the topological fracture index of the free surface at the gas-liquid interface based on the flow field evolution data; determining whether preset optimization conditions are met based on the flow field evolution data and thermal field evolution data; if met, determining the initial filling parameters as target filling parameters and outputting them; if not met, correcting the initial filling parameters and returning to the step of numerically simulating the filling process of the target fluid within the target cavity. This invention effectively reduces the risk of sand erosion caused by local kinetic energy overload.

Description

Technical Field

[0001] This invention relates to the fields of metal casting and computer-aided engineering technology, specifically to a method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation. Background Technology

[0002] In the current casting process environment for large variable wall thickness castings such as wind turbine housings, the cavity usually contains a complex structure with thick hot spots and thin-walled reinforcing ribs interspersed, and the huge amount of molten metal has a great initial kinetic energy under the action of gravity. In order to control the filling process, existing solutions generally rely on traditional static parameter settings or only rely on the global average flow rate to determine the filling state. When faced with insufficient kinetic energy dissipation capacity, the temperature or initial pouring flow rate is often adjusted first.

[0003] Although such solutions have basic parameter setting capabilities, they easily mask extreme local jet phenomena caused by complex cavity topologies, and a single wavefront velocity index cannot truly reflect the risk of gas entrapment caused by chaos at the gas-liquid interface. This leads to a strong coupling contradiction between local kinetic energy overload of the melt and end-heat decay. At the same time, the need to control the flow rate in the early stage to prevent scouring and the need to maintain the flow rate in the later stage to prevent cold shut-off are inherently conflicting on the time axis. Traditional adjustment methods are very likely to disrupt the overall process cycle, thereby causing secondary defects such as sand flushing, gas entrapment, cold shut-off, or shrinkage porosity.

[0004] Therefore, how to break through the limitations of static parameter settings in complex multi-branch flow channels, and achieve adaptive dissipation of fluid kinetic energy, effective suppression of gas-liquid interface disturbances, and balanced heat-fluid co-evolution has become an urgent technical problem to be solved. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation, thereby solving the following technical problems:

[0006] Breaking the strong coupling contradiction between local kinetic energy overload and end-heat decay caused by traditional static parameter settings, it achieves adaptive dissipation of fluid kinetic energy and effective suppression of gas-liquid interface disturbance in complex multi-branch flow channels, and finally achieves thermal-fluid synergistic evolution balance, thereby significantly reducing the risk of actual casting defects such as sand flushing, air entrapment, cold shut, or shrinkage porosity.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] A method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation, the method includes:

[0009] Acquire the spatial topology data of the target cavity and the preset initial filling parameters, wherein the initial filling parameters include spatial topology parameters, initial kinetic parameters and initial thermodynamic parameters;

[0010] Based on the spatial topology data, the target cavity is spatially discretized to generate a mesh model composed of multiple spatial nodes;

[0011] Based on the mesh model and the initial filling parameters, the filling process of the target fluid in the target cavity is numerically simulated to obtain flow field evolution data and thermal field evolution data.

[0012] Based on the flow field evolution data, the gas-liquid two-phase interface and the filling wavefront of the target fluid during the filling process are tracked, and the local kinetic energy dissipation rate of each spatial node in the grid model and the free surface topological rupture index of the gas-liquid two-phase interface are calculated.

[0013] Based on the flow field evolution data and the thermal field evolution data, calculate the spatiotemporal thermal flow field co-variance of the filling wavefront;

[0014] Determine whether the preset optimization conditions are met. The preset optimization conditions are: the local kinetic energy dissipation rate is greater than or equal to the preset kinetic energy dissipation threshold, the free surface topological rupture index is lower than the preset rupture index threshold, and the spatiotemporal thermal flow field cooperative variance is lower than the preset cooperative variance threshold.

[0015] If satisfied, the initial filling parameters are determined as the target filling parameters and output; if not satisfied, the initial filling parameters are corrected and the process returns to the step of numerically simulating the filling process of the target fluid in the target cavity.

[0016] Optionally, calculating the local kinetic energy dissipation rate of each spatial node in the grid model includes: extracting the peak velocity data of each spatial node from the flow field evolution data; calculating the velocity reduction of each spatial node on the target fluid based on the peak velocity data and a preset fluid attenuation model; and determining the local kinetic energy dissipation rate based on the velocity reduction.

[0017] Optionally, calculating the free surface topological fracturing index of the gas-liquid two-phase interface includes: extracting morphological feature data of the gas-liquid two-phase interface from the flow field evolution data; calculating the degree of fracturing and the degree of rolling of the gas-liquid two-phase interface based on the morphological feature data; and performing weighted summation of the degree of fracturing and the degree of rolling to generate the free surface topological fracturing index.

[0018] Optionally, calculating the spatiotemporal thermal flux field co-variance of the filling wavefront includes: extracting the temperature retention rate of the filling wavefront from the thermal field evolution data; extracting the local velocity of the filling wavefront from the flow field evolution data; calculating the ratio of the temperature retention rate to the local velocity as a matching degree, and solving the spatiotemporal thermal flux field co-variance based on the matching degree.

[0019] Optionally, modifying the initial filling parameters includes: reducing the feature size value in the spatial topology parameters when the local kinetic energy dissipation rate is less than the kinetic energy dissipation threshold; and keeping the spatial topology parameters unchanged when the local kinetic energy dissipation rate is greater than or equal to the kinetic energy dissipation threshold.

[0020] Optionally, modifying the initial filling parameters further includes: reducing the initial kinetic parameters if the free surface topological fracture index is not lower than the fracture index threshold; and keeping the initial kinetic parameters unchanged if the free surface topological fracture index is lower than the fracture index threshold.

[0021] Optionally, modifying the initial filling parameters further includes: increasing the initial thermodynamic parameters when the spatiotemporal heat flow field cooperative variance is not lower than the cooperative variance threshold; and keeping the initial thermodynamic parameters unchanged when the spatiotemporal heat flow field cooperative variance is lower than the cooperative variance threshold.

[0022] Optionally, the target fluid is molten metal, and the target cavity is a mold cavity for casting wind turbine housings; wherein, the spatial topology parameters include the cross-sectional area ratio of the ingate, and the initial thermodynamic parameters include the pouring temperature profile.

[0023] Optionally, the spatial topology data includes the dimensions of the hot spot and the dimensions of the thin-walled reinforcing ribs in the mold cavity used for casting the wind turbine housing.

[0024] Optionally, numerical simulation of the filling process of the target fluid within the target cavity is performed, including: using a preset finite volume method model, and based on the mesh model and the initial filling parameters, discretizing and iteratively calculating the filling process to generate the flow field evolution data and the thermal field evolution data.

[0025] The beneficial effects of this invention are:

[0026] 1. This invention constructs a closed-loop optimization mechanism to conduct multi-dimensional evaluation and parameter correction by comprehensively considering local kinetic energy dissipation rate, free liquid surface topological rupture index, and spatiotemporal thermal flow field cooperative variance; it breaks through the limitations of traditional static parameter settings, effectively solves the strong coupling contradiction between local kinetic energy overload and terminal heat decay in complex flow channels, and realizes adaptive kinetic energy dissipation and heat flow cooperative balance during the filling process.

[0027] 2. This invention extracts the peak flow velocity of spatial nodes and combines it with a fluid attenuation model to calculate the local kinetic energy dissipation rate, thereby achieving a precise quantitative assessment of the microscopic local jet phenomenon in complex cavities. It overcomes the drawback of relying solely on the global average flow velocity to mask extreme local scouring defects, ensuring that high-inertia melts complete the smoothing of kinetic energy and the elimination of extreme values, effectively reducing the risk of sand scouring caused by local kinetic energy overload.

[0028] 3. This invention extracts the morphological characteristics of the gas-liquid two-phase interface and generates the free liquid surface topological fracture index by calculating the degree of fragmentation and the degree of entrapment. This mechanism objectively evaluates the chaotic and oscillating state of the interface in unsteady-state simulation, making up for the insufficiency of a single wavefront velocity index in accurately reflecting the degree of air entrapment, and significantly improving the early warning and suppression capabilities for interface disturbances and air entrapment defects under complex structures.

[0029] 4. This invention calculates the spatiotemporal thermal flow field coordinating variance by calculating the ratio of the filling wavefront temperature retention rate to the local flow velocity, and establishes an evaluation system that comprehensively measures the matching degree of wavefront temperature and flow velocity. This index accurately determines the ability of the flow field to provide enthalpy to prevent premature solidification of thin-walled regions under the premise of no turbulence, and resolves the intrinsic conflict on the time axis between controlling the flow velocity in the early stage to prevent scouring and maintaining the flow velocity in the later stage to prevent cold insulation.

[0030] 5. This invention establishes a hierarchical priority intervention parameter correction mechanism, which sequentially optimizes and controls the spatial topological feature size, initial kinetic parameters, and initial thermodynamic parameters; it avoids the problem of blindly prioritizing temperature or flow rate adjustment and disrupting the process cycle, and uses enthalpy dynamic compensation for the temperature drop caused by kinetic energy loss to achieve thermal flow synergistic evolution balance without inducing secondary defects. Attached Figure Description

[0031] The invention will now be further described with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart illustrating the method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation, as provided in the embodiments of this application. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Please see Figure 1A method for optimizing the filling parameters of wind turbine box casting based on numerical simulation. The method includes: acquiring the spatial topology data of the target cavity and the preset initial filling parameters, which include spatial topology parameters, initial kinetic parameters and initial thermodynamic parameters.

[0035] The target cavity is spatially discretized based on spatial topology data to generate a mesh model composed of multiple spatial nodes. Based on the mesh model and initial filling parameters, the filling process of the target fluid in the target cavity is numerically simulated to obtain flow field evolution data and thermal field evolution data.

[0036] Based on the flow field evolution data, the gas-liquid two-phase interface and the filling wavefront of the target fluid during the filling process are tracked, and the local kinetic energy dissipation rate of each spatial node in the grid model and the free surface topological rupture index of the gas-liquid two-phase interface are calculated.

[0037] Based on the flow field evolution data and thermal field evolution data, the spatiotemporal thermal flow field co-variance variance of the filling wavefront is calculated; the local kinetic energy dissipation rate is compared with the preset kinetic energy dissipation threshold, the free surface topological rupture index is compared with the preset rupture index threshold, and the spatiotemporal thermal flow field co-variance variance is compared with the preset co-variance threshold.

[0038] When the local kinetic energy dissipation rate is greater than or equal to the kinetic energy dissipation threshold, the free surface topological rupture index is lower than the rupture index threshold, and the spatiotemporal thermal flow field cooperative variance is lower than the cooperative variance threshold, the initial filling parameters are determined as the target filling parameters and output.

[0039] If the local kinetic energy dissipation rate is less than the kinetic energy dissipation threshold, the free surface topological rupture index is not less than the rupture index threshold, or the spatiotemporal thermal flow field cooperative variance is not less than the cooperative variance threshold, then the initial filling parameters are corrected and the process of numerically simulating the filling process of the target fluid in the target cavity is returned.

[0040] This embodiment provides a global closed-loop optimization mechanism for the method of optimizing the filling parameters of wind turbine housing casting based on numerical simulation; specifically, the system acquires the spatial topology data of the target cavity and the initial filling parameters including spatial topology, dynamics and thermodynamics dimensions;

[0041] The system generates a grid model by spatial discretization based on spatial topology data, and uses the initial filling parameters to conduct numerical simulation of the filling process of the target fluid in the target cavity. During the simulation evolution, the system continuously acquires flow field evolution data and thermal field evolution data, tracks the gas-liquid two-phase interface and the filling wavefront, and calculates the local kinetic energy dissipation rate, the free liquid surface topological rupture index, and the spatiotemporal thermal flow field co-variance.

[0042] The system compares the above three calculation results with their corresponding preset thresholds. Before the comparison, the system needs to determine each preset threshold. The corresponding preset thresholds, including the kinetic energy dissipation threshold, the fracture index threshold, and the cooperative variance threshold, are not fixed constants, but are dynamically determined by the system based on the specific parameters of the target cavity.

[0043] Specifically, the system acquires the material properties of the casting, including density and liquidus temperature, overall wall thickness gradient, and expected casting quality grade parameters. These parameters are used as input vectors and input into a pre-established multinomial regression model or expert experience data table to adaptively calculate the applicable kinetic energy dissipation threshold, fracture index threshold, and cooperative variance threshold under the current working conditions.

[0044] For example, the expert experience data table contains a basic threshold mapping relationship for different casting quality grades and wall thickness gradients: when the expected casting quality grade is high and the overall wall thickness gradient is greater than 2.0, the preset kinetic energy dissipation threshold is 30%, the fracture index threshold is 0.4, and the cooperative variance threshold is 0.15; when the expected casting quality grade is ordinary and the overall wall thickness gradient is less than or equal to 2.0, the preset kinetic energy dissipation threshold is 25%, the fracture index threshold is 0.6, and the cooperative variance threshold is 0.25.

[0045] If the local kinetic energy dissipation rate is not lower than the preset kinetic energy dissipation threshold, and the free surface topological rupture index and the spatiotemporal thermal flow field co-variance are both lower than the corresponding rupture index threshold and co-variance threshold, then the current flow field steady state and thermal field uniformity are determined to meet the standard, and the initial filling parameters are directly determined as the target filling parameters and output.

[0046] In terms of anomaly and boundary handling, if any of the following four situations occur: low local kinetic energy dissipation rate, excessive fracture index, excessive cooperative variance, or abnormal local solidification shrinkage rate in a closed area, it is determined that the current filling and solidification process has an extremely high risk of sand flushing, air entrapment, cold shut, or shrinkage porosity. The system will automatically refuse to output and correct the initial filling parameters, and force a return to re-execute the numerical simulation steps until all parameters meet the standards.

[0047] For example, taking the metal melt casting process of a large variable wall thickness wind turbine box as an example, its interior has a complex structure of thick hot sections and thin-walled reinforcing ribs; the initial filling parameters are set to a specific inlet runner cross-sectional area ratio and casting temperature curve; the system divides this complex cavity into millions of microscopic spatial nodes; in the first round of simulation calculation, if the cooperative variance of the wavefront spatiotemporal heat flow field flowing through a certain thin-walled reinforcing rib region is detected to increase sharply, the system will determine that there is a serious risk of end heat attenuation and cold shut, reject the current casting scheme and enter the parameter correction loop;

[0048] The purpose of this mechanism is to break the strong coupling contradiction between local kinetic energy overload and end heat decay caused by traditional static parameter settings, and to achieve adaptive dissipation of fluid kinetic energy and effective suppression of gas-liquid interface disturbance in complex multi-branch flow channels.

[0049] In a preferred embodiment of the present invention, calculating the local kinetic energy dissipation rate of each spatial node in the grid model includes: extracting the peak velocity data of each spatial node from the flow field evolution data;

[0050] Based on the peak velocity data and the preset fluid attenuation model, the velocity reduction of the target fluid at each spatial node is calculated; based on the fluid dynamics kinetic energy theorem, the velocity reduction and its corresponding node mass are substituted into the preset kinetic energy-velocity mapping equation to solve and determine the local kinetic energy dissipation rate; the preset fluid attenuation model expression is as follows:

[0051]

[0052] in, For flow rate reduction, For peak flow rate data, The attenuation coefficient is related to the geometry of the flow obstruction structure at the spatial node, for a right-angle turn flow channel. The value ranges from 0.6 to 0.8 for circular arc transition channels. The value range is from 0.2 to 0.4; The spatial damping attenuation constant is calculated from the target fluid dynamic viscosity, and its unit is . To ensure that the unit is Feature length The product is dimensionless. Let be the characteristic length of the flow-blocking structure. Further, the kinetic energy-velocity mapping equation is specifically:

[0053]

[0054] in, To obtain the local kinetic energy dissipation rate, This represents the fluid mass corresponding to that spatial node;

[0055] If the filling state is determined solely based on the global average flow velocity, it often masks the extreme local jet phenomenon caused by the complex cavity topology, resulting in the scouring defects being ignored. This embodiment provides a microscopic evaluation step for quantifying the local kinetic energy dissipation rate of each spatial node in the mesh model.

[0056] Specifically, the system delves into the flow field evolution data and extracts the peak velocity data for each spatial node; it inputs this peak velocity data into a preset fluid attenuation model to calculate the velocity reduction that the corresponding spatial node can produce on the target fluid; finally, the system accurately determines the local kinetic energy dissipation rate of the node based on the mapping ratio between the input peak value and the velocity reduction.

[0057] When dealing with boundary conditions, if the peak velocity data extracted by the system is already lower than the critical scour threshold of the molding sand, it is determined that there is no kinetic energy overload threat at the node, the velocity reduction output by the fluid attenuation model is recorded as zero, and the corresponding local kinetic energy dissipation rate is directly assigned as the safe bottom value; if the peak velocity undergoes a step-like sudden overload, the high-order nonlinear attenuation model is activated to perform damping calculation.

[0058] For example, at the intersection of the straight and horizontal gating channels at the bottom of the wind turbine casing, designated as node A, the system extracts a peak flow velocity of 3.0 m / s for the target fluid. Substituting this into the fluid attenuation model, the system calculates that the physical form of this flow-blocking structure can reduce the flow velocity by 1.2 m / s. Therefore, based on the kinetic energy-velocity mapping equation, the system calculates the local kinetic energy dissipation rate of node A as follows:

[0059]

[0060] By accurately calculating kinetic energy loss rather than simply the decrease in flow rate, the absorption effect of the gating system on the destructive kinetic energy of the melt can be more accurately reflected.

[0061] The purpose of this step is to precisely quantify the ability of the gating system to reduce the peak flow velocity of the melt at a specific spatial node by using the indicator of local kinetic energy dissipation rate, so as to ensure that the kinetic energy is smoothed and the extreme value is eliminated before the high-inertia melt enters the cavity.

[0062] In a preferred embodiment of the present invention, the calculation of the free surface topological fracture index of the gas-liquid two-phase interface includes: extracting morphological feature data of the gas-liquid two-phase interface from flow field evolution data; calculating the degree of fragmentation and the degree of rollover of the gas-liquid two-phase interface based on the morphological feature data; and performing weighted summation of the degree of fragmentation and the degree of rollover to generate the free surface topological fracture index.

[0063] The massive melt has extremely high initial kinetic energy under the action of gravity, resulting in an extremely chaotic morphology at the gas-liquid interface. The simple wavefront velocity index cannot truly reflect the degree of gas entrainment risk. This embodiment provides a quantitative mechanism for calculating the topological fracture index of the free liquid surface. Specifically, the system performs dimensionality reduction analysis on the flow field evolution data and extracts the morphological feature data of the gas-liquid interface.

[0064] The system decouples the morphological feature data and calculates the degree of fragmentation representing the interface discontinuity and the degree of roll-up representing the interface's tendency to vortex, respectively.

[0065] The calculation logic for the degree of fragmentation is as follows: the system counts the total volume of isolated droplets in the local grid domain of the target fluid within a preset time step, calculates the ratio of this volume to the total volume of the fluid in that region, and then linearly normalizes it to obtain a degree of fragmentation value between 0 and 1.

[0066] The calculation logic for the degree of rollover is as follows: extract the velocity divergence and vorticity amplitude of each spatial node on the gas-liquid two-phase interface. When the vorticity amplitude is greater than the preset vorticity threshold, calculate the average spatial curvature of the interface as the rollover feature quantity, and use the Sigmoid function to map the feature quantity to the interval between 0 and 1, thereby obtaining the degree of rollover.

[0067] The fragmentation degree value and the rollover degree value are weighted and summed according to preset weights to generate a comprehensive free surface topological fracture index.

[0068] In extreme morphology determination, if the extracted morphological feature data indicates that the liquid surface is in a completely discrete splash state, that is, the number of topological islands exceeds the limit, the system will directly assign the degree of breakage value to the preset highest extreme value, skip the subsequent linear calculation of weights and determine that the breakage warning condition is met; if the interface remains macroscopically continuous, the calculation will be performed according to the conventional weighted summation logic.

[0069] For example, in a stepped flow channel with a significant elevation difference within the wind turbine housing cavity, the target fluid surface is disturbed. Based on grid morphology data, the system quantifies the current fragmentation degree as 0.3 and the turbulence degree as 0.7. Assuming the system imposes a greater penalty on turbulence behavior that induces deep entrainment, assigning a weighting coefficient of 0.8 to turbulence and 0.2 to fragmentation, the final calculated free surface topology fracture index is: ;

[0070] The purpose of this mechanism is to objectively evaluate the degree of breakage and turbulence at the gas-liquid interface in unsteady numerical simulations. The lower the index, the better the stability of the interface and the lower the probability of gas turbulence defects.

[0071] In a preferred embodiment of the present invention, the calculation of the spatiotemporal thermal flow field co-variance of the filling wavefront includes: extracting the temperature retention rate of the filling wavefront from the thermal field evolution data; extracting the local velocity of the filling wavefront from the flow field evolution data; calculating the ratio of the temperature retention rate to the local velocity as the matching degree, and statistically analyzing the degree of discrete deviation of the matching degree from the continuous spatial grid domain at a selected time step, and finally calculating the spatiotemporal thermal flow field co-variance based on this statistical deviation value.

[0072] In the filling process of large variable wall thickness castings, the melt viscosity increases exponentially with the loss of temperature, which leads to an intrinsic conflict on the time axis between controlling the flow rate in the early stage to prevent erosion and maintaining the flow rate in the later stage to prevent cold insulation; this embodiment introduces an evaluation mechanism for solving the co-variance of the spatiotemporal thermal flow field.

[0073] Specifically, the system extracts the temperature retention rate and local velocity of the filling wavefront from the thermal field evolution data and the flow field evolution data in parallel. The system calculates the ratio of the temperature retention rate to the local velocity, uses it as a matching degree to measure the heat transfer efficiency, and calculates the spatiotemporal thermal flow field cooperative variance based on the spatiotemporal discrete distribution of the matching degree.

[0074] In terms of capturing abnormal flow fields, if the extracted local velocity approaches zero, i.e., it falls into a flow dead zone, it will cause the ratio calculation to have a division-to-zero anomaly or a meaningless maximum value. At this time, the system directly assigns a penalty coefficient with a very high cooperative variance to the region and marks it as an extremely high-risk cold gap region; within the normal flow range, the standard ratio variance statistics are performed.

[0075] For example, when the target fluid-filled wavefront advances to the thin-walled reinforcing rib area at the top of the wind turbine housing, although the wavefront temperature retention rate is still maintained at 92%, the local flow velocity drops sharply to 0.05 m / s due to the thin-walled resistance. This extremely high temperature and extremely low flow velocity result in a severe imbalance between the two. The system calculates that the spatiotemporal thermal flow field co-variance is higher than the co-variance threshold at this time, and determines that there is a potential cold insulation risk.

[0076] The purpose of this step is to establish a variance system that comprehensively measures the matching degree between wavefront temperature and flow velocity, so as to accurately determine whether the current flow field can provide sufficient enthalpy to prevent premature solidification of the thin-walled region while avoiding turbulence.

[0077] In a preferred embodiment of the present invention, modifying the initial filling parameters includes: reducing the feature size value in the spatial topology parameters when the local kinetic energy dissipation rate is less than the kinetic energy dissipation threshold; and keeping the spatial topology parameters unchanged when the local kinetic energy dissipation rate is greater than or equal to the kinetic energy dissipation threshold.

[0078] When the system detects insufficient kinetic energy dissipation capacity, blindly prioritizing the adjustment of temperature or initial pouring flow rate can easily disrupt the overall process rhythm and cause secondary defects. This embodiment provides a parameter correction mechanism that prioritizes intervention in physical boundary constraints.

[0079] Specifically, when the local kinetic energy dissipation rate is less than the kinetic energy dissipation threshold, the system determines that the current throttling structure is too broad and directly reduces the characteristic size value in the spatial topology parameters to enhance the physical flow obstruction effect.

[0080] When reducing the feature size value, the correction step size is determined using an error-based adaptive gradient descent method. The specific mathematical logic is as follows:

[0081]

[0082] in, This is the correction step size for the feature size value, and a preset topology correction coefficient with length dimensions. Its unit is consistent with the unit of the feature size value, for example, mm, and its value range is typically 0.01 mm to 0.1 mm. The preset kinetic energy dissipation threshold, The local kinetic energy dissipation rate is calculated at the current time. Through the above adaptive calculation, it is ensured that as the local kinetic energy dissipation rate approaches the threshold, the correction step size gradually decreases, thereby ensuring the convergence and stability of multiple iterations of the numerical simulation.

[0083] Conversely, if the local kinetic energy dissipation rate is greater than or equal to the kinetic energy dissipation threshold, the existing spatial topology parameters remain unchanged.

[0084] For example, a simulation of a wind turbine housing model shows that the local kinetic energy dissipation rate of its secondary inlet channel is only 15%, which is significantly lower than the set 30% kinetic energy dissipation threshold. At this time, the system triggers a correction mechanism to reduce the characteristic size value of the spatial topology parameter of the inlet channel cross-sectional area from 200 square centimeters to 150 square centimeters, forcing the fluid to be subjected to stronger geometric resistance when passing through the area.

[0085] The purpose of this mechanism is to follow the objective laws of fluid mechanics and prioritize the peak-shaving action from the rigid spatial topology dimension to ensure that the casting system itself has the basic ability to intercept high-inertia jets.

[0086] In a preferred embodiment of the present invention, the modification of the initial filling parameters further includes: reducing the initial kinetic parameters when the free surface topological fracture index is not lower than the fracture index threshold; and keeping the initial kinetic parameters unchanged when the free surface topological fracture index is lower than the fracture index threshold.

[0087] If the system detects severe oscillations on the filled liquid surface even after the spatial topology has been adjusted for diameter reduction, it means that the kinetic energy of the input source exceeds the physical limit that the structure can dissipate. This embodiment further supplements a correction mechanism for the source dynamics. Specifically, when the free liquid surface topological fracture index is not lower than the fracture index threshold, the system actively reduces the initial dynamic parameters to weaken the source potential energy. Conversely, when the free liquid surface topological fracture index is lower than the fracture index threshold, the initial dynamic parameters remain unchanged.

[0088] For example, after the runner diameter reduction optimization, the local kinetic energy dissipation rate has indeed met the standard, but the topological fracture index of the free liquid surface at the bottom of the large plane of the main box is still as high as 0.7, which is higher than the threshold of 0.5. At this time, the system keeps the topology unchanged and instead reduces the initial pouring flow rate from 2.5m / s to 1.8m / s, thereby directly reducing the Reynolds number of the injected target fluid.

[0089] The purpose of this mechanism is to further mitigate the risks of interface breakage and severe gas entrapment by dynamically suppressing the initial fluid kinetic energy of the input source, under the premise that the spatial constraint capability is close to the physical limit.

[0090] In a preferred embodiment of the present invention, the modification of the initial filling parameters further includes: increasing the initial thermodynamic parameters when the spatiotemporal heat flow field cooperative variance is not lower than the cooperative variance threshold; and keeping the initial thermodynamic parameters unchanged when the spatiotemporal heat flow field cooperative variance is lower than the cooperative variance threshold.

[0091] After the dual forced constraints of reduced flow rate and reduced topology diameter, the disturbance of the flow field is suppressed, but this usually leads to a slow flow rate at the end, which causes the cooling rate at the thin wall to increase dramatically, and the technical disadvantage of maintaining flow rate and preventing cold insulation is exposed. This embodiment adds a temperature compensation mechanism at the end of the optimization link.

[0092] Specifically, if the spatiotemporal thermal flow field cooperative variance is not lower than the cooperative variance threshold, it means that the current smooth flow field can no longer maintain sufficient heat supply at the far end, and the system will increase the initial thermodynamic parameters to inject additional enthalpy; conversely, if the spatiotemporal thermal flow field cooperative variance is lower than the cooperative variance threshold, the initial thermodynamic parameters will remain unchanged.

[0093] For example, although the turbulence in the large cavity was resolved by effectively suppressing the initial flow velocity in the early stage, the cooperative variance at the thin-walled reinforcing rib at the far end of the wind turbine box exceeded the standard alarm; without changing the existing low flow velocity and topology parameters, the system specifically increased the initial pouring temperature from 1380°C to 1410°C.

[0094] The purpose of this step is to use the increased enthalpy as the final closed-loop compensation measure of the optimization mechanism to dynamically compensate for the sudden drop in temperature caused by the lack of kinetic energy without disrupting the established flow field stability, and ultimately achieve the thermal-fluid co-evolution balance.

[0095] In a preferred embodiment of the present invention, the target fluid is a molten metal, and the target cavity is a mold cavity for casting wind turbine housings; wherein, the spatial topology parameters include the cross-sectional area ratio of the ingate, and the initial thermodynamic parameters include the pouring temperature curve; the spatial topology data includes the thermal section size data and the thin-walled reinforcing rib size data in the mold cavity for casting wind turbine housings;

[0096] Numerical simulation of the filling process of the target fluid in the target cavity is performed, including: using a preset finite volume method model, discretizing and iteratively calculating the filling process based on the mesh model and initial filling parameters, and generating flow field evolution data and thermal field evolution data.

[0097] This embodiment maps and encapsulates the highly abstract data optimization model and logical links described above into the underlying physical entity and core algorithm framework of large-scale wind power equipment manufacturing. Specifically, the target fluid is clearly defined as a molten metal with high inertia and thermal shock effect; the target cavity is the mold cavity used for casting wind turbine housings; in this scenario, the spatial topology parameters are visualized as the cross-sectional area ratio of the ingate that determines the flow velocity distribution, while the initial thermodynamic parameters are visualized as the pouring temperature curve with time-series characteristics.

[0098] Meanwhile, the spatial topology data accurately contains the size data of the thick hot section and the thin-walled stiffener that are prone to nonlinear mutations in the complex cavity; in the core layer numerical solution stage, the system adopts the preset finite volume method model, and performs high-frequency discretization iterative calculation based on the divided grid model and parameter combination to approximate the real physical state and output flow field evolution data and thermal field evolution data.

[0099] In the fault-tolerant backup logic for solving anomalies at the bottom layer, if the finite volume method model encounters an excessive Coulomb number in the extremely distorted thin-walled stiffened mesh region, causing iterative divergence, the system will pause the simulation and forcibly start the mesh adaptive re-meshing program to repair the mesh quality before resuming the discretization iterative calculation; under normal mesh quality, the scalar field and vector field results will be output continuously and stably.

[0100] For example, in the finite volume method underlying derivation of a 5-ton-class large-mass wind turbine enclosure, the system concretizes the extreme variable wall thickness structure as the cross-sectional area ratio of the straight-lined inlet gating system, such as the ratio of the cross-sectional area of ​​the straight gating system to the horizontal gating system to the inlet gating system. This will output a dynamic pouring temperature curve to guide on-site operations. For example, the initial pouring temperature will be controlled at 1420°C to facilitate filling, and the temperature will be reduced to 1390°C in the middle and later stages to reduce thermal shock.

[0101] The purpose of this combined mechanism is to connect the fundamental laws of fluid mechanics and thermodynamics with engineering solutions that significantly reduce actual casting defects, ensuring that the computational cost of the algorithm is completely controllable, and that the combination of key output parameters has strong significance for industrial field guidance and execution.

[0102] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation, characterized in that, The method includes: Acquire the spatial topology data of the target cavity and the preset initial filling parameters, wherein the initial filling parameters include spatial topology parameters, initial kinetic parameters and initial thermodynamic parameters; Based on the spatial topology data, the target cavity is spatially discretized to generate a mesh model composed of multiple spatial nodes; Based on the mesh model and the initial filling parameters, the filling process of the target fluid in the target cavity is numerically simulated to obtain flow field evolution data and thermal field evolution data. Based on the flow field evolution data, the gas-liquid two-phase interface and the filling wavefront of the target fluid during the filling process are tracked, and the local kinetic energy dissipation rate of each spatial node in the grid model and the free surface topological rupture index of the gas-liquid two-phase interface are calculated. Based on the flow field evolution data and the thermal field evolution data, calculate the spatiotemporal thermal flow field co-variance of the filling wavefront; Determine whether the preset optimization conditions are met. The preset optimization conditions are: the local kinetic energy dissipation rate is greater than or equal to the preset kinetic energy dissipation threshold, the free surface topological rupture index is lower than the preset rupture index threshold, and the spatiotemporal thermal flow field cooperative variance is lower than the preset cooperative variance threshold. If satisfied, the initial filling parameters are determined as the target filling parameters and output; if not satisfied, the initial filling parameters are corrected and the process returns to the step of numerically simulating the filling process of the target fluid in the target cavity.

2. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The calculation of the local kinetic energy dissipation rate of each spatial node in the grid model includes: extracting the peak velocity data of each spatial node from the flow field evolution data; calculating the velocity reduction of each spatial node on the target fluid based on the peak velocity data and a preset fluid attenuation model; and determining the local kinetic energy dissipation rate based on the velocity reduction.

3. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The calculation of the free surface topological fracture index of the gas-liquid two-phase interface includes: extracting morphological feature data of the gas-liquid two-phase interface from the flow field evolution data; calculating the degree of fragmentation and the degree of rollover of the gas-liquid two-phase interface based on the morphological feature data; and performing a weighted summation of the degree of fragmentation and the degree of rollover to generate the free surface topological fracture index.

4. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The calculation of the spatiotemporal thermal flux field co-variance of the molding wavefront includes: extracting the temperature retention rate of the molding wavefront from the thermal field evolution data; extracting the local velocity of the molding wavefront from the flow field evolution data; calculating the ratio of the temperature retention rate to the local velocity as a matching degree, and solving the spatiotemporal thermal flux field co-variance based on the matching degree.

5. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The modification of the initial filling parameters includes: reducing the feature size value in the spatial topology parameters when the local kinetic energy dissipation rate is less than the kinetic energy dissipation threshold; and keeping the spatial topology parameters unchanged when the local kinetic energy dissipation rate is greater than or equal to the kinetic energy dissipation threshold.

6. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 5, characterized in that, The modification of the initial filling parameters further includes: reducing the initial kinetic parameters when the free surface topological fracture index is not lower than the fracture index threshold; and keeping the initial kinetic parameters unchanged when the free surface topological fracture index is lower than the fracture index threshold.

7. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 6, characterized in that, The modification of the initial filling parameters further includes: increasing the initial thermodynamic parameters when the spatiotemporal heat flow field cooperative variance is not lower than the cooperative variance threshold; and keeping the initial thermodynamic parameters unchanged when the spatiotemporal heat flow field cooperative variance is lower than the cooperative variance threshold.

8. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The target fluid is a molten metal, and the target cavity is a mold cavity for casting wind turbine housings; wherein, the spatial topology parameters include the cross-sectional area ratio of the ingate, and the initial thermodynamic parameters include the pouring temperature profile.

9. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 8, characterized in that, The spatial topology data includes the thermal section size data and thin-walled reinforcing rib size data in the mold cavity used for casting the wind turbine housing.

10. The method for optimizing the filling parameters of wind turbine housing casting based on numerical simulation according to claim 1, characterized in that, The numerical simulation of the filling process of the target fluid in the target cavity includes: using a preset finite volume method model, and based on the mesh model and the initial filling parameters, performing discretized iterative calculations on the filling process to generate the flow field evolution data and the thermal field evolution data.

Images