Full-process deformation accumulation simulation and tolerance allocation method for ultra-high strength steel seat frame

By constructing a finite element simulation model and verifying it using measured force-displacement curves, virtual disturbance sensitivity analysis was conducted to identify the wave-dominant process. This solved the problems of simulation prediction deviation and high measured cost of multi-process stamping deformation accumulation in ultra-high strength steel seat frames, achieving accurate tolerance allocation and cost reduction.

CN123069816APending Publication Date: 2026-11-10WORLD ASIA (TIANJIN) AUTO PARTS CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611593647.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-15
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing technologies suffer from problems such as simulation prediction bias and high cost of actual measurement methods when analyzing and allocating tolerances for the cumulative deformation of ultra-high strength steel seat frames during multi-process stamping.

Method used

By constructing a finite element simulation model, using measured force-displacement curves to verify and correct boundary conditions, conducting virtual disturbance sensitivity analysis, identifying the wave-dominant process, and allocating tolerances.

Benefits of technology

It improves the accuracy of deformation prediction, reduces testing costs and time, captures the nonlinear interaction amplification effect between processes, and ensures that tolerance allocation matches the actual contribution level.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123069816A_ABST
    Figure CN123069816A_ABST
Patent Text Reader

Abstract

This application provides a method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frames, relating to the field of stamping simulation technology. A full-process stamping finite element simulation model of the seat frame is constructed, and sheet metal parameters are set based on the mechanical property data of ultra-high strength steel. The simulation boundary conditions are verified and corrected through experimental force-displacement curves to ensure consistency between the simulation and actual stamping. In the corrected model, virtual perturbations are applied to key process parameters of each process, and the change in dimensional deviation is calculated to obtain the initial sensitivity coefficient, identifying the dominant process of fluctuation. Two sets of different virtual perturbations are set for the dominant process to obtain the sensitivity offset of the remaining processes, and the corrected sensitivity coefficients for each process are obtained. Tolerances for each process are allocated according to the proportion of the corrected sensitivity coefficients; the higher the sensitivity, the larger the tolerance allocation. Simulation quantifies the deformation coupling effect between processes, improves the accuracy of tolerance allocation, and adapts to the requirements of multi-process stamping deformation accumulation control of ultra-high strength steel.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of stamping simulation technology, specifically involving a method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame. Background Technology

[0002] Ultra-high strength steel seat frames typically require multiple stamping processes, including blanking, drawing, trimming and punching, and flanging. The deformation from each process accumulates and is transmitted sequentially, resulting in dimensional deviations in the final part caused by the sum of independent forming errors from each process and the datum transmission errors from each stage. To ensure the final part meets dimensional standards, tolerances for each process need to be rationally allocated during the design phase.

[0003] Currently, the analysis and tolerance allocation of cumulative deformation in multi-process stamping in engineering mainly rely on two methods. One is the full-process deformation prediction method based on finite element simulation. This method establishes simulation models for each process, simulates the deformation behavior of the sheet metal in each process, predicts the springback and dimensional deviation of the final part, and then allocates the tolerances for each process accordingly. However, ultra-high strength steel exhibits complex Bauschinger effect and hardening attenuation behavior under cyclic loading and reverse loading, which is difficult to accurately describe by the material constitutive model. This results in a large deviation between the deformation predicted by simulation and the actual situation, leading to unreasonable tolerance allocation. The other method involves taking intermediate parts from each process for actual measurement during the trial molding stage, and using the measured data to infer the deformation contribution of each process. However, this method requires repeated mold disassembly and measurement, which is time-consuming and costly.

[0004] Therefore, existing technologies lack a comprehensive deformation accumulation analysis and tolerance allocation method that can both avoid deformation prediction deviations caused by inaccurate material models in pure simulation and reduce the trial molding costs of pure experimental methods. Summary of the Invention

[0005] In view of the above-mentioned defects or deficiencies in the existing technology, a method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame is provided, including the following steps: A finite element simulation model of the entire stamping process of the seat frame is constructed. The finite element simulation model includes mold models and sheet metal models for each process. The material parameters of the sheet metal model are set based on the mechanical property test data of ultra-high strength steel. During the trial molding stage, the measured force-displacement curves of each process are obtained under the condition of no disturbance to the process parameters during actual stamping. The measured force-displacement curves are used to verify and correct the boundary conditions of each process in the finite element simulation model so that the force-displacement curves output by the simulation model are consistent with the measured force-displacement curves. In the modified finite element simulation model, under the condition that the process parameters of the remaining processes other than the target process remain unchanged, a virtual disturbance of a predetermined amplitude is applied to at least one key process parameter of the target process. The finite element simulation model calculates the change in the key dimension deviation of the final part, and uses the change in deviation as the initial sensitivity coefficient of the process to the final deformation. The process with the largest change in deviation among all processes is determined as the fluctuation-dominant process. Under the condition that the process parameters of all processes except the fluctuation-dominant process remain unchanged, two sets of virtual perturbations of different amplitudes are applied to the fluctuation-dominant process to obtain two process states of the fluctuation-dominant process; under each process state, the virtual perturbation of the predetermined amplitude is applied to each process other than the fluctuation-dominant process, and the final part critical dimension deviation change under each combination is calculated by the finite element simulation model; for the same process, the difference between the deviation changes obtained under the two process states is taken as the sensitivity offset of the process. The initial sensitivity coefficient is corrected based on the sensitivity offset of each process to obtain the corrected sensitivity coefficient of each process; The critical dimension tolerance values ​​of the final part are allocated to the process tolerances of each process according to the proportion of the correction sensitivity coefficient of each process to the total correction sensitivity coefficient. The process with the larger the correction sensitivity coefficient is, the larger the tolerance value is allocated.

[0006] According to the technical solution provided in this application, applying a virtual disturbance of a predetermined magnitude to at least one key process parameter of the target process includes the following steps: From the corrected finite element simulation model, extract the force-displacement curve of the target process under the condition of no disturbance process parameters, and calculate the plastic deformation work done by the target process on the sheet metal based on the force-displacement curve; A preset plastic work disturbance value is set, which is not greater than a preset proportion of the plastic deformation work; Using the plastic work disturbance value as an increment, and combining the characteristics of the force-displacement curve, the change in key process parameters required to increase the plastic work disturbance value is calculated in reverse, and this change is taken as the predetermined amplitude.

[0007] According to the technical solution provided in this application, in the modified finite element simulation model, the virtual disturbance is applied to each process and the deviation change is calculated, including the following steps: All processes are arranged in the stamping sequence as the first to the Nth process, where N is an integer greater than or equal to 2; First, according to the stamping sequence from the first to the Nth, the virtual disturbance is applied to each process in turn. The deviation change of the key dimensions of the final part is calculated by the finite element simulation model, and the first set of deviation changes of each process in the sequentially increasing direction is obtained. Then, following the reverse stamping sequence from the Nth to the first, the virtual disturbance is applied to each process sequentially. The finite element simulation model is used to calculate the deviation changes of the key dimensions of the final part, and the second set of deviation changes of each process in the sequential decreasing direction is obtained. The step of using the change in deviation as the initial sensitivity coefficient of the process to the final deformation includes the following steps: For each process, the average of the first set of deviation changes and the second set of deviation changes is used as the initial sensitivity coefficient of that process to the final deformation.

[0008] According to the technical solution provided in this application, before applying two sets of virtual disturbances of different amplitudes to the fluctuation-dominant process, the following steps are also included: Obtain the upper and lower limits of the key process parameters of the process with the most fluctuating parameters in historical qualified production batches. The upper and lower limits of the parameter values ​​define the normal production allowable range of the key process parameters. The normal production allowable range is divided into two non-overlapping sub-intervals, with the first sub-interval closer to the upper limit of the parameter value and the second sub-interval closer to the lower limit of the parameter value. Among the two sets of virtual disturbances of different amplitudes, one set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the first sub-interval, and the other set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the second sub-interval.

[0009] According to the technical solution provided in this application, obtaining the first set of deviation changes in the sequentially increasing direction and the second set of deviation changes in the sequentially decreasing direction for each process includes the following steps: For any direction between the sequentially increasing direction and the sequentially decreasing direction, take it as the current test direction; When applying the virtual disturbance to the target process, in the finite element simulation model, the random fluctuation range of the sheet material parameters and / or boundary conditions is set, at least two virtual sheet samples are randomly generated and the deviation values ​​of the key dimensions of the final part are calculated respectively. The arithmetic mean of these deviation values ​​is taken as the average value of the disturbance state deviation of the process in the current test direction. Under undisturbed conditions, at least two virtual sheet metal samples are randomly generated and the arithmetic mean of the final critical dimension deviation values ​​of the parts is calculated as the undisturbed deviation benchmark value. The difference between the mean deviation of the disturbance state of the process and the reference value of the undisturbed deviation is taken as the deviation change of the process in the current test direction.

[0010] According to the technical solution provided in this application, before applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage, the method further includes the following steps: Obtain the measured temperature change curves of the mold from a cold state to a thermal equilibrium state during continuous production of each process. Based on the measured temperature change curves, determine the temperature rise amplitude of each process during continuous production. Mark the process with the largest temperature rise amplitude among all processes as a temperature drift sensitive process. The moment when the temperature rise rate first falls below the preset steady-state rate in the temperature rise curve of the temperature drift-sensitive process is taken as the moment when the mold temperature of the temperature drift-sensitive process stabilizes. The process of applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage includes setting the temperature boundary conditions of the mold for each process in the finite element simulation model to the thermal equilibrium temperature after the mold temperature stabilizes.

[0011] According to the technical solution provided in this application, before applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage, the method further includes the following steps: In the finite element simulation model, the initial stress state and initial temperature state of the mold for each process are set as the initial reference state; It also includes the following steps: After completing the virtual disturbance analysis for each process, the current stress state or current temperature state of the mold for that process is extracted from the simulation results, compared with the initial reference state, and the cumulative offset of the mold state is calculated. If the cumulative offset of the mold state is less than the preset offset threshold, the current state is used as the new initial reference state for subsequent processes to determine the cumulative offset of the mold state, and the virtual disturbance is applied to the next process.

[0012] According to the technical solution provided in this application, after calculating the cumulative offset of the mold state, the following steps are also included: If the cumulative offset of the mold state is greater than or equal to the preset offset threshold, the virtual disturbance of the subsequent process is paused, the mold state in the finite element simulation model is reset to the initial reference state, and the virtual disturbance of the subsequent process is continued after the reset.

[0013] According to the technical solution provided in this application, the step of using the measured force-displacement curve to verify and correct the boundary conditions of each process in the finite element simulation model includes the following steps: The measured force-displacement curve and the simulated force-displacement curve output by the simulation model are segmented along the displacement axis using the same displacement interval division method; By comparing the deviations between the measured force values ​​and the simulated force values ​​within the same displacement range one by one, and based on the distribution of the deviations in each displacement range, the inaccuracy starting range where the simulation model begins to show significant deviations relative to the actual stamping process is located. Based on the aforementioned inaccuracy initiation interval, the boundary condition parameters of the finite element simulation model are corrected in segments.

[0014] According to the technical solution provided in this application, the step of segmentally correcting the boundary condition parameters of the finite element simulation model based on the inaccuracy initiation interval includes the following steps: Using the displacement value corresponding to the inaccuracy initiation interval as the boundary, only the boundary condition parameters corresponding to the simulated force-displacement curve after that displacement value are corrected, so that the subsequent segments of the corrected simulated force-displacement curve are consistent with the measured force-displacement curve; the boundary condition parameters corresponding to the displacement interval before the inaccuracy initiation interval are not corrected.

[0015] Compared with the prior art, the beneficial effects of this application are as follows: First, the simulation model was verified and corrected using measured force-displacement curves, ensuring consistency between the output force-displacement curves and the measured curves. This resolved the issue of deformation prediction deviations caused by inaccurate constitutive models of ultra-high-strength steel materials in pure simulation methods. Force-displacement curves are the most direct and readily available measured data during the stamping process. Using them as a verification benchmark, the corrected simulation model can more accurately reflect material flow and springback behavior under actual stamping conditions.

[0016] Second, virtual disturbance sensitivity analysis is performed on the simulation model after actual measurement, verification, and correction, replacing the physical disturbance testing on the physical mold during the traditional trial molding stage. Virtual disturbances do not consume sheet metal or mold lifespan, and are not affected by temperature drift or equipment status fluctuations at the trial molding site, significantly reducing the cost and time of sensitivity testing.

[0017] Third, identify the process that dominates the fluctuation and test the sensitivity offset of other processes under different amplitude disturbances to capture the nonlinear interaction amplification effect between processes. The initial sensitivity coefficient is corrected by the sensitivity offset, so that the corrected sensitivity coefficient more accurately reflects the real deformation contribution of each process under different states of the process that dominates the fluctuation.

[0018] Fourth, tolerances are allocated based on the proportion of the sensitivity coefficient of each process to the total tolerance. Processes with higher sensitivity coefficients receive larger tolerance values. This allocation method matches the tolerance zone with the actual contribution of each process to the final deformation, avoiding the problems of over-constraining critical processes and over-relaxing non-critical processes in traditional proportional or experience-based allocations. Attached Figure Description

[0019] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1The flowchart shows the steps of the full-process deformation accumulation simulation and tolerance allocation method for the ultra-high strength steel seat frame provided in this application. Detailed Implementation

[0020] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] As mentioned in the background section, this application proposes a method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frames, such as... Figure 1 As shown, it includes the following steps: S1. Construct a finite element simulation model of the entire stamping process of the seat frame. The finite element simulation model includes mold models and sheet metal models for each process. The material parameters of the sheet metal model are set based on the mechanical property test data of ultra-high strength steel. S2. During the trial molding stage, the measured force-displacement curves of each process are obtained under the condition of no disturbance to the process parameters during actual stamping. The measured force-displacement curves are used to verify and correct the boundary conditions of each process in the finite element simulation model so that the force-displacement curves output by the simulation model are consistent with the measured force-displacement curves. S3. In the modified finite element simulation model, under the condition that the process parameters of the other processes except the target process remain unchanged, a virtual disturbance of a predetermined amplitude is applied to at least one key process parameter of the target process. The finite element simulation model calculates the change in the key dimension deviation of the final part, and uses the change in deviation as the initial sensitivity coefficient of the process to the final deformation. The process with the largest change in deviation among all processes is determined as the fluctuation-dominant process. S4. Under the condition that the process parameters of the remaining processes, except for the fluctuation-dominant process, remain unchanged, apply two sets of virtual disturbances of different amplitudes to the fluctuation-dominant process to obtain two process states of the fluctuation-dominant process; under each process state, apply the virtual disturbance of the predetermined amplitude to each process other than the fluctuation-dominant process, and calculate the change in the critical dimension deviation of the final part under each combination by the finite element simulation model; for the same process, the difference between the deviation changes obtained under the two process states is taken as the sensitivity offset of the process. S5. Based on the sensitivity offset of each process, the initial sensitivity coefficient is corrected to obtain the corrected sensitivity coefficient of each process. S6. Allocate the critical dimension tolerance values ​​of the final part to the process tolerances of each process according to the proportion of the correction sensitivity coefficient of each process to the total correction sensitivity coefficient. The process with the larger correction sensitivity coefficient is allocated a larger tolerance value.

[0023] Specifically, in the description of this application, ultra-high strength steel refers to high-strength steel plates with a yield strength of over 550 MPa and a tensile strength of over 1000 MPa, typical grades including 22MnB5, DP980, and QP1180. A seat frame refers to the load-bearing frame structure of an automotive seat, typically formed by welding multiple steel plates or integral stamping, possessing complex spatial curved surfaces and localized reinforcement features. The entire process refers to the complete sequence of stamping operations from blanking to final springback stabilization. Deformation accumulation refers to the sequential transmission and superposition of deformation and residual stress generated in each process during multi-process stamping, with the final part's dimensional deviation being the vector sum of the deformation contributions from each process. A finite element simulation model refers to a computational model that discretizes the geometric model of the part and mold into a finite number of elements, and uses numerical methods to solve for the stress-strain distribution and geometric changes of the sheet metal during the stamping process by setting material constitutive models, contact friction conditions, and load boundary conditions. The force-displacement curve is a continuous curve showing the change in forming force on the die as the die closes during stamping. It is the most direct process monitoring quantity in stamping. Virtual disturbance refers to applying a predetermined change to a process parameter in a finite element simulation model to simulate the fluctuation of that parameter in actual production, rather than adjusting the parameter on a physical die before actual stamping. Sensitivity coefficient refers to the change in the critical dimension deviation of the final part caused by a unit disturbance in the process parameter of a certain process, reflecting the degree of influence of that process on the final deformation. The fluctuation-dominant process is the process with the highest sensitivity coefficient among all processes; the fluctuation of this process has the most significant impact on the final deformation. Sensitivity offset refers to the difference in the sensitivity coefficient of the remaining processes under different process states of the fluctuation-dominant process, reflecting the strength of the nonlinear interaction between processes. Process tolerance refers to the allowable dimensional deviation range assigned to a certain process, serving as the basis for die manufacturing and process control.

[0024] This implementation method includes four main stages: simulation model construction and verification, virtual disturbance sensitivity analysis, interaction testing, and tolerance allocation.

[0025] The first stage involves constructing a finite element simulation model of the entire stamping process for the seat frame. The complete stamping process for a seat frame typically includes four to six steps, with a typical sequence being blanking, drawing, trimming and punching, and flanging / forming. The finite element simulation model is constructed using general-purpose stamping simulation software, such as AutoForm or Dynaform. First, the 3D models of the dies for each step are imported, including the geometry of the punch, die, and blank holder. Then, the 3D model of the sheet metal is imported, or a flat blank is generated directly in the software. The material parameters of the sheet metal model are set based on the mechanical property test data of ultra-high strength steel. The test data includes yield strength, tensile strength, elongation, and thickness anisotropy coefficient obtained from uniaxial tensile tests, as well as Bauschinger effect parameters and hardening curves obtained from cyclic loading tests. The material constitutive model can be either the Yoshida-Uemori model or the Chaboche model; these two hybrid hardening models can describe the Bauschinger effect and hardening decay behavior of ultra-high strength steel under cyclic and reverse loading. The boundary conditions, such as the motion curves of the dies for each step, blank holder force, and friction coefficient, are initially set according to the design values.

[0026] The second stage involves acquiring measured data and verifying and correcting the simulation model during the trial molding phase. During trial molding, a small number of blanks are continuously stamped using the same set of dies according to the formal stamping process parameters. The press's force and displacement sensors automatically record the complete force-displacement curves for each process. The force sensors are typically strain gauge force sensors mounted on the press slide or die, and the displacement sensors are typically optical or magnetic scales. Both are synchronously acquired by the press control system, with a sampling frequency of no less than 100 Hz. After acquiring the measured force-displacement curves for each process under undisturbed process parameters, they are compared with the simulated force-displacement curves output by the simulation model under the same process parameters. If discrepancies are found, boundary condition parameters such as the friction coefficient, die clearance, and heat transfer coefficient in the simulation model are adjusted, recalculated, and compared again. This iterative correction continues until the simulated force-displacement curves and the measured force-displacement curves are consistent across the entire displacement range. The consistency criterion can be that the envelope deviation between the two curves does not exceed a preset percentage, such as five percent or ten percent. The revised simulation model accurately reflects the actual mechanical behavior of the mold and sheet metal during the stamping process, laying a reliable foundation for subsequent virtual disturbance analysis.

[0027] In the third stage, virtual disturbance sensitivity analysis is performed in the modified simulation model. Each process is selected as the target process in turn. Under the condition that the process parameters of all processes except the target process remain unchanged, a virtual disturbance of a predetermined amplitude is applied to a key process parameter of the target process. The key process parameter refers to an independently adjustable parameter that has a significant impact on the forming behavior of the process. The key process parameter of the drawing process is usually the blank holder force or the coefficient of friction; the key process parameter of the flanging process is usually the flanging gap or the stamping speed; and the key process parameter of the trimming process is usually the blanking gap. The method for determining the predetermined amplitude will be detailed later. After applying the virtual disturbance, the finite element simulation model calculates the critical dimension deviation value of the final part under the disturbance condition. The critical dimension refers to the functional dimension of the seat frame that has an important impact on the assembly and performance, such as the coordinates of the mounting holes, the profile of the mating surface, and the position of the critical boundary. The deviation value calculated under the disturbance condition is subtracted from the reference deviation value calculated under the undisturbed condition. The difference is the change in the critical dimension deviation of the final part caused by the disturbance of this amplitude. This change in deviation is used as the initial sensitivity coefficient of the process to the final deformation. The larger the deviation change, the stronger the amplification effect of the process parameter fluctuation on the final dimensional deviation. After all processes have been tested, the deviation changes of each process are compared, and the process with the largest deviation change is identified as the process dominated by the fluctuation.

[0028] In the fourth stage, the interactions between processes are tested and tolerances are allocated. With the process parameters of all processes except the one dominated by the fluctuation remaining constant, two sets of virtual perturbations of different amplitudes are applied to the fluctuation-dominant process, resulting in two different process states. Under each process state, virtual perturbations of the same amplitude as in the third stage are applied to each process other than the fluctuation-dominant process. The simulation model calculates the deviation changes of the final part's critical dimensions under each combination of conditions. For the same process, the deviation changes measured under the two process states of the fluctuation-dominant process are subtracted, and the difference is used as the sensitivity offset for that process. The sensitivity offset reflects the nonlinear amplification or reduction effect of the fluctuation-dominant process on the sensitivity of other processes. The larger the sensitivity offset, the more significantly the sensitivity of that process is affected by the changes in the state of the fluctuation-dominant process. Based on the sensitivity offset of each process, the initial sensitivity coefficient is corrected to obtain the corrected sensitivity coefficient. The correction method can be to add the sensitivity offset to the initial sensitivity coefficient, or multiply it by an adjustment coefficient positively correlated with the sensitivity offset. The corrected sensitivity coefficient comprehensively reflects the independent sensitivity of each process and its interactive sensitivity with the fluctuation-dominant process. The critical dimensional tolerances of the final part are allocated to the process tolerances of each process according to the proportion of each process's correction sensitivity coefficient to the total correction sensitivity coefficient. The larger the correction sensitivity coefficient of a process, the greater its overall contribution to the final deformation, and the larger the tolerance value allocated to it, thus granting it a wider process window; conversely, the smaller the tolerance value allocated to it, the more stringent the process requirements.

[0029] The technical effect and principle of this implementation method lies in verifying and correcting the simulation model using measured force-displacement curves, thus solving the deformation prediction deviation problem caused by the inaccuracy of the constitutive model of ultra-high strength steel materials in pure simulation. Force-displacement curves are the most direct process monitoring quantities during stamping, comprehensively reflecting the actual behavior of multiple factors such as material flow, frictional contact, and die deformation. Correcting the simulation model based on measured force-displacement curves significantly improves the model's approximation of actual working conditions. Virtual disturbance analysis is performed in the corrected simulation model, replacing the extensive physical disturbance testing on the physical die in traditional trial molding, greatly reducing testing costs and time. By identifying the dominant processes of fluctuation and testing their interactions, the nonlinear coupling effect ignored by traditional tolerance allocation methods is captured. Finally, tolerances are allocated according to the correction sensitivity coefficient ratio, ensuring that the process control requirements of each process match their actual impact on the final quality.

[0030] In a preferred embodiment, applying a virtual perturbation of a predetermined magnitude to at least one key process parameter of the target process includes the following steps: From the corrected finite element simulation model, extract the force-displacement curve of the target process under the condition of no disturbance process parameters, and calculate the plastic deformation work done by the target process on the sheet metal based on the force-displacement curve; A preset plastic work disturbance value is set, which is not greater than a preset proportion of the plastic deformation work; Using the plastic work disturbance value as an increment, and combining the characteristics of the force-displacement curve, the change in key process parameters required to increase the plastic work disturbance value is calculated in reverse, and this change is taken as the predetermined amplitude.

[0031] Specifically, it is necessary to apply a predetermined amplitude of perturbation to each process to test the sensitivity coefficient of each process. However, the key process parameters of different processes are of different types. The blank holder force in the drawing process, the stamping speed in the flanging process, and the blanking clearance in the trimming process have significantly different dimensions and normal fluctuation ranges. If the same absolute amplitude or the same percentage of perturbation is applied to each process, the resulting physical effects will be completely different, and the sensitivity coefficients measured for each process will not be on the same comparison benchmark, making it impossible to make fair rankings between processes. This embodiment uses plastic deformation work as a unified perturbation measurement standard, making the perturbations received by each process equivalent on an energy scale, thus solving the problem of normalizing the perturbation amplitude across processes.

[0032] The specific operation steps of this implementation method are as follows. First, extract the force-displacement curve of the target process under undisturbed process parameters from the modified finite element simulation model. This curve is calculated and output in the modified simulation model under the condition that all process parameters are set to the design baseline values. The horizontal axis of the force-displacement curve represents the mold closing displacement, and the vertical axis represents the forming force borne by the mold. Second, calculate the plastic deformation work done on the sheet metal by the target process based on the force-displacement curve. The plastic deformation work is defined as the area remaining after deducting the elastic recovery portion after unloading from the area below the force-displacement curve. Specifically, the force-displacement curve is integrated from the zero displacement point to the maximum displacement point to obtain the total deformation work; then, the elastic recovery work corresponding to the area below the unloading curve is subtracted, and the difference is the plastic deformation work. The integration calculation can be performed directly using the curve integration tool in the simulation software, or the curve data can be exported and calculated using the trapezoidal numerical integration method in spreadsheet software. The physical meaning of the plastic deformation work is the energy consumed by the process to cause permanent plastic deformation of the sheet metal, comprehensively reflecting the forming strength and the degree of influence on the sheet metal of the process.

[0033] The third step is to set a preset plastic work disturbance value. The plastic work disturbance value is the energy-scale representation of the disturbance intensity that the process should achieve, and its value is no greater than a preset proportion of the unperturbed plastic deformation work of that process. The preset proportion is determined based on the forming characteristics and fluctuation tolerance of the process. For example, the drawing process, as a forming process, has a relatively large plastic deformation work, so the preset proportion can be 3% to 5%; the trimming process, as a blanking process, has a relatively small plastic deformation work, so the preset proportion can be 5% to 10%. The plastic work disturbance value is equal to the plastic deformation work multiplied by the preset proportion. Since all processes determine their disturbance values ​​using the same preset proportion of their respective plastic deformation work, the disturbances experienced by each process are proportional on their own energy scale, and the measured sensitivity coefficients are comparable across processes.

[0034] The fourth step involves using the plastic work disturbance value as the increment, and combining it with the characteristics of the force-displacement curve, to inversely calculate the change in key process parameters required to increase the plastic deformation work by that disturbance value. The inverse calculation method involves using the key process parameters as variables and the plastic work disturbance value as the target value in the simulation model, solving the problem through numerical iteration. Specifically, an initial, guessed change in the key process parameter is given; the force-displacement curve after the change is calculated in the simulation model, and the plastic deformation work is determined; the calculated increment of plastic deformation work is compared with the target plastic work disturbance value. If it is too small, the parameter change is increased; if it is too large, the parameter change is decreased. This process is iterated repeatedly until the deviation between the increment of plastic deformation work and the plastic work disturbance value is within a preset tolerance range. The final parameter change is the predetermined amplitude for this process.

[0035] The technical effect and principle of this implementation method lies in shifting the normalization problem of cross-process disturbance amplitude from the process parameters themselves to the energy scale. Plastic deformation work is a measure of the energy required for plastic deformation of the sheet metal during stamping; it is independent of the type of process parameters and has comparability across processes. Testing each process with disturbances of the same energy proportion ensures that the measured sensitivity coefficients are based on a fair comparison benchmark.

[0036] In a preferred embodiment, in the modified finite element simulation model, the virtual disturbance is applied to each process step and the deviation change is calculated, including the following steps: All processes are arranged in the stamping sequence as the first to the Nth process, where N is an integer greater than or equal to 2; First, according to the stamping sequence from the first to the Nth, the virtual disturbance is applied to each process in turn. The deviation change of the key dimensions of the final part is calculated by the finite element simulation model, and the first set of deviation changes of each process in the sequentially increasing direction is obtained. Then, following the reverse stamping sequence from the Nth to the first, the virtual disturbance is applied to each process sequentially. The finite element simulation model is used to calculate the deviation changes of the key dimensions of the final part, and the second set of deviation changes of each process in the sequential decreasing direction is obtained. The step of using the change in deviation as the initial sensitivity coefficient of the process to the final deformation includes the following steps: For each process, the average of the first set of deviation changes and the second set of deviation changes is used as the initial sensitivity coefficient of that process to the final deformation.

[0037] Specifically, in the virtual disturbance test of multi-process stamping, when a disturbance is applied to a process according to the stamping sequence, the test blank has already undergone the virtual stamping process of all preceding processes. Although the process parameters of the preceding processes remain unchanged, the virtual stamping state of the preceding processes also exhibits slight random fluctuations between different test rounds due to the different disturbance processes. These fluctuations accumulate sequentially, resulting in the virtual sheet state reaching a later process being mixed with the accumulated components of the preceding fluctuations. If the test is only performed once according to the stamping sequence, the test results of the earlier processes are less affected by the preceding fluctuations, while the test results of the later processes are more affected, resulting in a systematic bias in the sensitivity coefficients of each process. This implementation scheme uses two rounds of testing, one in forward order and one in reverse order, and takes the average value. This ensures that the interference direction of the accumulated fluctuations from the preceding processes experienced by each process in the two tests is opposite, and the average value cancels each other out, thus obtaining the true sensitivity of the process to the final deformation.

[0038] The specific operation steps of this implementation method are as follows. First, arrange all processes according to the actual stamping sequence as processes 1 to N, where N is an integer greater than or equal to 2. For example, a typical process sequence for a seat frame is blanking as the first process, drawing as the second process, trimming and punching as the third process, and flanging and shaping as the fourth process, where N equals 4. Second, apply virtual disturbances to each process sequentially from the first to the Nth process according to the stamping sequence and calculate the deviation changes. Specifically, take the first process as the target process and apply a virtual disturbance of a predetermined magnitude to its process parameters. Keep the process parameters unchanged for the remaining N-1 processes. Run the simulation model to calculate the critical dimension deviation value of the final part. Subtract this deviation value from the baseline deviation value under the undisturbed condition to obtain the first set of deviation changes for the first process. Then, remove the disturbance from the first process, restore its process parameters, and apply the disturbance to the second process as the target process, again keeping the other process parameters unchanged, to obtain the first set of deviation changes for the second process. This process continues until the Nth process, at which point the first set of deviation changes for each process in the sequentially increasing direction is obtained.

[0039] The third step involves applying virtual perturbations to each process sequentially from the Nth to the first stamping stage, following the reverse stamping order, and calculating the deviation changes. During the reverse-order test, the actual stamping sequence in the simulation model remains from the first to the Nth stage; the change is in the order of the target processes. Specifically, a perturbation is first applied to the Nth process as the target process, while the other processes remain unchanged, resulting in the second set of deviation changes for the Nth process. Then, the perturbation is removed from the Nth process, and the perturbation is applied to the (N-1)th process as the target process, obtaining the second set of deviation changes for the (N-1)th process. This process is repeated until the first process, obtaining the second set of deviation changes for each process in the decreasing order. The fourth step involves taking the arithmetic mean of the first and second sets of deviation changes for each process, using this average as the initial sensitivity coefficient for the final deformation of that process.

[0040] The principle of this method is illustrated using the third process as an example. In the sequentially increasing direction test, the third process, as the target process, is disturbed, while the preceding first and second processes remain undisturbed. However, these two processes also undergo virtual stamping in this test. In the sequentially decreasing direction test, the third process is still the target process, but it is tested after the fourth process has been disturbed. In the two tests, the cumulative fluctuations of the preceding processes interfere with the third process in different directions. After averaging, this interference component is canceled out, and the remaining average value is closer to the true independent sensitivity of the third process.

[0041] The technical effect and principle of this implementation method lies in eliminating the cumulative fluctuation bias introduced by the test order by averaging two rounds of tests, one in forward order and one in reverse order. Each process is located in the first half and the other in the second half of the test sequence in the two tests, respectively. The directions of the preceding fluctuation interference are opposite, and the interference components cancel each other out after averaging. This method does not add any steps other than additional simulation calculations; it only changes the selection order of the target processes, which can be easily implemented in simulation software by modifying the process disturbance flags.

[0042] In a preferred embodiment, before applying two sets of virtual disturbances of different amplitudes to the fluctuation-dominant process, the following steps are further included: Obtain the upper and lower limits of the key process parameters of the process with the most fluctuating parameters in historical qualified production batches. The upper and lower limits of the parameter values ​​define the normal production allowable range of the key process parameters. The normal production allowable range is divided into two non-overlapping sub-intervals, with the first sub-interval closer to the upper limit of the parameter value and the second sub-interval closer to the lower limit of the parameter value. Among the two sets of virtual disturbances of different amplitudes, one set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the first sub-interval, and the other set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the second sub-interval.

[0043] Specifically, applying two sets of virtual disturbances of different amplitudes to the process that dominates the fluctuation is to obtain two different process states for that process, and then test the sensitivity changes of other processes under these two states. The values ​​of the two sets of disturbance amplitudes directly determine the degree of difference between the two process states. If the difference between the two sets of disturbance amplitudes is too small, the two states are too similar, and the measured sensitivity offset is easily overwhelmed by numerical calculation errors, failing to effectively identify the nonlinear interactions between processes. If the difference between the two sets of disturbance amplitudes is too large, one set may exceed the normal fluctuation range that the process may experience in actual production, causing the measured sensitivity offset to reflect behavior under extreme abnormal conditions, thus having limited guiding significance for mass production tolerance allocation. Therefore, the two sets of disturbance amplitudes should cover both the two typical fluctuation states that the process may experience in actual production, and not exceed its allowable range for normal production.

[0044] The specific operation steps of this implementation method are as follows.

[0045] The first step is to obtain the upper and lower limits of the critical process parameters for the fluctuation-dominant process in historical qualified production batches. Historical qualified production batches refer to batches of the seat frame product that passed final part dimensional inspection during early trial production or similar product production. From the process parameter records of these batches, extract the set or actual values ​​of the critical process parameters for the fluctuation-dominant process during the production of each qualified part. Process parameter records are usually automatically stored by the press control system or production management system and can be obtained through database queries or report exports. The type of critical process parameter has been determined above. For example, when the fluctuation-dominant process is a drawing process, the critical process parameter is usually the blank holder force. Therefore, extract the maximum and minimum values ​​of the blank holder force in qualified batches from historical data. The upper limit of the parameter value is the maximum value in qualified batches, and the lower limit is the minimum value. The upper and lower limits together define the allowable range for normal production of the critical process parameter while ensuring part qualification. Parameter values ​​exceeding this range have resulted in parts being out of tolerance or scrapped in historical production.

[0046] The second step is to divide the normal production allowable range into two non-overlapping sub-intervals. The normal production allowable range is a continuous numerical interval from the lower limit to the upper limit of the parameter value. The division method is to use the midpoint of the range or a division point specified by the process engineer as the boundary, dividing the range into a first sub-interval closer to the upper limit of the parameter value and a second sub-interval closer to the lower limit of the parameter value. The two sub-intervals do not overlap; that is, the lower limit of the first sub-interval is greater than or equal to the upper limit of the second sub-interval, ensuring that the two sets of disturbances correspond to two different process states. The size of the sub-intervals can be set according to the control accuracy requirements of the parameter; typically, each sub-interval covers half the width of the normal production allowable range.

[0047] The third step is to determine the values ​​of two sets of disturbance amplitudes based on the two sub-intervals. One set of disturbance amplitudes is the value that makes the critical process parameter fall within the first sub-interval, that is, adding an offset to the baseline value of the process parameter for this process, so that the disturbed parameter value is within the first sub-interval. The other set of disturbance amplitudes is the value that makes the critical process parameter fall within the second sub-interval, that is, the disturbed parameter value is within the second sub-interval. The specific values ​​can be the midpoint of each sub-interval or the endpoint value within the sub-interval closest to the baseline value. The two sets of disturbance amplitudes correspond to the two typical fluctuation states that the fluctuation-dominant process may exhibit in actual production: deviation from the upper limit and deviation from the lower limit. Both states are within the allowable range of normal production and can represent the process drift scenarios that may actually occur in this process during mass production.

[0048] The technical effect and principle of this implementation method lies in transforming the determination of the two sets of disturbance amplitudes from subjective setting to objective extraction based on historical qualified production data. The upper and lower limits of parameter values ​​are derived from the range of process parameters that have been proven to produce qualified parts in actual production. Two sub-intervals within this range are used as two process states, ensuring the rationality and representativeness of the two sets of disturbance states in engineering. The sensitivity offset measured under these two states can accurately reflect the nonlinear impact of the fluctuation-dominant process on the sensitivity of other processes when the fluctuation is within the allowable range of actual production, making the final allocated process tolerances more suitable for mass production.

[0049] In a preferred embodiment, obtaining the first set of deviation changes in the sequentially increasing direction and the second set of deviation changes in the sequentially decreasing direction for each process both include the following steps: For any direction between the sequentially increasing direction and the sequentially decreasing direction, take it as the current test direction; When applying the virtual disturbance to the target process, in the finite element simulation model, the random fluctuation range of the sheet material parameters and / or boundary conditions is set, at least two virtual sheet samples are randomly generated and the deviation values ​​of the key dimensions of the final part are calculated respectively. The arithmetic mean of these deviation values ​​is taken as the average value of the disturbance state deviation of the process in the current test direction. Under undisturbed conditions, at least two virtual sheet metal samples are randomly generated and the arithmetic mean of the final critical dimension deviation values ​​of the parts is calculated as the undisturbed deviation benchmark value. The difference between the mean deviation of the disturbance state of the process and the reference value of the undisturbed deviation is taken as the deviation change of the process in the current test direction.

[0050] Specifically, each round of virtual disturbance testing requires calculating the deviation change of the final critical dimension of the part, i.e., the difference between the deviation values ​​of the final critical dimension of the part under disturbance and undisturbed conditions. In the finite element simulation model, if only one calculation is run, the result corresponds to a single simulation value under a specific set of sheet material parameters and boundary conditions. However, in actual production, sheet materials are not completely consistent. Sheets from the same batch exhibit slight fluctuations in material parameters such as thickness, yield strength, and elongation, and the friction coefficient between the mold and the sheet material also varies between each piece. These random fluctuations can lead to deviations between the single simulation calculation result and the statistical mean of batch production. If only the single simulation value is used as the deviation change, the measured sensitivity coefficient may be unstable due to the influence of random fluctuations. This solution introduces random fluctuations in material parameters and boundary conditions into the simulation model, generating multiple virtual sheet material samples. The arithmetic mean of the calculation results from multiple samples is used as the deviation change, and the interference of random fluctuations is suppressed by statistical averaging, thus improving the stability of the sensitivity coefficient calculation.

[0051] The specific operation steps of this implementation method are as follows.

[0052] The first step is to set the random fluctuation range of sheet material parameters and boundary conditions in the finite element simulation model when applying virtual perturbation to the target process. The fluctuating terms of sheet material parameters include sheet thickness, yield strength, tensile strength, and elongation. The fluctuating terms of boundary conditions include the friction coefficient between the die and the sheet material in each process, and the slight deviation of the blank holder force near the set value. The random fluctuation range is set based on the measured statistical distribution of the parameter in normal production batches. For example, the random fluctuation range of sheet thickness can be set to the nominal thickness plus or minus half a tolerance band, and the random fluctuation range of yield strength can be set to the nominal value plus or minus two standard deviations. In actual operation, these parameters can be set as random variables following a normal distribution in the simulation software, with the mean being the nominal value and the standard deviation being the historical batch statistical value or the value given by the technical standard.

[0053] The second step involves randomly generating at least two virtual sheet metal samples and calculating the deviation values ​​of the critical dimensions of the final part for each. Each virtual sheet metal sample corresponds to a set of material parameters and boundary conditions randomly selected according to the above random distribution. The simulation software independently runs a full-process simulation calculation for each virtual sheet metal sample, outputting the deviation value of the critical dimensions of the final part corresponding to that sample. The arithmetic mean of the deviation values ​​of at least two samples is taken, and this average value is the mean deviation of the disturbance state in this test direction for this process. The statistical principle of taking the arithmetic mean is that the mean of multiple independent random samples is closer to the true mean of the population than the value of a single sample, and the more samples there are, the smaller the standard error of the mean. At least two samples can suppress the influence of a single random fluctuation to a certain extent. In actual operation, three to five samples can be taken depending on the computing resources and accuracy requirements.

[0054] The third step, under undisturbed conditions, involves randomly generating at least two virtual sheet metal samples and calculating the arithmetic mean of the final critical dimension deviations of the parts. This arithmetic mean is used as the baseline deviation value under undisturbed conditions. The method for generating virtual samples under undisturbed conditions is exactly the same as under disturbed conditions, using the same random distribution parameters and the same number of samples. The only difference is that the process parameters for the target process remain at the design baseline value instead of applying a disturbance value. Subtracting the mean deviation value under disturbed conditions from the baseline deviation value under undisturbed conditions yields the change in deviation for that process in that test direction.

[0055] The technical effect and principle of this implementation method lies in improving the deterministic single calculation in the simulation model to a multi-sample statistical calculation that considers the fluctuations in actual production. By simulating the random fluctuations of sheet material parameters and boundary conditions in the simulation, the arithmetic mean of multiple virtual samples can more stably represent the expected deviation level of batch production under this process condition, avoiding the fluctuations in the sensitivity coefficient calculation caused by the deviation of specific random samples in a single simulation value. This method does not change the basic architecture and calculation process of the simulation model, but only adds parameter randomization settings and multi-sample loops before the calculation, which can be automatically completed in the batch processing function of the simulation software.

[0056] In a preferred embodiment, before applying the virtual disturbance to each process sequentially from the first to the Nth stamping pass according to the stamping sequence, the method further includes the following steps: Obtain the measured temperature change curves of the mold from a cold state to a thermal equilibrium state during continuous production of each process. Based on the measured temperature change curves, determine the temperature rise amplitude of each process during continuous production. Mark the process with the largest temperature rise amplitude among all processes as a temperature drift sensitive process. The moment when the temperature rise rate first falls below the preset steady-state rate in the temperature rise curve of the temperature drift-sensitive process is taken as the moment when the mold temperature of the temperature drift-sensitive process stabilizes. The process of applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage includes setting the temperature boundary conditions of the mold for each process in the finite element simulation model to the thermal equilibrium temperature after the mold temperature stabilizes.

[0057] Specifically, a complete virtual perturbation analysis needs to be performed both sequentially and in reverse order of stamping. In the simulation model, the temperature boundary conditions of the dies for each process are crucial factors affecting the sheet metal forming behavior. Die temperature influences the coefficient of friction between the sheet metal and the die contact surface, the thermal expansion of the sheet metal, and the die's own surface dimensions. In actual continuous production, the die starts stamping from a cold die state, gradually heating up after multiple stamping cycles until it reaches thermal equilibrium. The temperature difference between the cold die state and the thermal equilibrium state can reach tens of degrees Celsius or even higher, resulting in significant differences in the sheet metal forming behavior and final springback. If the die temperature boundary conditions set in the virtual perturbation analysis are inconsistent with the actual mass production conditions, the sensitivity coefficient and tolerance allocation determined based on the simulation analysis results will be difficult to apply in mass production. Traditionally, engineers often select a temperature value as the simulation boundary condition based on experience, lacking objective basis. This solution analyzes the measured temperature change curves of the molds in each process to determine the temperature drift-sensitive processes and their temperature stabilization moments. After these moments, virtual perturbation analysis is initiated using the thermal equilibrium temperature as the simulation boundary condition to ensure that the simulation state is consistent with the mass production thermal equilibrium state.

[0058] The specific operation steps of this implementation method are as follows.

[0059] The first step is to obtain the measured temperature change curves of the mold from a cold mold state to a thermal equilibrium state during continuous production for each process. The measured temperature change curves are obtained by pre-embedding armored thermocouples inside the mold body for each process. The thermocouple measuring ends are at a certain depth from the mold surface, typically five to ten millimeters, to measure the base temperature of the mold body rather than the instantaneous temperature of the mold surface. During continuous production, starting from the cold mold state, the thermocouples record temperature values ​​at fixed time intervals, typically five to thirty seconds, until the temperature stabilizes and no longer rises significantly. The recorded data are arranged in chronological order to obtain the temperature change curves of the mold from cold mold to thermal equilibrium for each process. If thermocouples are not pre-embedded in the mold, a mold surface thermometer or thermal imager can be used during the trial molding stage to measure the mold surface temperature during each stamping interval to approximate the temperature change trend. The stable temperature value at thermal equilibrium is subtracted from the initial cold mold temperature value to obtain the temperature rise amplitude for each process during continuous production. The temperature rise amplitudes of each process are compared, and the process with the largest temperature rise amplitude is marked as a temperature drift-sensitive process. Temperature drift sensitive processes are the processes most significantly affected by changes in mold temperature. The difference in sheet metal forming behavior between cold mold and thermal equilibrium states is the greatest.

[0060] The second step is to obtain the moment when the temperature rise rate first falls below the preset stable rate in the temperature rise curve of the temperature drift-sensitive process. The temperature rise rate is the first difference or first derivative of the temperature change curve, reflecting the rate at which the mold temperature rises per unit time. The calculation method involves discretizing the temperature change curve as a time series, calculating the temperature difference for each adjacent time interval, and dividing it by the time interval to obtain the temperature rise rate sequence. Starting from time zero, the sequence is scanned backward to find the moment when the temperature rise rate first falls below the preset stable rate. The preset stable rate is a small value close to zero, set according to the mold temperature measurement accuracy and thermal balance assessment standards, for example, 0.05 degrees Celsius per minute or 0.1 degrees Celsius per minute. The physical meaning of this moment is that the mold temperature has basically stabilized; although thereafter, the temperature may fluctuate slightly but no longer rises significantly, indicating that the mold has entered a thermal equilibrium state. This moment is the stable mold temperature moment for the temperature drift-sensitive process.

[0061] The third step involves using the thermal equilibrium temperature after the mold temperature stabilizes as the mold temperature boundary condition for the virtual perturbation analysis. In the simulation model, the temperature boundary condition for each process is uniformly set to the thermal equilibrium temperature after its respective mold temperature stabilizes. For temperature drift-sensitive processes, this temperature is the measured temperature value corresponding to its temperature stabilization moment. For other processes, if their temperature stabilization moment differs from that of the temperature drift-sensitive process, the temperature value corresponding to their respective temperature stabilization moment is used. If the temperature stabilization moment of a certain process is later than that of the temperature drift-sensitive process, for safety reasons, the stabilization moment of the temperature drift-sensitive process can be used as a unified time reference point, and each process uses its own temperature value corresponding to that moment.

[0062] The fourth step involves setting the temperature boundary conditions for each process mold, then applying virtual perturbations to each process sequentially from the first to the Nth stamping pass. Subsequent reverse-order virtual perturbations are also performed under the same temperature boundary conditions to ensure consistency in the temperature baseline between the forward and reverse analyses.

[0063] The technical effect and principle of this implementation method lies in objectively determining the start time of virtual disturbance analysis and the mold temperature boundary conditions through measured temperature data, thus avoiding the disconnect between the simulation state and the actual mass production state caused by setting the temperature based on experience. By using the temperature stabilization time of the temperature drift-sensitive process as a benchmark, it ensures that the virtual disturbance analysis of all processes is carried out under the condition that the most sensitive temperature drift factors have been eliminated, and the measured sensitivity coefficient is closer to the true value under the thermal equilibrium state of mass production.

[0064] In a preferred embodiment, before applying the virtual disturbance to each process sequentially from the first to the Nth stamping pass according to the stamping sequence, the method further includes the following steps: In the finite element simulation model, the initial stress state and initial temperature state of the mold for each process are set as the initial reference state; It also includes the following steps: After completing the virtual disturbance analysis for each process, the current stress state or current temperature state of the mold for that process is extracted from the simulation results, compared with the initial reference state, and the cumulative offset of the mold state is calculated. If the cumulative offset of the mold state is less than the preset offset threshold, the current state is used as the new initial reference state for subsequent processes to determine the cumulative offset of the mold state, and the virtual disturbance is applied to the next process.

[0065] Specifically, the forward-sequence virtual disturbance analysis requires applying disturbances sequentially to the first through Nth process and calculating the deviation changes. While this process doesn't involve the physical wear and tear of the physical mold in the simulation model, each virtual disturbance analysis leaves a record of the mold's stress and temperature states in the simulation model for that process. After the virtual disturbance analysis of a process is completed, the mold's state in the simulation model may have slightly deviated from the initial state. If this state difference is not monitored and controlled, and the simulation proceeds directly to the next process's disturbance analysis, the simulation results for subsequent processes will be calculated under mold conditions that have deviated from the initial state. As the number of test processes increases, the cumulative shift in the mold's state may cause significant differences between the disturbance analysis results of later processes and the results of independent testing under the initial state, resulting in a lack of a unified state benchmark for the sensitivity coefficients of each process. This solution achieves closed-loop monitoring of mold state drift during the simulation process by detecting the cumulative shift in the mold's state after each process's disturbance analysis and comparing it with the initial benchmark state. Within a controllable range of shift, the benchmark is updated and testing continues.

[0066] The specific operation steps of this implementation method are as follows.

[0067] The first step, before starting the forward-sequence virtual perturbation analysis, is to set the initial reference state of the mold for each process in the finite element simulation model. The initial reference state includes two dimensions: the initial stress state and the initial temperature state of the mold. The initial stress state of the mold refers to the stress distribution state of the mold when it is not subjected to stamping load, usually a zero-stress state, i.e., the mold is in a free state without preload. The initial temperature state of the mold refers to the temperature distribution state of the mold before stamping begins, usually the ambient temperature of the workshop or the temperature at thermal equilibrium. The initial reference state is set in the simulation software by applying initial conditions to the mold model. The temperature state is assigned in the form of nodal temperatures or element temperatures, and the stress state is zero by default. The initial reference state serves as a unified reference for comparing state drift in all subsequent processes.

[0068] The second step involves extracting the current stress or temperature state of the mold for each process after completing the virtual perturbation analysis. The virtual perturbation analysis process involves applying a perturbation to the target process, running a full-process simulation, and outputting the stress and temperature field results for each process mold after simulation convergence. The state parameters corresponding to the initial baseline state of the mold for that process are extracted from the results file. If the initial baseline state includes a stress state, the current stress state is extracted, specifically the maximum equivalent stress of the mold or the peak contact stress in key areas of the mold. If the initial baseline state includes a temperature state, the current temperature state is extracted, specifically the average surface temperature of the mold or the highest temperature of the mold. This extraction can be performed using the post-processing function of the simulation software or automatically read from the data fields in the results file using a script.

[0069] The third step is to compare the extracted current stress state or current temperature state with the initial reference state to calculate the cumulative offset of the mold state. The comparison method depends on the type of state extracted. If the current stress state is extracted, the current stress state value is subtracted from the initial reference stress state value, and the absolute value is taken as the stress offset. If the current temperature state is extracted, the current temperature state value is subtracted from the initial reference temperature state value, and the absolute value is taken as the temperature offset. If both stress and temperature states are extracted, the stress offset and temperature offset are each divided by their respective allowable tolerances, and the maximum value is taken, or they are combined into a single comprehensive offset using a preset weighting coefficient. The cumulative offset of the mold state characterizes the degree of deviation of the mold from the initial reference state at the current moment; the larger the value, the more severe the deviation.

[0070] The fourth step involves comparing the calculated cumulative offset of the mold state with a preset offset threshold. The preset offset threshold is an upper limit of allowable offset set based on the process sensitivity and simulation accuracy requirements of this step. If the stress offset is used as a comparison indicator, the offset threshold can be set as a preset percentage of the mold material's yield strength, such as 5%. If the temperature offset is used as a comparison indicator, the offset threshold can be set as a preset percentage of the thermal equilibrium temperature, such as 2% or 3% of Celsius. If the cumulative offset of the mold state is less than the preset offset threshold, it indicates that the current mold state is still within the acceptable initial state range. The current state is then used as the new initial reference state for subsequent steps to judge the cumulative offset of the mold state. That is, the current stress state value or temperature state value replaces the original initial reference value, and the offset calculation for subsequent steps is based on this new reference. This reference update operation is equivalent to acknowledging the small offset that has occurred in the mold state and resetting it to zero, avoiding unnecessary exceedances triggered by the accumulation of small offsets from multiple steps. Then, virtual perturbation is applied to the next step.

[0071] The technical effect and principle of this implementation method lies in establishing a closed loop for monitoring and controlling mold state drift during simulation. By extracting and comparing the state step by step, the cumulative shift of the mold state during virtual disturbance analysis can be detected in a timely manner. When the shift is controllable, the benchmark is updated and testing continues, which ensures the continuity of testing and avoids interference from the cumulative shift of multiple processes on the results of subsequent tests. The entire monitoring process is completed in the post-processing stage of the simulation software, without involving physical operations, and the implementation cost is zero.

[0072] In a preferred embodiment, after calculating the cumulative offset of the mold state, the method further includes the following step: If the cumulative offset of the mold state is greater than or equal to the preset offset threshold, the virtual disturbance of the subsequent process is paused, the mold state in the finite element simulation model is reset to the initial reference state, and the virtual disturbance of the subsequent process is continued after the reset.

[0073] Specifically, when the cumulative offset of the mold state is less than a preset offset threshold, the system updates the initial baseline state and continues testing for the next process. However, after the disturbance analysis for a certain process is completed, the cumulative offset of the mold state may exceed the preset offset threshold. This may occur when the disturbance amplitude of that process is large, or when the forming behavior of that process has a strong impact on the mold state. An offset exceeding the threshold means that the current mold state has significantly deviated from the initial baseline state. If the disturbance analysis for the next process continues based on this offset state, the accuracy of subsequent test results cannot be guaranteed. In this case, the mold state in the simulation model needs to be reset to restore it to the initial baseline state before subsequent tests can continue.

[0074] The specific operation steps of this implementation method are as follows.

[0075] The first step is to calculate the cumulative offset of the mold state and then compare it with a preset offset threshold. When the offset is greater than or equal to the preset offset threshold, the state reset process is triggered.

[0076] The second step is to pause the virtual disturbance of subsequent processes. Subsequent processes refer to all processes that follow the current process in the stamping sequence and have not yet undergone disturbance analysis. Pausing is achieved in the simulation software by stopping the batch processing task for subsequent processes or postponing the parameter disturbance settings for the next process. The purpose of pausing is to prevent the test from continuing under conditions where the mold state has deviated, thus rendering the data invalid.

[0077] The third step is to reset the mold state in the finite element simulation model to the initial baseline state. Specifically, in the simulation software, the current stress and temperature states of each mold process are overwritten with the previously set initial baseline stress and temperature states, respectively. For the stress state, the stress values ​​of all mold elements are cleared to zero or set to the initial zero-stress state. For the temperature state, the temperature values ​​of all mold nodes are reset to the initial thermal equilibrium temperature. This overwrite operation can be achieved through the simulation software's state reset function or by restarting the simulation analysis task. After resetting, the mold state is completely restored to the initial baseline state, perfectly consistent with the state before the start of the first process disturbance analysis.

[0078] Fourth, after the mold state is reset, continue with the virtual disturbance of subsequent processes. The tests for subsequent processes are performed sequentially from the paused position, following the previously described forward or reverse order. Since the mold state has been reset, the test results for subsequent processes will be consistent with the results of independent tests in the initial state, unaffected by previous tests.

[0079] The technical effect and principle of this implementation method lies in providing a direct solution to the problem of excessive cumulative deviation in mold state. In physical mold trials, exceeding the mold state requirements may require mold disassembly, waiting for cooling, or readjustment, which is time-consuming. In the simulation model, state reset only needs to restore the state variables to their initial values, which is completed instantly without additional cost. The reset mold state is completely consistent with the initial baseline state, ensuring a unified state baseline for subsequent process tests, and allowing the sensitivity coefficients of each process to be measured under comparable state conditions.

[0080] In a preferred embodiment, the step of verifying and correcting the boundary conditions of each process in the finite element simulation model using the measured force-displacement curve includes the following steps: The measured force-displacement curve and the simulated force-displacement curve output by the simulation model are segmented along the displacement axis using the same displacement interval division method; By comparing the deviations between the measured force values ​​and the simulated force values ​​within the same displacement range one by one, and based on the distribution of the deviations in each displacement range, the inaccuracy starting range where the simulation model begins to show significant deviations relative to the actual stamping process is located. Based on the aforementioned inaccuracy initiation interval, the boundary condition parameters of the finite element simulation model are corrected in segments.

[0081] Specifically, the simulation model is verified and corrected using measured force-displacement curves to ensure consistency between the output force-displacement curves of the simulation model and the measured ones. This step is a crucial bridge connecting physical mold testing and virtual simulation analysis, directly affecting the accuracy of subsequent virtual disturbance analysis results. Traditional verification and correction methods typically treat the entire force-displacement curve as a whole, calculating the overall deviation between the simulation curve and the measured curve across the entire displacement range, and then globally adjusting boundary condition parameters such as the friction coefficient and die clearance of the simulation model to minimize the overall deviation. However, this global correction method has a problem: the force-displacement curve of the stamping process has a clear temporal sequence. The small displacement segment in the early stage of the curve corresponds to the elastic deformation and initial yielding stage of the sheet metal, mainly affected by the die closing clearance and the initial contact state of the sheet metal. The large displacement segment in the middle and later stages of the curve corresponds to the stage of fully plastic flow of the sheet metal and elastic deformation of the die, mainly affected by the friction coefficient and material hardening behavior. If the simulation model matches the measured values ​​well in the early stage but deviates in the middle and later stages, the global correction will adjust the correct parameters from the early stage as well, causing the corrected model to become inaccurate in the early stage. This scheme divides the force-displacement curve into segments along the displacement axis, compares the deviation between the simulated and measured values ​​segment by segment, locates the inaccuracy starting interval where the simulation model begins to show significant deviation, and then performs segmented correction based on this interval, thus avoiding the problem of the accurate part in the early stage being misadjusted in the global correction.

[0082] The specific operation steps of this implementation method are as follows.

[0083] The first step is to segment the measured force-displacement curve and the simulated force-displacement curve output from the simulation model along the displacement axis using the same displacement interval division method. The displacement coordinate axis range of the two curves should be consistent, both extending from the initial displacement at mold closure to the maximum closed displacement. The displacement interval division method is equidistant segmentation, dividing the entire displacement range into several displacement intervals of equal length. The number of displacement intervals is determined based on the total length of the curve and the characteristic details, typically ranging from ten to twenty intervals. Too few intervals result in insufficient segmentation accuracy, making it impossible to accurately capture local deviations; too many intervals increase the computational load, and force fluctuations within short intervals may be amplified by noise. Each displacement interval has a starting displacement value and an ending displacement value.

[0084] The second step involves comparing the deviations of the measured force values ​​and the simulated force values ​​within the same displacement interval. The comparison is performed by calculating the arithmetic mean of the measured force-displacement curve within each interval, and simultaneously calculating the arithmetic mean of the simulated force-displacement curve within the same displacement interval. The deviation of the force value for that displacement interval is obtained by subtracting the mean of the measured force value from the mean of the simulated force value. The absolute value of this deviation is then compared with a preset verification threshold. The preset verification threshold is set based on the measurement accuracy of the force sensor and the expected fit of the simulation model, typically ranging from 1% to 3% of the maximum measured force value. This comparison is performed sequentially for each displacement interval in ascending order of displacement during the stamping process.

[0085] The third step involves locating the inaccuracy initiation interval based on the distribution of deviations across different displacement intervals. This inaccuracy begins when the simulation model starts to deviate significantly from the actual stamping process. The force deviations in each displacement interval are scanned, starting from the first interval and proceeding backwards. If the absolute value of the deviation in the preceding intervals is less than a preset verification threshold, it indicates that the simulation model matches the actual measurement well in the early stages of stamping. When a displacement interval is scanned and its absolute deviation first exceeds or equals the preset verification threshold, this interval is marked as the inaccuracy initiation interval. The physical meaning of the inaccuracy initiation interval is that, starting from the stamping process stage corresponding to this displacement interval, the behavior of the simulation model deviates significantly from the actual situation. If the absolute value of the deviation in all intervals across the entire displacement range is less than the preset verification threshold, it indicates that the simulation model does not require correction and the verification is successful.

[0086] The fourth step involves segmentally correcting the boundary condition parameters of the finite element simulation model based on the initial misalignment interval. The correction strategy involves adjusting the boundary condition parameters only for the portion of the simulation behavior following the displacement value corresponding to the initial misalignment interval, ensuring that the corrected simulated force-displacement curve matches the measured force-displacement curve within and after the initial misalignment interval. Boundary condition parameters include the friction coefficient between the mold and the sheet metal at each stage, and the mold's elastic modulus or coefficient of thermal expansion. Correction can be achieved by manually adjusting the parameters in the simulation software and then recalculating and comparing, or by using the simulation software's parameter optimization module for automatic iterative solution.

[0087] In a preferred embodiment, the step of segmentally correcting the boundary condition parameters of the finite element simulation model based on the initial misalignment interval includes the following steps: Using the displacement value corresponding to the inaccuracy initiation interval as the boundary, only the boundary condition parameters corresponding to the simulated force-displacement curve after that displacement value are corrected, so that the subsequent segments of the corrected simulated force-displacement curve are consistent with the measured force-displacement curve; the boundary condition parameters corresponding to the displacement interval before the inaccuracy initiation interval are not corrected.

[0088] Specifically, the specific operation steps of this implementation method are as follows.

[0089] The first step is to use the displacement value corresponding to the initial misalignment interval as the boundary. This displacement value is taken as the initial displacement value of the misalignment interval, that is, the left endpoint value of this interval on the displacement axis. This displacement value divides the entire stamping process into two stages: before and after this displacement value is the early stage of stamping, in which the simulation model's behavior is consistent with the actual measurement. After this displacement value is the middle and late stages of stamping, in which the simulation model's behavior deviates significantly from the actual measurement.

[0090] The second step involves correcting only the boundary condition parameters corresponding to the simulated force-displacement curves after the displacement value. Among the boundary condition parameters, those significantly affecting the shape of the force-displacement curves in the later stages of stamping include the friction coefficient between the die and the sheet metal in each process, the heat transfer coefficient between the sheet metal and the die, and the elastic deformation compensation coefficient of the die after being subjected to force. Parameters that significantly affect the shape of the force-displacement curves in the early stages of stamping but have a smaller impact on the later stages include the die closing clearance, the initial positioning deviation of the sheet metal, and the initial contact stiffness. Since the simulation and actual measurements match before the initial misalignment interval, the parameter settings for the early stages are correct and do not require adjustment. Therefore, only the friction coefficient and heat transfer coefficient, which have a significant impact on the behavior in the later stages, are selected as the parameters to be corrected, while parameters related to the early stages, such as the die closing clearance, remain unchanged.

[0091] The third step is to perform a correction to ensure that subsequent segments of the corrected simulated force-displacement curve are consistent with the measured force-displacement curve. This correction can be achieved in the simulation software by keeping the initial parameters unchanged in the simulation model's parameter settings, adjusting only parameters such as the friction coefficient in the later stages. Then, rerun the simulation calculation and output the corrected simulated force-displacement curve. Compare the corrected curve with the measured curve again, segment by segment. If the deviations in the initial misalignment interval and subsequent intervals have all fallen back within the preset verification threshold, the correction is complete. If there are still intervals with excessive deviations, continue fine-tuning the parameters and iterating until the consistency requirements are met. The boundary condition parameters corresponding to the displacement intervals before the initial misalignment interval are not corrected.

[0092] Fourth, after the correction is completed, the corrected simulation model can be used for subsequent virtual disturbance sensitivity analysis and tolerance allocation. Since the parameters were not changed in the early stage, the simulation accuracy of the corrected model is guaranteed throughout the stamping process.

[0093] The technical effect and principle of this implementation method lies in proposing an asymmetric segmented boundary condition correction strategy. This strategy uses the initial misalignment interval as the boundary, correcting only later parameters while freezing earlier parameters. The physical basis of this strategy is that the dominant physical factors differ at different stages of the stamping process, and the causes of early and later misalignments typically stem from different boundary condition parameters. Correcting only later parameters solves the misalignment problem while protecting the previously calibrated parts, avoiding the seesaw effect common in traditional global correction methods where correcting later parameters sacrifices earlier ones. This ensures that the final simulation model achieves high accuracy consistent with actual measurements across the entire displacement range.

[0094] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frames, characterized in that, Includes the following steps: A finite element simulation model of the entire stamping process of the seat frame is constructed. The finite element simulation model includes mold models and sheet metal models for each process. The material parameters of the sheet metal model are set based on the mechanical property test data of ultra-high strength steel. During the trial molding stage, the measured force-displacement curves of each process are obtained under the condition of no disturbance to the process parameters during actual stamping. The measured force-displacement curves are used to verify and correct the boundary conditions of each process in the finite element simulation model so that the force-displacement curves output by the simulation model are consistent with the measured force-displacement curves. In the modified finite element simulation model, under the condition that the process parameters of the remaining processes other than the target process remain unchanged, a virtual disturbance of a predetermined amplitude is applied to at least one key process parameter of the target process. The finite element simulation model calculates the change in the key dimension deviation of the final part, and uses the change in deviation as the initial sensitivity coefficient of the process to the final deformation. The process with the largest change in deviation among all processes is determined as the fluctuation-dominant process. Under the condition that the process parameters of all processes except the fluctuation-dominant process remain unchanged, two sets of virtual perturbations of different amplitudes are applied to the fluctuation-dominant process to obtain two process states of the fluctuation-dominant process; under each process state, the virtual perturbation of the predetermined amplitude is applied to each process other than the fluctuation-dominant process, and the final part critical dimension deviation change under each combination is calculated by the finite element simulation model; for the same process, the difference between the deviation changes obtained under the two process states is taken as the sensitivity offset of the process. The initial sensitivity coefficient is corrected based on the sensitivity offset of each process to obtain the corrected sensitivity coefficient of each process; The critical dimension tolerance values ​​of the final part are allocated to the process tolerances of each process according to the proportion of the correction sensitivity coefficient of each process to the total correction sensitivity coefficient. The process with the larger the correction sensitivity coefficient is, the larger the tolerance value is allocated.

2. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 1, characterized in that, Applying a virtual disturbance of a predetermined magnitude to at least one key process parameter of the target process includes the following steps: From the corrected finite element simulation model, extract the force-displacement curve of the target process under the condition of no disturbance process parameters, and calculate the plastic deformation work done by the target process on the sheet metal based on the force-displacement curve; A preset plastic work disturbance value is set, which is not greater than a preset proportion of the plastic deformation work; Using the plastic work disturbance value as an increment, and combining the characteristics of the force-displacement curve, the change in key process parameters required to increase the plastic work disturbance value is calculated in reverse, and this change is taken as the predetermined amplitude.

3. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 1, characterized in that, In the revised finite element simulation model, the virtual disturbance is applied to each process step and the deviation change is calculated, including the following steps: All processes are arranged in the stamping sequence as the first to the Nth process, where N is an integer greater than or equal to 2; First, according to the stamping sequence from the first to the Nth, the virtual disturbance is applied to each process in turn. The deviation change of the key dimensions of the final part is calculated by the finite element simulation model, and the first set of deviation changes of each process in the sequentially increasing direction is obtained. Then, following the reverse stamping sequence from the Nth to the first, the virtual disturbance is applied to each process sequentially. The finite element simulation model is used to calculate the deviation changes of the key dimensions of the final part, and the second set of deviation changes of each process in the sequential decreasing direction is obtained. The step of using the change in deviation as the initial sensitivity coefficient of the process to the final deformation includes the following steps: For each process, the average of the first set of deviation changes and the second set of deviation changes is used as the initial sensitivity coefficient of that process to the final deformation.

4. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 1, characterized in that, Before applying two sets of virtual disturbances of different amplitudes to the fluctuation-dominant process, the following steps are also included: Obtain the upper and lower limits of the key process parameters of the process with the most fluctuating parameters in historical qualified production batches. The upper and lower limits of the parameter values ​​define the normal production allowable range of the key process parameters. The normal production allowable range is divided into two non-overlapping sub-intervals, with the first sub-interval closer to the upper limit of the parameter value and the second sub-interval closer to the lower limit of the parameter value. Among the two sets of virtual disturbances of different amplitudes, one set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the first sub-interval, and the other set of disturbance amplitudes is set to values ​​that make the key process parameter fall within the second sub-interval.

5. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 3, characterized in that, Obtaining the first set of deviation changes in the sequentially increasing direction and the second set of deviation changes in the sequentially decreasing direction for each process includes the following steps: For any direction between the sequentially increasing direction and the sequentially decreasing direction, take it as the current test direction; When applying the virtual disturbance to the target process, in the finite element simulation model, the random fluctuation range of the sheet material parameters and / or boundary conditions is set, at least two virtual sheet samples are randomly generated and the deviation values ​​of the key dimensions of the final part are calculated respectively. The arithmetic mean of these deviation values ​​is taken as the average value of the disturbance state deviation of the process in the current test direction. Under undisturbed conditions, at least two virtual sheet metal samples are randomly generated and the arithmetic mean of the final critical dimension deviation values ​​of the parts is calculated as the undisturbed deviation benchmark value. The difference between the mean deviation of the disturbance state of the process and the reference value of the undisturbed deviation is taken as the deviation change of the process in the current test direction.

6. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 3, characterized in that, Before applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage, the process further includes the following steps: Obtain the measured temperature change curves of the mold from a cold state to a thermal equilibrium state during continuous production of each process. Based on the measured temperature change curves, determine the temperature rise amplitude of each process during continuous production. Mark the process with the largest temperature rise amplitude among all processes as a temperature drift sensitive process. The moment when the temperature rise rate first falls below the preset steady-state rate in the temperature rise curve of the temperature drift-sensitive process is taken as the moment when the mold temperature of the temperature drift-sensitive process stabilizes. The process of applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage includes setting the temperature boundary conditions of the mold for each process in the finite element simulation model to the thermal equilibrium temperature after the mold temperature stabilizes.

7. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 3, characterized in that, Before applying the virtual disturbance to each process sequentially from the first to the Nth stamping stage, the process further includes the following steps: In the finite element simulation model, the initial stress state and initial temperature state of the mold for each process are set as the initial reference state; It also includes the following steps: After completing the virtual disturbance analysis for each process, the current stress state or current temperature state of the mold for that process is extracted from the simulation results, compared with the initial reference state, and the cumulative offset of the mold state is calculated. If the cumulative offset of the mold state is less than the preset offset threshold, the current state is used as the new initial reference state for subsequent processes to determine the cumulative offset of the mold state, and the virtual disturbance is applied to the next process.

8. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 7, characterized in that, After calculating the cumulative offset of the mold state, the following steps are also included: If the cumulative offset of the mold state is greater than or equal to the preset offset threshold, the virtual disturbance of the subsequent process is paused, the mold state in the finite element simulation model is reset to the initial reference state, and the virtual disturbance of the subsequent process is continued after the reset.

9. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 1, characterized in that, The process of verifying and correcting the boundary conditions of each process in the finite element simulation model using the measured force-displacement curve includes the following steps: The measured force-displacement curve and the simulated force-displacement curve output by the simulation model are segmented along the displacement axis using the same displacement interval division method; By comparing the deviations between the measured force values ​​and the simulated force values ​​within the same displacement range one by one, and based on the distribution of the deviations in each displacement range, the inaccuracy starting range where the simulation model begins to show significant deviations relative to the actual stamping process is located. Based on the aforementioned inaccuracy initiation interval, the boundary condition parameters of the finite element simulation model are corrected in segments.

10. The method for full-process deformation accumulation simulation and tolerance allocation of ultra-high strength steel seat frame according to claim 9, characterized in that, The step of segmentally correcting the boundary condition parameters of the finite element simulation model based on the initial misalignment interval includes the following steps: Using the displacement value corresponding to the inaccuracy initiation interval as the boundary, only the boundary condition parameters corresponding to the simulated force-displacement curve after that displacement value are corrected, so that the subsequent segments of the corrected simulated force-displacement curve are consistent with the measured force-displacement curve; the boundary condition parameters corresponding to the displacement interval before the inaccuracy initiation interval are not corrected.