Vibration aging control method for large machine tool structural component
By employing a vibration aging method based on multiphysics coupling modeling and adaptive intelligent control, the problem of residual stress control in large machine tool structural components was solved, thereby improving the stress relief rate and extending the component life.
Patent Information
- Application Number
- CN202511351103.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-22
- Publication Date
- 2026-01-23
AI Technical Summary
Existing vibration aging processes cannot adapt to the complex and varied residual stress distribution of large machine tool structural components. They lack accurate simulation of the multi-physics field coupling effect during the casting process and cannot monitor the vibration state in real time, resulting in low stress relief efficiency and damage risk.
By employing multi-physics coupling modeling combined with measured data, the simulated residual stress field is iteratively corrected using the Kriging interpolation method to identify stress concentration areas, perform dynamic modal analysis, design a graded frequency sweep excitation mode, and use an acceleration sensor network for real-time monitoring. Adaptive control technology is then used to dynamically adjust the exciter parameters.
It achieves precise control of residual stress in large machine tool structural components, significantly improves stress relief rate, inhibits aging deformation, extends component life, and avoids the propagation of microcracks.
Smart Images
Figure CN121386599A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field related to mechanical manufacturing, and more particularly, to a vibration aging control method for a large machine tool structural member. BACKGROUND
[0002] The structural members such as cross beams and columns are the core load-bearing components of large machine tools, and their performance directly determines the precision and stability of the machine tools. Such structural members are mostly formed by casting process, but during the casting process, the metal is unevenly cooled and solidified, inevitably generating large residual stress. The residual stress not only affects the dimensional accuracy and mechanical properties of the components, but also may induce the propagation of micro cracks, and even cause aging deformation, which seriously restricts the machining precision and service life of large machine tools.
[0003] The vibration aging process is widely used in the field of mechanical manufacturing due to its efficient and environmentally friendly characteristics, and through periodic dynamic load, it promotes the redistribution of internal residual stress of the component. However, the traditional vibration aging process has many limitations and cannot meet the needs of modern high-end equipment manufacturing, and the specific reasons are as follows: 1. Fixed vibration frequency, amplitude and loading mode cannot adapt to the complex and variable residual stress distribution of large machine tool structural members, resulting in difficulty in covering key parts for stress relief; 2. Lack of precise simulation of the formation mechanism of residual stress under the coupling action of thermal field, flow field, stress field and other multi-physical fields during the casting process, making the process parameter setting lack scientific basis; 3. Without dynamic control capability, it cannot adjust the process parameters in time in the face of changes in residual stress caused by casting process fluctuations and material property differences, and the stress relief efficiency and effect are not good; 4. Cannot real-time monitor the vibration state of the structural member, making it difficult to ensure effective use of vibration energy, and there is a risk of new damage caused by unreasonable vibration.
[0004] With the increasing demand for high precision and high reliability of large machine tool structural members in modern high-end equipment manufacturing, it has become a key technical problem to be solved in the field of mechanical manufacturing to develop a method that can deeply optimize the vibration aging process and achieve precise control of residual stress of large machine tool structural members. SUMMARY
[0005] In view of the above defects or improvement needs of the prior art, the present application provides a vibration aging control method for a large machine tool structural member, which aims to solve the problem of limitations of the existing vibration aging process in dealing with residual stress of large machine tool structural members.
[0006] To achieve the above-mentioned purpose, according to one aspect of the present application, a vibration aging control method for a large machine tool structural member is provided, which comprises the following steps: S1. Construct a multi-physics coupled model of residual stress in the casting of machine tool structural parts, and then establish a simulation residual stress field for the filling-solidification process. S2 uses the Kriging interpolation method and combines measured stress data to iteratively correct the simulated residual stress field. S3, based on the corrected simulated residual stress field, identifies stress concentration areas and stress change areas of machine tool structural components through stress gradient analysis to obtain the stress gradient distribution; S4. Perform dynamic modal analysis on machine tool structural components to determine the dynamic sensitive frequency band; S5, based on stress gradient distribution and dynamic sensitive frequency band, designs a graded frequency sweep excitation mode and generates the initial control curve of the vibration aging excitation system; S6, during the operation of the vibration aging excitation system, uses an acceleration sensor network to realize real-time monitoring of the vibration mode of the machine tool structural components; S7 calculates the mode shape fit based on the modal confidence criterion, and then dynamically adjusts the output process parameters of the exciter based on the obtained mode shape fit using adaptive control technology based on a combination of prediction and fuzzy control, thereby realizing adaptive intelligent control of vibration aging.
[0007] Furthermore, step S1 employs a phased multiphysics coupling simulation strategy, specifically as follows: (a) Simulation stage of filling process: Focus on dynamic simulation of temperature field and flow field, ignore stress field calculation; (b) Solidification process simulation stage: Using the temperature field and flow field data when the filling is completed as the initial boundary conditions, a coupled simulation model of temperature field-flow field-stress field is constructed; (c) Simulation stage of natural cooling process: When the temperature of the casting system drops to the temperature set by the unpacking process, the convection and radiation heat exchange and stress relaxation behavior of the machine tool structural parts and the environment are simulated until the temperature drops to room temperature, forming a complete simulation process of casting stress evolution.
[0008] Further, step S2 includes the following sub-steps: S21: Using a blind hole stress tester and a residual stress ultrasonic non-destructive testing instrument, the surface and internal parts of the cast machine tool structural parts are measured at multiple locations to obtain stress data covering different depths and key parts. S22: By comparing the measured and simulated residual stress fields, the Kriging interpolation method is used to quantify the differences, and a spatial correlation model is constructed to optimize the parameters of the residual stress distribution field.
[0009] Further, the Kriging interpolation method combines the measured stress data to iteratively correct the residual stress distribution model of the simulation residual stress field, and the steps are as follows: a, calculating the measured and simulated stress ratio to construct a correction coefficient field; b, using the Gaussian variation function to establish the spatial correlation between the measuring points; c, solving the optimal weight equation to determine the influence degree of each measuring point on the prediction point; d, multiplying the full-field stress based on the optimal weight.
[0010] Further, the correction coefficient calculation formula is:
[0011] is the correction coefficient of the i th measuring point; is the measured stress value of the i th measuring point; is the simulated stress value of the i th measuring point; The Gaussian variation function is:
[0012] describes the change rule of the correlation between two points in space with the distance; is the Euclidean distance between two points; is the range parameter, which controls the decay rate of correlation; The weight equation is:
[0013] is the covariance matrix between the measuring points, ; is the weight vector to be solved, which determines the influence degree of each measuring point on the simulation prediction point; is the covariance vector of the prediction point and all measuring points, ; The full-field stress correction formula is:
[0014] is the corrected full-field stress value; is the correction coefficient of the prediction point, which is obtained by weighted average of the measuring point correction coefficients; is the original simulation stress field.
[0015] Further, step S3 includes the following sub-steps: S31: using a numerical differential algorithm to calculate the stress gradient of each part of the machine tool structure, and analyzing the stress concentration area and the stress change area; S32: visualizing the stress gradient distribution through a color cloud chart, marking the key processing parts, and providing a targeted basis for subsequent sweep mode design Further, the stress gradient calculation formula is:
[0016] Wherein, x, y, z are X, Y, Z axis direction coordinates; , , X, Y, Z three direction stress components.
[0017] Further, step S4 comprises the following steps: S41: Considering the fixed constraint condition and environmental temperature of the machine tool structure during the vibration aging process, the modal analysis of the machine tool structure is carried out to analyze the natural frequency and vibration mode of the machine tool structure; S42: The actual natural frequency of the machine tool structure is detected by professional modal analysis equipment, and the sensitive frequency band is identified.
[0018] Further, step S5 is specifically: S51: In the dynamic sensitive frequency band, full frequency sweep is carried out with fixed frequency step, the vibration response of the machine tool structure under each frequency is monitored in real time by means of acceleration sensor network, and the change of vibration characteristic of the machine tool structure is analyzed; S52: For the stress concentration area, the exciter sweep step is refined in the specific natural frequency interval, the dynamic response characteristic of the machine tool structure in the stress concentration area is captured, and the optimal excitation frequency is identified through the judgment of vibration response; S53: Based on the vibration response data collected by sweep, the optimization goal of maximizing elimination efficiency is used, the genetic algorithm is used for parameter optimization, the initial control curve of frequency-time (f-t) and excitation force-time (F-t) is generated, and the optimal frequency and excitation force parameters at each time are determined.
[0019] Further, step S7 comprises the following steps: S71: Based on modal assurance criterion (MAC), the coincidence degree of the measured vibration mode of the machine tool structure and the theoretical vibration mode is calculated quantitatively, and whether the actual vibration state of the machine tool structure is consistent with the theoretical calculation vibration mode is evaluated; S72: The adaptive control technology of fusion of prediction and fuzzy control is adopted, the error between the vibration mode coincidence degree and the target value is calculated, and the two-dimensional fuzzy controller is input, the excitation process parameters of the exciter are output through fuzzy rule reasoning, and the dynamic adjustment is carried out by using coarse and fine control strategy, the exciter is driven to adaptively adjust the operation parameters, the vibration mode coincidence degree reaches the set standard, the adaptive intelligent control of vibration aging process is realized; The modal assurance criterion (MAC) is:
[0020] Wherein, , is two vibration mode vectors to be compared, is a measured mode shape, is a theoretical mode shape.
[0021] Overall, compared with the prior art, the vibration aging control method of the large machine tool structural part provided by the present application mainly has the following beneficial effects: 1. The present application uses multi-physical field coupling modeling combined with measured data iterative optimization to accurately invert the residual stress distribution, providing quantitative basis for vibration aging parameter design, and the stress relief rate is significantly higher than that of traditional process level.
[0022] 2. The present application is based on stress gradient analysis and hierarchical frequency scanning strategy of dynamic sensitive frequency band, realizes accurate targeted application of vibration energy, effectively suppresses the aging deformation of large components, and greatly improves the dimensional stability of machine tool structural parts.
[0023] 3. The present application uses adaptive intelligent control technology to dynamically adjust the excitation parameters in real time, promotes the vibration aging from "experience driven" optimization to "model driven", avoids the expansion of micro cracks caused by poor vibration matching, and significantly prolongs the fatigue life of the component. BRIEF DESCRIPTION OF DRAWINGS
[0024] Figure 1 is a flow chart of a vibration aging control method of a large machine tool structural part provided by an embodiment of the present application; Figure 2 is a phased flow chart of a multi-physical field coupling residual stress simulation strategy of an embodiment of the present application; Figure 3 is a schematic diagram of a geometric model of a large machine tool structural part of an embodiment of the present application. DETAILED DESCRIPTION
[0025] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.
[0026] Please refer to Figure 1 , the present application provides a vibration aging control method of a large machine tool structural part, which significantly improves the elimination effect of residual stress, effectively suppresses the expansion of micro cracks, solves the aging deformation problem of large castings, and greatly improves the dimensional stability and fatigue life of large machine tool structural parts by using multi-physical field coupling modeling and adaptive intelligent control.
[0027] The control method comprises the following steps: S1, a multi-physical field coupling model of casting residual stress of the machine tool structure part is constructed, and then a simulation residual stress field of the filling-solidification process is established.
[0028] Wherein, referring to Figure 3 , comprises the following specific steps: S11: three-dimensional modeling of the machine tool structure part casting system is performed; S12: finite element meshing of the casting system is completed; S13: material thermal physical property parameters, pouring temperature, boundary conditions and other key parameters are set.
[0029] Step S1 adopts a multi-stage multi-physical field coupling simulation strategy, specifically: (a) filling process simulation stage: dynamic simulation of the temperature field and flow field of the casting system is carried out.
[0030] (b) solidification process simulation stage: the temperature field and flow field data at the time of filling completion (100% filling rate) are extracted as the initial boundary conditions of the solidification stage, and a temperature field-flow field-stress field coupling simulation model, i.e. a multi-physical field coupling model, is constructed.
[0031] (c) natural cooling process simulation stage: when the temperature of the casting system drops to the opening box process setting temperature, based on the multi-physical field coupling simulation model, the convection, radiation heat exchange and stress relaxation behavior of the component and the environment are simulated, until the temperature drops to room temperature, forming a complete casting stress evolution simulation process.
[0032] In one embodiment, step S1 comprises the following steps: S11: a professional CAD software is used to construct a three-dimensional model of the machine tool structure part casting system, and the geometric shape and layout are accurately restored. Based on the casting process requirements, the three-dimensional model is simplified, including parameterization simplification of chamfer, fillet and other non-key features, to form a geometric model meeting the simulation requirements.
[0033] S12: the three-dimensional model is exported to the ProCAST simulation platform in IGES format, a sand box model is constructed and geometric repair is performed to eliminate intersection and overlapping areas. Tetrahedral mesh is used for finite element discretization of the machine tool structure part, the pouring system, the riser and the sand box, to generate a high-quality mesh model.
[0034] S13: the pouring direction and gravity acceleration of gravity casting are set, the material thermal physical property parameters of each component are defined, and the boundary conditions such as interface heat transfer coefficient, cooling condition, pouring temperature are configured. Through the multi-physical field coupling solver of ProCAST, the whole process of filling-solidification-cooling is simulated, and the temperature field and stress field are obtained.
[0035] The embodiment adopts a multi-physics field coupling simulation strategy (as shown in Figure 2 Specifically, the simulation strategy includes the following stages: (a) Simulation stage of filling process: focus on dynamic simulation of temperature field and flow field, and ignore stress field calculation. Since the shear modulus of liquid metal approaches zero, it cannot form an effective stress bearing structure. This stage only calculates the flow path of metal melt and the temperature distribution, providing initial data for the subsequent solidification stage.
[0036] (b) Simulation stage of solidification process: use the temperature field and flow field data at the completion of filling (100% filling rate) as the initial boundary conditions to build a temperature field-flow field-stress field coupling simulation model. Through the multi-physics field coupling model, the evolution law of dendrite growth, shrinkage hole and residual stress in the machine tool structure during casting process is revealed, and the influence of solid-state shrinkage and latent heat release on residual stress is focused on.
[0037] (c) Simulation stage of natural cooling process: when the temperature of the casting system drops to the opening box process setting temperature (such as 300℃), simulate the convective and radiative heat exchange between the component and the environment, as well as the stress relaxation behavior, until the temperature drops to room temperature, forming a complete casting stress evolution simulation process. This strategy can improve the calculation efficiency while ensuring the accuracy by eliminating secondary factors.
[0038] S2, using Kriging interpolation method, combining with the measured stress data to iteratively correct the simulated residual stress field.
[0039] Step S2 includes the following sub-steps: S21: Use blind hole method stress tester, residual stress ultrasonic nondestructive testing instrument and other equipment to measure the surface and internal multiple positions of the machine tool structure after casting, obtain stress data covering different depths and key positions, and lay a data foundation for model correction; S22: Compare the measured and simulated residual stress fields, use the Kriging interpolation method to quantify the differences, build a spatial correlation model to optimize the residual stress distribution field parameters, so that the simulated residual stress field accurately fits the actual stress state, and improves the model accuracy and accuracy.
[0040] The Kriging interpolation method combines measured stress data to iteratively correct the residual stress distribution model of the simulated residual stress field, which includes the following steps: calculate the ratio of measured and simulated stress, build a correction coefficient field; use Gaussian variation function to establish the spatial correlation between measurement points; solve the optimal weight equation to determine the influence degree of each measurement point on the prediction point; multiply the full-field stress based on the optimal weight.
[0041] Specifically: 1. Calculate the ratio of measured and simulated stress, build a correction coefficient, quantify the difference between the two, and the correction coefficient calculation formula is:
[0042] wherein, is the correction coefficient of the i-th measurement point; is the measured stress value of the i-th measurement point; is the simulated stress value of the i-th measurement point. If , it indicates that the simulation underestimates the stress; if , it indicates that the simulation overestimates the stress.
[0043] 2. The stress-space correlation between each measurement point is established by using a Gaussian variation function to describe the stress variation law, and the Gaussian variation function is:
[0044] wherein, describes the variation law of the correlation between two points in space with distance; is the Euclidean distance between two points; is the range parameter, which controls the decay rate of the correlation. The influence degree between measurement points is quantified, and the closer the measurement points, the more similar the correction coefficients.
[0045] 3. The optimal stress correction weight equation is solved to determine the influence degree of each measurement point on the simulation prediction point, and the Kriging weight equation is:
[0046] wherein, is the covariance matrix between measurement points, ; is the weight vector to be solved, which determines the influence degree of each measurement point on the simulation prediction point; is the covariance vector of the prediction point and all measurement points, . By solving the optimal stress correction weight equation, the optimal weight is obtained, so that the correction coefficient field is smoothly transitioned in space.
[0047] 4. The full-field stress is multiplied based on the optimal stress correction weight, and the full-field stress correction formula is:
[0048] wherein, is the corrected full-field stress value; is the correction coefficient at the prediction point, which is obtained by weighted average of the measurement point correction coefficients; is the original simulation stress field. By spatial interpolation to calculate the correction coefficient of each point, and then multiplied by the simulation stress, the corrected stress field closer to the measured value is obtained.
[0049] In one specific embodiment, step S2 comprises the following steps: S21: Using blind hole method stress tester and residual stress ultrasonic nondestructive testing instrument, the residual stress of different positions on the surface (measuring points are arranged at intervals of 1-2 m) and inside (detected by drilling layer by layer) of the actual casting completed machine tool structure is measured to obtain comprehensive data containing X, Y and Z direction stress components.
[0050] S22: Comparing the measured data with the simulation stress field, the Kriging interpolation method is used to correct the simulation residual stress field model value. The specific steps are as follows: 1. Through the measured stress and the simulation stress ratio to calculate the correction coefficient . If , it means that the simulation underestimates the stress; if , it means that the simulation overestimates the stress.
[0051] 2. The stress-space correlation between each measured point is established by using the Gaussian variation function . Among them, describes the change law of the correlation of the stress between two points in space with the distance; is the Euclidean distance between two points; is the range parameter, which controls the decay rate of the correlation. This function is used to quantify the influence degree between the measuring points. The closer the measuring points, the more similar the correction coefficients.
[0052] 3. The weight vector is solved by the Kriging weight equation to ensure smooth transition of the correction coefficient field. Among them, is the covariance matrix between the measuring points, ; is the weight vector to be solved, which determines the influence degree of each measuring point on the prediction point; is the covariance vector of the prediction point and all measuring points, . By solving the optimal stress correction weight equation, the optimal weight is obtained, so that the correction coefficient field is smoothly transitioned in space.
[0053] 4. The corrected stress field is calculated by the full-field stress correction formula to improve the consistency between the simulation results and the measured data. Among them, is the corrected full-field stress value; is the correction coefficient at the prediction point, which is obtained by weighted average of the measuring point correction coefficients; is the original simulation stress field. By spatial interpolation to calculate the correction coefficient of each point, and then multiplied by the simulation stress, the corrected stress field closer to the measured value is obtained.
[0054] S3, based on the corrected simulation residual stress field, identifying stress concentration areas and stress change areas of the machine tool structure through stress gradient analysis to obtain stress gradient distribution.
[0055] Step S3 includes the following sub-steps: S31: using numerical differentiation algorithm to calculate the stress gradient of each part of the machine tool structure, analyzing the stress concentration area and the stress change area; S32: visualizing the stress gradient distribution through color cloud chart, marking the key processing parts, and providing targeted basis for subsequent sweep mode design.
[0056] The stress gradient calculation formula is:
[0057] Wherein, x, y, z are the X, Y, Z axis direction coordinates; , , X, Y, Z three direction stress components.
[0058] In one embodiment, step S3 includes the following steps: S31: using numerical differentiation algorithm to calculate the stress gradient of each part of the structure , setting the gradient threshold (such as 100 MPa / mm) to identify high stress concentration area and change area.
[0059] S32: visualizing the stress gradient distribution through color cloud chart, providing targeted basis for subsequent sweep mode design.
[0060] S4, carrying out dynamic modal analysis on the machine tool structure to determine the dynamic sensitive frequency band.
[0061] Step S4 is specifically: S41: considering the fixed constraint conditions and environmental temperature of the machine tool structure during vibration aging process, carrying out modal analysis of the machine tool structure to analyze the natural frequency and vibration mode of the machine tool structure; S42: detecting the actual natural frequency of the machine tool structure through professional modal analysis equipment, identifying the sensitive frequency band.
[0062] In one embodiment, step S4 includes the following steps: S41: using Abaqus or Ansys and other finite element software to establish the dynamic model of the machine tool structure, setting the actual constraint conditions such as bolt fixing, support pad, considering the influence of temperature field on the elastic modulus and Poisson's ratio of the material, calculating the natural frequency and vibration mode of the machine tool structure, and obtaining the theoretical vibration characteristics.
[0063] S42: Constructing an experimental modal testing system with the shaker and the acceleration sensor network, obtaining the frequency response function curve through the hammering method or the steady-state sinusoidal excitation, and determining the sensitive frequency band (usually the inherent frequency ± 10% range) combined with the finite element calculation result to form a vibration parameter database combining theory and experiment.
[0064] S5, designing a hierarchical swept frequency excitation mode based on the stress gradient distribution and the dynamic sensitive frequency band, and generating an initial control curve of the vibration aging excitation system.
[0065] Step S5 is specifically: S51: In the dynamic sensitive frequency band, full-range swept frequency is carried out with a fixed frequency step, the vibration response of the machine tool structural member at each frequency is monitored in real time by means of the acceleration sensor network, and the change of the vibration characteristic of the machine tool structural member is analyzed; S52: For the stress concentration area, the swept frequency step of the shaker is refined in a specific inherent frequency interval, the dynamic response characteristic of the machine tool structural member in the stress concentration area is accurately captured, and the optimal excitation frequency is identified through the judgment of the vibration response; S53: Based on the vibration response data collected by swept frequency, the initial control curve of frequency-time (f-t) and excitation force-time (F-t) is generated by using genetic algorithm for parameter optimization with the optimization target of maximum elimination efficiency, and the optimal frequency and excitation force parameters at each time are accurately determined.
[0066] In one specific embodiment, step S5 includes the following steps: S51: Full-range swept frequency is carried out with a step of 1 Hz in the sensitive frequency band, the vibration response at each frequency point is monitored in real time by means of the acceleration sensor network, the vibration mode change characteristic of the structural member is obtained, and a global vibration characteristic map is formed.
[0067] S52: For the high stress concentration area, the swept frequency step is refined to 0.1 Hz in the ± 10 Hz sub-interval of the sensitive frequency band, and the dynamic response peak value is accurately captured through high-density frequency sampling.
[0068] S53: Based on the swept frequency data, a residual stress elimination efficiency evaluation model is constructed, the optimal shaker process parameter combination is solved by using genetic algorithm with the optimization target of maximum elimination efficiency, and a vibration aging control curve is generated: f-t frequency-time curve: 5-7 swept frequency stages are divided, the frequency range and residence time of each stage are dynamically adjusted according to the stress gradient distribution, and intelligent allocation of frequency resources is realized; F-t excitation force-time curve: combined with the mass (10-50 t) of the structural member and the modal response amplitude, the excitation force amplitude interval of 5-50 kN is set, the excitation force is smoothly adjusted through piecewise linear interpolation, and it is ensured that the vibration energy is targeted to act on the high stress area.
[0069] S6, in the working process of the vibration aging excitation system, an acceleration sensor network is used to realize real-time monitoring of the vibration mode of the machine tool structure.
[0070] Step S6 is specifically: S61: based on the results of finite element modal analysis, accurately positioning the stress concentration area and other key positions of the machine tool structure, and formulating an optimal arrangement scheme for the acceleration sensor network, to ensure that the sensors cover all sensitive modes and key positions; S62: rapidly transmitting the collected signals to a data processing unit, using professional signal processing algorithms such as Fourier transform and low-pass filter, and combining ERA algorithm to identify modal parameters such as natural frequency and vibration mode of the machine tool structure, to provide data support for subsequent process control.
[0071] In one specific embodiment, step S6 includes the following steps: S61: based on the results of finite element modal analysis, 5-10 triaxial acceleration sensors are deployed in the area with the largest stress gradient of the structure, forming a spatial orthogonal monitoring matrix, to ensure that the sensors cover all sensitive modes.
[0072] S62: the collected time domain signals are filtered by a low-pass filter (cutoff frequency 1000Hz) to eliminate noise interference, and after being converted to the frequency domain by FFT transform, the ERA algorithm is combined to identify modal parameters such as natural frequency and vibration mode of the machine tool structure.
[0073] S7, according to the modal confidence criterion to calculate the vibration mode coincidence degree, and then based on the obtained vibration mode coincidence degree, using the adaptive control technology based on the combination of prediction and fuzzy control to dynamically adjust the output process parameters of the exciter, to realize adaptive intelligent control of vibration aging.
[0074] Step S7 is specifically: S71: based on the modal confidence criterion (MAC), the coincidence degree of the measured vibration mode of the structure and the theoretical vibration mode is quantitatively calculated, to evaluate whether the actual vibration state of the machine tool structure is consistent with the theoretical calculated vibration mode; S72: using the adaptive control technology combining prediction and fuzzy control, taking the vibration mode coincidence degree as the input variable, calculating the error with the target value, and inputting it into a two-dimensional fuzzy controller, through fuzzy rule reasoning, outputting the excitation process parameters of the exciter such as excitation frequency, and using coarse and fine control strategy for dynamic adjustment, to drive the exciter to adaptively adjust the operating parameters, to ensure that the vibration mode coincidence degree reaches the set standard, to realize adaptive intelligent control of the vibration aging process.
[0075] The modal confidence criterion (MAC) is:
[0076] wherein, , is the measured mode shape vector, is the measured mode shape, is the theoretical mode shape. When the modal assurance criterion MAC=1, it means that the two mode shapes are identical; when the modal assurance criterion MAC=0, it means that the two mode shapes are orthogonal and independent.
[0077] Preferably, the two-dimensional fuzzy controller takes the mode shape coincidence error and its rate of change as input variables, and through a preset fuzzy rule base, the Mamdani reasoning is carried out, and the weighted average method is combined to solve the defuzzification to output the excitation frequency adjustment amount and the force adjustment amount ; in combination with the coarse-fine control strategy, when , a large step size coefficient is adopted for fast adjustment, when , a small step size coefficient is switched to for fine tuning, and finally the dynamic adaptive optimization of the excitation parameters is realized to ensure that the mode shape coincidence error meets the standard and to improve the residual stress relief efficiency.
[0078] The fuzzy reasoning rules are:
[0079] The weighted average method is:
[0080] The coarse-fine control strategy is specifically: when , ; when , . Wherein, and are control coefficients, ; and are threshold parameters.
[0081] In one specific embodiment, step S7 includes the following steps: S71: Based on the modal assurance criterion (MAC), the coincidence of the measured mode shape of the structure and the theoretical mode shape is quantitatively calculated, and the normalized inner product of the measured mode shape and the theoretical mode shape is calculated. The MAC value sequence is calculated in real time through a sliding window, and the mean and standard deviation are extracted as vibration state characteristic indexes.
[0082] S72: Design a prediction-fuzzy fusion adaptive controller to realize three-stage regulation: (a) Error modeling: define the mode shape coincidence error and the error rate of change , wherein For sampling interval, , construct a two-dimensional input space (X, Y) E , EC ); (b) Fuzzy inference: use Mamdani type fuzzy controller, define fuzzy subsets NL (negative large), NM (negative medium), NS (negative small), ZE (zero), PS (positive small), PM (positive medium), PL (positive large), design a 7x7 fuzzy rule table, as shown in Table 1.
[0083] Table 1 Fuzzy rule
[0084] (c) Coarse and fine dual-mode regulation: set threshold (coarse adjustment), (fine adjustment), dynamically switch control coefficients. When , enable large step coefficient fast correction ; when , switch to small step coefficient fine adjustment, to ensure that the MAC value converges to the target interval (X, Y) ).
[0085] Those skilled in the art will readily understand that the above description is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A vibration aging control method of a large machine tool structural member, characterized by, The control method includes the following steps: S1. Construct a multi-physics coupled model of residual stress in the casting of machine tool structural parts, and then establish a simulation residual stress field for the filling-solidification process. S2 uses the Kriging interpolation method and combines measured stress data to iteratively correct the simulated residual stress field. S3, based on the corrected simulated residual stress field, identifies stress concentration areas and stress change areas of machine tool structural components through stress gradient analysis to obtain the stress gradient distribution; S4. Perform dynamic modal analysis on machine tool structural components to determine the dynamic sensitive frequency band; S5, based on stress gradient distribution and dynamic sensitive frequency band, designs a graded frequency sweep excitation mode and generates the initial control curve of the vibration aging excitation system; S6, during the operation of the vibration aging excitation system, uses an acceleration sensor network to realize real-time monitoring of the vibration mode of the machine tool structural components; S7 calculates the mode shape fit based on the modal confidence criterion, and then dynamically adjusts the output process parameters of the exciter based on the obtained mode shape fit using adaptive control technology based on a combination of prediction and fuzzy control, thereby realizing adaptive intelligent control of vibration aging.
2. The method of claim 1, wherein: Step S1 employs a phased multiphysics coupling simulation strategy, specifically as follows: (a) Simulation stage of filling process: Focus on dynamic simulation of temperature field and flow field, ignore stress field calculation; (b) Solidification process simulation stage: Using the temperature field and flow field data when the filling is completed as the initial boundary conditions, a coupled simulation model of temperature field-flow field-stress field is constructed; (c) Simulation stage of natural cooling process: When the temperature of the casting system drops to the temperature set by the unpacking process, the convection and radiation heat exchange and stress relaxation behavior of the machine tool structural parts and the environment are simulated until the temperature drops to room temperature, forming a complete simulation process of casting stress evolution.
3. The method of claim 1, wherein: the large machine tool structure is a large machine tool structure having a length of 2 meters or more in at least one direction. Step S2 includes the following sub-steps: S21: Using a blind hole stress tester and a residual stress ultrasonic non-destructive testing instrument, the surface and internal parts of the cast machine tool structural parts are measured at multiple locations to obtain stress data covering different depths and key parts. S22: By comparing the measured and simulated residual stress fields, the Kriging interpolation method is used to quantify the differences, and a spatial correlation model is constructed to optimize the parameters of the residual stress distribution field.
4. The method of claim 3, wherein: The steps of the Kriging interpolation method to iteratively correct the residual stress distribution model of the simulated residual stress field by combining measured stress data are as follows: a) calculate the ratio of measured stress to simulated stress to construct a correction coefficient field; b) use the Gaussian variogram to establish the spatial correlation between measurement points. c. Solve the optimal weight equation to determine the degree of influence of each measuring point on the prediction point; d. Perform multiplicative correction on the stress across the entire field based on the optimal weight.
5. The method of claim 4, wherein: the step of applying a stress to the large machine tool structure is performed by applying a stress to the large machine tool structure by applying a force to the large machine tool structure. The formula for calculating the correction factor is: wherein, is a correction factor for the i-th measurement point; is a measured stress value for the i-th measurement point; is a simulated stress value for the i-th measurement point; The Gaussian mutability function is: wherein, The correlation of stress between two points in a space is described with the change rule of distance; is the Euclidean distance between two points; is the variable range parameter, controlling the decay rate of correlation; The weighting equation is: wherein, is the covariance matrix between the measurement points, ; is the weight vector to be solved, which determines the influence degree of each measurement point on the simulation prediction point; is the covariance vector between the prediction point and all measurement points, ; The formula for correcting stress across the entire field is: wherein, is the modified full-field stress value; is the modification coefficient at the prediction point, obtained by weighted averaging of the measurement point modification coefficients; is the original simulated stress field.
6. The method of claim 1, wherein: Step S3 includes the following sub-steps: S31: The stress gradient of each part of the computer tool structure is calculated using a numerical differential algorithm, and the stress concentration area and the stress change area are analyzed. S32: Visualizes stress gradient distribution through color cloud maps, marks key processing areas, and provides a targeted basis for subsequent frequency sweep mode design.
7. The method of claim 6, wherein: the step of applying a stress to the large machine tool structure is performed by applying a stress to the large machine tool structure by applying a force to the large machine tool structure. The formula for calculating the stress gradient is: Wherein, x, y, z are X, Y, Z axis direction coordinates; , , are X, Y, Z three direction stress components.
8. The method of claim 1, wherein: Step S4 includes the following steps: S41: Considering the fixed constraints and ambient temperature of the machine tool structural components during the vibration aging process, perform modal analysis on the machine tool structural components to analyze their natural frequencies and mode shapes. S42: Detect the actual natural frequencies of machine tool structural components using professional modal analysis equipment to identify sensitive frequency bands.
9. The method of claim 1-8, wherein: Step S5 is as follows: S51: Within the dynamic sensitive frequency band, a full-range frequency sweep is carried out with a fixed frequency step size. The vibration response of the machine tool structural components at each frequency is monitored in real time with the help of an acceleration sensor network, and the changes in the vibration characteristics of the machine tool structural components are analyzed. S52: For stress concentration areas, the exciter sweep frequency step size is refined in a specific natural frequency range to capture the dynamic response characteristics of machine tool structural components in stress concentration areas, and the optimal excitation frequency is identified by judging the vibration response. S53: Based on the vibration response data acquired by frequency sweep, with the optimization objective of maximizing elimination efficiency, a genetic algorithm is used to optimize parameters to generate initial control curves of frequency-time (ft) and excitation force-time (Ft) to determine the optimal frequency and excitation force parameters at each moment.
10. The vibration aging control method for large machine tool structural components as described in any one of claims 1-8, characterized in that: Step S7 includes the following steps: S71: Based on the Modal Confidence Criterion (MAC), quantitatively calculate the fit between the measured vibration modes and theoretical vibration modes of machine tool structural components, and evaluate whether the actual vibration state of the machine tool structural components matches the theoretically calculated vibration modes; S72: Adaptive control technology that integrates prediction and fuzzy control is adopted. The mode shape fit is used as the input variable. The error between the mode shape fit and the target value is calculated and input into the two-dimensional fuzzy controller. Through fuzzy rule reasoning, the excitation process parameters of the exciter are output. The coarse and fine control strategy is used for dynamic adjustment to drive the exciter to adaptively adjust the operating parameters, ensuring that the mode shape fit reaches the set standard, and realizing adaptive intelligent control of the vibration aging process. The Modal Confidence Criterion (MAC) is as follows: in, , Let these be the two mode shape vectors to be compared. For the measured vibration mode, This is the theoretical vibration mode.