A precision casting mold life prediction method based on digital twinning

Abstract

The application relates to the technical field of digital twinning, and discloses a precision casting mold life prediction method based on digital twinning, which comprises a mold data acquisition module, a mold data processing module, a mold coefficient generation module, a prediction model establishment module and a mold life prediction module; the physical data, operation data and maintenance records are synchronously collected by the multi-source sensor, real-time monitoring of the whole life cycle is realized, four types of quantitative coefficients, a thermal fatigue coefficient, a comprehensive mechanical coefficient, a comprehensive wear coefficient and a comprehensive maintenance coefficient are established, comprehensive evaluation is realized, and multi-field damage coupling analysis is realized; the life prediction index unifies the multi-dimensional damage into 0-1 standard, outputs an intuitive life percentage, sets a graded early warning threshold, and improves the prediction accuracy; all the coefficients are calculated through a standardization formula, a full-digital process from data acquisition-index generation-model prediction-decision support is constructed, and the detection requirements of new materials and complex structure molds are met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, and more particularly to a precision casting mold life prediction method based on digital twinning. BACKGROUND

[0002] With the development of manufacturing industry towards intelligentization and digitization, the life of precision casting mold as a key production tool directly affects the production efficiency and cost.

[0003] The traditional technology adopts empirical formula method or artificial detection method to predict the life of precision casting mold, wherein the empirical formula method establishes a statistical model based on historical data, inputs material hardness, load and other parameters to calculate the remaining life, and quantifies the damage degree by microscopic observation or 3D scanning according to industry standards; the artificial detection method regularly stops checking the mold state, including: surface damage such as tearing, sticking, collapsing, structural integrity such as small core bending, push rod fracture, size deviation such as using micrometer, profilometer detection combined with casting defects to infer the mold health state reversely. The following defects exist:

[0004] Only data is obtained by stopping checking, and the damage change under dynamic working condition cannot be captured;

[0005] The thermal fatigue or mechanical wear is analyzed in isolation, and the synergistic effect of multiple physical fields is ignored;

[0006] The experience of maintenance personnel directly affects the detection result, and the prediction accuracy is insufficient;

[0007] Different enterprises adopt different evaluation standards, and it is also difficult to adapt to the detection needs of new materials and complex structure molds.

[0008] Therefore, a method of real-time detection, multi-parameter coupling, accurate prediction and perfect evaluation system is needed to solve the above problems. SUMMARY

[0009] In order to overcome the above defects of the prior art, the present application provides a precision casting mold life prediction method based on digital twinning to solve the problems existing in the background art.

[0010] To achieve the above purpose, the present application provides the following technical scheme: a precision casting mold life prediction method based on digital twinning, comprising:

[0011] S1, deploying multiple source sensors through a mold data acquisition module to monitor and collect physical data, running data and maintenance data of the precision casting mold in real time;

[0012] S2, the original data collected by the mold data processing module are denoised, cleaned and time-aligned, to generate thermal fatigue related indexes, mechanical load related indexes, mold wear related indexes and historical maintenance related indexes;

[0013] The thermal fatigue related indexes include temperature fluctuation coefficients, cooling rate coefficients and thermal cycle number coefficients; the mechanical load related indexes include static load coefficients and dynamic load coefficients; the mold wear related indexes include surface wear coefficients and crack propagation coefficients; and the historical maintenance related indexes include repair frequency coefficients and repair type coefficients;

[0014] S3, the related indexes obtained are mathematically modeled by a mold coefficient generation module to generate thermal fatigue coefficients, mechanical load coefficients, mold wear coefficients and historical maintenance coefficients;

[0015] S4, a life prediction model is established by a prediction model establishment module to fuse the coefficients to generate a life prediction index and convert it into a life percentage;

[0016] S5, the life prediction index is set with a threshold value by a mold life prediction module for graded early warning, and the maintenance decision is guided according to the grading result.

[0017] The present application quantifies thermal fatigue parameters such as temperature fluctuation and cooling rate by real-time collection of mold physical, operation and maintenance data through multi-source sensors, reflects mechanical damage by combining static / dynamic load coefficients, evaluates the wear degree by weighted calculation of surface wear and crack propagation, and introduces repair frequency and component replacement coefficients to represent maintenance influence to generate thermal fatigue, mechanical load, wear and other indexes and model them into coefficients, unify multi-physical field damage into a life prediction index, realize closed-loop management from data collection to graded early warning, and provide digital evaluation basis for mold health status.

[0018] Technical effects and advantages of the present application:

[0019] 1. The present application synchronously collects physical data, operation data and maintenance records through multi-source sensors to realize real-time monitoring throughout the whole life cycle and eliminate detection blind area;

[0020] 2. The present application establishes four types of quantitative coefficients, integrates temperature fluctuation, cooling rate and thermal cycle number into thermal fatigue coefficients, comprehensively weights and fuses static load and dynamic load into mechanical coefficients, dynamically weights and calculates surface wear and crack propagation into wear coefficients, and comprehensively evaluates repair frequency, component replacement and maintenance interval into maintenance coefficients to realize multi-field damage coupling analysis;

[0021] 3. The life prediction index of the present application unifies multi-dimensional damage into 0-1 standard, outputs intuitive life percentage and sets graded early warning threshold, directly relates to maintenance decision, and improves prediction accuracy.

[0022] 4、All the coefficients of the present application are calculated through a standardization formula, and a full-digital process from data acquisition-index generation-model prediction-decision support is constructed, which is suitable for the detection needs of new materials and complex structure molds. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 is a structural diagram of the present application.

[0024] Figure 2 is a flowchart of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the present application will be described clearly and completely in combination with the drawings in the present application, and in addition, the forms of each structure described in the following embodiments are only examples, and the automatic discharging device of the rotary furnace with the self-cooling function is not limited to each structure described in the following embodiments, and all other embodiments obtained by the ordinary skilled in the art without making creative efforts belong to the scope of protection of the present application.

[0026] Referring to Figure 1 The present application provides a precision casting mold life prediction method based on digital twinning, which comprises a mold data acquisition module, a mold data processing module, a mold coefficient generation module, a prediction model establishment module and a mold life prediction module.

[0027] Referring to Figure 2 The specific implementation steps of the present application include the following steps:

[0028] S1, deploy multiple source sensors through the mold data acquisition module, and real-time monitor and collect physical data, running data and maintenance data of the precision casting mold.

[0029] It needs to be specifically pointed out that the physical property data specifically includes:

[0030] Geometry parameters: the CAD model data of the mold includes cavity size, parting surface structure, etc.

[0031] Material properties: hardness, thermal expansion coefficient, fatigue limit of mold steel;

[0032] Process parameters: pouring temperature, cooling rate, injection pressure and other casting process parameters.

[0033] Real-time running state data includes:

[0034] Mechanical load: collect mold force, demolding resistance and vibration spectrum through stress sensors;

[0035] Wear: 3D scanning or industrial camera records surface cracks, erosion depth.

[0036] Maintenance data includes repair frequency, replacement component type and timestamp.

[0037] S2, the original data collected by the mold data processing module is denoised, cleaned and time aligned, and the thermal fatigue related index, the mechanical load related index, the mold wear related index and the historical maintenance related index of the mold are generated.

[0038] Need to be specified is that the thermal fatigue related index includes temperature fluctuation coefficient and cooling rate coefficient.

[0039] The temperature fluctuation coefficient is specifically:

[0040] ;

[0041] Where T is the temperature fluctuation coefficient, reflecting the linear relationship between thermal fatigue damage and temperature fluctuation.

[0042] K is the material sensitivity coefficient, indicating the response strength of a specific mold steel to temperature fluctuation, such as H13 steel with a value between 0.01 and 0.03; T j is the pouring temperature, reflecting the thermal shock strength of molten metal on the mold, the higher the pouring temperature, the faster the thermal fatigue damage rate; Tb is the mold reference temperature, taking the mold preheating temperature as the reference for thermal fatigue damage; △Tc is the critical temperature difference threshold, which is the lower limit of the temperature difference for the material to occur thermal fatigue.

[0043] Need to explain is that the temperature difference ratio (T j -T b ) / △T quantifies the amplitude of thermal stress, when the ratio is >1, that is, when the temperature difference T j -T b exceeds the critical value △T c , the greater the temperature difference, the closer the exponential term to 0, the temperature fluctuation coefficient close to 1, the higher the thermal fatigue damage; T j =T b , the exponential term is 1, and the coefficient is 0, indicating no thermal fatigue accumulation.

[0044] The cooling rate coefficient is specifically:

[0045] ;

[0046] Where V is the cooling rate coefficient, reflecting the influence degree of cooling efficiency on mold thermal fatigue life.

[0047] m is the material sensitivity constant, determined by fatigue test, reflecting the sensitivity of the material to the change of cooling rate, the greater the sensitivity, the value range is between 0.05 and 0.08; Va The measured cooling rate; V b The reference cooling rate, corresponding to the critical point of mold thermal fatigue, as a threshold for judging cooling efficiency, calibrated by experiment.

[0048] Need to explain is, no matter the cooling rate exceeds the reference value or below the reference value, deviation from the reference value will lead to thermal stress accumulation, cooling rate is higher than the reference value, rapid cooling leads to high residual stress, cooling rate is lower than the reference value, slow cooling causes the organization coarsening; When the cooling rate is equal to the reference value, the exponential term is 1, the coefficient is minimum; The greater the difference between the cooling rate and the reference value, the closer the exponential term to 0, the cooling rate coefficient tends to 1, the more serious the thermal fatigue.

[0049] Need to explain is, the mechanical load related index includes static load coefficient and dynamic load coefficient.

[0050] The static load coefficient calculation is specifically the clamping force divided by the mold pressure area and then divided by the yield strength of the mold material, reflecting the average stress level of the mold in the casting process; wherein the clamping force reflects the mechanical pressure that the mold bears in the casting process, directly affecting the closing accuracy of the cavity; The mold pressure area is the effective stress area region of the mold cavity in contact with the molten metal, which converts concentrated force into average distributed stress to avoid plastic deformation caused by local overload; The material yield strength is the plastic deformation column cutoff of the mold steel, which is used as a safety margin reference to judge whether it enters the plastic deformation stage.

[0051] Need to explain is, the above results are normalized, take the calculated value compared with the preset safety threshold, get the normalized static load coefficient J, when J = 0, it means that the mold is in the empty state, no mechanical stress; When J is close to 1, it shows that the local stress is close to the material yield strength, and further exceeding will cause permanent deformation.

[0052] The dynamic load coefficient calculation is specifically the n-th power of the ratio of the stress intensity factor amplitude and the material fracture toughness threshold, and then summing up, reflecting the cumulative life loss of the mold under periodic vibration and impact load; wherein the stress intensity factor amplitude reflects the local stress concentration degree of load fluctuation, directly affecting the crack initiation and development rate; The material fracture toughness threshold is the critical threshold, indicating the crack resistance of the material, exceeding this value will lead to rapid fracture; N is a material constant, controlling the sensitivity of crack propagation rate to stress intensity factor amplitude, the value range is between 3-5.

[0053] Need to explain is, the above results are normalized, using logarithmic transformation to get the normalized dynamic load coefficient D, wherein , D a is the calculated value, D maxD is the critical value; when D = 0, it means no alternating load, only constant stress; when D approaches 1, it means that high-frequency vibration and impact load are significant, and the crack propagation rate increases.

[0054] Need to be specific, the mold wear-related indicators include surface wear coefficient and crack propagation coefficient.

[0055] The surface wear coefficient is calculated as the product of the contact pressure and the relative sliding distance divided by the product of the mold material hardness and the initial contact area, and then multiplied by the material wear constant, which reflects the wear rate of the mold under sliding contact conditions; wherein the contact pressure is the load force per unit area, and the increase in pressure directly leads to an increase in wear; the sliding distance is the length of the relative motion trajectory of the mold and the metal, and the longer the cumulative sliding distance, the greater the wear; the material hardness is the property of the material itself, and the higher the hardness, the lower the wear coefficient; the initial contact area is the effective contact area between the metal and the mold, and the larger the unit area, the smaller the pressure; the wear constant is determined by the material properties and reflects the chemical properties of the interface friction.

[0056] Need to explain that the calculated value is normalized, that is, the current calculated value is subtracted from the minimum value in the historical data, and the normalized surface wear coefficient B is obtained by subtracting the minimum value in the historical data from the critical failure threshold; the coefficient is 0, which means no contact load or material hardness close to infinity; the larger the coefficient, the greater the wear and the shorter the mold life.

[0057] The crack propagation coefficient is specifically:

[0058] ;

[0059] Where L is the crack propagation coefficient, which quantifies the extent of crack propagation of the mold material under cyclic loading.

[0060] a0 and a t are the initial crack depth and the current crack depth, respectively, reflecting the cumulative damage degree of the crack; △K is the stress intensity factor amplitude, which represents the strength of the stress field at the crack tip, and the larger the value, the faster the crack propagation rate, △K = K max -K min , wherein K max and K min are the maximum and minimum stress intensity factors, respectively; C is a material constant obtained by experimental fitting, and n is an index reflecting the sensitivity of the crack propagation coefficient to △K, with a value range of 2-4.

[0061] It needs to be explained that the crack propagation coefficient is normalized, and the difference between the current value and the reference value is divided by the difference between the critical failure value and the reference value; When a0=at, it means no crack or stable crack, such as the initial value of the new mold close to 0; When the coefficient is close to 1, it means that the crack is about to lose stability and expand, and the mold life is shorter.

[0062] Need to be explained in detail is that the historical maintenance related indicators include maintenance frequency coefficient and maintenance type coefficient.

[0063] The maintenance frequency coefficient is specifically:

[0064] ;

[0065] Wherein X is the maintenance frequency coefficient, reflecting the influence degree of each maintenance on the remaining life of the equipment.

[0066] λ is the attenuation coefficient, reflecting the influence intensity of single maintenance on the mold life, and the value range is between 0.15-0.3; N is the cumulative maintenance frequency, recording the mold maintenance frequency.

[0067] It needs to be explained that the exponential function is used to reflect the nonlinear relationship between maintenance frequency and performance degradation, and with the increase of maintenance frequency, the mold performance degradation is more serious, the coefficient value is larger, and the mold life is shorter.

[0068] The component replacement coefficient is specifically:

[0069] ;

[0070] Wherein G is the component replacement coefficient, reflecting the influence degree of different component replacement on the mold life.

[0071] W i The weight of the replaced component, which quantifies the importance of different components, the value range of the core component is between 0.8-1, the value range of the key part is between 0.5-0.7, and the value range of the conventional part is between 0.2-0.4; G is the replacement mark, 0 represents not replaced, and 1 represents replaced.

[0072] It needs to be explained that the replacement of core components has the greatest influence on life, and the replacement of core components, the larger the coefficient value, the shorter the mold life.

[0073] S3, the related indicators obtained by the mold coefficient generation module are mathematically modeled to generate thermal fatigue coefficient, mechanical load coefficient, mold wear coefficient and historical maintenance coefficient.

[0074] Need to be explained in detail is that the thermal fatigue coefficient is specifically:

[0075] R=T*V*N;

[0076] Wherein R is the thermal fatigue coefficient, quantifying the cumulative degree of damage of the mold under periodic thermal load.

[0077] T is the temperature fluctuation coefficient, reflecting the influence of the deviation of pouring temperature from the reference temperature on the sensitivity of the material to thermal fatigue. The greater the temperature fluctuation coefficient, the more serious the thermal fatigue damage, and the shorter the service life of the mold.

[0078] V is the cooling rate coefficient, reflecting the contribution of the difference between the cooling rate and the reference rate to the accumulation of thermal stress. The greater the cooling rate coefficient, the more serious the thermal fatigue damage, and the shorter the service life of the mold.

[0079] N is the thermal cycle number coefficient, reflecting the cumulative damage of thermal cycle number to fatigue life. The irreversible damage caused by each cycle is accumulated through the number of thermal cycles. The more the number of cycles, the more serious the thermal fatigue damage.

[0080] It needs to be explained that the product relationship shows that the temperature fluctuation and the cooling rate jointly determine the damage degree of a single cycle, and the number of thermal cycles controls the total amount of damage. The greater the thermal fatigue coefficient, the greater the damage to the mold, and the shorter the service life of the mold.

[0081] It needs to be specifically explained that the comprehensive mechanical coefficient is specifically:

[0082] F=α*J+(1-α)*D;

[0083] Wherein F is the comprehensive mechanical coefficient, quantifying the comprehensive influence of the mechanical load on the precision casting mold in work, reflecting the cumulative damage of the mold under mechanical stress. The coefficient is 0, which represents that the mold has no any mechanical load; the greater the coefficient, the faster the service life of the mold under mechanical load.

[0084] It needs to be explained that J is the static load coefficient, which is the constant mechanical stress level that the mold bears, determines the plastic deformation risk of the mold under stable working conditions, and the greater the static load coefficient, the closer the mold to the yield limit, the shorter the service life; α is the contribution weight of static load in comprehensive mechanical damage, the value range is between 0.4-0.6, but it needs to be noted that in injection mold, the value range of α can be increased to 0.8.

[0085] D is the dynamic load coefficient, reflecting the cumulative damage of the mold under dynamic load, determining the damage propagation rate of the mold under vibration load. The greater the dynamic load coefficient, the greater the risk of crack propagation, and the shorter the service life of the mold; 1-α is the contribution weight of dynamic load in comprehensive mechanical damage, the value range is between 0.4-0.6, but it needs to be noted that in stamping mold, the value of 1-α can be increased to 0.7.

[0086] It needs to be specifically explained that the comprehensive wear coefficient is specifically:

[0087] M=β*B+(1-β)*L;

[0088] wherein M is the comprehensive wear coefficient, quantifying the degree of die wear by weighting surface wear and crack propagation, the higher the coefficient value, the more serious the wear and the shorter the remaining life.

[0089] It needs to be explained that B is the surface wear coefficient, quantifying the wear rate, determining the amount of material loss under unit contact pressure and sliding distance, the larger the surface wear coefficient, the more serious the wear and the shorter the remaining life of the die.

[0090] β is the surface wear coefficient weight, specifically:

[0091] ;

[0092] t is the current number of use cycles, determining the main stage of wear, including early wear and later crack dominated; t0 is the number of wear turning point cycles, calibrated by historical data; b is the decay rate coefficient, controlling the weight switching.

[0093] L is the crack propagation coefficient, which is the fatigue damage accumulation, converting the stress intensity factor amplitude into crack depth increment, reflecting the damage accumulation under load, embodying the effect of internal defects of the die material on crack propagation, the larger the crack propagation coefficient, the more serious the crack and the shorter the die life; 1-β is the crack propagation coefficient weight.

[0094] It needs to be specified that the comprehensive maintenance coefficient is specifically:

[0095] ;

[0096] wherein W is the comprehensive maintenance coefficient, evaluating the maintenance state of the precision casting die, quantifying the comprehensive influence of repair behavior and component replacement on the die life, the larger the coefficient value, the shorter the remaining life of the die.

[0097] It needs to be explained that X is the maintenance frequency coefficient, reflecting the cumulative damage of maintenance frequency on the performance of the die, the more the maintenance frequency, the larger the coefficient, and the shorter the die life.

[0098] G is the component replacement coefficient, reflecting the influence of replacing different components on the die life, the core components are all replaced, the coefficient value is the largest, and the die life is shorter.

[0099] Z is the number of cycles since the last maintenance, Z a is the historical average maintenance interval, the ratio term reflects the deviation degree of the current maintenance cycle from the historical average level, if Z>Z a , indicating that the current maintenance interval exceeds the historical average, indicating that the maintenance is delayed, and the coefficient value increases.

[0100] S4, the life prediction model is established by the prediction model establishment module, and a life prediction index is generated and converted into a life percentage by fusing each coefficient.

[0101] It needs to be specifically pointed out that the life prediction index is specifically:

[0102] ;

[0103] Wherein Z is the life prediction index, which reflects the mold health state by fusing multiple physical damages, 0 represents a brand-new state, and 1 represents complete failure.

[0104] R is a thermal fatigue coefficient, which reflects the damage degree of the mold material under cold and hot cycles; since the failure of the precision casting mold accounts for the highest proportion in the case of thermal fatigue, the highest weight is given, and therefore the weight ω1 is in the range of 0.35-0.4.

[0105] F is a comprehensive mechanical coefficient, which quantifies the cumulative damage caused by alternating stress; the proportion of mechanical load causing mold failure is second, so the value range of the weight ω2 is between 0.2-0.25.

[0106] M is a comprehensive wear coefficient, which reflects the influence of mold surface wear area and crack depth on mold life; wear causes the lowest proportion of mold failure, and the value range of the weight ω3 is between 0.15-0.2.

[0107] W is a comprehensive maintenance coefficient, which quantifies the damage of maintenance times and component replacement to the life of the mold; the maintenance factor directly affects the evolution rate of other coefficients; the value range of the weight ω4 is between 0.2-0.25.

[0108] The life percentage is specifically:

[0109] S=(1-Z)*100%;

[0110] Wherein S is the mold life percentage, which reflects the proportion of the current precision casting mold in the theoretical life, and the lower the percentage, the more it indicates that it is in the end of life.

[0111] S5, the life prediction index is set to a threshold for graded early warning by the mold life prediction module, and the maintenance decision is guided according to the grading result.

[0112] It needs to be specifically pointed out that the life prediction index is set to a threshold for grading, which is specifically:

[0113] When 0≤Z<z1, the remaining life of the mold is >s1, and routine inspection is performed;

[0114] When z1≤Z<z2, the remaining life of the mold is between s2-s1, and the detection period is shortened;

[0115] When z2≤Z<z3, the remaining life of the mold is between s3-s2, spare parts are prepared, and the next maintenance is arranged;

[0116] When z3≤Z<z4, the remaining life of the mold is between s4-s3, forced maintenance within 72 hours;

[0117] When z4≤Z≤1, the remaining life of the mold is <s4, and immediate shutdown and replacement;

[0118] Wherein z1 is valued between 0.25-0.35, z2 is valued between 0.45-0.55, z3 is valued between 0.65-0.75, and z4 is valued between 0.85-0.95;

[0119] Wherein s1 is valued between 65%-75%, s2 is valued between 45%-55%, s3 is valued between 25%-35%, and s4 is valued between 5%-15%.

[0120] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of the present application can be realized by means of software or software combined with necessary general hardware platforms, and of course can also be realized by hardware functions; based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium, and includes a plurality of instructions for causing a computer device, such as but not limited to a personal computer, a server, or a network device, to execute all or part of the steps of the method described in any embodiment of the present application.

[0121] The above describes exemplary embodiments of the present application, and it should be understood that the above exemplary embodiments are not restrictive but illustrative, and the protection scope of the present application is not limited thereto; it should be understood that those skilled in the art can modify and vary the embodiments of the present application without departing from the spirit and scope of the present application, and these modifications and variations should be within the protection scope of the present application.

Claims

1. A method for predicting the lifespan of precision casting molds based on digital twins, characterized in that, Specifically, it includes: S1. Deploy multi-source sensors through the mold data acquisition module to monitor and collect physical data, operation data and maintenance data of precision casting molds in real time; S2. The raw data collected is denoised, cleaned and time-aligned by the mold data processing module to generate thermal fatigue related indicators, mechanical load related indicators, mold wear related indicators and historical maintenance related indicators of the mold. The thermal fatigue-related indicators include temperature fluctuation coefficient, cooling rate coefficient, and thermal cycle number coefficient; the mechanical load-related indicators include static load coefficient and dynamic load coefficient; the mold wear-related indicators include surface wear coefficient and crack propagation coefficient; and the historical maintenance-related indicators include repair frequency coefficient and component replacement coefficient. S3. The relevant indicators obtained are mathematically modeled through the mold coefficient generation module to generate thermal fatigue coefficient, comprehensive mechanical coefficient, comprehensive wear coefficient and comprehensive maintenance coefficient. S4. By integrating various coefficients through the prediction model building module, a life prediction model is constructed to generate a life prediction index and convert it into a life percentage. S5. Set thresholds for the life prediction index through the mold life prediction module to provide graded early warnings, and guide maintenance decisions based on the graded results.

2. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: The temperature fluctuation coefficient is specifically: ； Where T is the temperature fluctuation coefficient, K is the material sensitivity coefficient, and T j Tb is the pouring temperature, Tc is the mold reference temperature, and ΔTc is the critical temperature difference threshold. The cooling rate coefficient is specifically: ； Where V is the cooling rate coefficient, m is the material sensitivity constant, and V a To measure the cooling rate, V b The reference cooling rate is used.

3. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: The static load factor is specifically calculated by dividing the clamping force by the mold bearing area and then by the yield strength of the mold material, which reflects the average stress level of the mold during the casting process. The dynamic load factor is specifically calculated by raising the ratio of the single stress intensity factor amplitude to the material fracture toughness threshold to the power of n, and then summing them to reflect the cumulative life loss of the mold under periodic vibration and impact loads.

4. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: The surface wear coefficient is specifically calculated by multiplying the product of contact pressure and relative sliding distance by the product of mold material hardness and initial contact area, and then multiplying by the material wear constant, reflecting the wear rate of the mold under sliding contact conditions. The crack propagation coefficient is specifically: ； Where L is the crack propagation coefficient, and a0 and a t These represent the initial crack depth and the current crack depth, respectively, where ΔK is the stress intensity factor amplitude, and ΔK = K. max -K min K max and K min are the maximum and minimum stress intensity factors, respectively, and C is a material constant.

5. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: The specific maintenance frequency coefficient is as follows: ； Where X is the maintenance frequency coefficient, λ is the attenuation coefficient, and N is the cumulative maintenance frequency. The mold maintenance frequency is recorded using an exponential function to reflect the non-linear relationship between the maintenance frequency and performance degradation. As the maintenance frequency increases, the mold performance degradation is more severe, the coefficient value is larger, and the mold life is shorter. The specific component replacement coefficient is as follows: ； Where G is the component replacement coefficient, w i The weight of the replaced parts is given by g, which is the replacement flag, where 0 indicates no replacement and 1 indicates replacement. The replacement of core components has the greatest impact on lifespan. If all core components are replaced, the larger the coefficient value, the shorter the mold lifespan.

6. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: The thermal fatigue coefficient is specifically: R = T * V * N; Where R is the thermal fatigue coefficient, which quantifies the degree of cumulative damage to the mold under cyclic thermal load; T is the temperature fluctuation coefficient; V is the cooling rate coefficient; and N is the thermal cycle number coefficient. The comprehensive mechanical coefficients are specifically as follows: F = α * J + (1 - α) * D; Where F is the comprehensive mechanical coefficient, which quantifies the comprehensive influence of mechanical loads on the precision casting mold during operation; J is the static load coefficient, D is the dynamic load coefficient, and α is the contribution weight of static load in comprehensive mechanical damage. The comprehensive wear coefficient is specifically as follows: M = β*B + (1-β)*L; Where M is the comprehensive wear coefficient, which quantifies the degree of mold wear by weighting surface wear and crack propagation; B is the surface wear coefficient, and β is the weight of the surface wear coefficient, specifically: ； t is the current number of usage cycles, which determines the main stage of wear, including early wear and later crack-dominated wear; t0 is the number of cycles at the wear inflection point, which is calibrated through historical data; b is the attenuation rate coefficient, which controls the weight switching; L is the crack propagation coefficient, and 1-β is the weight of the crack propagation coefficient. The comprehensive maintenance factor is specifically as follows: ； Where W is the comprehensive maintenance coefficient, which assesses the maintenance status of precision casting molds; X is the repair frequency coefficient; G is the component replacement coefficient; and Z is the number of cycles since the last repair. a This represents the historical average maintenance interval.

7. The method for predicting the life of precision casting molds based on digital twins according to claim 1, characterized in that: Life Prediction Index (WPI) details: ； Where Z is the life prediction index, which reflects the health status of the mold by integrating multiple physical damages, with 0 representing a brand new state and 1 representing complete failure; R is the thermal fatigue coefficient, F is the comprehensive mechanical coefficient, M is the comprehensive wear coefficient, W is the comprehensive maintenance coefficient, and ω1, ω2, ω3 and ω4 are the weighting coefficients of the thermal fatigue coefficient, comprehensive mechanical coefficient, comprehensive wear coefficient and comprehensive maintenance coefficient, respectively. The percentage of lifespan is specifically: S = (1-Z) * 100%; Where S represents the mold life percentage, reflecting the proportion of the current precision casting mold that is within its theoretical lifespan. The lower the percentage, the more likely it is to be nearing the end of its lifespan.

Images