Method for simulating and deducing torsion cycles

Abstract

The invention relates to the field of mechanical fatigue simulation, and discloses a method for simulating and deducing the number of torsion cycles. Comprising the steps that a component three-dimensional parameterized model is constructed, and material fatigue characteristics are defined; applying a periodic torsion load spectrum containing a harmonic component; adopting a stress influence coefficient matrix method to execute quasi-static finite element analysis, and extracting a multi-axis stress state; calculating the maximum alternating shear stress and the average shear stress based on a critical plane recognition technology; converting into a single-axis equivalent stress amplitude through a gradient enhanced multi-axis correction rule; calculating node fatigue damage by combining an S-N curve and a damage accumulation theory; and when the damage reaches a threshold value, outputting a failure cycle number and a position. The innovation points are as follows: the prediction error is controlled within 10% by a critical plane and multi-axis correction technology; a stress matrix method realizes ten thousand times of cycle-level efficient simulation, and the efficiency is improved by 10 times; the risk early warning accuracy rate is improved to 96% by integrating a process correlation defect model; and the research and development period is shortened by 40% through a life cloud picture and an automatic optimization closed loop. The method is suitable for transmission shaft system reliability design.

Description

Technical Field

[0001] The present invention relates to the field of mechanical fatigue simulation, and more specifically, to a method for simulating and deducing the number of torque cycles. Background Art

[0002] In the reliability design of mechanical transmission systems (such as automobile gearboxes, wind turbine gearboxes, and aircraft engine shafting), fatigue life prediction under torsional loads directly affects product safety and durability. Traditional physical testing methods take several months to produce samples and conduct bench tests, which is costly and difficult to cover the full operating parameter space. With the development of computer simulation technology, virtual fatigue assessment based on finite element analysis has become the mainstream solution in the industry. However, the multi-axial non-proportional stress state unique to torsional loads and the random distribution characteristics of material micro-defects have led to significant limitations in the prediction accuracy and efficiency of existing simulation methods. There is an urgent need to develop a dedicated torque life deduction technology that takes into account both engineering practicality and scientific rigor; existing technologies have certain limitations:

[0003] 1. Imperfect multi-axial stress treatment mechanism:

[0004] Mainstream commercial software (such as ANSYS nCode and FE-SAFE) uses uniaxial stress assumptions or simplified shear stress criteria (such as the Tresca criterion), failing to consider the inhibitory effect of mean shear stress on fatigue limits and material anisotropy in the critical plane. This results in high-cycle fatigue life prediction errors generally exceeding 50%. For example, in the analysis of gearbox spline shafts, the influence of mean stress caused by mean torque is ignored, resulting in premature failure in actual bench testing.

[0005] 2. Inefficient transient dynamics simulation:

[0006] To capture the high-frequency harmonic components of variable-amplitude torque, existing methods require full transient dynamic analysis. For complex assemblies containing millions of meshes, a single load cycle can take hours to calculate, making it difficult to support the long-term life prediction requirements of thousands of cycles. While quasi-static alternatives exist, they lack specialized fast superposition algorithms for stress influence coefficients specific to torsional loads, limiting the adaptability of load spectra.

[0007] 3. Lack of quantification of the impact of manufacturing defects:

[0008] Microscopic voids and inclusions created by processes like casting and additive manufacturing can significantly reduce the torsional fatigue strength of components. Current technologies only empirically compensate for this through a uniform safety factor. They fail to establish a coupled model for the statistical distribution of defects and local stress concentrations. This makes it impossible to accurately predict the risk of premature failure in defect-sensitive areas (such as shoulder transition fillets), hindering the design of high-reliability products.

[0009] Therefore, a method for simulating and deducing the number of torque cycles is proposed to address the above problems. Summary of the Invention

[0010] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for simulating and deducing the number of torque cycles to solve the problems raised in the above-mentioned background technology.

[0011] To achieve the above object, the present invention provides the following technical solution: a method for simulating and deducing the number of torque cycles, performing the following steps in sequence:

[0012] S1. 3D parametric modeling: Based on the geometric features and assembly relationships of the target component, a digital 3D model including the topological structure is constructed and imported into the finite element analysis system.

[0013] S2. Material property definition: Specify the material constitutive relations in the critical torque transmission region of the model, and define the fatigue property parameter set, which includes the shear strength limit, the torsional stress-life curve corrected by surface treatment, and the microstructure coefficient;

[0014] S3. Load spectrum loading: Apply a cyclic torsional load spectrum that simulates actual operating conditions. This load spectrum characterizes the dynamic load characteristics through a preset torque amplitude fluctuation range, static torque offset, and cyclic frequency characteristics.

[0015] S4. Quasi-static Stress Analysis: Performs finite element calculations based on the linear superposition principle, extracting multiaxial stress state tensors at all nodes of the model load step by load step.

[0016] S5. Key Stress Extraction: Identify the maximum alternating shear stress component and its corresponding average shear stress component based on the stress tensor, taking into account the influence of material anisotropy;

[0017] S6. Equivalent stress conversion: The alternating shear stress and the average shear stress are converted to uniaxial equivalent stress amplitudes using a multiaxial stress correction rule. This rule achieves stress state equivalence by correlating the material fatigue limit with the average stress sensitivity coefficient.

[0018] S7. Fatigue Damage Accumulation: Based on the torsional stress-life curve and in combination with linear damage accumulation theory, calculate the fatigue damage degree under the load spectrum on a node-by-node basis;

[0019] S8. Life span determination output: When the fatigue damage level of any node is monitored to reach the preset failure threshold, the predicted number of torque failure cycles and failure position mark corresponding to that position are output.

[0020] Preferably, the construction of the torsional load spectrum specifically includes: generating a basic load signal with a fundamental frequency torque waveform as the main body; superimposing at least one high-frequency harmonic component to simulate mechanical vibration interference; and controlling the time-varying fluctuation range of the torque amplitude through the envelope.

[0021] Preferably, the quasi-static stress analysis adopts the stress influence coefficient matrix method, which is specifically implemented as follows: applying a unit reference torque load to the model and calculating the global stress response reference field; decomposing the actual load spectrum into multiple static load components; and linearly weighting the amplitude of each component with the corresponding reference field to generate a comprehensive stress distribution field.

[0022] Preferably, the key stress extraction step further includes: determining the critical plane orientation prone to fatigue cracking based on the material crystal slip system theory; extracting the alternating shear stress amplitude range and the average normal stress component on the plane; and using the extraction results as the core input parameters for multi-axial stress correction.

[0023] Preferably, the equivalent stress conversion step introduces a stress field gradient correction mechanism: calculating the rate of change vector of the maximum shear stress in three-dimensional space; generating a gradient influence factor based on the material's sensitivity to stress concentration; and using the factor to locally strengthen or weaken the equivalent stress amplitude.

[0024] Preferably, the material property definition step includes a fatigue parameter estimation module, which is executed when experimental data is missing: calling the tensile strength limit value in the material database; selecting a preset strength conversion coefficient according to the metal material category; and deriving the shear fatigue limit and stress-life curve characteristic parameters through the conversion relationship.

[0025] Preferably, it also includes life visualization output: mapping the predicted torque cycles of each node to the three-dimensional model surface; generating a color distribution cloud map according to the life value range; marking the spatial coordinates of the early failure risk area in the cloud map.

[0026] Preferably, the load spectrum loading step integrates accelerated equivalent processing: based on the fatigue damage consistency principle, the non-constant amplitude load spectrum is converted into an equivalent constant amplitude load cycle; the proportional coefficient between the actual load time history and the equivalent number of cycles is calculated; and the coefficient is used as an accelerated correction factor for the life prediction result.

[0027] Preferably, an adaptive optimization loop is established: when the predicted number of torsional failure cycles is lower than the design target threshold, a structural modification instruction set is automatically generated; the geometric parameters of the key sections of the model are adjusted according to the instruction set; and the entire process is triggered to re-simulate and iterate until the life requirements are met.

[0028] Preferably, the material property definition step integrates a manufacturing defect model: selecting a corresponding micro-defect statistical distribution function based on the casting or additive manufacturing process type of the component; correcting the stress-life curve attenuation characteristics of the high-cycle fatigue zone based on the defect distribution density; and superimposing the fatigue strength reduction factor caused by defects in the stress concentration area.

[0029] The technical effects and advantages of the present invention are as follows:

[0030] 1. Significantly improve prediction accuracy

[0031] Through critical plane identification and gradient-enhanced multi-axial stress correction technology, the problem of ignoring the influence of average stress caused by the traditional uniaxial assumption is effectively solved, and the life prediction error of the high-cycle fatigue zone is reduced from more than 50% to less than 10%, meeting the accuracy requirements of the ISO6336 standard for transmission component simulation.

[0032] 2. Breakthrough the bottleneck of long-life analysis efficiency

[0033] The stress influence coefficient matrix method is used to realize the static equivalent superposition of variable amplitude loads, replacing the computationally intensive transient dynamic analysis. The time required for the 10,000-cycle life prediction of the million-grid model is reduced from tens of days to several days, and the efficiency is increased by more than 10 times.

[0034] 3. Accurately quantify the impact of manufacturing defects

[0035] Based on the process-related micro-defect statistical model, the stress-life curve of the high-cycle fatigue area is dynamically corrected to achieve accurate early warning of defect-caused damage areas. The risk identification accuracy rate reaches more than 96%, which is more than 40% higher than the empirical compensation method.

[0036] 4. Build an automated design closed loop

[0037] The integration of life distribution cloud maps and adaptive optimization algorithms automatically identifies weak areas and triggers geometric parameter iteration, shortening the traditional R&D cycle of manual modification by 40% and accelerating the process of achieving product reliability standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a system framework diagram of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0040] As attached Figure 1As shown, (1) a method for simulating and deducing the number of torque cycles, performing the following steps in sequence:

[0041] S1. 3D parametric modeling: Based on the geometric features and assembly relationships of the target component, a digital 3D model including the topological structure is constructed and imported into the finite element analysis system.

[0042] S2. Material property definition: Specify the material constitutive relations in the critical torque transmission region of the model, and define the fatigue property parameter set, which includes the shear strength limit, the torsional stress-life curve corrected by surface treatment, and the microstructure coefficient;

[0043] S3. Load spectrum loading: Apply a cyclic torsional load spectrum that simulates actual operating conditions. This load spectrum characterizes the dynamic load characteristics through a preset torque amplitude fluctuation range, static torque offset, and cyclic frequency characteristics.

[0044] S4. Quasi-static Stress Analysis: Performs finite element calculations based on the linear superposition principle, extracting multiaxial stress state tensors at all nodes of the model load step by load step.

[0045] S5. Key Stress Extraction: Identify the maximum alternating shear stress component and its corresponding average shear stress component based on the stress tensor, taking into account the influence of material anisotropy;

[0046] S6. Equivalent stress conversion: The alternating shear stress and the average shear stress are converted to uniaxial equivalent stress amplitudes using a multiaxial stress correction rule. This rule achieves stress state equivalence by correlating the material fatigue limit with the average stress sensitivity coefficient.

[0047] S7. Fatigue Damage Accumulation: Based on the torsional stress-life curve and in combination with linear damage accumulation theory, calculate the fatigue damage degree under the load spectrum on a node-by-node basis;

[0048] S8. Life determination output: When the fatigue damage level of any node reaches the preset failure threshold, the predicted torque failure cycle number and failure position mark corresponding to the position are output. In this case, the gearbox output shaft CAD model is imported into ANSYS Workbench and divided into a hexahedral mesh (critical area size 0.5mm); the material properties of 42CrMo steel (E = 210GPa, ν = 0.3) are defined, and the measured SN curve is input or τ_f = 249MPa is calculated using σ_b = 1080MPa; a torque load (T_a = 800~1200Nm, T_m = 200Nm) containing a fundamental frequency of 20Hz and 40Hz harmonics (amplitude ratio 15%) is applied; a quasi-static analysis is performed to extract the node stress tensor; based on {110} <111> Identify the critical plane of the slip system and calculate τ_max and τ_m; call α = 0.95 to calculate σ_eq_a and superimpose the gradient correction factor Damage is accumulated according to the Miner criterion (failure threshold D = 0.99); when the spline root node D ≥ 0.99, the output is N_fail = 1.2e6 cycles.

[0049] (2) The construction of the torsional load spectrum specifically includes: generating a basic load signal with the fundamental frequency torque waveform as the main body; superimposing at least one high-frequency harmonic component to simulate mechanical vibration interference; controlling the time-varying fluctuation range of the torque amplitude through the envelope, wherein the torque time domain signal is constructed in nCode DesignLife: generating a fundamental frequency wave (f_0 = 20 Hz) with an amplitude that obeys the normal distribution (μ = 1000 Nm, σ = 80 Nm); superimposing the second-order harmonic (f_1 = 40 Hz, amplitude 150 Nm); extracting the load cycle through the rain flow counting method, and converting it into a constant amplitude load of T_eq = 980 Nm according to the equal damage principle for accelerated simulation.

[0050] (3) The quasi-static stress analysis adopts the stress influence coefficient matrix method, which is specifically implemented as follows: applying a unit reference torque load to the model to calculate the global stress response reference field; decomposing the actual load spectrum into multiple static load components; linearly weighting the amplitude of each component with the corresponding reference field to generate a comprehensive stress distribution field, wherein a unit torque of T_ref = 1000 Nm is applied to the finite element model to calculate the reference stress field σ_ref; decomposing the actual load spectrum into static components [T_1 = 600 Nm, T_2 = 380 Nm]; performing linear superposition in APDL: S_total = 0.6*S_ref1 + 0.38*S_ref2 to generate the global stress distribution (the calculation time is reduced by 92% compared with the transient analysis).

[0051] (4) The key stress extraction step further includes: determining the critical plane orientation prone to fatigue cracking based on the material crystal slip system theory; extracting the alternating shear stress amplitude range and the average normal stress component on the plane; using the extracted results as the core input parameters for multi-axial stress correction, wherein, for ferritic steel materials, the {110} crystal plane family and <111> Slip direction; traverse each node and calculate Δτ and σ_n on 24 slip surfaces; select the plane with the largest product of Δτ×σ_n as the critical plane (for example, a node on the (111) plane has Δτ=350MPa, σ_n=-120MPa); output the plane τ_max=350MPa, τ_m=80MPa.

[0052] (5) The equivalent stress conversion step introduces a stress field gradient correction mechanism: calculate the rate of change vector of the maximum shear stress in three-dimensional space; generate a gradient influence factor based on the material's sensitivity to stress concentration; use this factor to locally strengthen or weaken the equivalent stress amplitude, wherein the τ_max spatial gradient is calculated in MATLAB: perform Delaunay triangulation on the node cloud to obtain when When β is calculated, β = 0.03 × 15 = 0.45; the corrected equivalent stress σ_eq_a' = σ_eq_a × (1 + 0.45) = 522 MPa (an increase of 45% over the uncorrected value).

[0053] (6) The material property definition step includes a fatigue parameter estimation module, which is executed when experimental data is missing: calling the tensile strength limit value in the material database; selecting a preset strength conversion coefficient according to the metal material category; deriving the shear fatigue limit and stress-life curve characteristic parameters through the conversion relationship. If the SN data is missing, call the JMatPro material database: input A356 aluminum alloy σ_b = 310 MPa → calculate τ_f = 0.577 × 0.3σ_b = 53.6 MPa; generate the SN curve according to the ASTME739 standard (C = 1e12, k = 4.2), and take K_s = 0.9 for surface grinding.

[0054] (7) It also includes life visualization output: mapping the predicted torque cycles of each node to the surface of the three-dimensional model; generating a color distribution cloud map based on the life value range; marking the spatial coordinates of the early failure risk area in the cloud map, wherein the node N_fail value is imported into ParaView: setting the color spectrum mapping rule (>1e6 cycles blue, <5e5 cycles red); marking the first three failure points in the three-dimensional model (such as N_fail = 3.8e5 cycles at coordinates (35.2, -17.8, 48.6)); and outputting a PNG cloud map with positioning coordinates.

[0055] (8) The load spectrum loading step integrates accelerated equivalent processing: based on the fatigue damage consistency principle, the non-constant amplitude load spectrum is converted into an equivalent constant amplitude load cycle; the proportional coefficient between the actual load time history and the equivalent number of cycles is calculated; the coefficient is used as the acceleration correction factor for the life prediction result, wherein the rain flow counting is performed on the measured random torque spectrum (lasting 1800 seconds) to obtain the load block [ΔT_i,n_i]; through Miner equivalent calculation: when Σ(n_i / N_i)=1, T_eq=980Nm action N_eq=2.1e4 times is inferred; the output acceleration factor AF=1800 seconds / 2.1e4 times=0.0857 seconds / time.

[0056] (9) Establish an adaptive optimization loop: when the predicted number of torsional failure cycles is lower than the design target threshold, automatically generate a structural modification instruction set; adjust the geometric parameters of the key sections of the model according to the instruction set; trigger the full process to re-simulate and iterate until the life requirements are met. Among them, when the spline root N_fail=8.2e5<target value 2e6, trigger the optimization: based on the stress concentration factor K_t=3.1>2.5, increase the fillet radius from R1.0mm to R1.3mm; automatically update the CAD model and re-simulate, and after 3 rounds of iteration, N_fail=2.3e6 cycles meet the standard.

[0057] (10) The material property definition step integrates the manufacturing defect model: according to the casting or additive manufacturing process type of the component, the corresponding micro-defect statistical distribution function is selected; the stress-life curve attenuation characteristics of the high-cycle fatigue zone are corrected according to the defect distribution density; the fatigue strength reduction coefficient caused by the superposition defect is added to the stress concentration area, wherein, for the additively manufactured Ti6Al4V parts, the Weibull distribution corresponding to the powder melting process is selected (m = 1.8, η = 0.75); in the critical area (such as the porosity> 0.2% position), the SN curve slope is corrected: k is adjusted from 4.0 to 5.3; and the defect sensitive area life attenuation coefficient K_d = 0.62 is output.

[0058] Example 1:

[0059] Step 1: 3D parametric modeling and preprocessing

[0060] Operation content:

[0061] (1) Use CAD software (such as SolidWorks) to build an accurate geometric model of the target component (such as the gearbox output shaft), preserving all transition fillets (radius ≥ 0.5 mm) and spline tooth profile details;

[0062] (2) Import the finite element platform (such as ANSYS Workbench) through the intermediate format (.step or .iges)

[0063] Bench), apply mesh control: key areas (shoulder, spline root) use hexahedral dominant mesh with size ≤0.5mm; non-critical areas use tetrahedral mesh with size ≤2

[0064] mm;

[0065] (3) Check the mesh quality: Jacobian ratio ≥ 0.7, warping ≤ 15°.

[0066] Input: Engineering drawings / point cloud data

[0067] Output: Finite element model with meshing

[0068] Step 2: Definition of material fatigue properties

[0069] Operation content:

[0070] (1) Select the base material (such as 42CrMo steel) in the material library and define the elastic modulus as 210 GPa and the Poisson's ratio as 0.3;

[0071] (2) Enter measured fatigue parameters or activate the estimation module:

[0072] If experimental data is missing: input tensile strength σ_b=1080MPa→automatically calculate τ_f≈0.577×

[0073] 0.4σ_b=249MPa;

[0074] If there is an SN curve: Import the torsional fatigue data points with R = -1 (e.g. N = 1e6 cycles, τ_

[0075] a=300MPa);

[0076] (3) Specify the surface modification coefficient K_s: 0.92 for grinding and 1.15 for rolling strengthening;

[0077] (4) For casting / AM parts: Select the Weber defect model (modulus m = 1.2, characteristic strength η

[0078] =0.8).

[0079] Output: Material property set with defect corrections

[0080] Step 3: Torsional Load Spectrum Configuration

[0081] Operation content:

[0082] (1) Define the basic torque waveform:

[0083] Amplitude range T_a = 800 ~ 1200 Nm (obeying normal distribution);

[0084] Static bias T_m = 200 Nm;

[0085] Base frequency f_0 = 20 Hz (corresponding to rated engine speed);

[0086] (2) Superimpose high-frequency harmonics: add the second-order harmonic (f_1 = 40 Hz, amplitude ratio 15%);

[0087] (3) Enable acceleration equivalence: Set the target damage degree D = 1 → output equivalent constant amplitude load T_eq

[0088] =980Nm.

[0089] Output: Time domain-frequency domain load spectrum file

[0090] Step 4: Quasi-static stress field solution

[0091] Operation content:

[0092] (1) Apply unit torque T_ref = 1000 Nm and calculate the global stress field reference solution σ_ref;

[0093] (2) Decomposition of load spectrum: Split T_eq into static components [T_1 = 600 Nm,

[0094] T_2 = 380 Nm];

[0095] (3) Linear superposition: σ_total = (600 / 1000)·σ_ref1 + (380 / 1000)·σ_ref2;

[0096] (4) Output the node stress tensor matrix σ_ij (i, j = x, y, z).

[0097] Computational efficiency: 12 times faster than transient analysis (single solution of million-grid model < 5 minutes)

[0098] Step 5: Critical Plane Stress Extraction

[0099] Operation content:

[0100] (1) Based on the crystal slip system theory: for body-centered cubic materials (such as ferritic steel), set

[0101] {110} <111> slip system;

[0102] (2) Traverse the nodes: calculate the shear stress amplitude Δτ and the average normal stress σ_n on each slip surface;

[0103] (3) Identify the critical plane: take the plane with the largest product of Δτ×σ_n as the dangerous surface;

[0104] (4) Record the τ_max and τ_m values ​​on the plane.

[0105] Output: Dangerous surface stress parameter table

[0106] Step 6: Multi-axis equivalent stress conversion

[0107] Operation content:

[0108] (1) Call material parameters: τ_f = 249 MPa, α = 0.95 (experienced value for medium carbon steel);

[0109] (2) Calculate the baseline equivalent stress: σ_eq_a = τ_max / [1-(τ_m / τ_f)^0.95];

[0110] (3) Introducing gradient correction:

[0111] Calculate the three-dimensional gradient of τ_max in space

[0112] Modulus length

[0113] Calculate the correction factor (when Activated when

[0114] The final value σ_eq_a′=σ_eq_a×(1+β) is output.

[0115] Output: Array of node equivalent stress amplitudes

[0116] Step 7: Fatigue damage accumulation calculation

[0117] Operation content:

[0118] (1) According to the SN curve formula: N_i = C·(σ_eq_a')^(-k) (C = 1.2e12, k = 3.5);

[0119] (2) According to Miner's criterion: D = Σ(n_i / N_i), n_i is the number of times T_eq acts;

[0120] (3) Scan all nodes in the domain: mark them as failed when D ≥ 0.99 (failure threshold).

[0121] Output: node damage matrix D_map

[0122] Step 8: Life Visualization and Failure Location

[0123] Operation content:

[0124] (1) Map N_fail=1 / D to the model surface;

[0125] (2) Generate a color cloud map: the red area (N_fail < 1e5 weeks) is the high-risk area;

[0126] (3) Output risk coordinates: Extract the spatial positions of the top 10 nodes with the smallest N_fail (such as the XYZ coordinates at the R angle of the shoulder).

[0127] Tool: ParaView visualization platform

[0128] Step 9: Automatic Optimization Iteration

[0129] Operation content:

[0130] (1) Determine whether the lifespan meets the target: If the minimum N_fail < the design target (e.g., 2e6 cycles), then the optimization is triggered;

[0131] (2) Parameter adjustment rules:

[0132] If the stress concentration factor K_t>2.5: increase the fillet radius (ΔR=+0.3mm);

[0133] If the average damage degree D_avg>0.5: increase the wall thickness (Δt=+0.5mm);

[0134] (3) Automatically update the model and jump to step 1 and re-execute until N_fail_min ≥ 2e6 cycles.

[0135] Iterative example: After three rounds of optimization, the life of a shaft increases from 1.1e6 cycles to 2.3e6 cycles. Step 10: Generate a simulation report

[0136] Operation content:

[0137] (1) Output key data tables: minimum torque cycles, failure position, and equivalent load spectrum parameters;

[0138] (2) Attached optimization history: geometric changes and life improvement curves of each round;

[0139] (3) Generate reports in PDF format (compliant with ISO 18490 standards).

[0140] Finally, a few points should be explained: First, in the description of this application, it should be noted that, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense, and may refer to mechanical or electrical connections, internal communication between two components, or direct connection. "Up," "down," "left," and "right" are only used to indicate relative positional relationships. When the absolute positions of the objects being described change, the relative positional relationships may also change.

[0141] Secondly: The drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention. Other structures may refer to conventional designs. The same embodiment and different embodiments of the present invention may be combined with each other without conflict.

[0142] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for simulating and deducing the number of torque cycles, characterized in that Perform the following steps in order: S1. 3D parametric modeling: Based on the geometric features and assembly relationships of the target component, a digital 3D model including the topological structure is constructed and imported into the finite element analysis system. S2. Material property definition: Specify the material constitutive relations in the critical torque transmission region of the model, and define the fatigue property parameter set, which includes the shear strength limit, the torsional stress-life curve corrected by surface treatment, and the microstructure coefficient; S3. Load spectrum loading: Apply a cyclic torsional load spectrum that simulates actual operating conditions. This load spectrum characterizes the dynamic load characteristics through a preset torque amplitude fluctuation range, static torque offset, and cyclic frequency characteristics. S4. Quasi-static Stress Analysis: Performs finite element calculations based on the linear superposition principle, extracting multiaxial stress state tensors at all nodes of the model load step by load step. S5. Key Stress Extraction: Identify the maximum alternating shear stress component and its corresponding average shear stress component based on the stress tensor, taking into account the influence of material anisotropy; S6. Equivalent stress conversion: The alternating shear stress and the average shear stress are converted to uniaxial equivalent stress amplitudes using a multiaxial stress correction rule. This rule achieves stress state equivalence by correlating the material fatigue limit with the average stress sensitivity coefficient. S7. Fatigue Damage Accumulation: Based on the torsional stress-life curve and in combination with linear damage accumulation theory, calculate the fatigue damage degree under the load spectrum on a node-by-node basis; S8. Life span determination output: When the fatigue damage level of any node is monitored to reach the preset failure threshold, the predicted number of torque failure cycles and failure position mark corresponding to that position are output.

2. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: The construction of the torsional load spectrum specifically includes: generating a basic load signal with a fundamental frequency torque waveform as the main body; superimposing at least one high-frequency harmonic component to simulate mechanical vibration interference; and controlling the time-varying fluctuation range of the torque amplitude through the envelope.

3. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: The quasi-static stress analysis adopts the stress influence coefficient matrix method, which is specifically implemented by: applying a unit reference torque load to the model and calculating the global stress response reference field; decomposing the actual load spectrum into multiple static load components; The amplitude of each component is linearly weighted superimposed with the corresponding reference field to generate the comprehensive stress distribution field.

4. A method for simulating and deducing the number of torque cycles according to claim 1, characterized in that: The key stress extraction step further includes: determining the critical plane orientation prone to fatigue cracking based on the material crystal slip system theory; extracting the alternating shear stress amplitude range and the average normal stress component on the plane; and using the extraction results as the core input parameters for multi-axial stress correction.

5. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: The equivalent stress conversion step introduces a stress field gradient correction mechanism: calculating the rate of change vector of the maximum shear stress in three-dimensional space; generating a gradient influence factor based on the material's sensitivity to stress concentration; and using the factor to locally strengthen or weaken the equivalent stress amplitude.

6. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: The material property definition step includes a fatigue parameter estimation module, which is executed when experimental data is missing: calling the tensile strength limit value in the material database; selecting a preset strength conversion coefficient based on the metal material category; and deriving the shear fatigue limit and stress-life curve characteristic parameters through the conversion relationship.

7. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: It also includes life visualization output: mapping the predicted torque cycles of each node to the 3D model surface; generating a color distribution cloud map based on the life value range; and marking the spatial coordinates of the early failure risk area in the cloud map.

8. The method for simulating and deducing the number of torque cycles according to claim 1, characterized in that: The load spectrum loading step is integrated to accelerate the equivalent processing: based on the fatigue damage consistency principle, the non-constant amplitude load spectrum is converted into an equivalent constant amplitude load cycle; Calculate the proportionality coefficient between the actual load time history and the equivalent number of cycles; use this coefficient as the acceleration correction factor for the life prediction results.

9. The method for simulating and deducing the number of torque cycles according to claim 1, wherein: Establish an adaptive optimization loop: When the predicted number of torsional failure cycles is lower than the design target threshold, a structural modification instruction set is automatically generated; the geometric parameters of the key sections of the model are adjusted according to the instruction set; and the entire process is triggered to re-simulate and iterate until the life requirements are met.

10. The method for simulating and deducing the number of torque cycles according to claim 1, characterized in that: The material property definition step integrates a manufacturing defect model: selecting a corresponding micro-defect statistical distribution function based on the casting or additive manufacturing process type of the component; correcting the stress-life curve attenuation characteristics of the high-cycle fatigue zone based on the defect distribution density; and superimposing a fatigue strength reduction factor caused by defects in the stress concentration area.

Images