Calculation system for tire snowfield performance evaluation

Abstract

The invention discloses a calculation system for tire snowfield performance evaluation, and belongs to the technical field of tire performance detection. According to the system, structured light scanning and dynamic mechanical analysis (DMA) equipment are integrated through a CATIA secondary development interface, and a multi-source data fusion framework including geometric feature extraction, rubber material physical property dynamic modeling and machine learning prediction is constructed. The key technology comprises the steps that an ICP algorithm is improved to achieve that the registration error of pattern point cloud and a CAD model is smaller than or equal to 0.1 mm; the temperature-sensitive model corrects rubber material parameters in real time, and the error is + / -2% at-30 DEG C to 0 DEG C; the three-layer neural network is used for predicting the performance score of the snowfield, and the correlation reaches 0.95 or above; a discrete element method simulates snow and tire effects, and a snow discharge efficiency index is output. The time consumption of system single-scheme evaluation is less than or equal to 1.5 hours, the prediction error is less than or equal to + / -6%, the efficiency is improved by 30 times compared with the traditional real tire test, the cost is reduced by 96%, the method is suitable for quick optimization of snowfield performance in the tire research and development stage, and the tire design efficiency and precision related to winter driving safety can be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of spraying technology, in particular to a calculation system for evaluating tire snow performance. Background Art

[0002] The tire tread is in direct contact with the ground. The performance of the tread pattern and tread compound is directly related to the tire's snow shedding, cold resistance, and wet grip. This is especially true in winter snowy conditions, where the tire's snow shedding and wet grip play a crucial role in vehicle driving safety. Tread pattern design and compound selection are extremely important in the tire product design process. Existing tire snow performance evaluation faces three major technical bottlenecks:

[0003] Geometric parameter distortion: Traditional CAD modeling ignores pattern microstructure (such as 0.2mm chamfers), resulting in a calculation error of the transverse groove projection length of ±0.5mm;

[0004] Static rubber parameter: Manual input of rubber hardness value does not take into account the temperature effect (hardness can vary by 40% between -30℃ and 0℃), resulting in performance prediction deviation of more than ±8%;

[0005] Insufficient environmental adaptability: A single formula cannot cover wet snow (water content > 30%), compacted snow (density > 0.6g / cm 3 ) and other 12 snow conditions, with the prediction error fluctuating by ±15%.

[0006] The industry urgently needs an intelligent evaluation system that integrates geometric accuracy improvement, dynamic calibration of physical properties and environmental adaptation.

[0007] Based on this, a calculation system for evaluating tire snow performance is now provided, which can eliminate the disadvantages of existing devices. Summary of the Invention

[0008] The purpose of the present invention is to provide a calculation system for evaluating tire snow performance, which solves the problem of inconvenience in use in the prior art.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] A calculation system for evaluating tire snow performance, comprising:

[0011] Data acquisition layer: obtain pattern 3D model data through CAD software interface, and integrate structured light scanner to obtain pattern surface point cloud data;

[0012] Analysis and calculation layer: includes point cloud registration module, rubber material property dynamic modeling module, machine learning prediction module and multi-physics field coupling simulation module;

[0013] Result output layer: Generates a dynamic report including snow performance scores and pattern parameter cloud maps; the system achieves collaborative calculation of pattern geometric characteristics, rubber material properties and environmental parameters through multi-source data fusion.

[0014] On the basis of the above technical solutions, the present invention also provides the following optional technical solutions:

[0015] In an optional solution, the point cloud registration module adopts an improved ICP algorithm, and the correction formula is:

[0016]

[0017] where p i is the point cloud coordinate of the pattern surface (unit: mm), acquired by a structured light scanner;

[0018] q i is the theoretical coordinate of the corresponding point in the CAD model (unit: mm);

[0019] R is the rotation matrix, which is used to adjust the spatial angle between the point cloud and the model;

[0020] t is the translation vector, which is used to adjust the spatial position of the point cloud and the model;

[0021] The registration error between point cloud and CAD model is ≤0.1mm.

[0022] In an optional solution, the dynamic modeling module for rubber compound properties is directly connected to the DMA device to collect the storage modulus E′ and loss factor tanδ in the range of -30°C to 0°C in real time, and establish a temperature-sensitive model:

[0023]

[0024] where K i (T) is the rubber correction coefficient at temperature T, The standard temperature coefficient is -15℃.

[0025] In an optional solution: the machine learning prediction module uses a three-layer fully connected neural network, the input layer contains 12-dimensional pattern geometric features and 5-dimensional environmental parameters, the output layer is the snow performance score P, the model prediction correlation R 2 ≥0.95.

[0026] In an optional solution, the multi-physics coupling simulation module simulates the collision process between snow particles and patterns based on the discrete element method (DEM), calculates the snow removal efficiency index (SRI), and uses the thermal-mechanical coupling formula:

[0027] ΔH=0.5×μ×F×v

[0028] Real-time correction of the effect of tread temperature on rubber hardness, where μ is the friction coefficient, F is the vertical load, and v is the vehicle speed.

[0029] In an optional solution, the system integrates a hardware-in-the-loop test module, obtains snow force feedback data through a six-component force testing machine, and uses an error back propagation algorithm:

[0030]

[0031] Automatically adjust model parameters with a calibration period of ≤30 minutes, achieving a reduction of ≥2% in annual system prediction error;

[0032] Δθ is the model parameter adjustment amount;

[0033] η is the learning rate (artificially set optimization step size);

[0034] is the gradient of the loss function (indicating the changing trend of the current prediction error).

[0035] In an optional solution, the structured light scanner has an accuracy of ≤0.05 mm and can obtain high-density point cloud data of ≥500,000 points / pattern, which is used to extract microstructural features such as the pattern edge radius r and the groove curvature k, and to modify the formula for the transverse groove projection length:

[0036] L 修正 is the corrected transverse groove projection length (unit: mm);

[0037] L 测量 The projected length of the transverse groove directly measured from the CAD model (unit: mm);

[0038] 0.2 is the microstructure influence coefficient (obtained by fitting experimental data).

[0039] In an optional solution, the environmental parameters include a snow condition sextuple S = {T, H, D, V, C, M} consisting of temperature, snow depth, humidity, compaction, and contamination. The system automatically adjusts the calculation formula using a weight matrix:

[0040] P=∑(L i ×W i ×K i ×ω j )

[0041] where ω j is the snow condition weight factor, covering 12 typical snow conditions, with a prediction error fluctuation of ≤±6%;

[0042] L i is the projected length of the i-th transverse groove (unit: mm);

[0043] Wi is the width of the i-th transverse groove (unit: mm);

[0044] K i is the rubber correction coefficient for the i-th region;

[0045] w i is the weight factor of the jth snow condition (determined by orthogonal test).

[0046] In an optional solution: the dynamic report includes a pattern performance cloud map with a resolution of ≤0.5mm×0.5mm, a full life cycle performance attenuation curve, and basic indicators such as transverse groove projection length (error ±0.1mm), steel sheet density (error ±2%), etc.

[0047] In one optional solution: the system is deployed on the CATIA V5 R28 platform, and CAD model analysis is achieved through the CAA secondary development interface. The evaluation of a single solution takes ≤1.5 hours, and supports simultaneous comparison of ≥10 pattern solutions.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] Significantly improved accuracy: Through structured light scanning and an improved ICP algorithm, the error in pattern geometric parameters has been reduced from ±0.5mm to ±0.1mm, and the degree of microstructure restoration has been increased by 60%. Dynamic temperature calibration has been introduced into the physical properties of the rubber, and the parameter error in the range of -30°C to 0°C has been narrowed from ±8% to ±2%. The correlation between snow performance prediction and actual vehicle testing has reached over 92%.

[0050] Significantly improved efficiency: The evaluation time for a single solution has been reduced from 48 hours using traditional methods to 1.5 hours. Simultaneous comparison of more than 10 tread patterns is supported, shortening the R&D cycle by more than 70%. No actual tire production is required, reducing the cost of a single evaluation by 96%.

[0051] Enhanced environmental adaptability: A six-tuple snow condition model was constructed, covering 12 typical scenarios, and the prediction error fluctuation was narrowed from ±15% to ±6%. Integrated hardware-in-the-loop calibration automatically reduced the system's annual prediction error by 2%, achieving self-evolution.

[0052] Ease of use and scalability: Developed on the CATIA platform, the operating steps have been simplified from more than 20 to 5, making it suitable for non-professionals. It supports the expansion of the rubber database and the customization of new pattern units, and is applicable to the entire range of winter tires. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION

[0054] 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without creative work are within the scope of protection of the present invention.

[0055] Hardware configuration;

[0056]

[0057]

[0058] Data collection phase

[0059] Step 1: 3D modeling of the pattern

[0060] Using the Generative Surface Design module in CATIA, a tire tread pattern model was constructed, including main grooves, transverse grooves, steel plates, and microstructures (such as 0.2mm chamfers).

[0061] The output is a CATPart format file containing the precise coordinate data of each pattern feature (accuracy 0.01mm).

[0062] Step 2: Structured light scanning point cloud acquisition

[0063] Fix the tire model on the scanning turntable and start scanning the tread surface with an Artec Eva scanner at a 0.1mm pitch.

[0064] The scanned data was processed through Artec Studio: noise removal → automatic stitching of multi-view point clouds → generation of triangulated mesh models, which were then exported in STL format (containing over 500,000 point cloud data).

[0065] Step 3: Dynamic physical property test of rubber compound

[0066] Take the tread rubber sample (size Φ25mm×2mm) and put it into the DMA equipment fixture;

[0067] Set up a temperature sweep program: increase the temperature from -30°C to 0°C at a rate of 2°C / min, hold each temperature point for 5 minutes, apply a 1 Hz sinusoidal load, and collect the storage modulus \(E'\) and loss factor \(\tan\delta\);

[0068] After the test is completed, the temperature-property curve is generated and imported into the system temperature-property database.

[0069] Step 4: Snow condition parameter input

[0070] Enter the six-tuple of snow conditions in the system interface:

[0071] Temperature (T = -20 ° C), snow depth (H = 5 cm\), humidity (D = 80\%\),

[0072] Compaction degree (V=70\%\), contamination degree (M=10%) (ice content).

[0073] Model calculation stage

[0074] Step 1: Align point cloud with CAD model

[0075] Load the STL point cloud file and CAD model in CATIA and start the point cloud registration module;

[0076] Automatically match feature points using the improved ICP algorithm:

[0077] Initial matching: roughly align the point cloud and the model through principal component analysis (PCA);

[0078] Iterative optimization: Calculate the minimum Euclidean distance between the point cloud and the corresponding points of the model, and update the rotation matrix R and translation vector t until the registration error is ≤ 0.1 mm;

[0079] Output the registered pattern model for subsequent geometric feature extraction.

[0080] Step 2: Microstructure feature extraction

[0081] Use the CATIA "Shape Analysis" module to automatically identify the pattern edge radius r and groove curvature k:

[0082] Edge radius r: Use the curvature analysis tool to detect the pattern edge transition area and take the average value of 5 sampling points;

[0083] Groove curvature k: Select a 10mm length curve at the bottom of the groove and calculate the curvature by fitting a quadratic function

[0084] Corrected cross groove projection length: Corrected measurement Example: If the length measurement L 测量 =40mm, edge radius r = 0.3mm, groove curvature k = 0.5 / mm, then the corrected length is:

[0085] Step 3: Calculate the rubber correction coefficient based on the current snow temperature ℃T = -20℃, call the temperature sensitivity model: ℃℃K i (-20) = K -15℃ ×(1-0.03·|-20+15|)=K -15℃ ×0.85(assuming ℃K -15℃ =0.9, then K i (-20) = 0.765);

[0086] Step 4: Machine Learning Prediction Input the following data into the neural network model: Geometric features: transverse groove length (41.7mm), steel sheet density (20 pieces / 100mm 2 ), groove depth (8mm) and other 12-dimensional data; environmental parameters: ℃T = -20℃, H = 5cm, D = 80% and other 5-dimensional data; the model output snow performance score P = 78.5 (out of 100), snow removal efficiency index SRI = 0.72.

[0087] Step 5: Multi-physics coupled simulation Run DEM simulation in MATLAB: Generate 100,000 snow particles (particle size 2-5 mm, density 0.4 g / cm 3 ), impacting the pattern surface at a speed of 10m / s; calculating the depth distribution of snow particles invading the grooves, and statistically analyzing the snow removal efficiency: the number of retained snow particles and the total number of particles Thermal-mechanical coupling calculation of tread temperature rise: ΔH = 0.5 × 0.3 × 5000 N × 10 m / s = 7500 J (assuming friction coefficient μ = 0.3, vertical load F = 5000 N, vehicle speed v = 10 m / s)

[0088] Closed-loop calibration phase

[0089] Step 1: Hardware-in-the-loop test: Install the tire to be evaluated on a six-component force testing machine and perform a braking test on a -20°C snow test bench. Measure the braking distance and S 实测 =35.2m; System predicted braking distance S 预测 =38.5m, calculation error:

[0090] Step 2: Model parameter adjustment starts the error back propagation algorithm: (Learning rate η=0.01, loss function prediction and measurement L=(S 预测 -S 实测 ) 2 ); automatically adjust the neural network weights, and re-predict after 15 minutes to get the prediction S 预测 =35.8m, and the error is reduced to +1.7%.

[0091] Calibration and maintenance mechanism:

[0092] 1. Daily calibration: Use a standard tread block (known transverse groove length L = 50.00mm, edge radius r = 0.5mm) to perform point cloud registration test: After scanning, calculate the registration error: If the error is greater than 0.1%, recalibrate the structured light scanner lens parameters.

[0093] 2. Weekly training: Collect actual vehicle test data (≥50 groups) for that week, including tread parameters, snow conditions, and measured performance indicators; perform incremental training on the neural network in TensorFlow, update the model weights, and ensure the prediction correlation R 2 ≥0.95.

[0094] 3. Annual maintenance upgrade of the DEM simulation particle count to 1 million to improve the accuracy of snow particle collision simulation; calibrate the DMA equipment temperature sensor to ensure that the temperature measurement error in the range of -30℃ to 0℃ is ≤0.5℃;

[0095]

[0096] The system of the present invention significantly improves the accuracy and efficiency of tire snow performance evaluation through multi-source data fusion and intelligent algorithms, meeting the rapid iteration requirements in the research and development stage.

[0097] 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 calculation system for evaluating tire snow performance, characterized by: include: Data acquisition layer: obtain pattern 3D model data through CAD software interface, and integrate structured light scanner to obtain pattern surface point cloud data; Analysis and calculation layer: includes point cloud registration module, rubber material property dynamic modeling module, machine learning prediction module and multi-physics field coupling simulation module; Result output layer: Generates a dynamic report including snow performance scores and pattern parameter cloud maps; the system achieves collaborative calculation of pattern geometric characteristics, rubber material properties and environmental parameters through multi-source data fusion.

2. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The point cloud registration module adopts the improved ICP algorithm, and the correction formula is: where p i is the point cloud coordinate of the pattern surface, acquired by a structured light scanner; q i are the theoretical coordinates of the corresponding points in the CAD model; R is the rotation matrix, which is used to adjust the spatial angle between the point cloud and the model; t is the translation vector, which is used to adjust the spatial position of the point cloud and the model; The registration error between point cloud and CAD model is ≤0.1mm.

3. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The dynamic modeling module of the rubber compound physical properties is directly connected to the DMA equipment to collect the storage modulus E' and loss factor tanδ in the range of -30℃~0℃ in real time, and establish a temperature-sensitive model: where K i (T) is the rubber correction coefficient at temperature T, The standard temperature coefficient is -15℃.

4. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The machine learning prediction module adopts a three-layer fully connected neural network. The input layer contains 12-dimensional pattern geometric features and 5-dimensional environmental parameters. The output layer is the snow performance score P and the model prediction correlation R. 2 ≥0.

95.

5. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The multi-physics coupling simulation module simulates the collision process between snow particles and patterns based on the discrete element method, calculates the snow removal efficiency index, and uses the thermal-mechanical coupling formula: ΔH=0.5×μ×F×v Real-time correction of the effect of tread temperature on rubber hardness, where μ is the friction coefficient, F is the vertical load, and v is the vehicle speed.

6. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The system integrates a hardware-in-the-loop test module, obtains snow force feedback data through a six-component force testing machine, and uses an error back propagation algorithm: Automatically adjust model parameters with a calibration period of ≤30 minutes, achieving a reduction of ≥2% in annual system prediction error; Δθ is the model parameter adjustment amount; η is the learning rate (artificially set optimization step size); is the gradient of the loss function (indicating the changing trend of the current prediction error).

7. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The structured light scanner has an accuracy of ≤0.05mm and can obtain high-density point cloud data of ≥500,000 points / pattern, which is used to extract microstructural features such as pattern edge radius r and groove curvature k, and to correct the formula for transverse groove projection length: L 修正 is the corrected transverse groove projection length; L 测量 The transverse groove projection length directly measured from the CAD model; 0.2 is the microstructure influence coefficient.

8. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The environmental parameters include the snow condition sextuple S = {T, H, D, V, C, M} consisting of temperature, snow depth, humidity, compaction, and contamination. The system automatically adjusts the calculation formula through the weight matrix: P=∑(L i ×W i ×K i ×ω j ) where ω j is the snow condition weight factor, covering 12 typical snow conditions, with a prediction error fluctuation of ≤±6%; L i is the projection length of the i-th transverse groove; W i is the width of the i-th transverse groove; K i is the rubber correction coefficient for the i-th region; w i is the weight factor of the jth snow condition.

9. The computing system according to claim 1, wherein: The dynamic report includes a pattern performance cloud map with a resolution of ≤0.5mm×0.5mm, a performance attenuation curve for the entire life cycle, and basic indicators such as transverse groove projection length and steel sheet density.

10. The calculation system for evaluating tire snow performance according to claim 1, characterized in that: The system is deployed on the CATIA V5 R28 platform and implements CAD model analysis through the CAA secondary development interface. The evaluation of a single scheme takes ≤1.5 hours and supports the simultaneous comparison of ≥10 pattern schemes.

Images