A method and system for characterizing tire low temperature rolling resistance using an rsa stretch mode deformation scanning test

Abstract

The present application relates to the technical field of tire rubber performance testing, and particularly relates to a method and system for testing low-temperature rolling resistance of tires by using RSA tensile mode deformation scanning. The method comprises pre-cycling, constant temperature and humidity control, peak automatic positioning, cross-validation and error checking, and can predict R 2 up to 0.97 or more, and the error of the sampling inspection is not greater than 5%. The system is composed of an RSA testing subsystem, an environment and clamp assembly, a data processing module, a model engine and a database, and can significantly shorten the testing period and reduce the cost without preparing a whole tire.

Description

Technical Field

[0001] This invention relates to the field of tire rubber performance testing technology, and in particular to a method and system for characterizing the low-temperature rolling resistance of tires using RSA tensile mode deformation scanning test. Background Technology

[0002] Tire rolling resistance is a key indicator affecting vehicle fuel economy and driving range. Especially in low-temperature environments, changes in the dynamic mechanical properties of rubber materials can significantly alter rolling resistance characteristics. Traditional rolling resistance testing relies on real-vehicle road tests or drum tests, requiring the preparation of complete tire samples. This process is time-consuming (usually 2-3 days) and costly, making it difficult to meet the rapid screening needs of tire compound development.

[0003] my country has established a standardized system for relevant testing methods to unify measurement conditions and the correlation between laboratory results. A commonly used standard is GB / T29040-2012, "Automobile Tire Rolling Resistance Test Method: Single-Point Test and Correlation of Measurement Results," which is adopted from ISO28580 and is widely used in engineering practice. In materials science, the industry has long used the loss factor obtained from dynamic mechanical analysis as a proxy indicator for rolling resistance, especially using tanδ at a single temperature point close to the operating temperature range for rapid formulation screening. Furthermore, Chinese patent CN101792545A clearly states that the lower the tanδ of the tread compound at approximately 60°C, the lower the tire rolling resistance, and provides a typical numerical range to guide tread formulation design and evaluation.

[0004] Recent formulation modification patents (see Chinese patents CN116160804B and CN116141876A) mostly evaluate fuel efficiency indicators at a fixed temperature in DMA and verify them using the drum method to measure tire rolling resistance according to ISO28580, reflecting the empirical correspondence between DMA single-point tanδ and overall tire rolling resistance. For example, tanδ is measured at a fixed temperature and strain amplitude and converted into a fuel efficiency index, while the rolling resistance coefficient measured by ISO28580 is used for comparison and verification. In addition, formulation routes to reduce tread hysteresis loss (such as solution-polymerized styrene-butadiene rubber and silica / silane systems) are used in Chinese patent CN106633258A to achieve low rolling resistance targets, and tanδ at a fixed temperature point is usually used as a proxy for screening and optimization.

[0005] In summary, existing technologies have provided: 1. Standardized rolling resistance indoor measurement devices and methods; 2. A conventional development paradigm using the fixed-temperature point DMA loss factor as a proxy index for rolling resistance; and 3. A verification path for empirically correlating the DMA index with actual test results from the entire tire drum. However, for rolling resistance in low-temperature environments, existing disclosures regarding how to establish a stable and verifiable quantitative linear mapping between the characteristic parameters extracted through deformation scanning and low-temperature rolling resistance under specific test modes and conditions, along with a clear statistical threshold and quality control process, remain insufficient and have room for improvement. Summary of the Invention

[0006] The technical objective of this invention is to provide a rapid laboratory characterization method based on RSA tensile mode deformation scanning. By obtaining the maximum tanδ value of the tread rubber at 41℃ and establishing a stable linear relationship with the measured rolling resistance of the drum under low-temperature conditions (such as -30℃ and -7℃), this method can achieve accurate prediction and rapid screening of low-temperature rolling resistance without the need to prepare a complete tire. This significantly shortens the R&D cycle, reduces testing costs, and improves the reliability and engineering guidance value of the evaluation results under repeatable quality control and statistical threshold constraints.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for characterizing the low-temperature rolling resistance of tires using RSA tensile mode deformation scanning test, the method comprising the following steps:

[0009] S1. Sample preparation: Take samples from the tire tread rubber to prepare dumbbell-shaped samples with a thickness of 2 mm, an effective gauge length of 20 mm, and a narrow area width of 4 to 6 mm. Roughen or sand the surface of the clamping area to prevent slippage.

[0010] S2. Pretreatment and temperature control: The sample is mounted on the tensile fixture of the RSA dynamic mechanical analyzer and subjected to 20 to 50 pre-cycles at 25℃, 1 Hz, and 0.1% strain. Then, the temperature is raised to 41℃ and held at a constant temperature for no less than 10 minutes, with temperature fluctuations not exceeding ±0.2℃.

[0011] S3. Deformation scanning and acquisition: Strain scanning is performed at 41℃ and 1 Hz, with a strain range of 0.01% to 5%. Gauge length closed-loop control is used to ensure that the strain control error is no greater than ±1%. Force, displacement and phase difference are acquired synchronously.

[0012] S4. Peak Extraction: The tanδ curve with strain is denoised and smoothed. The peak value is located by the joint criterion that the first derivative is zero and the second derivative is less than zero. The maximum value of tanδ at 41℃ is obtained and the average value is taken in three independent repetitions. The standard deviation of repeatability is no greater than 0.01.

[0013] S5. Modeling: Establish a linear relationship between the maximum tanδ value at 41℃ for different formulation samples as the independent variable and the rolling resistance of the corresponding whole tire measured under low-temperature conditions in the drum test as the dependent variable. The coefficient of determination R squared is not less than 0.97, and the low temperature condition is -30℃ to 10℃.

[0014] S6. Prediction and Verification: For unknown formulas, obtain independent variables according to steps S3 to S4 and substitute them into the model obtained in step S5 to calculate low-temperature rolling resistance. When the sampled samples are retested in a drum, the relative error between the predicted value and the measured value is no greater than 5%.

[0015] As a preferred option, the low-temperature operating conditions include at least two points: -30℃ and -7℃. Independent linear models are established for each point, and the average R-squared values ​​of the cross-validation are not less than 0.97 and 0.98, respectively.

[0016] Preferably, the strain upper limit is adaptively set based on the boundary of the linear viscoelastic region at 41°C. The criterion for entering nonlinearity is to monitor that the proportion of the third harmonic does not exceed 1% of the main harmonic. Once the limit is exceeded, the strain upper limit is automatically reduced and the measurement is repeated. And / or, the fixture has an online monitoring function for the normal pressure at the clamping end and the relative sliding displacement. When the relative sliding is detected to exceed five micrometers, the measurement is terminated and an alarm is triggered.

[0017] As a preferred option, compliance and phase calibration are performed on the instrument before modeling, and zero-point and gain dual calibration is performed using the standard tanδ value of the reference viscoelastic material at 41℃ and 1 Hz. After calibration, the reference deviation is no greater than ±1%.

[0018] As a preferred approach, linear modeling employs ordinary least squares combined with three-fold or higher cross-validation, while simultaneously performing residual normality and homoscedasticity tests. Data that fail the tests are not included in the modeling process. When new samples cause R-squared to fall below the threshold or the sampling error exceeds 5%, the model recalibration process is triggered.

[0019] Furthermore, the present invention also provides the application of the method in rapid screening during the tire formulation development stage, particularly for quantitative prediction and formulation ranking of rolling resistance in environments ranging from −30℃ to 10℃ by obtaining the maximum tanδ value through deformation scanning in a 41℃ stretching mode without preparing a complete tire.

[0020] Furthermore, the present invention also provides a rapid characterization system for low-temperature rolling resistance for implementing the method, the system comprising:

[0021] The RSA test subsystem is used to perform strain scanning at a frequency of 1 Hz in tensile mode and achieve gauge length closed-loop control.

[0022] Environmental control and clamping components are used to provide a constant temperature of 41°C and anti-slip clamping, and to monitor clamping pressure and sliding displacement;

[0023] The data acquisition and processing module is used to acquire mechanical and phase data and execute peak localization algorithms to extract the maximum value of tanδ at 41℃;

[0024] The model engine is used to build and store linear models, calculate R-squared, perform cross-validation and residual testing, and issue recalibration prompts when the standards are not met.

[0025] The database is used to store raw experimental data, calibration records, model parameters, and prediction results.

[0026] The human-computer interaction interface is used for task setting, curve display, and result output.

[0027] As a preferred option, the data acquisition and processing module employs a peak extraction algorithm based on the joint criteria of derivative and curvature, combined with moving average or Savitzky-Golay filtering for noise reduction, with peak repeatability deviation not exceeding 0.02% of the upper limit of strain;

[0028] And / or, the model engine has two independent linear models for low-temperature conditions, corresponding to -30℃ and -7℃ respectively, and records the acceptable range of their slope and intercept. If the range is exceeded, batch prediction is prohibited and additional samples or recalibration is required.

[0029] Furthermore, the present invention also provides a computer-readable storage medium storing a program that, when executed by a processor, causes a computer to perform peak extraction, linear regression modeling, cross-validation, error assessment, recalibration triggering, and the generation and storage of prediction results in the method.

[0030] Furthermore, the present invention also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement peak extraction, linear regression modeling, cross-validation, error assessment, recalibration triggering, and generation and storage of prediction results in the method.

[0031] This invention, by employing the aforementioned technical solution, achieves rapid quantitative prediction of tire low-temperature rolling resistance at the material level by performing deformation scanning at 41℃ in RSA tensile mode, extracting the maximum tanδ value, and establishing a linear model with the low-temperature drum rolling resistance. Compared to the traditional whole-tire drum method, the test object is reduced from a whole tire to a small sample, and a single measurement and modeling can be completed within approximately one hour, significantly shortening the R&D cycle and reducing sample and tooling costs. Under quality control conditions such as unified fixtures, temperature control, and pre-cycling, the prediction results show a high correlation with the measured rolling resistance at -30℃ and -7℃. It can stably reach the preset threshold), and the relative error of the sampled samples is controlled within the target range, thereby supporting rapid sorting and agile iteration of different tread formulations. At the same time, the method is based on a linear mapping of a single independent variable, which is convenient for cross-project reuse and database accumulation. It can also maintain the long-term stability and engineering transferability of the model through cross-validation and recalibration mechanisms. Attached Figure Description

[0032] Figure 1 The relationship between rolling resistance at ambient temperatures of -30℃ and -7℃ and the maximum value of tanδ at 41℃ is given. Detailed Implementation

[0033] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present invention.

[0034] I. System Overall Structure

[0035] A rapid characterization system for low-temperature rolling resistance for implementing the method of the present invention includes:

[0036] 1. RSA testing subsystem: It features tensile mode, strain closed loop and 1 Hz frequency control, and real-time gauge length feedback;

[0037] 2. Environmental control and clamping components: The constant temperature chamber or environmental hood achieves temperature control at 41℃ and relative humidity control at 40% to 60%; anti-slip clamps monitor clamping pressure and relative slippage at the ends;

[0038] 3. Data acquisition and processing module: synchronously acquires force, displacement and phase difference, and outputs energy storage modulus, loss modulus and tanδ curve;

[0039] 4. Model Engine: Completes linear regression modeling, cross-validation, R-squared and residual tests, error assessment, and recalibration triggering;

[0040] 5. Database: Stores raw experimental data, calibration records, model parameters, prediction results, and quality status;

[0041] 6. Human-Machine Interface: Used for setting test plans, displaying curve and scatter plots, and outputting predicted values ​​and alarm information.

[0042] II. Methods, Steps, and Key Parameters

[0043] S1. Sample Preparation

[0044] Samples were taken from the tire tread compound and cut into dumbbell-shaped specimens: 2 mm thick, 20 mm effective gauge length, and 4 to 6 mm narrow section width. The clamping area was roughened with sandpaper or sandblasted to ensure a stable clamping friction coefficient. The samples were shaped and numbered according to the GB / T528 procedure and allowed to stand for at least 2 hours.

[0045] S2, Pretreatment and Temperature Control

[0046] The specimen was clamped in an RSA tensile fixture and pre-cycled 20 to 50 times at 25°C, 1 Hz, and 0.1% strain to eliminate the effects of clamping and stress history. The temperature was then raised to 41°C and held at this temperature for at least 10 minutes, with temperature fluctuations not exceeding ±0.2°C. Relative humidity was controlled between 40% and 60%.

[0047] S3, Deformation Scanning and Acquisition

[0048] Strain scanning was performed at 41℃ and 1 Hz, with a strain range of 0.01% to 5%. Gauge length displacement closed-loop control was used to ensure that the strain control error was no greater than ±1%. Force, displacement, and phase difference were recorded in real time, and the tanδ curve as a function of strain was calculated.

[0049] S4, Peak Extraction and Repeatability

[0050] The tanδ curve was denoised using a moving average or Savitzky-Golay filter. The peak position was located by a joint criterion that the first derivative is zero and the second derivative is less than zero, and the maximum tanδ value at 41℃ was obtained. Three independent measurements were repeated, and the mean was taken; the standard deviation of repeatability was no greater than 0.01.

[0051] S5, Modeling and Thresholding

[0052] Collect the independent variable x (maximum tanδ at 41℃) of at least three different tread compound samples, and the measured dependent variable y of the rolling resistance of the corresponding whole tire under low-temperature conditions in the drum test, and establish a linear model. .

[0053] Preferably, such as Figure 1 As shown, two independent models are established for -30℃ and -7℃ respectively:

[0054] y(-30℃) = 16.05x + 8.6642, R squared is 0.9714;

[0055] y(-7℃) = 15.959x + 5.5264, R squared is 0.9852.

[0056] When undertaking new projects, the average R-squared value for cross-validation is required to be no less than 0.97 and 0.98, respectively.

[0057] S6. Prediction and Verification

[0058] For the unknown formula, obtain x according to S3 to S4, and substitute it into the corresponding linear model to output the predicted value of low-temperature rolling resistance. Perform drum retesting on samples from each batch. The relative error between the predicted value and the measured value should not exceed 5%; if it exceeds the limit or R-squared drops below the threshold, recalibration is triggered.

[0059] III. Quality Control and Calibration

[0060] Criterion for linear viscoelastic region: If the proportion of the third harmonic does not exceed 1% of the main harmonic, it is considered to still be in the linear viscoelastic region; if it exceeds the limit, the strain upper limit will be automatically reduced and the test will be repeated.

[0061] Fixture stability: The system monitors the normal pressure at the clamping end and the relative slippage at the end online. When the slippage exceeds five micrometers, an alarm is automatically triggered and the data is discarded.

[0062] Instrument calibration: Zero point and gain are calibrated daily using the standard tanδ of a reference viscoelastic material at 41℃ and 1 Hz, with a reference deviation of no more than ±1%.

[0063] Statistical tests: Linear regression uses ordinary least squares and at least three-fold cross-validation; residuals must satisfy approximate normality and homoscedasticity, and those that fail to meet these requirements will not be included in the model.

[0064] IV. Data Processing Algorithm Flow

[0065] A) Synchronization and phase correction of the original signal;

[0066] B) Noise reduction and smoothing;

[0067] C) Automatic peak optimization: Initial location of derivative zero, confirmation of peak characteristics by negative second derivative value; refinement of peak position and peak value by quadratic polynomial fitting in the neighborhood;

[0068] D) Repeatability evaluation and outlier removal;

[0069] E) Regression modeling and cross-validation;

[0070] F) Predicted output, error assessment, and recalibration criteria.

[0071] V. Examples and Comparative Examples

[0072] Example 1: Modeling and Validation of Three Formulations (41℃, Tensile Mode, Deformation Scan)

[0073] 1. Samples and Equipment

[0074] Formulas 1, 2, and 3 were used to prepare dumbbell-shaped tread compounds according to GB / T528; RSA-G2 was used in a constant temperature chamber; the rolling resistance test was conducted at -30℃ and -7℃ according to GB / T29040 / ISO28580.

[0075] 2. RSA test parameters

[0076] Pre-cycle: 25℃, 1Hz, 0.1% strain, 30 cycles; Isothermal: 41℃, equilibration for 12 min; Scan: 1Hz, 0.01%–5% strain, gauge length closed loop.

[0077] 3. Verification of the maximum value of tanδ and rolling resistance at 41℃

[0078] The mean value was taken after three independent replicates, with repeatability SD ≤ 0.01. A linear model was established using the data from the three formulations and validated using their respective measured rolling resistances.

[0079] Table 1 shows the maximum tanδ value and low-temperature rolling resistance under tensile conditions at 41℃.

[0080]

[0081] Note: The prediction formula is obtained by scatter plot fitting.

[0082]

[0083]

[0084] The results showed that the relative error between prediction and actual measurement was within 0.05% for the three representative formulations (and remained stable at ≤5% during batch verification). If the preset threshold is met, the method can be used to make highly correlated quantitative predictions of low-temperature rolling resistance at the material level.

[0085] Example 2: Adaptive Upper Limit of Linear Viscoelastic Region

[0086] For a high-filler formulation, the proportion of the third harmonic exceeds 1% in the strain range of 4%-5%. The system automatically lowers the upper limit to 3.5% and remeasures. The peak position is stable. After adding modeling, the overall R² remains above 0.98. If the upper limit is not lowered, the peak position shift of 0.006–0.010 causes the prediction error to rise to about 6% (see Comparative Example 2).

[0087] Example 3: Batch Filtering Scenario

[0088] New formulations A, B, and C were introduced, yielding maximum tanδ values ​​of 0.188, 0.205, and 0.230 at 41℃, respectively. The rolling resistance at -7℃ was directly predicted to be 8.53, 8.79, and 9.20 using the model from Example 1; retesting of the sampled drum yielded values ​​of 8.58, 8.83, and 9.16, with relative errors of 0.58%, 0.45%, and 0.44%. The three formulations were tested and ranked within one day, significantly shortening the R&D cycle.

[0089] Comparative Example 1: tanδ at a single point at 35℃ as a characteristic quantity

[0090] Without deformation scanning, only a fixed strain of 0.5% was measured at 35℃ and 1Hz; the corresponding tanδ values ​​for the three formulations were 0.247, 0.166, and 0.262. Leave-one-out cross-validation (LOOCV) was used to evaluate the predictive performance after establishing a linear relationship.

[0091] Table 2 shows the prediction error (LOOCV) of the single-point method at 35℃.

[0092]

[0093] The method exhibited a maximum error of approximately 4.94% in its −7℃ prediction, overall... The stability is inferior to the 41°C deformation scanning peak method of the present invention; and the migration capability is insufficient for different rubber systems.

[0094] Comparative Example 2: Linear viscoelastic region criterion not enabled

[0095] Maintaining all conditions of Example 1, but without monitoring the third harmonic proportion and without limiting the upper limit of strain. For the highly filled sample, the tanδ curve exhibits nonlinearity in the high strain region, with the peak position being raised by approximately 0.008–0.012, causing the relative error predicted by the -7℃ model to increase to 6%–8%. Decreased to 0.94–0.95; recovered to ≤5% after enabling the criterion and .

[0096] Comparative Example 3: Compression mode replaces stretching mode

[0097] Three formulations were tested at 41℃ using compression mode and the same strain procedure. A linear model was established by taking the tanδ peak value and cross-validated. Due to differences in fixture compliance and boundary conditions, the peak repeatability SD increased to 0.02-0.03, and the mean absolute error of the prediction at -7℃ was 6.7%. In comparison, the present invention exhibits SD≤0.01 and mean absolute error≤3% in the stretching mode. .

[0098] Comparative Example 4: Temperature scan peak value replaces deformation scan peak value

[0099] Peak finding was performed in temperature scanning mode (1Hz, -20℃→60℃ heating rate of 2℃ / min), and the peak value near 41℃ was taken as the characteristic value. Due to the coupling effect of thermal hysteresis and temperature range, the peak position drift was 0.005-0.009, and the average prediction error for -30℃ was approximately 5.8%. The present invention eliminates the effect of thermal hysteresis by performing deformation scanning at an isothermal temperature of 41°C, with an error ≤5%. .

[0100] VI. Summary of Results

[0101] This application uses the maximum tanδ value extracted under isothermal conditions at 41℃, tensile mode, and deformation scanning as a single independent variable, and can establish a stable linear relationship between rolling resistance at -30℃ / -7℃ across different formulations. The results reached 0.9714 and 0.9852 respectively, and the sampling error could be stably controlled within the target of 5%. However, if the 35℃ single-point method, compression mode, or temperature scanning peak finding is used, repeatability and correlation decrease due to factors such as linear region disruption, fixture compliance, and thermal hysteresis. When the value drops to the 0.90-0.95 range, the maximum error rises to approximately 6%-8%.

[0102] This invention ensures the long-term stability of the model through quality control measures such as linear viscoelastic region criteria, fixture slip monitoring, daily dual calibration and cross-validation; in batch screening, prediction and sorting can be completed in hours, which is significantly better than the cycle and cost of the whole tire drum test.

[0103] The foregoing description of embodiments of the present invention, through which those skilled in the art are able to implement or use the present invention, will be readily apparent to those skilled in the art. Various modifications to these embodiments will be readily apparent to those skilled in the art. The general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novelty disclosed herein.

Claims

1. A method for characterizing the low-temperature rolling resistance of tires using RSA tensile mode deformation scanning test, characterized in that, The method includes the following steps: S1. Sample preparation: Take samples from the tire tread rubber to prepare dumbbell-shaped samples with a thickness of 2 mm, an effective gauge length of 20 mm, and a narrow area width of 4 to 6 mm. Roughen or sand the surface of the clamping area to prevent slippage. S2. Pretreatment and temperature control: The sample is mounted on the tensile fixture of the RSA dynamic mechanical analyzer and subjected to 20 to 50 pre-cycles at 25℃, 1 Hz, and 0.1% strain. Then, the temperature is raised to 41℃ and held at a constant temperature for no less than 10 minutes, with temperature fluctuations not exceeding ±0.2℃. S3. Deformation scanning and acquisition: Strain scanning is performed at 41℃ and 1 Hz, with a strain range of 0.01% to 5%. Gauge length closed-loop control is used to ensure that the strain control error is no greater than ±1%. Force, displacement and phase difference are acquired synchronously. S4. Peak Extraction: The tanδ curve with strain is denoised and smoothed. The peak value is located by the joint criterion that the first derivative is zero and the second derivative is less than zero. The maximum value of tanδ at 41℃ is obtained and the average value is taken in three independent repetitions. The standard deviation of repeatability is no greater than 0.

01. S5. Modeling: The maximum value of tanδ at 41℃ for different formulation samples is taken as the independent variable, and the rolling resistance of the corresponding whole tire measured in the low temperature condition of the drum method is taken as the dependent variable to establish a linear relationship. The coefficient of determination R square is not less than 0.

97. The low temperature condition is -30℃ to 10℃. S6. Prediction and Verification: For unknown formulas, obtain independent variables according to steps S3 to S4 and substitute them into the model obtained in step S5 to calculate low-temperature rolling resistance. When the sampled samples are retested in a drum, the relative error between the predicted value and the measured value is no greater than 5%.

2. The method according to claim 1, characterized in that, The low-temperature operating conditions include at least two points: -30℃ and -7℃. Independent linear models are established for each point, and the average R-squared of the cross-validation is not less than 0.97 and 0.98, respectively.

3. The method according to claim 1, characterized in that, The strain upper limit is adaptively set based on the boundary of the linear viscoelastic region at 41℃. The criterion for entering nonlinearity is to monitor that the proportion of the third harmonic does not exceed 1% of the main harmonic. Once the limit is exceeded, the strain upper limit is automatically reduced and the measurement is repeated. And / or, the fixture has an online monitoring function for the normal pressure at the clamping end and the relative sliding displacement. When the relative sliding is detected to exceed five micrometers, the measurement is terminated and an alarm is triggered.

4. The method according to claim 1, characterized in that, Before modeling, compliance and phase calibration were performed on the instrument, and zero-point and gain dual calibration were performed using the standard tanδ value of the reference viscoelastic material at 41℃ and 1 Hz. After calibration, the reference deviation was no greater than ±1%.

5. The method according to claim 1, characterized in that, Linear modeling employs ordinary least squares combined with three-fold or higher cross-validation, while also performing residual normality and homoscedasticity tests. Data that fail the tests are not included in the modeling process. When new samples cause R-squared to fall below the threshold or the sampling error exceeds 5%, the model recalibration process is triggered.

6. The use of the method of any one of claims 1 to 5 in rapid screening during the tire formulation development stage, particularly for quantitative prediction and formulation ranking of rolling resistance in environments from -30 ℃ to 10 ℃ by obtaining the maximum value of tan δ through deformation scanning in a 41 ℃ tensile mode without preparing a complete tire.

7. A rapid characterization system for low-temperature rolling resistance for implementing the method of any one of claims 1 to 5, characterized in that, The system includes: The RSA test subsystem is used to perform strain scanning at a frequency of 1 Hz in tensile mode and achieve gauge length closed-loop control. Environmental control and clamping components are used to provide a constant temperature of 41°C and anti-slip clamping, and to monitor clamping pressure and sliding displacement; The data acquisition and processing module is used to acquire mechanical and phase data and execute peak localization algorithms to extract the maximum value of tanδ at 41℃; The model engine is used to build and store linear models, calculate R-squared, perform cross-validation and residual testing, and issue recalibration prompts when the standards are not met. The database is used to store raw experimental data, calibration records, model parameters, and prediction results. The human-computer interaction interface is used for task setting, curve display, and result output.

8. The system according to claim 7, characterized in that, The data acquisition and processing module employs a peak extraction algorithm based on the joint criteria of derivative and curvature, combined with moving average or Savitzky-Golay filtering for noise reduction, with peak repeatability deviation not exceeding 0.02% of the upper limit of strain. And / or, the model engine has two independent linear models for low-temperature conditions, corresponding to -30℃ and -7℃ respectively, and records the acceptable range of their slope and intercept. If the range is exceeded, batch prediction is prohibited and additional samples or recalibration is required.

9. A computer-readable storage medium storing a program that, when executed by a processor, causes a computer to perform peak extraction, linear regression modeling, cross-validation, error assessment, recalibration triggering, and generation and storage of prediction results as described in any one of claims 1 to 5.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the processor, they implement the peak extraction, linear regression modeling, cross-validation, error assessment, recalibration triggering, and generation and storage of prediction results as described in any one of claims 1 to 5.

Images