Non-intrusive epoxy casting curing residual strain testing and calibrating method

Abstract

The invention discloses a non-intrusive epoxy casting curing residual strain test and calibration method, which comprises the following steps: S1, preparing a dumbbell-shaped sample in combination with an actual basin-type insulator epoxy alumina composite material formula and a casting curing process; s2, testing strain sensitivity coefficient calibration of the Bragg grating aiming at the dumbbell-shaped sample to obtain wavelength and strain test data; s3, testing temperature sensitivity coefficient calibration of the Bragg grating according to the dumbbell-shaped sample to obtain test data of wavelength and temperature; and S4, respectively establishing Bayesian regression models of the strain sensitivity coefficient and the temperature sensitivity coefficient by using a Bayesian regression method, and realizing Bayesian regression to obtain a mean value and a confidence interval of the strain sensitivity coefficient and the temperature sensitivity coefficient. Accurate calibration of the strain and temperature sensitivity coefficient is realized, and the method has important significance in accurately measuring the strain and ensuring the quality of parts.

Description

Technical Field

[0001] The present invention belongs to the field of quality control of insulating parts of high-voltage switchgear, and in particular relates to a non-invasive epoxy casting curing residual strain testing and calibration method. Background Art

[0002] Epoxy castings are widely used in aerospace, automotive, electronics and other industries. As an important high-performance composite material, they have excellent mechanical properties and thermal stability. However, epoxy resin often produces residual strain during the curing process, which affects its subsequent performance. During the curing process, due to changes in temperature and humidity and uneven distribution of reactive substances, curing strain is inevitable. These residual strains may cause deformation of the casting and even affect the stability of its overall structure. Therefore, accurately measuring the curing residual strain of epoxy castings is crucial to evaluating material properties and ensuring component quality.

[0003] At present, the commonly used residual strain test methods mainly include physical cutting method, mechanical testing method and optical measurement method. However, the physical cutting method will damage the material and affect the accuracy and reliability of the test results; the mechanical testing method requires the installation of strain sensors on the specimen, which may introduce external interference and cannot achieve global and real-time monitoring; although the optical measurement method can provide more accurate surface deformation information, it is a big challenge due to the complex structure of epoxy castings. Traditional methods generally have problems such as strong invasiveness, low measurement accuracy, and great influence on the material during the test process, making it difficult to achieve efficient and accurate testing of curing residual strain. At the same time, for the calibration of coefficients in the strain function relationship, the traditional average method usually only provides a single point estimate, ignoring data variability and possible errors, resulting in the calibration results being overly simplified and lacking the quantification of uncertainty.

[0004] In order to overcome the shortcomings of existing methods in testing and calibrating residual strain of epoxy castings during curing, a non-invasive testing and calibration method for residual strain of epoxy castings during curing is proposed, which has not been publicly reported yet. Summary of the invention

[0005] The purpose of the present invention is to propose a non-invasive epoxy casting curing residual strain testing and calibration method. The method uses a Bragg grating sensor to test the strain of the epoxy casting curing process. The test data uses the Bayesian regression method to calibrate the coefficients in the strain function relationship, and calculates the posterior mean as the best estimate, as well as a 95% confidence interval to quantify the uncertainty of the estimate, which fills the gap in the existing methods.

[0006] In one aspect of the present invention, the present invention provides a non-invasive epoxy casting curing residual strain testing and calibration method. According to an embodiment of the present invention, the method comprises the following steps:

[0007] S1. Combined with the actual formula of epoxy-alumina composite material for pot-type insulator and the pouring and curing process, dumbbell-shaped specimens were prepared;

[0008] S2. Test the calibration of strain sensitivity coefficient of Bragg grating on dumbbell-shaped specimens to obtain test data of wavelength and strain;

[0009] S3. Calibrate the temperature sensitivity coefficient of the Bragg grating for the dumbbell-shaped sample and obtain the test data of wavelength and temperature;

[0010] S4. Based on the test data obtained in S2 and S3, the Bayesian regression model of the strain sensitivity coefficient and the temperature sensitivity coefficient was established respectively using the Bayesian regression method to obtain the mean value and confidence interval of the strain sensitivity coefficient and the temperature sensitivity coefficient.

[0011] In addition, the non-invasive epoxy casting curing residual strain testing and calibration method according to the above embodiment of the present invention may also have the following additional technical features:

[0012] In some embodiments of the present invention, in step S1, the epoxy-aluminum oxide composite material formula includes epoxy resin, curing agent, and aluminum oxide filler.

[0013] In some embodiments of the present invention, in step S2, the dumbbell-shaped specimen is placed at the same temperature and subjected to different limit loads to obtain test data of wavelength and strain.

[0014] In some embodiments of the present invention, step S2 is specifically as follows: a Bragg grating is arranged in the middle of a dumbbell-shaped specimen, and a tensile test is performed on a high and low temperature tensile tester. During the measurement process, the internal temperature of the high and low temperature test chamber is kept constant, and an extensometer is used to measure the strain of the grating area under different limit loads of the dumbbell-shaped tensile specimen, and its wavelength is recorded.

[0015] In some embodiments of the present invention, in step S3, the dumbbell-shaped specimen is placed at different temperatures and subjected to the same limit load to obtain test data of wavelength and temperature.

[0016] In some embodiments of the present invention, step S3 is specifically as follows: keeping the limit load of the dumbbell-shaped tensile specimen unchanged, raising the internal temperature of the high and low temperature test box, measuring the strain at different temperatures respectively, and recording its wavelength.

[0017] In some embodiments of the present invention, in step S4:

[0018] The Bayesian regression model of the strain sensitivity coefficient is as follows:

[0019] Δλ=K ε Δε+ε1

[0020] Where Δλ represents the change in the optical fiber reflection wavelength; K ε represents the temperature sensitivity coefficient, nm / ε; Δε represents the measured strain change / ε; ε1 represents the error term in the Bayesian regression model of the strain sensitivity coefficient, satisfying ε1~N(0,σ1 2 )’s normal distribution;

[0021] The Bayesian regression model of the temperature sensitivity coefficient is as follows:

[0022] Δλ=K T ΔT+ε2

[0023] In the formula, Δλ represents the change in the optical fiber reflection wavelength, K T represents the temperature sensitivity coefficient / (K / ε), ΔT represents the measured temperature change / K, ε2 represents the error term in the Bayesian regression model of the temperature sensitivity coefficient, and satisfies ε2~N(0,σ2 2 ) is a normal distribution.

[0024] In some embodiments of the present invention, the mean and confidence interval of the strain sensitivity coefficient and the temperature sensitivity coefficient are calculated as follows:

[0025] (1) are strain sensitivity coefficients K ε , Temperature sensitivity coefficient K T Set the prior distribution and set K ε , K T Follows normal distribution;

[0026] (2) are strain sensitivity coefficients K ε , K T Calculate the posterior distribution, substitute the test data of strain and wavelength, and the test data of temperature and wavelength into the likelihood function respectively. According to Bayes' theorem, the posterior distribution is the product of the prior distribution and the likelihood function.

[0027] (3) A series of sample values ​​in the posterior distribution results of the strain sensitivity coefficient and the temperature sensitivity coefficient are sampled using the Hamiltonian Monte Carlo method, and the posterior means of the strain sensitivity coefficient and the temperature sensitivity coefficient are calculated as the best estimate, as well as a 95% confidence interval.

[0028] In some embodiments of the present invention, in step (1):

[0029] K ε It obeys normal distribution, denoted as Strain sensitivity coefficient K ε The probability density function is as follows:

[0030]

[0031] Where f1(x) represents the strain sensitivity coefficient K ε Probability density function, x represents the strain sensitivity coefficient K ε The possible values ​​of σ ε Indicates the strain sensitivity coefficient K ε The prior standard deviation, μ ε Indicates the strain sensitivity coefficient K ε The prior mean, σ ε 2 Indicates the strain sensitivity coefficient K ε The variance of the prior;

[0032] K T It obeys normal distribution, denoted as Temperature sensitivity coefficient K T The probability density function is as follows:

[0033]

[0034] Where f2(x) represents the temperature sensitivity coefficient K T Probability density function, x represents the temperature sensitivity coefficient K T The possible values ​​of σ T Indicates the temperature sensitivity coefficient K T The prior standard deviation, μ T Indicates the temperature sensitivity coefficient K T The prior mean, σ T 2 Indicates the temperature sensitivity coefficient K T The variance of the prior.

[0035] In some embodiments of the present invention, in step (2):

[0036] Strain sensitivity coefficient K ε The posterior distribution of is as follows:

[0037] P(K ε |Δλ,Δε)∝P(Δλ|K ε ,Δε)·P(K ε )

[0038] In the formula, P(K ε |Δλ,Δε) is the strain sensitivity coefficient K ε The posterior distribution of P(Δλ|K ε ,Δε) is the likelihood function, which means N(Δλ|K ε Δε,σ12 ), P(K ε ) is the probability corresponding to the probability density function f1(x).

[0039] Temperature sensitivity coefficient K T The posterior distribution of is as follows:

[0040] P(K T |Δλ,ΔT)∝P(Δλ|K T ,ΔT)·P(K T )

[0041] In the formula, P(K T |Δλ,ΔT) is the temperature sensitivity coefficient K T The posterior distribution of P(Δλ|K T ,ΔT) is the likelihood function, which means N(Δλ|K T ΔT,σ2 2 ), P(K T ) is the probability corresponding to the probability density function f2(x).

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] (1) The present invention provides a non-invasive testing method that can measure the residual strain during the curing process of epoxy castings without causing physical damage to them, thereby avoiding the measurement deviation and inconvenience caused by sample damage in traditional methods, making the testing process safer, simpler and more efficient.

[0044] (2) The present invention uses a Bragg grating sensor to test the strain of epoxy castings during the curing process, which can capture the residual strain generated by the epoxy castings during the curing process in real time and accurately. Compared with traditional testing methods, it has higher accuracy and reliability.

[0045] (3) The present invention uses the Bayesian regression method to calibrate the coefficients in the strain function relationship, which can effectively integrate the uncertainty in the measurement data and provide a more accurate calibration result by combining prior knowledge with test data.

[0046] (4) The present invention uses Bayesian regression to not only provide point estimates of coefficients, but also generate probability distributions of coefficients. Compared with the traditional coefficient averaging method, Bayesian regression can reduce the bias caused by data noise or measurement errors and avoid the potential risks caused by ignoring uncertainty in traditional methods.

[0047] (5) The residual strain testing and calibration method proposed in the present invention can be widely used in the testing and calibration of other large epoxy castings, providing a new technical solution for related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 The Bragg grating test principle diagram of the non-invasive epoxy casting curing residual strain test in the embodiment of the present invention is shown;

[0049] Figure 2 This is a flow chart of a non-invasive epoxy casting curing residual strain testing and calibration method in an embodiment of the present invention;

[0050] Figure 3 Schematic diagram of the sample pouring process in step S1 in an embodiment of the present invention;

[0051] Figure 4 This is a schematic diagram of the sample curing process in step S1 in an embodiment of the present invention;

[0052] Figure 5 This is a schematic diagram of the layout of the dumbbell-shaped sample in step S1 in an embodiment of the present invention;

[0053] Figure 6 This is a flow chart of calibrating the strain sensitivity coefficient and the temperature sensitivity coefficient using the Bayesian regression method in step S4 in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0055] like Figure 2 As shown, a non-invasive epoxy casting curing residual strain testing and calibration method comprises the following steps:

[0056] S1. Based on the actual formula of epoxy-alumina composite material for pot-type insulator and the casting and curing process, dumbbell-shaped specimens were prepared.

[0057] The composite sample material is an epoxy resin system used for UHV basin insulators: CT5531Cl (solid at room temperature, solvent-free bisphenol A epoxy resin) and HT903Cl (carboxylic acid anhydride curing agent), the mass ratio of the two is 1:0.38, and the filler used is alumina filler for electrical use produced by China Aluminum Corporation. Figure 3As shown in the figure, according to the actual application of the UHV basin insulator process, the epoxy resin and alumina filler for the sample are fully mixed and degassed, and then the preheated curing agent is added to the mixture and stirred, and finally the final mixture is injected into the mold. The mold is cleaned with ethanol and coated with a release agent in advance. The mold is preheated at 100°C for more than 5 hours before pouring. After pouring, the sample is sent to the oven for curing. Figure 4 As shown, the residual strain monitoring starts at the same time.

[0058] like Figure 1 As shown, the Bragg grating used in the measurement is made on an ordinary quartz optical fiber for communication. The fiber diameter is 125m, the grating length is 12mm, the central wavelength is 1550nm and 1555nm, and the reflectivity is greater than 99%. A 30mm optical fiber is left at the end of the grating for fixing to the opening of the end cover of the mold. The sample is a cylindrical sample with a length of 70mm and a diameter of 80mm. The container for the reaction is an aluminum mold with a central opening on both side end covers so that the optical fiber can be buried along the central axis of the sample. The grating is arranged in the middle of the sample. Figure 5 As shown, an unpackaged FBG sensor, i.e., a fiber Bragg grating sensor, is arranged in a dumbbell-shaped tensile specimen mold with through holes at both ends to measure strain and wavelength data.

[0059] S2. Carry out the strain sensitivity coefficient calibration test of Bragg grating on dumbbell-shaped specimens and obtain the test data of wavelength and strain.

[0060] The Bragg grating was arranged in the middle of the dumbbell-shaped tensile specimen, and the tensile test was carried out on the high and low temperature tensile tester. During the measurement, the internal temperature of the high and low temperature test box was set to 20°C. The extensometer was used to measure the grating area. 10%, 20%, 30%, and 40% of the ultimate load were applied to the dumbbell-shaped tensile specimens 1, 2, 3, and 4, respectively, and their wavelengths λ1, λ2, λ3, and λ4 were recorded. The results are shown in Table 1.

[0061] Table 1 Strain and wavelength of samples at the same temperature and different limit loads

[0062]

[0063] S3. Carry out the temperature sensitivity coefficient calibration test of Bragg grating on dumbbell-shaped samples and obtain the test data of wavelength and temperature.

[0064] Keep the dumbbell-shaped tensile specimen at 40% of the ultimate load, increase the temperature inside the high and low temperature test chamber, read the wavelength and temperature when the temperature is stable, and measure the corresponding wavelengths at 5 temperatures. The results are shown in Table 2.

[0065] Table 2 Temperature and wavelength of samples under the same limit load

[0066]

[0067] S4. Based on the test data obtained from S2 and S3, the Bayesian regression model of the strain sensitivity coefficient and the temperature sensitivity coefficient were established respectively using the Bayesian regression method, and the mean and confidence interval of the strain sensitivity coefficient and the temperature sensitivity coefficient were obtained using Bayesian regression.

[0068] The change in wavelength is a linear function of the change in temperature and the change in strain, and the relationship is shown in the following equation:

[0069] Δλ=K ε Δε+K T ΔT

[0070] In the formula, Δλ represents the change in the optical fiber reflection wavelength, K ε represents the temperature sensitivity coefficient / (nm / ε), Δε represents the measured strain change / ε, K T represents the temperature sensitivity coefficient / (K / ε), and ΔT represents the measured temperature change / K. Based on this relationship, the strain sensitivity coefficient and the temperature sensitivity coefficient can be made independent of each other by controlling the test variables, and calibrated separately to reduce the influence of other variables in the test on the calibration and improve the accuracy of the coefficient calibration.

[0071] like Figure 6 As shown, the method of calibrating the strain sensitivity coefficient and the temperature sensitivity coefficient using Bayesian regression specifically includes the following steps:

[0072] 1) Establish a Bayesian regression model of strain sensitivity coefficient, and its model expression is:

[0073] Δλ=K ε Δε+ε1

[0074] In the formula, Δλ represents the change in the optical fiber reflection wavelength, K ε represents the temperature sensitivity coefficient / (nm / ε), Δε represents the measured strain change / ε, and ε1 represents the error term in the Bayesian regression model of the strain sensitivity coefficient, satisfying ε1~N(0,σ1 2 ) is a normal distribution.

[0075] 2) is the strain sensitivity coefficient K ε Set the prior distribution and set K ε It obeys normal distribution, denoted as Satisfy the following formula:

[0076]

[0077] Where f1(x) represents the strain sensitivity coefficient K εProbability density function, x represents the strain sensitivity coefficient K ε The possible values ​​of σ ε Indicates the strain sensitivity coefficient K ε The prior standard deviation, μ ε Indicates the strain sensitivity coefficient K ε The prior mean, σ ε 2 Indicates the strain sensitivity coefficient K ε The variance of the prior.

[0078] 3) is the strain sensitivity coefficient K ε Calculate the posterior distribution and substitute the strain and wavelength test data into the likelihood function. According to Bayes' theorem, the posterior distribution is the product of the prior distribution and the likelihood function, that is:

[0079] P(K ε |Δλ,Δε)∝P(Δλ|K ε ,Δε)·P(K ε )

[0080] In the formula, P(K ε |Δλ,Δε) is the strain sensitivity coefficient K ε The posterior distribution of P(Δλ|K ε ,Δε) is the likelihood function, which means N(Δλ|K ε Δε,σ1 2 ), P(K ε ) is the probability corresponding to the probability density function f1(x).

[0081] 4) Establish a Bayesian regression model of temperature sensitivity coefficient, and its model expression is:

[0082] Δλ=K T ΔT+ε2

[0083] In the formula, Δλ represents the change in the optical fiber reflection wavelength, K T represents the temperature sensitivity coefficient / (K / ε), ΔT represents the measured temperature change / K, ε2 represents the error term in the Bayesian regression model of the temperature sensitivity coefficient, and satisfies ε2~N(0,σ2 2 ) is a normal distribution.

[0084] 5) is the temperature sensitivity coefficient K T Set the prior distribution and set K T It obeys normal distribution, denoted as Satisfy the following formula:

[0085]

[0086] Where f2(x) represents the temperature sensitivity coefficient K TProbability density function, x represents the temperature sensitivity coefficient K T The possible values ​​of σ T Indicates the temperature sensitivity coefficient K T The prior standard deviation, μ T Indicates the temperature sensitivity coefficient K T The prior mean, σ T 2 Indicates the temperature sensitivity coefficient K T The variance of the prior.

[0087] 6) is the temperature sensitivity coefficient K T Calculate the posterior distribution and substitute the test data of temperature and wavelength into the likelihood function. According to Bayes' theorem, the posterior distribution is the product of the prior distribution and the likelihood function, that is:

[0088] P(K T |Δλ,ΔT)∝P(Δλ|K T ,ΔT)·P(K T )

[0089] In the formula, P(K T |Δλ,ΔT) is the temperature sensitivity coefficient K T The posterior distribution of P(Δλ|K T ,ΔT) is the likelihood function, which means N(Δλ|K T ΔT,σ2 2 ), P(K T ) is the probability corresponding to the probability density function f2(x).

[0090] 7) A series of sample values ​​in the posterior distribution results of the strain sensitivity coefficient and the temperature sensitivity coefficient are sampled using the Hamiltonian Monte Carlo (HMC) method, and the posterior mean is calculated as the best estimate, as well as a 95% confidence interval to quantify the uncertainty of the estimate.

[0091] Substituting the wavelength and strain test data obtained by S2 into the Bayesian regression method, the mean strain sensitivity coefficient calculated by the HMC method is 1.083×10 -3 nm / ε, with a 95% confidence interval of [1.075×10 -3 ,1.091×10 -3 ]nm / ε.

[0092] The test data of wavelength and temperature obtained by S3 were substituted into the Bayesian regression method, and finally the mean value of the strain sensitivity coefficient calculated by the HMC method was 0.01055 K / ε, and the 95% confidence interval was [0.01047, 0.01063] K / ε.

[0093] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the present invention or exceed the scope defined by the claims, they shall all fall within the protection scope of the present invention.

Claims

1. A non-invasive epoxy casting curing residual strain testing and calibration method, characterized in that: The following steps are involved: S1. Combined with the actual formula of epoxy-alumina composite material for pot-type insulator and the pouring and curing process, dumbbell-shaped specimens were prepared; S2. Test the calibration of strain sensitivity coefficient of Bragg grating on dumbbell-shaped specimens to obtain test data of wavelength and strain; S3. Calibrate the temperature sensitivity coefficient of the Bragg grating for the dumbbell-shaped sample and obtain the test data of wavelength and temperature; S4. Based on the test data obtained in S2 and S3, the Bayesian regression model of the strain sensitivity coefficient and the temperature sensitivity coefficient was established respectively using the Bayesian regression method to obtain the mean value and confidence interval of the strain sensitivity coefficient and the temperature sensitivity coefficient.

2. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 1, characterized in that: In step S1, the epoxy-alumina composite material formula includes epoxy resin, curing agent, and alumina filler.

3. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 1, characterized in that: In step S2, the dumbbell-shaped specimen is placed at the same temperature and different limit loads are applied to obtain test data of wavelength and strain.

4. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 3, characterized in that: Step S2 is specifically as follows: a Bragg grating is arranged in the middle of a dumbbell-shaped specimen, and a tensile test is performed on a high and low temperature tensile tester. During the measurement process, the temperature inside the high and low temperature test chamber is kept constant. An extensometer is used to measure the strain of the grating area under different limit loads of the dumbbell-shaped tensile specimen, and its wavelength is recorded.

5. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 1, characterized in that: In step S3, the dumbbell-shaped specimen is placed at different temperatures and the same limit load is applied to obtain test data of wavelength and temperature.

6. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 5, characterized in that: Step S3 is specifically as follows: keeping the limit load of the dumbbell-shaped tensile specimen unchanged, raising the temperature inside the high and low temperature test box, measuring the strain at different temperatures respectively, and recording the wavelength.

7. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 1, characterized in that: In step S4: The Bayesian regression model of the strain sensitivity coefficient is as follows: Δλ=K ε De+e1 Where Δλ represents the change in the optical fiber reflection wavelength; K ε represents the temperature sensitivity coefficient, nm / ε; Δε represents the measured strain change / ε; ε1 represents the error term in the Bayesian regression model of the strain sensitivity coefficient, satisfying ε1~N(0,σ1 2 )’s normal distribution; The Bayesian regression model of the temperature sensitivity coefficient is as follows: Δλ=K T ΔT+ε2 In the formula, Δλ represents the change in the optical fiber reflection wavelength, K T represents the temperature sensitivity coefficient / (K / ε), ΔT represents the measured temperature change / K, ε2 represents the error term in the Bayesian regression model of the temperature sensitivity coefficient, and satisfies ε2~N(0,σ2 2 ) is a normal distribution.

8. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 7, characterized in that: The mean and confidence interval of the strain sensitivity coefficient and temperature sensitivity coefficient are calculated as follows: (1) are strain sensitivity coefficients K ε , Temperature sensitivity coefficient K T Set the prior distribution and set K ε , K T Follows normal distribution; (2) are strain sensitivity coefficients K ε , K T Calculate the posterior distribution, substitute the test data of strain and wavelength, and the test data of temperature and wavelength into the likelihood function respectively. According to Bayes' theorem, the posterior distribution is the product of the prior distribution and the likelihood function. (3) A series of sample values ​​in the posterior distribution results of the strain sensitivity coefficient and the temperature sensitivity coefficient are sampled using the Hamiltonian Monte Carlo method, and the posterior means of the strain sensitivity coefficient and the temperature sensitivity coefficient are calculated as the best estimate, as well as a 95% confidence interval.

9. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 8, characterized in that: In step (1): K ε It obeys normal distribution, denoted as Strain sensitivity coefficient K ε The probability density function is as follows: Where f1(x) represents the strain sensitivity coefficient K ε Probability density function, x represents the strain sensitivity coefficient K ε The possible values ​​of σ ε Indicates the strain sensitivity coefficient K ε The prior standard deviation, μ ε Indicates the strain sensitivity coefficient K ε The prior mean, σ ε 2 Indicates the strain sensitivity coefficient K ε The variance of the prior; K T It obeys normal distribution, denoted as Temperature sensitivity coefficient K T The probability density function is as follows: Where f2(x) represents the temperature sensitivity coefficient K T Probability density function, x represents the temperature sensitivity coefficient K T The possible values ​​of σ T Indicates the temperature sensitivity coefficient K T The prior standard deviation, μ T Indicates the temperature sensitivity coefficient K T The prior mean, σ T 2 Indicates the temperature sensitivity coefficient K T The variance of the prior.

10. A non-invasive epoxy casting curing residual strain testing and calibration method according to claim 9, characterized in that: In step (2): Strain sensitivity coefficient K ε The posterior distribution of is as follows: P(K ε |Δλ,Δε)∝P(Δλ|K ε ,D)·P(K ε ) In the formula, P(K ε |Δλ,Δε) is the strain sensitivity coefficient K ε The posterior distribution of P(Δλ|K ε ,Δε) is the likelihood function, which means N(Δλ|K ε Δε,σ1 2 ), P(K ε ) is the probability corresponding to the probability density function f1(x); Temperature sensitivity coefficient K T The posterior distribution of is as follows: P(K T |Δλ,ΔT)∝P(Δλ|K T ,ΔT)·P(K T ) In the formula, P(K T |Δλ,ΔT) is the temperature sensitivity coefficient K T The posterior distribution of P(Δλ|K T ,ΔT) is the likelihood function, which means N(Δλ|K T ΔT,σ2 2 ), P(K T ) is the probability corresponding to the probability density function f2(x).

Images