CN112255113B - High-temperature elastic constant measuring method for thin strip - Google Patents

Abstract

The invention discloses a high-temperature elastic constant measuring method for a thin strip, which comprises the following steps: (1) Preparing a flat high-temperature tensile sample, and preparing high-temperature speckles on the surface of the sample; (2) Respectively acquiring a reference image and a loading image of a plate type high-temperature tensile sample; (3) Identifying a high reliability square sub-region in the reference image; (4) Comparing the reference image with the loaded image by using a digital image correlation method, and calculating the horizontal displacement and the vertical displacement of the high-reliability sub-region after loading; (5) And obtaining the elastic modulus and the poisson ratio of the sample material by weighted least square fitting based on the calculated horizontal displacement, the vertical displacement and the corresponding load of the high-reliability sub-region after loading. The method can not only overcome the problem of reduced correlation coefficient of the image partial region caused by the decorrelation effect, but also effectively eliminate the randomness error caused by the high-temperature heat flow disturbance, and can finish the high-efficiency and accurate measurement of the high-temperature elastic constant.

Description

A Method for Measuring High Temperature Elastic Constant of Thin Strip

technical field

The invention belongs to the technical field of measurement, and in particular relates to a high-temperature elastic constant measurement method for thin strips.

Background technique

Strips and foils are one of the metallurgical products of deformed superalloys, which have the advantages of good toughness, high service temperature, and corrosion resistance. In the modern aerospace industry, components such as aero-engine turbine cooling ducts and turbine casing seals made of foil materials need to withstand high temperatures above 800°C under working conditions, and high-temperature materials used in these structures will be subject to thermal deformation. Due to the coupling effect with mechanical deformation, the service life is seriously reduced, so it is very important to ensure its service reliability. Elastic constants such as elastic modulus and Poisson's ratio reflect the stiffness of the material and are very important properties for design. The measurement methods of elastic constant are mainly divided into dynamic method and static method. The dynamic method uses resonance, ultrasonic and other methods to complete the measurement of the elastic constant of the material, but it is difficult to obtain the dynamic response of the ultra-thin material. The static method uses the transverse and longitudinal strain of the material in the quasi-static mechanical test to measure the elastic constant, and the deformation measurement is an important part of the static method test. In the high temperature environment above 1000 ℃, the application of the traditional contact method will be greatly restricted, the specific performance is as follows: the high temperature extensometer has a short life and is expensive, and cannot measure small gauge length and large deformation conditions; high temperature strain gauge It is difficult to paste in a high temperature environment, the reliability is poor, and the price is expensive, resulting in high testing costs. Non-contact deformation measurement methods are usually based on optical principles, and have the advantages of full-field measurement, high precision, and easy automation, and have good application prospects in high-temperature testing. The optical non-contact method is mainly divided into an interference method and a non-interference method. Interferometry is generally based on the principle of light diffraction and has higher precision. Among them, the moiré interferometry method has become a standard method for measuring the elastic constant of high-temperature materials, but the testing process and data processing process of this method are complicated, and its grating etching cannot be applied to In the measurement of ultra-thin materials. The digital image correlation (DIC) method in the non-interferometric measurement method has a wider range of applications, and shows a good application prospect in the measurement of the elastic constant of materials at high temperatures.

The principle of the two-dimensional digital image correlation method is: collect the images of the surface of the object before and after deformation, and calculate the correlation degree of the gray features in the images before and after the deformation of the object, and find the largest difference between the deformed image and a certain sample sub-area on the reference image. The target sub-area of the correlation degree, so as to determine the displacement vector of the sample sub-area.

When the digital image correlation method is applied to the high temperature deformation measurement, the calculated correlation coefficient in some areas will decrease, which is called "decorrelation effect". The main reasons for this situation are: (1) High-temperature speckles fall off, oxidize, and melt at high temperatures, resulting in a decrease in matching degree before and after deformation; (2) Thermal radiation on the sample surface leads to oversaturation of digital image brightness.

In addition, high-temperature deformation measurements using digital image correlation methods in atmospheric environments will be affected by heat flow disturbances. Due to the non-uniform distribution of air density caused by air thermal convection on the optical path, the refractive index of air on different optical paths is inconsistent, resulting in distortion of the collected images. Due to decorrelation effects and heat flow perturbations

In order to solve the decorrelation effect and the influence of heat flow disturbance, a large number of scholars at home and abroad have carried out research on digital image correlation methods at high temperatures. Experiments have proved that introducing a cool-color light source and a narrow band-pass filter into the measurement optical path can effectively eliminate the influence of thermal radiation. Scholars at home and abroad have also proposed a variety of high-temperature speckle production methods and processes. At present, the most common speckle production method is to use alumina ceramic powder and organic solvent two-component speckle material to mix, and use intermolecular force to bond on the surface of the sample through the heating and curing process. This method has the advantages of wide applicability, high operating temperature, small speckle layer thickness, simple curing process and generally no bubbles, but has some disadvantages, such as the large particles of alumina ceramic powder, which are difficult to be evenly sprayed by general spray guns. The ejection; the intermolecular force is weak, and it is easy to fall off in the high temperature test. Corresponding to the digital image, the quality of speckle in some areas is reduced due to problems such as speckle production and speckle drop-off, which in turn reduces the correlation coefficient and affects the test accuracy. The airflow field of hot air has spatial randomness and temporal randomness. Spatial randomness causes images to fluctuate randomly on the horizontal and vertical coordinates around a certain origin under the influence of thermal convection. Temporal randomness is manifested as a sequence of images in the same state. is a random jitter over time. At present, general solutions to heat flow disturbance include gray-scale averaging, median filtering, mean filtering, and so on.

Contents of the invention

In view of the above-mentioned situation of the prior art, the purpose of the present invention is to provide a kind of high-temperature elastic constant measurement method for thin strips, to solve the problems caused by de-correlation effect and heat flow disturbance when using digital image correlation method to measure deformation under high-temperature atmospheric environment. In order to reduce the problem of accuracy, this method can effectively improve the measurement accuracy of deformation in high temperature environment.

Above-mentioned purpose of the present invention utilizes following technical scheme to realize:

A method for measuring high-temperature elastic constants for thin strips, comprising the following steps:

(1) Prepare a flat-plate high-temperature tensile sample, and make high-temperature speckles on the surface of the sample;

(2) Collect the reference image (hereinafter referred to as the reference image) of the flat high-temperature tensile specimen in the unloaded state, and the images in each loaded state (hereinafter referred to as the loaded image);

(3) identifying high-reliability square sub-regions in the reference image;

(4) Utilize the digital image correlation method to compare the reference image and the loaded image, and calculate the horizontal and vertical displacements of the high-reliability sub-region after loading;

(5) Based on the calculated transverse and longitudinal displacements and corresponding loads of the high-reliability sub-region after loading, the elastic modulus E and Poisson's ratio μ of the sample material are obtained through weighted least squares fitting.

Further, in the step (1), aluminum oxide powder and alcohol mixed according to a mass ratio of 3:1 are used to make high-temperature speckles. Among them, the cutting and milling method is used to process the material into a flat tensile sample. In addition, if the sample is too thin, a reinforcing sheet can be set in the clamping part.

Further, wherein in step (2), the collection of the reference image includes continuously collecting no less than 20 frames of images at the target temperature, and performing grayscale averaging on the collected images to obtain a reference image; Acquisition includes loading with the beam displacement control method, the tensile speed is not greater than 0.6mm/min, and the loading images under the corresponding load are obtained continuously.

Further, in the step (4), the evaluation parameters of the reliability include the average gray gradient sum of squares, temporal correlation and spatial correlation.

Further, wherein identifying the high-reliability square sub-region in the reference image includes:

Translate the sample, use the neural network algorithm to train, and establish the reliability parameter Q of the image. When p increases, the reliability parameter Q of all p×p sub-regions in the reference image is calculated cyclically, and the reliable parameters are selected. The sub-area of the maximum value,

Judging whether the selected reliability maximum value sub-area contains the previously determined small-size high-reliability sub-area, and judging whether the reliability parameter Q of the reliability maximum value sub-area is larger than the contained small-size high-reliability sub-area , if it is not included, then the maximum reliability sub-area will be used as a new high-reliability sub-area; if it is included, and the reliability parameter Q of the reliability maximum value sub-area is larger than the included high-reliability sub-area, Then replace the included high reliability sub-area with the reliability maximum value sub-area as a new high-reliability sub-area; If the area is small, the maximum reliability sub-area will not be used as a new high-reliability sub-area.

Further, in step (5), the weighting factor is the product of the reliability parameter of the sub-area and the area of the sub-area.

The present invention considers that for a uniform strain field such as tension, there is a large amount of full-field information redundancy, and only needs to calculate the displacement of the region with the highest reliability of the speckle field to obtain the elastic constant of the material, and utilizes multiple high-reliability sub-regions The fitting between displacement and load eliminates the dispersion of heat flow disturbance in time and space. Therefore, the present invention establishes a composite speckle evaluation parameter that considers the speckle average gray gradient, spatial correlation and temporal correlation, and obtains it using a neural network algorithm in a translation test. The method of the invention can not only overcome the problem of the decrease of the correlation coefficient of some image regions caused by the decorrelation effect, but also effectively eliminate the random error caused by the high-temperature heat flow disturbance, and complete the high-efficiency measurement of the high-temperature elastic constant by non-contact means.

Description of drawings

Figure 1 is a flow chart illustrating the steps of the method of the present invention;

Fig. 2 is the schematic diagram of the measurement system that builds when realizing the method of the present invention;

Fig. 3 is the schematic diagram of the plate sample that the inventive method adopts;

Fig. 4 is a schematic diagram of a high-reliability sub-region in an image at 900°C in the method of the present invention.

Detailed ways

In order to understand the purpose, technical solutions and advantages of the present invention more clearly, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.

The present invention provides a kind of method for measuring the high temperature elastic constant of thin strip, comprising steps as follows (see Fig. 1):

S1: Sample processing and high temperature speckle production:

S11: use the cutting and milling method to process it into a flat tensile sample. If the sample is too thin, braze the reinforcing sheet in the clamping part to ensure that the pin hole will not be pulled out during the stretching process. If the initial surface state is too smooth, use frosting to increase the roughness of the surface.

S12: Aluminum oxide powder and alcohol solvent are used to make high-temperature speckles. Mix alumina powder and alcohol solvent at a mass ratio of 3:1, and spray evenly on the surface of the specimen with a pressure spray gun to form a layer of speckle with uniform distribution and high image contrast, ensuring that the speckle particle size is about 3~ 5 pixels, the speckle density is about 50%; dry the speckle-sprayed specimen at room temperature for 2 hours, then put it in a high-temperature environment box at 93°C for 2 hours, take it out and cool it naturally; Blow off loose speckles to complete the preparation of high-temperature speckles.

S2: Build a high-temperature non-contact deformation measurement system:

The system consists of a high-resolution digital CCD camera, a bi-telecentric lens, a blue LED area array lighting source, a narrow band-pass filter installed in front of the lens, a signal controller, image analysis equipment, a fan, and a high temperature sensor with an optical observation window. Composition of muffle furnace, testing machine and sample. Among them, the bi-telecentric lens can greatly reduce the influence of out-of-plane displacement and lens distortion on the experimental results, and the blue LED light source and narrow band-pass filter can greatly eliminate the error caused by the thermal radiation of the sample. The signal controller is connected with the tensile testing machine and the computer, and is used for synchronizing the load signal of the tensile testing machine with the image signal collected by the computer. The heating control system is used to control the temperature rise and heat preservation of the high temperature muffle furnace. The test device is shown in Figure 2.

S3: Loading and Image Acquisition

S31: After the sample is heated to the set temperature in the high-temperature furnace and kept warm until the heat is stable, no less than 20 images are continuously collected under the initial load state, and the collection interval is 5s. These images are gray-scale averaged to obtain a reference image of the flat-type high-temperature tensile specimen.

S32: Use the beam displacement control method to load, the tensile speed does not exceed 0.6mm/min, and use the signal controller to coordinate the load Fn and the digital image of the sample surface collected under this load to obtain the loading of the flat high-temperature tensile sample image.

S4: Identification of high reliability sub-areas

The reference image after gray level averaging is selected as the processing object to identify high-reliability sub-regions. The identification of the high-reliability sub-area is divided into the following steps:

S41: Using mean intensity gradient (MIG), temporal correlation (ST), and spatial correlation (SS) as intermediate parameters for reliability evaluation. Translate the sample, use the neural network algorithm to train, and establish the reliability parameters of the image, specifically including moving the sample up and down once, and collecting 50 images in each state. Using the neural network algorithm, the mean intensity gradient (MIG), time correlation, and spatial correlation are used as the convolution intermediate parameters, and the translation error is used as the objective function to train the reliability evaluation parameters and solve the linear Composite reliability evaluation parameter Q.

时间相关度表示在一段时间内子区位置的方差，在T个时间点里，时间相关度可以表示为

空间相关度表示子区平移后位移误差的方差，在G次位移后，/>

其中f_x(x_ij)和f_y(x_ij)是点x_ij处沿x向和y向的偏导数，采用中心差分算法进行计算。where the average gray gradient of the sub-region of p×p

The temporal correlation represents the variance of the position of the sub-area within a period of time. In T time points, the temporal correlation can be expressed as

Spatial correlation indicates the variance of the displacement error after sub-area translation, after G times of displacement, />

Where f _x (x _ij ) and f _y (x _ij ) are the partial derivatives along the x and y directions at the point x _ij , which are calculated by the central difference algorithm.

S42: When p is incremented, cyclically calculate the Q of all p×p pixel sub-areas in the image, and screen out the sub-areas with maximum reliability. The initial value of p is set to 5. Judging whether the selected reliability maximum value sub-area contains the previously determined small-size high-reliability sub-area, and judging whether the reliability parameter Q of the reliability maximum value sub-area is larger than the contained small-size high-reliability sub-area , if it is not included, then the maximum reliability sub-area will be used as a new high-reliability sub-area; if it is included, and the reliability parameter Q of the reliability maximum value sub-area is larger than the included high-reliability sub-area, Then replace the included high-reliability sub-area with the reliability maximum value sub-area; if it is included, but the reliability parameter Q of the reliability maximum value sub-area is smaller than the included high-reliability sub-area, then the reliable The sex maximum sub-area is used as a new high-reliability sub-area. After the calculation, assign p to p+1, and repeat the calculation until p=min(H,W). Where H is the pixel length of the sample, and W is the pixel width of the sample. Furthermore, n high-reliability regions are obtained on the reference image.

S5: Displacement calculation of high reliability sub-area

Using the digital image correlation algorithm, comparing the reference image and the loaded image, the corresponding position of the high-reliability sub-region in the deformed image is obtained, and then the displacement (u _n,i ,v _n,i ) of the center point of the sub-region is obtained.

S6: Calculation of elastic constant

The displacement (u _n,i ,v _n,i ) of the center point (x _n,i ,y _n,i ) of the high-reliability sub-area and the load level F _n are subjected to the least square calculation, and the area of the sub-area and the reliability parameter The product of Q is used as a weighting factor to obtain the elastic modulus E and Poisson's ratio μ, where c ₁ and c ₂ are constant items, and S is the cross-sectional area of the working section of the sample.

Specific example:

GH605 nickel-cobalt-based deformed superalloy cold-rolled strip is used to process flat tensile specimens (see Figure 3), the initial thickness of which is 0.2mm, and 0.4mm-thick reinforcing sheets are welded on both sides of the clamping section to ensure tensile strength You will not be exempted during the process.

Aluminum oxide powder and alcohol solvent are used to make high-temperature speckles. Mix alumina powder and alcohol solvent at a mass ratio of 3:1, and spray evenly on the surface of the test piece by the brush dripping method to form a layer of speckle with uniform distribution and high image contrast. The collected images are captured by a CCD camera, the speckle density is about 50%, and the size of a single speckle is about 3-5 pixels. Dry the speckle-sprayed specimen naturally at room temperature for 2 hours, then heat it in a high-temperature environment box at 93°C for 2 hours, take it out, and cool it naturally; Spot preparation.

Build a high-temperature non-contact deformation measurement system: the system consists of a high-resolution Baumer digital CCD camera, a Schneider bi-telecentric lens (working distance 368mm), a blue LED area array lighting source (2000 lumens), and a narrow bandpass filter installed in front of the lens (Filter range 450~455nm), signal controller, image analysis equipment, pneumatic device, high temperature muffle furnace with optical observation window, heating control system, tensile testing machine and sample. The signal controller is connected with the tensile testing machine and the computer, and is used for synchronizing the load signal of the tensile testing machine with the image signal collected by the computer. The heating control system is used to control the temperature rise and heat preservation of the high temperature muffle furnace.

After the pre-tightening force is applied to the sample, it is heated to the target temperature in a high-temperature furnace and held for 20 minutes. In this state, 20 images are collected continuously with an interval of 5 seconds. These images are gray averaged to obtain the reference image Q0.

The beam displacement control method is used for loading, the tensile speed is 0.6mm/min, and the signal controller is used to coordinate the load Fn and the digital image of the sample surface collected under this load.

Taking the reference image Q0 as the object, the sub-regions with high reliability in the image are identified.

Among them, the identification of the high-reliability sub-area is divided into the following three steps:

In the first step, the mean intensity gradient (MIG), temporal correlation (ST), and spatial correlation (SS) are used as reliability evaluation parameters. Translate the sample up and down once, and collect 50 images in each state. Using the neural network algorithm, the mean intensity gradient (MIG), time correlation, and spatial correlation are used as the convolution intermediate parameters, and the translation error is used as the objective function to train the reliability evaluation parameters and solve the linear Composite reliability evaluation parameter expression Q.

The second step is to rely on the composite reliability parameter Q as the reliability evaluation standard. Calculate the Q of all 5pixel×5pixel subregions in the image, and screen out the subregions with the maximum reliability.

其中f_x(x_ij)和f_y(x_ij)是点x_ij处沿x向和y向的偏导数，采用中心差分算法进行计算。时间相关度表示在一段时间内子区位置的方差，在T个时间点里，时间相关度可以表示为

空间相关度表示子区平移后位移误差的方差，在G次位移后，/>

average gray gradient

Where f _x (x _ij ) and f _y (x _ij ) are the partial derivatives along the x and y directions at the point x _ij , which are calculated by the central difference algorithm. The temporal correlation represents the variance of the position of the sub-area within a period of time. In T time points, the temporal correlation can be expressed as

Spatial correlation indicates the variance of the displacement error after sub-area translation, after G times of displacement, />

The third step is to recursively calculate the Q of all p×p sub-regions in the image, and screen out the sub-regions with the maximum reliability. The initial value of p is set to 5. Judging whether the selected reliability maximum value sub-area contains the previously determined small-size high-reliability sub-area, and judging whether the reliability parameter Q of the reliability maximum value sub-area is larger than the contained small-size high-reliability sub-area , if it is not included, then the maximum reliability sub-area will be used as a new high-reliability sub-area; if it is included, and the reliability parameter Q of the reliability maximum value sub-area is larger than the included high-reliability sub-area, then Replace the included high-reliability sub-area with the reliability maximum value sub-area; if it is included, but the reliability parameter Q of the reliability maximum value sub-area is smaller than the included high-reliability sub-area, then the The maximum reliability sub-area is used as a new high-reliability sub-area. After the calculation, assign p to p+1, and repeat the calculation until p=min(H,W). Where H is the pixel length of the sample, and W is the pixel width of the sample. Furthermore, n high-reliability sub-regions are obtained on the initial image, as shown in FIG. 4 , where the boxed part in the figure is the selected high-reliability sub-region.

The digital image correlation algorithm is used to correlate the loaded images one by one, and the sub-pixel search algorithm uses the inverse synthesis-Gauss-Newton (IC-GN) algorithm to obtain the corresponding position of the high-reliability sub-region in the deformed image, and then obtain the sub-pixel Displacement (u _n,i ,v _n,i ) of the zone center point.

The displacement (u _n,i ,v _n,i ) of the center point (x _n,i ,y _n,i ) of the high-reliability sub-area and the load level F _n are subjected to the least square operation, and the product of the area of the sub-area and Q As a weighting factor, the elastic modulus E and Poisson's ratio μ are obtained, where c ₁ and c ₂ are constant items, and S is the cross-sectional area of the working section of the sample.

This method eliminates the errors caused by the decorrelation effect and heat flow disturbance when the digital image correlation method is used to measure the deformation in a high-temperature atmospheric environment, and effectively improves the elastic modulus E and Poisson's ratio of the GH605 nickel-cobalt-based deformed superalloy under high-temperature conditions The measurement accuracy of μ.

Claims (6)

1. A method for measuring the high temperature spring constant of a thin strip comprising the steps of:

(1) Preparing a flat high-temperature tensile sample, and preparing high-temperature speckles on the surface of the sample;

(2) Respectively acquiring a reference image and a loading image of a flat high-temperature tensile sample, wherein the reference image is obtained by continuously acquiring at least 20 frames of images at a target temperature and carrying out gray level average on the acquired images;

(3) Identifying a square subarea with high reliability in a reference image, wherein the square subarea with high reliability comprises the steps of performing translational motion on a sample, training by using a neural network algorithm, establishing reliability parameters of the image, circularly calculating the reliability parameters of all p multiplied by p pixel subareas in the reference image under the condition of increasing p, screening out the subarea with the maximum reliability as the square subarea with high reliability, judging whether the screened reliability maximum subarea contains a small-size high-reliability subarea determined before, judging whether the reliability parameters of the reliability maximum subarea are larger than the contained small-size high-reliability subarea, and taking the reliability maximum subarea as a new high-reliability subarea if the reliability maximum subarea is not contained; if the reliability parameter of the reliability maximum sub-area is larger than that of the included high reliability sub-area, replacing the included high reliability sub-area with the reliability maximum sub-area as a new high reliability sub-area; if included, but the reliability parameter of the reliability maximum subregion is smaller than the included high reliability subregion, the reliability maximum subregion is not taken as a new high reliability subregion;

(4) Comparing the reference image with the loaded image by using a digital image correlation method, and calculating the horizontal displacement and the vertical displacement of the high-reliability sub-region after loading;

(5) Based on the calculated horizontal and vertical displacement and corresponding load of the high-reliability sub-region after loading, taking the product of the area of the sub-region and the reliability parameter as a weighting factor, and obtaining the elastic modulus and poisson ratio of the sample material by least square fitting for weighting the horizontal and vertical displacement and the load level of the high-reliability sub-region.

2. The method according to claim 1, wherein in step (1), a mass ratio of 3:1 mixing alumina powder and alcohol to make high temperature speckle.

3. The method of claim 1, wherein in step (2), the acquisition of the loading image comprises loading by using a beam displacement control method, wherein the stretching speed is not more than 0.6mm/min, and the loading image under the corresponding load is continuously obtained.

4. The method of claim 1, wherein in step (4), the reliability evaluation parameters include average gray gradient sum-squares, temporal correlation, and spatial correlation.

5. The method of claim 1, wherein in step (1), the material is processed into a flat tensile specimen using a cutting and milling process.

6. The method of claim 5, wherein step (1) further comprises providing a reinforcing sheet at the clamping portion if the specimen is too thin.

Images