JP6240206B2 - Displacement field and strain field measurement method and material testing machine - Google Patents

Description

  The present invention relates to a displacement field and strain field measurement method in a static material test, and a material testing machine.

  In a static test in which a tensile load or a compressive load is applied to a specimen along a unidirectional axis, the load deformation of the specimen is measured by a contact extensometer such as a strain gauge or a non-contact type extensometer using a video camera. Measured as displacement and strain in the load axis direction.

  In addition, by analyzing the image of the test piece acquired by the video camera, not only the displacement and strain in the load axis direction, but also the displacement and strain fields, which are continuous deformations that occur in the test piece due to the load, are measured. In addition, a material testing machine for evaluating the mechanical properties of a test piece has been proposed (see Patent Document 1 and Patent Document 2).

  As described in Patent Document 1 and Patent Document 2, a displacement field and a strain field generated in a test piece to which a load is applied are calculated using image data (for example, gradation values of each pixel). . In the measurement using the conventional digital image correlation method, since the displacement is identified in units of pixels, the measurement accuracy of the displacement field and the strain field as a measurement result is limited by the image resolution. The digital image correlation method is a method for determining the displacement of a test piece by attaching a plurality of reference points to the surface of the test piece and comparing the reference point patterns in the test piece before and after the load is applied. It is.

  For this reason, in recent years, it has also been proposed to measure the displacement field and the strain field with a higher accuracy than the image resolution by adopting a technique called a dot centroid tracking method. FIG. 10 is a flowchart showing a conventional procedure for calculating a displacement field and a strain field by the dot centroid tracking method.

  First, before the test is run, a test piece with random marks (dots) on the surface is attached to the main body of the material testing machine, and the video camera's layout, field of view, sampling frequency, etc. are collected. , Setting of light and test environment, and setting of load applied to the test piece (steps S101 and S102).

  In the test piece image obtained by photographing with the video camera, each dot attached to the surface of the test piece is recognized, and the center of gravity position of each dot is calculated (step S103). Then, during the test in which a load is applied to the test piece, dot centroid measurement is performed to track the movement of the centroid position of each dot (step S104). At this time, the displacement amount of the dot, which is the amount of movement of the center of gravity position of each dot under test, is calculated (step S105). Based on the displacement amount and the shape function of each dot, the dot at any point on the test piece is calculated. A displacement amount and a strain amount are calculated, and a displacement field and a strain field are derived (steps S106 and S107).

  A displacement field or strain field, which is a two-dimensional continuous change occurring in a loaded specimen, is calculated based on the above-described movement amount of the dot centroid, and the dot centroid is calculated for each pixel constituting the dot in the image. It is calculated from a gradation value (for example, a gray scale of 256 gradations). In this way, the dot centroid tracking method can measure the movement of dots at a resolution exceeding the image resolution, so the measurement accuracy of the displacement field and strain field is improved compared to the digital image correlation method, Achieves measurement accuracy close to that when using strain gauges.

JP 2004-184152 A Registered Utility Model No. 3119564

  On the other hand, image data acquired by a video camera includes various noises such as environmental noise such as chromatic light and vibration at the location where the material testing machine is installed, mechanical noise such as camera lens distortion, and electrical noise in the image sensor. Noise is included. The magnitudes of these noises vary depending on the environment in which the material testing machine is installed, the selection of cameras arranged in the material testing machine, and parameter settings related to them.

  Data such as pixel gradation values correspond to the observed values in such measurement tests, and the noise contained in these data is the displacement and strain fields that are physical quantities calculated from the observed values. Measurement accuracy will be affected. Furthermore, in the evaluation of displacement and strain fields using the dot centroid tracking method, the measurement results are not only affected by the approximation error of the numerical analysis method and the rounding error of the computer, but also the dot size and dot density on the image. The effect of is also a problem.

  In the measurement method using a video camera, it was difficult to consider how reliable the measured value uniquely obtained from the information obtained through image processing was. Was not done.

  The present invention has been made to solve the above-described problems, and is a displacement field capable of presenting measurement uncertainty of a displacement field and a strain field caused by noise included in image data acquired by an imaging unit. An object of the present invention is to provide a strain field measurement method and a material testing machine.

In the first aspect of the present invention, the load acquired by the imaging unit is applied using a material testing machine including a load mechanism for applying a load to the test piece and an imaging unit for imaging the test piece. by analyzing the image of the test piece, a displacement field and strain field measurement method for measuring the displacement field and strain field generated in the specimen, prior to execution test, obtained by the shooting means and the A fixed noise model creation step for stochastically modeling the noise contained in each pixel of the image data of the test piece, and a plurality of marks randomly assigned to the surface of the test piece are recognized, and calculates the center of gravity as an average value, by estimating the standard deviation of the center of gravity of the target point based on the noise of the pixels modeled in the fixed noise model creation process, given the load on the specimen The center-of-gravity measurement step of probabilistically measuring the center of gravity of each target point and the average value and standard deviation of the center of gravity calculated in the center-of-gravity measurement step are used to calculate the displacement amount of the center of gravity of each target point In addition, by estimating the uncertainty of the displacement amount of the center of gravity, a center-of-gravity displacement amount calculation step for calculating the displacement amount of the center of gravity of each target point probabilistically, and the center of gravity calculated in the gravity center displacement amount calculation step Using the amount of displacement, calculate the amount of displacement and strain at an arbitrary point on the specimen, and estimate the uncertainty of the displacement field and strain field using the uncertainty of the displacement of the center of gravity. By using the displacement field and strain field calculation step for stochastically calculating the displacement field and strain field and the calculation results in the displacement field and strain field calculation step, the error range of the displacement field and the error range of the strain field are calculated. It characterized by having a a error range estimating step of estimating.

The invention according to claim 2, in the invention described in claim 1, fixed noise model creation step, darkness of each pixel repeatedly acquired in grayscale by the shooting means against the specimen with a different shade The average value of the darkness and the standard deviation in each pixel are calculated by statistically processing the measured values, and a correlation between the darkness and the standard deviation is derived, and a probability distribution of arbitrary darkness is estimated.

  According to a third aspect of the present invention, in the second aspect of the present invention, the center-of-gravity measurement step recognizes the plurality of reference points, and the darkness of each pixel constituting each reference point, and the fixed noise model. Using the probability distribution at an arbitrary darkness estimated in the creation step, the uncertainty of the center of gravity of each gage calculated as an average value is estimated.

  According to a fourth aspect of the present invention, in the first aspect of the present invention, the error range estimating step assumes that the displacement amount of the center of gravity of the gauge point follows a Gaussian process in the displacement field and strain field calculation step. An integration range in which a predetermined probability is obtained when integrating the probability density function of the displacement amount and strain amount at the arbitrary point obtained in this way is defined as a measurement error range.

The invention according to claim 5 comprises a load mechanism for applying a load to the test piece, and a photographing means for photographing the test piece, and analyzes the image of the test piece to which the load acquired by the photographing means is applied. by, a material testing machine for measuring the displacement field and strain field generated in the test piece, the noise contained in each pixel of the image data of the test piece obtained by the shooting means before executing test In the fixed noise model creation unit, the fixed noise model creation unit that recognizes a plurality of random marks randomly attached to the surface of the test piece and calculates the center of gravity of the standard points as an average value. Center-of-gravity measurement that probabilistically measures the center of gravity of each target point when a load is applied to the test piece by estimating the standard deviation of the center of gravity of the target point based on the noise of each modeled pixel And Using the average value and standard deviation of the center of gravity calculated in the center-of-gravity measurement unit, calculating the displacement amount of the center of gravity of each target point, and estimating the uncertainty of the displacement amount of the center of gravity of the target point, and the center of gravity displacement calculation unit that calculates a displacement amount of the center of gravity stochastically each gauge, the uncertainty in the amount of displacement of the center of gravity of the displacement amount and the target point of the center of gravity of the target point calculated in said center-of-gravity displacement calculation unit The amount of displacement and strain at any point on the surface of the test piece is calculated using this, and the displacement and strain fields are calculated stochastically by estimating the uncertainty of the displacement and strain fields. An image analysis comprising: a displacement field and strain field calculation unit; and an error range estimation unit that estimates an error range of the displacement field and an error range of the strain field using a calculation result in the displacement field and strain field calculation unit With means And wherein the door.

According to claim 1, the invention of claim 5, the noise included in the image data obtained by shooting means modeled, incorporating statistical techniques conventional dot centroid tracking method, with the test specimen By measuring the center of gravity of the measured point probabilistically and by estimating the uncertainty of the displacement field and strain field using the propagation rule of uncertainty, generally the approximate value or estimated value of the measurand value is used. Uncertainty can be shown together with the measurement results that are considered to be present. Therefore, it is possible to improve the reliability of the measured value of the displacement field and strain field derived by processing the image acquired by the shooting unit.

It is a schematic diagram which shows the main structures of the material testing machine of this invention. It is a block diagram which shows the main control systems of the material testing machine of this invention. It is a flowchart which shows the measurement procedure of the displacement field and strain field in the material testing machine of this invention. It is a graph explaining modeling of the fixed noise of each pixel. It is a graph explaining modeling of the fixed noise of each pixel. 2 is a schematic diagram illustrating an example of dots attached to a test piece 10. FIG. It is a conceptual diagram of the stochastic measurement of the dot center of gravity. It is a conceptual diagram of the stochastic calculation of the amount of dot displacement. It is an example of a display after the stochastic calculation of a displacement field and a strain field. It is a flowchart which shows the calculation procedure of the conventional displacement field and strain field by a dot gravity center tracking method.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram showing the configuration of a material testing machine according to the present invention.

  This material testing machine includes a table 16, a pair of screw rods 11 erected on the table 16 so as to be vertically oriented, a crosshead 13 movable along these screw rods 11, A load mechanism 30 for applying a test force to the test piece 10 by moving the crosshead 13 and a video camera 23 for measuring a displacement field and a distortion field of the test piece 10 are provided.

  The cross head 13 is connected to the pair of screw rods 11 via nuts (not shown). A lower end portion of each screw rod 11 is connected to the load mechanism 30, and power from a motor 31 (see FIG. 2) as a power source in the load mechanism 30 is transmitted to the pair of screw rods 11. ing. As the pair of screw rods 11 rotate in synchronization, the cross head 13 moves up and down along the pair of screw rods 11.

  The crosshead 13 is provided with an upper gripping tool 21 for gripping the upper end portion of the test piece 10. On the other hand, the table 16 is provided with a lower gripping tool 22 for gripping the lower end portion of the test piece 10. When performing a tensile test, the test force (tensile load) is applied to the test piece 10 by raising the cross head 13 in a state where both ends of the test piece 10 are held by the upper gripping tool 21 and the lower gripping tool 22. ). At this time, the test force acting on the test piece 10 is detected by the load cell 14 and input to the control unit 35. Further, the physical quantity such as the displacement amount of the test piece 10 is calculated by the control unit 35 based on the image acquired by the video camera 23.

  The control unit 35 is configured by a computer including a storage device such as a RAM and a ROM that can store programs and various data, and an arithmetic device such as a CPU, and serves as a control device that controls the overall operation of the material testing machine. In addition to functioning, it functions as image analysis means of the present invention. A display unit 36 and an input unit 37 are connected to the control unit 35. The input unit 37 accepts user operations such as changing test conditions. The display unit 36 displays test conditions, displacement of the test piece 10 under test, and the like.

  FIG. 2 is a block diagram showing a main control system of the material testing machine of the present invention.

  The control unit 35 of the material testing machine includes a load control unit 41, an image processing unit 50, a display control unit 42, and a storage unit 43 as functional configurations. The load control unit 41 operates the motor 31 in the load mechanism 30 based on the test conditions set by the user via the input unit 37, and performs the test under the test based on the load detected by the load cell 14 under test. The operation of the motor 31 is adjusted.

  The image processing unit 50 processes the image data of the test piece 10 by processing a fixed noise model creating unit 51 that models fixed noise at each pixel in the image data acquired by the video camera 23 before the test, and the test piece. The center of gravity of a target point (dot) randomly assigned to the surface of 10 is measured, and the dot center of gravity measuring unit 52 that calculates the uncertainty of measurement, and the amount of displacement of the center of gravity of each dot is calculated. The dot displacement amount calculation unit 53, the displacement field and the strain field, and the like, and the displacement field and the strain field calculation unit 54 for estimating the uncertainty of the calculated physical quantity and the physical field, and the estimated uncertainty Then, an error range estimation unit 55 that estimates the error range of the displacement field and the strain field is provided. The error range of the displacement field and strain field is, for example, the level of confidence in the probability distribution of the displacement amount and strain amount at an arbitrary point on the surface of the test piece 10 calculated by the displacement field and strain field calculation unit 54. For example, the confidence interval is 90 to 95%.

  The display control unit 42 displays the image of the test piece 10 taken by the video camera 23 on the display unit 36 and also displays the image data and various calculation results in the image processing unit 50.

  The storage unit 43 stores image data acquired by the video camera 23 and stores various processing results in the image analysis unit 50.

  Apply the displacement field and strain field measurement method described below to the material testing machine with the above-described configuration, execute the material test from the material test preparation, and calculate the displacement field and strain field based on the image data In addition, the operation until the error range of the calculated value is estimated will be described. FIG. 3 is a flowchart showing a procedure for measuring a displacement field and a strain field in the material testing machine of the present invention. 4 and 5 are graphs for explaining modeling of fixed noise of each pixel in an image. FIG. 6 is a schematic diagram illustrating an example of dots attached to the test piece 10. FIG. 7 is a conceptual diagram of the probabilistic measurement of the dot center of gravity, and FIG. 8 is a conceptual diagram of the stochastic calculation of the dot displacement amount.

  In the measurement of the displacement field and strain field by this material testing machine, fixed noise in an image acquired in gray scale by the video camera 23 is stochastically modeled based on the darkness, and this stochastic model of fixed noise is used. Then, the uncertainty of the measured value of the center of gravity of the dot applied to the surface of the test piece when a load is applied is evaluated. Then, from the measurement result of the center of gravity of the dot, the dot displacement amount, the displacement field, and the strain field are stochastically calculated by the uncertainty propagation law. In other words, the reliability of the measured value is presented to the user by probabilistically performing the measurement by the dot centroid tracking method and propagating the uncertainty to each physical quantity derived by calculation from the observed value.

  First, as preparation for the material test, the upper gripping tool 21 and the lower gripping tool 22 are held at both ends of the test piece 10, the setting of the arrangement of the video camera 23, the field of view, the sampling frequency of data, the setting of the light and the test environment, etc. (Step S1). Before setting the load to be applied to the test piece 10 (step S2) and performing the test, the stochastic modeling of the fixed noise of each pixel of the image acquired in gray scale from the video camera 23 is performed (step S3).

  Stochastic modeling of the fixed noise of each pixel is performed in the following procedure. First, about 10 sheets of paper with different shades of white to black are applied to the surface of the test piece 10 in sequence, and the darkness measurement is repeated at least 2000 times for each different density of paper. , Get the tone value. Normally, the gray scale value of 256 gray scales is 255 for white and 0 for black. However, in the stochastic modeling of fixed noise, this gray scale value is inverted so that white is 0, It is stored in the storage unit 43 as a measured value representing the darkness with black being 255. Note that the measurement value (darkness value) representing the darkness is based on a luminance value that is one of the elements expressing the color. That is, by converting the color image of the test piece 10 acquired from the video camera 23 into an image having only a luminance value, a grayscale image is acquired and a darkness value is obtained.

  And the average value and standard deviation of the darkness of each pixel are calculated, and the graph of FIG. 4 is completed. As shown in FIG. 4, white and bright colors with smaller darkness values are more likely to contain noise than black and darker colors and tend to increase measurement uncertainty. The graph of FIG. 4 is represented by the function shown in the following equation (1), in which the average value of the darkness of the pixels is input and the standard deviation is output.

  In this way, in the present invention, a stochastic model of fixed noise of each pixel is constructed as shown in the graph of FIG. 4 by statistically processing the measured darkness of each pixel. As a result, as shown in FIG. 5, it is possible to estimate a probability distribution in an arbitrary darkness, in other words, a probability density function of a normal distribution.

  When the graph shown in FIG. 4 is completed and Expression (1) is obtained, next, a material test for applying a load to the test piece 10 is executed. In this embodiment, as shown in FIG. 6, dots having different sizes radially from the circular holes are attached to the surface of the test piece 10 in which a circular hole is formed in the central portion. In addition, how to attach a dot is not limited to what is shown in this embodiment, You may attach the dot of the same size at random.

  During the execution of the material test, the video camera 23 sequentially acquires images, calculates the center of gravity of each dot attached to the surface of the test piece 10, and tracks the movement of the center of gravity. That is, recognition of each dot shape on the image (step S4), probabilistic measurement of the dot center of gravity in the recognized dot shape (step S5), and probability calculation of the dot displacement amount (step S6) 10 is repeated according to the change in the load applied to 10. As shown in FIG. 6, each dot attached to the surface of the test piece 10 is composed of a plurality of pixels, and the measured darkness in each pixel constituting each dot is different. For this reason, from the equation (1), the estimated standard deviation with respect to the measured darkness of the j-th pixel of the i-th dot at step k (an image at an arbitrary time) is expressed by the following equation (2). .

  Note that the barycentric position of the i-th dot is calculated as an average value by the following equation (3), assuming that the measurement is repeated infinitely.

  Using the equation (3) of the average value of the dot centroid given as a weighted sum of the normal distribution, the measurement uncertainty of the dot centroid is analytically derived by the following equation (4).

The measurement uncertainty of the dot centroid shown in equation (4) can be estimated by propagating the estimated standard deviation of equation (2) according to the uncertainty propagation law using equation (3). That is, as shown in the upper part of FIG. 7, the standard deviation of the position x _i to be estimated with the center of each pixel constituting a certain single dot shapes, using the relationship between the shade and the standard deviation shown in FIG. 4 To estimate. That is, the probability distribution of darkness observed at each pixel constituting each dot is estimated (see FIG. 5). Then, as shown in the lower part of FIG. 7, the uncertainty of the dot centroid can be estimated by combining the estimated standard deviation of each pixel in two dimensions in the x direction and the y direction. In general, since it has been experimentally known that a dot as shown in this embodiment has the same uncertainty in the x direction and the y direction, as shown in the following equation (5), x The standard deviation in the direction and the standard deviation in the y direction can be treated as the same.

  Next, from the dot centroid at step k-1 and the dot centroid at step k of the i-th dot, the displacement amount of the i-th dot centroid at the time between step k-1 and step k is This is expressed as Equation (6) (see the upper part of FIG. 8).

  The uncertainty of the dot centroid of the i-th dot at step k-1 and the uncertainty of the dot centroid at step k are given by equation (4). The uncertainties of the respective terms in the equation (6) are the uncertainties of the dot centroids in the step k-1 and the uncertainties of the dot centroids in the step k by analytical or simulation using the equation (6). It estimates by propagating (see the lower part of FIG. 8). The uncertainty of the displacement amount of the dot centroid is expressed as equation (7).

  In general, since it has been experimentally known that a dot as shown in this embodiment has the same uncertainty in the x direction and the y direction, as shown in equation (8), The standard deviation and the standard deviation in the y direction can be treated as the same thing.

  The displacement amount of the dot centroid in all the dots attached to the test piece 10 is calculated by the above-described equation (6) (step S6). That is, calculating the displacement field and the strain field stochastically means that the displacement amount of the dot center of gravity, which is the average value represented by the equation (6), is known, and the displacement amount and strain at an arbitrary point that is an unknown value. This corresponds to obtaining the average value and variance of the quantities. The displacement amount at an arbitrary point is expressed by Equation (9) from the displacement amount of the dot centroid and the approximate matrix of the shape function in the neighboring dots, regardless of whether the finite element method or the mesh free method is used.

  The strain amount at an arbitrary point is calculated by the following equation (10) obtained by adding a Laplace differential operator matrix to the equation (9).

The amount of displacement at an arbitrary point calculated as an average value by Equation (9) and the uncertainty of the amount of strain at an arbitrary point calculated as an average value by Equation (10) are expressed by Equation (7). It can be expressed by the uncertainty of the amount of dot displacement. Therefore, the uncertainty ⁱ σ ^k _u of the displacement field can be estimated by propagating the uncertainty of the dot displacement amount by the equation (9) (step S7), and the uncertainty ⁱ σ ^k _ε of the strain field is the dot displacement amount. Can be estimated by propagating the uncertainties in accordance with equation (10) (step S8). That is, the probabilistic calculation of displacement and strain estimates the uncertainty of the displacement and strain at any point, assuming that the dot displacement derived from the dot centroid measurement follows a Gaussian process. That is.

  FIG. 9 is a display example after the stochastic calculation of the displacement field and the strain field.

  The displacement field and strain field obtained by the probabilistic calculation method described with reference to the equations (1) to (10) help the user understand the two-dimensional deformation state occurring on the surface of the test piece 10. For this reason, the image is processed and displayed on the display unit 36 by the action of the image processing unit 50 and the display control unit 42.

  When a tensile load is applied to the test piece 10 in which a circular hole is formed in the center of the plate shown in FIG. 6, deformation due to shearing force (deformation having an angle with respect to the load axis) occurs in the vicinity of the circular hole. For this reason, in FIG. 9, the displacement and strain fields on the surface of the test piece 10 observed around the circular hole are represented as a shear strain distribution. The value of the shear strain, which is an average value inserted on the surface image of the test piece 10, is visualized by showing different hatching values for each predetermined numerical range in FIG. In addition, the uncertainity of the shear strain distribution (Uncertainty) is displayed on the screen displaying the shear strain distribution on the surface of the test piece 10 with an isoline diagram showing different line types for each predetermined numerical value in FIG. doing. The uncertainty shown in the isoline diagram in FIG. 9 is the standard deviation of the shear strain derived as an average value.

  As shown in FIG. 9, it can be seen that large shear strain is generated in the upper right and lower left directions from the circular hole, and the uncertainty value is small in the area of the isoline indicated by the solid line from the circular hole to the upper right. It is understood that the accuracy is high. In FIG. 9, for convenience, the numerical range of the shear strain is indicated by different hatching, and the uncertainty is expressed using different line types, but it is easier to understand for the user by color coding using a large number of colors. Information can be presented in a manner.

  In this embodiment, the measurement error range is presented to the user as a numerical value as a representation of the uncertainty of each measurement value calculated as an average value. Since the properties of the measured values such as displacement and strain are expressed as a probability density function, the measurement error range of the displacement field and strain field has a probability close to 1 by integrating the probability density function (for example, , 0.95 and 0.90) can be estimated (range of confidence) (steps S9 and S10).

  Further, in the measurement of the displacement field and the strain field in this embodiment, the numerical value of the error range of the measurement value can be provided to the user of the material testing machine together with the image display on the display unit 36 and the numerical display of the physical quantity. The reliability of the measured value can be improved.

  As described above, in the measurement of the displacement field and the strain field to which the present invention is applied, the fixed noise in the image acquired in gray scale by the video camera 23 is probabilistically modeled based on the darkness, and the probability of this fixed noise. A statistical model is used to evaluate the uncertainty of the measured value of the center of gravity of the dots applied to the surface of the specimen when a load is applied. Then, by propagating the uncertainty of the measured value of the dot centroid from the measurement result of the dot centroid to the dot displacement, displacement field, and strain field determined by calculation, these physical quantities are calculated stochastically. doing. Therefore, in the measurement of the displacement field and strain field by the dot center of gravity tracking method by the conventional material testing machine, the measured value is output as the definite value, whereas in the material testing machine to which this invention is applied, Uncertainty can be shown together with the measurement result that is understood to be an approximate value or an estimated value of the value of the measured quantity.

DESCRIPTION OF SYMBOLS 10 Test piece 11 Screw head 13 Cross head 14 Load cell 16 Table 19 Support | pillar 21 Upper grip 22 Lower grip 23 Video camera 30 Load mechanism 31 Motor 35 Control part 36 Display part 37 Input part 41 Load control part 42 Display control part 43 Memory | storage Unit 50 image processing unit 51 fixed noise model creation unit 52 dot center of gravity calculation unit 53 dot displacement amount calculation unit 54 displacement field and strain field calculation unit 55 error range estimation unit

Claims (5)

Analyzing the image of the test piece to which the load acquired by the photographing means is applied, using a material testing machine including a load mechanism for applying a load to the test piece and a photographing means for photographing the test piece. A displacement field and strain field measuring method for measuring a displacement field and a strain field generated in the test piece,
Before test execution, the fixed noise model creation step for stochastically modeling a noise included in each pixel of the image data of the acquired test piece by the shooting means,
Recognizing a plurality of marks randomly attached to the surface of the test piece, calculating the center of gravity of each mark as an average value, and based on the noise of each pixel modeled in the fixed noise model creation step A center-of-gravity measurement step for probabilistically measuring the center of gravity of each test point when a load is applied to the test piece by estimating the standard deviation of the center of gravity of each test point;
By using the average value and standard deviation of the center of gravity calculated in the center-of-gravity measurement step, calculating the displacement amount of the center of gravity of each target point, and estimating the uncertainty of the displacement amount of the center of gravity, A center-of-gravity displacement amount calculating step for stochastically calculating a displacement amount of the center of gravity;
Using the displacement amount of the center of gravity calculated in the center-of-gravity displacement calculation step, the displacement amount and strain amount at an arbitrary point in the test piece are calculated, and the uncertainty of the displacement amount of the center of gravity is used. A displacement field and strain field calculating step for stochastically calculating the displacement field and strain field by estimating the uncertainty of the displacement field and strain field;
An error range estimation step of estimating an error range of the displacement field and an error range of the strain field using the calculation result in the displacement field and strain field calculation step;
Displacement field and strain field measurement method having The displacement field and strain field measurement method according to claim 1,
Fixed noise model creation process, by statistically processing the measurement values of the shade of each pixel repeatedly acquired in grayscale by the shooting means against the specimen with different shade, the average value of the darkness of each pixel A displacement field and strain field measurement method that calculates the standard deviation and the correlation between the darkness and the standard deviation, and estimates the probability distribution of any darkness. In the displacement field and strain field measuring method according to claim 2,
The center-of-gravity measurement step uses the probability distribution at the darkness estimated at the fixed noise model creation step and the darkness of each pixel constituting each of the target points by recognizing the plurality of points. A displacement field and strain field measurement method for estimating the uncertainty of the center of gravity of each gage calculated as an average value. The displacement field and strain field measurement method according to claim 1,
The error range estimating step is a probability density of the displacement amount and the strain amount at the arbitrary point obtained on the assumption that the displacement amount of the center of gravity of the gauge point follows a Gaussian process in the displacement field and strain field calculation step. A displacement field and strain field measurement method in which a measurement error range is an integration range in which a predetermined probability is obtained when a function is integrated. A load mechanism that applies a load to the test piece, and a photographing unit that photographs the test piece, and is generated in the test piece by analyzing an image of the test piece to which the load acquired by the photographing unit is applied A material testing machine for measuring displacement and strain fields,
A fixed noise modeling unit that models the noise contained in each pixel of the image data of the acquired test piece by the shooting means prior test execution,
Recognizing a plurality of marks randomly attached to the surface of the test piece, calculating the center of gravity of the marks as an average value, and based on the noise of each pixel modeled in the fixed noise model creation unit A center-of-gravity measuring unit that probabilistically measures the center of gravity of each of the test points when a load is applied to the test piece, by estimating the standard deviation of the center of gravity of the test points;
Using the average value and standard deviation of the center of gravity calculated in the center of gravity measuring unit, calculating the displacement amount of the center of gravity of each target point, and estimating the uncertainty of the displacement amount of the center of gravity of the target point, A center-of-gravity displacement calculation unit that calculates the displacement of the center of gravity of each gauge point, and
Using the uncertainties of the displacement of the center of gravity of the gauge and the displacement of the center of gravity of the gauge calculated in the gravity center displacement calculation unit, the displacement and strain at any point on the surface of the test piece And a displacement field and strain field calculation unit for probabilistically calculating the displacement field and the strain field by estimating the uncertainty of the displacement field and the strain field,
Using the calculation results in the displacement field and strain field calculation unit, an error range estimation unit for estimating the error range of the displacement field and the error range of the strain field,
A material testing machine provided with an image analysis means.

Images