A method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis

Abstract

The present invention discloses a method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis, comprising: 1) acquiring an image sequence simultaneously containing load or nominal stress values and the sample surface; 2) identifying the values in the image, and if they are load values, converting them into stress values based on the sample size; setting an inspection point in the image area where strain needs to be calculated, and obtaining corresponding points of the inspection point on other images in the sequence through image matching; 3) if the image sequence is acquired by a single device, calculating its two-dimensional displacement based on the coordinates of the inspection point and its corresponding points at different times; if the image sequence is acquired by multiple devices, reconstructing the three-dimensional coordinates of the inspection point, and calculating its three-dimensional displacement based on the three-dimensional coordinates of the inspection point at different times; 4) obtaining a strain result synchronized with the nominal stress time from the displacement result. The present invention can complete the synchronous acquisition of the sample surface strain field and the nominal stress measurement results without the need for communication between the load measurement device and the image acquisition device.

Description

Technical Field

[0001] The present invention relates to the technical field of solid material mechanical property testing, and in particular to a method for synchronously acquiring stress and strain data of material mechanics experiments based on image analysis. Background Art

[0002] Mechanical testing of solid materials to determine their stress-strain relationship is a primary experimental method for studying and evaluating the mechanical properties of materials. Stress is typically calculated from load data collected by the force sensor of a materials testing machine. Currently, non-contact optical measurement methods are commonly used to measure the surface strain field of material specimens. Image acquisition systems are used to capture a sequence of images of the specimen surface before and after deformation. Image analysis is then used to determine the specimen surface displacement field at each moment of image acquisition. Based on this, the surface strain field at each moment is calculated.

[0003] The image acquisition system and the material testing machine used to obtain the image sequence are two independent systems. If the current stress and surface strain field of the specimen need to be acquired synchronously, the two independent systems need to communicate and unify the time point of data acquisition. For newer material testing machines, manufacturers often develop supporting modules for optical measurement, including image acquisition systems and image analysis algorithms. However, their versatility is generally low and they cannot be used with testing machines of other brands or earlier models of the same brand. They are also relatively expensive. If the experimenter builds an image acquisition device for optical measurement by himself, in order to synchronously acquire the current stress and surface strain field of the specimen, secondary development is required based on the hardware and software interfaces provided by the testing machine manufacturer. This requires a high level of technical expertise and is often limited by the interface of the testing machine's software development toolkit. Summary of the Invention

[0004] The purpose of the present invention is to overcome the shortcomings and deficiencies of the existing technology and propose a method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis. The method collects an image sequence that contains both the load or stress value and the sample surface, analyzes the image sequence to extract the stress value and the sample surface strain field, and can achieve precise synchronization of the acquisition time of the stress and sample surface strain field measurement results without information interaction between the image acquisition system and the material testing machine.

[0005] To achieve the above-mentioned purpose, the technical solution provided by the present invention is: a method for synchronously acquiring stress and strain data of material mechanics experiments based on image analysis, comprising the following steps:

[0006] 1) During the experiment, an image sequence containing both the load or nominal stress value and the specimen surface is acquired;

[0007] 2) Identify the load or nominal stress value of each image in the image sequence. If the identified value is a load, it needs to be converted into a nominal stress value based on the pre-determined specimen geometry. At the same time, set an inspection point in a certain image and obtain the corresponding point of the inspection point in other images in the sequence through image matching;

[0008] 3) If the image sequence is acquired by a single device, the displacement of the observation point is directly calculated based on the coordinates of the observation point and its corresponding point in step 2). If the image sequence is acquired by multiple devices, the three-dimensional coordinates of the observation point must first be reconstructed based on the coordinates of the observation point in step 2) and the coordinates of the corresponding points in images acquired at other times. The displacement is then calculated based on the three-dimensional coordinates of the observation point at different times.

[0009] 4) The displacement calculated in step 3) is further used to obtain a strain value that is synchronized with the nominal stress value in time.

[0010] Furthermore, in step 1), the load or nominal stress value and the image sequence are acquired by two independent systems respectively, and there is no information communication between the two systems; wherein, the image sequence acquisition system is composed of at least one image acquisition device. If it is composed of multiple image acquisition devices, each device acquires images synchronously, and each device needs to be calibrated to obtain calibration parameters, including imaging parameters and relative positional relationships between devices. There must be a device that displays the load or nominal stress value within the field of view of the image acquisition device.

[0011] Furthermore, in step 1), the random texture pattern naturally existing on the sample surface or artificially prepared is used to carry the deformation information in the sample surface image.

[0012] Furthermore, in step 2), if the value in the image is the load value F, the nominal stress value σ of the sample is converted based on the pre-measured average width w and average thickness h of the sample cross section:

[0013]

[0014] Further, the step 2) includes the following steps:

[0015] 2.1) Define the image area where strain calculation is required in the image at time a, and set observation points within this area. The observation points are pixels and are expressed as:

[0016]

[0017] Where, Refers to the i-th observation point in the image at time a, x i and y i They are The pixel coordinates of the X-axis and Y-axis in image coordinates, N is the total number of observation points;

[0018] 2.2) If a single image acquisition device is used, the image matching The corresponding point in the image at time j Where M is the total number of images; if multiple image acquisition devices are used, image matching is performed to obtain The corresponding point in the image captured by the kth device at time j as well as The corresponding points in the image collected at time j on the same device Where k = 2, 3, 4, ..., S, and S is the total number of image acquisition devices.

[0019] Furthermore, in step 3), the i-th inspection point The displacement between time b and time c Depend on The coordinates of the corresponding points at two moments are subtracted. If the image sequence is acquired by a single device, the displacement For inspection points The two-dimensional pixel coordinates in the image coordinate system are expressed as:

[0020]

[0021] Where, b∈1,2,3,...,M, c=1,2,3,...,M, and They are At the corresponding points at time b and time c;

[0022] If the image sequence is acquired by multiple devices, the calibration parameters of the image acquisition devices and and The pixel coordinates of any two points in the The three-dimensional coordinates at time j in the world coordinate system The world coordinate system is a three-dimensional coordinate system with the optical center of the lens of the image acquisition device as the origin, the optical axis as the Z axis, and the horizontal and vertical arrangement directions of the photosensitive elements as the X axis and Y axis, which can be used to calculate the inspection point. The three-dimensional displacement from time b to time c is expressed as:

[0023]

[0024] Where, and They are The three-dimensional coordinates obtained by reconstruction at time b and time c.

[0025] Furthermore, in step 4), the continuous displacement field function is fitted by the discrete displacement data to obtain the displacement gradient, and then the displacement gradient of each observation point is calculated. strain.

[0026] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0027] 1. Compared with the method of synchronously acquiring the strain field and stress measurement results of the specimen surface through information interaction between the material testing machine and the image acquisition system, the present invention provides a method for synchronously acquiring the stress and specimen surface strain field measurement results without the need for communication between two independent systems. The method is easy to implement.

[0028] 2. Compared with using the optical measurement module provided by the testing machine manufacturer, the present invention has no specific matching requirements for the material testing machine and the image acquisition system of the optical measurement module, has strong versatility and is relatively low in cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a schematic diagram of the hardware structure used in the present invention.

[0030] Figure 2 This is the camera optical path diagram of the image acquisition system.

[0031] Figure 3 This is the image taken by Cam1 at the initial moment.

[0032] Figure 4 Flowchart for digital image analysis.

[0033] Figure 5 Schematic diagram of image processing for extracting stress information.

[0034] Figure 6 Schematic diagram of the image area and inspection point distribution where strain needs to be calculated.

[0035] Figure 7 Schematic diagram of the sample surface displacement field calculation. DETAILED DESCRIPTION

[0036] The present invention will be further described in detail below with reference to the embodiments and the accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0037] This embodiment provides a method for synchronously acquiring stress and strain data from a material mechanics experiment based on image analysis, comprising the following steps:

[0038] 1) During the experiment, an image sequence containing both the load or nominal stress values and the specimen surface is acquired, as follows:

[0039] The load or nominal stress values and image sequences are acquired by two independent systems, with no information communication between the two systems. The image sequence acquisition system consists of at least one image acquisition device. If it consists of multiple image acquisition devices, each device acquires images synchronously and is calibrated to obtain calibration parameters, including imaging parameters and the relative position relationship between devices. A device that displays the load or nominal stress values must be within the field of view of the image acquisition device.

[0040] The random texture patterns naturally existing or artificially prepared on the sample surface are used to carry the deformation information in the sample surface image.

[0041] 2) Identify the load or stress value of each image in the image sequence. If the identified value is a load, it needs to be converted into a stress value based on the pre-determined geometric dimensions of the specimen. At the same time, set an inspection point in a certain image and obtain the corresponding point of the inspection point on other images in the sequence through image matching. The specific situation is as follows:

[0042] If the value in the image is the load value F, the nominal stress value σ of the specimen can be obtained by converting it according to the pre-measured average width w and average thickness h of the specimen cross section:

[0043]

[0044] In the image at time a, the image area where strain calculation is required is defined, and the observation points are set in this area. The observation points are pixel points, which are expressed as:

[0045]

[0046] Where, Refers to the i-th observation point in the image at time a, x i and y i They are The pixel coordinates of the X-axis and Y-axis in image coordinates, N is the total number of observation points;

[0047] If a single image acquisition device is used, the image matching The corresponding point in the image at time j Where M is the total number of images; if multiple image acquisition devices are used, image matching is performed to obtain The corresponding point in the image captured by the kth device at time j as well as The corresponding points in the image collected at time j on the same device Where k = 2, 3, 4, ..., S, and S is the total number of image acquisition devices.

[0048] 3) If the image sequence is acquired by a single device, the displacement of the observation point is directly calculated based on the coordinates of the observation point and its corresponding point in step 2). If the image sequence is acquired by multiple devices, the three-dimensional coordinates of the observation point must first be reconstructed based on the coordinates of the observation point in step 2) and the coordinates of the corresponding points in the images acquired at other times. The displacement is then calculated based on the three-dimensional coordinates of the observation point at different times. The specific situation is as follows:

[0049] The i-th inspection point The displacement between time b and time c Depend on The coordinates of the corresponding points at two moments are subtracted. If the image sequence is acquired by a single device, the displacement For inspection points The two-dimensional pixel coordinates in the image coordinate system are expressed as:

[0050]

[0051] Where, b∈1,2,3,...,M, c=1,2,3,...,M, and They are At the corresponding points at time b and time c;

[0052] If the image sequence is acquired by multiple devices, the calibration parameters of the image acquisition devices and and The pixel coordinates of any two points in the The three-dimensional coordinates at time j in the world coordinate system The world coordinate system is a three-dimensional coordinate system with the optical center of the lens of the image acquisition device as the origin, the optical axis as the Z axis, and the horizontal and vertical arrangement directions of the photosensitive elements as the X axis and Y axis, which can be used to calculate the inspection point. The three-dimensional displacement from time b to time c is expressed as:

[0053]

[0054] Where, and They are The three-dimensional coordinates obtained by reconstruction at time b and time c.

[0055] 4) The displacement calculated in step 3) is further used to obtain the strain value synchronized with the nominal stress value, as follows:

[0056] The continuous displacement field function is fitted by discrete displacement data to obtain the displacement gradient, and then the displacement gradient of each observation point is calculated. strain.

[0057] Let's take the uniaxial tensile test as an example. Figures 1 to 7 The above-mentioned method of synchronously obtaining stress and strain data of material mechanics experiments based on image analysis is described in detail.

[0058] 1. Hardware system construction and image acquisition

[0059] Figure 1 The hardware required to implement the method of the present invention is shown, including: an image acquisition system 1, a host 2 for image sequence acquisition and image analysis, a small display 3, and a computer 4 for controlling the material testing machine and collecting load data. Figure 2 This is a camera optical path diagram of the image acquisition system. In this experiment, the image acquisition system consists of two industrial cameras, Cam1 and Cam2, each equipped with a 16mm fixed-focus lens. The optical axes of the two cameras are angled at 30°. Universal input and output interfaces are used to connect the two cameras for synchronized triggering of image acquisition. The industrial cameras are mounted on a tripod and placed in front of the test sample. A camera calibration algorithm is used to calibrate the image acquisition system, obtaining its calibration parameters, including the imaging parameters of each industrial camera and the relative positional relationship between the two cameras.

[0060] A small monitor is configured for the computer that controls the material testing machine and collects load data. Use split-screen software to display the load data from the testing machine's main control software on this monitor. The load data display area on this monitor should be located within the space that the image acquisition system can clearly image. Figure 3 The position of the display relative to the test specimen is shown, with the display on the left and the test specimen on the right.

[0061] Use spray paint to create randomly distributed spots on the sample surface to carry information about the sample's surface deformation. Clamp the sample in the testing machine fixture, load it, and use Cam1 and Cam2 to simultaneously capture image sequences from different perspectives.

[0062] 2. Digital Image Analysis

[0063] Figure 4 This is a flowchart of digital image processing. The left and right sides show the process of extracting stress information and the strain field of the specimen surface, respectively.

[0064] First, the image sequence is cropped to obtain the area containing the load information. After rotating the image, the image contrast is improved by adaptive histogram equalization with limited contrast. Figure 5 The image processing process is demonstrated; then, the load value F in the image is identified using a deep learning-based optical character recognition algorithm. The nominal stress value σ of the specimen at the time of image acquisition is then calculated based on the average width w and average thickness h of the specimen cross section:

[0065]

[0066] After obtaining the nominal stress value of the specimen, it is necessary to calculate the deformation field of the specimen surface based on the image sequence. The specific steps are as follows:

[0067] 1) In the image collected at the initial moment of Cam1, the image area where strain calculation is required is defined, with a size of 180×630 pixels. Within this area, a series of observation points are evenly set at intervals of 30 pixels along the x and y directions of the image coordinate system. Where N is the total number of inspection points set. Figure 6 The image area where strain needs to be calculated and the set observation points are displayed.

[0068] 2) Based on the inspection point A rectangular area with a size of 41×41 pixels is selected as the center, and the corresponding position of the image in the area is searched in the deformed sequence image by image matching, so as to obtain The corresponding point. Figure 7 The image matching strategy is shown. Through image matching, we can get The corresponding points in the image collected at the initial moment of Cam2 And the corresponding points on the image collected by Cam1 and Cam2 at time j and

[0069] 3) According to and as well as and Combined with the imaging system parameters, the survey point can be calculated by applying the triangulation method The three-dimensional point coordinates corresponding to the initial time and time j and The observation point at time j can be obtained The three-dimensional displacement is as follows:

[0070]

[0071] In order to ensure the simplicity of the subsequent formula expression, the symbols of formula (2) are simplified to the following form:

[0072]

[0073] 4) Take Each calculation point within a 90-pixel radius is the center Three-dimensional coordinates at the initial moment and the three-dimensional displacement at time j Perform local fitting to calculate strain. First calculate the displacement gradient, and assemble the matrices A and b and the displacement gradient vector X according to the following formulaU :

[0074]

[0075] Solve AX U =b, and the displacement gradient is obtained. Then the Green strain tensor is calculated from the displacement gradient according to the following formula:

[0076]

[0077]

[0078]

[0079] e xy =0.5(u y +v x +u y u x +v y v x +w y w x )

[0080] e yz =0.5(v z +w y +u z u y +v z v y +w z w y )

[0081] e zx =0.5(w x +u z +u x u z +v x v z +w x w z )

[0082] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis, characterized in that: The following steps are involved: 1) During the experiment, an image sequence containing both the load or nominal stress value and the specimen surface is acquired; 2) Identify the load or nominal stress value of each image in the image sequence. If the identified value is a load, it needs to be converted into a nominal stress value based on the pre-determined geometric dimensions of the specimen. At the same time, set an inspection point in a certain image and obtain the corresponding point of the inspection point on other images in the sequence through image matching. The following steps are included: 2.1) Define the image area where strain calculation is required in the image at time a, and set observation points within this area. The observation points are pixels and are expressed as: Where, Refers to the i-th observation point in the image at time a, and They are The pixel coordinates of the X-axis and Y-axis in image coordinates, N is the total number of observation points; 2.2) If a single image acquisition device is used, the image matching The corresponding point in the image at time j j=1,2,3,...,M, where M is the total number of images; if multiple image acquisition devices are used, image matching is performed to obtain The corresponding point in the image captured by the kth device at time j as well as The corresponding points in the image collected at time j on the same device Where k = 2, 3, 4, ..., S, S is the total number of image acquisition devices; 3) If the image sequence is acquired by a single device, the displacement of the observation point is directly calculated based on the coordinates of the observation point and its corresponding point in step 2). If the image sequence is acquired by multiple devices, the three-dimensional coordinates of the observation point must first be reconstructed based on the coordinates of the observation point in step 2) and the coordinates of the corresponding points in images acquired at other times. The displacement is then calculated based on the three-dimensional coordinates of the observation point at different times. The i-th inspection point The displacement between time b and time c Depend on The coordinates of the corresponding points at two moments are subtracted. If the image sequence is acquired by a single device, the displacement For inspection points The two-dimensional pixel coordinates in the image coordinate system are expressed as: Where, b=1,2,3,...,M, c=1,2,3,...,M, and They are At the corresponding points at time b and time c; If the image sequence is acquired by multiple devices, the calibration parameters of the image acquisition devices and and The pixel coordinates of any two points in the The three-dimensional coordinates at time j in the world coordinate system The world coordinate system is a three-dimensional coordinate system with the optical center of the lens of the image acquisition device as the origin, the optical axis as the Z axis, and the horizontal and vertical arrangement directions of the photosensitive elements as the X axis and Y axis, which can be used to calculate the inspection point. The three-dimensional displacement from time b to time c is expressed as: Where, and They are The three-dimensional coordinates obtained by reconstruction at time b and time c; 4) The displacement calculated in step 3) is further used to obtain a strain value that is synchronized with the nominal stress value in time.

2. The method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis according to claim 1, characterized in that: In step 1), the load or nominal stress value and the image sequence are acquired by two independent systems, respectively, and there is no information communication between the two systems; wherein, the image sequence acquisition system is composed of at least one image acquisition device. If it is composed of multiple image acquisition devices, each device acquires images synchronously, and each device needs to be calibrated to obtain calibration parameters, including imaging parameters and the relative position relationship between devices. A device that displays the load or nominal stress value must be within the field of view of the image acquisition device.

3. The method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis according to claim 2, characterized in that: In step 1), the random texture pattern naturally existing on the sample surface or artificially prepared is used to carry the deformation information in the sample surface image.

4. The method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis according to claim 3, characterized in that: In step 2), if the value in the image is the load value F, the nominal stress value σ of the sample is converted based on the pre-measured average width w and average thickness h of the sample cross section:

5. The method for synchronously acquiring stress and strain data from material mechanics experiments based on image analysis according to claim 4, characterized in that: In step 4), the continuous displacement field function is fitted by the discrete displacement data to obtain the displacement gradient, and then the displacement gradient of each observation point is calculated. strain.

Images