Rotor blade torsional rigidity testing device and method

Abstract

The invention provides a rotor blade torsional rigidity testing device, and the device comprises a blade fixing tool which is used for fixing a blade; the circular coding marks are adhered to the two ends of the measured section of the paddle; the loading tray is used for applying force to the paddle; the image acquisition system is arranged above the paddles, and the image acquisition system is used for acquiring images; meanwhile, the invention further provides a rotor blade torsional rigidity testing method, coding marks are pasted on the surfaces of the blades, three-dimensional plane information is fitted through the MPS industrial camera, the torsional rigidity of the blades is obtained through calculation, and the rotor blade torsional rigidity testing method has the advantages of being easy to install, high in precision and high in stability.

Description

Technical Field

[0001] This application belongs to the technical field of rotor blade testing, and specifically relates to a rotor blade torsional stiffness testing device and method. Background Art

[0002] As a key component of a helicopter, the static characteristics of the blade directly affect the flight performance of the helicopter and have an important impact on flight safety. The helicopter design manual has clear requirements for the static characteristics of the blade, and there are also specific test requirements in relevant standards and specifications. Torsional stiffness, as an important characteristic of the blade's static characteristics, needs to be obtained through experimental testing. In traditional blade torsional stiffness testing experiments, generally, laser displacement sensors are arranged on both sides of the blade pitch axis to measure the displacement changes of the leading and trailing edges in the measurement profile section to achieve the measurement of torsional stiffness. However, this method has several disadvantages:

[0003] (1) At least 4 displacement sensors and related auxiliary facilities need to be installed, and the installation workload is relatively large;

[0004] (2) Since the measurement reference point changes with the deformation of the test piece, the measurement results in a deviation between the measured value and the true value;

[0005] (3) Affected by the installation space of the sensors, the length of the measured profile section cannot be small enough; due to the small measurement contact points of the sensors, high requirements are imposed on the surface smoothness and flatness of the test piece;

[0006] (4) Since the torsional deformation of the blade is a multi-directional vector deformation, and the measurement method of traditional sensors can only perform single-direction measurement, the measurement accuracy is affected. Summary of the Invention

[0007] Object of the Invention: Provided is a rotor blade torsional stiffness measurement device and method based on image recognition technology. Coding marks are pasted on the blade surface, and the three-dimensional plane information is fitted by an MPS industrial camera to calculate the torsional stiffness of the blade, which has the characteristics of simple installation, high accuracy, and high stability.

[0008] This application provides a rotor blade torsional stiffness testing device, and the device includes:

[0009] A blade fixing tooling for fixing the blade;

[0010] Circular coding marks pasted at both ends of the measured profile of the blade;

[0011] A loading tray for applying force to the blade;

[0012] An image acquisition system arranged above the blade, and the image acquisition system is used for acquiring images.

[0013] Preferably, the image acquisition system includes:

[0014] Two industrial cameras, arranged above the blade, and the industrial cameras are used to acquire images.

[0015] In a second aspect, the present application further provides a method for testing the torsional stiffness of a rotor blade, and the method includes:

[0016] Paste circular coded markers on the blade surface;

[0017] Arrange two industrial cameras for high-precision real-time measurement above the circular coded markers;

[0018] Perform single-object calibration on the left and right industrial cameras respectively to obtain the internal parameter matrix K of the cameras;

[0019] Among them,

[0020] f x and f y are the focal lengths of the camera in the horizontal and vertical directions respectively; c x and c y are the coordinates of the principal point on the image plane respectively;

[0021] Perform binocular calibration, calculate the reprojection matrix Q according to the results obtained from single-object calibration, and realize the conversion from pixel coordinates to three-dimensional coordinates;

[0022] Detect the circular coded markers, obtain the center coordinates of all circular codes, fit them into a plane, and calculate the normal vector of this plane After step-by-step loading, when the blade is twisted, calculate the new plane normal vector at this time

[0023] Calculate the torsional stiffness of the blade k = L / θ;

[0024] Among them, k is the torsional stiffness of the blade, L is the applied torque, and θ is the torsional angle, which can be obtained according to the normal vectors of the two planes before and after torsion:

[0025]

[0026] Preferably, the performing binocular calibration, calculating the reprojection matrix Q according to the results obtained from single-object calibration, and realizing the conversion from pixel coordinates to three-dimensional coordinates includes:

[0027] The specific conversion is as follows: P is a point in the world coordinate system, L is the length of the camera plane, PL and PR are the points projected onto the two calibrated camera planes respectively, XL and OR represent the distances of PL and PR from the left edges of their respective mirror planes, and OL and OR represent the intersection points of the midpoints of the two camera planes and the projection points. According to the similarity triangle theorem, the coordinate z of P in the world coordinate system is:

[0028]

[0029] b is the actual distance between the two optical centers of the cameras, and K x is the internal parameter matrix obtained by calibrating the left camera, and u L , u R respectively represent the pixel distances of PL and PR from the left edges of their respective image planes. Let d = u L - u R . Similarly, the x and y coordinates can also be obtained by the similarity principle:

[0030]

[0031] K x , K y respectively represent the internal parameter matrices of the two cameras. Since the two industrial cameras used have the same specifications, therefore, K x = K y ;

[0032] u and v represent the coordinates of a point in the pixel coordinate system, and u0 and v0 are the coordinate values of the origin of the left camera image plane in the pixel coordinate system;

[0033] Therefore, x and y can be expressed as:

[0034]

[0035] Combining the coordinates of x, y, and z together, we get:

[0036]

[0037] Then the three-dimensional coordinates of the point in space are:

[0038]

[0039] Preferably, the normal vectors of the two planes are represented by the three-dimensional coordinates of the circular marker points. Assume that are the normal vectors before and after the twist of the blade respectively. The three-dimensional coordinates of any three non-collinear circular marker points on the blade before the twist are p a (x a , y a , z a ), p b (xb , y b , z b ), p c (x c , y c , z c ), after torsion, the coordinates of these three points are p' a (x' a , y' a , z' a ), p' b (x' b , y' b , z' b ), p' c (x' c , y' c , z' c ), then there is:

[0040]

[0041] Preferably, the circular coding marks are respectively pasted at both ends of the measured section.

[0042] Preferably, before pasting the circular coding marks on the blade surface, it further includes:

[0043] Constructing an image acquisition system.

[0044] Preferably, the image acquisition system includes two industrial cameras.

[0045] This application has the following technical effects:

[0046] The method of the present invention does not require any auxiliary equipment, only two photogrammetry cameras need to be installed at appropriate positions; the measurement points can be arranged at any position on the blade surface, without being restricted by space and the number of measurement points, and has the advantages of simple installation and high precision. Description of the Drawings

[0047] Figure 1 is a flowchart of a test method provided by an embodiment of this application;

[0048] Figure 2 is a field diagram of torsion loading;

[0049] Figure 3 is a physical diagram of a planar target;

[0050] Figure 4 is a coordinate schematic diagram after binocular calibration;

[0051] Figure 5 is a schematic diagram of blade torsion. Detailed Embodiments

[0052] Please refer toFigures 1 - 5 , based on image recognition technology, the present invention converts the displacements of the leading and trailing edges of the blade section interval into three-dimensional plane XYZ coordinate values, obtains the change in the twist angle from the three-dimensional plane information before and after deformation, and then measures the torsional stiffness of the blade.

[0053] The present invention provides a method for measuring the torsional stiffness of a rotor blade based on image recognition technology. By pasting coded markers on the blade surface and fitting the three-dimensional plane information through an MPS industrial camera, the torsional stiffness of the blade is calculated, which has the characteristics of simple installation, high precision, and high stability.

[0054] Its main steps include:

[0055] Step 1, construct an image acquisition system;

[0056] Among them, the image acquisition system includes two industrial cameras, a computer, a lighting source, a controller, etc., as Figure 1 shown.

[0057] Step 2, paste circular coded markers on the blade surface;

[0058] Among them, the circular codes are respectively pasted at both ends of the measured section, as Figure 1 shown.

[0059] Step 3, arrange two industrial cameras for high-precision real-time measurement above the coded markers.

[0060] Step 4, perform single calibration on the left and right industrial cameras respectively to obtain the internal parameter matrix K of the cameras;

[0061] Among them,

[0062] f x and f y are the focal lengths of the camera in the horizontal and vertical directions respectively; c x and c y are the coordinates of the principal point on the image plane respectively.

[0063] Step 5, perform binocular calibration, calculate the reprojection matrix Q according to the results obtained from single calibration, and realize the conversion from pixel coordinates to three-dimensional coordinates;

[0064] Among them, the specific conversion is as Figure 4 shown. P is a point in the world coordinate system, L is the length of the camera plane, PL and PR are the points projected onto the two calibrated camera planes respectively, XL and OR respectively represent the distances of PL and PR from the left edge of their respective mirror planes, and OL and OR respectively represent the intersection points of the midpoints of the two camera planes and the projection points. According to the similarity triangle theorem, the coordinate z of P in the world coordinate system is:

[0065]

[0066] b is the actual distance between the two optical centers of the cameras, and K x is the internal parameter matrix obtained by calibrating the left camera, and u L , u R respectively represent the pixel distances of PL and PR from the left edge of their respective image planes. Let d = u L - u R . Similarly, the x and y coordinates can also be obtained by the principle of similarity:

[0067]

[0068] K x , K y respectively represent the internal parameter matrices of the two cameras. Since the two industrial cameras used have the same specifications, K x = K y ;

[0069] u and v represent the coordinates of a point in the pixel coordinate system, and u0 and v0 are the coordinate values of the origin of the left camera image plane in the pixel coordinate system.

[0070] Therefore, x and y can be expressed as:

[0071]

[0072] By combining the coordinates of x, y, and z together, we can obtain:

[0073]

[0074] Then the three-dimensional coordinates of the point in space are:

[0075]

[0076] Step 6: Detect the circular coding marks, obtain the center coordinates of all circular codings, fit them into a plane, and calculate the normal vector of this plane After step-by-step loading, when the blade twists, calculate the new plane normal vector at this time As Figure 5 shown;

[0077] Among them, the normal vectors of the two planes can be represented by the three-dimensional coordinates of the circular marking points. Assume are the normal vectors before and after the blade twists respectively. The three-dimensional coordinates of any three non-collinear circular marking points on the blade before twisting are p a (x a , y a , z a ), p b (x b , yb , z b ), p c (x c , y c , z c ), after torsion, the coordinates of these three points are p' a (x' a , y' a , z' a ), p' b (x' b , y' b , z' b ), p' c (x' c , y' c , z' c ), then there is:

[0078]

[0079] Step 7: Calculate the blade torsional stiffness k = L / θ

[0080] where k is the blade torsional stiffness, L is the applied torque, and θ is the angle of torsion, which can be obtained from the normal vectors of the two planes before and after torsion:

[0081]

[0082] Technical effects:

[0083] Measurement area: ≥500 mm × 500 mm;

[0084] Resolution: ≥0.05 mm;

[0085] Measurement speed: 3 frames / s;

[0086] Error: <2% FS.

[0087] In a feasible implementation, the blade is fixed on the index plate, circular coding marks are pasted at both ends of the section to be measured, and two industrial cameras are arranged above the circular coding, as Figure 1 shown. Before the formal test, the two cameras need to be calibrated. First, individual camera calibration is performed. A precision ceramic plane target with circular coding marks is used as a template, as Figure 3 shown, including a guiding circle and the remaining 56 circles arranged in 8 rows and 7 columns, for a total of 57 circles. The distance between the centers of each circle is 35 mm. The left and right cameras each take 28 pictures, and the internal parameter matrices K of the two cameras are directly obtained through the OpenCV library functions;

[0088] Subsequently, stereo calibration is performed to obtain the reprojection matrix and achieve the conversion from pixel coordinates to three-dimensional coordinates. AsFigure 4 As shown in the figure, P is a point in the world coordinate system, L is the length of the camera plane, PL and PR are the points projected onto two calibrated camera planes respectively, XL and OR represent the distances from PL and PR to the left edges of their respective mirror planes, and OL and OR represent the intersection points of the midpoints of the two camera planes and the projection points. According to the similarity theorem of triangles, the coordinate z of P in the world coordinate system is:

[0089]

[0090] b is the actual distance between the two optical centers of the cameras, K x is the internal parameter matrix obtained by calibrating the left camera, u L , u R represent the pixel distances from PL and PR to the left edges of their respective image planes respectively. Let d = u L - u R . Similarly, the x and y coordinates can also be obtained by the similarity principle:

[0091]

[0092] K x , K y represent the internal parameter matrices of the two cameras respectively. Since the two industrial cameras used have the same specifications, K x = K y ;

[0093] u, v represent the coordinates of a point in the pixel coordinate system, and u0, v0 are the coordinate values of the origin of the left camera image plane in the pixel coordinate system.

[0094] Therefore, x and y can be expressed as:

[0095]

[0096] By combining the coordinates of x, y, and z together, we can obtain:

[0097]

[0098] Place the calibration board in the fields of view of the two cameras, take 13 groups of pictures, and the reprojection matrix Q can be directly obtained through OpenCV function processing. Therefore, the three-dimensional coordinates of a point in space can be expressed as:

[0099]

[0100] Subsequently, detect the circular coding marks pasted on the blade. Shoot the pasted area with two industrial cameras, obtain the three-dimensional coordinates of the centers of all circular codings based on coordinate transformation, fit them into a plane, and calculate the normal vector of this plane The specific calculation method is as follows: Assume that the three-dimensional coordinates of any three non-collinear circular marking points on the blade before torsion are p a (x a ,y a ,z a ), p b (x b ,y b ,z b ), p c (x c ,y c ,z c ). Then:

[0101]

[0102] Take this as the initial state image of the blade, apply torque step by step. After the blade is stable, take a picture of the circular coding mark image at this time. Assume that the coordinates of these three points are p' a (x' a ,y' a ,z' a ), p' b (x' b ,y' b ,z' b ), p' c (x' c ,y' c ,z' c ). Then there is:

[0103]

[0104] Therefore, the torsion angle of the blade can be obtained:

[0105]

[0106] The torsional stiffness of the blade is:

[0107]

[0108] L is the applied torque.

Claims

1. A rotor blade torsional stiffness test device, characterized in that The device includes: A blade fixing tooling for fixing the blade; A circular coding mark pasted at both ends of the measured section of the blade; A loading tray for applying force to the blade; An image acquisition system arranged above the blade, and the image acquisition system is used for acquiring images.

2. The device according to claim 1, characterized in that The image acquisition system includes: Two industrial cameras arranged above the blade, and the industrial cameras are used for acquiring images.

3. A method for testing the torsional stiffness of a rotor blade, characterized in that, The method includes: Pasting circular coding marks on the blade surface; Arranging two industrial cameras for high-precision real-time measurement above the circular coding marks; Performing single-object calibration on the left and right industrial cameras respectively to obtain the internal parameter matrix K of the cameras; Among them, f x and f y are the focal lengths of the camera in the horizontal and vertical directions, respectively; c x and c y are the coordinates of the principal point on the image plane, respectively; Performing binocular calibration, calculating the reprojection matrix Q according to the results obtained from single-object calibration, and realizing the conversion from pixel coordinates to three-dimensional coordinates; Detect the circular coding marks to obtain the center coordinates of all circular codings, fit them into a plane, and calculate the normal vector of this plane After step-by-step loading, when the blade twists, calculate the new plane normal vector at this time Calculating the blade torsional stiffness k = L / θ; Wherein, k is the blade torsional stiffness, L is the applied torque, and θ is the torsional angle, which can be obtained according to the normal vectors of the two planes before and after torsion:

4. The method according to claim 3, wherein The performing binocular calibration, calculating the reprojection matrix Q according to the results obtained from single-object calibration, and realizing the conversion from pixel coordinates to three-dimensional coordinates includes: The specific conversion is as follows: P is a point in the world coordinate system, L is the length of the camera plane, PL and PR are respectively the points projected onto the two calibrated camera planes, XL and OR respectively represent the distances from PL and PR to the left edges of their respective mirror planes, and OL and OR respectively represent the intersection points of the midpoints of the two camera planes and the projection points. According to the similarity triangle theorem, the coordinate z of P in the world coordinate system is: b is the actual distance between the two optical centers of the cameras, K x is the internal parameter matrix obtained by calibrating the left camera, u L , u R respectively represent the pixel distances of PL and PR from the left edge of their respective image planes. Let d = u L - u R , similarly, the x and y coordinates can also be obtained by the similarity principle: K x and K y represent the internal parameter matrices of two cameras respectively. Since the two industrial cameras used have the same specifications, therefore, K x = K y ; u and v represent the coordinates of a point in the pixel coordinate system, and u0 and v0 are the coordinate values of the origin of the left camera image plane in the pixel coordinate system; Therefore, x and y can be expressed as: Combining the coordinates of x, y, and z together, we get: Then the three-dimensional coordinates of the point in space are:

5. The method according to claim 4, characterized in that, The normal vectors of two planes are represented by the three-dimensional coordinates of circular marked points. Assume that are the normal vectors before and after the blade twists respectively. The three-dimensional coordinates of any three non-collinear circular marked points on the blade before twisting are p a (x a , y a , z a ), p b (x b , y b , z b ), p c (x c , y c , z c ). After twisting, the coordinates of these three points are p' a (x' a , y' a , z' a ), p' b (x' b , y' b , z' b ), p' c (x' c , y' c , z' c ). Then there are:

6. The method according to claim 4, characterized in that, The circular coding marks are respectively pasted at both ends of the measured section.

7. The method according to claim 4, characterized in that Before pasting the circular coding marks on the blade surface, it further includes: Constructing an image acquisition system.

8. The method according to claim 7, wherein The image acquisition system includes two industrial cameras.

Images