Iron tower anti-falling steel strand clamp bearing capacity evaluation method

Abstract

The application provides a kind of tower anti-falling steel strand clamp bearing capacity evaluation method, comprising: obtaining the groove surface pressure distribution and the outer layer steel wire surface image in the process of continuously applying bolt pretightening of anti-falling steel strand clamp;Through image recognition, the outer layer steel wire surface image is processed, the geometric profile change of the outer layer steel wire is extracted, and the plastic flattening morphology is obtained;According to the groove surface pressure distribution and the plastic flattening morphology, the actual contact area between the groove surface and the outer layer steel wire is calculated, the part that is no longer continuously expanded along the groove length direction in the actual contact area is identified, and is marked as the effective contact area;The slip critical load of the steel strand under different bolt pretightening conditions is obtained, the damage morphology and the complete morphology are combined, and the evaluation basis of the anti-slip bearing capacity of the steel strand is established.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for evaluating the load-bearing capacity of steel strand clamps used for tower fall prevention. Background Technology

[0002] Steel strand clamps for tower fall arrest are critical protective devices ensuring the safety of personnel working at heights in power transmission lines and communication towers. Their load-bearing reliability directly affects the lives of workers. With increasingly frequent tower maintenance, accurate assessment of clamp load-bearing capacity has become an unavoidable core issue in high-altitude work safety management. Current load-bearing capacity assessment methods are generally based on static test data under a single working condition, simply assuming a positive correlation between preload and load-bearing capacity, believing that continuously increasing the preload of the clamping bolts will result in higher anti-slipping capacity. This understanding ignores the differences in the mechanical response of steel strands as multi-strand twisted structures under different pressure conditions, leading to the frequent anomaly in practical applications where excessive preload actually reduces load-bearing capacity, a phenomenon that cannot be reasonably explained by the assessment system. The root of the problem lies in the fact that when preload is applied to the steel strand, moderate preload can indeed increase the normal pressure between the clamping groove surface and the steel strand, thereby increasing the static friction resistance of the contact surface, requiring the steel strand to withstand greater axial tensile force before slippage occurs. The mechanical behavior at this stage conforms to classical friction theory; the frictional force on the contact surface is proportional to the normal pressure, and the increase in preload directly translates into an improvement in anti-slip capability. However, the steel strand is not a solid structure; it is made up of a central steel wire and several layers of outer steel wires twisted at a specific angle, with natural gaps between layers and strands. When the pressure applied to the clamping groove exceeds a certain critical value, the outer single-strand steel wire will undergo plastic flattening deformation under concentrated load. The circular cross-section of the steel wire is compressed into an ellipse or even an approximate plane, and the deformed steel wire is forced into the gap between adjacent strands. This local embedding causes the clamping contact area, which was originally evenly distributed along the circumference of the steel strand, to shrink drastically. Only the contact area between the flattened steel wire and the groove surface bears the frictional force, while other undeformed areas are suspended and lose contact due to the embedded steel wire. For example, during the maintenance of a 220 kV transmission line, after technicians increased the preload by 30% from the standard value according to the manual, the critical slip load of the steel strand clamp in the simulated drop test decreased by 18% instead of increasing. Disassembly and inspection revealed obvious indentations and localized strand breakage in the outer steel wires. This non-monotonic variation in load-bearing capacity renders traditional assessment methods ineffective, failing to identify the critical inflection point where the preload transitions from a favorable to a harmful range. A key breakthrough in the load-bearing capacity assessment technology for tower anti-fall clamps lies in how to simultaneously capture the pressure distribution on the channel surface and the degree of steel wire deformation during the load-bearing capacity assessment process, and establish a dynamic correlation between these two factors and the critical slip load, thereby accurately locating the safe range for preload application. Summary of the Invention

[0003] This invention provides a method for evaluating the load-bearing capacity of steel strand clamps used for tower fall protection, the method comprising: Obtain images of the pressure distribution on the groove surface and the surface of the outer steel wire during the continuous application of bolt pre-tightening of the anti-fall steel strand clamp; By processing the surface image of the outer steel wire through image recognition, the geometric contour changes of the outer steel wire are extracted to obtain the plastic flattening morphology; Based on the pressure distribution and plastic flattening morphology of the groove surface, the actual contact area between the groove surface and the outer steel wire is calculated. The part that no longer expands continuously along the groove length direction but narrows is marked as the effective contact area. When the effective contact area is broken into discontinuous patches along the twist direction of the wire and the gaps between adjacent strands are opened, the damage morphology when the twist structure is damaged and the complete morphology when the twist structure is intact are extracted from the surface image of the outer wire of the same twist pitch section. The critical slip load of steel strand under different bolt preload conditions was obtained, and the evaluation basis for the anti-slip bearing capacity of steel strand was established by combining the failure morphology and the intact morphology. Based on the evaluation criteria, the pressure distribution on the groove surface is accumulated segmentally along the groove length to obtain the clamping resultant force and its concentration along the groove length. It is then classified into the corresponding failure morphology or integrity morphology to evaluate and obtain the slip critical load of the steel strand. The inflection point of the slip critical load changing from rising to falling during the continuous increase of bolt preload is analyzed. The bolt preload force recorded synchronously at the inflection point is determined as the safe range of the fastening operation. A stop command is generated and sent to the automatic fastening equipment to execute the stop fastening action.

[0004] Preferably, acquiring the groove surface pressure distribution and outer steel wire surface image of the anti-fall steel strand clamp during the continuous application of bolt pre-tightening includes: Pressure sensor collection points are arranged along the length of the groove on the inner wall of the clamping groove of the anti-fall steel strand clamp. Real-time pressure values ​​are obtained from each collection point when the bolts are continuously tightened. The pressure distribution data of the groove surface is obtained by summarizing the collection point locations. An industrial camera is installed at the lateral opening of the steel strand clamp, and a ring light source is used to continuously photograph the surface of the outer steel wire to obtain images of the outer steel wire surface. The time nodes of the bolt pre-tightening process are aligned with the pressure distribution data of the groove surface to obtain a synchronization sequence.

[0005] Preferably, the step of processing the surface image of the outer steel wire through image recognition to extract the geometric contour changes of the outer steel wire and obtain the plastic flattening morphology includes: The outer steel wire surface image is retrieved from the synchronization sequence, and Gaussian filtering is applied to each frame to denoise it, resulting in a denoised steel wire surface image. The Canny edge detection method is used to extract the light and dark transition lines from the denoised steel wire surface image, resulting in a set of edge pixels arranged along the direction of the steel wire. The edge pixel set is grouped and assigned to each single-strand outer layer steel wire segment according to the wire twist direction. For each segment, least squares ellipse fitting is used to obtain the cross-sectional ellipse parameters containing the major axis length and the minor axis length as the geometric profile. The cross-sectional flatness is obtained by calculating the ratio of the major axis length to the minor axis length of the same segment frame by frame, and the average value of the cross-sectional flatness of each segment in the initial stage of bolt pre-tightening is used as the circular reference flatness. Arrange the cross-sectional flattening of each frame along the timestamp, extract the segments that exceed the preset threshold of the circular reference flattening and mark them as plastic flattening segments, and summarize the cross-sectional elliptical parameters and positions of each plastic flattening segment to obtain the plastic flattening morphology.

[0006] Preferably, the step of calculating the actual contact area between the groove surface and the outer steel wire based on the pressure distribution and plastic flattening morphology of the groove surface, and marking the portion that no longer continuously expands and narrows along the groove length direction as the effective contact area, includes: Read the real-time pressure values ​​of each collection point from the pressure distribution data on the tank surface, and mark the collection points that exceed the contact pressure threshold as pressure contact points; At the same timestamp, the pressure contact point and the plastic flattening section are aligned and matched along the groove length direction. The short axis length of the cross section of the flattened outer steel wire in the matching section is taken as the contact arc width. Multiply it by the extension length of the section to obtain the contact surface element and accumulate to obtain the actual contact area. When the length of the contact surface element in adjacent segments stops increasing and the contact arc width narrows from the outside to the inside, the adjacent segments are merged and marked as the effective contact area.

[0007] Preferably, when the wire breaks into discontinuous patches along the twist direction in the effective contact area and adjacent strands open with gaps, extracting the damage morphology when the twist structure is damaged and the complete morphology when the twist structure is intact from the surface image of the outer wire in the same twist pitch section includes: The contact arc width within the effective contact area is scanned frame by frame along the twist direction. When the contact arc width drops from a non-zero value to zero and reappears as a non-zero value at a more distant twist direction position, it is determined that the fracture is into a discontinuous patch and the patch discontinuity point is recorded. If the edge lines of adjacent outer strands on both sides of the same twist pitch section are retrieved from the synchronous sequence, and the lateral pixel distance between the edge lines of adjacent strands exceeds the preset gap threshold and the gray level of the middle pixel falls back to the background gray level, it is determined that the adjacent strands have opened a gap, and the twist pitch section is marked as the twist structure destruction section.

[0008] Preferably, the twisted structure failure section includes: Based on the fracture timestamp of the broken section of the twisted structure, image frames of adjacent strands closely attached and outer steel wires continuously attached along the groove surface in the same twisted section before the fracture timestamp are retrieved from the synchronous sequence as complete morphology. Image frames acquired after the fracture timestamp, showing increased loosening of adjacent strands and embedding of outer strands into the gaps between adjacent strands, are retrieved as the damage morphology. This yields a morphology record where the complete morphology and the damage morphology of the same twist pitch segment are stored in pairs.

[0009] Preferably, the step of obtaining the slip critical load of the steel strand under different bolt preload conditions, and establishing an evaluation basis for the slip resistance bearing capacity of the steel strand by combining the failure morphology and the intact morphology, includes: For each morphology record, a tensile force is applied along the axial direction of the steel strand under the corresponding bolt preload condition and gradually increased. The relative displacement of the steel strand relative to the clamping groove surface is monitored simultaneously. The axial tensile force value when the relative displacement jumps from a slow linear increase to a rapid increase is recorded as the slip critical load of the working condition. The load record set is obtained by corresponding the bolt preload value one by one. According to the bolt preload values ​​in the load record set, corresponding to the complete or damaged morphology of the same lay length section, the slip critical load in the complete morphology section is classified into the rising section, and the slip critical load in the damaged morphology section is classified into the falling section. The resulting two-segment comparison relationship serves as the basis for evaluating the slip resistance capacity of the steel strand.

[0010] Preferably, the evaluation criteria for the anti-slip bearing capacity of the steel strand include: For each complete morphology frame, scan along the twist direction of the steel wire to find the seamless bonding pixel strip between the edge of the outer steel wire and the edge of the groove surface. Accumulate the length of the continuous and uninterrupted twist direction pixels to obtain the continuous bonding length. Read the discontinuous patches that have opened between adjacent strands in the frame of the damaged morphology, and measure the pixel spacing between the start and end points of two adjacent patches in the vertical twist direction to obtain the gap width between the discontinuous patches. Arrange the continuous bonding length and the gap width between discontinuous patches according to the bolt preload values ​​from small to large. Identify the bolt preload section that is covered by the continuous bonding length entering a downward trend and the gap width entering a synchronous upward trend. Obtain the comparison interval where the bonding length narrows, the gap width increases and the slip critical load decreases.

[0011] Preferably, the pressure distribution on the groove surface is accumulated segmentally along the groove length based on the evaluation criteria to obtain the clamping resultant force and its concentration along the groove length, which is then classified into the corresponding failure morphology or intact morphology to evaluate the slip critical load of the steel strand, including: The pressure distribution data of the tank surface under the current evaluation condition is divided into equal-length sub-segments along the tank length direction according to the preset segment length. The real-time pressure value of the current timestamp in each sub-segment is multiplied by the product of the distance between the collection points and the width of the tank, and then added point by point to obtain the segment clamping force. The segment clamping forces of each sub-segment are accumulated to obtain the clamping resultant force. The concentration is obtained by dividing the force value of the segment with the highest segment clamping force by the arithmetic mean of the clamping forces of all segments. The bolt preload value range is located in the corresponding reference range table of the evaluation basis based on the bolt preload value under the current evaluation condition. When the morphology range is marked as complete morphology, the sliding critical load is matched with the same key value in the rising segment by the clamping resultant force and concentration. When the morphology range is marked as damaged morphology, the sliding critical load is matched with the same key value in the falling segment.

[0012] Preferably, the analysis of the slip critical load at the inflection point where it changes from increasing to decreasing during the continuous increase of bolt preload, and the bolt preload recorded synchronously at the inflection point, is determined as the safe range for the tightening operation. A stop command is generated and sent to the automatic tightening equipment to execute the stop tightening action, including: The evaluation curve is obtained by connecting the slip critical load of each working condition in series from small to large according to the synchronously recorded bolt preload values, and the slope sequence is obtained by calculating the difference between adjacent points segment by segment. The bolt preload value corresponding to the position in the slope sequence where the positive value turns into a negative value and the negative value is continuously maintained for no less than a preset number of frames is determined as the inflection point, and the bolt preload value at the inflection point is used as the upper limit of the safe range. When the bolt preload applied by the automatic fastening equipment reaches the upper limit value, a stop command is generated and sent to the automatic fastening equipment, which then cuts off the tightening torque output to stop the fastening action.

[0013] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for evaluating the bearing capacity of steel strand clamps used in tower fall arrestors. Addressing the challenge of determining clamping effectiveness and the safe tightening range during bolt pre-tightening, the method simultaneously acquires images of the groove surface pressure distribution and the outer steel wire surface during continuous bolt pre-tightening. Image recognition extracts the geometric contour changes of the outer steel wire to obtain its plastic flattening morphology. Combined with the pressure distribution, the actual contact area between the groove surface and the outer steel wire is calculated. The portion that no longer continuously expands and narrows along the groove length is marked as the effective contact area. When the effective contact area breaks into discontinuous patches along the twist direction and adjacent strands open up, the bearing capacity is assessed from the same... The loose embedding morphology when the twisted structure is damaged and the stable bonding morphology when it is intact are extracted from the twisted section. The slip critical load under different working conditions is collected to establish a comparative evaluation basis for the decrease of slip critical load corresponding to the narrowing of bonding length and the increase of gap width. Then, the groove surface pressure is accumulated in segments along the groove length to obtain the clamping resultant force and concentration, and is included in the corresponding morphology to evaluate the slip critical load. The bolt preload corresponding to the inflection point of its rise to fall is identified as the safe range of the fastening operation. A stop command is generated and sent to the automatic fastening equipment, realizing the quantitative evaluation of the load-bearing capacity of the anti-fall clamp and the automatic safety control of the fastening process. Attached Figure Description

[0014] Figure 1 This is a flowchart of a method for evaluating the load-bearing capacity of steel strand clamps used for preventing falls from iron towers, according to the present invention.

[0015] Figure 2 This is a schematic diagram of a method for evaluating the load-bearing capacity of steel strand clamps for tower fall prevention according to the present invention.

[0016] Figure 3 This is another schematic diagram of the method for evaluating the bearing capacity of steel strand clamps for tower fall prevention according to the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] like Figures 1-3 This embodiment of a method for evaluating the load-bearing capacity of a steel strand clamp for tower fall protection may specifically include: S101. Obtain images of the groove pressure distribution and the outer steel wire surface during the continuous application of bolt pre-tightening on the anti-fall steel strand clamp.

[0019] Pressure sensor data collection points are arranged along the length of the groove on the inner wall of the clamping groove of the anti-fall steel strand clamp. Real-time pressure values ​​are acquired from each collection point as the bolts are continuously tightened. These values ​​are then aggregated according to the location of the collection points to obtain groove surface pressure distribution data. An industrial camera is installed at the lateral opening of the steel strand clamp, and a ring light source is used to continuously photograph the surface of the outer steel wire, obtaining images of the outer steel wire surface. These images are then aligned with the time points of the bolt pre-tightening process and the groove surface pressure distribution data to obtain a synchronization sequence of the groove surface pressure distribution and the outer steel wire surface images.

[0020] In one embodiment, the clamping groove of the anti-fall steel strand clamp adopts a two-part structure, with the upper and lower parts clamping the steel strand after being fastened together with bolts. Several shallow grooves are evenly formed on the inner wall of the upper part of the groove along its length. Each shallow groove embeds a flexible thin-film pressure sensor, and the lead wire of the pressure sensor extends along the edge of the groove to an external data acquisition box. Throughout the continuous tightening of the bolts, each acquisition point continuously outputs a real-time pressure value. The data acquisition box summarizes the pressure values ​​according to the location number of the acquisition points, forming pressure distribution data of the groove surface arranged along its length.

[0021] Specifically, the pressure sensor operates at a constant sampling frequency, ranging from 10Hz to 50Hz, covering the entire propulsion process from a loose state to a fully tightened state. Each sampling outputs a frame of pressure distribution data arranged along the length of the groove.

[0022] In one possible implementation, the number of sampling points is determined by the length of the wire clamp slot. The sampling points are arranged at equal intervals of 5mm to 10mm along the length of the slot; 8 sampling points are arranged for shorter slots, and up to 16 for longer slots. They are arranged in a single row along the width of the slot to capture the pressure gradient along the axial direction of the steel strand. An industrial camera is mounted at the lateral opening of the steel strand clamp, with the camera lens facing the exposed section of the outer steel wire. A ring light source surrounds the camera, illuminating the steel wire surface at an angle of 15 to 30 degrees to the wire axis to avoid metallic reflection and create bright spots, ensuring clear imaging of the surface texture and contour edges of the outer steel wire. The camera continuously captures images of the outer steel wire surface at the same sampling frequency as the pressure sensor, outputting one frame of the outer steel wire surface image with each exposure.

[0023] It should be noted that the pressure acquisition box and the industrial camera share the same time base. Their sampling clocks are driven by the same trigger signal. At each sampling moment, while the acquisition box records a frame of pressure distribution data, the camera simultaneously exposes and outputs a frame of image of the outer steel wire surface. The two data streams are numbered and paired according to their timestamps to obtain a synchronization sequence of the pressure distribution on the trench surface and the image of the outer steel wire surface. This synchronization sequence, indexed by the timestamp, is used for subsequent image recognition and morphology extraction.

[0024] S102. By processing the surface image of the outer steel wire through image recognition, the geometric contour changes of the outer steel wire are extracted to obtain the plastic flattening morphology.

[0025] The surface image of the outer steel wire is retrieved from the synchronization sequence. Gaussian filtering is applied to each frame to denoise and suppress speckle noise caused by metallic reflection, resulting in a denoised steel wire surface image. Canny edge detection is then used to extract the light-dark transition lines on the steel wire surface, yielding a set of edge pixels arranged along the wire's direction. Based on this set of edge pixels, the pixels are grouped according to the wire twist direction and assigned to the segments containing each single strand of the outer steel wire. Least-squares ellipse fitting is applied to the edge pixels of each segment to obtain the cross-sectional ellipse parameters of the outer steel wire in that segment. These parameters include the major axis length and the minor axis length, serving as the geometric contour of the outer steel wire. For the geometric profile, the ratio of the major axis length to the minor axis length within the same segment is calculated frame by frame to obtain the cross-sectional flatness of the outer steel wire in that segment. The average value of the cross-sectional flatness of each segment in the initial stage of bolt pre-tightening is taken as the circular reference flatness. The cross-sectional flatness of each frame is arranged along the timestamp. Segments exceeding the preset threshold of the circular reference flatness are extracted from the cross-sectional flatness and marked as plastic flattening segments. The cross-sectional ellipse parameters and locations of each plastic flattening segment are summarized to obtain the plastic flattening morphology of the outer steel wire.

[0026] In one implementation, images of the outer steel wire surface are retrieved sequentially from the synchronization sequence according to timestamps. For each frame of the original image, grayscale processing is performed first, compressing the color pixel channels into single-channel grayscale values. Due to the narrow opening of the clamping groove, the industrial camera is easily affected by specular reflection from the steel wire surface during imaging, resulting in scattered bright spots and random noise in the image. Direct edge extraction would misidentify a large number of false edges. Therefore, Gaussian filtering is used to smooth the grayscale image, with a Gaussian kernel size of 5×5 and a standard deviation of 1.2. This suppresses spot noise while preserving the low-frequency transition information of the steel wire contour, resulting in a denoised steel wire surface image.

[0027] Specifically, Canny edge detection is used to extract the transition lines between light and dark areas from the denoised wire surface image. The Canny operator first calculates the pixel gradient magnitude and direction, then refines the edges through non-maximum suppression, and finally obtains the closed contour using double-threshold hysteresis connection.

[0028] For example, the lower threshold of the dual threshold is set to the 10% quantile of the gradient histogram, and the higher threshold is set to the 40% quantile, resulting in a set of edge pixels arranged along the direction of the wire, with each pixel in the edge pixel set carrying its row and column coordinates.

[0029] It should be noted that steel strand is made by twisting a central steel wire with several outer strands of steel wire at a fixed helix angle. Each outer strand appears as a diagonally extending strip in the image. The lay length and helix angle of the steel strand are determined by the product model and specifications at the time of manufacture.

[0030] For example, the lay length of a standard 7-strand steel strand is 12 to 16 times the outer diameter of the steel strand, and the helix angle is 70 to 80 degrees. These specifications are read from the steel strand model file as the grouping basis. The image is divided into segments along the horizontal direction with a lay length as the period, and then the strips are cut along the diagonal direction of the lay. The pixels in the edge pixel set are assigned to the corresponding single-strand outer steel wire segment according to the strip they fall into, resulting in the grouped edge pixel set.

[0031] In one possible implementation, an industrial camera is mounted at the lateral opening of the clamping slot, with the lens facing the exposed section of the outer steel wire, acquiring a lateral image extending along the axial direction. For the strip-shaped edge pixels extracted by Canny, a sampling section is set every 5 pixels along the axial direction of the steel wire. The vertical distance between the upper and lower edge pixels at this section is measured as the projected width of the section at that location. The projected widths of all sampling sections within this section are counted, and the maximum value is taken as the major axis length 'a', and the minimum value as the minor axis length 'b'. Only the major axis length 'a' and the minor axis length 'b' are retained as the elliptical parameters of the outer steel wire section for this section; the principal axis tilt angle is not involved in subsequent determination. The elliptical parameters are stored using a dual index of section number and timestamp, forming the geometric contour of the outer steel wire. Further, for the geometric contour, the flattening e of the section for the same section is calculated frame by frame, equal to a divided by b. The closer e is to 1.0, the closer the cross-section of the outer steel wire in that section is to a circle; the larger e is, the higher the degree of flattening of that section.

[0032] Preferably, in the initial stage when the bolt pre-tightening has just begun, before the clamping groove surface has applied significant normal pressure to the steel strand, the arithmetic mean of the cross-sectional flatness of each segment of the first 10 frames of images in the initial stage is taken as the circular reference flatness e0.

[0033] For example, e0 is typically between 1.02 and 1.08, serving as the baseline for the cross-sectional flatness of the strand of that specification in an uncompressed state.

[0034] For example, the plastic flattening threshold is set to 1.3 times the circular reference flattening e0. This coefficient is based on the elastic-plastic deformation characteristics of the steel wire material: when the cross-sectional flattening reaches 1.3 times, according to finite element analysis of material mechanics, the surface of the steel wire has entered the yield plateau and begins to produce irreversible plastic flattening. This value is consistent with the statistical results of flattening tests on steel wires of the same specification. That is, when the cross-sectional flattening of a certain segment at a certain time point exceeds 1.3 times e0, the segment is determined to have entered the plastic flattening stage at that time point. The value of 1.3 times corresponds to the empirical range where the outer steel wire enters yield deformation. The cross-sectional flattening sequence of each segment is arranged along the time points. The difference between the flattening of adjacent frames is used to obtain the flattening increment sequence. When the value of the flattening increment sequence jumps from less than 0.01 to continuously greater than 0.03 and this state is maintained for more than 3 frames, the starting time point of this jump is identified as the lifting inflection point tp. tp corresponds to the moment when the outer steel wire transitions from elastic deformation to plastic deformation. Using tp as the starting point of the plastic segment, only frames with a timestamp greater than or equal to tp and a flattening ratio not less than 1.3 times e0 are included in the plastic flattening segment. Both conditions must be met simultaneously to exclude pseudo-plastic frames in the elastic rebound stage where the flattening ratio briefly exceeds 1.3 times e0 but the increment is still less than 0.01. The cross-sectional ellipse parameters, segment number, position coordinates along the groove length direction, and the corresponding tp value of all plastic flattening segments are summarized to obtain the plastic flattening morphology of the outer steel wire. This plastic flattening morphology is stored in a ternary record format with position, flattening ratio, and timestamp, with a tp field attached to each record. When retrieving this morphology by segment in the subsequent actual contact area calculation and effective contact area marking stages, the time axis is first truncated by tp, discarding the elastic data before tp. Then, within the time window after tp, effective contact frames are filtered according to the 1.3 times e0 threshold, thus ensuring that the inflection point identification result is integrated into both the plastic segment division and effective area marking stages.

[0035] S103. Based on the pressure distribution and plastic flattening morphology of the groove surface, calculate the actual contact area between the groove surface and the outer steel wire, identify the part of the actual contact area that no longer expands continuously along the groove length and narrows, and mark it as the effective contact area.

[0036] The location of each sampling point and its corresponding real-time pressure value are read from the pressure distribution data of the groove surface. Sampling points with pressure values ​​exceeding a preset contact pressure threshold are marked as pressure contact points, resulting in a pressure contact point distribution along the groove length. Based on the pressure contact point distribution and the plastic flattening morphology, the position coordinates of the pressure contact points and the position coordinates of the plastic flattening segment are aligned and matched along the groove length at the same timestamp. For the matched position segment, the minor axis length of the cross-section of the flattened outer steel wire in that segment is taken as the contact arc width. The contact arc width is multiplied by the extension length of that segment along the groove length to obtain the contact surface element of that segment. The contact surface elements of all matched segments are accumulated to obtain the actual contact area between the groove surface and the outer steel wire. Based on the distribution of the actual contact area along the groove length, the changes in the contact surface elements between adjacent sections are compared one by one. If, during the continuous increase of bolt preload, the expansion length of the contact surface elements of adjacent sections along the groove length stops increasing, and at the same time, the contact arc width narrows from the outside to the inside, then the adjacent sections are merged and marked as the effective contact area, thus obtaining the effective contact area distributed along the groove length.

[0037] In one implementation, each frame of data is sequentially read from the pressure distribution data of the trench surface according to its timestamp. Each frame of data includes the locations of several sampling points arranged along the length of the trench and their corresponding real-time pressure values. A preset contact pressure threshold is set to 0.5 MPa. This threshold is determined based on the upper limit of the background noise of the flexible thin-film pressure sensor under no-load contact. Only when the pressure value exceeds this threshold can it be determined that a substantial pressing contact has been formed between the outer steel wire of the steel strand and the trench surface. The real-time pressure value is compared with the contact pressure threshold point by point, and the sampling points with pressure values ​​exceeding 0.5 MPa are marked as pressure contact points, thus obtaining the distribution of pressure contact points arranged along the length of the trench.

[0038] Specifically, adjacent pressure contact points within the same frame are connected in series from near to far according to their position coordinates along the groove length. If the coordinate distance between two adjacent pressure contact points does not exceed the distance between one acquisition point, they are classified into the same pressure contact segment. The pressure contact segment is marked with the start and end coordinates along the groove length, thus obtaining several pressure contact segments arranged along the groove length at the current timestamp.

[0039] It should be noted that the pressure contact segments are arranged along the length of the groove, and the plastic flattening morphology stores the position coordinates of each plastic flattening segment along the groove length in the form of a ternary record of position, flattening ratio, and timestamp. Under the same timestamp, the start and end coordinates of the pressure contact segments and the start and end coordinates of the plastic flattening segments are mapped onto the same position axis along the groove length. The position axis takes the clamping groove surface as the origin and the millimeter as the scale unit. The coordinate intervals of each pressure contact segment and the coordinate intervals of the plastic flattening segments are compared one by one to see if there is any overlap. The intersection of the overlapping coordinate intervals is taken, and the coordinate range covered by the intersection is determined as the matching position segment.

[0040] In one possible implementation, for the matched position segment, the minor axis length *b* of the flattened outer steel wire within that segment is read from the plastic flattening morphology. The projection of the minor axis length *b* onto the groove plane is taken as the contact arc width *w*, which reflects the lateral dimension of the actual contact between the flattened outer steel wire and the groove surface. The extension length *L* of the segment along the groove length direction is obtained from the difference between the start and end coordinates of the matched position segment. The contact surface element *s* = *w* × *L* is calculated, where *s* is the contact surface element of the segment, *w* is the contact arc width, and *L* is the extension length along the groove length direction. The contact surface elements of all matched segments along the groove length direction at the current timestamp are accumulated segment by segment to obtain the actual contact area between the groove surface and the outer steel wire. Furthermore, the actual contact area changes dynamically as the bolt preload increases. In the initial stage of bolt preload, the actual contact area continues to expand to both sides along the groove length direction, and the extension length L of each matching section increases frame by frame. When the bolt preload reaches a certain time stamp, the extension length L of some matching sections no longer increases, and at the same time, unmatched discontinuous sections appear between adjacent sections along the groove length direction.

[0041] Preferably, the changes in contact surface elements between adjacent segments are compared one by one, and five consecutive frames are taken as the observation window. If the increment of the extension length L between any two adjacent frames within the observation window is equal to 0, it is determined that the extension length of the segment has stopped increasing. At the same time, the contact arc width w of the segment is read frame by frame, and the contact arc width w is defined as the transverse dimension along the normal direction of the groove surface in the cross section of the flattened steel wire. In order to avoid the overlap of the criteria with the flattening process in S102 where the short axis b decreases monotonically with the continuous increase of the preload, the absolute decrease of w is not used as the narrowing criterion. Instead, a relative rate attenuation ratio r = dw / dw0 is introduced, where dw is the decrease of w between the current two adjacent frames, and dw0 is the average value of the decrease of w between frames in the segment from the beginning of the flattening segment to the midpoint of the plastic segment in S102, which is used as the flattening reference rate of the segment itself. The range Δw of w within the observation window is statistically analyzed simultaneously. When r < 0.2 and Δw < 0.02 mm are simultaneously satisfied within 5 consecutive observation windows, it indicates that w has moved away from the flattening-dominated stage where it decreases synchronously with the preload and entered a steady-state narrowing stage dominated by the balance of contact surface reaction forces. This indicates that an effective contact arc width narrowing event has occurred in this section. Conversely, even if w continues to decrease, as long as r is close to 1, it is still considered a general attenuation of b during the flattening process and is not accepted. At the timestamp when the extension length stops increasing and the effective contact arc width narrowing occurs simultaneously, adjacent sections that meet the composite condition are merged along the groove length direction. The merged section is marked on the position axis with start and end coordinates, and is designated as the effective contact area. The effective contact area is saved as a four-element record containing the start coordinate, end coordinate, contact arc width w, and the timestamp, resulting in an effective contact area distributed along the groove length direction, which is then used for subsequent extraction of the twisted structure's damage and integrity morphology.

[0042] S104. When the bolt preload continues to increase, and the effective contact area breaks into discontinuous patches along the wire twist direction and the adjacent strands open gaps, extract the destruction morphology of the outer strands when the twist structure is destroyed and the complete morphology of the outer strands when the twist structure is intact from the outer wire surface images collected in the same twist pitch section during the continuous application of bolt preload.

[0043] Based on the quaternary record of the effective contact area, the morphological changes of the same effective contact area along the wire twist direction are tracked during the continuous increase of bolt preload. Within the same effective contact area, the contact arc width is scanned frame by frame along the twist direction. If the contact arc width at an adjacent twist position within the same effective contact area drops from a non-zero value to zero and then reappears at a more distant twist position, it is determined that the effective contact area is broken into discontinuous patches along the wire twist direction. The twist position where the contact arc width drops to zero is recorded as the patch discontinuity point. Based on the patch discontinuity point, the outer wire surface images of the same twist pitch segment continuously acquired after the fracture timestamp are retrieved from the synchronous sequence. The edge lines of adjacent outer strands on both sides of the same twist pitch segment are re-extracted along the twist direction in the image. The lateral pixel distance between the edge lines of adjacent strands is taken as the strand spacing. If the strand spacing expands from close to the baseline to exceed the preset gap threshold and the gray level of the middle pixel falls back to the background gray level, it is determined that the adjacent strands have opened a gap, and the twist pitch segment is marked as the twist structure failure segment. Based on the fracture timestamp of the fractured section of the twisted structure, image frames of the same twisted section acquired before the fracture timestamp, in which adjacent strands are closely attached and the outer steel wire is continuously attached along the groove surface, are retrieved from the synchronization sequence as complete morphology; image frames acquired after the fracture timestamp, in which adjacent strands are increasingly loose and the outer strand is embedded in the gap between adjacent strands, are retrieved as fracture morphology, thus obtaining a morphology record in which the complete morphology and fracture morphology of the same twisted section are stored in pairs.

[0044] In one embodiment, the steel strand is formed by twisting a central steel wire and several outer strands of steel wire at a fixed helix angle. Each outer strand of steel wire is tightly attached to the groove surface along the twist direction under bolt preload. The effective contact area is stored as a four-element record containing the start coordinate, end coordinate, contact arc width w, and the corresponding timestamp. The morphological changes of the same effective contact area are tracked as the bolt preload continuously increases. Specifically, the four-element record of the effective contact area is retrieved frame by frame from the current timestamp forward, and the changes in the contact arc width w and the drift of the start and end coordinates between adjacent timestamps are compared.

[0045] Specifically, the lay length of the steel strand is determined by the product model and specifications when it leaves the factory.

[0046] For example, the lay length of a standard 7-strand steel strand is 12 to 16 times the outer diameter of the steel strand. Within the same effective contact area, the contact arc width w is scanned point by point along the lay direction at lay-direction unit steps, with each lay-direction unit step being 1 / 20 of the lay length. The scanning sequence proceeds sequentially according to the spiral of the outer layer steel wire. The contact arc widths w at each scanning point are strung together to form a contact arc width curve arranged along the lay direction, with the horizontal axis of the contact arc width curve representing the lay direction position and the vertical axis representing the contact arc width w value.

[0047] It should be noted that when the bolt preload continues to increase until the outer steel wire enters the excessive flattening stage, the contact arc width curve no longer maintains a smooth, non-zero shape along the twist direction. Instead, it exhibits a dip to zero at several twist direction positions, followed by non-zero rising sections before and after the dips. If the contact arc width w at adjacent twist direction positions within the same effective contact area drops from a non-zero value to zero, and then reappears at a more distant twist direction position, the effective contact area is determined to have broken into discontinuous patches along the steel wire twist direction. The twist direction position where the contact arc width w drops to zero is recorded as the patch discontinuity point. The twist direction interval covered by the non-zero rising sections on both sides of the patch discontinuity point is marked as the discontinuous patch. The timestamp of the first appearance of the patch discontinuity point is recorded as the breakage timestamp. Further, based on the patch discontinuity points, images of the outer steel wire surface continuously acquired after the breakage timestamp are retrieved from the synchronous sequence for the same twist pitch segment. The length of the same twist pitch segment along the groove length corresponds to the length of one complete spiral twist of the outer steel wire.

[0048] In one possible implementation, the edge lines of adjacent outer strands on both sides of the same twist pitch segment are re-extracted along the twist direction in the image. Canny edge detection is then used to rescan the outer contour boundaries of the adjacent strands on the image frame after the break timestamp, resulting in two strand edge lines extending obliquely along the twist direction. The lateral pixel distance between adjacent strand edge lines along the direction perpendicular to the twist direction is taken as the strand spacing d. The strand spacing d is composed of two states: the close-fitting baseline value when the outer strands are tightly attached and the expanded value when they are loosely spread apart.

[0049] Preferably, the arithmetic mean d0 of the strand spacing d in each of the first 10 frames during the initial stage of bolt pre-tightening is taken close to the baseline. Considering the sub-pixel level error in visual inspection and the possible slight oxide scale undulations on the steel wire surface, the gap threshold is set to 3 times d0. When the strand spacing d exceeds 3 times d0, surface texture interference can be effectively eliminated, confirming that substantial structural loosening and gap opening have occurred. At the same time, the gray value of the intermediate pixel between adjacent strands is read frame by frame. If the gray value of the intermediate pixel between adjacent strands falls back to the gray level of the image background, it indicates that the gap between adjacent strands has penetrated into the background, and it is determined that the adjacent strands have opened a gap.

[0050] Specifically, a twist section that simultaneously meets both the conditions of the strand spacing d exceeding the gap threshold and the gray level of the middle pixel falling back to the background gray level is marked as a twist structure failure section. The twist structure failure section is stored in the form of a four-element record containing the start and end twist positions, the breakage timestamp, and the strand spacing d. Further, based on the breakage timestamp of the twist structure failure section, several frames of the same twist section before the breakage timestamp are retrieved from the synchronization sequence. From these, image frames with a strand spacing d not exceeding 1.2×d0 and a contact arc width curve maintaining a continuous non-zero shape along the twist direction are selected. This image frame is taken as the complete morphology, which reflects the stable adhesion of the outer steel wire along the groove surface when the twist structure is complete.

[0051] Preferably, several frames of images of the same twist length segment after the fracture timestamp are retrieved from the synchronization sequence. From these frames, images are selected where the strand spacing d exceeds 3 times d0 and the flattened outer strand end pixels extend beyond the boundary of the adjacent strands into the gap by at least 2 pixel units. This image frame is used as the fracture morphology, reflecting the state of increased loosening of the outer strands and embedding into the gap between adjacent strands. Two pixel units correspond to an actual extension distance of 0.1 mm. This value is determined based on the morphology statistics of 100 fracture samples from the previous period. When the embedding depth of the strand end reaches 0.1 mm, a significant stress concentration phenomenon can be observed. At this point, the embedding depth is approximately 7% to 12% of the standard strand diameter, sufficient to characterize the significant intrusion of loose strands into the gap. Below this threshold, the strand end only shows slight deformation and does not form substantial embedding.

[0052] In one embodiment, the complete morphology and the failure morphology of the same twist pitch section are indexed according to the twist pitch section number, the fracture timestamp, and the strand spacing d, and stored in pairs in the morphology record library. Each morphology record in the morphology record library contains a complete morphology image frame, a failure morphology image frame, the twist pitch section number, the fracture timestamp, and the corresponding strand spacing d, which can be called upon in the subsequent establishment of the slip critical load assessment basis.

[0053] S105. Obtain the critical slip load of steel strand under different bolt preload conditions, and establish an evaluation basis for the anti-slip bearing capacity of steel strand by combining the failure morphology and the intact morphology.

[0054] For each morphology record in the morphology record library, a tensile force along the axial direction of the steel strand is applied to the steel strand clamp under the corresponding bolt preload condition. The tensile force is gradually increased while the relative displacement of the steel strand relative to the clamping groove surface is monitored simultaneously. If the relative displacement jumps from a slow linear increase to a sudden rapid increase while the tensile force is continuously increased, the axial tensile force value corresponding to the jump is recorded as the slip critical load under the bolt preload condition. At the same time, the bolt preload value corresponding to this condition is recorded, resulting in a load record set with a one-to-one correspondence between the slip critical load and the bolt preload. Based on the load record set, each load record in the load record set is correlated with its bolt preload value and the complete morphology and failure morphology of the same lay length segment in the morphology record library. It is determined whether the same lay length segment under the bolt preload value is in the complete morphology range where the outer steel wire is stably attached or in the failure morphology range where the outer strand is loosely embedded, thus obtaining a reference set of changes in slip critical load with morphology range. Based on the aforementioned comparison set, the slip critical load in the complete morphology range is assigned to the slip critical load rising segment corresponding to the complete morphology range, and the slip critical load in the failure morphology range is assigned to the slip critical load falling segment corresponding to the failure morphology range. This results in a two-segment comparison relationship where the morphology range and the slip critical load range are mapped one-to-one, which serves as the basis for evaluating the slip resistance capacity of the steel strand.

[0055] In one embodiment, for each morphology record in the morphology record library, an axial tension loading device is constructed to conduct a simulated drop test on the steel strand clamp. The axial tension loading device consists of a servo hydraulic actuator, a tension sensor, and a laser displacement sensor. The servo hydraulic actuator stretches the free end of the steel strand along the axial direction of the steel strand. The tension sensor is connected in series between the actuator and the free end of the steel strand and outputs the axial tension value in real time. The laser displacement sensor points to a displacement observation point marked on the steel strand and outputs the relative displacement of the steel strand with respect to the clamping groove surface in real time.

[0056] Specifically, under the bolt pre-tightening condition corresponding to the morphological record, after the steel strand is clamped in place, the actuator is activated, and the actuator gradually increases the tension along the axis of the steel strand at a constant loading rate.

[0057] For example, the loading rate is taken as 0.1kN / s to 0.5kN / s, and the axial tensile force and relative displacement are recorded synchronously at a sampling frequency of 100Hz.

[0058] It should be noted that in the initial stage of tensile loading, the steel strand is in an elastic deformation state within the clamping groove. The relative displacement increases slowly and linearly with the increase of axial tension, and the slope of the displacement curve remains at a low level. When the axial tension exceeds the static friction limit that the clamping groove can provide, the steel strand begins to slide macroscopically relative to the clamping groove. The slope of the relative displacement increases suddenly and rapidly.

[0059] Preferably, the relative displacement sequence is differentially analyzed between two adjacent frames to obtain a displacement increment sequence. Each term in the displacement increment sequence is denoted as Δp. The ratio k of two adjacent displacement increments Δp is taken as k = Δp current term / Δp previous term. When the ratio k continuously exceeds a preset jump factor of 5, it is determined that the relative displacement has jumped from a slow linear increase to a sudden rapid increase. The axial tensile force value corresponding to the jump instant is recorded as the slip critical load under the bolt preload condition. At the same time, the bolt preload force value acting on the steel strand clamping bolt under this condition is read to form a load record. For each morphology record in the morphology record library, the above loading and recording process is repeated to cover multiple conditions from the loosening stage to the over-preload stage of the bolt preload, resulting in a load record set with a one-to-one correspondence between the slip critical load and the bolt preload force. Further, according to the load record set, each load record in the load record set is matched with the complete morphology and failure morphology of the same lay length segment in the morphology record library according to its bolt preload force value.

[0060] It is understood that the synchronization sequence is stored using an ordered array structure with timestamps as index keys and bolt preload values ​​as data values, with a timestamp accuracy of 0.01 seconds. The fracture timestamp of each morphology record in the morphology record library is located in the synchronization sequence using a binary search algorithm. If a perfectly matching timestamp exists, the corresponding preload value is directly obtained; otherwise, the preload values ​​from the two most recent times are used for linear interpolation to calculate the preload at that moment. The interpolation formula is as follows: F represents the target preload, t represents the fracture timestamp, t1 and t2 represent the preceding and following time points, and F1 and F2 represent the corresponding preload values. After obtaining the preload at the fracture time through reverse lookup, this preload is used as the dividing point. The interval below this value is determined to be the low preload interval covered by the intact morphology, and the interval above this value is determined to be the high preload interval covered by the failure morphology. The bolt preload value in the load record is compared with the two intervals in the morphology record to determine whether the same lay length segment under the bolt preload value is in the intact morphology interval where the outer steel wire is stably attached, or in the failure morphology interval where the outer strands are loosely embedded. The determination result is stored together with the load record in the comparison set to obtain the comparison set of the slip critical load changing with the morphology interval.

[0061] In one possible implementation, based on the reference set, each load record is traversed sequentially according to the bolt preload value from smallest to largest. In the load records within the complete morphology range, the slip critical load shows a monotonically increasing trend with the increase of the bolt preload value, and these records are classified into the slip critical load rising segment corresponding to the complete morphology range. In the load records within the failure morphology range, the slip critical load shows a monotonically decreasing trend with the further increase of the bolt preload value, and these records are classified into the slip critical load falling segment corresponding to the failure morphology range.

[0062] In one embodiment, the slip critical load rising segment and slip critical load falling segment are stored in a two-segment lookup table in the form of key-value pairs of bolt preload value range and slip critical load value sequence. Each record in the two-segment lookup table contains a morphological range identifier, the start and end values ​​of the bolt preload value range, and the corresponding slip critical load value sequence. The morphological range and the slip critical load range are thus mapped one-to-one, and the two-segment lookup table serves as the basis for evaluating the slip resistance capacity of the steel strand.

[0063] The slippage critical load under different bolt preload conditions was collected. The continuous bonding length when the outer steel wire is stably bonded and the gap width between the discontinuous patches when the outer strand is loosely embedded were identified. The control range in which the slippage critical load decreases when the bonding length narrows and the gap width increases was obtained, which served as the basis for evaluating the slippage resistance capacity of the steel strand.

[0064] For each complete morphology image frame in the morphology record library, the tightly bonded pixel strips between the outer steel wire edge and the groove edge are scanned point by point along the wire twist direction. The twist direction pixel lengths of the continuous and uninterrupted bonded pixel strips are accumulated to obtain the continuous bonding length of the twist pitch segment corresponding to the complete morphology frame. For the damaged morphology image frame in the same morphology record, the discontinuous patches that have opened between adjacent strands are read along the twist direction. The pixel interval between the start and end points of two adjacent discontinuous patches in the direction perpendicular to the twist direction is measured to obtain the gap width between the discontinuous patches in the twist pitch segment corresponding to the damaged morphology frame. Based on the continuous bonding length and the gap width between the discontinuous patches, the continuous bonding length and the gap width between the discontinuous patches under different bolt preload conditions in the same twist pitch section are arranged in ascending order of bolt preload value to obtain a narrowing sequence in which the continuous bonding length narrows under each working condition as the bolt preload increases, and an expanding sequence in which the gap width between the discontinuous patches expands under each working condition as the bolt preload increases. The slip critical loads in the load record set are marked on the narrowing sequence and the widening sequence according to the corresponding bolt preload values. The bolt preload section covered by the continuous bonding length entering a downward trend in the narrowing sequence and the gap width between the discontinuous patches in the widening sequence entering a synchronous upward trend is identified. Within the bolt preload section, the slip critical load changes from rising to falling. The bolt preload section is bound to the value sequence of the falling slip critical load within it to obtain the comparison interval of narrowing bonding length, increasing gap width and decreasing slip critical load, which serves as the evaluation basis for the slip resistance capacity of the steel strand.

[0065] In one embodiment, the contact shape between the outer steel wire of the steel strand and the clamping groove surface during the bolt pre-tightening process changes with the value of the bolt pre-tightening force under different working conditions. To quantify this change, each morphology record is retrieved from the morphology record library, and pixel-level measurements are performed along the twist direction on its complete morphology image frame and damaged morphology image frame, respectively.

[0066] Specifically, for each complete morphology image frame in a morphology record, the lower edge of the outer steel wire and the upper edge of the groove surface are compared point by point along the twist direction. The vertical coordinate of the lower edge of the outer steel wire is y1, and the vertical coordinate of the upper edge of the groove surface is y2. The difference d = y1 - y2 is calculated point by point. The judgment threshold uses the single pixel equivalent k obtained from the calibration target in the image acquisition stage. k is the conversion factor between the pixel output by the calibration process and the actual size. In this case, k is taken as 0.05 mm / pixel. When the absolute value of d does not exceed 1 pixel, that is, the corresponding actual gap does not exceed 1 times k, the pixel position in the twist direction is considered to be in a tight fit without gap. The threshold of 1 pixel is taken because the lower limit of the fit tolerance between the outer steel wire and the groove surface is about 0.05 mm. Gaps smaller than this amount are considered to be fit in the assembly process. When d is greater than 1 pixel and does not exceed 2 pixels, it enters the embedding judgment branch described later, thus forming a three-level progressive judgment of fit, embedding, and separation. The twisted pixels that are continuously in close contact without gaps are strung together to form a bonding pixel strip. The bonding pixel strip is scanned point by point along the twist direction of the steel wire. The twisted pixel length of the bonding pixel strip that is continuous without interruption is multiplied by k and accumulated to obtain the continuous bonding length of the twist pitch segment corresponding to the complete morphology frame.

[0067] It should be noted that the continuous bonding length reflects the extension dimension along the twist direction of the outer steel wire to maintain effective frictional bonding with the groove surface. The larger the continuous bonding length, the more fully the outer steel wire is bonded along the twist direction under the bolt preload condition. Furthermore, for the destruction morphology image frame in the same morphology record, the discontinuous patches that have opened up between adjacent strands are read along the twist direction. The discontinuous patches are bonding blocks that appear at intervals along the twist direction. The interrupted area located between two adjacent discontinuous patches along the twist direction is taken as the gap between patches. The difference between the starting pixel coordinates and the ending pixel coordinates of the gap between patches along the twist direction is measured to obtain the width of the gap between discontinuous patches in the corresponding twist pitch segment of the destruction morphology frame. The larger the gap width, the more significant the gap that has opened up along the twist direction between adjacent strands.

[0068] In one possible implementation, the continuous bonding length and the gap width between discontinuous patches under different bolt preload conditions for the same pitch segment are arranged in ascending order according to their corresponding bolt preload values, resulting in two data sequences arranged according to bolt preload values. The continuous bonding length maintains a higher value when the bolt preload value is low, and decreases with each working condition as the bolt preload value continues to increase, resulting in a narrowing sequence. The gap width between discontinuous patches is close to zero when the bolt preload value is low, and increases with each working condition as the bolt preload value continues to increase, resulting in an expanding sequence.

[0069] Preferably, the slip critical loads in the load record set are marked on the narrowing sequence and the expanding sequence according to the corresponding bolt preload values. That is, the slip critical load value corresponding to each bolt preload scale in the narrowing sequence is attached, and the same slip critical load value is attached to the same bolt preload scale in the expanding sequence, forming a morphology and load tripartite table arranged along the bolt preload values. The morphology and load tripartite table uses the bolt preload value as the primary key, and each primary key simultaneously includes three values: continuous bonding length, gap width between discontinuous patches, and slip critical load.

[0070] Specifically, the morphology and load triplicate table is traversed along the direction of the bolt preload value. For each working condition, the percentage decrease in the continuous bonding length compared to the previous working condition and the percentage increase in the gap width between discontinuous patches compared to the previous working condition are calculated. The percentage decrease is recorded as α and the percentage increase is recorded as β. When α is greater than 10% and β is greater than 10%, the working condition is determined to be a trend reversal working condition. The bolt preload value segments covered by several adjacent trend reversal working conditions are merged and recorded as a common coverage segment.

[0071] In one embodiment, when the slippage critical load changes from rising to falling within the common coverage section, the bolt preload section and the slippage critical load falling numerical sequence within it are bound in the form of key-value pairs and stored in a reference interval table. Each record in the reference interval table contains the start and end values ​​of the bolt preload numerical interval, the falling numerical sequence of the continuous bonding length, the rising numerical sequence of the gap width between discontinuous patches, and the falling numerical sequence of the slippage critical load. This results in a reference interval that compares the narrowing of the bonding length, the increase of the gap width, and the decrease of the slippage critical load, serving as the evaluation basis for the slippage resistance capacity of the steel strand.

[0072] S106. Based on the evaluation criteria for the anti-slip bearing capacity of steel strands, the pressure distribution on the groove surface is accumulated segment by segment along the groove length to obtain the clamping resultant force between the groove surface and the outer steel wire and its concentration along the groove length. The clamping resultant force and concentration are assigned to the corresponding failure morphology or integrity morphology in the evaluation criteria to evaluate and obtain the slip critical load of the steel strand.

[0073] Based on the pressure distribution data of the groove surface under the current evaluation conditions, the groove surface is divided into several equal-length sub-segments along the groove length. Real-time pressure values ​​at all collection points within each sub-segment are read at the current timestamp. Each pressure value is multiplied by the product of the collection point spacing and the groove width, and then summed point by point to obtain the segmented clamping force of that sub-segment. The segmented clamping forces of all sub-segments are accumulated along the groove length to obtain the clamping resultant force between the groove surface and the outer steel wire. Based on the distribution of the segmented clamping forces along the groove length, the force value of the sub-segment with the highest segmented clamping force is divided by the arithmetic mean of the segmented clamping forces of all sub-segments to obtain the concentration of the clamping resultant force along the groove length. The larger the concentration value, the more the clamping resultant force is concentrated in a few sub-segments. Based on the clamping force and the concentration, the bolt preload value of the current evaluation condition is located in the corresponding reference interval table of the evaluation criteria for the anti-slip bearing capacity of the steel strand to the corresponding bolt preload value interval. If the morphology interval of the bolt preload value interval is marked as a complete morphology, the clamping force and the concentration are used as query keys to match the slip critical load value under the same key value in the slip critical load rising segment corresponding to the complete morphology. If the morphology interval is marked as a failure morphology, the slip critical load value under the same key value is matched with the clamping force and the concentration in the slip critical load falling segment corresponding to the failure morphology to obtain the evaluated steel strand slip critical load.

[0074] In one embodiment, during the clamping operation of the steel strand, the pressure on the groove surface is not uniformly distributed along the groove length. In order to convert this distribution data into a quantitative index that can be used for load assessment, the groove surface is divided into several equal-length sub-segments along the groove length according to a preset segment length based on the groove surface pressure distribution data under the current assessment condition.

[0075] For example, the segment length is 5mm. If the wire clamp groove is 80mm long, it is divided into 16 equal-length sub-segments.

[0076] Specifically, the real-time pressure values ​​of all collection points at the current timestamp within each sub-segment are read. Each pressure value is multiplied by the product of the collection point spacing and the groove width, and then summed point by point. That is, the segmented force of a micro-element is obtained by multiplying the pressure value by the area of ​​the groove surface covered by the collection point. The segmented forces of all micro-elements within the sub-segment are then summed to obtain the segmented clamping force Fi of the sub-segment, where i represents the sub-segment number and Fi represents the segmented clamping force of the i-th sub-segment. The segmented clamping forces Fi of all sub-segments are accumulated along the groove length to obtain the resultant clamping force F = ∑Fi between the groove surface and the outer steel wire. F reflects the total clamping force applied by the groove surface to the steel strand under the current evaluation condition.

[0077] It should be noted that the clamping force F only reflects the overall clamping strength and does not reflect whether the pressure is evenly distributed along the length of the groove. Under the same clamping force F, if the pressure is concentrated in a few segments, the steel wire in those segments will be subjected to much higher compression than in other areas, which can easily trigger the plastic flattening of the outer steel wire; if the pressure is evenly distributed along the length of the groove, the outer steel wire will be subjected to gentler force.

[0078] In one possible implementation, based on the distribution of the segmented clamping forces Fi along the groove length, the force value of the segment with the highest segmented clamping force Fi is denoted as Fmax, and the arithmetic mean of the segmented clamping forces of all segments is denoted as Favg. The quotient of the two, C = Fmax / Favg, is used to obtain the concentration C of the clamping resultant force along the groove length. The larger the concentration C value, the more the clamping resultant force is concentrated in a few segments; when the concentration C is close to 1, it indicates that the segmented clamping forces of each segment are equivalent, and the pressure is evenly distributed along the groove length. Further, based on the clamping resultant force F and the concentration C, the bolt preload value acting on the fastening bolt under the current evaluation condition is substituted into the corresponding reference interval table of the evaluation basis. The start and end values ​​of the bolt preload value interval of each record in the reference interval table are traversed to find the bolt preload value interval into which the current bolt preload value falls, and the morphological interval identifier of the record is read.

[0079] Preferably, if the morphology interval is identified as a complete morphology, the clamping resultant force F and the concentration C are used as query keys to enter the slip critical load rising segment corresponding to the complete morphology. The clamping resultant force F is normalized to a dimensionless quantity F′ according to its maximum value, and together with the concentration C, which is also a dimensionless quantity, they form a two-dimensional matching vector.

[0080] For example, the Euclidean distance between F′ and C is taken as the matching degree measure. In the rising segment of the slip critical load, the historical working condition record with the smallest Euclidean distance to the query key is retrieved, and the slip critical load value in the historical working condition record is read as the slip critical load obtained under the current working condition.

[0081] In one embodiment, if the morphology interval is identified as a failure morphology, the clamping resultant force F and the concentration C are also normalized in the above manner to form a two-dimensional matching vector, and enter the slip critical load falling segment corresponding to the failure morphology. In the slip critical load falling segment, historical working condition records are retrieved according to the same Euclidean distance minimum principle, and the slip critical load value in the retrieved historical working condition records is output as the slip critical load obtained under the current working condition.

[0082] It is understandable that the concentration C value is significantly larger when the morphology is in the damaged range, and the slip critical load obtained from the evaluation under the same clamping force F falls back to the level below the end of the rising segment, reflecting that the outer steel wire has entered the flattened embedding state.

[0083] In one embodiment, the entire above evaluation process is treated as an evaluation call, with the groove surface pressure distribution data and bolt preload value under the current evaluation condition as the processing objects, and the evaluation results of the steel strand slip critical load under the current condition are output.

[0084] S107. Analyze the inflection point of the slip critical load from rising to falling during the continuous increase of bolt preload. Determine the bolt preload force corresponding to the inflection point and recorded synchronously during the continuous application of bolt preload as the safe range for the anti-fall steel strand clamp fastening operation. Generate a stop command and send it to the automatic fastening equipment to execute the stop fastening action.

[0085] Based on the slip critical load obtained during the continuous increase of bolt preload, the slip critical loads under various working conditions are connected in series according to the synchronously recorded bolt preload values ​​from smallest to largest, resulting in an evaluation curve showing the change of slip critical load with bolt preload. The difference between the slip critical loads of adjacent points in the evaluation curve is calculated segment by segment to obtain a slope sequence of the evaluation curve along the bolt preload direction. Based on the slope sequence, the bolt preload is traversed to identify the position in the slope sequence where the value changes from positive to negative and remains negative for at least a preset number of frames. The bolt preload value corresponding to this position is determined as the inflection point where the slip critical load changes from increasing to decreasing. The bolt preload value at the inflection point is taken as the upper limit of the safe range for the anti-fall steel strand clamp fastening operation. The bolt preload range where the bolt preload value is less than the upper limit value is determined as the safe range for the fastening operation. According to the safety range, while the automatic fastening equipment continues to tighten, the current bolt preload value applied by the automatic fastening equipment is compared with the upper limit value in real time. If the current bolt preload value reaches the upper limit value, a stop command is generated and sent to the automatic fastening equipment. After receiving the stop command, the automatic fastening equipment cuts off the tightening torque output and executes the stop tightening action.

[0086] In one implementation, during the continuous increase of bolt preload, each time the steel strand clamp completes a preload increase, it outputs a sliding critical load estimate through the aforementioned evaluation process. The estimates are then connected in series according to the bolt preload values ​​recorded synchronously from smallest to largest to obtain an evaluation curve of the sliding critical load changing with the bolt preload. The horizontal axis of the evaluation curve is the bolt preload value, and the vertical axis is the sliding critical load estimate.

[0087] Specifically, the difference between the estimated sliding critical loads of two adjacent points in the evaluation curve is calculated segment by segment. That is, the difference between the estimated sliding critical loads of frame j and frame j-1 is calculated and then divided by the difference between the corresponding bolt preload values ​​of the two frames to obtain the slope kj of that segment. j represents the frame number in the increasing bolt preload sequence, and kj represents the slope of the evaluation curve at frame j along the bolt preload direction. The slopes kj of all frames are arranged into a slope sequence along the bolt preload direction according to the frame number.

[0088] It should be noted that the evaluation curve shows a monotonically rising trend in the initial stage of bolt preload, with all slopes kj remaining positive. As the bolt preload continues to increase and enters the over-preload stage, the evaluation curve declines, and the slope kj turns negative. The position where the slope sign changes is the physical inflection point where the slip critical load changes from rising to falling.

[0089] Preferably, the slope sequence is traversed from the side with smaller bolt preload to the side with larger preload to identify candidate inflection point positions where the slope sequence changes from positive to negative. To avoid misjudgment caused by noise in individual frames, a preset number of consecutive frames is set to 3 frames. That is, after the slope kj at the candidate inflection point turns negative, the slope kj in the subsequent 3 consecutive frames must remain negative in order to confirm the candidate inflection point as the inflection point where the sliding critical load changes from rising to falling.

[0090] In one possible implementation, the bolt preload value of the frame corresponding to the confirmed inflection point is read. This bolt preload value is used as the upper limit of the safe range for the anti-fall steel strand clamp tightening operation. The bolt preload range from zero to this upper limit value is defined as the safe range for the tightening operation. The lower limit of the safe range is zero, indicating that the bolt is in a loose state. Further, the upper limit value is written in digital form into the real-time comparison module of the automatic tightening device. The automatic tightening device includes a servo tightening gun, a torque and preload conversion module mounted on the tightening gun, and the real-time comparison module.

[0091] For example, the torque and preload conversion module converts the real-time output torque T of the tightening gun into the bolt preload value P currently applied to the fastening bolt through a pre-calibrated bolt torque coefficient K. The conversion relationship is P=T / K, where the bolt torque coefficient K reflects the friction characteristics of the thread pair. The value of K is pre-calibrated by the thread specification and lubrication status of the fastening bolt, and the conversion frequency is 100Hz.

[0092] In one embodiment, while the automatic fastening device continues to tighten, the real-time comparison module compares the current bolt preload value P output by the torque and preload conversion module with the stored upper limit value frame by frame at a comparison frequency of 100Hz. If the current bolt preload value P reaches the upper limit value, a stop command is triggered.

[0093] Specifically, the stop command is sent to the automatic tightening device in the form of a digital signal. After receiving the stop command, the automatic tightening device immediately cuts off the tightening torque output, the servo tightening gun stops rotating, and then the tightening gun retracts in the opposite direction by a preset small angle to release the residual torque on the bolt, thus performing the stop tightening action.

[0094] It is understood that when the aforementioned stop tightening action occurs, the bolt preload is exactly one frame before the inflection point where the slip critical load changes from rising to falling, and the bolt preload of the steel strand clamp at this time falls within the aforementioned safe range.

[0095] In one embodiment, the above-mentioned real-time comparison and stop command issuance process is repeatedly invoked as a closed-loop control in each tightening operation of the anti-fall steel strand clamp, and the tightening operation of the anti-fall steel strand clamp is completed once within the safety range.

[0096] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for evaluating the load-bearing capacity of steel strand clamps used for tower fall arrestors, characterized in that, The method includes: Obtain images of the pressure distribution on the groove surface and the surface of the outer steel wire during the continuous application of bolt pre-tightening of the anti-fall steel strand clamp; By processing the surface image of the outer steel wire through image recognition, the geometric contour changes of the outer steel wire are extracted to obtain the plastic flattening morphology; Based on the pressure distribution and plastic flattening morphology of the groove surface, the actual contact area between the groove surface and the outer steel wire is calculated. The part that no longer expands continuously along the groove length direction but narrows is marked as the effective contact area. When the effective contact area is broken into discontinuous patches along the twist direction of the wire and the gaps between adjacent strands are opened, the damage morphology when the twist structure is damaged and the complete morphology when the twist structure is intact are extracted from the surface image of the outer wire of the same twist pitch section. The critical slip load of steel strand under different bolt preload conditions was obtained, and the evaluation basis for the anti-slip bearing capacity of steel strand was established by combining the failure morphology and the intact morphology. Based on the evaluation criteria, the pressure distribution on the groove surface is accumulated segmentally along the groove length to obtain the clamping resultant force and its concentration along the groove length. It is then classified into the corresponding failure morphology or integrity morphology to evaluate and obtain the slip critical load of the steel strand. The inflection point of the slip critical load changing from rising to falling during the continuous increase of bolt preload is analyzed. The bolt preload force recorded synchronously at the inflection point is determined as the safe range of the fastening operation. A stop command is generated and sent to the automatic fastening equipment to execute the stop fastening action.

2. The method according to claim 1, characterized in that, The acquisition of the groove surface pressure distribution and outer steel wire surface image of the anti-fall steel strand clamp during the continuous application of bolt pre-tightening includes: Pressure sensor collection points are arranged along the length of the groove on the inner wall of the clamping groove of the anti-fall steel strand clamp. Real-time pressure values ​​are obtained from each collection point when the bolts are continuously tightened. The pressure distribution data of the groove surface is obtained by summarizing the collection point locations. An industrial camera is installed at the lateral opening of the steel strand clamp, and a ring light source is used to continuously photograph the surface of the outer steel wire to obtain images of the outer steel wire surface. The time nodes of the bolt pre-tightening process are aligned with the pressure distribution data of the groove surface to obtain a synchronization sequence.

3. The method according to claim 1, characterized in that, The process of processing the surface image of the outer steel wire through image recognition to extract the geometric contour changes of the outer steel wire and obtain the plastic flattening morphology includes: The outer steel wire surface image is retrieved from the synchronization sequence, and Gaussian filtering is applied to each frame to denoise it, resulting in a denoised steel wire surface image. The Canny edge detection method is used to extract the light and dark transition lines from the denoised steel wire surface image, resulting in a set of edge pixels arranged along the direction of the steel wire. The edge pixel set is grouped and assigned to each single-strand outer layer steel wire segment according to the wire twist direction. For each segment, least squares ellipse fitting is used to obtain the cross-sectional ellipse parameters containing the major axis length and the minor axis length as the geometric profile. The cross-sectional flatness is obtained by calculating the ratio of the major axis length to the minor axis length of the same segment frame by frame, and the average value of the cross-sectional flatness of each segment in the initial stage of bolt pre-tightening is used as the circular reference flatness. Arrange the cross-sectional flattening of each frame along the timestamp, extract the segments that exceed the preset threshold of the circular reference flattening and mark them as plastic flattening segments, and summarize the cross-sectional elliptical parameters and positions of each plastic flattening segment to obtain the plastic flattening morphology.

4. The method according to claim 1, characterized in that, The actual contact area between the groove surface and the outer steel wire is calculated based on the pressure distribution and plastic flattening morphology of the groove surface. The portion that no longer extends continuously along the groove length and narrows is marked as the effective contact area, including: Read the real-time pressure values ​​of each collection point from the pressure distribution data on the tank surface, and mark the collection points that exceed the contact pressure threshold as pressure contact points; At the same timestamp, the pressure contact point and the plastic flattening section are aligned and matched along the groove length direction. The short axis length of the cross section of the flattened outer steel wire in the matching section is taken as the contact arc width. Multiply it by the extension length of the section to obtain the contact surface element and accumulate to obtain the actual contact area. When the length of the contact surface element in adjacent segments stops increasing and the contact arc width narrows from the outside to the inside, the adjacent segments are merged and marked as the effective contact area.

5. The method according to claim 1, characterized in that, When the wire breaks into discontinuous patches along the twist direction in the effective contact area and adjacent strands open with gaps, the method of extracting the damage morphology of the twisted structure when it is damaged and the complete morphology of the twisted structure when it is intact from the surface image of the outer wire in the same twist pitch section includes: The contact arc width within the effective contact area is scanned frame by frame along the twist direction. When the contact arc width drops from a non-zero value to zero and reappears as a non-zero value at a more distant twist direction position, it is determined that the fracture is into a discontinuous patch and the patch discontinuity point is recorded. If the edge lines of adjacent outer strands on both sides of the same twist pitch section are retrieved from the synchronous sequence, and the lateral pixel distance between the edge lines of adjacent strands exceeds the preset gap threshold and the gray level of the middle pixel falls back to the background gray level, it is determined that the adjacent strands have opened a gap, and the twist pitch section is marked as the twist structure destruction section.

6. The method according to claim 5, characterized in that, The section where the twisting structure is damaged includes: Based on the fracture timestamp of the broken section of the twisted structure, image frames of adjacent strands closely attached and outer steel wires continuously attached along the groove surface in the same twisted section before the fracture timestamp are retrieved from the synchronous sequence as complete morphology. Image frames acquired after the fracture timestamp, showing increased loosening of adjacent strands and embedding of outer strands into the gaps between adjacent strands, are retrieved as the damage morphology. This yields a morphology record where the complete morphology and the damage morphology of the same twist pitch segment are stored in pairs.

7. The method according to claim 1, characterized in that, The method involves obtaining the critical slip load of steel strands under different bolt preload conditions, and establishing an evaluation basis for the slip resistance capacity of steel strands by combining the failure morphology and the intact morphology. This includes: For each morphology record, a tensile force is applied along the axial direction of the steel strand under the corresponding bolt preload condition and gradually increased. The relative displacement of the steel strand relative to the clamping groove surface is monitored simultaneously. The axial tensile force value when the relative displacement jumps from a slow linear increase to a rapid increase is recorded as the slip critical load of the working condition. The load record set is obtained by corresponding the bolt preload value one by one. According to the bolt preload values ​​in the load record set, corresponding to the complete or damaged morphology of the same lay length section, the slip critical load in the complete morphology section is classified into the rising section, and the slip critical load in the damaged morphology section is classified into the falling section. The resulting two-segment comparison relationship serves as the basis for evaluating the slip resistance capacity of the steel strand.

8. The method according to claim 7, characterized in that, The evaluation criteria for the anti-slip bearing capacity of the steel strand include: For each complete morphology frame, scan along the twist direction of the steel wire to find the seamless bonding pixel strip between the edge of the outer steel wire and the edge of the groove surface. Accumulate the length of the continuous and uninterrupted twist direction pixels to obtain the continuous bonding length. Read the discontinuous patches that have opened between adjacent strands in the frame of the damaged morphology, and measure the pixel spacing between the start and end points of two adjacent patches in the vertical twist direction to obtain the gap width between the discontinuous patches. Arrange the continuous bonding length and the gap width between discontinuous patches according to the bolt preload values ​​from small to large. Identify the bolt preload section that is covered by the continuous bonding length entering a downward trend and the gap width entering a synchronous upward trend. Obtain the comparison interval where the bonding length narrows, the gap width increases and the slip critical load decreases.

9. The method according to claim 1, characterized in that, The evaluation criteria are based on the segmented accumulation of the pressure distribution along the length of the groove to obtain the clamping resultant force and its concentration along the groove length. This is then categorized into the corresponding failure or integrity morphology to evaluate the slip critical load of the steel strand, including: The pressure distribution data of the tank surface under the current evaluation condition is divided into equal-length sub-segments along the tank length direction according to the preset segment length. The real-time pressure value of the current timestamp in each sub-segment is multiplied by the product of the distance between the collection points and the width of the tank, and then added point by point to obtain the segment clamping force. The segment clamping forces of each sub-segment are accumulated to obtain the clamping resultant force. The concentration is obtained by dividing the force value of the segment with the highest segment clamping force by the arithmetic mean of the clamping forces of all segments. The bolt preload value range is located in the corresponding reference range table of the evaluation basis based on the bolt preload value under the current evaluation condition. When the morphology range is marked as complete morphology, the sliding critical load is matched with the same key value in the rising segment by the clamping resultant force and concentration. When the morphology range is marked as damaged morphology, the sliding critical load is matched with the same key value in the falling segment.

10. The method according to claim 1, characterized in that, The analysis identifies the inflection point where the slip critical load changes from increasing to decreasing during the continuous increase of bolt preload. The bolt preload force recorded synchronously at the inflection point is determined as the safe range for the tightening operation. A stop command is generated and sent to the automatic tightening equipment to execute the stop tightening action, including: The evaluation curve is obtained by connecting the slip critical load of each working condition in series from small to large according to the synchronously recorded bolt preload values, and the slope sequence is obtained by calculating the difference between adjacent points segment by segment. The bolt preload value corresponding to the position in the slope sequence where the positive value turns into a negative value and the negative value is continuously maintained for no less than a preset number of frames is determined as the inflection point, and the bolt preload value at the inflection point is used as the upper limit of the safe range. When the bolt preload applied by the automatic fastening equipment reaches the upper limit value, a stop command is generated and sent to the automatic fastening equipment, which then cuts off the tightening torque output to stop the fastening action.

Images