An automatic detection and reliable interpretation method and system for flexible welded joints of reinforcing bars
By collecting and analyzing three-dimensional point cloud data of steel bar welded joints using an industrial inspection processor, inverting weld area information, calculating strength correction coefficients and adaptive clamping forces, and adjusting the loading strain rate, high-precision automatic inspection and reliable interpretation of steel bar welded joints are achieved. This solves the strength measurement deviation and clamping problems existing in the prior art, and improves the reliability and traceability of the inspection results.
Patent Information
- Application Number
- CN202611205159.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-10
- Publication Date
- 2026-11-17
AI Technical Summary
Existing automatic testing systems fail to effectively consider the coupling effect of non-uniform residual stress field and loading rate introduced during the welding process in mechanical property tests of welded steel bar joints. This results in large deviations in strength measurement, problems such as clamping slippage and weld damage, large dispersion in elongation measurement after fracture, and a lack of a reliable quantitative mechanism for the entire process results.
A flexible automatic inspection method executed by an industrial inspection processor is used to collect three-dimensional point cloud data and material grade information, invert the peak residual tensile stress and heat-affected zone width of the weld area, calculate the strength correction coefficient and adaptive clamping force, adjust the loading strain rate, simultaneously collect test data and calculate the confidence level, and generate an inspection report.
It eliminates the bias of the strength measurement system, solves the problems of clamping slippage and weld damage caused by the attenuation of the friction coefficient, reduces the measurement dispersion and the misjudgment rate of abnormal results, and realizes the reliable quantification and traceability of the test results.
Smart Images

Figure CN123111808A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of building material testing and intelligent testing equipment technology, and discloses a method and system for automatic detection and reliable interpretation of the flexibility of steel bar welded joints. Background Technology
[0002] Mechanical property testing of welded steel rebar joints is a core component of building construction quality inspection, and the test results directly determine the structural safety of the project. Existing automated testing systems, when conducting tensile tests, all use a standardized strain rate to control the loading process, failing to consider the coupling effect between the non-uniform residual stress field introduced into the joint area during welding and the loading rate. Welding residual stress, as an internal stress field, superimposes with the stress field generated by the external load, altering the critical stress for dislocation activation. Furthermore, this effect exhibits non-linear characteristics with strain rate changes. For high-carbon equivalent steel rebar, this coupling effect can lead to strength measurement deviations exceeding 5%, directly impacting the accuracy of the test results. Current clamping force calculations use a linear friction model, assuming a constant friction coefficient, without considering the non-linear decay of the friction coefficient caused by plastic deformation of the surface oxide film and rib micro-protrusions under high pressure, easily leading to clamping slippage or weld damage. Post-fracture elongation measurements do not consider the plastic gradient and temperature effects of the heat-affected zone, resulting in large measurement dispersion and a lack of a reliable quantitative mechanism for the entire process, leading to a high misjudgment rate for abnormal samples. Therefore, an automated detection and interpretation technology for welded steel rebar joints that can solve the above problems is urgently needed. Summary of the Invention
[0003] To address the problems existing in the prior art, this invention proposes an automatic detection and reliable interpretation method for the flexibility of welded steel bar joints, executed by an industrial detection processor, comprising: S1: Collect three-dimensional point cloud data of the surface of the steel bar welded joint, weld location information, heat-affected zone location and material grade information; S2: Based on the three-dimensional point cloud data and material grade information, invert the peak residual tensile stress, heat-affected zone width, and initial friction coefficient of the jaw interface in the weld area; calculate the strength correction coefficient coupled with the residual stress and strain rate; S3: Iteratively calculate the target value of the adaptive clamping force based on the nonlinear evolution of interface friction according to the strength correction coefficient; S4: Center the specimen, control the variable pressure jaws to clamp the specimen according to the adaptive clamping force target value, and adjust the tensile test loading strain rate according to the strength correction coefficient. S5: Simultaneously collect force data, displacement data, and ambient temperature data during the test process; S6: Collect three-dimensional point cloud data of the fracture surface after the specimen fractures, and calculate the true value of the elongation after fracture after correction of the plastic gradient and temperature coupling in the heat-affected zone. S7: Extract the centering error factor, clamping slip factor, curve integrity factor, image quality factor, and temperature stability factor, and calculate the confidence level of the detection result fused from the multi-source uncertainty. S8: Determine the validity of the test results based on the stated confidence level and generate a test report.
[0004] Preferably, the acquisition of three-dimensional point cloud data of the welded joint surface of the reinforcing bar, weld location information, heat-affected zone location and material grade information includes: calling a multi-camera array and structured light three-dimensional reconstruction unit to acquire surface texture images and three-dimensional point clouds of the reinforcing bar; identifying the weld boundary location, heat-affected zone color characteristics and reinforcing bar rib characteristics; reading the sample identification through the barcode scanning unit, and calling the material database to obtain the corresponding grade of base material yield strength, base material tensile strength and base material uniform elongation.
[0005] Further preferably, the adaptive clamping force target value calculated iteratively based on the strength correction coefficient to determine the nonlinear evolution of interface friction includes: real-time acquisition of the actual contact area between the clamp jaws and the reinforcing bar; The measured slip amount was collected during the iterative process. Using the initial clamping force as the starting point of the iteration, the target value of the clamping force is calculated in three consecutive iterations. The iteration results are subject to upper and lower limits, and the final adaptive clamping force target value is output. The amount of slippage of the sample relative to the jaws is monitored in real time during the clamping process. When the slippage exceeds the threshold, a re-clamping action is triggered.
[0006] Further preferably, after the specimen fractures, three-dimensional point cloud data of the fracture surface is collected, and the corrected post-fracture elongation value after the coupling of the plastic gradient and temperature in the heat-affected zone is calculated, including: The robot was used to move the two fractured sections of the sample to the post-fracture measurement station. The total gauge length after fracture, the axial splicing gap of the fracture surface, and the arithmetic mean deviation of the three-dimensional morphology of the fracture surface were measured by a three-dimensional vision unit. Obtain the average ambient temperature during the test; The uniform elongation and elongation temperature sensitivity coefficient of the heat-affected zone are obtained by calling the material database. By successively deducting the spurious elongation introduced by the fracture morphology, correcting the systematic deviation introduced by the plastic gradient in the heat-affected zone, and performing temperature compensation calculations, the true value of the corrected elongation after fracture is obtained.
[0007] Further preferably, the strength correction factor for calculating the coupling between residual stress and strain rate includes: Obtain the material carbon equivalent-related coupling coefficient, the nominal yield strength of the parent material, the standard-specified reference strain rate, and the material strain rate sensitivity index; By combining the peak residual tensile stress in the weld area with the actual controlled strain rate in the experiment, the amplification effect of residual stress on strain rate sensitivity is quantified, and a dimensionless strength correction coefficient is obtained.
[0008] A further preferred embodiment of the adaptive clamping force target value, calculated iteratively based on the strength correction coefficient to determine the nonlinear evolution of interfacial friction, includes: Obtain the nominal diameter of the reinforcing bar, the tensile strength of the base material, the attenuation coefficient of the interfacial friction pressure, and the historical slip feedback gain coefficient; The friction coefficient is expressed as an exponential decay function of the normal pressure. A hyperbolic tangent feedback term of the historical slip is introduced, and the target clamping force is obtained by iteratively solving the implicit equation.
[0009] Further preferred methods for calculating the corrected post-fracture elongation value after coupling the plastic gradient and temperature in the heat-affected zone include: Obtain the original gauge length, average width of the heat-affected zone on one side, and standard reference temperature before the test; Using the nominal diameter of the steel bar as the characteristic length of the fracture, the errors introduced by the fracture splicing error, the plasticity difference error of the heat-affected zone, and the temperature deviation are successively corrected to obtain the true elongation after fracture.
[0010] Further preferred methods for calculating the confidence level of the detection result based on multi-source uncertainty fusion include: Obtain the weight coefficients of each influencing factor and the penalty coefficients for abnormal events; Five confidence factors were fused using a geometric mean, and the total number of anomalous events triggered throughout the experiment was combined with an exponential penalty to determine the impact of anomalous events on the reliability of the results, resulting in a confidence level ranging from zero to one. When the confidence level is greater than or equal to the set threshold, a formal test report is generated; When the confidence level is less than the set threshold, a review task is generated and the sample is returned to the buffer station. Choose a route based on the type of anomaly, such as secondary positioning, re-clamping test, image verification, or manual confirmation. The sample identification, equipment calibration status, raw data acquisition, parameter correction process, and calculation results are bound into a trusted data chain through hash operations.
[0011] A flexible automatic detection and reliable interpretation system for welded steel bar joints, applied to any of the above-mentioned flexible automatic detection and reliable interpretation methods for welded steel bar joints, includes a multimodal visual perception module, a material parameter inversion module, a loading parameter generation module, a flexible execution and closed-loop control module, a multi-sensor data synchronous acquisition module, a post-fracture parameter correction module, and a reliable interpretation and report generation module. The multimodal visual perception module is electrically connected to the material parameter inversion module, the flexible execution and closed-loop control module, and the post-fracture parameter correction module. It is used to collect surface texture images of welded joints, three-dimensional point cloud data, weld location information, heat-affected zone width, and three-dimensional fracture morphology data. The material parameter inversion module is electrically connected to the multimodal visual perception module and the loading parameter generation module, and is used to invert the peak value of welding residual stress, plastic parameters of heat-affected zone and initial friction coefficient of interface based on visual features and material grade. The loading parameter generation module is electrically connected to the material parameter inversion module, the flexible execution and closed-loop control module, and the multi-sensor data synchronous acquisition module. It is used to calculate the strength correction coefficient of the coupling between residual stress and strain rate, and to iteratively calculate the target value of friction nonlinear adaptive clamping force. The flexible execution and closed-loop control module is electrically connected to the multimodal visual perception module, the loading parameter generation module, and the multi-sensor data synchronous acquisition module. It includes a dual-zone obstacle avoidance gripper unit, an adaptive V-shaped centering unit, a variable pressure closed-loop jaw unit, and a follow-up temperature compensation unit, which are used to perform sample picking and placing, centering adjustment, closed-loop clamping, and rate loading. The multi-sensor data synchronization acquisition module is electrically connected to the loading parameter generation module, the flexible execution and closed-loop control module, and the post-fault parameter correction module, and is used for hardware time synchronization acquisition of force, displacement, deformation, temperature and vibration time series data. The post-fracture parameter correction module is electrically connected to the multimodal visual perception module, the multi-sensor data synchronous acquisition module, and the reliable interpretation and report generation module. It is used to calculate the corrected post-fracture elongation value after the coupling of plastic gradient and temperature in the heat-affected zone. The credible interpretation and report generation module is electrically connected to the post-failure parameter correction module. It is used to calculate the credible confidence level of the detection result fused from multiple sources of uncertainty, determine the validity of the result, and generate a traceable detection report.
[0012] More preferably, the multimodal visual perception module includes a multi-camera array unit, a structured light 3D reconstruction unit, and a barcode scanning and recognition unit; The multi-camera array unit is electrically connected to the structured light 3D reconstruction unit and is used to acquire images of the steel reinforcement surface from different angles. The structured light 3D reconstruction unit is used to generate 3D point cloud data of the steel bar surface and fracture surface; The barcode scanning and identification unit is used to read the sample identification information; The flexible execution and closed-loop control module also includes a slip detection unit, which is electrically connected to the variable pressure closed-loop jaw unit and is used to monitor the amount of slippage of the sample relative to the jaw in real time. The multi-sensor data synchronization acquisition module includes a hardware time synchronization unit, which is electrically connected to all sensors and is used to unify the time reference of all acquired data. The trusted interpretation and report generation module includes an abnormal route generation unit, which is used to generate routes for secondary positioning, re-clamping test, image verification or manual confirmation when the trust confidence is insufficient. The trusted interpretation and report generation module also includes a hash chain generation unit, which is used to bind the sample identity, device status, raw data and calculation results into a traceable data chain.
[0013] Technical effects: This scheme establishes a strength correction model coupling welding residual stress and strain rate to quantify the nonlinear influence of residual stress on dislocation motion, thus eliminating the bias in strength measurement systems for specimens with different weld qualities. Combined with frictional nonlinear adaptive clamping force calculation, it simultaneously addresses the problems of clamping slippage and weld damage caused by friction coefficient attenuation under high pressure. Through heat-affected zone and temperature-coupled elongation at break correction and full-process confidence level calculation, it reduces measurement dispersion and the misjudgment rate of abnormal results. All test data are bound via a hash chain, ensuring traceability and recalculation of results, meeting the accreditation requirements of testing institutions. Attached Figure Description
[0014] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0015] Figure 1 This is a flowchart of the visual model testing method for the track fastener missing detection system of the present invention; Figure 2 This is a graph showing the variation of the strength correction coefficient with strain rate under different carbon equivalents and residual stress levels of the present invention. Figure 3 This is a comparison chart showing the effects of multi-level correction of elongation after fracture and temperature compensation in this invention. Detailed Implementation
[0016] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.
[0017] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.
[0018] Please see Figures 1-3 This technical solution describes an automated detection and reliable interpretation method for the flexibility of welded steel bar joints, executed by an industrial testing processor. The method covers the entire process of automated detection and reliable data interpretation, from sample input to report generation.
[0019] The specific implementation method for collecting three-dimensional point cloud data of the surface of the welded joint of the steel bar, weld location information, heat-affected zone location and material grade information is as follows.
[0020] A multi-camera array unit and a structured light 3D reconstruction unit work together to acquire surface texture images and 3D point cloud data of the reinforcing steel. The multi-camera array unit consists of four industrial cameras, evenly arranged at 90-degree angles along the circumference of the reinforcing steel. Each camera covers a field of view of more than 90 degrees on the steel surface, with at least a 15-degree overlap between the four cameras' fields of view. This arrangement avoids target occlusion caused by a single viewpoint. The structured light 3D reconstruction unit projects coded sinusoidal fringe structured light onto the steel surface. The fringe pattern is modulated and deformed on the curved surface of the steel, and the cameras simultaneously acquire images of the deformed fringe patterns. Based on the principle of triangulation, the calibrated camera-projector system converts the fringe phase offset into height information, generating 3D point cloud data of the steel surface. The resolution of the point cloud is determined by the projected fringe density and the camera pixels, with a minimum of 5 million points in a single reconstruction.
[0021] The acquired image and point cloud data were used to identify the weld boundary location, heat-affected zone (HAZ) color features, and rebar rib features. Weld boundary location identification was achieved by analyzing the curvature changes of the 3D point cloud along the axial direction. The curvature of the point cloud in the weld reinforcement region exhibits abrupt changes compared to the base metal region. By sliding a curvature gradient operator along the rebar axis, the location of the curvature abrupt change is extracted, which is the weld boundary. The color features of the HAZ were identified based on the difference in surface oxide color. The welding thermal cycle forms a dense oxide film on the surface of the HAZ metal, which appears as a specific blue or purple in the color image, distinct from the silvery-white base metal and the deep black weld. The image was converted from RGB to HSV space, and the hue threshold range of the H channel was set to segment the HAZ pixel region. Morphological closing operations were then performed to eliminate noise, resulting in a continuous HAZ boundary. Rebar rib features were identified by performing a circumferential unfolding operation on the point cloud to extract the periodic fluctuation components of the radial cross-sectional profile, thereby determining the rib spacing, direction, and height parameters.
[0022] The barcode scanning unit reads the QR code marking on the sample surface. This QR code is laser-marked and attached to one end of the rebar at a fixed distance from the end before welding. The decoded unique sample identification number is transmitted to the industrial testing processor. The processor retrieves material parameters pre-stored in its local material database, which stores the mechanical property parameters of different grades of rebar in a relational table format. Based on the grade information carried in the marking, the corresponding yield strength, tensile strength, and uniform elongation of the base material are matched and obtained.
[0023] The specific implementation method for inverting the peak residual tensile stress, heat-affected zone width, and initial friction coefficient of the jaw interface in the weld area based on 3D point cloud data and material grade information is as follows.
[0024] The inversion of peak residual tensile stress in the weld region is based on the thermo-elastic-plastic relationship between welding deformation and residual stress. The 3D point cloud data contains macroscopic deformation information of the weld and its surrounding area after cooling to room temperature. Residual stress release leads to deformation rebound in a specific direction, and there is a corresponding relationship between the longitudinal residual tensile stress in the weld region and the lateral shrinkage and angular deformation of the weld joint. By extracting multiple cross-sectional contour lines perpendicular to the weld direction from the 3D point cloud, the radial offset of each point on each contour line relative to the reference cylindrical surface is calculated, and the lateral shrinkage value and angular deformation angle are fitted. Using a pre-established finite element parametric inverse model, the lateral shrinkage value and angular deformation angle are used as inputs to inversely solve for the peak residual tensile stress in the weld region. This inverse model has been calibrated through numerical simulation and field measurements on a preset number of weld joints with different process parameters, and the inverse solution error is controlled within ±15%.
[0025] The width of the heat-affected zone (HAZ) is inverted by analyzing the boundary spacing of the color feature regions within the HAZ. On the mask image obtained from the previous color feature segmentation, an integral projection along the rebar axis is performed to obtain the axial distribution curve of the color changes. The half-width at half-maximum (WHM) of this curve is used as the estimated value of the width of one side of the HAZ.
[0026] The inversion of the initial friction coefficient at the jaw interface is based on the relationship between surface microstructure and tribological properties. The micro-roughness parameters of the 3D point cloud of the reinforcing bar surface, especially the arithmetic mean deviation and kurtosis coefficient, exhibit a statistical mapping relationship with the initial friction coefficient. The surface morphology of the base material far from the weld in the acquired 3D point cloud is extracted, and the surface arithmetic mean deviation in this region is calculated. and skewness coefficient Using a regression-calibrated mapping model, and Convert to initial friction coefficient of jaw interface The regression model was established by fitting the results of sliding friction tests on multiple sets of steel bars with different surface conditions and standard clamp specimens.
[0027] The specific implementation method for calculating the strength correction coefficient coupled with residual stress and strain rate is as follows.
[0028] Obtaining the carbon equivalent-related coupling coefficient of the material This coefficient is determined based on the carbon equivalent value calculated from the chemical composition of the reinforcing steel. The carbon equivalent is calculated using the formula recommended by the International Institute of Welding. ,in Represents the mass percentage of element X. Coupling coefficient. With carbon equivalent The regression relationship was calibrated experimentally. , Taking 0.15, which corresponds to a carbon equivalent of 0.43 for HRB400E steel reinforcement in this example, the calculation yields... It is 0.0645.
[0029] Obtain the nominal yield strength of the base material The value is retrieved from the materials database; in this example, 400 MPa is used.
[0030] Obtain the reference strain rate specified in the standard According to the test standard, the value is 0.00025 per second.
[0031] Obtaining the material strain rate sensitivity index This index was calibrated through tensile tests on the base material at different strain rates. The tests were conducted at four rate levels: 0.0001 s, 0.00025 s, 0.001 s, and 0.0025 s. The yield strength at each rate was recorded, and linear regression was performed on the data points. This refers to the slope of the flow stress-strain rate curve on a logarithmic coordinate system. In this example, the HRB400E steel reinforcement... The value is calibrated to 0.015.
[0032] Obtain the peak residual tensile stress in the weld area The value is obtained from the aforementioned inversion steps, and in this example it is 120 MPa.
[0033] Obtain the actual controlled strain rate of the experiment The initial strain rate is dynamically set by the test loading control system based on the strength correction coefficient. The calculation process is as follows: first, set the initial strain rate to... The final execution value is adjusted after strength correction calculation.
[0034] Substitute the above parameters into the coupling calculation formula:
[0035] In the formula, The strength correction factor is dimensionless and represents the yield strength correction factor caused by the coupling effect of welding residual stress and strain rate. Multiplying this value by the strength value at the standard strain rate yields the true strength value. The carbon equivalent of the material is a dimensionless coupling coefficient, and its value is positively correlated with the carbon equivalent of the steel reinforcement. It represents the weight of the influence of the material's weld hardening tendency on the sensitivity to strain rate. The peak residual tensile stress in the weld region is expressed in Pascals. It represents the maximum value of the residual tensile stress field introduced by the welding thermal cycle in the weld region of the joint. It is the internal stress source that drives stress concentration and affects the dislocation motion barrier. Let be the nominal yield strength of the base material, in Pascals; be the benchmark yield strength of the base material at standard strain rate; and be the ratio of residual stress to yield strength. Characterizes the relative level of residual stress. The actual controlled strain rate is measured in units of seconds, and represents the actual strain rate value during the plastic loading stage in the testing process. The standard specifies a reference strain rate, measured in units of seconds, which is the strain rate value for the plastic stage as specified in the testing standards of this field, and serves as the benchmark rate for strength evaluation. The strain rate sensitivity index is dimensionless and represents the intrinsic thermal activation parameter of dislocation motion in the material. It reflects the power-law relationship between flow stress and strain rate, i.e. .
[0036] The physical basis of this calculation formula is the dislocation thermal activation theory. When the applied strain rate deviates from the standard value, the thermal activation waiting time for dislocation motion changes, and the material flow stress changes accordingly according to a power law relationship. The residual tensile stress from welding, acting as an internal stress field, lowers the effective barrier for dislocation motion. As the strain rate increases, the effective stress required for dislocations to overcome the barrier further increases, and the superposition of residual stress and applied stress amplifies the increase in flow stress. This amplification effect is positively correlated with the residual stress level and the carbon equivalent. When... When the value in parentheses is 0, No corrections will be made. When the value in parentheses is positive, and the greater the residual stress, the better. The greater the deviation from 1, the better. All the above parameters have been clearly defined and their acquisition methods have been given; the intensity correction coefficient... It can be calculated in real time by the detection processor, without the need for additional hardware.
[0037] The specific implementation method of the adaptive clamping force target value based on the iterative calculation of the nonlinear evolution of interface friction using the strength correction coefficient is as follows.
[0038] Obtaining the nominal diameter of the reinforcing bar The material is retrieved from the material database based on the grade information. In this example, it is 25 mm, which is equivalent to 0.025 m.
[0039] Obtaining the tensile strength of the base material The tensile strength of HRB400E steel bars in this example is 540 MPa, retrieved from the materials database.
[0040] Obtain the interface friction pressure attenuation coefficient This coefficient is calibrated through a jaw-reinforcing bar friction test under different pressures. The calibration method is as follows: On a dedicated testing apparatus, normal compressive stresses ranging from 10 MPa to 100 MPa are applied to the jaws and the reinforcing bar sample. The static friction coefficient is measured under each pressure, and a semi-logarithmic curve of friction coefficient versus normal pressure is plotted. The absolute value of the curve's slope is the coefficient of friction. In this example, The calibration value is 3.2 times 10 to the power of negative 8 per pascal.
[0041] Obtain historical slip feedback gain coefficient This coefficient is determined by designing the response characteristics of the clamping system. Taking an empirical value of 0.3 can achieve a moderate increase in clamping force after the slippage exceeds the threshold, and the saturation characteristics of the hyperbolic tangent function can avoid overcompensation.
[0042] Obtain the initial static friction coefficient The value is obtained from the aforementioned inversion steps; in this example, it is 0.35.
[0043] Obtain the effective contact area between the jaws and the reinforcing bar. The contact area is provided in real time by the force-sensitive sensor and contact state monitoring subsystem of the variable pressure closed-loop jaw unit. The contact area is indirectly measured using impedance imaging: an electrode array is arranged on the back of the jaw teeth, the impedance between each electrode pair is measured, and the contact area is calculated in real time based on the mapping relationship between impedance and contact area. .
[0044] Obtain the measured slip during the iterative process The slip detection unit records and provides feedback during each test or adjustment process.
[0045] Get the allowed slip threshold Based on the test standard, the tolerance requirement for slippage is set, and in this example it is set to 0.1 mm, which is equivalent to 0.0001 m.
[0046] Substitute the above parameters into the implicit iterative calculation formula:
[0047] In the formula, Let be the target clamping force, in Newtons, and let be the target value of the normal clamping force required to be applied to the variable pressure jaws. is the nominal diameter of the steel bar, measured in meters, and is used to calculate the nominal cross-sectional area of the steel bar. is the tensile strength of the base material, in Pascals, representing the maximum tensile stress required for the steel bar to break, and is used to estimate the maximum axial tensile load. This is a dimensionless strength correction factor, derived from the previous calculation step, reflecting the correction of the actual strength of the reinforcing steel by the coupling of welding residual stress and strain rate. denoted as the initial static friction coefficient at the interface between the jaws and the reinforcing bar, dimensionless, representing the frictional characteristics of the interface under low pressure. The coefficient of friction attenuation is expressed in Pascals. It reflects the physical law that as the normal compressive stress increases, the microscopic protrusions at the interface are gradually flattened, the actual contact area growth rate slows down, and the macroscopic friction coefficient exhibits an exponential decay. The effective contact area between the jaws and the reinforcing bar, measured in square meters, is a key geometric parameter that determines the level of normal compressive stress at the interface. The historical slip feedback gain coefficient is dimensionless and determines the magnitude of the increase in clamping force after slippage occurs. The measured slip amount during the iterative process is in meters. It is the measured slip value from the feedback of the previous test or from previous tests of the same specification specimen, and is used to predictively adjust the clamping force of the current specimen. The allowable slip threshold is measured in meters. If the slip exceeds this value, the clamping failure is determined. It is a hyperbolic tangent function, and its output range is between -1 and 1. When the output is small, it is close to 0. Much larger When the output approaches 1, it provides smooth saturation characteristics for the slip feedback.
[0048] The implicit equation is solved using a fixed-point iteration method. The initial clamping force is used as the starting point. As the starting point of the iteration The calculation is based on the linear friction model: .Will Substituting the values into the right-hand side of the iterative equation, we obtain the first iteration value. ,Will Substituting the right end yields ,Will Substitute to get After 3 iterations, and A relative change of less than 1% is considered convergence. Set upper and lower limits, with the lower limit being... Ensure that the frictional force is not less than the maximum axial tensile force; the upper limit is The upper limit is the smaller of the maximum clamping pressure and the contact stress being less than the compressive strength of the weld zone. The target adaptive clamping force is calculated based on the maximum output capacity of the clamping hydraulic system and the compressive strength of the material in the weld area of the clamped sample. The smaller value is selected to prevent damage to the weld. The value after limiting is the target value of the output adaptive clamping force. .
[0049] The specific implementation method of centering and adjusting the specimen, controlling the variable pressure jaws to clamp the specimen according to the adaptive clamping force target value, and adjusting the tensile test loading strain rate according to the strength correction coefficient is as follows.
[0050] The sample is retrieved from the feed buffer station by the robot and transferred to the tensile fixture station of the testing machine. The adaptive V-shaped centering unit begins its centering operation. The adaptive V-shaped centering unit consists of a six-degree-of-freedom parallel micro-motion platform and a laser centering measurement system. The laser centering measurement system projects a laser line around the sample circumferentially, measuring the deviation angle and offset distance between the steel bar axis and the loading axis of the testing machine from four directions: front, back, left, and right. The offset data is fed back to the control system of the six-degree-of-freedom parallel micro-motion platform, which performs translational and rotational fine adjustments until the coaxiality deviation angle is less than 0.1 degrees and the offset distance is less than 0.1 millimeters.
[0051] After centering is completed, the variable pressure closed-loop jaw unit uses the adaptive clamping force target value calculated above. The sample is clamped. The variable pressure closed-loop jaw unit consists of a hydraulic servo booster system, a high-precision force sensor, and an embedded controller. The embedded controller receives... The command controls the servo valve to drive the hydraulic cylinder piston, closing the jaws, and the force sensor provides real-time feedback of the actual clamping force signal. The controller employs an incremental PID control algorithm with a control cycle of 10 milliseconds and a steady-state clamping force control accuracy of ±1%. During clamping, the slip detection unit starts synchronously, using a high-resolution laser displacement sensor to monitor the axial displacement of the sample relative to the jaw surface in real time. The slip detection unit records the zero displacement reference point at the beginning of the clamping phase, and acquires displacement signals at a frequency of 200 Hz throughout the test. When the detected slip exceeds the allowable slip threshold... When the test is completed, the re-clamping action is triggered: the controller releases the jaw pressure, the sample returns to the free state, the adaptive V-shaped centering unit re-centers, the variable pressure closed-loop jaw unit re-executes the clamping force loading, the slip detection benchmark is reset to zero, and the test restarts.
[0052] The tensile test loading process is divided into two control segments: the elastic stage and the plastic stage. The elastic stage employs a stress rate control mode, with the stress rate set at 20 MPa per second. The controller adjusts the opening of the hydraulic loading valve based on feedback from the force sensor, ensuring a uniform increase in stress across the specimen cross-section. The slope change in the elastic segment is monitored in real-time on the stress-strain curve. When the slope deviates from linearity by more than a set threshold, the material is deemed to have entered the yielding stage, and the control system automatically switches to the plastic stage.
[0053] The strain rate control mode is used during the plastic stage. The control system is based on the previously calculated strength correction coefficient. Adjust the strain rate setting. When A value greater than 1 indicates that residual tensile stress causes the material to exhibit a higher yield strength than the nominal value. In this case, the strain rate during the plastic stage should be appropriately reduced to determine the strength value at an equivalent rate closer to the standard condition. The strain rate is set to... In this example, The calculated value is 1.032, the standard strain rate. The initial strain rate was 0.00025 per second, and the adjusted actual strain rate was set to 0.00024 per second. The control system acquires the strain of the gauge length section in real time through an extensometer or video extensometer, and adjusts the loading valve to achieve closed-loop control. The yield plateau and the maximum force point are identified through real-time force value signals. A force value fluctuation of less than 1.5% in the plateau section is considered to be at the yield plateau, and the inflection point where the force value reaches its peak and then begins to decline is determined to be the maximum force point.
[0054] The specific implementation method for simultaneously acquiring force data, displacement data, and ambient temperature data during the test is as follows.
[0055] The multi-sensor data synchronous acquisition module is based on a hardware time synchronization unit. This unit emits a periodic synchronization pulse signal, which is used as a reference for timestamping all sensor data acquisition cards. Force data is acquired by a high-precision strain gauge force sensor with a range of 1000 kN, an accuracy class of 0.5, and a sampling frequency of 1000 Hz. Displacement data is acquired by a grating ruler displacement sensor with a resolution of 0.001 mm and a sampling frequency of 1000 Hz, used to record the displacement of the fixture beam and the total elongation of the specimen. Gauge length deformation data is acquired non-contactly by a video extensometer. The video extensometer uses an ultra-high resolution industrial camera to capture the gauge length markers at a frame rate of 100 frames per second. The real-time distance between gauge length points is calculated using a digital image correlation algorithm to obtain the strain of the specimen gauge length. Ambient temperature data is acquired by a servo temperature compensation unit, which includes a platinum resistance temperature sensor installed near the specimen at a distance of 50 mm from the specimen surface. The measurement range is -20 to +80 degrees Celsius, with an accuracy of ±0.1 degrees Celsius and a sampling frequency of 1 Hz. Vibration data is acquired by a triaxial accelerometer mounted on the testing machine frame to monitor and identify abnormal vibration events during testing. All data is aggregated to an industrial detection processor via an industrial Ethernet bus, and the data stream is stored in timestamp alignment.
[0056] The specific implementation method for collecting three-dimensional point cloud data of the fracture surface after the specimen fractures and calculating the corrected post-fracture elongation value after coupling of plastic gradient and temperature in the heat-affected zone is as follows.
[0057] When the testing unit detects that the force value has dropped from the maximum force point to below 0.3 times the maximum force value for 0.2 seconds, it determines that the sample has broken, stops loading, and releases the jaws. The robot then performs the transfer operation of the broken sample. The robot grasps the two broken segments of the sample, maintaining their original axial orientation in the testing machine fixture, and transfers them to the post-fracture measurement station.
[0058] The post-fracture measurement station is equipped with a fracture surface docking platform and a 3D vision measurement unit. One side of the docking platform is a fixed V-groove, and the other side is a precision linear slide. The linear slide is driven by a stepper motor and can move precisely along the axial direction. The lower fractured sample is placed in the fixed V-groove, and the upper fractured sample is placed in the V-block on the linear slide. The stepper motor pushes the upper fractured sample slowly towards the lower fractured sample along the axial direction until the two fracture surfaces come into contact. The contact force is detected by a force-sensitive sensor. When the contact force reaches 1 Newton, the advancement stops, achieving gapless docking.
[0059] The 3D vision measurement unit scans and measures the docked specimen. The unit consists of two high-precision structured light scanning heads, which scan from both sides of the fractured specimen joint, covering the entire gauge length. The scan generates a complete 3D point cloud of the docked specimen, from which the following measurements are extracted: Total length of gauge length after break Automatically identifies pre-marked gauge points on the sample surface. Gauge points are tiny conical holes laser-marked, appearing as localized deep depressions in the point cloud. Their positions are identified by the abrupt changes in the point cloud's normal vector. The axial distance between two gauge points is... The measurement accuracy is 0.01 mm.
[0060] Axial splicing gap of the fracture : Extract the best-fit plane from the point clouds of the two fracture surfaces after docking, and calculate the average distance between the two planes. This distance is the optimal distance. .
[0061] Arithmetic mean deviation of three-dimensional morphology of fracture surface Samples are taken from both fracture surfaces, with the sampling area being a circular region centered at the center of the fracture surface and having a diameter equal to 80% of the nominal diameter. The arithmetic mean of the absolute values of the height deviations of each point in the point cloud within this region relative to the best-fit plane is calculated. .
[0062] After obtaining the above measurements, a correction calculation for the elongation at fracture is performed.
[0063] Obtain the original gauge length before the experiment According to the standards in this field, for steel bars with a diameter of 25 mm, Take 125 millimeters, which is equivalent to 0.125 meters.
[0064] Obtain the average width of the heat-affected zone on one side The value is obtained from the aforementioned visual recognition steps; in this example, it is 8 millimeters, which is equivalent to 0.008 meters.
[0065] Obtaining the nominal diameter of the reinforcing bar It is 0.025 meters.
[0066] Obtaining uniform elongation of the base material Retrieve the HRB400E steel reinforcement from the materials database in this example. It is 18 percent.
[0067] Obtain the uniform elongation of the heat-affected zone The parameters are estimated by the material parameter inversion module using welding thermal simulation parameters from the material database. Based on the carbon equivalent of the reinforcing steel and the peak temperature, and combined with the elongation degradation model of the heat-affected zone microstructure, the results are calculated. Take 14%.
[0068] Obtaining the temperature sensitivity coefficient of elongation The model was calibrated through tensile tests on the base material at different temperatures. Tensile tests were conducted at five temperature levels: 10°C, 20°C, 23°C, 30°C, and 40°C. The elongation after fracture was measured, and a linear fit was performed on the elongation-temperature data. The absolute value of the slope divided by the elongation value at 23°C yielded the model. This example is calibrated. Take 0.0018 per degree Celsius.
[0069] Obtain the average ambient temperature during the test. The temperature is obtained by taking the arithmetic mean of the temperature time series data collected by the servo temperature compensation unit; in this example, it is 28 degrees Celsius.
[0070] Obtain standard reference temperature According to the standard, the temperature should be 23 degrees Celsius.
[0071] Substitute the above parameters into the corrected calculation formula:
[0072] In the formula, The corrected true elongation after fracture is dimensionless, while the elongation after fracture value reflects the true plastic deformation capacity of the base material in the gauge length section after deducting systematic errors. is the total gauge length measured after the specimen was broken and spliced, in meters, and is the measured axial distance between the two gauge points after the specimen was broken and spliced. is the original gauge length before the test, in meters, and is the original distance between gauge points on the sample surface before the test. The axial gap between the fracture surfaces is measured in meters. It represents the length of the axial gap caused by the unevenness of the fracture surfaces during butt jointing, resulting in the inability of the two fracture surfaces to fit together completely. The arithmetic mean deviation of the three-dimensional morphology of the fracture surface is expressed in meters. It characterizes the roughness of the fracture surface, and the larger the value, the rougher the fracture surface. The nominal diameter of the reinforcing bar, measured in meters, is used as a characteristic dimension of the fracture surface to normalize the roughness parameter. (Item in brackets) This represents the proportion of the actual ineffective gap caused by roughness in the fracture joint gap. When the fracture surface is an ideally smooth plane... All splicing gaps must be deducted; when the roughness of the fracture surface increases, the actual ineffective gap decreases, and the deduction amount decreases accordingly. The average width of the heat-affected zone on one side is expressed in meters. Since heat-affected zones exist on both sides of the welded joint, the total width of the heat-affected zone is... . is the uniform elongation of the base material, dimensionless, and represents the uniform plastic deformation capacity of the original region of the base material unaffected by welding heat. is the uniform elongation of the heat-affected zone, dimensionless, and represents the uniform plastic deformation capacity of the material after the microstructure changes in the heat-affected zone; its value is usually lower than that of the parent material. The elongation temperature sensitivity coefficient, with dimensions per degree Celsius, reflects the linear response rate of material elongation to temperature changes. The average ambient temperature during the test is expressed in degrees Celsius. The mean temperature is the arithmetic mean of all temperature data collected by the servo temperature compensation unit from the start of the test to the fracture time. The standard reference temperature is measured in degrees Celsius and is the standard ambient temperature for evaluating material properties.
[0073] The calculation formula consists of three parts of corrective logic. The first term within the curly braces is the total apparent elongation minus the spurious deformation from the fracture joint. Fracture joint gap. The gauge length deformation was mistakenly included after the fractured butt joint, including the roughness-related portion. It is not caused by plastic flow of materials, and needs to be determined from... The second item within the curly braces is a correction for the systematic bias introduced by the plastic gradient in the heat-affected zone (HAZ). Because the HAZ has undergone microstructural changes at peak temperatures, its plastic deformation capacity is typically inferior to that of the base metal. Tensile deformation is concentrated within the HAZ, contributing significantly to local elongation. If calculated based on the homogeneous gauge length assumption, the additional deformation of the HAZ would be evenly distributed across the gauge length, leading to an overestimation of the elongation measurement. This item deducts the true contribution of this non-base metal elongation based on the HAZ width ratio and the difference in elongation between the base metal and the HAZ. The third item is the temperature compensation coefficient. The elongation measured at the test temperature is converted to its equivalent value at the standard reference temperature. Elongation increases with increasing temperature. When the compensation coefficient is greater than 1, the correction value is increased accordingly; when When the compensation coefficient is less than 1, the correction value is reduced accordingly.
[0074] The specific implementation method for extracting the centering error factor, clamping slip factor, curve integrity factor, image quality factor, and temperature stability factor, and calculating the confidence level of the detection result through multi-source uncertainty fusion is as follows.
[0075] Confidence factor for mean square error Calculated based on the final alignment deviation recorded by the laser alignment measurement system. Let the alignment angle deviation be... and center offset The centering error factor is defined as... ,in Take 0.5 degrees. Take 0.5 mm. ,limited Between 0 and 1. In this example It is 0.05 degrees. It is 0.05 mm. It is 0.1. Take 0.9.
[0076] Clamping Slip Confidence Factor The calculation is based on the total slippage accumulated by the slippage detection unit during the test. Let the total slippage be... Define the slip factor. ,but In this example, the total slip is 0.005 mm. It is 0.1 mm. It is 0.05. Take 0.951.
[0077] Curve complete confidence factor Calculations are made based on the continuity, smoothness, and completeness of characteristic points of the force curve. The synchronously acquired force time series data is segmented for analysis. Let the proportion of missing points in the elastic segment data be . The proportion of missing data points in the plastic segment is The proportion of missing breakpoint data is Define complete factors. Weight It is 0.3. It is 0.5. It is 0.2. Therefore... In this example, the curve data is complete and uninterrupted. , , All are 0. Take 1.0.
[0078] Image quality confidence factor Calculations are made based on the quality indicators of the 3D point cloud of the fracture surface. Point cloud completeness is defined. The ratio of the effective number of point clouds to the theoretically required number of point clouds; point cloud noise level. This represents the proportion of discrete noise points in the point cloud to the total number of points. Image quality factor. Weight It is 0.6. It is 0.4. In this example It is 0.99. It is 0.01. It is 0.01. Take 0.99.
[0079] Temperature stability confidence factor Calculated based on the coefficient of variation of the temperature time series during the experiment. Let the standard deviation of the temperature time series be... The coefficient of variation is Define the temperature stability factor. , Take 0.05. Then... In this example, the temperature standard deviation is 0.5 degrees Celsius, and the average is 28 degrees Celsius. It is 0.0179. It is 0.357. Take 0.7.
[0080] Obtain the weight coefficients of each influencing factor The weighting coefficients were determined using the analytic hierarchy process (AHP) combined with historical experimental data, and their values were as follows: , , , , The sum of all weights is 1.
[0081] Get the penalty coefficient for abnormal events and the total number of abnormal events . An empirical value of 0.05 was used. Abnormal events were recorded and counted in real time by each module during the experiment, including centering errors, clamp readjustment, curve data interruption, image acquisition failure, and temperature exceeding limits. No abnormal events were triggered in this example. .
[0082] Substitute the above parameters into the fusion calculation formula:
[0083] In the formula, The confidence level of the final detection result is dimensionless and ranges from 0 to 1, representing the degree to which the current detection result can be accepted as a valid result. The weight coefficients for each influencing factor are dimensionless and determined by the importance of each testing step to the accuracy of the final result, reflecting the influence of that step in the overall confidence assessment. to The above five confidence factors are dimensionless and correspond to the quality assessment results of five independent stages, namely, centering, clamping, force curve, fracture image and temperature environment. This indicates a series of multiplication operations. It is a geometrically weighted average. The logic of using a geometric average instead of an arithmetic average is that the reliability of the test results is constrained by the weakest link effect. If the confidence level of any link is 0, the overall reliability will also be 0. The geometric average can accurately reflect this irreparable characteristic. This is the penalty coefficient for abnormal events, which is dimensionless and used to quantify the degree to which a single abnormal event reduces the reliability of the result. The total number of abnormal events triggered during the entire test is dimensionless, and is the sum of the abnormal counters of each module. This is an exponential penalty term. When there are no abnormal events, this term is 1, and no penalty is deducted for credibility. As the number of abnormal events increases, the penalty term decays exponentially, and credibility decreases rapidly.
[0084] Calculate the confidence level. Substituting the factor values, this example yields the following result. It is approximately 0.94.
[0085] The specific implementation method for determining the validity of test results and generating test reports based on the confidence level is as follows.
[0086] Preset confidence threshold in industrial inspection processor Take 0.85. When the calculated value is... Upon successful completion of the test, the system determines the test results to be valid and automatically generates a formal test report. The test report includes the sample identification information, test date and time, equipment number and calibration status, hash values of all collected raw data files, calculated strength value, corrected true value of elongation after fracture, strength correction factor, final value of adaptive clamping force, 5 confidence factor values, confidence level, and test conclusion.
[0087] when If the system determines that the validity of the test result is insufficient, it will not generate a formal report but will automatically generate a review task. The review task records the specific factors and types of abnormal events that led to the low confidence level. The abnormal route generation unit selects the review path based on the abnormality type: if both the centering error factor and the clamping slip factor are low, the re-clamping test path after secondary positioning is selected; if only the clamping slip factor is low, the re-clamping test path is selected; if the image quality factor is low, the image review path is selected; if the temperature stability factor is abnormal or other non-automatically recoverable abnormalities occur, the manual confirmation route is selected. The sample is returned from the sample output station to the buffer station by the robot and enters the corresponding review process according to the review path.
[0088] The hash chain generation unit in the trusted interpretation and report generation module performs a hash operation once after each test. The data involved in the hash operation includes: the sample's unique identifier, the testing equipment number, the hash value of the valid calibration certificate, the binary stream of all raw sensor data files, all intermediate calculation result parameters, each confidence factor value, and the confidence level. The SM3 cryptographic hash algorithm is used to generate a fixed-length hash value, which is written as a data chain fingerprint into the test report and the local database. Any modification to the above data will inevitably lead to a change in the hash value, ensuring the traceability of the test results.
[0089] In another embodiment of this technical solution, HRB500E steel bars with a diameter of 32 mm were inspected. After entering the line, the vision system identified a weld reinforcement height of 3.0 mm and a heat-affected zone width of 9.5 mm on one side. The material parameter inversion module obtained a residual stress peak of 160 MPa and an initial friction coefficient of 0.32. The calculated strength correction factor was 1.048, and the system adjusted the strain rate in the plastic stage from 0.00025 s / s to 0.00024 s / s. After three iterations, the target clamping force was calculated to be 312 kN. The average ambient temperature during the test was 35 degrees Celsius, which was 12 degrees Celsius higher than the standard temperature. After the specimen fractured, the vision system measured a post-fracture gauge length of 132.5 mm, a fracture gap of 0.9 mm, and a fracture surface roughness of 0.35 mm. The calculated corrected true post-fracture elongation was 17.5 percent. The experiment triggered one abnormal event in the fine-tuning process, and the final calculated confidence level was 0.89, which is higher than the threshold, so the result was deemed valid.
[0090] A flexible automatic detection and reliable interpretation system for welded steel bar joints includes a multimodal visual perception module, a material parameter inversion module, a loading parameter generation module, a flexible execution and closed-loop control module, a multi-sensor data synchronous acquisition module, a post-fracture parameter correction module, and a reliable interpretation and report generation module.
[0091] The multimodal vision perception module comprises a multi-camera array unit, a structured light 3D reconstruction unit, and a barcode scanning and recognition unit. The multi-camera array unit consists of four industrial cameras, synchronously triggered with the projector of the structured light 3D reconstruction unit to simultaneously acquire images of the rebar surface modulated by structured light from four angles. The barcode scanning and recognition unit operates independently of the camera system, reading the sample markings using infrared laser scanning. The output of the multimodal vision perception module includes rebar surface texture images, 3D point cloud data, and sample identification information, which are transmitted to the material parameter inversion module, the flexible execution and closed-loop control module, and the post-fracture parameter correction module.
[0092] The material parameter inversion module runs within the industrial inspection processor and receives visual data transmitted from the multimodal visual perception module. The module embeds three sub-models: a thermo-elastic-plastic inverse model, a heat-affected zone (HAZ) identification algorithm, and a roughness-friction coefficient mapping model. This module outputs the inverted peak residual tensile stress in the weld region, the width of the HAZ on one side, and the initial friction coefficient of the jaw interface, and transmits the results to the loading parameter generation module.
[0093] The loading parameter generation module receives parameters transmitted from the material parameter inversion module, as well as the mechanical property parameters of the parent material retrieved from the material database. The module performs iterative calculations of the strength correction coefficient and the adaptive clamping force target value, and transmits the calculation results, namely the strength correction coefficient and the target clamping force, to the flexible actuation and closed-loop control module and the multi-sensor data synchronous acquisition module.
[0094] The flexible execution and closed-loop control module includes a dual-zone obstacle avoidance gripper unit, an adaptive V-shaped centering unit, a variable pressure closed-loop jaw unit, a slip detection unit, and a follow-up temperature compensation unit. The dual-zone obstacle avoidance gripper unit has a 10 mm deep and 30 mm wide clearance groove milled at the end of the gripper corresponding to the weld seam, ensuring the weld seam is suspended and unloaded during clamping. The adaptive V-shaped centering unit includes a six-degree-of-freedom parallel micro-motion platform and a laser centering measurement system. The variable pressure closed-loop jaw unit is loaded via a hydraulic servo system, with clamping force sensor feedback in a closed loop. The slip detection unit uses a laser displacement sensor to monitor slippage through high-frequency sampling. The follow-up temperature compensation unit uses a platinum resistance temperature sensor. This module performs the entire process of automated sample handling, high-precision centering, closed-loop clamping, and rate-controlled loading.
[0095] The multi-sensor data synchronization acquisition module uses a hardware time synchronization unit as its central hub. This unit employs a GPS-disciplined, temperature-controlled crystal oscillator to provide a clock reference, outputting a second pulse signal and an IRIG-B time code. All sensor data acquisition cards are phase-locked to this time reference, recording sampled data using the same time standard. The force, displacement, deformation, temperature, and vibration data acquired by the module are aggregated and transmitted to the post-fracture parameter correction module and the reliable interpretation and report generation module.
[0096] The post-fracture parameter correction module receives 3D point cloud data of the fracture surface from the multimodal visual perception module and temperature and force data from the multi-sensor data synchronous acquisition module. The module performs post-fracture gauge point identification, fracture morphology parameter calculation, and calculation of the corrected post-fracture elongation value, and transmits the results to the reliable interpretation and report generation module.
[0097] The trusted interpretation and report generation module executes confidence factor extraction, trusted confidence fusion calculation, and judgment logic, as well as anomaly route generation and hash chain generation units. The judgment results drive the automatic generation of detection reports or the creation and distribution of review tasks. The hash chain is written to the report file header to ensure that the data is tamper-proof throughout the entire process.
[0098] Based on the above technical solution, simulation verification experiments were conducted on the two core aspects of this invention: strength correction and elongation correction. The simulations used typical hot-rolled ribbed steel bars HRB400E and HRB500E for construction as examples. Joint samples were prepared using actual welding processes. Surface morphology and material parameters were collected through an industrial testing system and substituted into the established residual stress-strain rate coupling model and heat-affected zone-temperature coupling correction model for calculation. The results were then compared with conventional test results under standard rates.
[0099] The strength correction factor simulation covered a wide range of strain rates from 0.0001 s / s to 0.0025 s / s, examining three typical conditions: low carbon equivalent with low residual stress, medium carbon equivalent with medium residual stress (corresponding to HRB400E), and high carbon equivalent with high residual stress (corresponding to HRB500E). Simulation results show that when the strain rate equals the standard reference rate of 0.00025 s / s, the correction factor for all three conditions approaches 1.0, verifying the model's self-consistency at the reference point. As the strain rate deviates from the reference value and increases upwards, the correction factor exhibits a non-linear upward trend, and this increase is positively correlated with the residual stress level and carbon equivalent. For the HRB400E specimen, the correction factor is approximately 1.032 at a rate of 0.001 s / s, while for the HRB500E specimen with higher residual stress, the correction factor reaches approximately 1.048 at the same rate. This trend quantitatively reveals the amplification effect of welding residual tensile stress on the sensitivity of material strain rate. That is, if high carbon equivalent steel bars are still evaluated according to the strength value under the standard rate at high strain rate, a non-negligible positive deviation will occur. The correction coefficient of the present invention can effectively eliminate this deviation, making the strength measurement results under different loading rates comparable.
[0100] The post-fracture elongation correction simulation was conducted based on measurement data from the two types of reinforcing bars under their respective actual test conditions. The HRB400E sample diameter was 25 mm, and the average ambient temperature was 28°C; the HRB500E sample diameter was 32 mm, and the average ambient temperature was 35°C. The simulation process sequentially presented the original apparent elongation, the median value after deducting the spurious elongation at the fracture surface, the value after further deducting the plasticity deviation in the heat-affected zone, and finally the true post-fracture elongation after temperature compensation. Data showed that the original apparent elongation of HRB400E was 6.56%, which decreased to 5.91% after correcting for the fracture gap, and further decreased to 5.38% after deducting the plasticity gradient effect in the heat-affected zone. Finally, after temperature compensation, the true value converted to the standard reference temperature of 23°C was 5.57%. For HRB500E, the original apparent elongation of 7.5% was corrected in three steps to obtain a true value of 6.24%. Each step of the correction process produced significant and physically meaningful numerical changes. The largest proportion of these changes was due to spurious elongation caused by fracture surface roughness, followed by differences in the plasticity of the heat-affected zone (HAZ). Temperature compensation was adjusted appropriately based on the deviation between the actual and standard temperatures. Simulation results show that the uncorrected apparent elongation is systematically overestimated, and the degree of overestimation is closely related to the width of the weld HAZ, the fracture morphology, and the ambient temperature. The three-level progressive correction logic of this invention can effectively eliminate interference factors from non-base material plastic deformation, allowing the final true value to more accurately reflect the intrinsic ductility properties of the joint base material region.
[0101] Based on the combined results of the two simulations, the corrected model proposed in this invention exhibits stability over a wide parameter range, with the correction magnitude consistent with theoretical expectations. This verifies the correctness and engineering applicability of the residual stress-strain rate coupled model and the heat-affected zone-temperature coupled corrected model. Simulations also demonstrate that the correction effect of this invention is particularly significant for welded steel joints tested under high carbon equivalent, large diameter, and high-temperature conditions. It can reduce the systematic bias of the test results to an acceptable range, providing accurate basic data for subsequent confidence assessment.
[0102] Unless otherwise specified, the equipment components involved in the above embodiments are all conventional equipment components, and the connection methods and control methods involved are all conventional connection methods and control methods unless otherwise specified.
[0103] The present invention has been described in detail above with reference to the embodiments. However, those skilled in the art will understand that, without departing from the spirit of the present invention, various specific parameters in the above embodiments can be changed to form multiple specific embodiments, all of which are common variations of the present invention, and will not be described in detail here.
Claims
1. A method for automatic detection and reliable interpretation of the flexibility of welded steel bar joints, executed by an industrial detection processor, characterized in that, include: S1: Collect three-dimensional point cloud data of the surface of the steel bar welded joint, weld location information, heat-affected zone location and material grade information; S2: Based on the three-dimensional point cloud data and material grade information, the peak residual tensile stress, heat-affected zone width, and initial friction coefficient of the jaw interface in the weld area are inverted; Calculate the strength correction factor for the coupling of residual stress and strain rate; S3: Iteratively calculate the target value of the adaptive clamping force based on the nonlinear evolution of interface friction according to the strength correction coefficient; S4: Center the specimen, control the variable pressure jaws to clamp the specimen according to the adaptive clamping force target value, and adjust the tensile test loading strain rate according to the strength correction coefficient. S5: Simultaneously collect force data, displacement data, and ambient temperature data during the test process; S6: Collect three-dimensional point cloud data of the fracture surface after the specimen fractures, and calculate the true value of the elongation after fracture after correction of the plastic gradient and temperature coupling in the heat-affected zone. S7: Extract the centering error factor, clamping slip factor, curve integrity factor, image quality factor, and temperature stability factor, and calculate the confidence level of the detection result fused from the multi-source uncertainty. S8: Determine the validity of the test results based on the stated confidence level and generate a test report.
2. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The acquisition of 3D point cloud data of the welded joint surface of the reinforcing bar, weld location information, heat-affected zone location and material grade information includes: calling a multi-camera array and structured light 3D reconstruction unit to acquire surface texture images and 3D point clouds of the reinforcing bar; identifying the weld boundary location, heat-affected zone color characteristics and reinforcing bar rib characteristics; reading the sample identification through the barcode scanning unit, and calling the material database to obtain the yield strength, tensile strength and uniform elongation of the base material of the corresponding grade.
3. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The adaptive clamping force target value based on the iterative calculation of the nonlinear evolution of interface friction based on the strength correction coefficient includes: real-time acquisition of the actual contact area between the clamp jaws and the reinforcing bar; The measured slip amount was collected during the iterative process. Using the initial clamping force as the starting point of the iteration, the target value of the clamping force is calculated in three consecutive iterations. The iteration results are subject to upper and lower limits, and the final adaptive clamping force target value is output. The amount of slippage of the sample relative to the jaws is monitored in real time during the clamping process. When the slippage exceeds the threshold, a re-clamping action is triggered.
4. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, Three-dimensional point cloud data of the fracture surface were collected after the specimen fractured, and the corrected post-fracture elongation after the coupling of plastic gradient and temperature in the heat-affected zone was calculated, including: The robot was used to move the two fractured sections of the sample to the post-fracture measurement station. The total gauge length after fracture, the axial splicing gap of the fracture surface, and the arithmetic mean deviation of the three-dimensional morphology of the fracture surface were measured by a three-dimensional vision unit. Obtain the average ambient temperature during the test; The uniform elongation and elongation temperature sensitivity coefficient of the heat-affected zone are obtained by calling the material database. By successively deducting the spurious elongation introduced by the fracture morphology, correcting the systematic deviation introduced by the plastic gradient in the heat-affected zone, and performing temperature compensation calculations, the true value of the corrected elongation after fracture is obtained.
5. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The strength correction factor for calculating the coupling of residual stress and strain rate includes: Obtain the material carbon equivalent-related coupling coefficient, the nominal yield strength of the parent material, the standard-specified reference strain rate, and the material strain rate sensitivity index; By combining the peak residual tensile stress in the weld area with the actual controlled strain rate in the experiment, the amplification effect of residual stress on strain rate sensitivity is quantified, and a dimensionless strength correction coefficient is obtained.
6. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The adaptive clamping force target value calculated iteratively based on the strength correction coefficient to determine the nonlinear evolution of interfacial friction includes: Obtain the nominal diameter of the reinforcing bar, the tensile strength of the base material, the attenuation coefficient of the interfacial friction pressure, and the historical slip feedback gain coefficient; The friction coefficient is expressed as an exponential decay function of the normal pressure. A hyperbolic tangent feedback term of the historical slip is introduced, and the target clamping force is obtained by iteratively solving the implicit equation.
7. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The calculation of the corrected post-fracture elongation after the coupling of the plastic gradient and temperature in the heat-affected zone includes: Obtain the original gauge length, average width of the heat-affected zone on one side, and standard reference temperature before the test; Using the nominal diameter of the steel bar as the characteristic length of the fracture, the errors introduced by the fracture splicing error, the plasticity difference error of the heat-affected zone, and the temperature deviation are successively corrected to obtain the true elongation after fracture.
8. The automatic detection and reliable interpretation method for the flexibility of welded steel bar joints according to claim 1, characterized in that, The confidence level of the detection result calculated by multi-source uncertainty fusion includes: Obtain the weight coefficients of each influencing factor and the penalty coefficients for abnormal events; Five confidence factors were fused using a geometric mean, and the total number of anomalous events triggered throughout the experiment was combined with an exponential penalty to determine the impact of anomalous events on the reliability of the results, resulting in a confidence level ranging from zero to one. When the confidence level is greater than or equal to the set threshold, a formal test report is generated; When the confidence level is less than the set threshold, a review task is generated and the sample is returned to the buffer station. Choose a route based on the type of anomaly, such as secondary positioning, re-clamping test, image verification, or manual confirmation. The sample identification, equipment calibration status, raw data acquisition, parameter correction process, and calculation results are bound into a trusted data chain through hash operations.
9. An automatic detection and reliable interpretation system for the flexibility of welded steel bar joints, applied to the automatic detection and reliable interpretation method for the flexibility of welded steel bar joints as described in any one of claims 1-8, characterized in that, It includes a multimodal visual perception module, a material parameter inversion module, a loading parameter generation module, a flexible execution and closed-loop control module, a multi-sensor data synchronous acquisition module, a post-fracture parameter correction module, and a reliable interpretation and report generation module; The multimodal visual perception module is electrically connected to the material parameter inversion module, the flexible execution and closed-loop control module, and the post-fracture parameter correction module. It is used to collect surface texture images of welded joints, three-dimensional point cloud data, weld location information, heat-affected zone width, and three-dimensional fracture morphology data. The material parameter inversion module is electrically connected to the multimodal visual perception module and the loading parameter generation module, and is used to invert the peak value of welding residual stress, plastic parameters of heat-affected zone and initial friction coefficient of interface based on visual features and material grade. The loading parameter generation module is electrically connected to the material parameter inversion module, the flexible execution and closed-loop control module, and the multi-sensor data synchronous acquisition module. It is used to calculate the strength correction coefficient of the coupling between residual stress and strain rate, and to iteratively calculate the target value of friction nonlinear adaptive clamping force. The flexible execution and closed-loop control module is electrically connected to the multimodal visual perception module, the loading parameter generation module, and the multi-sensor data synchronous acquisition module. It includes a dual-zone obstacle avoidance gripper unit, an adaptive V-shaped centering unit, a variable pressure closed-loop jaw unit, and a follow-up temperature compensation unit, which are used to perform sample picking and placing, centering adjustment, closed-loop clamping, and rate loading. The multi-sensor data synchronization acquisition module is electrically connected to the loading parameter generation module, the flexible execution and closed-loop control module, and the post-fault parameter correction module, and is used for hardware time synchronization acquisition of force, displacement, deformation, temperature and vibration time series data. The post-fracture parameter correction module is electrically connected to the multimodal visual perception module, the multi-sensor data synchronous acquisition module, and the reliable interpretation and report generation module. It is used to calculate the corrected post-fracture elongation value after the coupling of plastic gradient and temperature in the heat-affected zone. The credible interpretation and report generation module is electrically connected to the post-failure parameter correction module. It is used to calculate the credible confidence level of the detection result fused from multiple sources of uncertainty, determine the validity of the result, and generate a traceable detection report.
10. The automatic detection and reliable interpretation system for flexible welded joints of reinforcing bars according to claim 9, characterized in that, The multimodal visual perception module includes a multi-camera array unit, a structured light 3D reconstruction unit, and a barcode scanning and recognition unit; The multi-camera array unit is electrically connected to the structured light 3D reconstruction unit and is used to acquire images of the steel reinforcement surface from different angles. The structured light 3D reconstruction unit is used to generate 3D point cloud data of the steel bar surface and fracture surface; The barcode scanning and identification unit is used to read the sample identification information; The flexible execution and closed-loop control module also includes a slip detection unit, which is electrically connected to the variable pressure closed-loop jaw unit and is used to monitor the amount of slippage of the sample relative to the jaw in real time. The multi-sensor data synchronization acquisition module includes a hardware time synchronization unit, which is electrically connected to all sensors and is used to unify the time reference of all acquired data. The trusted interpretation and report generation module includes an abnormal route generation unit, which is used to generate routes for secondary positioning, re-clamping test, image verification or manual confirmation when the trust confidence is insufficient. The trusted interpretation and report generation module also includes a hash chain generation unit, which is used to bind the sample identity, device status, raw data and calculation results into a traceable data chain.
