Robotic Arc Welding Research Articles

A preliminary study on gas metal arc welding-based additive manufacturing of metal parts

Introduction: In the past three decades, additive manufacturing (AM), also known as 3D printing, has emerged as a promising technology, which allows the manufacture of complex parts by adding material layer upon layer. In comparison, with other metal-based AM technologies, gas metal arc welding-based additive manufacturing (GMAW-based AM) presents a high deposition rate and has the potential for producing medium and large metal components. To validate the technological performance of such a manufacturing process, the internal quality of manufactured parts needs to be analyzed, particularly in the cases of manufacturing the parts working in a critical load-bearing condition. Therefore, this paper aims at investigating the internal quality (i.e., and mechanical properties) of components manufactured by the GMAW-based AM technology.
 Method: A gas metal arc welding robot was used to build a thin-walled component made of mild steel on a low-carbon substrate according to the AM principle. Thereafter, the specimens for observing and mechanical properties were extracted from the built thin-walled component. The of the specimen were observed by an optical microscope; the hardness was measured by a digital tester, and the tensile tests were carried out on a tensile test machine.
 Results: The results show that the GMAW-based AM-built thin-walled components possess an adequate that varies from the top to the bottom of the built component: structures with primary dendrites in the upper zone; granular structure of with small regions of at grain boundaries in the middle zone, and grains of in the lower zone. The hardness (ranged between 164±3.46 HV to 192±3.81 HV), yield strength (YS offset of 0.2% ranged from 340±2 to 349.67±1.53 ), and ultimate tensile strength (UTS ranged from 429±1 to 477±2 ) of the GMAW-based AM-built components were comparable to those of wrought mild steel.
 Conclusions: The results obtained in this study demonstrate that the GMAW-based AM-built components possess adequate and good mechanical properties for real applications. This allows us to confirm the feasibility of using a conventional gas metal arc welding robot for additive manufacturing or repairing/re-manufacturing of metal components.

Real-time seam defect identification for Al alloys in robotic arc welding using optical spectroscopy and integrating learning

Accurate on-line weld defect detection in robotic arc welding manufacturing is still challenging, due to the complexity and diversity of weld defects. In this study, a new real-time defect identification method is proposed for Al alloys in robotic arc welding, using arc optical spectroscopy and an integrated learning method. Spectrum feature was extracted, based on the absolute coefficients of the principal components. Feature importance was quantitatively evaluated using the mean decrease accuracy of Principal Component Analysis-Random Forest (PCA-RF). A new indicator, e.g., Importance Factor, was proposed, based on the variance of the out-of-bag test error of RF to select the optimal feature subset. The proposed PCA-RF proved to effectively identify five classes of weld defects with better performance than support vector machine and back propagation neural network. Finally, the selection pattern of spectrum feature subset was investigated, before revealing the correlation mechanism of the selected lines spectrum and weld process.

A novel system for off-line 3D seam extraction and path planning based on point cloud segmentation for arc welding robot

Due to ever increasing precision and automation demands in robotic welding, the automatic and robust 3D seam extraction has become a research hot-spot of the welding robots. At present, most of the research work about seam extraction is aimed at butt joints. Nevertheless,too little work has been devoted to fillet joints and lap joints. Consequently,to achieve robust 3D seam extraction of different weld seams, a novel seam extraction system is proposed according to the 3D structures of welding work pieces. Firstly, a fringe projection system based on Digital Light Processing(DLP) projector is designed to measure the appearance of welding work pieces. Secondly, fusion of the shape information of welding work piece, a 3D seam extraction algorithm is proposed based on point cloud segmentation. Finally, according to the space structure of weld seams, the 3D seam path model and pose estimation are solved based on the established mathematical model of weld seams. Experiments show that the proposed algorithm could well solve different weld seams, such as fillet joints, butt joints and lap joints. Meanwhile, it could well overcome the influence of the materials of welding work pieces, scratch and rust.

A study on seam tracking in robotic GMA welding process

Abstract In recent years, an image processing and sensor technology for welding automation, intelligence and information with rapid development of welding automation have become an inevitable trend. Owing to the features of mass information, high reliability and applicability in weld inspection for robotic arc welding, the sensor-based weld seam detection has been widely applied and gradually become a hot topic. Furthermore, the robotic arc welding can use its own machine vision system to obtain workpiece positioning information, weld space position, bead geometry and weld seam tracking. The most important task in the machine vision system is how to process the collected visual information to accurately locate the center of the weld and achieve weld tracking. During the welding process, the images taken are often unclear and the features are not obvious due to the interference of arc, splash and smoke. Therefore, it is necessary to carefully design the image acquisition process and the image processing algorithm to complete the weld inspection. In this study, the robotic GMA(Gas Metal Arc) welding has been applied to bead-on-plate welding with vertical position. Thermal images of two different angles (0°, 30°) based on infrared camera through experiments might be obtained for developing an optimal image processing algorithm which was applied in seam tracking. In image processing procedure, the first step is image pre-processing including noise removal, contrast enhancement and binarization. The obtained image adopts multi-scale edge detection method, which can extract complex image edges well, then segment the image, accurately locate the weld part, perform image size measurement and centerline extraction, and finally convert the image information of the weld into the world calibration. The calculated data is then fed to a controller which controls the welding electrode movement. Use of machine vision system has eliminated the requirement to pre-feed the workpiece data for the robot to detect the dimension and position of the workpieces. The developed algorithm can detect weld image effectively, and extract the edge and the center line of the weld successfully

Investigating the Effects of Welding Processes on Tensile Stress and Strain Properties of Welded Mild Steel Plates Using Statistical Analysis

As test samples, mild steel plates with thicknesses of 0.5 mm, 0.6 mm, 0.7 mm, 0.8 mm, 0.9 mm, and 1.0 mm were made. After welding these test samples underwent Tensile Stress and strain tests with the Built Welding Robot and Manual Electric Arc Welding Machine. Both data collected from tensile stress and tensile stress were analyzed and the data produced from Electric Arc welding operations, the Robot welding operations and un-welded plates (control) were compared with one another. The analyses of the data obtained from the developed welding robot, manual electric arc welding and un-welded (control) mild steel plates of different thicknesses were carried out for tensile stress and strain. The descriptive statistics, ANOVA analysis, test of homogeneity of Variances and Post Hoc test (Least Significant Differences) were the statistical tools deployed using Statistical Package of Social Sciences (SPSS version 2016). The results showed that the robot welding sample produced gave the lowest tensile stress while the un-welded samples (CONTROL) gave the highest. The un-welded (CONTROL) samples gave the highest tensile strain values while the lowest was given by the developed robot welding samples. Finally, it was evident from the analyses results that the welding processes have significant impact on the tensile stress and strain properties of the welded mild steel plates and that good welding quality can be achieved more with the developed welding robot.

Multi-objective path optimization for arc welding robot based on discrete DN multi-objective particle swarm optimization

Rational weld seams sequence is important for improving welding productivity and quality. Hence, intelligent path optimization strategy is introduced to obtain optimized weld seams sequence in this article. The path length and total welding deformation are considered for multi-objective path planning. First, the optimization problem description is presented. At the same time, the path length and the total welding deformation of some sequences are calculated. Then, improved multi-objective particle swarm optimization algorithm is studied. In addition, the agent model is obtained based on the sample data and experiment design. At last, the proposed algorithm is applied to optimize the welding path length and total welding deformation. The simulation results show the effectiveness of the optimization strategy.

Optimal trajectory generation of an industrial welding robot with kinematic and dynamic constraints

PurposeThis paper aims to present an optimal trajectory planning for industrial MOTOMAN MA1440A gas metal arc welding system. A new and efficient evolutionary algorithm, enhanced multi-objective teaching learning-based optimization (EMOTLBO) method, i.e. TLBO with non-dominated sorting approach has been proposed to obtain the optimal joint trajectory for the defined weld seam path.Design/methodology/approachThe joint trajectory of the welding robot need to be computed in an optimal manner for proper torch orientation, smooth travel of the robot along the weld path and for achieving higher positional accuracy. This can be achieved by limiting the kinematic and dynamic variations of the robot joints like joint jerks, squared acceleration and torque induced in the joints while travel of the robot along the weld path. Also, the robot travel should be done within minimum possible time for maintaining productivity. This leads to a multi-objective optimization problem which needs to be solved for maintaining proper orientation of the robot end effector. EMOTLBO has been proposed to obtain the Pareto front consisting of optimal solutions. The fuzzy membership function has been used to obtain the optimal solution from the Pareto front with best trade-off between objectives.FindingsThe proposed method has been implanted in MATLAB R2017a for simulation results. The joint positions have been used to program the robot for performing welding operation along the weld seam. From the simulation and experimental results, it can be concluded that the proposed approach can be effectively used for optimal trajectory planning of MOTOMAN MA 1440 A arc welding robot system as a very smooth and uniform weld bead has been obtained with maximum weld quality.Originality/valueIn this paper, a novel approach for optimal trajectory planning welding arc robot has been performed. Though trajectory planning of industrial robots has been done before, it has not been done yet for welding robot. The objectives are formulated taking in consideration of requirement of welding process like minimization of joint jerks and torques induced during welding operation due to travel of robot with the effect of arc spatter, minimization of squared acceleration for maintaining constant joint velocity and finally minimization of total travel time for maintaining productivity.

An adaptive feature extraction algorithm for multiple typical seam tracking based on vision sensor in robotic arc welding

Intelligent robotic welding is an indispensable part of modern welding manufacturing, and vision-based seam tracking is one of the key technologies to realize intelligent welding. However, the adaptability and robustness of most image processing algorithms are deficient during welding practice. To address this problem, an adaptive feature extraction algorithm based on laser vision sensor is proposed. According to laser stripe images, typical welding seams are classified into continuous and discontinuous welding seams. A Faster R-CNN model is trained to identify welding seam type and locate laser stripe ROI automatically. Before welding, initial welding point is determined through point cloud processing to realize welding guidance. During seam tracking process, the seam edges are achieved by a two-step extraction algorithm, and the laser stripe is detected by Steger algorithm. Based on the characteristics of two kinds of welding seams, the corresponding seam center extraction algorithms are designed. And a prior model is proposed to ensure the stability of the algorithms. Test results prove that the algorithm has good adaptability for multiple typical welding seams and can maintain satisfying robustness and precision even under complex working conditions.

Weld image deep learning-based on-line defects detection using convolutional neural networks for Al alloy in robotic arc welding

Accurate on-line weld defects detection is still challenging for robotic welding manufacturing due to the complexity of weld defects. This paper studied deep learning–based on-line defects detection for aluminum alloy in robotic arc welding using Convolutional Neural Networks (CNN) and weld images. Firstly, an image acquisition system was developed to simultaneously collect weld images, which can provide more information of the real-time weld images from different angles including top front, top back and back seam. Then, a new CNN classification model with 11 layers based on weld image was designed to identify weld penetration defects. In order to improve the robustness and generalization ability of the CNN model, weld images from different welding current and feeding speed were captured for the CNN model. Based on the actual industry challenges such as the instability of welding arc, the complexity of the welding environment and the random changing of plate gap condition, two kinds of data augmentation including noise adding and image rotation were used to boost the CNN dataset while parameters optimization was carried out. Finally, non-zero pixel method was proposed to quantitatively evaluate and visualize the deep learning features. Furthermore, their physical meaning were clearly explained. Instead of decreasing the interference from arc light as in traditional way, the CNN model has taken full use of those arc lights by combining them in a various way to form the complementary features. Test results shows that the CNN model has better performance than our previous work with the mean classification accuracy of 99.38%. This paper can provide some guidance for on-line detection of manufacturing quality in metal additive manufacturing (AM) and laser welding.

Random forest-based real-time defect detection of Al alloy in robotic arc welding using optical spectrum

Aluminum alloy of arc welding is the main technology for the key components manufacturing in aerospace, nuclear power, ship and so on. Real-time weld defects detection is still challenging due to the complexity and diversity of weld defects. Arc optical Spectroscopy emission is the key information generated during arc welding process. However, how to select the effective spectrum feature from high dimension of arc spectrum is crucial for improving the accuracy of defects recognition. This paper proposed an on-line defects detection method for aluminum alloy in robotic arc welding based on random forest and arc spectrum. Firstly, preprocessing of arc spectrum was carried out before 50 features were extracted. Then, a quantitative index of feature importance is proposed based on mean decrease accuracy and mean decrease Gini to reduce the feature redundancy. Six spectral features were selected and analyzed in terms of the construction pattern. Furthermore, the defect identification model was established based on random forest and the optimal feature subset. Comparing with RBF and BP models, it can achieve better performance in identifying three typical defects, including incomplete penetration, burn-through and porosity. This paper can provide some guidance for data mining of optical information and intelligent manufacturing.

Corrections to “A High-Speed Seam Extraction Method Based on the Novel Structured-Light Sensor for Arc Welding Robot: A Review”

there was an error in the title of [1] ; it is a regular paper, not a review paper. The right title of this article is “A High-Speed Seam Extraction Method Based on the Novel Structured-Light Sensor for Arc Welding Robot.”

Coordinated motion analysis and simulation of arc welding robot and positioner

Direct at the problem of coordinated motion of the arc welding robot positioner, the motion between the arc welding robot and the positioner is required to be strongly coupled during the process of planning weld movement. In order to express this characteristic, in this article the coupling mathematical model of arc welding robot and positioner is established. The motion equation is simulated by MATLAB neural network In the MATLAB simulation software environment. The simulation results show that the constructed model can accurately reflect the accuracy of the arc welding robot positioner coupling motion.

Wire Arc Additive Manufacturing by Robot Manipulator: Towards Creating Complex Geometries

Abstract Additive manufacturing (AM) is the umbrella term that covers a variety of techniques that build up structures layer-by-layer as opposed to machining and other subtracting methods. It keeps evolving as an important technology in prototyping and the development of new devices. However, using AM on a larger scale is still challenging, as traditional methods require the AM machines to be larger than the manufactured structure. The research presented in this paper is a continuation of our work on assessing the possibility of large-scale robotic AM. The focus in this paper is the feasibility of large-scale AM of metallic materials by arc welding. A series of experiments with robotic arc welding using an ABB IRB2400/10 robot are presented and discussed. These experiment will help map some of the challenges that need to be addressed in future work.

Application of robotics in rail car manufacturing learning factory: A case of welding complex joints

Abstract The use of a learning factory creates a platform for putting to use research innovations, directly into industrial practice. Railcar manufacturers are migrating to fully automated and innovative solutions with industrial welding robot systems playing a pivotal role in the paradigm shift for tasks such as the production of bogie frames and their components. Furthermore, the skills required for bringing about the automation of dangerous and dirty task in the railcar industry are scarce. There is pressure from stakeholders on how researchers and manufacturers can be more efficient in developing a theoretical and practical application framework for a real-time production in the railcar manufacturing industry. Therefore, researchers must look for ways to transfer and demonstrate the application of robotics research in the shop floor with causing disruptions. In addition, there is a need to develop appropriate and customised skill transfer solutions for both smaller and large business owners especially for tasks that are boring for humans. This paper presents an offline-line scheme for building robotic arc welding cell using device task definition and workbenches of the Delmia V5 Robotics, within the learning factory framework.

Research on Rapid Prototyping Control System of Robotic Arc Welding

Rapid prototyping; Additive manufacturing; Robot; Arc welding

Image Segmentation Approaches for Weld Pool Monitoring during Robotic Arc Welding

There is a strong correlation between the geometry of the weld pool surface and the degree of penetration in arc welding. To measure the geometry of the weld pool surface robustly, many structured light laser line based monitoring systems have been proposed in recent years. The geometry of the specular weld pool could be computed from the reflected laser lines based on different principles. The prerequisite of accurate computation of the weld pool surface is to segment the reflected laser lines robustly and efficiently. To find the most effective segmentation solutions for the images captured with different welding parameters, different image processing algorithms are combined to form eight approaches and these approaches are compared both qualitatively and quantitatively in this paper. In particular, the gradient detection filter, the difference method and the GLCM (grey level co-occurrence matrix) are used to remove the uneven background. The spline fitting enhancement method is used to remove the fuzziness. The slope difference distribution-based threshold selection method is used to segment the laser lines from the background. Both qualitative and quantitative experiments are conducted to evaluate the accuracy and the efficiency of the proposed approaches extensively.

A High-Speed Seam Extraction Method Based on the Novel Structured-Light Sensor for Arc Welding Robot: A Review

In the modern manufacturing, the teaching-playback mode and the off-line programming mode of the welding robots still play the important roles. However, these modes cannot meet the autonomous and adaptive capabilities of the welding robots. To improve the flexibility of the welding robots, fast and accurate seam extraction is the key link to realize the intelligent welding robots. A fast and accurate seam extraction algorithm based on the novel structured light vision system is proposed in this paper. First, the digital light processing projector is used to construct the multi-function structured light vision sensor in this paper. It could generate different pattern images to adapt different welding tasks. Second, during the current research work about seam extraction, the seam extraction algorithm based on morphological image processing has been widely used and improved. However, these algorithms include much machine vision algorithms and the speed of these algorithms is too slow to meet real-time requirements. Meanwhile, these algorithms are difficult to adapt to different types of weld seam. In order to achieve fast and accurate extraction of various types of weld seams, the target tracking algorithm based on the kernelized correlation filters algorithm is applied into this paper. Experimental results show that the proposed method could well realize fast and accurate seam extraction of various types of weld seams.

From Wire to Seamless Flow-Formed Tube: Leveraging the Combination of Wire Arc Additive Manufacturing and Metal Forming

Flow forming is a metal-forming process for producing axisymmetric hollowed tubes starting from cylindrical workpieces. In cases where the preformed tube includes defects, impurities or inconsistencies, the final product may not withstand the required standard. In this work, the preformed tubes were produced by metal wires melted into the desired geometry using a gas-metal arc-welding robot. Two geometries were built to allow either forward or backward tube spinning of parts from ER-70S-6 and AWS ER316L welding wires. Following the building process, the parts were machined to fit a flow-forming operation. The mechanical and metallurgical properties of the wires, preformed cylinders and final tubes were compared with conventionally manufactured material. Results indicate that the mechanical properties of the products of the proposed process were superior to those of conventional processing. Applying flow forming to wire-arc additive manufacturing yields a deeper understanding of the material qualities over a large building envelope and can be used as valid quality criteria for the additive manufacturing process qualification.

Hand–eye calibration of arc welding robot and laser vision sensor through semidefinite programming

Purpose This paper aims to propose a hand–eye calibration method of arc welding robot and laser vision sensor by using semidefinite programming (SDP). Design/methodology/approach The conversion relationship between the pixel coordinate system and laser plane coordinate system is established on the basis of the mathematical model of three-dimensional measurement of laser vision sensor. In addition, the conversion relationship between the arc welding robot coordinate system and the laser vision sensor measurement coordinate system is also established on the basis of the hand–eye calibration model. The ordinary least square (OLS) is used to calculate the rotation matrix, and the SDP is used to identify the direction vectors of the rotation matrix to ensure their orthogonality. Findings The feasibility identification can reduce the calibration error, and ensure the orthogonality of the calibration results. More accurate calibration results can be obtained by combining OLS + SDP. Originality/value A set of advanced calibration methods is systematically established, which includes parameters calibration of laser vision sensor and hand–eye calibration of robots and sensors. For the hand–eye calibration, the physics feasibility problem of rotating matrix is creatively put forward, and is solved through SDP algorithm. High-precision calibration results provide a good foundation for future research on seam tracking.

Feature point extraction based on contour detection and corner detection for weld seam

In an arc welding robot system with vision sensor, the extraction precision of the weld seam position has an important effect on welding quality. However, due to disturbances such as camera resolution, scanning posture, reflections, etc., it is a great challenge to identify weld seam accurately. In this paper, weld seam is considered as a special curve which consists of feature points, based on that, an effective method is proposed to extract the feature points by using contour detection and corner detection. Experiment results show that feature points can be extracted successfully by using this method, so that the robot can weld the work part along the feature points detected. The limitation of this method is that the detected feature points are sometimes inconsistent with the weld center points, which may affect the welding quality.