Weld Pool Image Research Articles

Observation of weld pool profiles in short-circuiting gas metal arc welding

Short-circuiting gas metal arc welding (GMAW), a main joining process for light-weight steel structures, is a complex, strongly non-linear, dynamic, and multi input/output process. It is essential to implement real-time monitoring and control of this process in order to ensure product quality and service reliability of the welded structures. In this paper, a low-cost and compact vision system is developed to observe the weld pool surface profiles in short-circuiting GMAW under different levels of welding current, welding speed, and contact tube-to-workpiece distance. Appropriate technical means are employed to lower the arc light interference and to synchronize the image grabber with the short-circuiting stage. Clear and complete weld pool surface images are captured by Observation from the Front of the Pool (OFP) and Observation from the Rear of the Pool (ORP), and the weld pool shape and size are determined. The two viewing methods (OFP and ORP) can detect the pool width reliably, but the detected pool length has a larger error. A modification method is proposed. The weld pool images are captured by Observation from the Side Position (OSP), and the tilted slope of the pool surface and the reinforcement height are detected in this way. Through modification, the measurement accuracy of the pool length is improved. This lays a foundation for the process control of short-circuiting GMAW, but further investigation is still required as a next step.

Super-Resolution Analysis on Molten Pool Image of Metal Active-Gas Welding Based on Wavelet with Fractal

According to the unique advantages in image processing combining wavelet and fractal and the different ways of combination, a super-resolution image processing methods are proposed. The methods are characterized by combining the wavelet transform, Wavelet Image Interpolation and FBM Fractal Image interpolation in a certain way to achieve super-resolution image reconstruction. Through processing MAG welding pool images polluted by noises seriously, the results show that: the method proposed in this paper, compared with the method based on wavelet bilinear interpolation, not only effectively raises MAG welding image resolution, but also PSNR of reconstruction images are enhanced 21.1049 dB.

Image processing of weld pool and keyhole in Nd:YAG laser welding of stainless steel based on visual sensing

In order to obtain good welding quality, it is necessary to apply quality control because there are many influencing factors in laser welding process. The key to realize welding quality control is to obtain the quality information. Abundant weld quality information is contained in weld pool and keyhole. Aiming at Nd:YAG laser welding of stainless steel, a coaxial visual sensing system was constructed. The images of weld pool and keyhole were obtained. Based on the gray character of weld pool and keyhole in images, an image processing algorithm was designed. The search start point and search criteria of weld pool and keyhole edge were determined respectively.

Weld-pool image centroid algorithm for seam-tracking vision model in arc-welding process

Visual sensing is an attractive approach to detect the weld position in an arc-welding process, which provides information for seam tracking. However, it is difficult to accurately detect the weld position adjacent to a molten pool because of strong arc disturbances. A novel algorithm based on the weld-pool image centroid is presented to improve the seam-tracking ability. The molten pool images are taken by a camera arranged ahead of the welding torch and the centroid is extracted as a parameter to detect the weld position. It is worth noting that the centroid corresponds to the thermal distribution of the molten pool affected by the offset between the arc and the seam centreline. Therefore the offset between the arc and the seam centreline can be estimated by this centroid. The least square linear regression method is employed to correlate the relationship between the centroid and the offset under different welding conditions. For further analysis of the centroid characteristics, a non-linear neural network is designed with three input variables which are the position, displacement and moving velocity of the centroid, respectively. This neural network model shows higher accuracy of weld detection. In comparison with directly detecting the weld position, the centroid can be measured and computed easily. This algorithm is expected to provide a promising vision model to improve the accuracy of seam tracking in real time, and subsequently to ensure good welding quality.

High Speed Weld Control System of Dual-Bypass MIG Based on LabVIEW

Dual-bypass MIG welding (DB-GMAW) is a new kind of high speed MIG welding with three arcs. In order to monitor the weld process and control it, a high speed weld system of DB-GMAW was built. The system was run by LabVIEW programs, including getting data of system and control output signals. The test result of system showed that all equipments could be used in the same time. Beside images of weld pool and arc, the weld voltages and currents of every part had been acquired. The signals of bypass current and weld speed also had been input TIG welding sources and worktable motor successfully. Meanwhile, the high speed weld formation had a good quality, and all of these established the closed-loop control of high speed DB-GMAW.

Multi‐sensor information fusion in pulsed GTAW based on fuzzy measure and fuzzy integral

PurposeThe status of welding process is difficult to monitor because of the intense disturbance during the process. The purpose of this paper is to use multiple sensors to obtain information about the process from different aspects and use multi‐sensor information fusion technology to fuse the information, to obtain more precise information about the process than using a single sensor alone.Design/methodology/approachArc sensor, visual sensor, and sound sensor were used simultaneously to obtain weld current, weld voltage, weld pool's image, and weld sound about the pulsed gas tungsten‐arc welding (GTAW) process. Then special algorithms were used to extract the signal features of different information. Fuzzy measure and fuzzy integral method were used to fuse the extracted signal features to predict the penetration status about the welding process.FindingsExperiment results show that fuzzy measure and fuzzy integral method can effectively utilize the information obtained by different sensors and obtain better prediction results than a single sensor.Originality/valueArc sensor, visual sensor, and sound sensor are used in pulsed GTAW at the same time to obtain information, and fuzzy measure and fuzzy integral method are used to fuse the different features in welding process for the first time; experiment results show that multi‐sensor information can obtain better results than single sensor, this provides a new method for monitoring welding status and to control the welding process more precisely.

자동차 차체 적용을 위한 레이저-아크 하이브리드 용접의 동축 모니터링 시스템 개발

In this paper, the coaxial monitoring system to capture image of weld pool was developed in laser-arc hybrid welding. In order to obtain the reliable image, green laser was used as a illumination system and measuring components such as band pass filter, ND (Neutral Density) filter and shutter speed was designed and optimized. Using this monitoring system, weld pool images were captured according to laser power, welding speed, welding current and interspace between laser and arc through the experiment. ANOVA (Analysis of Variation) was carried out to identify the influence of process variables on bead widths extracted from captured images of monitoring system. Welding speed and current were major factor to affect weld pool.

Prediction of pulsed GTAW penetration status based on BP neural network and D-S evidence theory information fusion

This paper used multi-sensor information fusion technology in pulsed gas tungsten arc welding. Arc sensor, visual sensor, and sound sensor were used simultaneously to obtain weld current, voltage, weld pool image, and weld sound information about the pulsed gas tungsten arc welding process, and special algorithms were designed to extract the respective signal features of different sensors’ information. Then D-S evidence theory was used to fuse the different signal features to predict the penetration status about the welding process. Aimed at the difficulty of obtaining basic probability assignment in D-S evidence theory, back-propagation (BP) neural network was used to obtain the basic probability assignment. Experiments were done to obtain data for training the BP neural network and test the prediction reliability of D-S evidence theory information fusion, and comparison results showed that D-S evidence theory could effectively use the information obtained by different sensors and obtain better prediction result than single sensor.

Modeling of pulsed GTAW based on multi‐sensor fusion

PurposeWelding process is a complicated process influenced by many interference factors, a single sensor cannot get information describing welding process roundly. This paper simultaneously uses different sensors to get different information about the welding process, and uses multi‐sensor information fusion technology to fuse the different information. By using multi‐sensors, this paper aims to describe the welding process more precisely.Design/methodology/approachElectronic and welding pool image information are, respectively, obtained by arc sensor and image sensor, then electronic signal processing and image processing algorithms are used to extract the features of the signals, the features are then fused by neural network to predict the backside width of weld pool.FindingsComparative experiments show that the multi‐sensor fusion technology can predict the weld pool backside width more precisely.Originality/valueThe multi‐sensor fusion technology is used to fuse the different information obtained by different sensors in a gas tungsten arc welding process. This method gives a new approach to obtaining information and describing the welding process.

Al alloy weld pool control of welding robot with passive vision

PurposeThe purpose of this paper is to describe work aimed to control the Al alloy welding penetration through the passive vision for welding robot.Design/methodology/approachFirst a passive vision system was established. The system can capture the Al alloy welding image. Based on the analysis of the characteristic of the welding image, the composite edge detectors were developed to recognize the shape of the weld seam and the weld pool. To realize the automatic control of the Al alloy‐weld process, the relation between the welding parameter and the quality of the weld appearance was established through the random welding experiment. The wire feed was chosen with PID controller adjusting the wire feed rate according to the weld gap variation.FindingsThis paper finds that the passive vision system can be captured the clear weld seam and weld pool image simultaneously. the method of composite edge detectors can be effectively and accurately recognize the weld seam edges. The wire feed rate controller ensured the welding robot to adjust the wire feed rate according to the gap variation.Research limitations/implicationsThis system has been applied to the industrial welding robot production.Originality/valueThe weld seam and weld pool image can be simultaneously captured by the passive vision system. The composite edge detectors have been developed for the passive vision method. The controller has been set up for Al alloy welding process based on the neural network.

Penetration Control of GMAW Based on Multi-informational Fusion from the Front Surface

The validity of welding molten pool image is judged by quantificational information fusion of welding pool image and temperature collected with multi-sensors from the front surface. The temperature field based on semi-ellipsoidal heat source is modified by considering the influence of droplets heat content, which comes into a novel three dimensional temperature field analytical model. The molten temperature of countless positions can be gained by the aid of welding pool image information. The analytical model of real time modified three dimensional transient temperature field is obtained by solving the novel temperature field model super-coupled equation through dampedleast square algorithm and identifying four variational coefficients from semielliptical heat source and droplet heat content. The molten width on the rear surface is controlled by PID controller with the aid of the transient temperature model which is proved in the welding process with satisfied result.

Visual sensing and penetration control in aluminum alloy pulsed GTA welding

In this paper, visual sensing and penetration control in aluminum (Al) alloy pulse gas tungsten arc welding were researched. Firstly, a three-optical-route visual sensor was designed. The sensor can capture the weld pool from three directions at the same time. After analyzing the influences of different factors on weld pool image, serials of clear and stable weld pool images were obtained. Then, image processing technologies were developed to compute back topside weld pool geometry parameters. Wavelet transform and Canny operator were synthesized to get all edges in the weld pool image. After noise removal and calibration, the breaking edges of weld pool were obtained, and then piecewise curve fitting based on polynomial function were used to recover the whole weld pool edge. Lastly, proportional–integral–differential and a multiplex controller were designed to control penetration in welding process. Experiments proved that visual-based penetration control can insure welding quality well from weld pool width and reinforcement.

Artificial neural network approach for estimating weld bead width and depth of penetration from infrared thermal image of weld pool

In this article an artificial neural network based system to predict weld bead geometry using features derived from the infrared thermal video of a welding process is proposed. The multilayer perceptron and radial basis function networks are used in the prediction model and an online feature selection technique prioritises the features used in the prediction model. The efficacy of the system is demonstrated with a number of welding experiments and using the leave one out cross-validation experiments.

Vision based measuring system for both weld pool and root gap in continuous current GTAW

AbstractA low cost vision based measuring system is developed to detect both the weld pool geometry and root gap simultaneously. Efforts are made from both hardware and software aspects to decrease the strong interference from arc light during gas tungsten arc welding (GTAW) process with continuous current. A normal charge coupled device (CCD) camera is cooperated with specially made composite filter lens and the video signals before A/D conversion are adjusted according to the welding current. Clear images of both root gap and weld pool are captured under different welding parameters. The grey level distribution characteristics of root gap edges and weld pool edges in images are analysed and utilised for developing the algorithms for image processing. The continuum of weld pool edge and the linear features of root gap are also utilised in the algorithm. The results show that the algorithms can extract the weld pool edge and root gap edge effectively. A solid foundation for weld seam tracking and weld pen...

A novel visual image sensor for CO2 short circuiting arc welding and its application in weld reinforcement detection

In this paper, a novel visual image sensor was designed to detect the weld pool images during CO2 short circuiting arc welding. Based on a charge coupled device (CCD) camera, the new sensor was smaller than the previous one. As a result, larger operation space is achieved for the welding torch which is attached to it. Moreover, the mismatching problem between the fixed shooting schedule of the CCD camera and the random short circuiting metal transfer during welding was successfully resolved. By eliminating the influences from welding arc and spatter, clear and complete welding pool images were acquired from different angles around the welding torch. Based on these images, the shape parameters of the welding pool were defined and the relationship between these parameters and weld reinforcement was analysed. In addition, the multiple linear regression (MLR) model and the artificial neutral network (ANN) model were proposed and trained, respectively, to detect weld reinforcement from weld pool images. Results of both the MLR model and the ANN model show that weld reinforcement can successfully be detected.

Sensing and seam tracking of welding line in backingless V groove welding

The formation of stable back beads in the first layer weld during one side multilayer welding is important to achieve high quality welded metal joints. The authors tried to apply the switch back welding method to butt welding to join discs with flat metal plates without backing plates. During the welding, the gap was detected by processing the weld pool image taken with a CCD camera. Moreover, the tip of the electrode wire is changing owing to a curve of its wire extension in accordance with a rotation of the torch axis. Even though teaching is achieved before the welding, the weaving centre is shifted from the gap centre. The controller of the seam tracking was designed to prevent any steady state error.

Reconstructing a three-dimensional P-GMAW weld pool shape from a two-dimensional visual image

The online detection of the three-dimensional (3D) shape of a weld pool is a key unsolved question for weld shape control. In this paper, a method to reconstruct the 3D shape of a weld pool boundary from a two-dimensional (2D) visual image is proposed. Firstly, a new 3D model was developed to describe the weld pool geometry in pulsed gas metal arc welding (P-GMAW). In this model, four parameters could be used to determine the weld pool shape: the maximum width W, the length of the pool tail L, the height of the rear of the pool H and the projection of the rear angle onto the work piece plane θL. Then, the clear weld pool images from the P-GMAW were captured from the upper side, and a series of algorithms were developed to extract the model geometrical parameters from these images. Finally, the 3D shape of the weld pool was reconstructed from the 2D images. The results were then validated with metallograph observations. The measurement error of this method was investigated, and methods for improvement were discussed.

Reconstruction of weld pool surface based on shape from shading

A valid image-processing algorithm of weld pool surface reconstruction according to an input image of weld pool based on shape from shading (SFS) in computer vision is presented. The weld pool surface information is related to the backside weld width, which is crucial to the quality of weld joint. The image of weld pool is recorded with an optical sensing method. Firstly, the reflectance map model, which specifies the imaging process, is estimated. Then, the algorithm of weld pool surface reconstruction based on SFS is implemented by iteration scheme and speeded by hierarchical structure. The results indicate the accuracy and effectiveness of the approach.

Vision-based neuro-fuzzy control of weld penetration in gas tungsten arc welding of thin sheets

This paper develops a vision-based neuro-fuzzy system to control the weld joint penetration in Gas Tungsten Arc Welding (GTAW) of thin sheets. To this end, a camera equipped with a specially designed composite light-filter is used to observe the weld pool from the topside of the workpiece so that comparatively distinct images of the weld pool are obtained. As the Backside weld Width (BW) reflects the degree of the weld joint penetration, a model describing the relationship between the weld pool surface geometrical parameters (which can be extracted from the weld pool images) and the backside weld width is constructed. A neuro-fuzzy controller and a learning algorithm are developed to address dynamic and non-linear characteristics of the welding process. The controller can learn fuzzy rules and adjust the fuzzy rules with the variation of welding conditions automatically. Simulation and control tests demonstrated the effectiveness of the developed control system.

Seam tracking based on estimation of weld position using Kalman filtering

A seam tracking method is presented based on the estimation of weld position during the gas tungsten arc welding process. Kalman filtering of the weld pool images from a visual sensor is applied to compute recursively the solution to the weld position equations which are established based on an estimation of the centroid position of the weld pool images. This centroid, the position of which corresponds with the weld position, is extracted as the measurement eigenvector. The evolution of the weld position data from the weld pool images can be described through an appropriate process model, so that the weld position can be detected by applying a Kalman filter. This allows adjustment of the welding torch position in real time, which may significantly reduce processing time and promote seam tracking accuracy. Simulations and actual welding experiments have demonstrated the effectiveness of the proposed algorithm in the presence of weld pool image noise and have demonstrated the robustness of weld position detection for seam tracking.