Weld Inspection Research Articles

An on-line weld inspection method for underwater offshore structure based on an improved deep convolutional network

ABSTRACT The acoustic signal analysing is a crucial evaluation method for welding repair of in-service marine structures. However, the complex underwater environment complicates the quality evaluations. In addition, the weak acoustic signal poses challenges for traditional methods during signal processing, such as low signal-to-noise ratio and difficulty in feature extraction. In this paper, an improved deep convolutional network method for online weld inspection of underwater offshore structures is proposed. First, a Variational Mode Decomposition (VMD) algorithm is used to extract weak signals. The raw acoustic signal is denoised, and the effective pulses are extracted. Next, the recurrence and gradient plot methods are applied to extract the pulse features and to generate corresponding images. Finally, a VGG16 deep convolutional network is used to classify image samples and determine weld quality online. Results indicate that the VMD algorithm can process and reduce noise, the recurrence and gradient plot provide quality related image features for the neural network. The VGG16 neural network outperforms the VGG19, RESNET34, and RESNET50 neural networks and can achieve a 97.5% recognition rate. This paper offers some avenues for online identification of weld quality in underwater welding.

On-line measurement of nugget diameter in automatic resistance spot welding based on embedded ultrasound probe

The inspection of resistance spot welding (RSW) quality is crucial in the sheet metal joining industries. This paper proposes an enhanced approach for the automated real-time measurement of weld nugget diameter in welded joints, utilizing an ultrasonic probe integrated within the electrode. The ultrasonic signals received by the probe are sequentially arranged to generate consecutive ultrasonic B-scan images, which effectively capture the entire weld nugget growth and solidification process. The corresponding maximum time of flight (TOF) for the liquid nugget is extracted from the B-scan image. In the condition of qualified welding quality, a mathematic equation between maximum TOF and weld nugget diameter is established, while the opposite is alarmed. The on-line quality inspection of RSW can autonomously determine the qualification of welding quality, as well as quantitatively evaluate the weld nugget diameter in qualified weld joints. The research is applicable to various processing and manufacturing fields.

Potential of Radioactive Isotopes Production in DEMO for Commercial Use

There is widespread use of nuclear radiation for medical imagery and treatments. Worldwide, almost 40 million treatments are performed per year. There are also applications of radiation sources in other commercial fields, e.g., for weld inspection or steelmaking processes, in consumer products, in the food industry, and in agriculture. The large number of neutrons generated in a fusion reactor such as DEMO could potentially contribute to the production of the required radioactive isotopes. The associated commercial value of these isotopes could mitigate the capital investments and operating costs of a large fusion plant. The potential of producing various radioactive isotopes was studied from material pieces arranged inside a DEMO equatorial port plug. In this location, they are exposed to an intensive neutron spectrum suitable for a high isotope production rate. For this purpose, the full 3D geometry of one DEMO toroidal sector with an irradiation chamber in the equatorial port plug was modeled with an MCNP code to perform neutron transport simulations. Subsequent activation calculations provide detailed information on the quality and composition of the produced radioactive isotopes. The technical feasibility and the commercial potential of the production of various isotopes in the DEMO port are reported.

Eddy current array developments for remote deployment on the ITER project

The inspection of welds and components for nuclear applications can be more challenging due to extreme levels of radiation and temperature. A development project in collaboration with the UK Atomic Energy Authority (UKAEA) is being conducted to develop non-destructive testing (NDT) procedures for inspecting the remotely welded joints within the ITER facility. Initially, a feasibility study has been carried out to identify the most suitable non-destructive technique. Eddy current array (ECA), phased array ultrasonics and guided wave electromagnetic acoustic transducers (EMATs) were considered as candidate solutions and, following the feasibility study, ECA was selected to be taken forward for further development. Subsequently, dedicated high- and low-frequency array probes have been developed and tested to detect and size surface and buried defects employing encoder and calibration curves. The research has been extended to characterise defects such as tungsten inclusions using the phase angle. To avoid damage to the electronics in extreme environments, the ECA probe coils were separated from the excitation and data acquisition unit with 60 m coaxial coils to achieve a high signal-to-noise ratio. The results revealed that the ECA technique can successfully be deployed remotely for the detection and sizing of defects down to 3 mm in extreme environments.

Aplikácia nedeštruktívnych metód kontroly na zvarové spoje vyrobené z plastu

The article describes the possibilities of non-destructive testing of welds made of plastic materials. Both surface and volume methods of non-destructive testing can be used for plastic welds. Non-destructive testing of plastics has certain limitations. The limitations are related to the physical-mechanical properties of the material, as well as the geometry of the tested parts. The most common inspection of welds on plastic materials is a visual inspection, which is carried out to the extent of 100 %. For the inspection of plastic welds, it is also possible to use the penetration method and, with limitations, also ultrasonic and radiographic inspection.

Edge Detection and Defects Checking of Binder Clip and Welded Joint using a Python-Based Algorithm: Applications in Quality Inspection

Machine vision is a computer vision system that enables a computer to work on image-based inspection and analysis for different applications. In this computer vision, a camera and sensor were used to view an image for its analysis with the help of some sort of algorithms, then processed to infer the image-based data. Machine vision systems along with Python programs can be used for many interdisciplinary applications like weld inspection, online monitoring in manufacturing auto components etc. In this study, the “Edge detection python algorithm” was developed and run through “Google Colab” notebook to inspect the edges and the boundaries of samples like faying surface-modified friction welded dissimilar joints and a binder clip (paper clamp) to check any defects or cracks and straightness etc. With the help of this Python algorithm, the edge detection was done by Sobel, Scharr, and Prewit operators. An input image of the weld joint and the binder clip were converted into Otsu’s binary threshold image. The matrix vision camera and the CMOS sensor were used in the machine vision set-up to take the images. This written algorithm is helpful to trace the edges of any kind of solids components. The edges of the binder clips and the weld joint/zone were detected. The binder clips were inspected under two different cases namely the clip in folding condition (Case I) and the binder clip in unfolding condition (Case II). The results showed a defect that was identified in the weld zone and no bending was in the binder clips. This kind of study is useful in manufacturing industries for quality inspection purposes with a new machine vision set up for online inspection of fabricated components like nuts and bolts etc.

Addressing the inspection selection challenges of in-service pipeline girth weld using ensemble tree models

Natural gas transportation predominantly utilizes pipelines, and the safety of these systems largely hinges on the integrity of girth welds. However, once pipelines are buried post-construction, pinpointing weld inspection sites becomes complex. Presently, weld risk assessments rely on costly and specialized inspections. Hence, enhancing the accuracy in identifying defective welds and discovering superior inspection techniques is paramount. Our research employed machine learning, specifically an ensemble model of RandomForest and CatBoost, to prioritize weld inspections. After analyzing the West-East Gas Pipeline Girth Weld Dataset and addressing data imbalances using the SMOTE technique, we refined the model parameters. The culminating model yielded an F1-score of 0.815 and an average accuracy of 0.836, outperforming standalone models. Employing the SHAP method improved the model’s transparency, facilitating a more informed decision-making process regarding pipeline inspections. This study underscores the potential of machine learning in elevating oil and gas pipeline safety measures.

Data fusion of multi-view plane wave imaging for nozzle weld inspection

Ultrasonic non-destructive testing of nozzle-to-pipe welds has always been a challenge due to the complex saddle-shaped geometry of this junction. Inspection from the nozzle side can minimize geometry effects, however, conventional ultrasonic phased array methods suffer from artifacts and inadequate resolution. This paper proposes a novel nozzle-side inspection method based on the data fusion of multi-view plane wave imaging (PWI). A semi-analytic ultrasonic measurement model is derived, and a sensitivity mapping algorithm is presented for optimizing the inspection setup and enabling data fusion. Experiments were conducted on reference blocks and a nozzle specimen containing three embedded defects. Results demonstrate the capability of sensitivity mappings to not only optimize the inspection configuration, but also serve as filters to fuse multi-view images. Nozzle sidewalls were positioned in fusion images, evidencing the ability to deduce probe position around nozzle circumference without encoder feedback. A 10 dB amplitude difference was observed between the tip diffraction and specular reflection, enabling the defect characterization along with position and sizing accuracy within 0.59 mm and angle deviation below 9°. Slag, crack and lack of penetration were also characterized in the nozzle specimen, validating the proposed method.

Defect identification method for ultrasonic inspection of pipeline welds based on cross-modal zero-shot learning

Ultrasonic inspection of pipeline welds still uses the traditional visual inspection signal method to identify pipeline defects. The identification of defects relies entirely on the subjective judgment of practitioners and is highly dependent on their level of experience. Deep learning models have achieved very good results in classification tasks, but they rely on a large number of annotated data samples for each category. However, it is difficult to collect a large number of samples with different defects and annotate them for the classification of pipe welding defects. Based on the idea of zero-shot learning (ZSL), which makes full use of experts’ semantic descriptions of defect categories, artificial semantic features are integrated cross-modally with ultrasonic inspection signal features. In this way, a common semantic space containing seen and unseen classes is constructed to achieve the detection of various defects. Meanwhile, to alleviate the problem of extreme imbalance of training data between the seen and unseen classes in ZSL model training, a ZSL model Feature-GAN-ZSL (FGZ) fused with a generative adversarial network (GAN) is proposed. The model utilizes a Feature-GAN network to generate unseen class features during training and adds a classifier to enhance the generation of features with stronger discriminative power. In the experiments, sample data for porosity, incomplete penetration, and cracks were used as visible classes, and samples for incomplete fusion and slag entrapment were used as unseen classes. Five state-of-the-art models in the ZSL domain were compared. The results show that the FGZ model has a good ability to recognize various defects, not only the types of defects that participated in the training but also the defects that did not participate in the training. This plays a perfect role in dealing with various pipeline welding defects.

Deep Network-Assisted Quality Inspection of Laser Welding on Power Battery.

Reliable quality control of laser welding on power batteries is an important issue due to random interference in the production process. In this paper, a quality inspection framework based on a two-branch network and conventional image processing is proposed to predict welding quality while outputting corresponding parameter information. The two-branch network consists of a segmentation network and a classification network, which alleviates the problem of large training sample size requirements for deep learning by sharing feature representations among two related tasks. Moreover, coordinate attention is introduced into feature learning modules of the network to effectively capture the subtle features of defective welds. Finally, a post-processing method based on the Hough transform is used to extract the information of the segmented weld region. Extensive experiments demonstrate that the proposed model can achieve a significant classification performance on the dataset collected on an actual production line. This study provides a valuable reference for an intelligent quality inspection system in the power battery manufacturing industry.

ANALISA HASIL PENGELASAN PADA POSISI 1G BAJA ST37 DENGAN METODE FAULT TREE ANALYSIS

Welding is an important process in the manufacturing, metal and machinery industries, the welding results of students in the 1G Steel ST7 position are studied with the aim of identifying potential weld defects, determining root causes, evaluating and developing repair strategies to improve overall welding quality. The methods used in this research are qualitative and quantitative methods. Measuring welding dimensions with visual weld inspection, as well as liquid penetrant testing to identify weld defects. The data is processed using a Pareto diagram to determine the dominant types of defects that occur in the welding process. Once the dominant defect type is known, analysis is then carried out using the Fault Tree Analysis method. The results of the liquid penetrant test show that there are 8 types of weld defects with a total of 62 defects. In the FTA, basic events have been identified that contribute to the occurrence of weld defects, namely factors such as welding speed instability, incorrect current settings, and moist electrode. Regular evaluation, continuous monitoring and supervision during the welding process to reduce the risk of defects and improve welding quality

Noncontact laser ultrasonic inspection of weld defect in lithium-ion battery cap

The continuous development of electric vehicles and electronic devices has increased the demand for lithium-ion batteries. In this study, a laser ultrasonic inspection system was developed for the noncontact and nondestructive inspection of the laser welding conditions of a cylindrical lithium-ion battery cap. An Nd: YAG pulse laser was used for Lamb wave generation on the battery cap. Lamb waves were measured from the front surface of the battery cap using a laser Doppler vibrometer. The bending stiffness of the battery cap decreased as the welding condition deteriorated, and the Lamb wave amplitude was amplified. Accordingly, an automated defect detection algorithm for inspecting welding conditions was developed. The effectiveness of the proposed inspection system was theoretically and experimentally evaluated using actual cylindrical battery caps. The battery cap weld inspection accuracy exceeded 89 %, correctly identifying seven defective and ten intact battery caps out of eight defective and 11 intact battery caps, respectively. The numerical simulation indicated that the Lamb wave amplitude increased at the weld joint, demonstrating consistency with the experiment results. The proposed system is promising for the in-line inspection of battery caps because it facilitates high-speed, noncontact, and nondestructive inspection. Furthermore, the proposed inspection system can detect invisible weld defects on the back surface of a battery cap by inspecting its front surface.

Automated IBEX crawler for PAUT inspection for in-service ferromagnetic assets

In order to guarantee the safety of the gas production plants, Air Liquide conducts NDT inspections regularly closely following the assets’ maintenance plans. One high criticality asset is the Pressure Swing Adsorber (PSA), found in the last stages of the hydrogen production process, where product purification takes place. These adsorbers are pressure vessels and need to be inspected for hydrogen enhanced fatigue cracks. Conventional inspection would require stopping site operation, building scaffolding for the totality of the height, which depending on design can be up to 15 meters tall and 6 meters in diameter, grinding of the inspection surface to remove the paint, conducting the inspection manually by 2 operators, and repainting once inspections are concluded. Commonly, inspection plans for these high-criticality assets outline a high inspection frequency, resulting in elevated costs associated with logistics, personnel and time and monetary resources required to plan and execute each inspection. This necessity is what drove Air Liquide R&D and INTACT to create a partnership and develop an automated crawler that can be deployed on in-service ferromagnetic assets for Phased Array Ultrasonic Testing (PAUT) inspection, using Total Focusing Method (TFM) with INTACT proprietary imaging and software. The robot, named IBEX Crawler, is deployed at the lowest part of the vessel and controlled safely at the ground level by a single operator. It is able to change directions to accommodate the inspection of longitudinal and circular welds. It includes an encoding system that enables operators to exactly locate any indications present. The crawler’s payload includes an electronic 128 elements PAUT board controlled by fiber optics, laser tracking and machine vision camera. The inspection results are analyzed by 3D reconstruction and managed by mechanical integrity assessments proprietary technology, calculating the stability of the found indications and using fracture mechanics and Fitness-For-Service (FFS) to estimate the residual lifetime of the indications and in consequence, the interval between inspections to develop a more accurate maintenance plan. Today INTACT’s IBEX crawlers travel worldwide between Air Liquide sites to ensure the integrity of the assets.

Virtual round robin 2 – Phased array inspection of dissimilar metal welds

Round-robin exercises have traditionally been laborious to arrange in non-destructive testing (NDT). The exercises have involved manufacturing of costly big mock-ups and then distributing them around the world to facilitate testing by numerous laboratories. This has limited both the number of such round robins and their scope. Often the round robins have contained small number of flaws and the representativeness of these flaws has been limited. Nevertheless, the few round robins that have been completed have yielded significant additional understanding on the capability of the used NDT methods and procedures.Recently, the increased use of automated inspections together with the development of virtual flaws (independently by Trueflaw and EPRI) has enabled a new type of round robin, where instead of moving samples around the world, the round robin is focused on the data analysis and only pre-acquired data files are distributed. In 2019–2020, first of a kind virtual round robin (VRR) was completed. The round-robin allowed for the first time to compare inspection performance from teams around the world with statistically significant number of flaws and with ultrasonic data representative for nuclear dissimilar metal weld inspection. The study resulted in important new insight into NDE reliability for nuclear applications.However, as a first of a kind study, the first virtual round robin also contained some significant limitations. In particular, the data sets distributed were limited in order to limit the effort needed from each participating inspector. The reduced amount of data acquired was compensated by using pre-optimized data gathering, possible only with prior knowledge on the flaws present. While these choices were well justified for the first round-robin, they also made direct comparison of VRR results and real-life inspector performance problematic. In addition, the first VRR focused primarily on flaw detection and the data was insufficient for sizing.To address these shortcomings of the first round robin, a second round robin was completed in 2021–2022. In this second round robin, more representative data was used for evaluation. In addition, increased emphasis on the hard-to-detect small flaws was put forward to get improved into detectability especially in the low end.The more representative data required much more significant effort from the inspectors, which reduced the participation as compared to the first round robin. Furthermore, the emphasis on difficult-to-detect cracks may have further deterred participation, as the exercise may have been seen as too challenging. While the number of downloaded data sets (23) was similar to previous exercise, the number of returned sets was reduced to 5, compared to previous 18. Despite the smaller than expected participation, the results revealed several interesting features. The results displayed marked variation. Also, the false call rate was significantly reduced, as compared to the previous study. This could be attributed to the more rich data set, which allowed more comprehensive evaluation and exclusion of potential false calls.The recent advances in machine learning (ML) for ultrasonics also introduced an interesting opportunity to compare machine learning results with the human inspectors. Developing an optimized machine learning model for the present data was outside the scope of this study. Instead, an independently developed model, if somewhat sub-optimal, was used. Thus, the results should not be taken as a measure of ML performance as such. Nevertheless, the comparison between human results and ML model are informative and illustrate the potential benefits of automated data evaluation.

A multiple scale spaces empowered approach for welding radiographic image defect segmentation

Welding radiographic image defect segmentation (WRIDS) is a key technology to promote the automation and standardization of quality inspection. However, the complexity of scale variability, aggregation and contextual relationships presented by welding defects pose a great challenge to WRIDS, such as porosity, slag and lack of penetration. To address the issue, a multiple scale spaces (MSSs) empowered segmentation method of complex welding defects is proposed. First, a multi-scale feature space is constructed by dilated convolution with different dilated rates. Second, a multi-scale semantic space is constructed by max pooling in different windows. Third, a multi-scale relational space is constructed through a self-attention mechanism. Finally, the proposed MSSs method is validated on the basis of own collected weld inspection data from an aerospace structural component. The results show that the proposed method can effectively segment multiple complex defect types with an average Acc of 99.46% and an average Miou of 74.65% on the test set.

Development of HOIS guidance for ultrasonic NDT for non-intrusive inspection at elevated temperatures

The non-intrusive inspection (NII) of pressure vessels using external in-service ultrasonic NDT has multiple advantages compared to internal visual inspection, including the significant reduction of down-time related costs and safety benefits from avoiding manentry into vessels. The HOIS joint industry project, with support from the Net Zero Technology Centre (NZTC), has recently been focussing on NII of components with operating temperatures of up to 250°C (typical of upstream applications). These inspections are challenging because they require probes to be in contact with the hot surface much longer than for spot thickness measurements. The aim of the work has been to develop recommendations for in-service inspection of components at relatively high operating temperatures, thereby extending the benefits of the NII process, including avoiding or deferring internal visual inspection, to a wider range of vessels. Experimental trials have been performed on test components at up to ~250°C, with an emphasis on scanning ultrasonic systems used for corrosion mapping and weld inspection using TOFD and angled-beam phased-array (PA). The most successful trials gave corrosion mapping and TOFD weld examination results that had similar differences from benchmark values to those achieved in earlier ambient temperature trials. For angled-beam PA, it was found that some trials appeared to have issues with accurate correction of the data for the effects of the elevated component temperature on both the refracted beam angle and signal arrival times. Other trials were more successful in this respect. The lessons learned from trials on both ex-service and manufactured samples are leading to the development of specific guidance for NII at elevated temperatures. The paper highlights the technical challenges associated with ultrasonic scanning systems at elevated temperatures, share the lessons learned from the experimental trials and summarise the guidance developed.

Easy to go and innovative validation process using the spot weld inspection system PHAsis and related software

In the industry 4.0 the need for a cost-saving and efficient production process is of high significance. This includes every area of the production processes and leads to appropriate testing procedures by means of the quality management. A combination of ultrasonic inspection devices and software tools for result post processing reduces the need of several cost-intensive validation circles with destructive testing to a very low level. This can be shown for example with the imaging phased array spot weld inspection system PHAsis in combination with a software app for post processing. This is including the software part for the data and result management as well as a device software for the ultrasonic inspection of the spot welds. In combination with the usage of an automation interface for robotic usage, even more cost reduction is possible. Several validation processes on typical resistance spot welding applications with plate pairings of two or three plates (aluminium as well as steel combinations) by the use of PHAsis and the connected PostProcessor will be exemplified. The general procedure of the validation process will be accentuated step by step. Therefore, the original data of phased array testing and a comparative method in the shape of a destructive testing (hammer & chisel test or horizontal respectively vertical cross section) will be presented. Furthermore, the steps of the post processing procedure to define the ultrasound system parameters for a best fit will be described. During these steps, the influence of the different sensitivity and system parameters on the spot weld size will be explained. Also, the measurement uncertainties given by the destructive testing that have an influence to the comparation with ultrasonic testing in general will be addressed. The efficiency of the post processing software tool due to reaching a high match between PHAsis and the comparative destructive method will be illustrated.

Automated Spot Weld Testing using a Smart Robotic System

Since spot welding is still the most important joining method in the automotive industry, manual ultrasound testing of spot welds on car bodies and parts is a well-known and widely used application in this industry segment. Current trends in process automation and optimization, as well as current social topics, have given rise to an increasing interest in automizing the previously manual ultrasound inspection of spot welds, especially over the last couple of years. The first step in achieving this goal seems obvious – to integrate a conventional ultrasound device and transducer with a robot. However, there are many steps in the manual inspection process that are performed by an operator and subsequently need to be automated in a process-safe manner: positioning the ultrasound transducer correctly on the surface, recording and automatically evaluating the ultrasound data, and communicating in a fail-safe manner over an industrial network are only some aspects for which a solution must be found. This presentation will discuss the requirements for an ultrasonic system for automized spot weld testing, as well as present an already implemented smart robotic system for this purpose, which allows for both out-of-line as well as inline testing in series production.

Robot-based spot weld inspection - almost couplant-free, imaging phased array based inspection with PHAsis, integrated and automated by ABB Robotics

Constant innovations and rethinking of existing systems and processes are essential for the fulfilment of business goals of automotive groups - also for the field of quality assurance. In the context of quality assurance of spot welded joints, the basic tasks of inspection processes are to support the definition of welding parameters, the monitoring of welding results, the optimization of the inspection process and, as a consequence, the reduction of costs and production time. Rigid and manual processes, as they are still used in many areas of quality assurance, are based on laboratory conditions. They are built on general assumptions and are a generic procedure without the possibility of learning processes. If these processes developed in the laboratory encounter production realities, they reach their limits and waste valuable efficiency potential. For example, conventional inspection processes are performed randomly and compare the actual with the target state and, in the event of a deviation, only provide the information of an existing process error. However, this pure information is no longer sustainable for intelligent production and requires a function extension in the sense of Industry 4.0. The goal in the sense of Industry 4.0 in this context is to recognize the root causes of these detected deviations, to provide information in real time and the subsequent intelligent reaction of the system. Artificial intelligence provides a learning system which, based on continuous analysis of production parameters, can not only detect predictive deviations of flexible environmental conditions, but also in a further step autonomously takes action. An example is the detection of weld spatter or sharp-edged areas of spot welds and the subsequent initiated grinding before inspection. This presentation focuses on the development of a robot-controlled module for the almost coupling agent-free inspection of spot welds with innovatively applied phased array technology in production (inline and offline) according to the demands of Industry 4.0.

Merged Mode TFM with Mode Conversion Artifact Suppression

The objective of the paper is to compare the merits of the multiplication and max value schemes for producing a total focusing method (TFM) representation that combines multiple acoustic modes. We have chosen to compound phase coherence imaging (PCI) images because it eliminates the difficulties associated with handling vastly different acoustic sensitivities across different propagation modes. We also introduce an apodization-based algorithm that can filter out unwanted mode-conversion artifacts that are caused by similarities in time of flights. Experimental data were gathered on an electric resistance welding (ERW) hook crack and a natural bottom-surface-connected vertical crack. Experimental results indicate the multiplication scheme suppresses defect echoes that can only be imaged by a subset of acoustic modes. This can potentially lead to false negatives or underestimation of the extent of the defect. In contrast, the max value scheme retains all indications that appear in a subset of acoustic modes. For the vertical crack configuration imaged by shear wave acoustic modes, the max value scheme retains all mode-conversion artifacts and therefore complicates the defect analysis. The apodization-based filtering method significantly suppressed mode-conversion artifacts without affecting the imaging of the vertical crack. The combination of PCI with max value merging and apodization-based mode-conversion filtering is a promising methodology for merged mode TFM in shear wave weld inspections.