Weld Quality Prediction Research Articles

Enhancing VPPA welding quality prediction: A hybrid model integrating prior physical knowledge and CNN analysis

In response to the inconsistency between the features obtained by deep learning models and the quality features reflected by the physical laws of the welding process, this study proposes a solution by integrating a physical prior information model with a CNN model. Initially, the physical laws of the welding process are utilized to annotate the arc, weld pool, and weld seam features relevant to quality, which are then acquired through image processing algorithms, thereby converting the physical laws into a prior information model. Subsequently, this prior information model guides the CNN model for quality recognition, and the CNN model's attention to features is explained through visualization methods to elucidate the relationship between features and quality recognition. Experimental results demonstrate that under the guidance of the prior information model, the CNN model not only automatically focuses on features relevant to quality but also achieves a differential feature attention strategy, thereby improving the recognition accuracy of different outcomes. This research provides a new perspective for deep learning in the field of welding quality recognition.

Optimization and prediction of resistance seam weld quality in secondary coated steels using machine learning

Optimization and prediction of resistance seam weld quality in secondary coated steels using machine learning

Federated Learning for Weld Quality Prediction

This study introduces a monitoring system that can accurately predict the quality of friction stir welds. The study involved four different sets of welding experiments using varying materials and tool configurations. The goal was to create a universal monitoring system capable of predicting weld quality across these diverse experimental sets. Real-time welding data was collected using instruments such as load cells and a power sensor. Signal processing methods were then used to analyze this data and extract essential information about welding quality. Subsequently, Federated Learning (FL) was used to develop the universal monitoring system. This method involves collective learning and model training, leading to a global model trained on decentralized data from the different experimental sets. The approach proved to be effective in forecasting weld quality, achieving a mean absolute error (MAE) of 10.44 for all welding experiments. Additionally, it offered benefits such as reduced latency and enhanced user data protection, while maintaining the accuracy of the global model. An artificial neural network (ANN) model was also developed for comparison with FL, achieving a MAE of 13.85 for the welding experiments. Overall, this study demonstrates the effectiveness of training global and more reliable models, with multiple devices sharing their knowledgebases to train the global model effectively. Conflict of Interest Statement The authors declare no conflicts of interest.

Prediction of Nugget Diameter and Analysis of Process Parameters of RSW with Machine Learning Based on Feature Fusion

The welding quality during welding body-in-white (BIW) determines the safety of automobiles. Due to the limitations of testing cost and cycle time, the prediction of welding quality has become an essential safety issue in the process of automobile production. Conventional prediction methods mainly consider the welding process parameters and ignore the material parameters, causing their results to be unrealistic. Upon identifying significant correlations between vehicle body materials, we utilize principal component analysis (PCA) to perform dimensionality reduction and extract the underlying principal components. Thereafter, we employ a greedy feature selection strategy to identify the most salient features. In this study, a welding quality prediction model integrating process parameters and material characteristics is proposed, following which the influence of material properties is analyzed. The model is verified based on actual production data, and the results show that the accuracy of the model is improved through integrating the production process characteristics and material characteristics. Moreover, the overfitting phenomenon can be effectively avoided in the prediction process.

Digital twin-based tig welding quality prediction using electrode tip angle degradation influencing Industry 5.0 in manufacturing sector

Automation in tungsten inert gas (TIG) welding is important to achieve high production rates and quality in manufacturing industries. To improve the welding process and quality inspection methodologies, the intelligent welding robot and vision-based inspection system have been researched and deployed in many engineering fields. Hence to enhance the performance and production, a digital twin-based welding system with the prediction of weld quality based on the consideration of electrode tip angle degradation. The proposed system will capture real-time electrode tip angle and weld pool temperature using a forward looking infrared (FLIR) camera along with welding current and speed correlated with tensile strength as the output parameter. To validate the analysis, support vector machine (SVM) and random forest (RF) algorithms were implemented in which the RF model performs well on the prediction of welding quality by mapping with tensile strength. RF model confirms maximum accuracy of 90% with 0.29 seconds computation time to perform prediction on the next execution of welding operation. It is inferred that if the tip angle degradation increases consecutively welding current decreases drastically impacting the weld quality from good to poor. To forecast the need for immediate or scheduled maintenance to reduce the tip angle degradation, a linear regression algorithm is implemented to enable the inspection engineer to perform maintenance without delay in production.

Prediction and optimization method for welding quality of components in ship construction

Welding process, as one of the crucial industrial technologies in ship construction, accounts for approximately 70% of the workload and costs account for approximately 40% of the total cost. The existing welding quality prediction methods have hypothetical premises and subjective factors, which cannot meet the dynamic control requirements of intelligent welding for processing quality. Aiming at the low efficiency of quality prediction problems poor timeliness and unpredictability of quality control in ship assembly-welding process, a data and model driven welding quality prediction method is proposed. Firstly, the influence factors of welding quality are analyzed and the correlation mechanism between process parameters and quality is determined. According to the analysis results, a stable and reliable data collection architecture is established. The elements of welding process monitoring are also determined based on the feature dimensionality reduction method. To improve the accuracy of welding quality prediction, the prediction model is constructed by fusing the adaptive simulated annealing, the particle swarm optimization, and the back propagation neural network algorithms. Finally, the effectiveness of the prediction method is verified through 74 sets of plate welding experiments, the prediction accuracy reaches over 90%.

Development of a fuzzy logic model for predicting the quality of micro friction stir spot welding (µFSSW) using particle swarm optimization

Micro Friction Stir Spot Welding (µFSSW) is crucial in microelectronics and precision manufacturing. It requires a comprehensive understanding of the complex connections between various parameters to achieve the highest quality welds. This study aims to improve the prediction of µFSSW weld quality by incorporating advanced optimization techniques. Fuzzy Logic Optimization is used to model uncertainties, and Particle Swarm Optimization (PSO) is employed to fine-tune parameters for improved accuracy. The fuzzy logic system utilizes Gaussian functions as membership functions, organized with nine rule bases. The results clearly demonstrate that the fuzzy logic model greatly enhances accuracy when combined with Particle Swarm Optimization. The refined model improves precision for pin diameter, shoulder diameter, Thermo-Mechanically Affected Zone (TMAZ) area, and cross-tensile strength. The PSO-optimized model shows lower accuracy in predicting plunge depth and shear tensile strength. The ongoing decline in Root Mean Square Error (RMSE) values highlights the complexity of the results. The optimization significantly improves the model’s ability to predict specific weld quality metrics, as demonstrated by the pin diameter’s reduced RMSE value of 0.07. The collective results showcase an optimized Fuzzy Logic System (FLS) model adept at accurately predicting µFSSW weld quality, demonstrating adaptability across diverse conditions. The discernible increase in accuracy, reaching up to 76 % following the optimization of the fuzzy logic model with PSO, serves as a testament to the efficacy of the employed methodologies in advancing the precision and reliability of µFSSW weld quality predictions

A review of quality monitoring for ultrasonic metal welding

Ultrasonic metal welding, an emerging solid state bonding method, has drawn extensive attention and been applied in various manufacturing scenarios in recent years. This paper reviews the quality monitoring system of ultrasonic metal welding from in-situ testing (online monitoring), non-in-situ testing and prediction of ultrasonic weld quality through machine learning methods. In-situ testing focuses on the acquisition of different process parameters and their relationship to weld quality, while non-in-situ testing focus on the monitoring indicators of the weld quality and the testing means. The exploration of machine learning methods concentrates on the influence of model selection and parameter input on prediction accuracy. Based on the analysis, the future development trend of ultrasonic weld quality monitoring is provided.

Application of a concept for ML-driven closed-loop quality control in laser beam welding

Laser beam welding is valued for the capability of contactless welding. Despite significant technological improvements to achieve stable processes and the required quality standards, quality determination is often done destructively and offline, with unavoidable expenses. Inline approaches address this problem by using and processing collected machine and sensor data to enable indirect conclusions about the welding quality in the process. This paper presents an inline approach that merges a set of complementary sensor signals to enable the prediction of weld quality by machine learning algorithms, paving the way for closed-loop quality control.

A Novel Period-Sensitive LSTM for Laser Welding Quality Prediction

A Novel Period-Sensitive LSTM for Laser Welding Quality Prediction

A novel surface temperature sensor and random forest-based welding quality prediction model

A novel surface temperature sensor and random forest-based welding quality prediction model

Development of an Arc Image Classification Algorithm in Tip-Rotating Arc Welding using Camera Image Monitoring and Deep Learning

In this study, we developed a welding system and welding monitoring system for GMA (Gas Metal Arc) aluminum welding using a tip-rotation arc welding torch. Welding was performed under various conditions, including suitable, low-heat, and high-heat input conditions. For welding monitoring, current and voltage were measured, and arc images were obtained. After the experiments, arc images were analyzed and classified based on the bead quality and cross-sectional analysis data. To develop a welding quality prediction model using arc images, a CNN-based model was developed. A total of 5,203 images were trained, and the quality was predicted for 631 images. The prediction accuracy was 99.88% for training data, and the possibility of assessing the quality of welded joints based on arc images during welding was confirmed.

Implementation of Machine Learning Algorithms for Weld Quality Prediction and Optimization in Resistance Spot Welding

Implementation of Machine Learning Algorithms for Weld Quality Prediction and Optimization in Resistance Spot Welding

Monitoring of friction stirred spot weld quality for dissimilar Al6061 to polycarbonate using tool assisted thrust-torque signatures

Friction stir processing procedures are often found to be an alternative choice for the joining of thermoplastic polymers without melting the base materials. In recent years, dissimilar metallic to thermoplastic assemblies are highly used in aerospace and automotive industries as mechanical fastener or fusion welding techniques are rather inadequate for extreme discrepancy in thermal properties. The present work addresses the feasibility of friction stirred spot welding for non-ferrous aluminium (Al 6061) to polycarbonate using taper cylindrical tool of hardened H13 steel along with machine integrated axial thrust and stirring torque signals. The real-time positional and axial velocity of the rotating tool has been acquired to distinguish between plunging and dwelling stage. The process stability that is, the deviation of thrust-torque in respective phases was processed. The interaction of plunge depth and dwell time on tool pin assisted keyhole with associated material expulsion for different tool revolving speed was established. The weld mechanical behaviour has been scanned using load versus elongation diagram with associated weld pair interface. Lastly, sensors’ based process monitoring extended to weld quality prediction have been made using various significant features of thrust-torque-velocity based approaches. The tool spinning speed followed by dwell time was found to be important on weld quality index. The joint efficiency was significantly upgraded (more than 60%) at intermediate dwell time (40 s) with low tool plunging (0.3 mm) or reversed parametric setting at high tool revolving speed (2500 rpm) due to substantial reduction in axial thrust with higher stirring torque. The local deviation in thrust-torque during plunging was equally responsible with global mean thrust-torque in dwelling to dictate the joint strength and the weld ductility. The material stirring torque was better indicator of weld shear strength while pin assisted slot size highly correlated with tool thrust though tool plunging speed was highly crucial.

Optimization and Improvement of the Projection Welding of Nut Based on Regression Analysis

Since it has the advantages of high production efficiency, no shunting effects, and the ability to arrange solder joints on narrow locations without the limitation of point distance, projection welding of nut has been widely used in the automotive industry. This paper presents a new welding quality prediction and analysis framework by using regression analysis techniques. The relationship between the welding parameters (welding time, welding current and electrode force) and the Pull-out load of the M6 welded square nut and the 1.5 mm thick pickled hot-rolled steel was obtained by multiple nonlinear regression analysis. The influence and interaction of welding parameters on the Pull-out load are discussed. Using this regression model, a better combination of process parameters can be obtained. The mechanical properties, microstructure and microhardness of welded joints were also investigated.

Comparison and explanation of data-driven modeling for weld quality prediction in resistance spot welding

Comparison and explanation of data-driven modeling for weld quality prediction in resistance spot welding

Interpreting learning models in manufacturing processes: Towards explainable AI methods to improve trust in classifier predictions

Smart manufacturing processes, building upon machine learning (ML) models could potentially reduce the pre-production testing and validation time for new processes. Beyond calculating accurate and reliable models, one critical challenge would be for users of these models (plant operators, engineers and technicians) to trust these models’ outputs. We propose to apply explainable AI methods to create trustworthy AI-based manufacturing systems. Consequently, these systems will be enriched with capabilities to explain their reasoning processes and outputs (e.g., predictions) automatically. This paper applies explainable AI methods to two problems in manufacturing: ultrasonic weld (USW) quality prediction and body-in-white (BIW) dimensional variability reduction. Class activation maps were computed to explain the effect of input signals and their patterns on the quality predictions of an ultrasonic weld yield by a neural network (good or bad). Contrastive gradient based saliency maps were also created to assess the robustness of this classifier. Furthermore, we explain a given connectionist network that predicts the dimensional quality of body-in-white framer points based on deviations in underbody points. Explaining these predictions help engineers understand which underbody points have more influence on deviations in the framer points. These two applications highlight the importance of explainable AI methods in the modern manufacturing industry.

Machine learning-based algorithm for the classification and prediction of multi-type spot weld quality

Machine learning-based algorithm for the classification and prediction of multi-type spot weld quality

Quality analysis for condition monitoring of a manufacturing process integrating acoustic processing based on three distinct ML algorithms

Abstract Over the years, it has been proven that all machines that run continuously or even for a short period of time experiences breakdowns, reduced efficiency and damage to the parts. Considering one such manufacturing process is welding process where strip welding is studied for condition monitoring and prediction of online quality of the weld. The proposed embodiment discusses a predictive condition monitoring of a welding process, using acoustic sensors and processing their data. Converting the processed data to a predictive model, comparing different Machine Learning (ML) algorithms like Decision Tree (DT), Artificial Neural Networks (ANN) and Convolutional Neural Networks (CNN) based on the validation and training results, selecting the suitable algorithm for online or real-time weld quality prediction is the objective of the proposed embodiment. According to the research, effectively executing the proposed solution can improve its life, efficiency, and utilization. The training accuracy is observed to be 97 % for DT, 50 % for ANN and 67 % for CNN. DT, on the other hand, has a validation accuracy of 94.7 %, ANN has 47.65 % and CNN has 65.8 %. DT produced the best results in the study, initially DT prediction model is used in offline testing of the new data-set for individual weld class and the system produced 97 % accuracy. Furthermore the DT prediction model is implemented in online weld quality prediction in three classes: Bad weld, OK weld and Good weld where the system produced 90 % accuracy in real-time prediction.

Quality Prediction and Parameter Optimisation of Resistance Spot Welding Using Machine Learning

In a small sample welding test space, and to achieve online prediction and self-optimisation of process parameters for the resistance welding joint quality of power lithium battery packs, this paper proposes a welding quality prediction model. The model combines a chaos game optimisation algorithm (CGO) with the multi-output least-squares support vector regression machine (MLSSVR), and a multi-objective process parameter optimisation method based on a particle swarm algorithm. First, the MLSSVR model was constructed, and a hyperparameter optimisation strategy based on CGO was designed. Next, the welding quality was predicted using the CGO–MLSSVR prediction model. Finally, the particle swarm algorithm (PSO) was used to obtain the optimal welding process parameters. The experimental results show that the CGO–MLSSVR prediction model can effectively predict the positive and negative electrode nugget diameters, and tensile shear loads, with root mean square errors of 0.024, 0.039, and 5.379, respectively, which is better than similar methods. The average relative error in weld quality for the optimal welding process parameters is within 4%, and the proposed method has a good application value in the resistance spot welding of power lithium battery packs.