Beam Offset Detection in Laser Stake Welding of Tee Joints Using Machine Learning and Spectrometer Measurements (original) (raw)
Related papers
Beam offset detection in laser stake welding of tee joints based on photodetector sensing
Procedia Manufacturing, 2019
Under the concept of "Industry 4.0", production processes will be pushed to be increasingly interconnected, information based on a real time basis and, necessarily, much more efficient. In this context, capacity optimization goes beyond the traditional aim of capacity maximization, contributing also for organization's profitability and value. Indeed, lean management and continuous improvement approaches suggest capacity optimization instead of maximization. The study of capacity optimization and costing models is an important research topic that deserves contributions from both the practical and theoretical perspectives. This paper presents and discusses a mathematical model for capacity management based on different costing models (ABC and TDABC). A generic model has been developed and it was used to analyze idle capacity and to design strategies towards the maximization of organization's value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity optimization might hide operational inefficiency.
Vision based beam offset detection in laser stake welding of T-joints using a neural network
Procedia Manufacturing, 2019
Under the concept of "Industry 4.0", production processes will be pushed to be increasingly interconnected, information based on a real time basis and, necessarily, much more efficient. In this context, capacity optimization goes beyond the traditional aim of capacity maximization, contributing also for organization's profitability and value. Indeed, lean management and continuous improvement approaches suggest capacity optimization instead of maximization. The study of capacity optimization and costing models is an important research topic that deserves contributions from both the practical and theoretical perspectives. This paper presents and discusses a mathematical model for capacity management based on different costing models (ABC and TDABC). A generic model has been developed and it was used to analyze idle capacity and to design strategies towards the maximization of organization's value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity optimization might hide operational inefficiency.
Journal of Laser Applications, 2020
Laser beam welding significantly outperforms conventional joining techniques in terms of flexibility and productivity. The process benefits, in particular, from the highly focused energy and thus from a well-defined heat input. The high intensities of brilliant laser radiation, however, induce very dynamic effects and complex processes within the interaction zone. The high process dynamics require a consistent and reliable quality assurance to ensure the required weld quality. A novel sensor concept for laser material processing based on optical coherence tomography (OCT) was used to measure the capillary depth of the keyhole during deep penetration welding. The OCT measurements were compared with analyses of the surface quality of the weld seams. A machine learning approach could be utilized to reveal correlations between the weld depth signal and the weld seam surface quality, underlining the high level of information contained in the OCT signal about characteristic process phenomena that affect the weld seam quality. Fundamental investigations on aluminum, copper, and galvanized steel were carried out to analyze the structure of the data recorded by the OCT sensor. Based on that, evaluation strategies focusing on quality characteristics were developed and validated to enable a valid interpretation of the OCT signal. The topography of the weld seams was used to classify the surface quality and correlated with the weld depth signal of the OCT system. For this purpose, a preprocessing of the OCT data and a detailed analysis of the topographic information were developed. The processed data were correlated using artificial neural networks. It was shown that by using adequate network structures and training methods, the inline process data of the capillary depth can be used to predict the surface quality with decent prediction accuracy.
Scientific Reports
Laser welding is a key technology for many industrial applications. However, its online quality monitoring is an open issue due to the highly complex nature of the process. this work aims at enriching existing approaches in this field. We propose a method for real-time detection of process instabilities that can lead to defects. Hard X-ray radiography is used for the ground truth observations of the sub-surface events that are critical for the quality. A deep artificial neural network is applied to reveal the unique signatures of those events in wavelet spectrograms from the laser back-reflection and acoustic emission signals. The autonomous classification of the revealed signatures is tested on reallife data, while the real-time performance is reached by means of parallel computing. The confidence of the quality classification ranges between 71% and 99%, with a temporal resolution down to 2 ms and a computation time per classification task as low as 2 ms. This approach is a new paradigm in the digitization of industrial processes and can be exploited to provide feedbacks in a closed-loop quality control system.
Monitoring undercut, blowouts and root sagging during laser beam welding
Using a 14 kW CO2-laser, 12mm thick cold-formed steel S420MC has been welded to a machined shaft pivot made of 25CrMo4 steel as part of a truck rear axel. A photodiode-based, on-line process monitoring system has been applied for detecting defects. However, the occurrence of certain defects, namely undercut, blowouts and root sagging is often not detectable from the sensor signal. The time dependent signal is collected from emissions from the melt pool surface as well as from the plasma plume. Based on the evaluation of high speed images, an explanation of the potential and limitations of detection of these defects has been generated. Although preliminary conclusions have been drawn, uncertainties regarding emissivity and the keyhole and plasma radiation characteristics require further studies.
This study aimed to explore the feasibility of using airborne acoustic emission in laser beam butt welding for the development of an automated classification system based on neural networks. The focus was on monitoring the formation of joint gaps during the welding process. To simulate various sizes of butt joint gaps, controlled welding experiments were conducted, and the emitted acoustic signals were captured using audible to ultrasonic microphones. To implement an automated monitoring system, a method based on short-time Fourier transformation was developed to extract audio features, and a convolutional neural network architecture with data augmentation was utilized. The results demonstrated that this non-destructive and non-invasive approach was highly effective in detecting joint gap formations, achieving an accuracy of 98%. Furthermore, the system exhibited promising potential for low latency monitoring of the welding process. The classification accuracy for various gap sizes ...
Vision and spectroscopic sensing for joint tracing in narrow gap laser butt welding
Optics and Laser Technology, 2017
The automated laser beam butt welding process is sensitive to positioning the laser beam with respect to the joint because a small offset may result in detrimental lack of sidewall fusion. This problem is even more pronounced in case of narrow gap butt welding, where most of the commercial automatic joint tracing systems fail to detect the exact position and size of the gap. In this work, a dual vision and spectroscopic sensing approach is proposed to trace narrow gap butt joints during laser welding. The system consists of a camera with suitable illumination and matched optical filters and a fast miniature spectrometer. An image processing algorithm of the camera recordings has been developed in order to estimate the laser spot position relative to the joint position. The spectral emissions from the laser induced plasma plume have been acquired by the spectrometer, and based on the measurements of the intensities of selected lines of the spectrum, the electron temperature signal has been calculated and correlated to variations of process conditions. The individual performances of these two systems have been experimentally investigated and evaluated offline by data from several welding experiments, where artificial abrupt as well as gradual deviations of the laser beam out of the joint were produced. Results indicate that a combination of the information provided by the vision and spectroscopic systems is beneficial for development of a hybrid sensing system for joint tracing.
A Multi-Sensor Data Fusion System for Laser Welding Process Monitoring
IEEE Access
Most existing laser welding process monitoring (LWPM) technologies focus on detecting post-process defects. However, in sheet metal laser welding applications such as welding of electronic consumer products during mass production, in-process defect detection is more important. In this article, a compact LWPM system using multi-sensor data fusion to detect in-process defects has been built. This system can collect the time series of plasma intensity, light intensity and temperature data for feature analysis. To verify the system's effectiveness, a plasma-light-temperature dataset has been compiled, which consists of 5,836 samples of nine classes, including one positive class and eight negative classes of typical in-process defects. A multi-sensor data fusion network based on a convolution neural network for in-process defect detection, called IDDNet, has also been proposed. Experimental results have demonstrated that IDDNet can achieve better multi-classification results than the support vector machine, with an overall accuracy of 97.57%. In particular, considering this monitoring process as a binary classification problem, IDDNet can achieve a 99.42% accuracy. Moreover, IDDNet can reach an average speed of 0.79ms per sample on a single GTX 1080ti graphics card, which meets the real-time requirement for industrial production. The proposed LWPM system has been successfully verified in real applications of sheet metal laser welding. INDEX TERMS Laser welding process monitoring, in-process defect detection, multi-sensor data fusion, convolution neural network. FENGBIN XIA received the B.S. degree in material shaping and control engineering from Nanchang Hangkong University, Nanchang, China, in 2010, and the M.S. degree from the National Key Laboratory of Preparation and Processing of Nonferrous Metal Materials, Beijing General Research Institute for Nonferrous Metals, Beijing, China, in 2013. He currently engages in the applications and development of laser welding at Han's Laser Technology Industry Group Company, Ltd. His research interests include metal material preparation technologies, electron beam welding technologies, and material analysis and testing techniques.