Laser Keyhole Welding Process Research Articles

In-situ monitoring system for weld geometry of laser welding based on multi-task convolutional neural network model

This paper presents a low-cost, robust, in-situ monitoring system for weld geometry that can achieve multi-task prediction. First, the system uses a low-cost CCD camera to monitor the melt pool in the laser keyhole welding process. Then, the proposed novel multi-task convolutional neural network (Multi-task CNN) model is used to simultaneously complete the two prediction tasks of weld depth and width. Furthermore, the learning process of the Multi-task CNN model is explored using a visual feature map approach and the robustness of the model is demonstrated. Compared with Support Vector Machine, K-Nearest Neighbor, Bayesian Ridge, Decision Tree, the proposed Multi-task CNN model has the highest prediction accuracy. The model predicts a mean absolute percentage error (MAPE, relative to ground truth) of 3.0% and 1.9% for weld depth and width. The in-situ monitoring results show that the system can achieve accurate predictions, and the average time-consuming of the system is 23.35 ms.

Influence of Magnesium on Spatter Behavior in Laser Deep Penetration Welding of Aluminum Alloys

The quality of welds, as well as the necessity of post-processing, is challenged by spatter generation during the laser keyhole welding process. In this study, the influence of the magnesium content on spatter behavior is studied for three aluminum alloys (Al99.5, AlMg3, and AlMg5). A synchronized dual high-speed camera system is used to observe the spatter behavior and to reconstruct 3D spatter trajectories as well as determine the characteristics of spatter velocity, flight path angle, and approximate spatter size. The mean spatter velocities and flight path angles of the welding experiments with the three alloys were in welding direction between 4.1 m/s and 4.6 m/s and 44.8° and 51.0°, respectively. Furthermore, the AlMg alloys show excessive spatter behavior with spray events of more than 50 spatters at a time, and less frequently spatter explosions. Spatter spray events show a character similar to spatter explosions. Volumetric evaporation is proposed as effecting these events. In contrast, and resulting from a different mechanism, pure aluminum (Al99.5) shows group ejection events with at least 10 spatters at a time. In this study, there are no correlations between spatter velocities and flight path angles, nor between velocities and approximate spatter sizes.

Prediction of angular distortion in the fiber laser keyhole welding process based on a variable-fidelity approximation modeling approach

The angular distortion is one of the most common types of distortions frequently observed in laser weld assembling processes, which leads to a decline in welding joints’ quality and additional costs of rework. Therefore, it is of great importance to control and reduce the welding-induced angular distortion by selecting appropriate welding process parameters. The challenge is how to predict the welding-induced angular distortion in the whole process parameter design domain accurately and efficiently. To address this challenge, a variable-fidelity approximation modeling approach is developed in this paper, where two different levels of fidelity data are integrated for predicting the angular distortion in the laser welding process. In the proposed approach, a three-dimensional thermo-mechanical finite element model is developed as a low-fidelity model, while the laser welding experiment is taken as a high-fidelity model. A low-fidelity radial basis function (RBF) model is constructed based on the sample data from the finite element simulation. Then a linear tuning strategy is introduced to bring the low-fidelity RBF model as close as possible to the data from the laser welding experiment. Finally, the variable-fidelity approximation model is constructed by adopting a scaling function to calibrate the remaining differences between the tuning low-fidelity approximation model and the high-fidelity data. Two types of validation approaches are adopted to compare the prediction accuracy of the variable-fidelity approximation model with those of the single-fidelity approximation models solely constructed with laser welding experiment or finite element simulation. Results illustrate that the prediction ability of the developed variable-fidelity approximation model outperforms those of the single-fidelity approximation models.

Modeling and robust controlling of laser welding process on high strength titanium alloy using fuzzy basis function networks and robust Takagi-Sugeno fuzzy controller

The purpose of this paper is to present and implement a novel approach to data-driven modeling and robust control of a laser keyhole welding process. The objective is to maintain the penetration depth of the laser welding process at a desired value under system uncertainties. A framework was proposed to estimate the keyhole diameter and the keyhole penetration depth in real time by type-1 and type-2 fuzzy basis function networks, and an adaptive divided difference filter. A robust Takagi-Sugeno fuzzy controller was implemented to adjust the laser power of the welding process. Experimental results demonstrated that the fuzzy models provided an accurate estimation of both the welding geometry and its variations due to uncertainties, and the robust Takagi-Sugeno fuzzy controller successfully reduced the penetration depth variation and improved the quality of the welding process.

Analysis of microstructure and mechanical properties change in laser welding of Ti6Al4V with a multiphysics prediction model

This paper reports on a study involving examination of the geometry, microstructure and microhardness of welds produced on a 1.625mm thick Ti6Al4V titanium alloy using the IPG YLR-1000 fiber laser. A numerical model has been developed to predict the fusion zone (FZ), heat affected zone (HAZ) and microhardness of the weld. The thermal history from the multi-physics model is incorporated into a two-dimensional phase prediction model to predict the α′ martensite formation in the heat-affected zone and fusion zone. The multi-physics phase prediction model is applied to simulate the laser keyhole welding process with varying laser power and welding speed. It was observed that the weld geometry and microhardness predictions from the multi-physics model were in good agreement with the experimental data. The results show that the peak hardness consistently appear at the FZ and HAZ boundaries due to the high cooling rate. In addition, it has been found that laser welding speed has a profound influence on acicular α′ martensite formation leading to hardness change. Consequently, to minimize the weld hardening, use of lower power and lower welding speed parameters is recommended in the power range of 800–1000W and welding speed range of 1.73–4m/min.

A study on ray tracing method for CFD simulations of laser keyhole welding: progressive search method

In the laser keyhole welding process, one ray can have several reflections and absorptions during the process. This is called multiple-reflection and is considered to be one of the important physical phenomena in the laser keyhole welding process. To calculate the multiple-reflection in laser keyhole welding simulations, a ‘Ray tracing’ methodology is used. Through ray tracing, several ray properties, such as the current location of a ray, the location of reflection, and the reflected direction, can be calculated. This study investigated the numerical simulations of two ray tracing methods, the Direct Search Method (DSM) and Progressive Search Method (PSM). There are differences between the two ray tracing methods. PSM uses a real equation for the surface and ray direction vector while DSM uses a discriminant. A reflected ray is moved step by step in PSM, but only once by the discriminant in DSM. The reflected ray starts from the real contact point in PSM, but from the center of the contacted sub-cell in DSM. PSM can depict the multi-path of the ray by employing a set of sub-rays concept. On the other hand, DSM cannot depict the multi-path of the ray, because it uses a single ray definition. Therefore, PSM is a more physical numerical method and can depict various phenomena, such as transmission and scattering. PSM has a higher accuracy in laser keyhole welding simulations. In PSM, two new models were applied, the transmission and scattering models. All simulations were performed with the computational fluid dynamics (CFD) method including the volume of fluid (VOF) technique. Simulations using the DSM and PSM were performed to compare these two methods.

Multi-scale modeling of solidification and microstructure development in laser keyhole welding process for austenitic stainless steel

A multi-scale model is developed in this paper to investigate the growth of grains and sub-grain dendrites during the molten pool solidification of austenitic stainless steel AISI304 in laser keyhole welding processes. A macro-scale dynamic model is used to predict the fluid flow and heat transfer in the keyhole as well as the molten pool. The generated thermal history is fed to a three-dimensional Cellular Automata model to predict the meso-scale grain growth and to a two-dimensional Cellular Automata–Phase Field model to predict the micro-scale dendrite morphology. The multi-scale model is applied to simulate multiple cases of different welding parameters. The model predictions are validated against experiments and studied further for detailed information. The effects of the welding parameters are discussed based on the numerical and experimental results.

X-Ray and Optical Videography for 3D Measurement of Capillary and Melt Pool Geometry in Laser Welding

Abstract This paper describes a method to reconstruct the 3D shape of the melt pool and the capillary of a laser keyhole welding process. Three different diagnostic methods, including X-Ray and optical videography as well as metallographic cross sections are combined to gain the three dimensional data of the solidus-liquidus-surface. A detailed description of the experimental setup and a discussion of different methods to combine the 2D data sets of the three different diagnostic methods to a 3D-model will be given. The result will be a static 3D description of the welding process.

Laser deep penetration welding simulation based on a wavelength dependent absorption model

Today laser sources for laser material processing are available at different wavelengths. In this paper it is depicted that the wavelength dependent Fresnel absorption of the laser at the keyhole front is responsible for a different propagation and amplification or attenuation of instabilities at the keyhole front and the keyhole dynamics. To understand the physics of laser deep penetration welding and the interaction of the melt pool dynamics with these instabilities a three dimensional transient simulation of laser keyhole welding process has been developed. The results obtained by this simulation help to understand the different aspects of the formation of instabilities in the keyhole front at different wavelengths. To investigate the implication of the wavelength dependent absorption effects a CO2-laser and an Yb:YAG-laser are simulated. The presented model includes the calculation of the heat flux in the solid and molten phase and a temperature dependent melting and solidification model. The 3D velocity field inside the melt pool and the keyhole is calculated. The free surface of the melt pool is modelled using the Volume of Fluid method.

Experimental observation of keyhole shapes in the laser welding of aluminum blanks

A camera-based in situ monitoring system for the Nd:yttrium–aluminum–garnet laser keyhole welding process of aluminium has been developed. The use of an external illumination source for image formation decouples the performance of this system from many process parameters. A prototype of the monitoring system was used to investigate two different welding modes, which have been observed during the laser welding process of AA5182 alloy sheets. One welding mode results in a circular keyhole shape, while the second mode has a more elliptical shape. Welds resulting from both welding modes also have a different surface structure at the top and bottom weld surface. Mechanical tests showed a marginal difference in mechanical properties of both welding types. However, the welding mode with an elliptical keyhole shape is more susceptible to holes in the weld seam. Experiments showed that this welding process can be visualized well with the developed monitoring system and that both welding modes can be distinguished clearly.