Despite the steadily improving methods for process diagnostics in laser beam welding the fluid dynamics within the weld pool responsible for defects like pore formation, humping or spiking are not understood sufficiently. The multiphysical simulation model presented within this contribution is not only a powerful tool to reliably predict processing results like weld depth and width or temperature distributions but also helps to reveal the reasons for the above mentioned welding defects.Reliable weather forecasts based on multiphysical simulations are state of the art today. These simulations include many physical phenomena like e.g. the heating and evaporation of water induced by the incidence of the solar radiation. Computational fluid dynamics are used to calculate pressure and flow fields and thus to predict e.g. local wind forces and the development of low-pressure areas. Furthermore, the geometry of the Earth's surface has to be considered for taking into account the influence of e.g. mountains on the weather development.From the physical point of view it is a very similar task to simulate laser material processing. Instead of solar radiation we have got laser radiation that induces evaporation and the Earth's surface has to be replaced by the geometry of the work piece. However, the basic physics behind all these phenomena keeps the same and is in principle (perhaps besides some effects on very short time scales) well known. Nevertheless, the development and use of reliable simulation tools for laser material processing is still in an early stage compared to weather forecasting models.In the following a general model for laser material processing will be presented that will prospectively help both scientific and industrial users to predict and understand processing results even without having the laser machine or conducting experiments (Fig. 1). The model will be verified on basis of experimental results from laser beam welding of aluminum and copper. Finally, more insight into the process dynamics will be given and the reasons for typical welding failures like humping and spiking will be explained. The simulation modelThe main conception of the simulation model for laser material processing consists in implementing basic physical principles (energy, mass and momentum conservation) in form of coupled differential equations (e.g. heat equation, Navier-Stokes equation) and complementing them by several modules to describe the special interaction mechanisms between the laser beam and the material [1]. The basic equations can, in principle, be solved by means of every advanced CFD software package. However, the implementation of Institute for Manufacturing and High Power Laser Technology TU Wien, AustriaResearch in the department of laser assisted manufacturing supports a wide range of applications. Among them and highly rated in the research activities are the development of optics and metrology systems, sensing technologies, laser cutting and joining, as well as laser assisted forming processes. Most recentl...
In this paper the technique of parameter identification is investigated to reconstruct the 3D transient temperature field for the simulation of laser beam welding. The reconstruction bases on volume heat source models and makes use of experimental data. The parameter identification leads to an inverse heat conduction problem which cannot be solved exactly but in terms of an optimal alignment of the simulation and experimental data. To solve the inverse problem, methods of nonlinear optimization are applied to minimize a problem dependent objective function.In particular the objective function is generated based on the Response Surface Model (RSM) technique. Sampling points on the RSM are determined by means of Finite-Element-Analysis (FEA). The scope of this research paper is the evaluation and comparison of gradient based and stochastic optimization algorithms. The proposed parameter identification makes it possible to determine the heat source model parameters in an automated way. The methodology is applied on welds of dissimilar material joints.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.