Modelling and simulation of welding and metal deposition

Abstract: Fusion welding is one of the most used methods for joining metals. This method has largely been developed by experiments, i.e. trial and error. The problem of distortion and residual stresses of a structure due to welding is important to control. This is especially important in the aerospace industry where the components are expensive and safety and quality are very important issues. The safety requirements and the high costs of performing experiments to find different manufacturing routes is the motivation to increase the use of simulations in design of components as well as its manufacturing. Thus, in the case of welding, one can evaluate the effect of different fixtures, welding parameters etc on the deformation of the component. The effects of previous processes are also important to consider, as well as it is important to bring forward the current state to subsequent processes.When creating a numerical model, the aim is to implement the physical behaviour of the process into the model. However, it may be necessary to compromise between accuracy of the model and the required computational time. The aim of the work presented in this thesis was to develop a method and model for simulation of welding and metal deposition of large and complex components using the finite element method. The model must be reliable and efficient to be usable in the designing and planning of the manufacturing of the component. In this thesis, the meaning of efficiency of a model is wider than just the computational efficiency. The time for creation and definition of the model should also be included. The developed methods enable the user to create a model for welding or metal deposition with a minimum of manual work. The method for defining weld paths and heat input together with activation of elements is now implemented in the commercial finite element software MSC.Marc. The implementation is based on the experience in this work and communication with the author. The approach has been validated against test cases. Naturally, this validation is dependent on sufficient accuracy of the heat input model and material model that are used. It is the first time a dislocation density model has been used to describe the flow stress in a welding simulation. The work has also demonstrated the possibility to calibrate heat input models with a physical based heat input model, thus relieving the need to calibrate the heat source versus measurements.Efficiency in terms of computing time has also been investigated in the course of this work. Three different methods has been explored and used, adaptive meshing, substructuring and parallel computation. The method that is found to be the most versatile and reduce the overall simulation time the most is parallel computation. It is straightforward for the user to employ and it introduces no reduction in the accuracy.