MULTI-PHYSICS SIMULATION IN THE 21st CENTURY.

Email: info@physica.co.uk

[ ALUMINIUM REDUCTION CELL ]   [ CASTING & SOLIDIFICATION ]
[ ELECTRONICS APPLICATIONS ]   [ FORMING ]   [ GRANULAR FLOWS ]
[ STEEL PROCESSING ]   [ WELDING ]

Welding

Traditionally, investigation of welding processes have primarily focused on the thermo-mechanical behaviour of the work piece, the goal being to predict the final distortions. However, accurate characterisation of weld properties and quality requires a detailed understanding of weld pool dynamics and thermal stress in the work piece. Hence, any predictive model demands simultaneous consideration of turbulent free surface fluid flow, convective heat-transfer, mass-transfer, phase-transfer and possibly the effect of electromagnetics forces as well.

A modelling scenario like this, where complex interactions among so many phenomena simultaneously take place, generally requires using functionalities from a number of different pieces of software where each may focus on solving individual physics. However, PHYSICA provides a comprehensive and integrated multi-physics solver where phenomena associated with the welding process can be accurately and efficiently modelled under a single unified environment.

Example of PHYSICA application in welding process

PHYSICA’s capabilities for modelling the welding process are illustrated in the following example of Gas Metal Arc (GMA) welding of a T-joint. The computational model implements buoyancy, surface tension and electromagnetic forces in the solution of convective heat transfer in the molten weld pool.

The heat flux from the arc causes density gradients in the molten metal which causes thermally driven flows in the weld pool. The welding current and induced magnetic field can influence fluid flow in the weld pool by way of electromagnetic forces, whilst the variation of surface tension, due to the concentration of surface active elements such as sulphur and oxygen can dramatically alter the weld pool convection. These three forces combine together to determine the flow and hence the shape, size and penetration of the molten pool. It is therefore imperative to model these phenomena so that accurate predictions of the heat affected zone (HAZ) can be made.

The HAZ produced by welding is shown [ IMAGE ], whilst the following image [ IMAGE ] shows the molten weld pool region predicted by the simulation. The velocity profiles within the molten zone can also be viewed [ IMAGE ]. The development of the surface temperatures in the T-joint over time can be viewed [ IMAGE ], and the deformation and surface temperatures at a particular time are contained in the following image [ IMAGE ].