Multiphysics Tube Welding using Induction Heating - ProductGuy2/Multiphysics-welding-examples GitHub Wiki

This example describes how to weld a tube using induction heating. In such processes, first a metal sheet is bent to a tube in a number of roll forming stages. In the final roll forming stage, the two sides of the tube are pushed together. A coil is located in front of the last rollers. An AC electric voltage is applied to this coil. This generates a harmonic magnetic field. Due to this field, induced currents are generated in the tube. The flow of these induced currents follows a closed loop which means that in this situation the current will cross from one side of the tube to the other where the tube is making self-contact. Here at this crossing, the current will be highly concentrated. This concentrated current leads to localized heating due to the Joule effect. When enough heat is generated here, this part of the tube will reach the melting temperature and so a weld forms.

The complete roll forming process will not be performed in this example. Instead, we start with flat plate which is rolled into a tube by applying fixed displacements on the nodes which will form the outside of the tube using user subroutine forcdt. First the plate is rolled into a tube, where the two sides which will be welded together can make touching contact. Then the front part of the tube is glued and the prescribed displacement is removed. This creates a tube which is partly connected and contains a V-shaped opening which will be welded together. The V-shape occurs due to stresses in the bent plate. The model contains the following four parts.

• The flat plate which will be rolled into a tube
• Two rollers used for the final pressing
• A box with finely meshed elements to represent the air box in which the induced current near the V-shaped tip of the tube will be computed. Note that this box is located where the V-shape in the tube will occur once the plate is bent into a tube.
• A box of surfaces which will be used to create the surrounding air mesh.

The dual mesh approach is used where we have separate meshes for the magnetodynamic pass and the thermal/structural pass. The magnetodynamic mesh consists of the coil, the box with finely meshed elements and the remaining region which is meshed as surrounding air. In the dual mesh approach, integration points from the magnetodynamic elements use the material properties of the thermal/structural elements when they are located in such an element; otherwise, they use their own properties. In this approach, the induced currents are not computed in the thermal/structural elements but in the magnetodynamic elements. For this reason, a box with finely meshed elements is located near the V-shaped opening of the tube. The magnetodynamic elements must be small enough in this region so that they can accurately compute the induced currents where they occur in the tube. The heat generated from these induced currents is computed and then transferred to thermal/structural elements. During the simulation the magnetodynamic mesh remains stationary while the thermal/structural mesh, which is the tube, is pushed through the rollers. Integration points from the magnetodynamic elements will, therefore, be located in different thermal/structural elements or outside these elements during this process. In each simulation step, the material properties of appropriate magnetodynamic elements are determined based on the current position, shape and temperature of the tube.

See tubebend.proc in the repo