Have you ever created a FEA model on diffusion of heat on a material? Patran is a great tool that will accurately create a FEA model on diffusion of heat. University of Texas take advantage of Patran’s capability in creating a coupled electro-thermal analysis on an aluminum field coil and copper center post.
The University of Texas at Austin Center for Electromechanics uses MSC PATRAN for thermal and stress analysis. In this analysis, PATRAN was used to determine the thermal temperature rise and the resulting stresses in the conceptual design of an actively cooled toroidal field coil.
The analysis was done in two steps:
- A coupled electro-thermal analysis of the structure with heat generation due to a 25 MA current passing through the aluminum field coil and copper center post. Current diffusion was analyzed, and based on the current densities the Ohmic heat generation calculated. Temperature rise was calculated based on the Ohmic losses and convective boundary conditions applied at cooling channel surfaces.
- The second analysis took the nodal temperatures from the electro-thermal analysis as input and calculated stresses due to temperature gradients, as well as electromagnetic body forces acting on the field coil.
PATRAN Highlights:
Automatic Tetrahedral Mesh Generation: Taking advantage of cyclic symmetry, a ten degree circumferential section was modeled using PATRAN’s automatic tetrahedral mesh generation capabilities, as seen in figure 1. The model incorporated cooling channels at which convective film coefficients were defined.
Boundary Conditions and Body Forces: A magnetic body force proportional to the square of the radial position was easily defined using PATRAN’s capability to define boundary conditions and body forces as analytical expressions of position. PATRAN was used to generate an input file for use with ABAQUS finite element software, and for post-processing. Plots of the temperature rise and thermal stresses were generated using PATRAN. Figure 2 shows the temperature rise due to the Ohmic losses, and figure 3 shows Von Mises stress.
 |
Steve Manifold
|
 |
Shannon Strank
|
