Speaker
Description
Today, thermoelastic treatment relies on providing maps of average temperatures by zone. This is a conservative approach because these temperatures gradients are maximized. The main objective is to develop an automated link between mechanical engineering models and thermal analysis ones, with different methods (finite elements vs nodal method) and significative geometrical representativeness differences (100000 vertices vs 1000 nodes).
One of the main challenge consists in having a precise enough thermal model in order to obtain satisfying temperatures at mechanical points in order to compute the thermoelastic deformation of the structure and depointing of the instruments.This operation while being coarse enough to perform simulation in an acceptable amount of time.
If the initial thermal model is not accurate enough, the tedious task of designing a more accurate one usually has to be performed. To tackle this issue, DOREA and TAS investigated the idea of determining temperatures at mechanical points from an existing e-Therm thermal model without the need of redesigning a finer-grained thermal model.
Thanks to e-Therm algorithms, especially for instrumented panels, e-Therm allows a detailed temperatures cartography to be a physical representation of the heat equation (without any interpolation or extrapolation algorithms).
For the remaining parts of the thermal models and for all mechanical geometrical parts which don’t have a thermal representativeness, we apply an interpolation/extrapolation method.
The method is currently being tested and validated on the complete EUTELSAT 10B satellite. E10B telecom program provided a detailed temperatures cartography which match directly on a FE mechanical model declined over time. This tool may also be used to determine the worth sizing cases for thermoelastic aspects.
