This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
The Use of an Expert System for the Feasibility Study and the Preliminary Design of an Experiment in a Space Furnace
Annotation ability available
Sector:
Language:
English
Abstract
CNES engineering department has a long experience in space furnace design (GHF, MEPHISTO, MULTIZONE, AGHF etc.) and in the development of their respective experiments. An Expert System (E.S.) known as “CARTOUCHE”** has been developed to check the feasibility and to assist the conception of fusion/solidification experiment cartridges assigned to any axisymmetrical Bridgman type furnace. It will be validated with an ESA furnace, the AGHF (Advanced Gradient Heating Facility), designed for crystal growth and generally for directional solidification experiments up to 1400°C. The furnace consists in a heated section separated from the cooled heat extraction section by a large adiabatic zone. The furnace assembly can be moved along the sample which remains static. So far, five European experiments are to be carried out.
The E.S. is based on a materials data base and the knowledge of multi-experts (scientists, furnace designers, safety specialists), and it includes sophisticated thermal modelling software (ESATAN) to take some constraints into account.
Then, the E.S. supplies the following results:
- The feasibility of the experiment according to the sample and the furnace.
- The geometrical characteristics of the sample and of the crucible, their optimum position in the furnace, the cartridge envelope material, the best way of cooling, the minimum power consumption, etc.
Following this, a nodal mesh of the cartridge and a list of the appropriate materials are proposed.
These will enable the designer to fill the cartridge core parts (spring, thermal conduction parts, super insulation, atmosphere etc.). The temperature and thermal flux at each node, as well as the thermal profile versus time required in the furnace, are then calculated by ESATAN. The designer is thus able to achieve, by successive iterations, a suitably optimised cartridge.