A novel algorithm created by ThinkTank Maths will reduce the duration of a critical phase in spacecraft testing from months to minutes.
Creating systems that survive in space is without a doubt one of the most difficult engineering challenges. For example, spacecraft and satellites are exposed to extreme hot and cold temperatures — resulting in vibrations, disruptions to satellite systems and even stress fractures. Extensive thermal testing before launch is essential for the success of a mission.
However, many subtle properties of spacecraft materials and structure cannot be measured directly. A full thermal analysis of a spacecraft requires European Space Agency’s (ESA) Thermal Control engineers to correlate complex numerical simulation models with test data from the Large Space Simulator (LSS), the largest vacuum chamber in Europe. This task of thermal model correlation (TMC) is a painstaking process that involves fine-tuning hundreds of parameters by hand. Even when carried out by a dedicated, experienced engineer, TMC can take several months to complete.
Finding the solution to this enormous puzzle is crucial: the spacecraft cannot be launched before the thermal testing is complete, and any delay to launch comes with an enormous cost. For this reason, ESA has been investigating the possibility of an automated TMC solution for several years — but standard optimisation approaches like genetic algorithms came up against an insurmountable complexity barrier. Looking for a fresh perspective, ESA turned to ThinkTank Maths (TTM).