Optimization of HVAC planning of large high-ceiling room for spacecrafts and satellites using CFD and optimization algorithm

Generally, a clean room consumes a lot of energy for air conditioning because the thermal environment and cleanliness are the highest priorities. In particular, the fairing assembly room, which is a clean room for spacecrafts and satellites and is the subject of this report, is a very large space with a ceiling height exceeding 40 m and therefore inevitably consumes a large amount of energy for air conditioning.

The fairing assembly room uses an overhead crane to suspend the fairing, which is cover for protecting satellites, from the upper space of the satellite while installing it, so the ceiling height must be at least the sum of the satellite and fairing heights. However, since any works except installation are performed in an area less than half the ceiling height, it is appropriate that the HVAC system condition only the lower half of the room from the viewpoint of energy saving. In this case, it was difficult to estimate the cooling load due to staying hot air at upper part of the room, and it was necessary to study the air conditioning plan by trial and error using the block zonal model or non-isothermal CFD analysis.

Recently, with the development of computers and analysis software, design exploration using an optimization algorithm with CFD results is becoming practical in the architectural field, and it is expected to save manpower in the detailed design and planning of air conditioning in the future. The goals of this study are to conserve energy and maintain a thermal environment, but these two objectives are often in conflict. Therefore, multi-objective optimization was performed to search for a Pareto front and the multi-objective optimization algorithm MO-SHERPA is employed in this study.

The design variables were supply temperature, air flowrate, and blowoff angle. CFD analysis was conducted for each design with different combinations of variables. From the results of the CFD analysis, A) input cold heat and B) temperature difference between the average temperature of the evaluation area and the set temperature were calculated, and combinations of design variables that minimize these variables were explored. Even within a short study period (about 2 weeks), Several best designs, for example, a design with the average temperature of work area closer to the set temperature than the base design, and a design with a smaller cold heat input within the range that satisfied the constraint conditions, were found.