Design optimization study for an infrared heater using CFD and sensitivity analysis

SUMMARY Sensitivity analysis (SA) has become a very popular technique for various application areas. However, not many studies concern SA using CFD (computational fluid dynamics). The main problem appears to be the large number of model executions when using the Monte Carlo method of SA. The computing resources and time needed for such a SA in combination with CFD (in which one model execution could take several hours) can be enormous. Nevertheless, this paper describes an approach of using SA in such cases. The recommendations are obvious: simplifying the model as much as possible; considering just a key part of the model; and take more care about the speed of convergence. On this basis a case study on “Optimization of gas infrared heater radiation geometry” has been performed. The most interesting result is that the shape of the heater and the reflector material do not significantly influence the heat delivery to the occupied zone. The main conclusion is that Monte Carlo sensitivity analysis is not suitable for large CFD models with low level of