Monitoring and optimization of temperature field of pre-oxidation furnace using CFD and surrogate model

Abstract The temperature distribution in the pre-oxidation furnace has a great influence on the stabilization stage of carbon fiber production. The actual temperature of the pre-oxidation furnace is measured by fiber Bragg grating temperature sensors in this study. Moreover, the error caused by friction to the FBG temperature sensor is compensated so that the accuracy of the temperature sensor is improved. Considering that it is difficult to get the influence of process parameters on the temperature field in production, a numerical calculation model of the pre-oxidation furnace is established by computational fluid dynamics. The effect of the heat release from the polyacrylonitrile precursors on the temperature field is considered. Considering the high calculation cost of the CFD method in an optimization environment, this study established a radial basis function model instead of numerical calculation. The results demonstrate that the surrogate model is valid and the suggested optimization approach significantly enhances the thermal distribution uniformity in the pre-oxidation furnace.