Dynamic prediction of fuel temperature in aircraft fuel tanks based on surrogate

Due to the extreme increase of thermal load and the constraint of ram-air utility by stealth requirements, fuel becomes an important heat sink in an aircraft and plays a role in the heat collection, transport, and dissipation. Additionally, the fuel temperature is limited by the engine inlet temperature and fuel stability. Therefore, dynamically predicting the fuel temperature in aircraft flight is essential for the thermal management and propulsion systems. This paper proposes a rapid dynamic prediction method to calculate the fuel temperature under a complete flight profile, which is based on surrogate modeling, heat transfer theory of a fuel tank, and CFD technique. The CFD results show that the distributions of fuel temperature and wall convective heat transfer coefficient are uniform, and thus employing the average values of these two quantities in the prediction method is reasonable. Based on a sample database from CFD simulations, a response surface method surrogate model performs well in predicting the wall convective heat transfer coefficient. Ultimately, combined with the theory of heat transfer, an efficient algorithm is designed to predict the fuel temperature under a complete flight profile. The fuel temperature is predicted to increase from 298 K to 317.3 K by the operation of thermal management system under the typical profile. The computation results are explicable and in qualitative agreement with engineering data.