Numerical modeling of combustion chamber material permeability change

One of the most significant challenges for reaching a Mach number greater than four when using Supersonic combustion ramjet (Scramjet) is the thermal management. To overcome this temperature withstanding issue of materials, the transpiration cooling technique is used. Fuel itself is used as coolant and flows through the walls (porous) of the combustion chamber. Beyond a certain temperature, the fuel is pyrolyzed. This can generate coke particles at the surface and inside the porous material. This progressive formation of coke decreases the material's permeability. Hence, predicting the Darcian permeability evolution of a porous material is very important for better understanding the transpiration cooling technique efficiency. Considering existing experimental data for development and validation, this paper proposes an Artificial Neural Networks (ANN) model for estimating the transient changes of the Darcian permeability of a metallic material during fuel pyrolysis conditions. The ANN architecture with 24 hidden neurons is shown to give the best choice. Good agreement was obtained between numerical and experimental results. The prediction ability of ANN was compared with that of linear regression model. This work is expected to be used by aerospace engineers in order to study the efficiency of the transpiration cooling technique.