Date of Award

8-2020

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Aerospace, Physics, and Space Sciences

First Advisor

Hamidreza Najafi

Second Advisor

Aldo Fabregas Ariza

Third Advisor

Chelakara Subramanian

Fourth Advisor

David Fleming

Abstract

Thermal protection systems (TPS) are critical for space vehicles to keep them safe against extensive aerothermal heating, particularly during the atmospheric entry. Ablative materials burn away gradually when subjected to extreme heat flux and therefore are considered as a practical option for TPS application. Measurement of TPS surface condition, particularly heat flux, is essential to reduce design margins and also improve active control strategies for damage avoidance. However, it is not possible to directly measure TPS surface heat flux, and alternative approaches must be utilized for this purpose. One method is to use temperature sensors at internal layers and solve the associated inverse heat conduction problem (IHCP) to evaluate the heat flux at the surface. The IHCP associated with TPS involves significant complexities, mainly due to the presence of the moving boundary and temperature-dependent material properties. A novel solution based on the application of artificial neural network (ANN) is developed to solve the IHCP in a one-dimensional medium with moving boundary and temperature-dependent material properties in a near real-time fashion. Two possible approaches are considered. The first one uses temperature data from a limited number of previous and future time steps, along with the ablation rate to calculate the surface input heat flux at the current time step (Network A). And a second network uses temperature data only to calculate the conducted heat flux (Network B). The networks are trained, tested, and validated through numerical test cases that are developed in COMSOL Multiphysics. The results demonstrated that the developed approach could successfully estimate the surface heat flux and conducted heat flux in a near real-time fashion and with reasonable accuracy. The findings from this research can be further developed into the design of sensors and health monitoring systems for the TPS application.

Download

Share

COinS