Inverse Methods for Near-Real Time Estimation of Surface Heat Flux in Thermal Protection Systems for Space Vehicles
Date of Award
Doctor of Philosophy (PhD)
Mechanical and Civil Engineering
Gnana Bhaskar Tenali
With the increased goal of manned explorations to outer space, safety during the atmospheric entry/reentry process is paramount as the incident heat fluxes on the surface of vehicle can reach critical values. Thermal Protection Systems (TPS) are therefore essential components in space vehicles as they ensure the safety and protection of the payload and cargo. In order to ensure improved active control capabilities and prolonged used of the vehicle, TPS surface conditions need to be known and determined. Since direct measurement of the surface heat flux using sensors is not feasible, one alternative involves the use of temperature sensors located at layers within the TPS to develop and solve the associated Inverse Heat Conduction Problem (IHCP). Through the solution of the IHCP, estimates of the surface heat flux can be obtained although these solutions are highly sensitive to measurement errors. In this dissertation, a novel solution for the estimation of surface heat flux is presented. The solution is developed based on a filter form of the Tikhonov Regularization method which allows for near-real time estimation of surface heat flux in a multi-layered TPS medium. The solution is evaluated through numerical test cases which have been developed in ANSYS using experimental data from the literature. A parametric study is also conducted in order to understand the effect of sensor location (two layers and three layers models) as well as the effect of temperature dependent material properties on the performance of the solution. The proposed solution technique is shown to be fast, accurate and very convenient to implement even for complex problems involving large temperature variations and temperature dependent material properties. The main advantage of the proposed solution is in its computationally efficient nature as well as possibility of implementation for monitoring and control purposes due to the near real-time operation of the method. The second part of this work involves the expansion of the proposed solution to encompass moving boundary TPS. Since a significant portion of TPS involve the use of ablative materials i.e. materials which undergo surface recession based on incident heat flux, the developed algorithm is tested for such cases. In this portion of the work, it is hypothesized that interpolation of pre-calculated filter coefficients for a medium with various thickness values can be used to calculate the filter coefficients for an IHCP with moving boundary knowing the current thickness of the domain. This can be used along with interpolation among the pre-calculated filter coefficients at different temperatures knowing the temperature value to find the correct filter coefficients at each time step and accurately estimate the surface heat flux in a near real-time fashion. The hypothesis is validated through several numerical test cases developed in COMSOL. It is shown that the proposed solution method can accurately estimate the surface heat flux on the moving boundary in a near real-time fashion. Finally, the use of Artificial Neural Networks (ANN) is investigated as a potential solution for the aforementioned IHCP, especially those involving ablation. By using temperature history data from within the medium, networks are trained to identify patterns within the data and provide estimates of the surface heat flux. It is shown that the use of ANN can provide accurate and computationally efficient estimates of the surface heat flux.
Uyanna, Obinna Oluchukwu Ugochukwu, "Inverse Methods for Near-Real Time Estimation of Surface Heat Flux in Thermal Protection Systems for Space Vehicles" (2021). Theses and Dissertations. 1084.
Copyright held by author