In rocket systems, the re-entry speed to atmosphere is very high which leads to compression of air molecules and appearance of strong bow shock waves in the leading edge; consequently, this yields aerodynamic heating. When ablating materials arrive in harsh environment, the ablation of surface occurs. Using ablating-dispensing materials on the leading edge surfaces, it is important to accurately determine heat flux on these moving boundaries. Measuring heat flux directly is very difficult or impossible in some situations. In the literature, two categories of methods have been undertaken as inverse computational tools for recovering heat flux components. These categories are called on-line and off-line methods. In the on-line methods, once the temperature is taken at the current time, the real time estimation for heat flux can be subsequently obtained. This is particularly important in real thermal systems that require specifying unknowns instantaneously. In this condition, the inverse heat transfer methods may be used to determine heat fluxes from temperature sensors. However, temperature measurements are noisy. To determine accurately the heat flux, the temperature noises must be removed using gradient or other methods. Among the methods for solving the inverse heat conduction problems, the conjugate gradient methods have been extensively used for estimating the boundary heat flux. The gradient methods are well accurate; however, they are computationally extensive. In the present study, the online Kalman filtering is used to determine heat flux accurately for charring and non-charring ablators. The Kalman filtering technique is a typical sequential estimation method consisting of repeatedly updating the estimates and a covariance matrix to indicate the reliability of the estimates. This technique is a recursive method of estimating a function or parameters out of noisy data. Since the heat flux is estimated in online (non-iterative) fashion, the optimum location of temperature sensors can be effectively determined. In addition, the results of this study can be used to design heat flux sensors. In this thesis, the optimum locations of three temperature sensors are calculated on the basis that the disturbances occur due to burning of sensors are reduced. More robust solutions are obtained for heat flux on the ablating surfaces. Keywords Heat Flux Estimation, Inverse Heat Conduction Problems, Kalman Filtering Technique, Moving Boundary, Pyrolysis Gas.
