Investigating Thermal Testing Methods in the Frequency Domain

Abstract

Thermal Mathematical Models (TMMs) are used to predict the thermal behavior of satellite structures in orbit. However, due to inherent uncertainties in these models, physical testing is necessary to achieve reliable predictions. While these tests are critical, they often introduce uncontrolled uncertainties, such as heat leaks and measurement errors, making the correlation process complex and time-consuming. To address these challenges, this thesis proposes an alternative testing methodology that reduces uncertainties in thermal test data by using the phase shifts between temperature responses from oscillatory heat loads for TMM correlation. By comparing the measured and predicted phase shifts, thermal model parameters such as conductance or capacitance can be effectively correlated. The results demonstrate that this methodology is largely insensitive to uncontrolled conductive heat leaks, allowing the correlation process to focus only on key thermal parameters. This approach improves the reliability of thermal test data and streamlines the correlation process. Moreover, its potential application in ambient conditions offers a promising testing solution for early-phase model correlation.