PFAT: Post-flight analysis toolkit

Kurzfassung

Throughout every phase of a space mission, from conceptual design until disposal, a large amount of data is generated for each different engineering discipline involved. Most of this data is produced during the design phases, but additional information is gathered during mission operations. This relates directly to flight performance, which needs to be evaluated by conducting a post-flight analysis. To this end, and to additionally identify and understand the cause of potential discrepancies between predicted and observed behaviour, a large amount of data needs to be analysed methodically. This is key to increase the fidelity of analysis and prediction tools, as their outcomes are used for future mission design iterations. PFAT (Post-Flight Analysis Toolkit) is an open-source software, developed under an ESA contract, which aims at increasing the automation of this process. It allows for the extraction of figures-of-merit and uncertainties and for the derivation of engineering criteria for further design. PFAT implements a Common Data Structure (CDS) to treat data from different engineering domains in a homogeneous way. The domains covered by PFAT are aero(thermo) dynamics, structural and thermal analysis, propulsion system modelling and trajectory simulation. All modules are written in Python3, compatibly with ESA’s Open Simulation Framework (openSF), allowing advanced processing chains definition. By virtue of its modular approach, the toolkit can be easily extended with more functionalities, bringing the postprocessing automatization, data exchange simplicity and processing robustness to other engineering disciplines. An extensive validation campaign has been designed and executed after the implementation phase to ensure that the modules compatibility is guaranteed and that the functionalities available for all the engineering domains provide the expected results when compared to external tools. These end-to-end tests are designed by leveraging the openSF capability to join different modules into complex processing chains that simulate real operational scenarios and analyses.