A Python-based Post-processing Toolset For Seismic Analyses
Abstract
This paper discusses the design and implementation of a Python-based toolset to aid in assessing the response of the UK's Advanced Gas Reactor nuclear power stations to earthquakes. The seismic analyses themselves are carried out with a commercial Finite Element solver, but understanding the raw model output this produces requires customised post-processing and visualisation tools. Extending the existing tools had become increasingly difficult and a decision was made to develop a new, Python-based toolset. This comprises of a post-processing framework (aftershock) which includes an embedded Python interpreter, and a plotting package (afterplot) based on numpy and matplotlib. The new toolset had to be significantly more flexible and easier to maintain than the existing code-base, while allowing the majority of development to be carried out by engineers with little training in software development. The resulting architecture will be described with a focus on exploring how the design drivers were met and the successes and challenges arising from the choices made.
Keywords
Cite
@article{arxiv.1412.6410,
title = {A Python-based Post-processing Toolset For Seismic Analyses},
author = {Steve Brasier and Fred Pollard},
journal= {arXiv preprint arXiv:1412.6410},
year = {2014}
}
Comments
Part of the Proceedings of the 7th European Conference on Python in Science (EuroSciPy 2014), Pierre de Buyl and Nelle Varoquaux editors, (2014)