The aim of platypus is to develop a robust simulation tool for hybrid rocket motors that aims to capture some of the unsteady dynamics within the motor. This should allow greater insight into potential transients during motor start-up as well and potential instabilities of the motor.
At the moment the project consists of helper functions for a 1D finite volume solver using Roe's approximate fluxes.
To install use:
pip install git+https://github.com/cuspaceflight/platypus.git
To perform validation of the solver a validation suite is present and can be run through:
tox
This tests the various flow relationships implemented in the helper functions as well as validating the 1D finite volume solver against the analytic solutions of Sod's shock tube and a flow through a nozzle.
The current simulation tool used by CUSF for motor simulation uses a quasi-steady process, treating the fuel grain as a single unit. At each step the oxidiser mass flow is calculated by using the tank pressure and current chamber pressure, the corresponding fuel flux is found, the gas properties based on the O/F ratio are computed and then the mass flow through the nozzle is calculated. An iterative scheme is then applied to adjust the chamber pressure so that these mass flows are equal. The tank pressure is then updated based on the oxidiser mass flow and time step and the process is repeated.
Platypus aims to investigate the unsteady dynamics that the above method masks. By investigating variation of regression rate and gas composition along the motor length then the effect of the varying O/F ratio on motor performance will be revealed.
In addition to developing a robust simulation code the other goal of this project is to provide decent documentation such that future CUSF projects and other groups can make use of and improve these tools.