Work in progress / incomplete documentation
Python 3.7 package that provides exactly what it sounds like by automating usage of the GPL-licensed CLI-utilities XFOIL and XROTOR.
Both utilities are published by professor Mark Drela (MIT).
- XFOIL: for arbitrary 2D airfoil analysis
- XROTOR: for arbitrary propeller design schemes
PDT seeks to provide the user a set of python3 utilities that can be used for arbitrary scripting efforts to automate usage of both XFOIL and XROTOR while implementing its own unique python3.7-native algorithms to maintain local input files, meta files, databases, and results files and weave everything together for the user in a simple, meaningful way to aid in the initial / investigatory stage of well-behaved propeller designs.
pip install propeller_design_tools
import propeller_design_tools as pdt
PDT operates on two different "database" directories, defined by the user with:
pdt.set_airfoil_database(path: str)
pdt.set_propeller_database(path: str)
The user must set these two directories at the top of every script right after the imports
The airfoil directory will be used to store any foil / XFOIL- related support files, and the propeller directory will be used similarly to store any propeller / XROTOR - related support files.
In order to utilize any PDT functionality that depends on running XFOIL, the "xfoil.exe" executable file needs to be in the user-set "airfoil_database" location. Current pip installations include the xfoil.exe file in the foil_database, there should theoretically be no need to download it manually.
Likewise, in order to utilize any PDT functionality that depends on running XROTOR, the "xrotor.exe" executable file needs to be in the user-set "propeller_database" location. Current pip installations include the xrotor.exe file in the default prop_database, there should theoretically be no need to download it manually.
XROTOR executable and docs (this is actually a link to "CROTOR", which I find is actually the easiest way to obtain a windows-executable of XROTOR)
At a high-level, the current concept for PDT workflow is as follows (after obtaining the required executables and pip-installing the PDT package):
-
Try out the (currently extremely buggy and incomplete) user interface! example0_user_interface.py
-
Obtain normalized airfoil coordinate files from UIUC Database -> save these files into the "airfoil_database" directory
-
Use PDT to run XFOIL across ranges of Reynolds Numbers in order to populate database data for the desired foil sections -> see example1_airfoil_analysis.py
-
Once the required 2D airfoil data is generated, PDT can then be used to automatically generate all the required 2D foil definition parameters required by XROTOR (these "station parameters" are essentially what allow XROTOR to model the performance of well-behaved, arbitrarily-lofted blade geometries) -> see example2_radialstation_creation.py
But this step is also automated & displayed by PDT when the user uses the builtin PDT propeller creation function -> see example3_prop_creation.py
NOTE: It is highly recommended to first run XROTOR using either the 'grad' or the 'pot' vortex formulation in order to get your design "tweaked in" -> these are much faster than the (more accurate) 'vrtx' formulation, which you can then move on to.
-
PDT's Propeller() object instances can generate 3D geometry files including profle xyz coordinate listings, and .stl 3D geometry files -> see example4_stl_generation.py
-
Analyze a given Propeller() instance across a sweep of operating points -> see example5_prop_analysis.py
-
WIP Prop optimization (grid-search style generic optimizer for "optimal" prop design generation by means of maximizing or minimizing a given output / calculated metric based on outputs, optionally taking into account different propeller operating points via the ability to define the propeller's "duty-cycle")