Exploration of the Pressure Relaxation phenomena with Particle Image Velocimetry, or pivpr for short.
Vortices are a common phenomenon in fluid flows that arise as kinetic
energy dissipates into heat via viscous interaction. They arise naturally
at large scales in the form of dust devils, tornadoes, and as a
counter-rotating vortex pair in the wake of
aircraft. It is important to understand the conditions leading to their
formation, their duration, and their dissipation in order to forecast or
prevent undesirable effects. Among these deleterious effects is a decrease
in safety of aircraft
operations in the wake of other aircraft, an extremely common situation at
airports around the world. A large number of mathematical models and
experimental data sets exists to help explain various aspects of axial wake
vortex behavior, but current models fail to explain why many vortices
remain tightly wound with slowly decaying azimuthal
velocities about their cores the length of time for which they
have been observed. The current study builds upon the theoretical work of
Ash, Zardadkhan and Zuckerwar, and tests specific attributes
of a turbulent axial vortex for agreement with non-equilibrium pressure
relaxation theory. This theory provides an exact solution to a modified
version of the Navier-Stokes equations for an
axial vortex, with a resulting velocity model that agrees with leading
empirical models. In the present investigation, axial wake vortices were
created with a bi-wing vortex
generator in a low speed wind tunnel, at free stream velocities between 15
and 33
This repository is for code and resources for my masters thesis in Aerospace Engineering at Old Dominion University. The code revolves around synthesis and display of test data taken of an axial wake vortex with Particle Image Velocitmetry (PIV) in the ODU Low Speed Wind Tunnel (LSWT).
The data is discussed in the context of the pressure relaxation phenomena which is theorized to allow the longevity of axial wake vortex flow structures. "Raw" vector data is calculated from raw stereo image data and instrument calibration parameters by comercial software using particle displacement tracking methods, and is stored in tabular v3d files. These v3d files are considered the starting point for this code base, though some additional code has been writen to demonstrate uncertainty principles behind correlation techniques used to calculate 3d vector fields from stereo frame stradled image pairs.
The full 3d vector dataset and working drafts of the thesis can be downloaded from the release page. The vector dataset is 800MB compressed, but unzips to 12GB. The raw images used to synthesize the vector fields are not available for download due to the large file size, but may be supplied upon request.
Statistics sets and plots can be generated from the vector set using controller scripts which interface with a small axial vortex API. Runing this code requires:
And, with easy installation using pip:
Living version of the codebase. Performs analysis and dumps files to texdocs. Built to make the analysis as repeatable as posisble, with full exposure to the methods used as part of this study.
Obsolete code base started in matlab, kept for good record keeping only.
TeX and resources used to produce PDF formatted publication document. Several subdirectories are automatically filled by py/controler scripts with TeX files, figure jpgs, and tables converted from .csv format.
Folder to place full set of processed 3d vector files. Controllers in the py\controler package will work if data is correctly placed in this folder.
Small limited subset of vector files initially used for testing and development.
The intention was to eventually use OpenPIV to perform computation, but that repository at the time of this research still needed a lot of work. This approach was abandoned, but I encourage future users of PIV and python to check it out.