Rapid Extraction of Propeller Geometry Using Photogrammetry

As small Uninhabited Aerial Vehicles (sUAS) increase in popularity, computational analysis is increasingly being used to model and improve their performance. However, most manufacturers of propellers for this size of vehicle do not publish geometric information, making simulation of propeller aerodynamics difficult. While techniques exist to accurately extract the geometry of a propeller, these methods are often expensive or time-consuming. The method of choice among sUAS researchers is to cast the propeller in resin, slice the resulting piece into thin sections, and scan each section to extract the airfoil at that station. This technique is destructive and labor intensive as a human must perform each step. 

This paper describes a method to produce point clouds using readily available photo equipment and software, and subsequently fit typical airfoil and propeller blade parameters with minimal human input to produce a 3D model suitable for use in techniques such as Blade Element Momentum Theory. The propeller geometry generated is compared to one extracted using established methods and good agreement is shown.