Overview

The 2023-2024 AIAA Design Build Fly competition constrained the takeoff field length (TOFL) of the models to only twenty feet. This constraint drove the sizing of the airplane considerably, as well as several design decisions regarding propulsion and configuration.

Late in the year, discussions about mission strategy started to appear. Questions such as how much payload can we carry for a given amount of wind to still meet TOFL, or the true impact of a change in wing sizing for the takeoff roll (considering the added weight, for example) were being asked.

TOFLAnalyzer

With these questions in mind, my colleague Nicholas and I set out to write a higher fidelity takeoff model that could be used to answer such questions. The idea was for a time-based takeoff sim of moderate fidelity, easy to run on-the-go (in fact, a lot of this development occured during test flights, or at competition!) and that could yield reasonable results.

We started with a constant drag, weight, and thrust models, and a geometry parametrization similar to that of Project Caladrius. With these basic models in place, we wrote the takeoff model with scipy’s numerical integration algorithms, which seemed to yield reasonable results (note the low time-span - keep in mind this aircraft is a handful of pounds, and has a thrust-to-weight greater than 1.6!)

We then built up on this, adding a parasitic drag buildup (which makes the CD0 a function of Reynold’s Number, and therefore flight speed) and a constant power model instead of a constant thrust model. We also wrote trade study functionality to sweep parameters, and experimented with varying payload weight. The wind models were also added, but we never had time to fine-tune them - some preliminary wind trade study results seemed off.

Finally, I implemented a QPROP thrust model - the user inputs a motor and propeller combination, and the time-based sim will run QPROP at every timestep, extract the thrust and propwash for the given RPM and flight speed, and use the thrust in the equation of motion. The propwash was extracted with the idea of modelling blown-lift. The QPROP functionality used codes previously developed by this author to run the tool from Python.