RC Hovercraft

UBC 2nd year design project


Personal contributions

  • Team leader
  • Lead mathematical model development and simulation
  • Lead experimental design and data collection
  • Managed overall scheduling
  • Co-editor for the final report

Team achievements

  • 6th place overall in the hovercraft competition
  • Most accurate and advanced mathematical models (Monte Carlo, Sensitivity Analysis, Fluid dynamic simulation, Risk optimization)
  • Highest overall presentation score (out of 20 teams)
  • Highest documentation score
  • Utilized Monte Carlo simulation to accurately predict overall score before the competition began 
  • Had a tonne of fun :)!

 

 

 

Team A4 from left to right: Jeffrey Chao, Paiman Parmei, Ecem Kahraman, Max horner, Owen Lu, Christopher Bandy

Background

Every year the mechanical engineer department at UBC has a design competition where teams of students create a vehicle to participate in various challenges with scoring based on various performance metrics. In 2013, the challenge was to build a remote controlled hovercraft to participate in sprint, endurance, maneuverability and payload capacity competitions. All scores also factored in the team's ability to predict the actual performance, the cost and weight of the vehicle. Notably, competition scores had very little representation in a student's overall mark. Project score was determined heavily by an oral presentation, documentation log and a final report.

Objective

Conceptualize, construct, simulate and test a remote controlled hovercraft to compete in a class-wide battle of speed, cargo capacity, and maneuverability in 2 weeks.

Testing

Three main experiments were conducted

  1. Propeller thrust test
  2. Fan/skirt payload test
  3. Linear speed test

The results were then combined into simulation for three main competitions

  1. Sprint competition
  2. Cargo carrying competition
  3. Endurance competition

Each of the tests fed data into a physics based model that allowed the prediction of performance using Monte Carlo simulation to generate the prediction for optimal scoring.

Simulation

Mathematical models were developed for every competition and their parameters calibrated with experimental data. The main simulations used Monte Carlo simulation, Non-dimensional analysis, Blade element theory, and DC motor theory.

Relative scoring made prediction of the best approach to maximization of performance difficult. Critical models built upon fluid dynamics and DC motor theory were combined to produce quantitative trade off analysis between different designs. An interesting part of the competition was that the courses were not fully defined before the competition so that teams could not simply test their design and predict based on experimental results.

The scoring formula, which was well defined but complex factored in vehicle weight, performance in five competitions, and prediction accuracy. Notably the prediction accuracy factor favoured underprediction to overprediction.