• Current opened records

  • Evaluating the Potential for Simulation of Formula Student Events in IPG CarMaker

Awards
Author(s):
  • Jonathan Jamison
Category:
  • Engineering
Institution:
  • Queen's University Belfast
Region:
  • Island of Ireland
Winner Category:
  • Regional Winner
Year:
  • 2020
Abstract:
  • High fidelity simulation software is revolutionising the way in which engineers create and validate the world around us, the benefits of which are perhaps most prominent within the automotive industry. Eradicating the need for resource intensive physical testing, simulation allows multiple configurations to be assessed at a fraction of the time and cost – making its use within a Formula Student team an obvious application.


    This paper documents the role in which an off-the-shelf, transient multibody dynamics simulation package (IPG CarMaker) played during a Formula Student team’s transition from a traditional internal combustion-based power train to an all-electric alternative. This was achieved by creating accurate depictions of the Formula Student dynamic event tracks, upon which a detailed model of the previous season’s car was simulated as to provide a means of validation. Said model was inclusive of collected empirical aerodynamic, driver, powertrain and tyre testing data and subsequently validated using real-world telemetry. This basis was then modified as to represent preliminary design estimations regarding the electric vehicle with simulations being used to assess potential performance.


    An innovate step was taken to transform the tool from passive to generative, in that optimal suspension characteristics were redetermined autonomously through the integration of machine learning. A Genetic Algorithm was incorporated which programmatically produced and evaluated thousands of potential setups and ultimately generated an improved design with minimal human input – a feat yet to be documented in any other publication regarding Formula Student.


    In conclusion, IPG software was crucial to conveying correlations between cause and effect with regards to vehicle modification and will continue to play an ever more prevalent role as the team aims towards a driverless vehicle by 2025. Whilst only altering one aspect of the vehicle in its first iteration, the Genetic Algorithm demonstrated its capability in streamlining the vehicle optimisation process with potential to refine many more aspects of vehicle design.
Attached Documents: