Lap Time Simulation Tool for the Development of an Electric Formula Student Car 2019-01-0163
This work details the development of lap time simulation (LTS) tool for use by Queen’s University Belfast in the Formula Student UK competition. The tool provides a user-friendly virtual test environment for the development of the team’s first electric vehicle. A block diagram vehicle model was created within Simulink, and a series of manoeuvres simulated to generate the performance envelope of the car in the form of maximum combined lateral/longitudinal accelerations against velocity (ggv diagram). A four-wheeled vehicle including load transfer was modelled, capturing shifts in traction between each tyre, which significantly influence performance in vehicles where the total tractive power is split between individual wheel motors. The acceleration limits in the ggv diagram were used to simulate the various track events at Formula Student as MATLAB code using a point mass, quasi-steady state, single tyre model by assuming the car was always been driven on the performance limit by a perfect driver. The use of a ggv diagram which was created considering all four wheels captures performance characteristics that simplistic point mass models normally do not, without the complexity and time required for more detailed LTS solutions including yaw movements and driver models. Additionally, as the LTS relies only on the output from the Simulink model, setup changes can be easily implemented within the block model without having to change any of the subsequent code. A Latin Hypercube sampling technique was employed to generate numerous input scenarios for the simulation, and a regression model used with the results to perform a sensitivity analysis – results from which indicate electrical efficiency and vehicle mass as efficient areas to target for future development. Optimum powertrain gear ratio and battery capacity were specified using the tool.
Darryl Alan Doyle, Geoffrey Cunningham, Gavin White, Juliana Early