Abstract
The study and development of automotive suspension is an active research topic. It is well-known that improving the ride comfort comes at the cost of vehicle handling. To solve this, active suspension systems have been suggested, developed, and implemented by industry and academia. However, with the introduction of the inerter, the ability to solve this trade-off passively has improved. This thesis studies the extent to which inertance-integrated absorbers improve the performance of two different vehicles: a linearised five-mass model of an in-wheel motor powered vehicle, and a full car model with air springs. The findings of this thesis provide a step towards the adoption of inertance-integrated suspension systems by the automotive industry.Firstly, a five-mass model of an in-wheel motor powered vehicle is modelled in MATLAB and Simulink. By incorporating and optimising inertance-integrated absorbers into the strut and bushes, the performance metrics for the vehicle (ride comfort, tyre dynamic load, and magnetic gap deformation between the rotor and stator) are improved without degrading the others. This study is the first to incorporate inertance-integrated absorbers in an in-wheel motor system, and shows the performance benefit they provide exceeds that possible with passive dampers without needing active control.
Next, MATLAB and Simulink are used to model air springs consisting of air bags connected via pipes to reservoirs. These air springs replace the coil springs of a passenger car modelled in CarMaker. A pneumatic network design methodology is devised and used to perform pneumatic network synthesis on the pipes. This pneumatic network design methodology allows optimal linear inertance-integrated pneumatic networks to be found and completes the existing mechanical-electrical-hydraulic domain coupling methodologies, which previously did not consider pneumatic networks. Real-time simulations are performed with MATLAB, Simulink, and CarMaker to optimise these networks for ride comfort, whilst handling is constrained to not degrade beyond a benchmark. It is found that these networks give greater performance benefits than pure resistances can.
Finally, Amesim is used to create physical pneumatic realisations of the previously identified optimal linear pneumatic networks. Real-time optimisations using the same passenger car as the previous step are performed using Amesim, MATLAB, Simulink, and CarMaker concurrently. It is found that whilst the optimal networks can be realised using pure pneumatic means, many of the resulting physical parameters are inappropriate for automotive applications. This suggests that either an iterative design approach, or multi-domain realisation methods are needed before prototypes can be built.
| Date of Award | 18 Jun 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Jason Zheng Jiang (Supervisor), Tom L Hill (Supervisor) & Simon A Neild (Supervisor) |