Abstract
In the modern world, reducing gaseous emissions has been a motivational driver and enabler of low-carbon technology development. The release of excessive carbon emissions into our atmosphere is considered a global issue, with academia and industry intensely engaged in developing a viable solution for its reduction.
On the question of lowering gaseous carbon emission in thermal machines that produce electrical energy, the Free-Piston Engine (FPE) can be considered a viable solution. The FPE is an inherently novel technology with higher thermal efficiency than its Conventional Reciprocating Piston Engine (CPE) counterpart. The unique piston motion of the FPE is not constrained kinematically like the CPE by way of its connecting crankshaft and rotational inertial mass. In contrast, it is being dynamically constrained by its translational coupled loads. Constraining the FPE in this fashion enables a vast range of piston motion trajectories to be explored during its development and considered during its operation, which may be essential to reduce in-cylinder gaseous emissions.
With that, this thesis will introduce the key motivations and drivers for continued FPE development and highlight its innate advantages and challenges. FPE fundamentals and its rich history will be presented, including how industry and academia have described, tested and controlled its highly non-linear behaviour.
Accordingly, a novel system-level and multi-domain approach that describes the characteristic multi-directional nature of the FPE is first presented, followed by a detailed non-linear description of the coupled loads. The main body of this thesis demonstrates the usefulness of the proposed modelling approach by capturing the dominant dynamics of the FPE gas expander and opposed-piston FPE during both motoring and power generation. Consequently, the proposed modelling approach is validated against experimental hardware with identical parameters. The performance is assessed with a thorough qualitative and quantitative analysis, showing that the proposed non-linear and multi-domain modelling approach captured the dominant dynamics of the machine and is a viable solution for the future development of the FPE.