Free piston linear engine alternators (FPLEA) may be designed following several different baseline configurations. Common designs include a translator that carries permanent magnets, with either one piston attached to one end of the translator, or a piston at each end of the translator. The single cylinder engine requires a reversing force from a spring so that it can operate, whereas, the dual cylinder version can operate without a spring. However, inclusion of stiff springs would serve to raise the operating frequency and reduces the cycle-to-cycle variations in a dual cylinder version. A matlab/simulink numerical model with translator rod dynamics and in-cylinder thermodynamics was employed to predict the overall performance and efficiency of a FPLEA. This numerical model allowed comparison of different FPLEA configurations for a variety of design variables. First, a dual cylinder FPLEA design was considered where the spring constant was varied, changing the frequency of operation and the motion of the translator. The simulation results showed that without springs the motion was far from sinusoidal, and low in frequency and power, whereas the presence of stiff springs in the system strongly dictated nearly sinusoidal motion and high power at high frequency. Effects of other parameters such as stroke and bore were also examined. Finally, the same tests were carried out for a single cylinder FPLEA design. The simulation results showed that the dual cylinder engine produced twice the electrical power output with higher efficiency than the single cylinder engine for the same cylinder geometry.
Free Piston Engine Generators (FPEG) directly convert the reciprocating piston motion into electricity by using a linear alternator. Unlike conventional engines with piston motion restricted by a crankshaft mechanism, the FPEG piston motion is constrained by the energy available in the system. When stiff springs are considered in the design, the FPEG system attains high frequency with high power and efficiency. The main objective of this research was to model stiff spring-assisted FPEG system dynamics and performance accurately, and to apply the modeling results to the development of a 1kW, spark ignited, natural gas fueled, FPEG experimental prototype. The experimental data was further utilized to refine and improve the existing model. First, a MATLAB®/Simulink based multi-cycle numerical model was developed for single and dual cylinder FPEG systems to study the effects of major design parameters on FPEG dynamics and performance. When stiff springs were added, the dynamics became more sinusoidal and symmetric with respect to the initial starting position. For a total displacement of 34 cc, trapped compression ratio of 8.25, and assumed combustion efficiency of 95%, the modeled frequency and electric power varied from 72.3 Hz to 80.8 Hz and 0.81 kW to 0.88 kW for a single cylinder FPEG as the spring stiffness changed from 372 kN/m to 744 kN/m. For a dual cylinder FPEG with the same conditions, these modeled values changed from 76.8 Hz to 84.1 Hz and 1.7 kW to 1.8 kW with increasing spring stiffness. The numerical model was then expanded for sensitivity studies of major design parameters. When FPEG operating conditions were considered, the effective stroke length was found to have a dominant effect on efficiency followed by compression ratio, cylinder bore, and spring stiffness respectively. The experimental FPEG prototype generating 550 W of electricity with indicated efficiencies exceeding 13.8% was used for model validation. Finally, the stable FPEG system requires a control strategy to match the power generated by the engine to the power demanded by the alternator. A model-based control strategy was developed in Stateflow® for alternator mode switching, calibration maps, energy management, ignition and fuel injection timings. With the proposed control strategy and stiff spring dominance, the modeled and experimental FPEG system maintained stable operation with cycle-to-cycle variations less than 5%.