The Free Piston Linear Generator - Development of an Innovative, Compact, Highly Efficient Range-Extender Module

<div class="section abstract"><div class="htmlview paragraph">The present paper introduces the Free Piston Linear Generator (FPLG) - a compact electricity generation unit, which is being developed at the German Aerospace Center (DLR). It is designed as a free-piston combustion engine with integrated linear generator. This combination allows for highly efficient conversion of the chemical energy stored in a fuel to electrical energy.</div><div class="htmlview paragraph">By combining a two stroke combustion chamber, a linear alternator and an adjustable gas spring the engine design results in a compact package. In comparison to conventional combustion engines, additional degrees of freedom are available for controlling the combustion process. In this context efficiency advantages are expected due to the missing mechanical link to a crank which leads to flexibility in terms of stroke and compression ratio.</div><div class="htmlview paragraph">Applied as a range-extender-unit, the system provides additional electric energy to electric vehicles in case of discharged batteries. In addition to this automotive application, the FPLG can be used as an auxiliary power unit for example in aircrafts as well as in decentralized combined heat and power plants (CHP).</div><div class="htmlview paragraph">A step-by-step development strategy leading to the first overall FPLG system has been pursued. In the past years, the DLR has developed and experimentally verified several demonstrator prototypes for all necessary subsystems. Development systems comprising of several of the subsystems and (in some cases) of an auxiliary hydraulic actuator have been investigated.</div><div class="htmlview paragraph">Free piston specific challenges have to be addressed, e.g. increased requirements concerning steadiness of the combustion process. In addition to the hardware development, active control of piston motion is a major technological challenge. Model based algorithms for this purpose have been developed and implemented on a real time system. Applied to one of the development systems, their performance is evaluated in experiments.</div></div>