Development of an HCCI engine for DME application

An HCCI engine combustion concept was evaluated theoretically and experimentally with dimethyl ether (DME) as the primary fuel. DME has very low lubricity and combustion concepts with low injection pressures, like HCCI, are thus preferable since it is difficult to design reliable high pressure DME fuel pumps. The idea behind the study was also that one of the main problems with HCCI operation, namely obtaining correct combustion phasing, could be addressed by mixing a high cetane fuel (DME) with a high octane fuel (methanol). This mixing should take place in the intake manifold just prior to induction of the charge into the engine cylinder in a variable and appropriate ration in order to provide combustion phasing control. Using methanol as the high octane fuel is particularly interesting since DME is typically produced by dehydration of methanol. therefore DME and methanol constitute a cost-effective two-fuel solution for HCCI engines. The main research focus was on combustion phasing control (auto-ignition) and combustion noise. A reduced kinetic scheme for oxidation of DME and methanol was developed based on the detailed scheme for combustion of DME by Lawrence Livermore. The scheme could theoretically validate the concept of methanol for retardation of auto-ignition. It was also used for CFD simulations of combustion knock. Experiments with combustion phasing control were conducted on a truck-sized diesel engine at NTSEL (National Traffic Safety and Environmental Laboratory) in tokyo, Japan. It was verified that combustion phasing could be retarded with methanol added to the inlet air. Maximum engine load in HCCI-mode was 50% of the maximum load for the engine in diesel-mode. Fuel consumption for the two modes of operation was comparable. EGR also had a retarding effect on combustion phasing and the equivalence ratio could be brought close to stoichiometric without compromising combustion efficience when using EGR. Experiments concerning combustion generated acoustic noise were conducted on a smallelr two-cylinder diesel engine at DTU in Lyngby, Denmark. Even though results indicated that multiple smaller combustion chambers in the piston top reduced acoustic noise the largest noise reduction was achieved with a traditional bowl-in-piston diesel engine design. Increased crevice volumes resulting from applying additional combustion chambers in the piston also reduced the thermal efficiency of the engine. Student projects at DTU, performed after the start of this project, showed that it is possible to design a DME fuel pump that could supply up to 150 bars of pressure without resorting to exotic and thereby expensive coatings of the contact surfaces. Based on these findings, the project group then decided to perform experiments with late injection HCCI operation. Late injection HCCI is a process where the fuel is introduced into the engine cylinder towards the end of the compression stroke but with fuel injection being completed before the onset of combustion. This is in contrast to normal diesel operation where the fuel is burning as it is introduced into the engine cylinder. In this way sufficient time is available before ignition to achieve a high degree of premixing with air and thereby achieve HCCI-like operation. Injection pressures of approximately 100 bars are necessary in order to do this. In these experiments pure DME was used. No methanol addition. Results have been high engine efficiencies and low NOx emissions but even though combustion phasing is controllable, injection timing is critical. Free control of the injection event is necessary to obtain satisfactory operation which in turn means that some form of common rail system is required. At the low injection pressures required (