Search Results

In this paper, dependence of liquid-DME viscosity on temperature and pressure was studied theoretically. It was found that in the saturated-liquid state, the DME viscosity is 0.37 cSt at - 40 ° C and it drops to 0.17 cSt when temperature increases to 80 ° C. In the subcooled-liquid state, viscosity varies linearly with pressure at a given temperature; at 20 ° C, viscosity of the subcooled liquid is 0.23 cSt at 5.3 bar and it increases to 0.33 cSt at 500 bar. The predicted liquid-DME viscosity and its pressure dependence agree with those obtained by measurement. Lubricity of liquid DME also was studied. Polar-headed, long-chain alcohols and fatty acids with chain length of C15 ∼ C22 were found to be candidates of lubricity additives for DME. Castor oil (chemically, it is basically a C18 fatty acid) was found to be a good additive for improving the DME lubricity.

On the basis of the molecular thermodynamics for fluids, the thermodynamic properties of DME are developed for pressure p ≤ 500 bar and temperature T ≤ 200 °C, which covers pressures and temperatures that a DME fuel system for the CI-engine application would experience. The properties cover subcooled, two-phase, and superheated/supercritical regions, including p-v-T properties, enthalpy, entropy, latent heat, heat capacity, speed of sound in vapor, liquid and two-phase mixtures, bulk modulus, and surface tension. A volume-cubic equation of state for DME also is developed, which allows calculating the DME density at any given pressure and temperature analytically. All the properties are given in equations as well as in charts. For convenience in two-phase-flow applications, e.g., design of the fuel tank and cavitation analysis, the saturated properties are also given in tables, listed in both pressure and temperature up to the critical point.

This paper analyzed chemical and thermophysical properties of dimethyl ether (DME) as an alternative fuel for compression-ignition engines. On the basis of the chemical structure of DME and the molecular thermodynamics of fluids, equations have been developed for most of the DME thermophysical properties that would influence the fuel-system performance. These equations are easy to use and accurate in the pressure and temperature ranges for CI engine applications. The paper also pointed out that the DME spray in the engine cylinder would differ significantly from that of diesel fuel due to the thermodynamic characteristics of DME. The DME spray pattern will affect the mixing and combustion processes in the engine cylinder, which, in turn, will influence emissions from combustion.

Turbocharged gasoline direct injection (TGDI) engines can achieve a very high level of brake mean effective pressure and thus the engines can be downsized. The biggest challenge in developing highly-boosted TGDI engines may be how to mitigate the pre-ignition (PI) triggered severe engine knocks at high loads and low engine speeds. Since magnitudes of cylinder pressure fluctuations during aforementioned engine knocks reach those for peak firing pressures in normal combustion, they are characterized as super knocks. It is widely believed that the root cause for super knocks is the oil particles entering the engine cylinder, which pre-ignite the cylinder mixture in late of the compression stroke. It is neither possible nor practical to completely eliminate the oil particles from the engine cylinder; a reasonable approach to mitigate super knocks is to weaken the conditions favoring super knocks.

The relationship between fuel dilution of the crankcase oil and low-speed pre-ignition (LSPI) was studied experimentally with a highly-boosted 1.8L turbocharged gasoline direct injection (TGDI) engine fueled with RON93 gasoline. It was found that properties of oil particles entered the engine cylinder were affected significantly by fuel dilution. The gasoline content in the oil represents those with long carbon chain or heavy species in gasoline, with much lower boiling points and auto ignition temperatures than those for the undiluted engine oil. Thus, dilution of the engine oil by these gasoline species lowers the volatility and the minimum auto ignition temperature of the engine oil. With 15% fuel content in the oil, the flash point and the fire point of the SAE 5W30 oil dropped from 245 °C to 90 °C and from 265 °C to 150 °C, respectively.

The fuel saving benefit is analyzed for a class-8 truck diesel engine equipped with a WHR system, which recovers the waste heat from the EGR. With this EGR-WHR system, the composite fuel savings over the ESC 13-mode test is up to 5%. The fuel economy benefit can be further improved if the charge air cooling is also integrated in the Rankine cycle loop. The influence of working fluid properties on the WHR efficiency is studied by operating the Rankine cycle with two different working fluids, R245fa and ethanol. The two working fluids are compared in the temperature-entropy and enthalpy-entropy diagrams for both subcritical and supercritical cycles. For R245fa, the subcritical cycle shows advantages over the supercritical cycle. For ethanol, the supercritical cycle has better performance than the subcritical cycle. The comparison indicates that ethanol can be an alternative for R245fa.

A new fuel injection strategy is proposed for DME engines. Under this strategy, a pre-injection up to 40% demand is conducted after intake valves closing. Due to high volatility of DME, a lean homogeneous mixture can be formed during the compression stroke. Near TDC, a pilot injection is conducted. Combined fuel mass for the pre-injection and pilot injection is under the lean combustion limit of DME. Thus, the mixture is enriched and combustion can take place only in the neighborhood of sprays of the pilot injection. The main injection is conducted after TDC. Because only about half of the demand needs to be injected and DME evaporates almost immediately, combustion duration for the main injection plus the unburnt fuel in the cylinder should not be long because a large portion of the fuel has been premixed with air. With a high EGR rate and proper timing for the main injection, low temperature combustion could be realized.

A variable-displacement, 275-bar dimethyl-ether pump for a common-rail injection system has been developed successfully. The pump is an inlet-throttled, wobble-plate-actuated, multi-plunger system. Results of the pump tests/simulations show that the pump can deliver fuel according to the engine requirement at different speeds due to its variable-displacement feature, which is obtained by controlling the discharge phase angle via the two-phase filling characteristic of the pump. Although the pump is designed for dimethyl ether, its concept is general and thus may be applied to the common-rail systems for other fuels.

A novel fuel tank for storing liquid dimethyl ether (DME) has been developed. This fuel tank was made of cast aluminum with a water capacity of 40 liters. It contains two fluids: liquid DME and a vapor-liquid mixture of propane. A diaphragm separates the two fluids. The propane in the tank is a pressurizing fluid that pressurizes DME into a subcooled-liquid state; and, it also functions as a driving fluid that pumps the liquid DME from the tank to the injection pump using its vapor pressure. These features characterize the tank as a thermodynamic pump. Several hundred hours of tank tests at various temperatures have been conducted. Results of tank filling-discharge cycles simulating those in vehicle applications demonstrated that the concept of the two-fluid thermodynamic pump works and that the tank design is successful.

Compression ignition delay of DME is studied theoretically. Physical phenomena that would influence the ignition delay, characteristics of the DME spray and evaporation of DME droplets in the spray, are analyzed. It is found that the short ignition delay of DME revealed in engine tests is due largely to the short physical delay of DME: The evaporation rate of DME droplets is about twice that of diesel-fuel droplets at the same cylinder condition and, the stoichiometric mixture in a DME spray can be established immediately - in comparison, the stoichiometric mixture in a diesel-fuel spray cannot be established before temperatures of diesel-fuel droplets become higher than 225 °C. The high droplet evaporation rate of DME is also responsible for the irregular boundary and tip of the DME spray as observed by many investigators. On the basis of experimental data reported in the literature, cetane number of DME is estimated to be 68.

A comparative study of characteristics of diesel fuel and dimethyl ether sprays was conducted on the basis of momentum conservation. The analysis reveals that the DME spray in the diesel combustion system may not develop as well as that of diesel fuel at high engine loads and speeds due primarily to the following reasons. (1) Because 42% more fuel volume must be injected into the engine to reach the diesel-fuel equivalent and because the DME injection pressure is lower than that of diesel fuel, longer injection duration for DME is needed even if with the enlarged orifice diameters.

In order to improve low speed torques, turbocharged gasoline direct injection (TGDI) engines often employ scavenging with a help of variable valve timing (VVT) controlled by the cam phasers. Scavenging improves the compressor performance at low flows and boosts low-speed-end torques of the engines. Characteristics of the engine combustion in the scavenging zone were studied with a highly-boosted 1.5L TGDI engine experimentally. It was found that the scavenging zone was associated with the highest blowby rates on the engine map. The blowby recirculation was with heavy oil loading, causing considerable hydrocarbon fouling on the intake ports as well as on the stem and the back of the intake valves after the engine was operated in this zone for a certain period of time. The low-speed pre-ignition (LSPI) events observed in the engine tests fell mainly in the scavenging zone.

Emissions from CI engines fueled with dimethyl ether (DME) were discussed in this paper. Thanks to its high content of fuel oxygen, DME combustion is virtually soot free. This characteristic of DME combustion indicates that the particulate filter will not be needed in the aftertreatment system for engines fueled with DME. NOx emissions from a CI engine fueled with DME can meet the US 2007 regulation with a high EGR rate. Because 49% more fuel mass must be delivered in each DME injection than the corresponding diesel-fuel injection, and the DME injection pressure is lower than 500 bar under the current fuel-system technology, the DME injection duration is generally longer than that of diesel-fuel injection. This is unfavorable to further NOx reduction. A multiple-injection strategy with timing for the primary injection determined by the cylinder temperature was proposed.

The biggest challenge in developing Turbocharged Gasoline Direct Injection (TGDI) engines may be the abnormal combustion phenomenon occurring at low speeds and high loads, known as low-speed pre-ignition (LSPI). LSPI can trigger severe engine knocks with intensities much greater than those of spark knocks and thus characterized as super knocks. In this study, behavior and patterns of LSPI were investigated experimentally with a highly-boosted 1.5L TGDI engine. It was found that LSPI could occur as an isolated event, a couple of events in sequence, or a trail of events. Although occurring randomly among the engine cylinders, LSPI took place frequently when the engine was operated at low speeds and high loads in the zone where scavenging was employed for boosting engine torques at low speeds, typically < 2500 rpm.

Waste heat recovery (WHR) has been recognized as a promising technology to achieve the fuel economy and green house gas reduction goals for future heavy-duty (HD) truck diesel engines. A Rankine cycle system with ethanol as the working fluid was developed at AVL Powertrain Engineering, Inc. to investigate the fuel economy benefit from recovering waste heat from a 10.8L HD truck diesel engine. Thermodynamic analysis on this WHR system demonstrated that 5% fuel saving could be achievable. The fuel economy benefit can be further improved by optimizing the design of the WHR system components and through better utilization of the available engine waste heat. Although the WHR system was designed for a stand-alone system for the laboratory testing, all the heat exchangers were sized such that their heat transfer areas are equivalent to compact heat exchangers suitable for installation on a HD truck diesel engine.

Good understanding of fuel sprays in the engine cylinder is crucial to optimizing the operation of direct injection gasoline engines. In this paper, a detailed analysis is conducted on direct gasoline injection sprays from a multi-hole injector. Penetrations and angles of the sprays are characterized with a homogeneous model for the fuel spray. The drop size distributions in the sprays are analyzed using an empirical distribution model. Predicted spray penetrations, spray angles, and drop size distributions under three different injection pressures are compared with the measurements for injection pressures = 40, 100 and 150 bar and good agreements are observed. Transient evaporation rates are also studied for fuel drops in an environment simulating the cylinder condition during the intake stroke of a direct injection gasoline engine.

The present paper is Part II of an investigation on the influences of the late intake valve closing (LIVC) and the early intake valve closing (EIVC) on the engine fuel consumptions at different loads and speeds. The investigation was conducted with two 1.5L turbo-charged gasoline direct injection (TGDI) engines, one with a low-lift intake cam and the other with a high-lift intake cam. The focus of this paper is the cylinder charge motion. Computational fluid dynamic (CFD) analyses were conducted on the characteristics of the cylinder charge motion for the load points 6 bar-bmep / 2000 rpm, 12 bar-bmep / 3000 rpm, and 19 bar-bmep / 1500 rpm, representing naturally aspirated and boost-mode operations without and with scavenging during the valve overlap.