Blade Electric Vehicles Conversion of the Hyundai Getz Picks Up Momentum
France, Italy Considering PSA-Fiat Merger to Survive Global Auto Crisis

DME Low-Temperature Combustion Concept Maintains Low NOx with Decreases in CO and HC Emissions

Researchers at Shanghai Jiao Tong University are exploring a new combustion concept for dimethyl ether (DME): low-temperature combustion (LTC) of a compound charge combining port aspiration and in-cylinder direct injection (DI). In comparison to a DME homogeneous charge compression (HCCI) combustion mode, DME LTC can extend the engine operating range with little change in NOx emissions and a considerable reduction in HC and CO emissions. A paper on their work was published online 8 December in the ACS journal Energy & Fuels.

The same research team earlier this published a paper on the development of a DME compound charge compression ignition (CCCI) process. (Earlier post.) The CCCI combustion process consists of HCCI combustion, premixing combustion, and diffusion combustion. The combustion characteristics are mainly decided by the premixed fuel ratio and CO2 concentration in the air charge.

There is particular interest in China in exploring the use of DME—an LPG-like synthetic fuel that is produced through gasification of coal or various renewable substances—as a substitute for diesel. (Earlier post.) Numerous studies have shown that diesel engines fueled with DME can achieve high thermal efficiency, ultra-low emissions, and soft and smokeless combustion.

Researchers have found, however, a tradeoff exists between NOx emissions and thermal efficiency for a diesel engine fueled with DME using conventional in-cylinder DI combustion. HCCI combustion with DME shows very low NOx emissions, but CO and HC emissions turn out to be high.

It can be found that conventional in-cylinder DI combustion and HCCI combustion with DME have opposite advantages and disadvantages.

On the basis of the characteristics of HCCI combustion and conventional in-cylinder DI combustion for a diesel engine fueled with DME, a new combustion concept, namely, low temperature combustion (LTC) of compound charge with port aspirated DME and in-cylinder injected DME, is proposed in this paper.

The LTC cycle differs from the ideal diesel engine cycle in that the isobaric combustion process is replaced with an isothermal combustion process during the power stroke, which significantly lowers the peak cylinder temperature. To make the actual engine cycle close to this theoretical LTC cycle, a proper fuel provision strategy must be applied in consideration of physicochemical properties of DME.

—Zhang et al. 2008

In the study, the team aspirated a portion of the fuel is into the combustion chamber via the air intake port to initiate HCCI combustion at the compression stroke. The rest is injected by a conventional in-line pump. The resulting combustion includes HCCI combustion initially, with in-cylinder spray combustion later.

DME HCCI combustion results in very low NOx emissions, and the in-cylinder injection can offer more engine output. The evaporation of fuel drops in DME spray is faster after HCCI combustion, shortening the ignition delay period and combustion duration and therefore suppressing the NOx formation in the phase of mixing controlled combustion. Relatively high levels of CO and HC after HCCI combustion are further oxidized during the in-cylinder spray combustion. Remaining can be treated with an oxidation catalyst converter (DOC). The study focused on improving combustion and reducing emissions by regulating port aspirated DME and DI fuel injection timing.

The team found that:

  • For a fixed port-aspirated DME mass, the peaks for gas temperature and pressure increase with the rise of direct-injection DME mass. The ignition timing of both cool-flame and thermal-flame reactions advances, and the ignition timing of the diffusion combustion slightly advances. The peak of a cool flame shows no change, and the thermal-flame peak value increases.

  • At the same engine load, with an increase in the DME mass via port aspiration, the peaks for gas temperature and pressure increase and the start of the thermal-flame reaction and the diffusion combustion advances.

  • For a fixed port-aspirated DME mass, HC and CO emissions decrease while NOx emissions increase as the load increases. At the same engine load, HC emissions decrease slightly and CO emissions first increase and the decrease with the increase of port-aspirated DME. Meanwhile, NOx increases slightly while the maximum is below 30 ppm.

  • At the same engine load, NOx emissions are lower with the LTC method than that of the conventional in-cylinder spray combustion.

  • For a fixed port-aspirated DME mass, the indicated specific fuel consumption decreased with the increase of the engine load. At the same engine load, the indicated specific fuel consumption increases with the rise of port-aspirated DME mass.

  • In comparison to HCCI combustion, DME LTC can extend the engine-operating range with little change in NOx emissions and a considerable reduction in HC and CO emissions.


  • Zhang Junjun, Qiao Xinqi, Wang Zhen, Guan Bin, and Huang Zhen (2008) Experimental Investigation of Low-Temperature Combustion (LTC) in an Engine Fueled with Dimethyl Ether (DME). Energy Fuels, Article ASAP doi: 10.1021/ef800674s


Pao Chi Pien

Any combustion process of the internal combustion engines must obey the first law of thermodynamics as follows:


The first law of thermodynamics relates changes in enthalpy to heat and work transfer interactions. At the beginning of a compression process, the whole cylinder gas mixture is divided into many small cubes by imaginary x, y, and z planes. The value PV/T of a unit mass is computed. Each cube and the remaining cylinder mixture form a thermodynamic system. The first law of thermodynamics requires that this value of PV/T does not change throughout a cycle. One arbitrary cube is made red. During an adiabatic compression process, the heat and work transfer between the red cube and adjacent cubes change signs and thus cancel each other. Two equations P2/P1 = (V1/V2)k and T2/T1 =(V1/V2)k-1 of the first law guarantee that PV/T does not change to make the net enthalpy increase of the gas mixture exactly equal to piston work done on the gas mixture.

The shape, mole number, and molecular weights of the red cube are changing continuously as it going through a combustion process. The heat transfer and work done between adjacent cubes change signs and thus cancel each other. The fuel chemical energy converted into heat energy Q+ is equal to the product of the fuel burned per cycle times the fuel heating value.

For a constant volume combustion process, the combustion temperature increase is equal to Q+/cv where cv is the heat capacity of the gas mixture at the end of a combustion process. Then the combustion pressure is computed from the values of T and V to make sure that the value of PV/T does not change so that not energy is created or destroyed. Because PV/T is a constant though out the whole cylinder mixture all the time, the changing composition of the gas mixture has no effect on the applicability of the first law. This fact greatly simplifies the analysis of a combustion process of the internal combustion engine.

The combustion pressure equalizes quickly, while the combustion temperature may vary with local fuel equivalence ratio and the earlier burned mixture will be compressed to a high temperature as the combustion pressure increases. Only for the homogeneous charge autoignition combustion, the combustion temperature is also uniform. During an adiabatic expansion process, the heat and work transfer between adjacent cubes change signs and thus cancel each other. Two equations P2/P1 = (V1/V2)k and T2/T1 =(V1/V2)k-1 where V1 is the cylinder volume at the end of the combustion process and V2 is the cylinder total volume, guarantee that the net enthalpy decrease of the gas mixture is equal to piston work output.

It should be noted that the pressure decreases is much faster than that of temperature T. When the pressure P reaches the atmospheric pressure as in the Atkinson cycle, the temperature T is still much higher than the atmospheric temperature. The enthalpy of the exhaust gas is rejected from the cylinder at the end of expansion process. The heat removal Q- required to reduce the enthalpy of cylinder gas mixture at the end of expansion process to that of the ambient air is computed. The thermal efficiency is equal to (Q+ - Q-)/Q+ by definition and has nothing to do with the second law of thermodynamics.


I'll take your word for it Pao Chi Pien.
I am glad someone hit this early - it did sound like junk science to me.


I'm not sure what Pao Chi Pien is getting at, as it does not directly address anything in the article.

The article makes plenty of sense to me.  HCCI is known to have limited utility at higher BMEP because the temperature rise for uniform combustion of the homogeneous charge is high while pressures are high, causing both high NOx and high NVH.  Using HCCI to provide the initial heat for the charge with DI of DME to burn during the expansion stroke limits the peak temperature and NOx formation while the HCCI's heat cuts the ignition delay of the DME.  The researchers have found that the later injection of DME reduces HC and CO, which is an advantage I did not find obvious a priori but which seems logical in retrospect.

'Tis a pity that I'll probably never be able to buy a DME port-injection system for my TDI to aid in cold starting and emissions reduction. ;)

Andrey Levin

I believe that GCC authors did not get the essence of the technology.

EGR-diluted charge of fresh air is mixed with gaseous DME, admitted and compressed in the cylinder to sub-critical HCCI state. Then direct injection of liquid DMI initiates local combustion, and consequent pressure wave initiates cold combustion of homogenous charge (thus injection event is positively controlling ignition timing). Ability to inject afterwards additional amount of fuel is an additional benefit of the technology.

Seems pretty neat.

The comments to this entry are closed.