RESEARCH INTO THREE-COMPONENT BIODIESEL FUELS COMBUSTION PROCESS USING A SINGLE DROPLET TECHNIQUE

. I n order to reduce the engine emission while at same time improving engine effi ciency, it is very impor-tant to clarify the combustion mechanism. Even if, there are many researches into investigating the mechanism of engine combustion, so that to clarify the relationship between complicated phenomena, it is very diffi cult to investigate due to the complicated process of both physical and chemical reaction from the start of fuel injection to the end of combustion event. Th e numerical simulations are based on a detailed vaporization model and detailed chemical kinetics. Th e infl uence of diff erent physical parameters like droplet temperature, gas phase temperature, ambient gas pressure and droplet burning velocity on the ignition delay process is investigated using fuel droplet combustion stand. Experimental results about their infl uence on ignition delay time were presented.


Introduction
Great demands are being placed on energy production, especially with regard to environmental impact. Restrictions have been set on sulphur and nitrogen emissions from diesel engines and the Kyoto International Climate Convention limits carbon dioxide emissions. Finding new technological solutions to meet tightening emission targets creates new challenges in the development of new types of multicomponent ecological fuels. Th e objective is to minimise combustion emissions while, at the same time, increasing effi ciency of fuels used. Accurate tests on combustion and burning is the key to achieving good chamber design. However, engine development gains signifi cantly not only from tests, but also from experiments using a single droplet technique to identify the ignition delay time of various types of fuels, which characterizes burning properties. Th is kind of experiments can replace expensive trials and tests requiring stringent and diffi cult controls, saving millions of euros each year.
A multicomponent droplet [1,2] is known to exhibit a signifi cantly diff erent gasifi cation behavior compared with that of a pure fuel droplet. Th ese diff erences have been attributed [2−7] to transient liquid mass transport in the droplet interior, volatility diff erential between the constituent fuels, phase equilibrium at the droplet surface, and thermo-transport properties that are functions of mixture composition, temperature, and pressure. In order to address these complex issues for fuels in a systematic manner, we have taken the fi rst step, i.e., to study the combustion behavior of a multicomponent fuel droplet using a fuel droplet combustion stand.
Combustion of single fuel droplets of dehydrated ethanol (A), Rapeseed methyl ester (RME), diesel fuel (MD) and their blends of various proportions were studied.

Objects and methods
In the diesel engine the liquid fuel is injected into the heated compressed air, the temperature of which is 773-873 K. Sprayed fuel takes a semblance of fog. However experimentally to investigate burning peculiarities of such a small fraction using classic methods is impossible.
For the experiment we used several times bigger, one-sized droplets of unmixed fuels and also their blends of various proportions. Because of the increased fuel droplet mass, it was imposible to watch combustion process at the imitated experiment temperature equal to combustion chamber. Th e droplet simply evaporates. Th is problem was solved by increasing ambient tempera-ture (900−1100 K) to identify ignition delay time of the investigated fuel samples.
Single suspended droplet technique, collectively developed by the Institute of Combustion & Advanced Technologies, the Odessa State I. I. Mechnikov University (Ukraine) and Lithuanian Institute of Agricultural Engineering, Lithuanian University of Agriculture (Lithuania), was used for the droplet combustion tests on diff erent fuel samples and blends of diff erent composition.
Investigated liquid fuel droplet was settled with the help of medicine syringe on the suspension device (2), manufactured from glass yarn and having a 800 μm radius ball (3) on the end. At the time of experimental measuring, heated furnace (1) was pushed via defl ective rails towards suspension device in the way, that the distance between fuel droplet and thermo-couple was 2−3 mm. Th is distance was controlled by synchronising contacts (6) which also operate a stopwatch (7) switch-on. Droplet combustion control was provided by optical system, equipped with radiant (8), screen (9) and binoculars (10) (Fig. 1).
Research into fuel droplet combustion process was performed according to the following methodology. Furnace was heated until initial temperature T ∞ = 900 K. Fuel droplet (see Fig. 2) with radius r 0 = 1,125±0,05 mm was settled on the glass yarn with the help of medicine syringe. Droplet size was variable with the sprayed fuel volume and controlled looking through binoculars at the graduated screan where droplet refl ection is visible. Initial fuel droplet temperature was T 0 = 293 K (T 0 = 20 °C ).
Fuel droplet hanging on the suspension device assumes vertically outstretched form. Th is peculiarity causes a necessity to measure droplet vertical dimension. Aft er the fi xing of exact droplet size, furnace was secured atop to console of suspension device. In this way a horizontal part of suspension device appears inside the furnace and the suspended droplet position is coincident with the centre of furnace. A time that it took the droplet to reach the centre of furnace (2−3 mm from thermo-couple) did not exceed 0,5±0,05 s. Th is moment synchronized a swich-on of the stopwatch. Interval between fuel droplet reaching the centre of furnace and appearing of fl ame around the droplet was fi xated as ignition delay time τ ind .
A stopwatch also fi xated a total time of droplet existence. Droplet burning time was calculated as a diff erence between total time of droplet existence and ignition delay time. Th ermo-couple infl exible mounted into the back-wall of furnace fi xated droplet combustion temperature. Measurement error did not exceed ±5 K.
When the droplet burnt out, the furnace was pulling away via defl ective rails. Experiment with the same type of fuel sample was repeated 5−10 times at the same value of ambient temperature. Ignition delay time identifying experiment continued with increasing ambient temperature from 10 K step by step to 1100 K, performing the same procedures. Experiment error is scheduled as 2-3 %.
Aft er completing experiments with one sample in all temperature ranges, that type of fuel was replaced with another one.

Results and discussion
While analyzing experimental data results (Table 1, Fig. 3) it is evident that single droplet ignition delay time of mineral diesel fuel (MD) marginally exceeds or is equal to analogous index of RME until the temperature T ∞ = 945 K what is not coincident with the earlier done τ ind calculations using a quasistationary model. Th ough, forthcoming enlargement of temperature reversed this relation. Under a temperature T ∞ = 955 K and 965 K ignition delay time of MD fuel droplet decreased to 12-18 % comparing to the RME droplet of the same radius. To verify whether this tendency endures at the higher values of temperature became impossible, because droplets of RME fall dawn from suspension device before the combustion process begins. We can observe combustion process of MD droplets at even higher temperature volumes, until 1005 K. Th is can be explained that liquid RME has a slightly upper value of density than MD and a likely harder surface tension coeffi cient depending on temperature.  Continuing experiments of measuring MD ignition delay time near the edge of infl ammation, when fuel droplet before the moment of combustion has almost evaporated, many times we observe appearing of irradiant blue zone of infl ammation outermost from suspension device, near-by the internal walls of furnace. Irradiant blue zone appeared at distance approximately equal to 30 initial droplet radius. Th is fact proves theoretical study results about liquid fuel droplet ignition delay time analysis in near the critical conditions [8].
We must admit that critical temperature of MD showed the higher value comparing to RME, 910 K and 860 K, accordingly. Th is result coincides very well with the data presented by various authors in their works about estimation of ignition delay time of MD and RME, where single droplet radius does not exceed r 0 = 1 mm.
Combustion experiments were performed using chromium-aluminium thermo-couple.
In this way, ignition delay time calculations of MD and RME droplets using a quasistationary model qualitatively concur with experimental results, beginning from temperature T ∞ = 955 K and above. Th is value exceeds the critical infl ammation temperature of RME by 95 degrees.
One more signifi cant conclusion sequential from the quasistationary model calculations of single droplet ignition delay time analysis -additive of dehydrated alcohol (A) to MD and RME reduces ignition delay time τ ind volume of mentioned fuels. And the opposite -enlarging RME percental amount in the basic fuel blend causes an icrease of τ ind volume.
Experimental results proved that small additive quantities to the basic fuel blends practically do not cause critical infl ammation conditions. Above mentioned investigated fuel blends showed critical infl ammation temperature measurement results, which did not exceed experimental error limits.
Obtained fi ndings about diesel fuel and MD/RME blends with dehydrated alcohol (A) additive and also the single droplet ignition delay time evaluation are presented in Table 2 as compared to calculated ones using a quasistationary model (ambient temperature T ∞ = 965 K). It is evident, that presented in Table 2 experimental data qualitatively fi t in with the quasistationary model results very well. Moreover, we must admit that small quantities of dehydrated alcohol (A) additive to mineral diesel fuel, aff ect single droplet ignition delay time almost two times better than RME additive to MD (Figs. 4, 5).
While using an integrated RME/A additive to mineral diesel fuel, ignition delay time τ ind of such a threecomponent fuel blend approximates to the values of pure MD (Fig. 6).

Conclusions
1. Experimental data qualitatively fi t in with the quasistationary model results very well beginning from the ambient temperature T ∞ = 965 K and above.