Low fuel consumption is one of the greatest merits of the Diesel engine, emission of nitrous oxides, NOx, and particulate matter, PM, are its most prominent disadvantages. For modern passenger car diesel engines, to achieve low fuel consumption targets, i.e. CO2 emissions, and to simultaneously fulfil the ambitious EU6 emissions legislation, a combination of intelligent injection strategies and high EGR rates with appropriate boost pressures is required. Generally, NOx emissions can efficiently be decreased via exhaust gas recirculation, EGR. An unwanted consequence of this is an increase in smoke, HC and CO emissions. This can be alleviated by a suitable application of injection strategies. Injections strategies are also employed to curb excessive emissions of unburnt hydro-carbons, HC, and carbon monoxide, CO, particularly in conditions of cold start. These strategies may also be used to facilitate fast light-off of the oxidation catalyst. In the current work the novel directly actuated piezo injector of Continental Automotive GmbH was used to evaluate the emissions potential not only of multiple injection patterns, but also of advanced rate shaping strategies. The application of this wealth of technology puts further cost onto the already expensive Diesel engine. Nevertheless, there is a need to explore the benefits and the limits of such technology, since the fuel consumption and CO2 emission target of 120g/km across the European passenger car fleet can only be achieved with a sufficiently high share of Diesel vehicles, also in the most dominant market segment of small and medium sized cars. Cost in this respect needs to be considered from an overall engine system point of view. Higher cost for measures to reduce engine out emissions may be well invested if exhaust gas aftertreatment can be avoided or at least be minimised.
The advent of the EU5 and EU6 [
Both, downsizing and downspeeding, require the increase of engine torque to provide the desired power for pleasurable driving. As a consequence engine boosting has to be increased massively and, to limit the resulting NOx production, simultaneously EGR rates have to be raised [3, 4, 5].
The presence of large portions of more or less inert gas in the combustion chamber converts the conventional Diesel combustion process such that prolonged ignition delays allow for a more extended mixture preparation period before ignition [6]. In contrast to the predominantly premixed "homogeneous charge compression ignition", HCCI, the combustion initiation still is controlled by injection timing. This means, that at a desired time with respect to engine top dead centre at least the minimum critical mass of fuel, necessary to create loci of ignition, is introduced. Nevertheless, both techniques share to some extent a problem, also well known from premixed gasoline engine: the presence of premixed fuel and air close to the combustion chamber walls, where wall quenching leaves a layer of unburnt mixture. If additionally the mixture is overly lean, the problem of hydrocarbons leaving the engine unused is aggravated by flame quenching in mid air. This can be overcome to some extent by stratifying the in-cylinder charge by means of smart injection strategies [7, 8, 9]. These strategies have to respect both boundary conditions, that of the location of the fuel in the combustion chamber and the available time for mixture preparation before the onset of the main ignition.
Once ignition has occurred, the feeding of more fuel to the combustion has to be done such, that excessive peak temperatures are avoided, since this helps to reduce NOx.
At the end of the combustion process,
incylinder pressure and temperature drop, and
charge motion decays. There, a post injection
introduces some additional momentum for mixing
and some thermal energy for further oxidation of
carbonaceous species [
State-of-the-art injection systems allow for the
application of multiple injections per working cycle
of the engine. Beside conventional injection
strategies, "Boot" and "Ramp" injections are discussed
in recent publications [
Premixing before ignition is limited on the one hand by the local extension of the cloud of gaseous fuel, propagating from each injected jet into the combustion chamber. The proper distribution of the fuel cloud requires a suitable combination of injection pressure, charge movement and charge density, as well as a suitable shape of the spray plume. On the other hand injection timing is crucial, since the formation and propagation of the air fuel mixture is a function of time.
These temporal and spatial restrictions for mixture preparation and ignition are dictated by the operating condition of the engine. The time available for propagation and mixing of fuel shrinks
Conclusively, it can be summarised, that injection strategies serve the following purposes: pilot main post time (b) Boot injection (c) Ramp injection linearly with increasing engine speed, and the ignition delay is determined by the pressure and temperature rise during the compression phase of the engine. Also, at higher engine speeds the heat loss to the walls will be smaller, i.e. the gas temperature for a given crank angle position will be higher, which shortens the ignition delay. Additionally, when the engine is boosted at higher loads, there will be a higher gas mass present in the cylinder. This increases compression pressure and, hence, the gas temperature. Therefore, there will be a limit in terms of engine speed and load up to which pre-main injection strategies can sensibly be applied. Figure 2 provides an overview of the boundary conditions for mixture formation and combustion at three engine operating points. Clearly visible is the drastic shortening of the ignition delay with increasing in-cylinder pressure and temperature.
N / rpm
pRail / bar pBoost / bar
Ignition delay / ms / °crk 1500 4.2 750 1.06 12.7 1.4 13 1.0 (0.5+0.5) 1.0 1500 6.5 900 1.16 13.5 0.66 6 2280 10.5 1650 1.7 20 1.0 0.23 3.2 TGas_Max / K 1760 1890 1990 Figure 2: Thermo-physical boundary conditions for mixture formation and chemical reaction at three engine operating points.
An additional limitation for the use of any injection strategy is engine power output. When high power output is demanded, then thermodynamics require the necessary amount of fuel to be injected within a sensible time window, unless a loss in engine efficiency, i.e. the deterioration in fuel consumption is accepted. Additionally the resulting extreme exhaust temperatures pose severe problems for engine and turbo charger.
The current paper will elucidate basic mechanisms of injection strategies, including multiple injection, MI, and rate shaping. For this the directly actuated Continental NG injector was applied in a single cylinder research engine, based on a 2 liter 4 cylinder and 4 valve engine. The influence of engine operating conditions on the use of injection strategies will be discussed.
Pilot injections have different purposes depending on the operating point. At low load they serve for the reduction of HC and CO emissions (see Figure 3). At low engine speed, small pilot injections have time to mix well in the combustion chamber. Because the general temperature level is 1 0.8 0.6 0.4 0.2 0 0 5 10 15 20
NOx [g/h] 0 5 10 15 20
NOx [g/h] 0 5 10 15 20
NOx [g/h] 40 35 30 ]h25 / [g20 C15 H10 5 0 0 5 0.510 15 0.2500.5
NOx [g/h] - 30% Figu5r0e 4: Effect of pilot injectio2n0fu0el mass on pollutant emissions and combustion noise at n = 2280rpm, 40 IMEP = 10.516b0ar 150 120 ]/h 90 g [ O 60 C 30 0 ] ] h230 0 N5OBx1o0[og/[hg/t31h0i5]nj2e0ction ]/[hg120 kW 2 ]khW222435000 ttehwveeeTenxNOm2hnc0aelpionaislopientprjelaicccnoatdiutoiponmlnimnaogianfyotifnhojeftehfecesrtCOoiopf-u48ncrre00ats-hllmueegdragiPnzeMesfrutoeree-dldd,mwuthceatalsilotsbnaetsno-. I[/gFSC222012000 010 20 30 [/g24200 50 Figur1e0105 2sh0ow30s t4h0e 5a0pplicatio01n0 o2f0bo3o0t i4n0jec5ti0on NOx [gFC/2h1]0 strategies NwOitxh [gd/ihff]erent boot lengNthOsx a[gn/dh] levels IS200 (see Figure 1). The pre-main fuel mass was app. 1900 5th1e0s1a5me20for all shown injection patterns.
NOx [g/h] low at low load, the pre-ignition chemistry is slow.
Therefore a relatively large amount of fuel can be introduced into the combustion chamber. In order to prevent early ignNitGioBn0.2-oNf13-0.5-x-0p.8re;p-_mraial[ibnar]i=n1je65c0ted
the
NGB0.2-N13-0.8-x-0.8; p_rail [bar] = 1650 fuel, it has to be intrNoGdBu0.c2-eNd13-i1n.2-x-0.8; p_rail [pbaorr]t=io16n5s0 at
several sufficiently long timeN GinBt0e.2r-vNa13ls-0..3-O0.3th-xe-0r.8w;pis_reail t[bhaer] =c1r6i5ti0cal fuel concentration for ignition will be exceeded.
If on the other hand the injection sequence starts too early before ignition, wall and mid-air quenching will be the result, which increases the HC and pRail [bar] = 750 CO em6issions massively. 40 350 +M; pRail [bar] = 750 Figu]5re 3 depicts the effect of the number of 3p4il0ot +P3+M; pRail [bar] = 750 injectiorckns on HC and CO 3e0missions. The cu]r33v23e00s show /r°a4n EGR rate varia/htion, with EGR rat e[°C31i0n
] crease[bs3 from right to left in[xg2t0he diagram. PM exhm30i0sa i 12 E290 sions whp2ere on a very lowOlevel for all shown_2s8e0ttings a/pdn1d far below EU6 enN1g0ineeri1n0g target. T270 d 260 0 250 Legend 0.5
0.8 12 10
At higher part load PM and NOx emissions are of main concern. There, the local temperature is high enough for good oxidation of HC, but also is favourable for NOx generation. The equivalence ratio is locally rich, promoting soot production.
Shown in Figure 4 is an EGR trade off. PM emission and combustion noise are plotted over NOx emission. It is clearly demonstrated, that the increase of the pilot fuel mass leads to a significant PM increase, but also to a reduction in combustion noise. Therefore, the trade off at this operating point is between PM and combustion noise. Both are a strong function of the gas temperature in the combustion chamber before the onset of the main combustion, Tbm [9]. In [9] it was also found - by means of 3D combustion simulation - that very small dwells between pilot and main injection would reduce PM emission further. boot+-lowpost
MI boot -lowboot -high-20%
+EGR
Both boot patterns without post injection achieve about the same emissions level as the optimised MI pattern. However, the MI pattern also included a post injection. If a post injection is combined with the better boot strategy, the PM-NOxtrade off improves beyond that achievable with MI.
A further option to modify the injection rate is the application of a ramp injection. This is the continuous increase of the initial injection ramp, rather then the stepped shape used in the boot injection. Shown in Figure 6 is the comparison of a slow injection rate increase (blue with circles) versus a fast one (black with triangles). M6 P 0 10 20 30 40
NOx [g/h] 0 ]h ]h 6700
At higher p/[ga10rt load the "fast ram/[gp5"0offers an advantage in thCe PM-NOx-trade offO. 4H0owever, this effect is also dHe5pendant on the opeCra3t0ing point and further work is required to fully exploi2t 0its benefits. 10 0 0 2.8 ] /h2.6 g [ 2O2.4 C 2.2