Low Temperature Combustion

There are different ways to reduce combustion temperature and soot and NOx

emission and this section outline the effects of different strategies. For example, the

45

increase of boost pressure advances autoignition timing and reduces pressure rise rate (Hasegawa and Yanagihara (2003)). Reduction of inlet air temperature first decreases then increases NOx (Akagawa, Miyamoto et al. (1999), Musculus (2004)).

For valve opening and closing timing, late intake valve closing has the equivalent effect of reducing compression ratio and can help with ignition control for early injection regimes (Duffy, Faulkner et al. (2004), He, Durrett et al. (2008), Murata, Kusaka et al. (2008)). Lowering compression ratio (Laguitton, Crua et al. (2007)), coupled with increased EGR (Noehre, Andersson et al. (2006)), reduces NOx and soot emission. Parametric sweeps of engine variables have been carried out to find regimes which reduce soot, if not to provide a completely soot free combustion (Polonowski, Mueller et al. (2011)). All these enable combustion to operate in the LTC region and a conceptual model for partially premixed low-temperature diesel combustion has been developed in Musculus, Miles et al. (2013) and an overview can be found in Andersson and Miles (2014). The following variables have greater scope of reducing soot and NOx emission.

1.3.3.1 Injection

Injection variables affecting combustion in a compression ignition engine include injection pressure, duration, timing, number of injections and injector design.

Increased injection pressure has been shown to reduce smoke (see Table 1-1) as it increases mixing rate (Pickett, Caton et al. (2006)). Effects of nozzle design, configuration and size of injector orifice (see Table 1-1), injection rate shape (see Table 1-1), injection duration (Wang, Han et al. (1999)) have also been investigated.

Injection timing has an effect on combustion temperature because it changes ignition delay and mixing time affecting NOx production (Iwabuchi, Kawai et al.

(1999), Kook, Bae et al. (2005), Fang, Coverdill et al. (2009)). Effects of injection timing on auto-ignition have been studied (Kashdan and Papagni (2005)). Early, or retarded, injection also reduces soot as premixing is increased (Iwabuchi, Kawai et al. (1999), Kook, Bae et al. (2005), Musculus (2007), Fang, Coverdill et al. (2009)).

46

Variable References

Increase of injection pressure

reduces smoke

Badami, Nuccio et al. (1999), Inagaki, Takasu et al. (1999), Wang, Han et al. (1999), Dodge, Simescu et al. (2002), Hountalas, Kouremenos et al. (2003), Morgan, Gold et al.

(2003), Okude, Mori et al. (2004), Atzler, Weigand et al.

(2007), Karra and Kong (2008), Weigand, Atzler et al. (2008), Fang, Lee et al. (2010), Horibe, Tanaka et al. (2011) Nozzle design,

configuration and size of injector

orifice

(Harada, Shimazaki et al. (1998), Akagawa, Miyamoto et al.

(1999), Wang, Han et al. (1999), Dodge, Simescu et al. (2002), Okude, Mori et al. (2004), Lechner, Jacobs et al. (2005),

Pickett, Caton et al. (2006)

Injection rate shape

Wang, Han et al. (1999), Juneja, Ra et al. (2004), Tanabe, Kohketsu et al. (2005), Kastner, Atzler et al. (2006), Kastner,

Atzler et al. (2008), Pickett, Manin et al. (2013) Narrow spray angle

and retarded injection timing reduces NOx and

Soot

(Kimura, Aoki et al. (1999), Kimura, Aoki et al. (2001), Kimura (2002), Lee and Reitz (2003), Okude, Mori et al. (2004), Choi, Miles et al. (2005), Kook, Bae et al. (2005), Fang, Coverdill et

al. (2009))

Table 1-1: Summary of references relating to studies conducted on different injection variables.

There are some disadvantages of using early injection which causes fuel to stick to the wall, reduces efficiency and causes oil dilution. Early injection may form rich mixture in the squish volume, increasing HC and CO emission (Kim, Ekoto et al.

(2008)). The disadvantages of early injection may be minimised by using some of the following strategies. Lower pressure and more disperse spray angle helps with mixing and reduced wall wetting (Akagawa, Miyamoto et al. (1999), Iwabuchi, Kawai et al. (1999), Yanagihara (2001)) but a narrow spray cone for early injection was found to reduce soot and NOx emission for low flow rate injector along with optimized EGR rate and split injection strategy (Lechner, Jacobs et al. (2005)).

Cylinder liner impingement is a problem for early injection but can be adjusted

47

with a narrower spray direction (Walter and Gatellier (2002)) or employ the lower pressure early injection (Duffy, Faulkner et al. (2004), Sun and Reitz (2008)).

Lowering of compression ratio helps with controlling ignition timing when early injection is used (Walter and Gatellier (2002), de Ojeda, Zoldak et al. (2008)). The effect of compression ratio on ignition delay has been investigated (Akagawa, Miyamoto et al. (1999)).

For injection closer to TDC, narrower spray angle together with high pressure and high mixing rate reduces smoke and NOx. Retarded injection timing or near TDC injection reduces smoke and NOx due to low temperature combustion (see Table 1-1), although there is an increase of HC emission (Musculus, Lachaux et al.

(2007), de Ojeda, Zoldak et al. (2008)). Retarded injection forms over-lean regions in the squish volume and increase HC in that region (Kim, Ekoto et al. (2008)). The so-called Modulated kinetics (MK) mode employs retarded injection so mixing occurs during the expansion stroke, thereby delaying ignition (Kimura, Aoki et al.

(2001)). Although another study found that retarded injection timing reduces soot and increases NOx emission and it was possibly caused by reduced heat transfer through soot’s radiative properties and increases temperature for NOx production (Musculus (2004)).

Application of dual injection helps to reduce combustion temperature and reduce smoke and NOx such as those used in UNIBUS (Hasegawa and Yanagihara (2003)) and other systems (Yokota, Kudo et al. (1997), Yanagihara (2001), Mueller, Martin et al. (2004)). Multiple injections and their effect on pollutant formation, has also been investigated (Kastner, Atzler et al. (2006), Kastner, Atzler et al. (2008)). The importance of post injections and its mechanism of pollutant reduction has been studied (Bobba, Musculus et al. (2010), Chartier, Andersson et al. (2011), O'Connor and Musculus (2013), O'Connor and Musculus (2014)). Even multiple injection using differently orientated injections has been tried (Akagawa, Miyamoto et al.

(1999)).

48

Apart from soot and NOx emissions, unburned hydrocarbons emission is also affected by injection timing (Harada, Shimazaki et al. (1998), Iwabuchi, Kawai et al.

(1999), Kashdan, Mendez et al. (2007), Dec, Davisson et al. (2008), Ekoto, Colban et al. (2008)). Early injection increases HC emission because longer ignition delay produces lean mixtures near the injector which is too lean to combust (Lachaux and Musculus (2007)). This may be caused by the transient ramp-down at the end of injection which has a low momentum producing the excessively lean mixture which is exhausted as unburned hydrocarbons (Musculus, Lachaux et al. (2007)).

Another study found the sources of unburned hydrocarbons comes from nozzle dribble and incomplete reactions within the cylinder core and side wall above the crevice region (Ekoto, Colban et al. (2008)).

There are gaps within the literature which can be investigated. Variation of injection pressure and advanced timing, coupled with multiple injections with a post injection to identify the effects of these on in-cylinder combustion processes and soot and emissions will be useful, with specific attention paid to HC and CO productions and load limitation.

1.3.3.2 EGR Rate

EGR rate affects the amount of soot and NOx in the cylinder (see Table 1-2). The influence of EGR on auto-ignition generally has been investigated (see Table 1-2).

EGR and its effect on auto ignition and soot formation was investigated in detail in Idicheria and Pickett (2011). High level of cooled EGR controls the ignition process and prevents premature ignition when early injection is used (see Table 1-2). The modular kinetics (MK) system also employs high EGR rate to delay auto ignition.

Increased EGR rate has been shown to reduce NOx by suppressing peak combustion temperature to accommodate early or retarded injection timing (Kook, Bae et al. (2005)). There is however the problem of high CO emission for high EGR rate with retarded injection or high EGR rate in general (see Table 1-2).

As EGR is increased, soot first increases and then decreases (Akihama, Takatori et al. (2001), Idicheria and Pickett (2005), Huestis, Erickson et al. (2007)). The initial

49

increase is aided by longer soot residence time in the reacting fuel jet for moderate EGR (Idicheria and Pickett (2005)) and lower soot oxidation due to lower oxygen level (Huestis, Erickson et al. (2007)). For higher EGR, the combustion temperature is reduced so soot formation is limited (Akihama, Takatori et al.

(2001), Idicheria and Pickett (2005), Huestis, Erickson et al. (2007)). Within a diesel spray, higher EGR causes soot to form in regions further away from the injector in the radial and axial direction (Idicheria and Pickett (2005)).

Variable References

EGR rate on soot

Baert, Beckman et al. (1996), Akihama, Takatori et al. (2001), Idicheria and Pickett (2005), Atzler, Weigand et al. (2007),

Huestis, Erickson et al. (2007), Jacobs and Assanis (2007), Weigand, Atzler et al. (2008), Aronsson, Sjöholm et al. (2010) EGR rate on

NOx

Atzler, Weigand et al. (2007), Jacobs and Assanis (2007), Weigand, Atzler et al. (2008), Sjoberg and Dec (2009) EGR on

autoignition

Akagawa, Miyamoto et al. (1999), Per Risberg (2004), Kashdan and Papagni (2005), Kalghatgi, Risberg et al. (2007) EGR to

prevent premature

ignition

Akagawa, Miyamoto et al. (1999), Walter and Gatellier (2002), Okude, Mori et al. (2004), Sjoberg, Dec et al. (2007)

High CO emission

for high EGR rate

Akihama, Takatori et al. (2001), Kook, Bae et al. (2005), Noehre, Andersson et al. (2006), Atzler, Weigand et al. (2007), Kim, Ekoto

et al. (2008)) and HC emission (Noehre, Andersson et al. (2006), Atzler, Weigand et al. (2007), Kim, Ekoto et al. (2008)

Table 1-2: Summary of references relating to studies conducted on different EGR variables.

Conventional EGR operates by inducting cooled exhaust gas into the intake manifold from the exhaust. EGR can also be supercharged into the cylinder by opening of exhaust gas valve during the intake stroke utilising the exhaust blow-down pressure wave from other cylinders (Kuboyama, Moriyoshi et al. (2009)).

This creates a large thermal stratification, reduces pressure rise rate and extends

50

load range in gasoline HCCI. Simulating the effect of increase EGR has been shown to reduce soot precursor formation rate but increased the residence time for soot precursor formation (Pickett, Caton et al. (2006)). The effect of the quality of EGR has on combustion has also been investigated (Sjoberg and Dec (2009)). It was found that maximum indicated mean effective pressure (IMEP) using simulated EGR in HCCI combustion is limited by the amount of oxygen. Whereas maximum IMEP using real EGR is limited by EGR, NOx and wall heating induced run-away.

More fundamental research on EGR’s effect on combustion has been done regarding low temperature diesel combustion (Kook, Bae et al. (2005), Pickett (2005)). Studies have been done to look at the effect of different oxygen contents (Pickett (2005)) and also combined with fuel injection timing variation (Kook, Bae et al. (2005)) on combustion phasing and emission. Distinctions have been made between phases of premixed and mixing-controlled combustion at these different settings.

The use of EGR in engine is important for the reduction of smoke and NOx and the understanding of the effects of EGR on spray and droplet formation, affected by injection parameters such as injection pressure, will provide information on the fundamental mechanism of emission formation.

1.3.3.3 Lean/Rich Mixture Combustion

Lean and rich combustion can be achieved by varying EGR, injection pressure and timing to reduce NOx and soot (Okude, Mori et al. (2004), Jacobs and Assanis (2007)) but this also changes the fuel consumption (Jacobs, Bohac et al. (2005)).

Mixture stoichiometry can also be altered by changing the amount of fuel injected which affects auto ignition timing (Akagawa, Miyamoto et al. (1999)). Retarded injection near top dead centre reduces mixing and prevents over rich zone to form and reduces NOx and soot (Okude, Mori et al. (2004)). Rich diesel combustion can be achieved by increased level of EGR and reduced combustion temperature to limit soot and NOx formation (Akihama, Takatori et al. (2001), Jacobs, Bohac et al.

51

(2005)) but this increased fuel consumption (Jacobs, Bohac et al. (2005), Jacobs and Assanis (2007)).

It was found that soot formation in low temperature combustion is relatively insensitive to air/fuel ratio (Jacobs and Assanis (2007)). Similarly, lean diesel combustion also reduces NOx and soot emission but also increases fuel consumption, carbon monoxide and hydrocarbons emission (Akagawa, Miyamoto et al. (1999), Jacobs, Bohac et al. (2005)). Near stoichiometric operation along with high EGR and low compression ratio also lowers NOx and soot (Noehre, Andersson et al. (2006)). But over-mixed and over lean fuel/air mixtures were found to be the main source of HC and CO emissions (Petersen, Miles et al. (2012)), so a balance will need to be achieved. More studies are required to find ways to maintain the reduction of soot and NOx and attempt to reduce HC and CO emission.

1.3.3.4 Bowl Shape and Geometry

Bowl shape and geometry has an impact on diesel engine combustion as these affect spray dynamics, mixing and mixture formation. Bowl shape and different injection pattern from different injectors were investigated for mixing and spray wall interaction (Harada, Shimazaki et al. (1998), Akagawa, Miyamoto et al. (1999), Kashdan, Docquier et al. (2004)). A larger bowl diameter reduces impingement and provides a leaner jet but a higher proportion of the jet becomes too lean to combust with soot forming in the centre of the jets (Genzale, Reitz et al. (2008)). A smaller bowl encourages more mixing and provides a generally leaner mixture with soot forming between the jets (Genzale, Reitz et al. (2008)). Turbulence and mixing within the combustion chamber is also affected by combustion bowl design and intake geometry. Swirl affects combustion processes such as ignition delay, rate of heat release and peak heat release (Miles (2000)).

Wall temperature will have an influence on the combustion of diesel fuel in the cylinder as it provides a thermal stratification forming a boundary layer near the wall. Increased wall quenching by changing the distance between the top ring and top surface of the piston was shown to reduce CO and HC emission (Akagawa,

52

Miyamoto et al. (1999)). Changing the injection timing in a wall guided diesel combustion chamber showed large variation of HC emission. This was caused by mixing induced bulk quenching and temperature stratification within the bowl (Kashdan, Mendez et al. (2007)). Bowl geometry played a significant part in combustion and emission for late injection HCCI mode when two bowl shapes were compared (Kashdan, Docquier et al. (2004)).

Variation of bowl shape and geometry is limited in its application as usually one shape is used for all load and speed of an engine and it is difficult for it to be optimised for all conditions. Further work can be done on the variation of injection pressure and combination of bowl shape and dimension to better understand the optimal combination of the two in terms of wall wetting and HC and CO emission.

The increase in turbulence by swirl and bowl design can also be included for further study.

1.3.3.5 Fuel Properties

Research into properties of diesel and other fuels has been investigated extensively for the purpose of diesel combustion control (Siebers (1985), Kalghatgi, Risberg et al. (2003), Kalghatgi and Head (2004), Per Risberg (2005), Pickett and Hoogterp (2008)). Computer simulation which makes use of detailed or global chemical kinetics is useful because modern fuels are blended with a large number of different hydrocarbons. Simplification made to the chemistry is necessary in order to produce a model which is close to the properties of the fuel being simulated, while retaining a necessary level of complexity to reproduce the underlying chemistry is sufficient detail..

Investigation of the auto-ignition properties of fuel is useful because auto-ignition is important for HCCI/low temperature combustion operations so the combustion can be controlled adequately. Changes to the fuel used will have an effect on auto ignition consequent on changes to the chemical kinetics (Hwang, Dec et al. (2008), Morgan, Smallbone et al. (2010)) and experiments on autoignition of different fuels have been carried out (Akagawa, Miyamoto et al. (1999), Kalghatgi, Risberg

53

et al. (2007), Kalghatgi, Risberg et al. (2007), Fikri, Herzler et al. (2008), Kalghatgi, Hildingsson et al. (2009)). The effect of fuel chemistry on combustion phasing has been investigated as a way to control ignition using fuel stratification (Dec and Sjoberg (2004), Kokjohn, Reitz et al. (2012)).

Blends of primary reference fuels (PRF) are a useful way, and perhaps the only way, to simplify the chemical kinetics involved in commercial fuel blends. The balance to be struck is to avoid the number of species and reactions becoming too numerous and computationally expensive but detailed enough to represent the fuel adequately. Also, application to internal combustion engines requires a more specific approach than general chemical kinetics. Specific models matching the range of conditions expected in an engine will be more productive. In addition, chemical kinetics provides information as to whether a particular reaction takes place under certain conditions (Andrae, Johansson et al. (2005), Naik, Pitz et al.

(2005), Andrae, Bjornbom et al. (2007), Andrae, Brinck et al. (2008), Smallbone, Morgan et al. (2010)).

Investigation of fuel properties is important to understand their effect on autoignition and combustion. But given commercially available fuels are generally manufactured to a specific standard and composition, its application to engine design is limited unless a major stakeholder, such as a large oil company, is involved to provide the infrastructure of a new fuel product. Nonetheless, this can be further developed and it will be discussed in more detail in Section 1.3.5 on gasoline compression ignition engine.