10.4271@2013 01 0917

Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Friday, August 10, 2018
2013-01-0917
Published 04/08/2013
Copyright © 2013 SAE International
doi:10.4271/2013-01-0917
saeeng.saejournals.org
Post Injections for Soot Reduction in Diesel Engines: A Review

of Current Understanding
Jacqueline O'Connor and Mark Musculus
Sandia National Laboratories
ABSTRACT
This work is a technical review of past research and a synthesis of current understanding of post injections for soot
reduction in diesel engines. A post injection, which is a short injection after a longer main injection, is an in-cylinder tool
to reduce engine-out soot to meet pollutant emissions standards while maintaining efficiency, and potentially to reduce or
eliminate exhaust aftertreatment. A sprawling literature on post injections documents the effects of post injections on
engine-out soot with variations in many engine operational parameters. Explanations of how post injections lead to engine-
out soot reduction vary and are sometimes inconsistent or contradictory, in part because supporting fundamental
experimental or modeling data are often not available. In this paper, we review the available data describing the efficacy of
post-injections and highlight several candidate in-cylinder mechanisms that may control their efficacy. We first discuss
three in-cylinder mechanisms that have been frequently proposed to explain how post injections reduce engine-out soot.
Thereafter, to provide a foundation for interpretation of past research, we briefly review basic soot formation and oxidation
chemistry, and soot/fluid processes in fuel sprays and engine flows. Next, we provide a comprehensive overview of the
literature on the efficacy of post-injections for soot reduction as a function of engine operational parameters including
injection duration and dwell, exhaust-gas recirculation, load, boost, speed, swirl, and spray targeting. We conclude by
identifying major remaining research questions that need to be addressed to help achieve a design-level understanding of
the mechanisms of soot reduction by post injections.
CITATION: O'Connor, J. and Musculus, M., "Post Injections for Soot Reduction in Diesel Engines: A Review of Current
Understanding," SAE Int. J. Engines 6(1):2013, doi:10.4271/2013-01-0917.
____________________________________
INTRODUCTION injections per cycle. Early pump-line-nozzle fuel-injection

system designs were incapable of multiple injections [9]. For
Exhaust soot is a heavily regulated emission for diesel those designs, pressure dynamics in the relatively long lines,
engines [1, 2, 3], and while effective aftertreatment systems including the pressure-wave transit time (lag) and reflections,
have been developed for its mitigation, in-cylinder soot would have made multiple injections difficult to control even
reduction techniques remain attractive alternatives to reduce if they were made possible by pump design. Later unit
or eliminate the aftertreatment burden. One commonly used injector designs, whether hydraulically or mechanically
and potentially very effective in-cylinder solution for soot actuated, shortened the length of the high-pressure fuel
reduction is the use of post injections. A post injection is a delivery to the nozzle, thereby allowing more precise control
shorter injection that follows the main fuel injection. While of injection rate. With electronic controls, both the timing and
post injections have been used for a variety of reasons, duration of injection could be controlled independently,
including management of exhaust aftertreatment [4, 5, 6] and paving the way for multiple injections. Likewise,
reduction of unburned hydrocarbons at low temperature electronically controlled common-rail systems can also
combustion conditions [7, 8], here, we focus on post deliver multiple injections. Multiple injection actuation for
injections for the engine-out soot reduction. This work is a early electronically controlled unit and common rail injectors
technical review of past research and a synthesis of current was relatively slow, with rise/fall times (time between
understanding on post injections for soot reduction in diesel beginning/end of injection to the steady-state level) of
engines. 300-500 microseconds (e.g., [10, 11]), which limited how
The literature on post injections for soot reduction begins closely the injections could be placed. Modern common-rail
in force in the early 1990's, coinciding with the development injection systems can achieve rise/fall times around 100
and implementation of injection systems capable of multiple
400
O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 401
microseconds, and similar dwell between injections (e.g., [12, though in this case the main fuel was still delivered by direct
13]). injection. The purpose of the CCD was to provide a high-
These modern technologies have allowed for more velocity post-jet of combustion products after the main
complicated, multiple-injection schedules to be implemented injection. Their results showed that over a large range of
for a variety of purposes, including soot reduction with post CCD injection dwells (varied by changes in injection time),
injections. Most of the studies discussed in this review used engine speeds (800-1200 RPM), excess air ratios (between 1
standard common-rail injectors, although other injector and 4), and CCD jet diameters, the CCD injection reduced
technologies can certainly be used with multiple-injection engine-out soot. Soot reduction was greatest at the highest
schedules. Further advances in injector technology, such as CCD jet momentums, from which Konno et al. concluded
direct-acting piezoelectric-driven injectors [12] and injection that the increased mixing and/or turbulence introduced by the
pressure control [14], could potentially lead to better post- jet enhanced soot oxidation. This was also supported by the
injection performance for soot reduction. accelerated combustion observed in the apparent heat release
There is a large literature on post injections for engine-out rate (AHRR) with the use of the CCD jet. Konno et al.
soot reduction for both light-duty and heavy-duty engines indicated that the combustion products of the CCD jet likely
over a wide range of engine operating conditions. played a smaller role in combustion chemistry than in a
Experimental studies have measured the changes in post- conventional post injection, but the enhanced mixing and/or
injection efficacy with changes in injection schedule, turbulence that resulted from the use of a CCD jet certainly
exhaust-gas recirculation (EGR), load, boost, speed, swirl, affected the combustion process and oxidation of soot. Even
and spray targeting, among others. These measurements so, a definitive conclusion about post-injection mixing effects
include both engine-out soot and in-cylinder soot using on oxidation has not been established.
optical/laser diagnostics. Complementary computer modeling Whereas many studies have pointed to mixing as a way to
predictions also provide insight into potential in-cylinder enhance oxidation, the computer model predictions of Yun et
processes affecting post-injection performance. al. [22] predicted that increased mixing can suppress soot
Despite the large number of studies that have investigated formation from the main injection if the post jet interacts with
how post injections can reduce soot emissions, there is little the burning fuel from the main injection. Three-dimensional
consensus within the literature about how, or even how well, Reynolds-averaged Navier-Stokes (RANS) computational
post injections work. To help frame a detailed discussion on fluid dynamics (CFD) simulations using KIVA, examples of
how various engine operational parameters affect engine-out which are shown in Figure 1, predicted that the post injection
soot reduction by post-injections, we first review three redistributes the fuel from the main injection, creating a more
explanations for soot reduction with post injections that have well-mixed fuel/air distribution with smaller and less fuel-
been offered in the literature. rich soot-forming zones. In these tests, the main-injection
duration was shortened when a post injection was added to
Enhanced mixing maintain constant load.
Several studies have presented the explanation that post The predicted increase in mixing from the post injection is
injections reduce engine-out soot by enhancing mixing within evident in Figure 1, which shows contours of equivalence
the cylinder [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, ratio for a single injection (left) compared to main-plus-post
28, 29]. Ultimate soot yield depends on a balance between injection (right). In the single-injection case, a large region of
formation and oxidation (discussed in more detail in the high equivalence ratio (maximum near 3.5) resided in the
following section), and enhanced mixing could conceivably bowl, whereas the maximum equivalence ratio in the main-
affect formation and/or oxidation. Many authors have plus post-injection case was approximately 2.7, and the fuel
hypothesized that enhanced mixing increases oxidation of was more evenly distributed in both the bowl and the squish
soot from the main injection [15, 16, 19, 25, 28]. In these regions. Additionally, this work showed that this fuel
explanations, enhanced mixing by the post injection brings redistribution mechanism could be particularly important at
fresh oxygen to the soot from the main injection, enhancing high rates of exhaust gas recirculation (EGR) where oxygen
oxidation of this soot while simultaneously burning the post- is limited and mixing is crucial to soot reduction. While these
injection fuel. modeling results pointed to a post-injection mixing effect on
It is difficult to isolate mixing from other effects of post soot formation for this engine and operating condition,
injections, but one unique study from Konno et al. [30] experimental data that isolates the post-injection mixing
attempts to do so by using an unconventional post injection. effect on formation is limited and not yet conclusive.
In their experiment, a “combustion chamber for disturbance”
(CCD), a small auxiliary combustion chamber located above
the firedeck and connected to the main combustion chamber
via a small, straight nozzle, provided an additional injection
of combustion products into the cylinder. The CCD is akin to
a prechamber of older indirect-injection diesel engines,
402 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)
injections can be correlated to increased in-cylinder

temperature, these studies also acknowledged the importance
of mixing and/or targeting of the post jet.
Figure 2. Soot temperature in the squish region showing

a rise in temperature during the burn of the post
injection. Copyright © SAE International. Reprinted
with permission [15].
Figure 1. Predicted contours of equivalence ratio for a

single main injection (top) and for a main plus post Injection duration effects
injection (bottom) from KIVA simulations. [22]. Finally, a smaller number of studies have noted that soot
formation is related to the duration of each injection [28, 38,
39, 40, 41, 42]. Researchers have attributed the engine-out
Increased temperature soot reduction to a lack of soot formation in the post injection
Other studies have argued that as the post-injection fuel and/or a non-linear decrease in soot formation with injected
burns, the increased temperature from the additional heat fuel mass in the shortened main injection. This mechanism is
release can enhance the oxidation of soot from the main most often referenced in studies with close-coupled post-
injection, thereby reducing engine-out soot [15, 27, 31, 32, injection schedules, where the dwell between the end of the
33, 34, 35, 36]. This is a difficult explanation to support main injection and the beginning of the post injection is very
experimentally, as accurate measurements of in-cylinder short. An example is work by Desantes and coworkers [39,
temperatures in diesel combustion are challenging, even in 40], in which they discussed a “split-flame” concept, pictured
optically accessible engines. Two-color soot pyrometry [37] in Figure 3. In this concept, it was posited that the fuel from
can provide some information about soot temperature, though the main injection and the post injection burned separately
the uncertainties are considerable. One example is the work without any interaction, and the reduction in exhaust soot
by Bobba et al. [15], which measured soot temperature in the stemmed from splitting the fuel delivery into multiple
squish region where much of the post-injection fuel burned injections.
after a relatively long dwell time between the end of the main Another common concept related to injection duration
injection and the start of the post injection. An example of effects, proposed by Han et al. [38], is the concept of “jet
these results is in Figure 2. replenishment.” Based on computer model simulations, these
Higher soot temperatures were measured in the squish authors concluded that soot formation is dependent on
region when the post injection was present, as evidenced by injection duration, not just total quantity of fuel delivered to
the peaks in temperature during the burn of the post injection the engine per cycle. This was because as the injection
in Figure 2. The engine-out soot decreased with the post duration increases, the model predicted that the head of the
injection at these conditions, suggesting that temperature may jet, a fuel-rich region where much of the soot is produced
play a role. Similar effects of increased temperature in the [43], was replenished by fresh fuel along the centerline of the
squish region were predicted in a computational effort by jet for as long as the injection was sustained. The longer the
Hotta et al. [19]. While engine-out soot reduction by post- injection, the more fuel was delivered to this head region,
creating a larger fuel-rich mixture that supported more soot

formation.
Figure 4. Ratio of engine-out soot with a post-injection

to that without a post injection versus timing of the post
Figure 3. “Split flame” visualization showing separate injection for ten studies in the literature [10, 15, 16, 17,
combustion events from the main combustion and the 19, 20, 25, 32, 40, 44, 45].
post injection. Copyright © SAE International. Reprinted
with permission [40]. Though each of the soot-reduction mechanisms described
above is certainly plausible, insufficient evidence exists to
This mechanism also agrees with the idea that splitting a quantify the relative importance among the mechanisms for
given amount of fuel into shorter injections reduces the different operating conditions. As such, this manuscript aims
overall soot formed as a result of the change in mixture to provide an overview of the state of post-injection research
formation due to jet replenishment. While injection duration and clearly define what we as a community do and do not
can be correlated to in-cylinder and/or exhaust soot, the know about post injections for soot reduction.
importance of this effect relative to other potential mixing The paper is organized as follows. This introductory
and thermal effects has not been established. section concludes with an overview of some important
Despite the large number of studies on post injections, terminology that will be used in the discussion of post
little consensus exists as to how the three mechanisms above, injections for soot reduction. Next, we provide an overview
and perhaps others, act to reduce soot in diesel engines. of soot formation and oxidation processes in diesel engines.
Before even considering the question of the mechanism(s) of This begins with an overview of soot chemistry that is
soot reduction by post injections, the degree to which post- pertinent to diesel combustion situations. We use this
injections reduce or even increase soot varies by nearly an foundation to build an understanding of soot formation in
order of magnitude in each direction in results presented in diesel jets, based on significant research in both
the literature. The inconsistency in engine-out soot results is computational studies and experimental work in spray
evident in Figure 4, which shows a compilation of data from facilities and in optical diesel engines. We close the soot
ten post-injection studies. Here, the ratio of engine-out soot formation/oxidation overview by discussing the role of
with a post injection to that without a post injection is plotted engine flow fields on soot formation and oxidation, and how
against the start of post injection. Note that Figure 4 is not post injections may interact with these flow features to reduce
from a single controlled study, but rather a compilation of engine-out soot. The third section presents a review of the
data; operating conditions such as main injection scheduling/ effects of various engine operating parameters (injection
duration, post-injection duration, load, speed, and EGR are scheduling, EGR, load, boost, speed, swirl, and spray
not held constant across these studies. targeting) on post-injection soot-reduction efficacy as
The somewhat even distribution of data above and below reported in the post-injection literature. Finally, we pose
unity in Figure 4 indicates that post-injection efficacy is not several remaining research questions that remain unanswered.
universal, but rather highly sensitive to the engine and/or Terminology
operating condition. This dependence will be discussed at
length, using the studies shown in Figure 4 among many Before we delve into a discussion of the technical aspects
others, in the third section of this paper. of post injections, it is helpful to define terminology that will
be used throughout this paper. Though not an exhaustive list,
the following definitions explain terms that are found
throughout the literature, yet may not be precisely defined commanded injection schedule and the actual injection
among studies. schedule.
Post injection OVERVIEW OF SOOT PROCESSES
A particular multiple-injection schedule where a quantity IN DIESEL ENGINES
of fuel is allocated into separate portions such that the second
injection duration is much shorter than the first, or main, Soot Formation and Oxidation Chemistry
duration. The specific cutoff for which the second injection is
As briefly reviewed above, the post-injection literature
small enough relative to the main injection to constitute a
proposes three overall mechanisms by which engine-out soot
post injection rather than a generic split injection is not well
might be reduced by the addition of a carefully selected post
defined, but a maximum of approximately 20% of the total
injection: enhanced mixing, increased temperature, and
fuel in the post injection is consistent with most existing self-
injection duration effects. The action of all three mechanisms
described post-injection studies.
ultimately involves soot formation and oxidation rates, and
Split injection depends on in-cylinder fluid-mechanical and chemical kinetic
A general multiple-injection schedule where a quantity of processes. The literature on soot formation and oxidation
fuel is split into separate portions. The individual split chemistry is vast and still evolving, and it is beyond the scope
injections are often described by the percentage of the total of this work to include all of the fundamental details. Instead,
fuel that each contains. Studies often explore effects of split in this section, we provide an engineering-level overview of a
injections through parameter sweeps with the allocation of few important aspects of in-cylinder soot formation and
fuel shifted between two injections, such as from 10% in the oxidation chemistry. In the sections to follow, the soot
first injection and 90% in the second injection, to 90% in the formation and oxidation processes will be applied to more
first and 10% in the second injection. Hence, split-injections practical aspects of in-cylinder processes, with discussion of
are broad category of multiple injections, of which post- many example studies from the post-injection literature from
injections are a subset. which some insight into post-injection effects on soot
formation and oxidation may be gained.
Close-coupled The evolution of the in-cylinder soot yield results from a
An injection scheme where the dwell between the end of battle between the chemical processes of formation and
one injection and the beginning of the next injection is short, oxidation. The relative rates of these two processes, each of
such that combustion phasing for the second injection is still which is comprised of several complicated steps, are highly
favorable for thermodynamic efficiency. Some studies have sensitive to ambient conditions, particularly temperature and
shown considerable engine-out soot reduction for post equivalence ratio. These conditions are non-uniform and
injections that are late in the expansion stroke, but the constantly changing during the highly unsteady event of
combustion phasing degrades the thermodynamic efficiency. diesel combustion. The discussion here provides a brief
Such late-injection schemes would not be characterized as overview of some of the key aspects of basic soot chemistry,
close-coupled. The threshold for characterizing a post- followed by a brief discussion of how soot chemistry can be
injection schedule as close-coupled is not well defined, but applied to diesel combustion. For a more detailed discussion
dwells of at most a few crank angle degrees are typical. of soot formation and oxidation, the effects of engine
An alternate definition for “close-coupled,” found in operating conditions, and the process of soot formation and
certain literature and common in industry, implies that all oxidation in diesel engines, see for instance the review by
post injections intended for in-cylinder soot reductions are Tree and Svensson [46].
considered “close-coupled.” According to this alternative Soot formation can be considered, in general, as a five-
definition, post-injections that are used for exhaust step process that starts with fuel and ends with fully formed
aftertreatment management are not close-coupled. This soot agglomerates [47, 48, 49, 50, 51]. The five general steps
aftertreatment-based definition is very different than the are summarized in Figure 5 [46]. The first step, pyrolysis
efficiency-based definition of close-coupled used here. and/or fuel decomposition, involves the splitting of fuel
molecules into smaller molecules such as acetylene and
SOIC, DOIC resonantly stabilized radicals (e.g., propargyl). These smaller
Commanded injection-schedule start of injection (SOIC) molecules serve as building blocks for soot precursors [49,
and duration of injection (DOIC). These descriptions 52], which include polycyclic aromatic hydrocarbons (PAH).
differentiate the commanded injection schedule, i.e., the Nucleation, the transition of gas-phase precursors to particles,
pulses sent to the drive electronics, from the actual injection creates the starting point for subsequent soot particle growth.
schedule, i.e., when fuel is actually emerging from the Nucleation occurs at flame temperatures above 1300 K [53].
injector nozzles. Injector dynamics, rail dynamics, and other
system issues can lead to significant differences between the
Figure 5. Process of soot formation in flames [46].

Figure 6. Soot oxidation pathways through
fragmentation and subsequent surface oxidation [57],
After the nucleation site has been created in a relatively with high soot-burnout pathway on the bottom and low
high-temperature zone, the soot particle mass increases soot-burnout pathway on the top.
through surface growth, coalescence, and agglomeration, all
of which may happen simultaneously. These later processes
may proceed at lower temperatures than the initial precursor Additionally, the extent to which soot burns, termed the
synthesis and nucleation steps, and indeed, they often happen “soot burnout” [57], determined the sites of the fracturing in
away from high-temperature zones. The progress rates of the these studies. In low-burnout conditions (less than 15% soot
soot formation processes, from fuel decomposition and oxidation), fracturing occurred predominantly at the
precursor synthesis to nucleation, coalescence, and “bridges” connecting primary soot particles, as shown in the
agglomeration, increase with the concentration of the upper oxidation path in Figure 6. For high soot-burnout
reactants involved, such that other factors being equal, more (greater than 15% of soot oxidation), there was further
fuel-rich mixtures generally form soot more quickly. fracturing of primary particles by O2 molecules as well as OH
At any time during this process, soot formation can be oxidation at the surface, shown in the bottom path in Figure
stymied by the presence of oxidizing species, typically O, O2, 6.
and especially OH. Several investigations of soot oxidation Many empirical soot models, especially those applied to
chemistry have been undertaken at atmospheric conditions diesel combustion modeling, do not account for these detailed
with a variety of fuels, from methane to heavier hydrocarbons nuances of soot oxidation chemistry. As described in the
such as ethylene, m-xylene, and n-dodecane [54, 55, 56, 57, review of soot models by Kennedy [59], commonly used soot
58]. These investigations outline two pathways of soot models use non-physical shortcuts for estimating soot
oxidation. In fuel-lean mixtures, O2 and OH pathways of soot oxidation rates. These include calculating soot oxidation rate
oxidation are important. OH molecules are highly reactive at as a function of the concentration of only O2, and modulating
the surface of soot particles [54, 56]. These surface reactions soot oxidation rate to scale with soot formation rate. As
form CO, a molecule that does not participate in any further described above, these empiricisms do not account for the
soot-formation reactions [46]. Simultaneously, O2 molecules, multiple pathways for soot oxidation and the different roles
more plentiful at fuel-lean conditions, are not as reactive at of O2 and OH. Several more detailed mechanisms, however,
the surface, and instead penetrate into the internal soot account for oxidation by both O2 and OH at different types of
structure, causing internal burning and subsequent fracturing carbon sites [60, 61, 62].
of the soot molecules [56, 57, 58]. Simultaneous This more detailed understanding of soot formation and
measurements of soot particle diameter and number density oxidation chemistry may help to improve the understanding
confirm that prior to bulk burnout of soot, the number density of post-injection processes in diesel engines.
increases significantly due to the fracturing of large
molecules. Late in the soot burnout process, soot molecules Chemical-Kinetic Modeling of Diesel Soot
of all sizes are oxidized through OH surface pathways [58]. A Formation/Oxidation
schematic of this process is shown in the bottom portion of
The dependence of soot formation and oxidation
Figure 6, taken from Ref. [57].
processes on both mixedness and temperature can be
At fuel-rich conditions, results vary regarding the impact
illustrated through chemical kinetic simulations of many of
of the soot-fracturing by the O2 pathway [57, 58].
the processes described above. Figure 7 shows the predictions
Experiments performed with ethylene fuel did not observe the of one such simulation [63], with soot yield as a function of
soot fracturing pathway, but instead saw the dominance of the mixture equivalence ratio and temperature. The closed reactor
OH oxidation pathway at all particle sizes through the flame simulations used detailed soot formation and oxidation
[58]. In the same experiment, however, soot produced with kinetics at each combination of equivalence ratio and
heavier hydrocarbon fuels such as m-xylene and n-dodecane, temperature, at a pressure of 60 bar, typical of TDC
did undergo a fracturing process in addition to the OH conditions in diesel engines. The soot yield is the net
burnout [57]. The reasons for these differences are unclear. formation after a simulation time of 2 milliseconds, which is
a relevant timescale for diesel engines. Note that not all
conditions are practically accessible in diesel combustion,
i.e., both high temperature and high equivalence ratio (top- structure and evolution of diesel spray combustion provides a
right portion) generally do not occur simultaneously. framework upon which the soot formation and oxidation
Figure 7 shows that according to detailed soot formation processes described above may be applied to gain insight into
and oxidation modeling, net soot formation is positive within net soot production in diesel jets. The conceptual model of
a peninsula, over a range of temperatures and above an conventional diesel combustion proposed by Dec [43]
equivalence ratio (ϕ) threshold near ϕ=2. Moving across the provides a description of the reacting diesel jet structure and
peninsula from left to right, soot formation in fuel-rich soot formation and oxidation processes. Figure 8 shows an
mixtures does not begin until the temperature is sufficiently illustration from the model during the “quasi-steady” period
high, and the net soot formation initially increases with of diesel-jet combustion, which occurs between autoignition
increasing temperature. Eventually, increasing temperature and the end of fuel injection.
reduces soot formation rates while increasing oxidation rates, During the quasi-steady period, and in the absence of
so that the soot yield peaks and then decreases at higher interactions with in-cylinder surfaces, the characteristic
temperatures [63]. In the equivalence-ratio direction, soot features of the reacting jet remain essentially unchanged,
yield increases monotonically moving upward as the mixtures even as the jet continues to penetrate into the combustion
become more fuel rich, owing to the increasing concentration chamber during the fuel injection event. As shown in Figure
of reactants, as described above. 8, the high-pressure liquid-fuel spray emerges from the
injector and entrains hot ambient gases as it penetrates into
the combustion chamber. At some downstream location, the
entrained gases provide sufficient thermal energy to vaporize
all of the fuel. Downstream of this “liquid length,” fuel exists
only in the vapor phase.
Figure 7. Closed-reactor simulation of soot formation in

Φ-T coordinates. The simulation was constant
temperature, for n-heptane at a pressure of 60 bar and a
reaction time of 2 ms. Soot contours are soot yield, the Figure 8. Conceptual model of diesel combustion during
percentage of fuel carbon formed into soot. (The final, the quasi-steady period. Copyright © SAE International.
definitive version of Figure 7 has been published in the Reprinted with permission [43].
International Journal of Engine Research, 3(4), August,
2002 by SAGE Publications Ltd., all rights reserved. ©
Diesel spray flames are typically lifted, such that
[63].)
significant exothermic reactions occur only downstream of
some flame lift-off length (see annotation in Figure 8).
Downstream of the lift-off length, the diesel jet is surrounded
Soot Formation and Oxidation in Diesel by a diffusion flame. A fuel-rich, partially-premixed reaction-
Jets zone also resides across the interior of the jet near the lift-off
While the ϕ-T plot of Figure 7 lends insight into soot length [64], and the downstream jet interior contains the hot
formation/oxidation tradeoffs under static conditions, products of the premixed reaction zone, as well as entrained
conventional diesel combustion occurs in a dynamic in- products from the diffusion flame. It is within these hot, fuel-
cylinder environment with considerable spatial and temporal rich mixtures in the jet interior that soot formation occurs for
variation of temperatures and mixing states. Hence, the diesel jets, as depicted in Figure 8. One of the key species for
balance between soot formation and oxidation processes also soot oxidation, OH, exists primarily in the hot diffusion flame
varies considerably throughout the chamber and during the surrounding the jet. Hence, in the conceptual model, once
combustion event. Nevertheless, current understanding of the
soot is formed, it is oxidized primarily at the boundary of the Mixing prior to reaction, such as that affected by the
diffusion flame and the interior soot, as indicated in Figure 8. flame lift-off, is clearly important for net soot-formation.
The net soot formation within the jet interior is strongly However, the quasi-steady jet features illustrated in Figure 8,
affected by mixing upstream of the lift-off length, and hence including the lift-off length, are only relevant until the end of
by the magnitude of the lift-off length. A longer lift-off the main injection. A conceptual model for conventional
allows more unreacted ambient gases to be entrained prior to diesel combustion after the end of injection has not yet been
combustion, thereby reducing the equivalence ratio of the proposed, but limited in-cylinder optical imaging does
mixtures as they enter the reaction zone. As described in the provide some guidance. Combined imaging of laser-induced
previous section, the soot formation processes depend on the incandescence of soot and OH fluorescence shows that after
reactant concentrations, so that a reduction of the fuel the end of injection, soot distributions gradually become
concentration through increased mixing upstream of the lift- more disperse and broken into separate pockets [66]. The soot
off length reduces the overall soot formation rate. Figure 9 pockets generally remain surrounded by an oxidizing
provides experimental data that illustrates the dependency of envelope of OH, at least in the early part of the cycle when
soot within the diesel jet on the degree of mixing upstream of close-coupled post-injections would be added [66].
the flame lift-off. The data points show the peak optical If soot after the main injection does reside in fuel-rich
density (KL) from laser-extinction measurements of soot in pockets surrounded by a diffusion flame with an oxidizing
diesel jets over a range of operating conditions, as indicated layer of OH, then the post injection cannot affect mixing
in the figure inset, plotted versus the mean equivalence ratio prior to reaction in the way that flame lift-off does.
at the lift-off length. To help collapse the data, the optical Nevertheless, one way that a post injection could affect
density measurements are scaled according to the ambient oxidation is by increasing local mixing or the surface area of
density. Figure 9 shows that as the lift-off length increases the diffusion flame that surrounds the remaining main-
(moving from left to right), the soot within the jet decreases, injection soot, breaking up larger soot pockets and providing
eventually reaching zero when the lift-off is long enough fresh oxidizer to these fuel-rich regions. In-cylinder optical
such that the mean equivalence ratio is no larger than 2. imaging shows that residual soot clouds naturally break into
smaller pockets as the soot oxidizes late in the cycle [66], and
if post-injections accelerate this process, they should aid soot
oxidation. Given that diesel combustion after the premixed
burn is generally mixing limited, an increase in mixing by the
post injection could help increase the mixing rate through an
increase in flame area, thereby increasing the rate at which
soot is oxidized. Indeed, some studies have shown an
acceleration of combustion with the addition of a post
injection [30, 40], which would be consistent with increased
mixing, and thus increased oxidation of the remaining soot
pockets from the main injection.
Another question is the difference in the surface oxidation
and interior fracturing mechanisms of soot oxidation by OH
and O2, respectively, as described above. At the boundary of
the residual soot pockets from the main injection, OH and O2
should be plentiful, similar to the conceptual model
description of the diffusion flame in Figure 8, such that both
Figure 9. Peak soot optical density (KL) in diesel jets
pathways could be active. It is unlikely that the highly
under various conditions, as measured by laser-
reactive OH could penetrate into the soot-filled, fuel-rich
extinction and scaled by ambient density [65].
interior, and indeed, OH is not detected there. The lower-
reactivity O2, however, could conceivably penetrate
The ϕ>2 threshold for soot formation in diesel jet somewhat into the soot pockets if it can survive passage
experiments is consistent with the simulations of Figure 7, through the diffusion flame. If so, then the fracturing pathway
which indicate a lower boundary for the soot yield peninsula by O2 could be active within the soot pockets. O2 is a
near ϕ>2. Starting from a hypothetical position within the
difficult species to detect within fuel-rich diesel jets, and no
soot peninsula in Figure 7 (indicated by the round marker), an
such measurements yet exist. Nevertheless, it is conceivable
increase in mixing prior to reaction resulting from an increase
that post injections could enhance soot oxidation by mixing
in the liftoff length corresponds to a downward movement in
O2 deeper into the soot pockets, perhaps by disrupting the
the ϕ-T plot. In a similar way, an increase in temperature
diffusion flame so that the O2 can survive into the soot
from the hypothetical starting point, such as that illustrated in
Figure 2, also results in a decrease of soot yield, in this case pockets. Additionally, O2 may be transported into these
due at least in part to a predicted increased oxidation [63]. pockets of fuel by diffusion through highly-strained flames.
Effect of Engine Flow Fields on Soot does not create the turbulence-generating interface of a more
optimal swirl/spray combination.
Formation and Oxidation It is also worth noting that in large-bore diesel engines
In addition to mixing effects by direct interactions of post- with high swirl, non-solid-body rotation has been observed
injection jets with main-injection soot pockets, post-injection with higher swirl rates in the center of the chamber [68]. As
interactions with in-cylinder bulk flow fields and surfaces can there is higher angular velocity in the center of the chamber,
also affect mixing, and hence soot oxidation. Mixing and the mean flow has globally high shear, which increases
combustion in diesel engines are dependent on both turbulence production. Also, spray-swirl interactions can
turbulence at smaller scales and bulk flow motion at larger affect the local shear rates by targeting the fuel spray into
scales. Hence, post-injection effects on either turbulence or either the bowl or the squish region, where momentum
bulk-flow motion can be important for soot oxidation during exchange between the ambient crossflow and the fuel spray
the mixing-limited period of combustion. decelerates the local swirl flow.
As reviewed by Miles [67], interactions of fuel-spray In addition to spray/swirl/wall interactions, swirl/squish
flows with in-cylinder bulk flows and in-cylinder surfaces interactions can also be important. Similar to the effect that
can have profound effects on mixing and turbulence. One the swirl flow has on the spray flow that is redirected inward
obvious source of spray-induced turbulence is the strong by the bowl geometry, squish flows can also be deflected by
velocity gradients in the vicinity of the fuel spray, which may centrifugal forces associated with the swirl. For low levels of
interact with or be transported to regions where soot is being flow swirl, the squish flow can penetrate to nearly the
oxidized. Perhaps more importantly, however, the spray can cylinder centerline before it turns down into the bowl. At
interact with in-cylinder flows, especially swirl, to promote moderate swirl, the inward penetration of the squish flow is
large-scale mixing and turbulence production. Spray-swirl reduced and the flow turns down into the bowl farther from
interactions can promote mixing and turbulence production the cylinder centerline, where the radial momentum imparted
by creating bulk flow motions that bring fuel-rich regions by the squish process is balanced by the centrifugal forces of
together with fresh charge, while also increasing turbulence. the swirl flow. For high swirl, the squish flow turns down
For example, as described by Miles [67], using single into the bowl almost immediately after passing the bowl lip.
injections in light-duty engines with appreciable swirl and a These three general cases create vastly different in-cylinder
re-entrant bowl geometry, as the spray is deflected downward flow patterns with different rotational directions [67], which
by the bowl-rim, it displaces high angular-momentum fluid can affect the “initial conditions” into which the fuel spray
from the outer bowl down into the bottom of the bowl and penetrates. Reverse-squish flows during the early expansion
then toward the center of the chamber. Centrifugal forces stroke could also be important, especially in regard to
from the swirl impede the inward flow, forcing it upward and turbulence production near the bowl lip and subsequent
back outward, thereby creating a large rotating flow structure effects on later flow development.
in the bowl. A complementary counter-rotating structure While the above discussion was based on observations
higher and closer to the center of the chamber forms to with single-injections, the general concept of creating
accompany the large bowl structure. Together, these two favorable bulk flows and turbulence generation through the
structures bring fuel-rich mixtures out of the bowl and into interaction of a fuel jet with in-cylinder surfaces and squish/
contact with fresh oxidizer at the interface between the two swirl or other flows can apply to post-injections as well. For
vortices. post injections, the flow-structure interactions might be
In addition to the bulk mixing provided by these large- different than the examples provided above, but some fluid-
scale structures, large deformation rates of the fluid between mechanical mechanism almost certainly contributes to
the structures generate significant turbulence in the same increased soot oxidation with post injections. If so, proper
interfacial region between the rich bowl mixtures and the matching of the post-injection event with in-cylinder
oxidizers in the upper vortex. Hence, the spray/swirl geometry and flows may be critical for optimizing post-
interaction creates bulk flow motion that brings fuel and air injection performance. Indeed, the disparity in post-injection
together while also generating turbulence at the same fuel/air efficacy among studies in Figure 4 and as reviewed in the
interface. next section may be explained to some degree by differences
The above description of spray/swirl/wall interaction is in spray/flow/wall interactions. Furthermore, a large portion
not guaranteed, however, and depends strongly on the of soot oxidation generally occurs after the peak pressure,
balance between spray momentum, chamber geometry, and when isotropic stresses combined with volume expansion
swirl ratio. For instance, higher swirl increases the centrifugal decrease the turbulence intensity. One beneficial effect of
forces that resist the inward flow induced by the spray, which post injections might be to increase turbulence production
leads to a smaller vortical structure in the bowl that may not during the period of otherwise decreasing turbulence, when
bring rich mixtures out of the bowl. Conversely, greater spray soot is also oxidizing.
momentum (e.g., higher injection pressure or longer duration) The concepts outlined in this section provide background
can push rich bowl-mixtures farther toward the chamber information to help interpret experimental observations of
center, thereby creating a single large rotating structure that post injections for engine-out soot reduction. In the next
section, we transition from fundamentals to applications, and variety of post-injection schedules showed that a post
discuss the effect of operational parameters on post-injection injection with the shortest dwell possible, 5 °CA, lead to the
efficacy. maximum decrease in soot at the same load as a single
injection, as shown in Figure 10. Other later post-injection
DEPENDENCIES ON ENGINE timings, with post-injection dwells from 7 to 13 °CA after the
OPERATING PARAMETERS end of the main injection, had either no effect on engine-out
soot or increased engine-out soot.
In this section we discuss how the efficacy of post
injections for soot reduction changes as engine operational
parameters are varied. These parameters include injection
scheduling, exhaust gas recirculation, load, boost, speed,
swirl, and spray targeting, all of which were commonly
varied throughout the literature. While each effect will be
discussed separately here, it is important to note that many
studies changed several of these parameters simultaneously,
and as such, the effect of each of these engine operational
parameters on post-injection efficacy may not be separable.
For example, several studies [33, 34, 69, 70] measured
engine-out soot with post injections at various points on
standard drive cycles, such as the New European Drive cycle
(NEDC) or the European Stationary Cycle (ESC) [71, 72]. In
this type of study, operational parameters like load, speed,
and boost are typically varied simultaneously to realize
differences between standard operating conditions. In the
descriptions below, we have identified these cases and tried
to separate the effects as much as possible.
Injection Scheduling
The effect of injection scheduling on post-injection
efficacy has been studied in several different ways. Broadly
speaking, variations in post-injection scheduling can be
broken down into two separate aspects. First, the dwell
between the end of the main injection and the beginning of
the post injection can be varied, thereby changing the
combustion phasing of the post injection relative to the main
Figure 10. Effect of post injection on FSN (and other
injection, as well as the targeting of the post jet into the
emissions and performance parameters) as a function of
combustion chamber and mixture field created by the main
“after interval,” or post-injection dwell, between the end
injection. Second, the duration of the post injection can be
of the main injection and the start of the post injection.
varied, which changes both the fuel quantity and penetration
Copyright © SAE International. Reprinted with
of the post injection and, for a constant main-injection
permission [19].
duration, also increases the load. Both of these aspects of
post-injection scheduling have been extensively studied
within the literature for a variety of engine configurations and These authors [19] attributed the reduction in soot from
operating conditions. the close-coupled post injection to both enhanced mixing and
Most studies showed a reduction in engine-out soot with increased temperature. CFD simulations from the same study
the addition of a post injection at some dwell and duration, predicted that the close-coupled post jet entered the squish
but there was no dwell and duration combination that clearly region, which contained soot from the main injection and
and universally lead to soot reduction [10, 15, 16, 17, 18, 19, fresh oxygen. As a result, the predicted temperature in the
20, 27, 28, 31, 38, 39, 40, 41, 73, 74]. Researchers reported squish region rose due to combustion from the post-injection,
that often a “sweet spot” could be reached where engine-out shown in Figure 11. Based on the model predictions and
soot was minimized [15, 19, 21, 22, 24, 28, 33, 34, 73]. On experimental trends, the authors argued that the engine-out
the whole, studies have shown that shorter-dwell injection soot reduction was due to a combination of the enhanced
schedules reduce engine-out soot more than longer-dwell mixing between the main-combustion soot and the oxygen in
schedules. For example, Hotta et al. [19] examined the effect the squish, and the increased temperature and chemical
of post-injection dwell on engine-out soot in a single- reaction resulting from the post-injection burn in the squish
cylinder, light-duty optical research engine. Experiments at a region. Both the availability of oxygen and higher ambient
temperature in the squish would enhance soot oxidation in operating conditions (from the ESC, the European Steady
this region. State Test Cycle), post-injection timings (SOI2C), and post-
injection durations (DOI2C). For almost all engine operating
conditions and a wide range of post-injection schedules,
smaller post injections were more effective at reducing soot.
Engine-out soot could be minimized at a post-injection
quantity of approximately 10-20% of the total fuel.
While post-injection scheduling can be adjusted to reduce
engine-out soot, effects on engine efficiency cannot be
ignored. Engine efficiency is one of the key concerns for
regulators, engine manufacturers, and consumers. Fuel
delivery and timing plays a key role in determining engine
efficiency through the influence of combustion phasing on
thermodynamic efficiency. For instance, later phasing of a
portion of combustion, such as might occur with post
injections, can decrease efficiency, or by another measure,
increase brake-specific fuel consumption (BSFC).
Several authors have addressed the issue of engine
Figure 11. Temperature field comparison between main- efficiency in cases where post injections are used for soot
injection only (top) and main- plus post-injection reduction [14, 17, 27, 33, 73, 77]. Many authors measured an
operation (bottom), showing increased temperature in increase in BSFC with the addition of a post injection. The
the squish region with post-injections. Copyright © SAE extent of this increase could be mitigated, however, by
International. Reprinted with permission [19]. varying both the post-injection duration and the dwell. These
trends were shown by several authors[33, 73, 77], but were
illustrated simultaneously by Benajas et al.[33]. Figure 12
The timing of the post jet with respect to the piston shows an example of these results at the B50 ESC engine test
position and subsequent squish height seems to be an cycle condition.
important parameter for the efficacy of this mechanism,
although the later timings that resulted in higher engine-out
soot were not reported in detail. Though enhanced soot
oxidation in the squish was reported in a few papers in the
literature, it is by no means universal [15, 19, 22]. Many of
the papers do not offer detailed explanations like that
provided in Hotta et al. [19].
Certain studies, however, showed that longer post-
injection dwells lead to greater engine-out-soot reduction.
Results from Chen [20], Shayler et al. [73], and Bobba et al.
[15] all showed greater reduction in engine-out soot with
post-injection dwells greater than 20 °CA, which may be
related to differences in interaction of the post injection with
soot and combustion products in the squish region. As
described earlier in the overview of soot processes,
differences among these studies in spray interaction with in-
cylinder surfaces and flows may contribute to differences in
the observed post-injection efficacy.
Changes in post-injection duration also affect the Figure 12. BSFC variation from a single injection with
performance of post injections. Many of the studies that post-injection mass (duration) and dwell at the B50
investigated the effects of post-injection duration included operating condition. Copyright © SAE International.
data that fall into the “split injection” category [10, 18, 31, Reprinted with permission [33].
38, 75, 76], Wherever possible, only the cases where the
split-injection schedule would qualify as post injections (20%
or less of the total fuel in the post-injection) are discussed In this study, as in others, increases in BSFC stemmed
here. from increasing dwell and post-injection duration. For
Many studies showed that post-injection duration could be example, even at the smallest post-injection fuel mass, 12
optimized to reduce soot [15, 21, 34]. For example, Payri et mg/cycle, the BSFC at a dwell of 10 °CA was 1.8 g/kWh
al. [34] measured engine-out soot for various engine greater than that at a dwell of 4 °CA. The difference
increased at longer post-injection durations. This trend was differ, however, in their assessment of whether EGR renders
observed at several engine test cycle conditions, including post injections more or less effective at reducing soot.
high- and low-load conditions, as well as high- and low-speed For example, results obtained by O'Connor and Musculus
conditions. [79] showed that post injections became more effective at
The literature also discusses efficiency with respect to the higher rates of EGR. In this study, a constant main-injection
improved mixing and fuel distribution possible with post- duration was chosen and the post-injection duration was
injection schedules. For example, the fuel vapor distribution varied to assess the effect of post-injection duration and post-
measurements by Mendez and Thirouard [17] not only jet penetration. The engine-out soot levels using the post-
illustrated the efficacy of post injections for soot reduction by injection schedules were compared to measurements of
elimination of rich pockets, but they also noted that the post- single-injection operation at a similar range of loads. This
injection schedule resulted in an 8% reduction in BSFC, was done for four intake-oxygen levels, 21%, 18%, 15%, and
presumably due to a better fuel distribution for higher 12.6% by volume, as is shown in Figure 13. At conditions
combustion efficiency. This study is discussed further in the with 21% intake oxygen, post injections were able to reduce
Spray Targeting portion of this section. engine-out soot as much as 40% versus a single injection at
These results and others like them have an important the same load. At 18% intake oxygen, a level typical of
impact on the manner in which fuel injection schedules can current production heavy-duty diesel engines, this reduction
be designed for soot reduction. For example, it is fortuitous increased to 52% at low load. Finally, at 12.6% intake
that several researchers found that short, close-coupled post oxygen, the soot reduction at the same low load as a single
injections are best for reducing engine-out soot, since they are injection increased to 62%. It should be noted that these
also effective at maintaining or even improving BSFC [19, results are presented in units of filter smoke number (FSN),
33, 34]. As discussed above, however, several studies which has a monotonic but somewhat nonlinear relationship
reported that longer dwell times resulted in greater soot with soot mass.
reduction [15, 20, 73]. In the cases where there is a
competition between BSFC and soot reduction, the optimal
post-injection dwells will depend on a balance of both
considerations. This type of optimization was discussed in
work by Montgomery and Reitz [21]; their work shows that
this optimization process is costly but certainly possible.
Exhaust Gas Recirculation

Exhaust gas recirculation (EGR) is a commonly varied
intake condition in many post-injection studies. EGR is a
process by which gases from the exhaust stream are mixed
with the intake stream, effectively diluting the intake mixture
- reducing the oxygen content by volume - and decreasing the
peak combustion temperatures. Through this temperature
effect, EGR can significantly suppress the production of NOx,
a toxic and highly regulated emission in diesel engines.
Despite the positive effects that EGR can have on NOx
emissions, EGR rates that are practical for normal engine
operation can also lead to severe increases in engine-out soot
due to the reduction in intake oxygen [46, 64]. Reduction in
oxygen content initially increases the soot yield by increasing
the overall equivalence ratio and allowing for richer mixtures
of fuel and air, both promoting soot formation and limiting
soot oxidation. Figure 13. Engine-out soot emissions for single injection
It can be advantageous to use post injections in (filled squares) and main plus post injection (open
conjunction with EGR to simultaneously reduce NOx and circles) schedules with SOI1C=347 CAD, DOI1C=1950
soot with in-cylinder techniques. Previous studies have microseconds, SOI2C=366 CAD, and varying DOI2C
shown that post-injection strategies can be implemented from 300 to 600 microseconds at four intake-oxygen
without a penalty in NOx [27], and as such, many researchers levels [79]. Note the differences in FSN scale on the
have studied the efficacy of post injections for soot reduction ordinate.
at various levels of EGR [19, 20, 21, 25, 26, 73, 78, 79], all
the way down to 12.6% intake O2 [15, 79]. These studies
All these results were at similar loads and injection Measurements by O'Connor and Musculus [79] have
schedules, although the mechanism by which post-injection shown that post injections are less effective at reducing soot
efficacy increases with decreasing oxygen content is not yet as duration of the main injection (load) increases. For 18%
clear. These findings were similar to those in studies by intake oxygen, the optimum post-injection schedule at a
Shayler et al. [73] and Pierpont et al. [25], but were not gIMEP of 500 kPa reduced engine-out soot by 52%, whereas
universal. For example, post injections in a study by Chen the optimal post-injection schedule at a gIMEP of 650 kPa
[20] were much less effective at high levels of EGR, resulting only reduced engine-out soot by 19% as compared to a single
in similar engine-out soot levels for single and main- plus injection at the same load.
post-injection schedules at 40% EGR. The reasons behind Similar results were measured by Yun et al. [23], who
these differences are not clear at this time. also changed load by increasing injection duration. These
results are in Figure 14. At constant engine-out NOx, post
Load injections at 3 bar gIMEP reduced engine-out soot by 33%,
Load is another common variable that has been while post injections at 4.5 bar gIMEP reduced soot by only
investigated in studies of post-injection efficacy [21, 22, 23, 16%, although the absolute reduction in soot in both cases
24, 32, 33, 34]; this is an important quantity to vary as the was relatively similar.
absolute level of engine-out soot (not normalized by load)
generally increases with load as more fuel is delivered during
each cycle. In these studies, researchers have often tested a
variety of loads according to a standard testing cycle. As
discussed above, changing load in a test-cycle format often
means that other engine operating parameters, speed in
particular, are simultaneously changed. In this discussion of
the effect of load on post-injection efficacy, we have tried to
separate these effects as much as possible.
In general, load has been changed in two different ways:
changes in injection duration [24, 77] and changes in
injection pressure [21, 32, 34]. These two methods of
changing load have very different effects on in-cylinder
processes important to soot production. Changing the
injection duration changes not only the amount of fuel
injected but also the distribution of fuel within the cylinder,
as was mentioned in the discussion of jet replenishment [38]
in the introduction. According to the jet-replenishment
concept, longer injections can form large regions of fuel-rich
mixture at the head of the jet that promote soot formation
through the constant delivery of new fuel to the rich region. Figure 14. Engine-out soot-NOx trade-off at four loads,
Alternatively, changing the injection pressure without comparing single-injection and main- plus post-injection
changing the injection duration alters not only the distribution operation [23].
of fuel within the cylinder, but also the jet entrainment during
injection, resulting in changes in the reacting jet structure
[80], fuel/air mixture distribution, and soot formation [65]. The reduction in post-injection efficacy with increasing
As was described at length above, soot formation in diesel load has not been clearly explained in the literature. Based on
jets is highly dependent on initial mixing characteristics in the experimental results, the reduced efficacy could be due to
the liftoff length region of the jet. Higher injection pressure the nonlinear dependence of soot formation on injection
results in a longer lift-off length and more mixing within the duration, described above as jet replenishment by Han et al.
lift-off region; this can result in suppressed soot formation in [38]. Load increases with more fuel delivery per injection,
the mixing-limited region of the jet. Additionally, higher resulting from a longer injection duration. This nonlinear
injection pressure of the post injection itself can change the growth in soot quantity with fuel mass per injection could
mixing characteristics of the post jet with the main-injection mean post injection are less effective at reducing soot on a
mixture, altering the efficacy of the post injection. percentage basis for longer main injections versus shorter
Despite these differences, measurements have shown that main injections. For example, while the absolute reduction in
load has an effect on the performance of post injections for soot was similar at 3.0 bar and 4.5 bar in Figure 14, the post
soot reduction. In general, post injections are less effective at injection was less “effective” on a percentage basis at the
reducing soot at higher loads over a range of injection higher load. While this explanation is certainly plausible
schedules and intake conditions. given the experimental results and discussions published in
the existing literature, it has yet to be shown definitively by Tanin et al. [77] reported that at a variety of engine
experiments or through modeling techniques. operating conditions, engine-out soot could be minimized at a
certain boost pressure, as is shown in Figure 15. The engine-
Boost out soot for the 50/50 split-injection schedule was
Increased intake pressure, or boost, in diesel engines both consistently lower than for the single-injection schedule, and
enables higher loads and helps to optimize emissions and both injection schedules displayed similar optimal boost
efficiency over a range of engine operating conditions. levels for minimum soot emissions for this engine and
Increasing the pressure and density of the intake charge injection system.
results in higher peak pressures and temperatures, and can No known post-injection studies have varied boost
lead to favorable decreases in brake-specific fuel without varying other engine operating parameters as well;
consumption (BSFC) [81]. Boost can also have significant most of the studies have followed a standard engine test
effects on the combustion process. For example, higher boost procedure, varying speed, load, and boost simultaneously [33,
increases the mass of oxygen in the charge, which can change 34, 69, 70]. Despite that, an important study that discussed
combustion and soot formation/oxidation characteristics. the effect of boost on post-injection performance was
Additionally, changes in ambient density alter spray and performed by Montgomery and Reitz [21]. This study
flame characteristics, including liquid length, jet penetration optimized fuel consumption and emissions relative to three
rate, entrainment, and liftoff length [82]. Soot formation and baseline conditions using a response-surface method with six
oxidation chemistry are sensitive to many of these spray and variables: injection pressure, boost pressure, combustion
flame changes, as well as to ambient pressure [83]. phasing, dwell between injections, fuel percentage in each
Unfortunately, no post-injection studies available in the injection of the split injection, and EGR. In this study, the
literature address the issues of boost independently of other authors compared the effects of EGR versus those of boost
variables. The 50/50 split-injection study of Tanin et al. [77] pressure, as both variables are means by which to change the
provides a thorough comparison of boost effects on engine- quantity of intake oxygen. The optimization process showed
out particulate matter (PM) for split injections (but not for that at certain conditions, post injections can reduce engine-
post injections) relative to single-injection operation. In the out soot with less boost pressure because the post injection
absence of post-injection studies focusing on boost, this split- enhances the mixing of soot with remaining oxygen.
injection study is the best available alternative.
Speed
Variations in engine speed can result in considerable
changes in engine flows and combustion. Engine speed
affects several key features of the flows within the
combustion chamber, including swirl, squish and reverse
squish flows, and tumble. In particular, higher speeds result
in higher in-cylinder bulk-flow velocities, which are
particularly important in high-swirl, light-duty diesel engines
[17, 32, 41, 84]. Experimental data demonstrating speed
effects on bulk flows is limited; however, its effect on the
fluid velocities in bulk flow structures is minimal when
normalized by piston speed [67]. Nevertheless, as discussed
in the soot process overview section, higher centrifugal forces
associated with increasing swirl at high engine speed, for
instance, can affect the development of complementary
structures that aid mixing and soot oxidation. Additionally,
speed changes the heat transfer throughout the cycle [85].
Faster speeds mean less time for heat rejection during the
compression stroke, resulting in higher temperatures and
pressures before combustion. Faster speeds also mean that
there is less time per cycle for chemical processes to reach
completion. While the main combustion events occur on time
Figure 15. Particulate emissions for 50/50 split operation
scales short enough to be completed at a variety of speeds,
as a function of boost pressure, showing non-monotonic
time available for late-cycle soot burnout is at higher speeds,
dependency of engine-out soot with boost for constant
which may increase the engine-out soot.
NOx of 3.3 g/bhp-hr. Percentages in the plot indicate the
Results differ regarding the effect of speed on post-
boost level relative to the baseline boost. Copyright © injection efficacy. For example, experiments performed by
SAE International. Reprinted with permission [77]. Badami et al. [32] in a light-duty engine showed that post
injections over a range of dwell times could result in greater
reductions in soot at higher speeds. This was particularly true effective at the later timing, 12 CAD. The efficacy of the post
for close-coupled post injections, where the command dwell injections are even lower, 45% reduction, for the highest
time was on the order of 500-1000 microseconds. speed case, C100 at 1800 RPM (not shown here). However,
In contrast, extensive tests by Payri et al. [34] indicate the operating condition is too different from A50 and B50 for
that post injections are less effective at higher speeds in a direct comparison.
heavy-duty engine. They tested several conditions from the Variation in post-injection efficacy with dwell times must
ESC, two of which were A50 and B50, which are both 50% be given special consideration when varying speed. Studies
load conditions but with different speeds. Several parameters, have reported dwell times both in terms of °CA (as in Payri et
such as injection pressure, fuel mass per cycle, injection al. [34]) and actual times (as in Badami et al. [32]), but the
scheduling and EGR, stayed relatively constant between these actual time of a crank angle degree is a function of the speed
two test conditions, but the speed changed by 300 RPM. of the engine. How a post-injection dwell time is defined in
Figure 16 shows the engine-out soot (FNS) measurements for an engine controller as speed is changed may result in very
the single-injection case (labeled nominal) and three post- different post-injection behavior at various speeds. If the
injection cases with dwells of 4 °CA, 8 °CA, and 12 °CA. post-injection duration and dwell are defined in terms of
crank angle degrees, the targeting of the post injection into
the main-injection mixture will be similar across engine
speeds. If these definitions are in terms of absolute times,
however, the targeting of the post jet could be very different,
changing its burning and mixing characteristics, and as a
result, its effectiveness at reducing soot.
Swirl
Swirl is an important flow feature with respect to the
targeting of the post jet into the products of the main
injection. Reviewed by Miles [67, 86] for light-duty engines,
swirl effects on combustion include increased mixing,
enhanced evaporation rates near engine surfaces, and spray/
swirl/squish interactions. Each of these flow effects can
change the soot formation and oxidation characteristics of the
fuel jet, and alter the performance of the post injection for
soot reduction.
Various authors have identified swirl as an important
parameter for post-injection efficacy as it relates to dwell
between the end of the main injection and the beginning of
the post injection [22, 28, 41]. The strength of the swirl and
the duration of this dwell determine the targeting of the post
jet into the residual main-injection mixture, changing how the
post injection interacts with the main-injection soot. This
relationship was measured by Barro et al. [28] in their recent
work on post-jet targeting. Figure 17 shows the measured
effect of swirl and dwell on relative net oxidation-rate of
soot. Here, the “net oxidation-rate” was defined as the rate of
change in the in-cylinder soot as determined from optical
two-color pyrometry data.
Figure 16. Engine-out soot in a heavy-duty engine for In this case, the sensitivity of the soot net oxidation rate to
1200 RPM (top) and 1500 RPM (bottom) and three post- swirl strength reversed between short and long dwell periods.
injection schedules. Copyright © SAE International. At short dwell times, the oxidation rate was greater for cases
Reprinted with permission [34]. with higher swirl, presumably indicating that the combination
of higher turbulence levels from swirl and close-coupled fuel
delivery enhanced the soot oxidation process. The benefit of
While overall the FSN levels were lower at the higher high swirl dropped off significantly, however, as the post-
engine speeds, the post injections were more effective at low injection dwell increased.
speeds. At the lower speed (A50), the post injection resulted This change in post-injection efficacy with dwell time
in an engine-out FSN reduction of 62%, while at the higher may also indicate that post-jet targeting was important. The
speed (B50), the post injection only reduced the engine-out authors indicated that post injections are most effective if the
FSN by 47%. In this case, the post injections are particularly post injection occurs before the optically-measured in-
cylinder soot peak from the main injection occurs. In this injection schedules. The authors expected that the injector
case, the optimal targeting of the post jet depends on the with a wider included angle, 140°, would have less
spacio-temporally evolving process of soot formation in the interaction with the piston for the specified injection schedule
main-injection mixture. Iin this study [28], the post-injection than that with the smaller angle, 125°. Figure 18 shows the
efficacy was much less sensitive to timing in the low-swirl results of their comparison, and clearly indicates that the
case. This may indicate that post-injection efficacy is a smaller included angle performs better for both NOx and soot.
coupled function of both turbulence and timing, and that the These results indicate that targeting of the sprays of the main
post injection is less effective at low swirl ratios. Their and/or the post injection can be an important mechanism soot
engine-out soot measurements indicated that a post injection reduction with post injections.
can reduce engine-out soot by up to 45% at high swirl ratios,
but only 30% for low swirl ratios.
Figure 17. Measured relative net soot oxidation rates for Figure 18. Soot-NOx trade-off for two injector included
two swirl levels and three equivalence ratios over a range angles showing generally lower particulate emissions for
of post-injection dwells, showing the relative roles of the smaller included-angle injector. Copyright © SAE
swirl and dwell on post-injection efficacy [28]. International. Reprinted with permission [25].
This example also speaks to the issue of dwell times Some of the light-duty literature includes discussion of
measured in units of time versus units of °CA, as was the role of engine geometry in post-injection efficacy, which
discussed in the Speed portion of this section. It was is to be expected given the highly contoured geometry of
important in the Barro et al. study that the dwell was most light-duty engines. In some of these cases, interaction of
measured in units of time, as the swirl ratio is normalized by the spray and the engine surfaces has been used to enhance
the motion of the piston, which determines the length of a the efficacy of the post injections for soot reduction. As
°CA. The results of this study further emphasize this described earlier, the work by Hotta et al. [19] nicely
difference. illustrated the role of post-jet targeting in the bowl versus
squish, and mirrored the results found in Bobba et al. [15].
Spray Targeting Simulations of post-injection and multiple-injection
Several studies have reported that the combination of schedules by Liu and Reitz [84] also showed the importance
spray targeting and engine geometry can play an important of post-jet targeting between the bowl and squish region for
role in the efficacy of post injections for soot reduction. In optimal soot reduction.
both light-duty and heavy-duty engines, which can have quite Researchers have found that the behavior of the post
different geometries, the targeting of the post jet with respect injection changes as the geometry of the engine changes
to the different regions in the combustion chamber seems to throughout the cycle. For example, spray targeting relative to
be important. In this section, we highlight three studies that the squish region has been shown to change soot oxidation
have varied the targeting of the post injection, either through characteristics. The study by Bobba et al. [15] emphasized
changes in injector spray angle or through changes in the importance of the spray targeting at the bowl versus the
injection scheduling. squish region. If the main injection was timed such that it
A study by Pierpont et al. [25] investigated the role of targeted the bowl, soot formed predominantly low in the
spray angle in post-injection efficacy over a variety of split- bowl. The late post jet, targeted into the squish region, did
Figure 19. Schematic showing the difference between main injection targeting in the bowl (top) versus the squish (bottom) and
how this targeting affects post-injection efficacy. Copyright © SAE International. Reprinted with permission [15].
little to reduce the soot from the main injection in this case. injections is a non-reacting study by Zhang and Nishida [18].
However, when the main injection was timed later and In this study, the authors varied the fuel percentage in each
allowed to penetrate into the squish region, the post jet, also portion of the split injection. A dual-excitation source laser
targeted into the squish, was able to enhance the net oxidation absorption scattering technique [89] was used to measure the
of the soot from the main injection that remained in the optical thickness of the liquid fuel and vapor fuel separately.
squish region. An overview of this explanation is shown in The geometry mimicked that of a light-duty engine, with a re-
Figure 19. entrant contour to the bowl. The results of this non-reacting
Finally, in studies where combustion is not initiated until study highlighted the role of mixing of the post jet with the
after both the main- and post-injection events (or is not main-injection mixture for soot reduction. The authors found
initiated at all), interaction of the spray with engine surfaces that not only did a short, properly-timed post injection
has been shown to change the fuel/air distribution before soot enhance mixing of the main-injection mixture, but also that
formation starts, decreasing the incidence of rich pockets that the impingement of the jets onto the bowl wall could be
would promote soot formation. Interaction of the fuel spray important. As the post jet impinged on the bowl wall, it
with the wall is common in small-bore engines, and often this flowed around the contour, pushing the main-injection
feature is exploited for better mixing of the fuel and air [87]. mixture away from the wall and enhancing the mixing of the
Even without wall interactions, however, variation in jet main-injection products with air from the center of the
development and mixing characteristics of main- plus-post- combustion chamber. The authors indicated that this post-jet
injection schedules have been measured. Two such examples mixing process could help to reduce fuel-rich zones that
are the unconfined diesel sprays (i.e., free jets) studied at promote soot formation.
engine conditions by Bruneaux and Maligne [88] and Parrish Similar results were discussed by Mendez and Thirouard
et al. [74]. Under certain close-coupled operating conditions [17], who showed that spray targeting and bowl shape could
in both these studies, the jet structure of the first injection was be designed together to enhance the efficacy of post
altered by the second injection, changing mixing injections at reducing soot and increasing burning efficiency.
characteristics even without the interaction of the jet with In this study, a narrow spray angle coupled with a highly
engine surfaces. contoured bowl shape guided the spray as it vaporized,
An example of the importance of the spray/wall depositing vapor in particular regions in the bowl. A planar
interaction for fuel/air mixture preparation with post laser-induced exciplex fluorescence (LIEF) technique [90]
was used with a fuel tracer to measure the density of fuel First, the general mechanism of late-cycle soot oxidation
vapor. needs to be clarified. This is an important first step, even
With a single injection, momentum carried much of the without considering post-injection strategies. The diesel
fuel out of the bowl and to the upper, outer edge of the piston conceptual model (Figure 8) proposed by Dec [43] describes
bowl. A properly-timed, short post-jet deposited more fuel in diesel combustion, but only up to the end of injection. Other
the bottom of the bowl, as it did not contain the momentum to work (e.g., [66]) provides some information about late-cycle
flow up toward the top of the bowl as the main injection did. soot oxidation, but a comprehensive conceptual-model
In this way, a more even fuel distribution between the bowl picture has not been established. To understand how post-
bottom and the upper, outer edge of the bowl was produced, injections alter late-cycle oxidation, it is first required to
eliminating fuel-rich zones that lead to soot formation. This understand late cycle oxidation without post-injections.
process is illustrated in their visualization, shown in Figure
• Immediately after the end of injection, does the diffusion
20.
flame progress upstream to enclose the entire soot cloud, as
postulated in the Soot Processes in Diesel Engines section of
this paper?
• Thereafter, how does the large soot cloud break up into
smaller pockets? Does a layer of OH remain on the periphery
of the soot pockets until they burn out, or is the OH depleted
before the soot is fully oxidized? Does soot oxidation occur
only at the diffusion flame, or is soot oxidized throughout the
pocket cross-section?
• What are the roles of O2 and OH in soot oxidation by
fracturing and surface reactions, respectively?
• Does soot formation continue within the pocket even as the
soot is being oxidized?
• What specific portion of the soot ultimately survives into
the exhaust? Can exhaust soot measurements be correlated to
optical observations of features of the burnout process?
• How does EGR, including high rates of EGR relevant to
low-temperature combustion, affect the late-cycle soot burn
out picture?
• How do other engine operating parameters, in particular
speed, affect the extent of both early-cycle and late-cycle soot
burnout?
With a clear understanding of late-cycle soot burnout for
single-injections, the effects of post-injections can be
Figure 20. Distribution of fuel vapor for a main- plus explored. Given that the post-injection is a fluid-mechanical
post-injection strategy that results in more evenly event, the foremost issue is the fluid mechanical
distributed fuel vapor. Copyright © SAE International. mechanism(s) by which post injections effect a reduction in
Reprinted with permission [17]. engine-out soot.
• To what degree does the post injection interact directly with
the main-injection soot pockets, i.e., by penetrating into them,
REMAINING RESEARCH and how important is that for net soot reduction?
QUESTIONS • How do post injections affect the OH and O2 oxidation
Despite the large literature on post injections for soot pathways, for instance by disrupting the diffusion flame
reduction, a clear, design-level understanding remains surrounding the residual soot pockets from the main
elusive. There are several remaining research questions that injection?
need to be answered before post-injection schemes can be
• Does the post injection increase the flame area of the
designed by applying understanding of fundamental chemical
diffusion flame surrounding the main-injection soot pockets?
and fluid-mechanical mechanisms. The following issues are
important for the advancement of this topic, but by no means • In what way does the post injection displace the main-
do they constitute an exhaustive list. injection soot, and what sort of displacement is beneficial for
soot reduction (e.g., into/out of bowl, into/out of squish, explored. Details of the soot reduction mechanism may
etc.)? change for different operating conditions and engine
architectures, and the details of this should be investigated.
• What, if any, bulk flows created by the post injection are
beneficial for soot reduction? • For example, how do the soot reduction mechanisms change
between light-duty and heavy-duty engines?
• Where does the post injection increase turbulence, and by
how much, and by what mechanisms? • How should combustion chambers for a given engine size
Additionally, many studies noted the importance of be designed to take advantage of the soot reduction benefits
injection duration on soot formation through the jet of post injections?
replenishment concept. The basis for this explanation comes
• Where should the post injection be targeted and at what
from basic understanding of how liquid jets penetrate through
time in the cycle or with what dwell relative to the main
gaseous media, which is a relatively well-understood
injection?
phenomenon in diesel applications. However, the importance
of this mechanism relative to other fluid-mechanic or thermal • What aspects of the mechanisms change at high EGR
mechanisms is unclear. conditions, where both temperature and oxygen content are
significantly reduced?
• What is the functional dependence of soot formation on
both main- and post-injection duration and how does this Finally, from an engine operational perspective, previous
change with operating condition? research has shown that post-injection efficacy is highly
dependent on several important engine operating parameters,
Beyond the fluid mechanical questions, chemical kinetics
such as load, speed, and boost. To date, very few studies have
issues are also unresolved. Foremost is the question of
isolated the effects of these parameters. With fundamental
temperature on soot oxidation, which has been argued to be
understanding, answers to these questions should be
an important mechanism by some [15, 19, 22] and dismissed
available, but confirming data for individual parameter
by others [28]. Ideally, the influence of temperature on soot
variations would be helpful.
reduction by post injections should be isolated from other
It is unlikely that all of these questions can be answered
effects, but it is unclear how this can be done, even with in-
entirely by experiments. Instead, many or most of these
cylinder diagnostics. Nevertheless, measurements of the
questions will likely be answered only by a combination of
effects of post-injection-induced temperature changes on soot
experiments and computer modeling. A good example of the
formation and oxidation would be helpful.
ways that computer modeling can complement and extend
• Is the temperature effect happening throughout the bulk of understanding provided by experiments is described in the
the charge, or only locally, in the vicinity of the post-injection work on engine fluid mechanics by Miles [67], which was
jet? used extensively in the Soot Processes in Diesel Engines
section. A similar approach using fluid mechanics with
• How does the main-plus-post injection create a temperature combustion and soot chemistry in computer models,
effect that is not otherwise realized by a single injection with combined with experimental data for validation wherever
the same total fuel mass? available, could provide the understanding required to answer
the questions posed here.
• Does temperature affect the OH and O2 soot oxidation
mechanisms differently? REFERENCES
1. Higgins, B. and Siebers, D., “Measurement of the Flame Lift-Off
• How important is the temperature effect relative to fluid- Location on DI Diesel Sprays Using OH Chemiluminescence,” SAE
mechanical effects? Technical Paper 2001-01-0918, 2001, doi:10.4271/2001-01-0918.
2. “Regulation (EC) No 595/2009 of 18 June 2009” E. Parliament (2009)
3. United States Federal Register, 76(179), (2011)
• Are there other chemical-kinetic effects of post injections 4. Hiranuma, S., Takeda, Y., Kawatani, T., Doumeki, R. et al.,
on soot reduction, such as effects on the fracturing and “Development of DPF System for Commercial Vehicle - Basic
Characteristic and Active Regenerating Performance,” SAE Technical
surface reactions of O2 and OH? Paper 2003-01-3182, 2003, doi:10.4271/2003-01-3182.
5. Salvat, O., Marez, P., and Belot, G., “Passenger Car Serial Application
More fundamentally, the balance between formation and of a Particulate Filter System on a Common Rail Direct Injection Diesel
oxidation also needs to be explored. The fluid mechanical and Engine,” SAE Technical Paper 2000-01-0473, 2000, doi:
10.4271/2000-01-0473.
chemical-kinetic/thermal effects on soot reduction by post- 6. Parks, J., Huff, S., Kass, M., and Storey, J., “Characterization of In-
injections could either reduce soot formation or increase soot Cylinder Techniques for Thermal Management of Diesel
Aftertreatment,” SAE Technical Paper 2007-01-3997, 2007, doi:
oxidation, or both. 10.4271/2007-01-3997.
7. Chartier, C., Andersson, O., Johansson, B., Musculus, M. et al., “Effects
• What are the relative magnitudes of changes to formation of Post-Injection Strategies on Near-Injector Over-Lean Mixtures and
Unburned Hydrocarbon Emission in a Heavy-Duty Optical Diesel
and oxidation by post injections? Engine,” SAE Int. J. Engines 4(1):1978-1992, 2011, doi:
10.4271/2011-01-1383.
With a better understanding of the fundamental aspects of 8. O'Connor, J. and Musculus, M., “Optical Investigation of the Reduction
post-injection soot reduction, more practical issues can be of UHC Emissions using Close-Coupled Post Injections at LTC
Conditions in a Heavy-Duty Diesel Engine,” SAE Int. J. Engines 6(1): brake specific fuel consumption of an automotive direct injection
379-399, 2013, doi:10.4271/2013-01-0910. common-rail diesel engine” Int. J. Eng. Res. 4(4):299-314 (2003)
9. Majewski, W. and Khair, M., “Diesel Emissions and Their Control,” 33. Benajes, J., Molina, S., and García, J., “Influence of Pre-and Post-
Society of Automotive Engineers, Inc., Warrendale, PA, ISBN Injection on the Performance and Pollutant Emissions in a HD Diesel
978-0-7680-0674-2, 2006, doi:10.4271/R-303. Engine,” SAE Technical Paper 2001-01-0526, 2001, doi:
10. Nehmer, D. and Reitz, R., “Measurement of the Effect of Injection Rate 10.4271/2001-01-0526.
and Split Injections on Diesel Engine Soot and NOx Emissions,” SAE 34. Payri, F., Benajes, J., Pastor, J., and Molina, S., “Influence of the Post-
Technical Paper 940668, 1994, doi: 10.4271/940668. Injection Pattern on Performance, Soot and NOx Emissions in a HD
11. Coldren, D., Schuricht, S., and Smith, R., “Hydraulic Electronic Unit Diesel Engine,” SAE Technical Paper 2002-01-0502, 2002, doi:
Injector with Rate Shaping Capability,” SAE Technical Paper 10.4271/2002-01-0502.
2003-01-1384, 2003, doi:10.4271/2003-01-1384. 35. Helmantel A, Somhorst J, Denbratt I, “Visualization of the effects of
12. Schoppe D, Zulch S, Hardy M, Geurts D, Jorach R, Baker N, “Common post injection and swirl on the combustion process of a passenger car
rail system with direct acting injector” MTZ 10 69 (2008) common rail DI diesel engine” ICES2003-622, Spring Technical
13. Dietsche K-H and Crepin J, Diesel engine management, 4th edition, ed. Conference of ASME ICE Division, Salzberg Austria (2003)
Bauer H., SAE International, Stuttgart Germany (2006) 36. Mancaruso E, Merola S, Vaglieco B, “Study of the multi-injection
14. Greeves, G., Tullis, S., and Barker, B., “Advanced Two-Actuator EUI combustion process in a transparent direct injection common rail diesel
and Emission Reduction for Heavy-Duty Diesel Engines,” SAE engine by means of optical techniques” Int. J. Eng. Res. 9(6):483-498
Technical Paper 2003-01-0698, 2003, doi:10.4271/2003-01-0698. (2008)
15. Bobba, M., Musculus, M., and Neel, W., “Effect of Post Injections on 37. Bobba, M., Genzale, C., and Musculus, M., “Effect of Ignition Delay on
In-Cylinder and Exhaust Soot for Low-Temperature Combustion in a In-Cylinder Soot Characteristics of a Heavy Duty Diesel Engine
Heavy-Duty Diesel Engine,” SAE Int. J. Engines 3(1):496-516, 2010, Operating at Low Temperature Conditions,” SAE Int. J. Engines 2(1):
doi:10.4271/2010-01-0612. 911-924, 2009, doi:10.4271/2009-01-0946.
16. Vanegas, A., Won, H., Felsch, C., Gauding, M. et al., “Experimental 38. Han, Z., Uludogan, A., Hampson, G., and Reitz, R., “Mechanism of
Investigation of the Effect of Multiple Injections on Pollutant Formation Soot and NOx Emission Reduction Using Multiple-injection in a Diesel
in a Common-Rail DI Diesel Engine,” SAE Technical Paper Engine,” SAE Technical Paper 960633, 1996, doi: 10.4271/960633.
2008-01-1191, 2008, doi:10.4271/2008-01-1191. 39. Molina, S., Desantes, J., Garcia, A., and Pastor, J., “A Numerical
17. Mendez, S. and Thirouard, B., “Using Multiple Injection Strategies in Investigation on Combustion Characteristics with the use of Post
Diesel Combustion: Potential to Improve Emissions, Noise and Fuel Injection in DI Diesel Engines,” SAE Technical Paper 2010-01-1260,
Economy Trade-Off in Low CR Engines,” SAE Int. J. Fuels Lubr. 1(1): 2010, doi:10.4271/2010-01-1260.
662-674, 2009, doi:10.4271/2008-01-1329. 40. Desantes, J., Arrègle, J., López, J., and García, A., “A Comprehensive
18. Zhang, Y. and Nishida, K., “Vapor/Liquid Behaviors in Split-Injection Study of Diesel Combustion and Emissions with Post-injection,” SAE
D.I. Diesel Sprays in a 2-D Model Combustion Chamber,” SAE Technical Paper 2007-01-0915, 2007, doi:10.4271/2007-01-0915.
Technical Paper 2003-01-1837, 2003, doi:10.4271/2003-01-1837. 41. Beatrice C, Belardini P, Bertoli C, Lisbona M, Rossi Sebastiano G,
19. Hotta, Y., Inayoshi, M., Nakakita, K., Fujiwara, K. et al., “Achieving “Diesel combustion control in common rail engines by new injection
Lower Exhaust Emissions and Better Performance in an HSDI Diesel strategies” Int. J. Eng. Res. 3(1):23-36 (2002)
Engine with Multiple Injection,” SAE Technical Paper 2005-01-0928, 42. Arrègle J, Pastor J V, López J J, García A, “Insights on post-injection-
2005, doi:10.4271/2005-01-0928. associated soot emissions in direct injection diesel engines” Combustion
20. Chen, S., “Simultaneous Reduction of NOx and Particulate Emissions and Flame 154(3):448-461 (2008)
by Using Multiple Injections in a Small Diesel Engine,” SAE Technical 43. Dec, J., “A Conceptual Model of DI Diesel Combustion Based on Laser-
Paper 2000-01-3084, 2000, doi:10.4271/2000-01-3084. Sheet Imaging,” SAE Technical Paper 970873, 1997, doi:
21. Montgomery, D. and Reitz, R., “Effects of Multiple Injections and 10.4271/970873.
Flexible Control of Boost and EGR on Emissions and Fuel Consumption 44. Anselmi, P., Kashdan, J., Bression, G., Ferrero-Lesur, E. et al.,
of a Heavy-Duty Diesel Engine,” SAE Technical Paper 2001-01-0195, “Improving Emissions, Noise and Fuel Economy Trade-Off by using
2001, doi:10.4271/2001-01-0195. Multiple Injection Strategies in Diesel Low Temperature Combustion
22. Yun H, Sun Y, Reitz R D, “An experimental and numerical (LTC) Mode,” SAE Technical Paper 2010-01-2162, 2010, doi:
investigation on the effect of post injection strategies on combustion and 10.4271/2010-01-2162.
emissions in the low-temperature diesel combustion regime” 45. Desantes J M, Bermúdez V, García A, Linares W G, “A comprehensive
ICES2005-1043, ASME Internal Combustion Engine Division 2005 study of particle size distributions with the use of post injection
Spring Technical Conference, Chicago IL (2005) strategies in DI diesel engines” Aerosol Sci. Tech. 45(10):1161-1175
23. Yun H and Reitz R D, “An experimental investigation on the effect of (2011)
post-injection strategies on combustion and emissions in the low- 46. Tree D R and Svensson K I, “Soot processes in compression ignition
temperature diesel combustion regime” J. Eng. Gas Turb. Power engines” Prog. Energy Combust. Sci. 33(3):272-309 (2007)
129:279-286 (2007) 47. Haynes B and Wagner H G, “Soot formation” Prog. Energy Combust.
24. Tow, T., Pierpont, D., and Reitz, R., “Reducing Particulate and NOx Sci. 7(4):229-273 (1981)
Emissions by Using Multiple Injections in a Heavy Duty D.I. Diesel 48. Glassman I, “Soot formation in combustion processes” Proc. Combust.
Engine,” SAE Technical Paper 940897, 1994, doi: 10.4271/940897. Inst. 22(1):317-378 (1989)
25. Pierpont, D., Montgomery, D., and Reitz, R., “Reducing Particulate and 49. Wang H, “Formation of nascent soot and other condensed-phase
NOx Using Multiple Injections and EGR in a D.I. Diesel,” SAE materials in flames” Proc. Combust. Inst. 33(1):41-67 (2011)
Technical Paper 950217, 1995, doi: 10.4271/950217. 50. Homann K, “Carbon formation in premixed flames” Combust. Flame
26. Ehleskog, R. and Ochoterena, R., “Soot Evolution in Multiple Injection 11(4):265-287 (1967)
Diesel Flames,” SAE Technical Paper 2008-01-2470, 2008, doi: 51. Richter H and Howard J, “Formation of polycyclic aromatic
10.4271/2008-01-2470. hydrocarbons and their growth to soot-a review of chemical reaction
27. Dronniou, N., Lejeune, M., Balloul, I., and Higelin, P., “Combination of pathways” Prog. Energy Combust. Sci. 26(4):565-608 (2000)
High EGR Rates and Multiple Injection Strategies to Reduce Pollutant 52. Miller J A and Melius C F, “Kinetic and thermodynamic issues in the
Emissions,” SAE Technical Paper 2005-01-3726, 2005, doi: formation of aromatic compounds in flames of aliphatic fuels” Combust.
10.4271/2005-01-3726. Flame 91(1):21-39 (1992)
28. Barro C, Tschanz F, Obrecht P, Boulouchos K, “Influence of post- 53. Bartok W and Sarofim A F, Fossil fuel combustion, John Wiley and
injection parameters on soot formation and oxidation in a common-rail- Sons Inc., New York NY (1990)
diesel engine using multicolor-pyrometry” ICEF2012-92075, ASME 54. Fenimore C P and Jones G W, “Oxidation of soot by hydroxyl radicals”
Internal Combustion Engine Division Fall Technical Conference, J. Phys. Chem. 71(3):593-597 (1967)
Vancouver BC (2012) 55. Garo A, Lahaye J, Prado G, “Mechanisms of formation and destruction
29. Sperl, A., “The Influence of Post-Injection Strategies on the Emissions of soot particles in a laminar methane-air diffusion flame” Proc.
of Soot and Particulate Matter in Heavy Duty Euro V Diesel Engine,” Combust. Inst. 21(1):1023-1031 (1988)
SAE Technical Paper 2011-36-0350, 2011, doi:10.4271/2011-36-0350. 56. Neoh K G, Howard J B, Sarofim A F, “Effect of oxidation on the
30. Konno, M., Chikahisa, T., and Murayama, T., “Reduction of Smoke and physical structure of soot” Proc. Combust. Inst. 20(1):951-957 (1985)
NOx by Strong Turbulence Generated During the Combustion Process 57. Echavarria C A, Jaramillo I C, Sarofim A F, Lighty J S, “Burnout of
in D.I. Diesel Engines,” SAE Technical Paper 920467, 1992, doi: soot particles in a two-stage burner with a JP-8 surrogate fuel” Combust.
10.4271/920467. Flame 159(7):2441-2448 (2012)
31. Bakenhus, M. and Reitz, R., “Two-Color Combustion Visualization of 58. Echavarria C A, Jaramillo I C, Sarofim A F, Lighty J S, “Studies of soot
Single and Split Injections in a Single-Cylinder Heavy-Duty D.I. Diesel oxidation and fragmentation in a two-stage burner under fuel-lean and
Engine Using an Endoscope-Based Imaging System,” SAE Technical fuel-rich conditions” Proc. Combust. Inst. 33(1):659-666 (2011)
Paper 1999-01-1112, 1999, doi:10.4271/1999-01-1112. 59. Kennedy I M, “Models of soot formation and oxidation” Prog. Energy
32. Badami M, Mallamo F, Millo F, Rossi E, “Experimental investigation Combust. Sci. 23(2):95-132 (1997)
on the effect of multiple injection strategies on emissions, noise and
60. Frenklach M, Clary D W, Gardiner W CJr., Stein S E, “Detailed kinetic 86. Miles, P., “The Influence of Swirl on HSDI Diesel Combustion at
modeling of soot formation in shock-tube pyrolysis of acetylene” Proc. Moderate Speed and Load,” SAE Technical Paper 2000-01-1829, 2000,
Combust. Inst. 20(1):887-901 (1985) doi:10.4271/2000-01-1829.
61. Frenklach M and Wang H, “Detailed modeling of soot particle 87. Sahoo, D., Petersen, B., and Miles, P., “Measurement of Equivalence
nucleation and growth” Proc. Combust. Inst. 23(1):1559-1566 (1991) Ratio in a Light-Duty Low Temperature Combustion Diesel Engine by
62. Frenklach M and Wang H, “Detailed mechanism and modeling of soot Planar Laser Induced Fluorescence of a Fuel Tracer,” SAE Int. J.
particle formation” Springer Series in Chemical Physics 59:162-162 Engines 4(2):2312-2325, 2011, doi:10.4271/2011-24-0064.
(1994) 88. Bruneaux, G. and Maligne, D., “Study of the Mixing and Combustion
63. Kitamura T, Ito T, Senda J, Fujimoto H, “Mechanism of smokeless Processes of Consecutive Short Double Diesel Injections,” SAE Int. J.
diesel combustion with oxygenated fuels based on the dependence of the Engines 2(1):1151-1169, 2009, doi:10.4271/2009-01-1352.
equivalence ration and temperature on soot particle formation” Int. J. 89. Zhang Y, Yoshizaki T, Nishida K, “Imaging of droplets and vapor
Eng. Res. 3(4):223-248 (2002) distributions in a diesel fuel spray by means of a laser absorption-
64. Pickett L and Siebers D, “Soot formation in diesel fuel jets near the lift- scattering technique” Appl. Opt. 39(33):6221-6229 (2000)
off length” Int. J. Eng. Res. 7(2):103-130 (2006) 90. Bruneaux G, “Liquid and vapor spray structure in high-pressure
65. Pickett L M and Siebers D L, “Soot in diesel fuel jets: effects of ambient common rail diesel injection” Atomization and Sprays 11(5):533-556
temperature, ambient density, and injection pressure” Combust. Flame (2001)
138(1):114-135 (2004)
66. Dec, J. and Kelly-Zion, P., “The Effects of Injection Timing and Diluent
Addition on Late-Combustion Soot Burnout in a DI Diesel Engine
CONTACT INFORMATION
Based on Simultaneous 2-D Imaging of OH and Soot,” SAE Technical
Paper 2000-01-0238, 2000, doi:10.4271/2000-01-0238. For more information, please contact:
67. Miles P C, “Turbulent flow structure in direct-injection, swirl-supported Jacqueline O'Connor
diesel engines flow and combustion in reciprocating engines” in Flow
and combustion in reciprocating engines, ed. Arcoumanis C. and Sandia National Laboratories
Kamimoto T., Springer Berlin, Heidelberg Germany (2009) joconnor@sandia.gov
68. Dembinski, H. and Angstrom, H., “Optical Study of Swirl during
Combustion in a CI Engine with Different Injection Pressures and Swirl
Ratios Compared with Calculations,” SAE Technical Paper Mark Musculus
2012-01-0682, 2012, doi:10.4271/2012-01-0682.
69. Atzler, F., Kastner, O., Rotondi, R., and Weigand, A., “Multiple Sandia National Laboratories
Injection and Rate Shaping Part 1: Emissions Reduction in Passenger mpmuscu@sandia.gov
Car Diesel Engines,” SAE Technical Paper 2009-24-0004, 2009, doi:
10.4271/2009-24-0004.
70. Atzler, F., Kastner, O., Rotondi, R., and Weigand, A., “Multiple
injection and rate shaping Part 2: Emissions Reduction in Passenger Car ACKNOWLEDGMENTS
Diesel Engines Computational Investigation,” SAE Technical Paper
2009-24-0012, 2009, doi:10.4271/2009-24-0012. Support for this research was provided by the U.S.
71. “Directive 1999/96/EC” O.J.o.t.E. Communities (1999)
72. “EEC Directive 90/C81/01” O.J.o.t.E. Communities (1990) Department of Energy, Office of Vehicle Technologies.
73. Shayler, P., Brooks, T., Pugh, G., and Gambrill, R., “The Influence of Sandia is a multi-program laboratory operated by Sandia
Pilot and Split-Main Injection Parameters on Diesel Emissions and Fuel
Consumption,” SAE Technical Paper 2005-01-0375, 2005, doi: Corporation, a Lockheed Martin Company for the United
10.4271/2005-01-0375. State Department of Energy's National Nuclear Security
74. Parrish, S., Zhang, G., and Zink, R., “Liquid and Vapor Envelopes of
Sprays from a Multi-Hole Fuel Injector Operating under Closely-Spaced Administration under contract DE-AC04-94AL85000. The
Double-Injection Conditions,” SAE Int. J. Engines 5(2):400-414, 2012, authors would like to thank Paul Miles and Scott Skeen of
doi:10.4271/2012-01-0462.
75. Babu, A. and Devaradjane, G., “Control of Diesel Engine Pollutants by Sandia National Laboratories and Philip Dingle of Delphi
Split Injection Method using Muitizone Model,” SAE Technical Paper Diesel Systems for their many helpful discussions.
2001-28-0007, 2001, doi:10.4271/2001-28-0007.
76. Bower, G. and Foster, D., “The Effect of Split Injection on Fuel
Distribution in an Engine-Fed Combustion Chamber,” SAE Technical
Paper 930864, 1993, doi: 10.4271/930864.
77. Tanin, K., Wickman, D., Montgomery, D., Das, S. et al., “The Influence
DEFINITIONS/ABBREVIATIONS
of Boost Pressure on Emissions and Fuel Consumption of a Heavy-Duty
Single-Cylinder D.I. Diesel Engine,” SAE Technical Paper AHRR - Apparent heat release rate
1999-01-0840, 1999, doi:10.4271/1999-01-0840.
78. Yang, B., Mellor, A., and Chen, S., “Multiple Injections with EGR ASOI - After start of injection
Effects on NOx Emissions for DI Diesel Engines Analyzed Using an
Engineering Model,” SAE Technical Paper 2002-01-2774, 2002, doi: ATDC - After top dead center
10.4271/2002-01-2774. BSFC - Break-specific fuel consumption
79. O'Connor J and Musculus M, “Effects of EGR and load on soot in a
heavy-duty optical diesel engine with close-coupled post-injections for CA/°CA - Degrees crank angle (duration)
high efficiency combustion phasing” submitted to Int. J. Eng. Res.
(2013) CAD - Crank angle degree
80. Siebers, D. and Higgins, B., “Flame Lift-Off on Direct-Injection Diesel
Sprays Under Quiescent Conditions,” SAE Technical Paper CCD - Combustion chamber for disturbance
2001-01-0530, 2001, doi:10.4271/2001-01-0530.
81. Heywood J B, Internal combustion engine fundamentals, McGraw-Hill, CFD - Computational fluid dynamics
New York, NY (1988)
82. Naber, J. and Siebers, D., “Effects of Gas Density and Vaporization on DOI - Actual injection duration
Penetration and Dispersion of Diesel Sprays,” SAE Technical Paper
960034, 1996, doi: 10.4271/960034. DOI1C - Commanded duration of main injection (in
83. Böhm H, Hesse D, Jander H, Lüers B, Pietscher J, Wagner H G G, microseconds)
Weiss M, “The influence of pressure and temperature on soot formation
in premixed flames” Proc. Combust. Inst. 22(1):403-411 (1989) DOI2C - Commanded duration of post injection (in
84. Liu, Y. and Reitz, R., “Optimizing HSDI Diesel Combustion and
Emissions Using Multiple Injection Strategies,” SAE Technical Paper microseconds)
2005-01-0212, 2005, doi:10.4271/2005-01-0212.
85. Ferguson C R and Kirkpatrick A T, Internal combustion engines, EGR - Exhaust gas recirculation
Applied Thermosciences, ed. Hayton J., John Wiley & Sons, Inc., New
York NY (2001) ESC - European steady-state test cycle
FSN - Filter smoke number
gIMEP - Gross indicated mean effective pressure

IMEP - Indicated mean effective pressure
KIVA - 3-dimensional computational fluid mechanics code
for engine simulation
kL - Soot optical density
LIEF - Laser-Induced Exciplex Fluorescence
LTC - Low temperature combustion
NEDC - New European Design Cycle
ΔP - Pressure drop across injector nozzle
PAH - Poly-cyclic aromatic hydrocarbon
PM - Particulate matter
Prail - Rail pressure
RPM - Rotations per minute
SCR - Selective catalytic reduction (exhaust aftertreatment)
SOI - Actual start of injection
SOI1C - Commanded start of main injection (in crank angle
degrees)
SOI2C - Commanded start of post injection (in crank angle
degrees)
T - Temperature
TDC - Top dead center
ρo - Ambient density
ϕ - Equivalence ratio

10.4271@2013 01 0917

Uploaded by

Copyright:

Available Formats

10.4271@2013 01 0917

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10.4271@2013 01 0917

Uploaded by

Copyright:

Available Formats

Downloaded from SAE International by Univ of Nottingham - Kings Meadow Campus, Friday, August 10, 2018

Post Injections for Soot Reduction in Diesel Engines: A Review

INTRODUCTION injections per cycle. Early pump-line-nozzle fuel-injection

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 401

402 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

injections can be correlated to increased in-cylinder

Figure 2. Soot temperature in the squish region showing

Figure 1. Predicted contours of equivalence ratio for a

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 403

creating a larger fuel-rich mixture that supported more soot

Figure 4. Ratio of engine-out soot with a post-injection

404 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 405

Figure 5. Process of soot formation in flames [46].

406 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

Figure 7. Closed-reactor simulation of soot formation in

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 407

408 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 409

410 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 411

Exhaust Gas Recirculation

412 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 413

414 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 415

416 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 417

418 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 419

420 O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013)

O'Connor et al / SAE Int. J. Engines / Volume 6, Issue 1(May 2013) 421

gIMEP - Gross indicated mean effective pressure

You might also like