Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Pulsed Regeneration for DPF Aftertreatment 2011-24-0182

Published
Devices 09/11/2011

Jian Gong and Christopher J. Rutland

Univ. of Wisconsin Madison

Copyright © 2011 SAE International

doi:10.4271/2011-24-0182

the US [5]. These requirements post great challenges for

ABSTRACT advanced DPF regenerations control and strategies.
DPF regenerations involve a trade-off between fuel economy
and DPF durability. High temperature regenerations of DPFs For DPF regeneration control, one critical problem is to
have fewer fuel penalties but simultaneously tend to give estimate the actual soot trapped inside the DPF. “Over-
higher substrate temperatures, which can reduce thermal loading” with soot could lead to DPF damage, while “under-
reliability. In order to weaken the trade-off, the integrated loading” may result in unnecessary DPF regenerations.
system-level model [1,2,3,4] is used to conduct optimization Pressure drop based models can be used to predict soot mass
studies and explore novel regeneration strategies for DPF in reasonable accuracy during DPF loading [6,7,8]. However,
aftertreatment devices. The integrated model developed in the they are quite challenging during DPF regenerations since
Engine Research Center (ERC) includes sub-models for soot oxidation in a DPF is complicated and involves surface
engines, emissions, aftertreatment devices and controllers. and gas phase chemistry, mass, momentum, and energy
Based on the engine and regeneration fuel economy, multiple transport phenomena.
and single cycle regeneration tests are performed and
analyzed. The optimal soot loadings to initiate and terminate However, recent real-time soot sensor technologies have
regenerations are discussed. A pulsed regeneration strategy, made accurate estimates of soot possible and provide great
which is characterized by injecting multiple pulses of fuel potential to sophisticated DPF regeneration control and
(upstream of a DOC) during regenerations, is investigated. It regeneration strategies development. An optical system based
is found that pulsed regeneration has the potential to reduce on light scattering techniques demonstrated the capability of
regeneration fuel penalties without generating significantly measuring temporal variations of cumulative soot mass and
high wall temperatures that can reduce DPF durability. filtration efficiency of DPFs [9]. Bosch developed a soot
sensor based on multi-layer ceramic sensor technology and
showed excellent signal reproducibility on an artificial soot
INTRODUCTION particle test bench [10].
Diesel particulate filters (DPFs) have been shown to be very
effective aftertreatment devices to control particulate In order to control the duration of DPF regenerations, the
emissions from diesel engines. One of the main problems ending time of active regenerations was investigated by
associated with low-cost and robust DPF systems Kodama et al. [11]. It was shown that the total quantity of
development is regeneration. Active regenerations of DPFs oxygen mass passing through the filter during active
lead to fuel penalties. The total amount of fuel used for regenerations had a strong correlation with the amount of PM
regenerations in a DPF's lifetime is very large. In addition, oxidized. By using change of oxygen mass rather than
DPF systems are expected to be as robust as possible. In constant time as a judgment of ending, a significant
Europe DPF aftertreatment systems are expected to meet the improvement of about 20% in fuel consumption was
durability target of 500,000 km, while, the durability target is observed during road tests.
allowed to be approximately 700,000 km (435,000 miles) in
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Regeneration strategies are very important to DPF

aftertreatment device operations. Bach et al. [12] enabled
regenerations to take place at low exhaust temperatures by
combining different strategies. An electric heating module
with low heat capacity was placed upstream of a DPF to
increase exhaust temperature, and fuel additives were used to
lower the particulate ignition temperature. The limitation for
this configuration was that soot loading should be kept low to
prevent the high risk of uncontrolled regenerations. Another
regeneration strategy based on two post injections (late fuel
injection into cylinders) was investigated by Hiranuma et al.
[13].

The first post injection supplied unburned fuel to the pre- Figure 1. Aftertreatment system model for DPF
catalyst to increase exhaust temperature. The second post regenerations studies
injection was used for filter heat-up resulting in sufficiently
high temperatures to combust soot in the filter. A close-loop
control employing pre-catalyst outlet temperature feedback The DOC is modeled as a typical 1-D channel flow device,
was used to control the post injection and stable active including kinetics to oxidize CO, hydrocarbon, and NO to
regenerations were observed in urban driving conditions. CO2, H2O and NO2 respectively. The DPF is modeled as a 1-
D two-layer wall flow device, including soot filtration and
A combination of a fuel vaporizer with early post-injection soot oxidation kinetics. This DPF model in the aftertreatment
for DPF regenerations was studied by Alano et al. [14] and system model was calibrated for the 1.9 GM multi-cylinder,
showed promise. However, a hydrocarbon slip was observed automotive diesel engine by Navtej et al [3]. Details about the
when a large quantity of diesel fuel was vaporized. Park et al. DOC and DPF models can be found in [16, 17]. The DPF
[15] found hydrogen could be used to decrease PM oxidation specification in this study is shown in Table 1.
temperatures significantly on a catalyzed DPF. By using a
temperature controlled partial flow reactor system, they found
Table 1. DPF specification
soot oxidation in a catalyzed DPF started when hydrogen
concentration was 3% and the temperature was as low as
180°C. Also, regenerations became faster if the hydrogen
concentration was increased.

MODEL DESCRIPTION
A standalone version of aftertreatment system models is used
in this study. The standalone aftertreatment system models
are part of the integrated models [1,2,3,4]. Figure 1 shows the
structure of the aftertreatment system models, including
Diesel Oxidation Catalyst (DOC), Diesel Particulate Filter
(DPF) and controller models. The model inputs include a
time resolved engine exhaust stream, including the exhaust Engine operating conditions at mode 4a in Table 2 are chosen
mass flow rate, temperature, and pollutant species. The in this study for several reasons. First, exhaust temperature is
exhaust stream is the output from a 1.9L turbocharged GM not high enough to perform intensive passive regenerations.
automotive diesel engine, which is running at specific This is very important so that most of the soot oxidation
operating conditions. The engine exhaust first goes into the occurs during active regenerations. Furthermore, most of the
DOC, then the DPF, and eventually exits the tail pipe. injected fuel ahead of a DOC could be oxidized at this
Meanwhile, the engine exhaust and the state of the DOC and temperature. Second, the oxygen concentration is high
DPF devices are monitored by controllers to control the enough to combust the injected exhaust fuel and trapped soot.
regeneration process. In other words, oxygen is not a limitation factor to active
regenerations in this study.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Table 2. Engine operating conditions used for DPF

regenerations

Figure 2. Regeneration fuel penalty and back pressure

fuel penalty at different maximum soot loadings [3]
REGENERATION OPTIMIZATION
The objective of regeneration optimization is to achieve
efficient and reliable regenerations without sacrificing the WHEN TO TERMINATE
DPFs' durability. “Efficient” means less extra energy is REGENERATIONS
required for regeneration. “Reliable” means regenerations
In the aftertreatment control model, when to terminate
should be controlled without generating significantly high
regenerations is determined by a specified minimum soot
wall temperatures; otherwise, uncontrolled regenerations will
loading. Unlike the maximum soot loading, which has a great
lead to runaway or permanent damage of a DPF.
effect on DPF durability, minimum soot loading affects
regeneration performance in terms of fuel economy.
For DPF regenerations control and optimization, several
control points need to be addressed:
a. when to initiate regenerations;
Multiple Cycle Test
b. when to terminate regenerations; Multiple regeneration cycle tests were performed to seek the
c. how to efficiently and safely burn trapped soot during optimal soot loading for terminating regenerations based on
regenerations the fuel economy. Initial conditions of DPF are shown in
Table 3 and minimum soot loading to terminate regenerations
varies from 1g/l to 4.5 g/l. Even though the total simulation
WHEN TO INITIATE REGENERATIONS time is about 18 hours, the time of 16.5 hours is chosen for
Basically, many studies [3, 18] have shown that less frequent evaluation to ensure that there is almost the same amount of
regenerations are better for engine thermal management since soot left in the DPF at the end of simulation for all of the
less frequent regenerations extended the lifetime of DPFs. tests. The amount of soot inside DPF at different minimum
The frequency of regenerations or the interval between soot loadings for multiple regeneration cycles can be seen
regenerations is directly related to when to initiate and from Figure 3.
terminate regenerations.
Table 3. DPF initial condition setup for terminating
Multiple regeneration cycle tests, which considered the fuel regeneration test
penalties due to an increased engine back pressure and active
regenerations, were conducted by Navtej et al. using the
integrated model [3]. Based on the fuel economy, five
different soot loadings, varying from 1.5∼5.75 g/l, were
studied to investigate optimal soot loading to initiate
regenerations. Figure 2 shows the regeneration fuel penalties
and back pressure fuel penalties at different maximum soot
loadings. At higher soot loadings, the regeneration fuel
penalty is lower, while the engine fuel penalty is larger due to
an increased engine back pressure. A trade-off point between
regeneration and back pressure fuel penalties is observed at
around 5.5 g/l in his study. Furthermore, experimental work
conducted by Boger et al. [19] showed that peak filter
temperature increased with maximum soot loading, and they
suggested that regenerations at very high soot loading are not
preferred for DPF systems.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Figure 3. The amount of soot trapped in the DPF at Figure 5. Regeneration fuel penalty and back pressure
different minimum soot loadings fuel penalty at different minimum soot loadings

To calculate the engine back pressure fuel penalty, the Single Regeneration Cycle Test
correlation in Figure 4 between the back pressure and fuel
injection rate of the engine is used. The fuel injection rate in To further understand the regeneration process at different
Figure 4 looks linear to the engine back pressure. This was soot loadings, the amount of exhaust fuel injected for
also found in experiments by Zhan et al. [20], which showed regenerations is varied from 75∼400g in a single
a linear relationship between fuel consumption and post-turbo regeneration event (without considering the engine back
pressure for loaded DPF at different engine operating pressure fuel penalty). The DPF initial conditions are almost
conditions. the same as Table 3, except for the DPF regeneration
temperatures. The exhaust fuel injection rate is fixed for a
specific target regeneration temperature.

Figure 6 shows that fuel efficiency increases as injected fuel

increases, then it begins to drop after reaching a peak. This
behavior can be explained with the help of Figure 7. When
injected fuel quantity is low, most of the fuel is used to warm
up the DPF rather than combust soot, which results in very
low fuel efficiency. This can be seen in Figure 7, which
shows a very low DPF temperature when only a small
amount of fuel is injected. As injected fuel increases, the DPF
wall temperature reaches the target; meanwhile, there is
enough soot in the DPF, resulting in a higher soot oxidation
rate and higher fuel efficiency. As more fuel is injected, fuel
Figure 4. Correlation between engine back pressure and efficiency decreases because of low soot loading even though
engine fuel rate the regeneration temperature is still high. These observations
indicate “deep” regeneration is not necessary, which was also
found by Arasappa et al. [23]. Furthermore, considering
Regeneration fuel penalties and engine back pressure fuel filtration efficiency, it is not recommended to completely
penalties at different minimum soot loadings are shown in regenerate DPFs as well, since soot particles can easily
Figure 5. The regeneration fuel penalty is very consistent escape from a very clean filter [23].
with the experimental work in [21, 22], which showed about
3∼4% of regeneration fuel penalty on light duty diesel
engines. The back pressure fuel penalty increases as
minimum soot loading increases. However, the regeneration
fuel penalty shows a concave shape as minimum soot loading
increases. These observations indicate that optimal soot
loading for terminating regenerations exists, which is around
3∼3.5 g/l. More discussion about optimal minimum soot
loading will be given in single regeneration cycle tests.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Figure 6. Fuel efficiency and soot loading with a Figure 8. Fuel efficiency and soot loading with a
different amount of exhaust fuel injected at regeneration different amount of exhaust fuel injected at regeneration
temperature of 550°C temperature of 575°C

Figure 7. Soot oxidation rate and DPF effective wall

Figure 9. Optimal minimum soot loading to terminate
temperature with a different amount of exhaust fuel
regenerations based on fuel efficiency at different
injected at regeneration temperature of 550°C
regeneration temperatures

Regeneration at a target temperature of 575°C in Figure 8

shows a similar trend as Figure 6. Figure 9 shows the optimal THERMAL INERTIA EFFECT
soot loading at different regeneration temperatures based on Another interesting phenomena observed from the
fuel efficiency. It can be seen that the optimal soot loading simulations is the thermal inertia effect, which could be used
points, which are linked by the red line, are strongly to further improve regeneration fuel efficiency. Typically,
dependent on the regeneration temperature. when fuel injection ends, a DPF begins to cool down.
However, the DPF wall temperature is still high enough to
combust soot. This can be seen from Figure 10. For a typical
regeneration control, fuel injections are terminated when soot
loading hits the minimum limit. (Soot is still oxidized for a
short time in this case but fuel efficiency is very low for deep
regenerations, as is discussed above). By advancing the end
of the injection time a little (black solid line), about 12 grams
of fuel (9% of total fuel) could be saved.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

PULSED REGENERATION CONCEPT

In traditional active regenerations of DPFs, exhaust fuel is
continuously injected upstream of a DOC for a period of time
when the engine back pressure reaches a high limit. Injection
is stopped when the engine back pressure drops to a low
limit. There is only a single fuel injection during an active
regeneration period.

The idea for multiple injections used for active regenerations

is inspired by Khan et al [24]. In their studies, thermal and
catalytic regeneration characteristics of a DPF under
simulated time varying exhaust conditions were investigated
by employing a one-dimensional DPF model. Their results
showed that an artificially imposed modulation of exhaust gas
Figure 10. Illustration of thermal inertia effect temperature had a positive effect on the filter fuel efficiency
for both thermal and catalytic assisted regeneration
mechanisms.
PULSED REGENERATION
After answering how to initiate and terminate regenerations,
the remaining question of how to efficiently and safely burn
trapped soot during regeneration naturally comes up. Many
researchers have shown that split fuel injection strategies
(pilot, main and post fuel injections) could be used in engine
combustion to weaken the trade-off between in-cylinder NOx
and soot emission in diesel engines. There is a similar trade-
off between fuel efficiency and DPF durability during
regenerations, so what about using multiple injections for
DPF regenerations?

Figure 12. Exhaust fuel rate and trapped soot in DPF

during pulsed regeneration

A similar modulation of exhaust gas temperature could be

achieved by applying pulsed regeneration strategies. Pulsed
regeneration is characterized by a magnitude of a pulse (fuel
injection rate), a pulse period or the number of pulses and
dwell time or dwell ratio. These characteristics are shown and
described in Figure 12. The magnitude of pulse is determined
Figure 11. Amount of fuel and DPF wall maximum by a target regeneration temperature. Each pulse period
temperature for regenerations at different regeneration consists of pulse width (injection time) and dwell time. A
temperatures longer injection period means less frequent injections or
fewer numbers of injections. The general idea of this concept
High temperature regenerations require less fuel compared to is using multiple injections to regenerate DPFs. During the
low temperature regenerations. However, they are much dwell time period between each injection, the DPF wall
easier to lead to very high DPF wall temperatures, which are temperature is still high enough to burn soot due to large
not favored considering DPF durability. An extremely high thermal capacity. This helps to save fuel and lower the
DPF temperature will not only deactivate the catalyst coated effective wall temperatures.
on the wall surface, but also result in permanent damage of
DPF substrate materials. The trade-off between fuel economy
and DPF durability can be seen from Figure 11.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

temperature at an acceptable level. Multi-pulsed regeneration

strategies are explored in this section.

Engine operating conditions are the same as shown in Table

2. DPF initial conditions are shown in Table 4. The number
of injections is changing from 2 to 10 by changing the period
of pulses while keeping the dwell time ratio at 50%. A dwell
ratio of 50% means fuel is injected for half time and then
terminated for another half time during a period of a pulse
injection. The single injection strategy is regarded as the
baseline for comparison. The magnitude of a pulse is
changing to achieve the different target regeneration
temperatures of 550°C, 575°C and 600°C.

Table 4. DPF initial conditions setup for pulsed

regenerations
Figure 13. Exhaust fuel rate and DOC and DPF wall
temperature during pulsed regeneration

Figure 14. Soot oxidation and deposition rate (engine

out soot flow rate is 3.5 mg/s)

The wall temperatures of DOC and DPF in Figure 13

oscillate (increase and decrease) as fuel injection is initiated Figure 15. Fuel efficiency for single and multi-pulsed
and terminated. Therefore, the average wall temperatures of regenerations at 550°C, 575°C, and 600°C (mode 4a)
DOC and DPF are greatly reduced because of the repeated
heating up and cooling down process.
Figure 15 shows that fuel efficiency decreases as the number
of injection increases at all three different regeneration
Figure 14 shows that the soot oxidation rate during dwell
temperatures. This indicates that too frequent injections will
time is still high compared to the engine out soot flow rate.
diminish the benefit on fuel efficiency. This is exactly the
This indicates soot is still oxidized during dwell time. The
same as what was found by Khan et al. [24]. They found that
peaks of the soot oxidation rate corresponding to the three
the positive effect on fuel efficiency for both thermal and
pulses decrease with time because the amount of soot left in
catalytic assisted regenerations from the artificial modulation
DPFs decreases with time as regeneration continues.
of exhaust gas temperature became gradually smaller with an
increase in the modulation frequencies.
MULTI-PULSED REGENERATION
Pulsed regeneration strategies are based on finding a balance To further study fuel efficiency for multiple injections, a
between minimizing fuel penalty and extending DPF different engine operating condition is studied, as is shown in
durability and reliability through keeping the substrate Table 5. The trend of fuel efficiency shown in Figure 16 is
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

the same as that shown in Figure 15. However, the Figure 17 depicts that the effective DPF wall temperature is
regeneration efficiencies for multiple injections in mode 2a significantly reduced for multiple injections under mode 4a.
are much higher than those of single injection, especially for The effective DPF temperature is time averaged for the DPF
regeneration temperatures at 550°C and 575°C. This is due to wall temperature during the regeneration. Even for two or
the differences of exhaust temperature and mass flow three injections, the effective DPF temperature is about
between mode 2a and mode 4a. At mode 2a, it takes a longer 100°C lower than that of the single injection case. This
time for DPF temperature to drop to engine exhaust advantage diminishes as the number of injection increases.
temperature because of lower exhaust mass flow; therefore, The result of more frequent injections is lower DPF
more soot is oxidized at comparatively high temperatures temperatures.
during the dwell period. Moreover, less fuel is required to
heat up the DPF to the target regeneration temperature at high The maximum DPF wall temperature of multiple injection
engine exhaust temperatures immediately after the dwell cases is similar to the single injection case at comparatively
period. low regeneration temperatures, which is shown in Figure 18.
At a lower temperature, regenerations take place more
Table 5. Engine operating conditions (moderate load) uniformly, while, at a higher temperature, regenerations take
place in an aggressive way, resulting in large temperature
gradients in the DPF. This was also found by Hannamura et
al. in the study of the combustion phenomenon during DPF
regenerations [25].

In Figure 18, the maximum wall temperature for the two

injection cases is a little higher than a single injection case.
This is because the exhaust flow rate is reduced when the first
injection is finished, resulting in an increase of the wall
temperature. There is not enough exhaust flow to take away
the heat released by extensive oxidation of soot. This is not
obvious for comparatively low temperature regeneration
cases since the fuel injection rates are low, which can be
neglected compared to engine exhaust flow rates.

Figure 16. Fuel efficiency for single and multi-pulsed

regenerations at 550°C, 575°C, and 600°C (mode 2a)

Figure 17. Effective DPF wall temperature for single

and multi-pulsed regenerations at 550°C, 575°C, and
600°C (mode 4a)
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Pulsed injection with large dwell ratio does not have benefits
on fuel efficiency due to low soot oxidation rate, and more
fuel is required to heat up the exhaust to desirable
regeneration temperatures. In Figure 19, the optimal dwell
ratio corresponding to peak fuel efficiency is about 5% to
10% at different regeneration temperatures. The DPF wall
temperature at the optimal dwell ratio is very close to the soot
ignition temperature of 500°C. This indicates that the soot
ignition temperature (depending on different DPF systems,
catalyzed or uncatalyzed) could be used to determine the
dwell time dynamically. Once the DPF wall temperature
drops below the soot ignition temperature, fuel injection is
initiated again in order to maintain a comparatively high soot
oxidation rate during the entire regeneration period.

Figure 18. Maximum DPF wall temperature for single Next, two pulse injections with optimal dwell times, which
and multi-pulsed regenerations at 550°C, 575°C, and are controlled by the soot ignition temperature, were selected
600°C (mode 4a) to be further investigated by changing the magnitude of the
pulsed injections or using different regeneration
temperatures.
The discussion above shows that the optimal number of
injections for pulsed regeneration is two to three. Two to
three injection strategies display a much lower effective DPF TWO-STAGE PULSED
wall temperature without the sacrifice of fuel efficiency REGENERATION
compared to the single injection case. However, the dwell The two-stage pulsed regeneration strategy is applied through
ratio was kept the same for the studies shown above. To two pulsed injections of fuel at two different injection rates.
further study the effect of dwell time on multi-pulsed The general idea for this strategy is to burn enough soot at
regeneration performance, only two injections are chosen for stage one (the first pulse) at an intermediate target
investigation. regeneration temperature. At stage two, the second pulse
targeting to a high regeneration temperature is used to
conduct more aggressive regenerations in order to maintain
high fuel efficiency. The peak temperature occurring in the
more aggressive regeneration at stage two (the second pulse)
will be limited since there is not much soot left inside the
DPF.

In addition to those critical variables, which are shown in

Table A1 of the Appendix, additional critical variables for
two-stage pulsed regeneration are summarized in Table 6.

Figure 20,21,22 illustrate how two-stage pulsed regeneration

works. The injection profiles for two-stage pulsed
regeneration and single injection regeneration at 575°C and
600°C are shown in 20. The value of critical variables is
shown in Table 7. The magnitude of the first pulse is
targeting to the regeneration temperature of 575°C. After a
Figure 19. Dwell ratio effect on fuel efficiency at short dwell period, fuel injection is initiated at a rate targeting
regeneration temperature of 550°C, 575°C, and 600°C to 600°C.
for two injections

The effect of dwell time between two injections at three

different target regeneration temperatures was investigated.
The number on each point shown in Figure 19 corresponds to
DPF wall temperature at the end of the dwell period. It can be
seen that the DPF wall temperature drops to the engine
exhaust temperature when the dwell ratio is beyond 30%.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Table 6. Additional critical variables for two-stage pulsed regeneration

Table 7. Critical variables set for two-stage pulsed regeneration) will be activated. However, the DPF wall
regeneration test 1 temperature is suppressed due to a limited soot reaction rate
resulting from the combustion of a large amount of soot
during the first stage of regeneration. The second stage of
regeneration is finished as fuel injection stops when soot in
DPF drops to a specific quantity (msoot,min · fthermal). The
DPF wall temperature then begins to cool down, while soot is
still oxidized due to the thermal inertia effect discussed
above.

Figure 20. Fuel injection profile for two-stage pulsed

Figure 21. DPF maximum wall temperature for two-
regeneration, single injection regeneration at 575°C and
stage pulsed regeneration, single injection regeneration
600°C
at 575°C and 600°C

Figure 21 shows five different periods during a two-stage

pulsed regeneration. In the heat-up period, the soot oxidation The reason for conducting mild regenerations rather than
rate in Figure 22 is low. Most of the fuel injected in this aggressive regenerations during the first stage is that a very
period was used to heat up DPF to initiate DPF regenerations. high temperature beyond what DPF material can tolerate will
At the first stage of regeneration (mild regeneration), soot is be generated if the initial reaction rate is too high. This is
combusted at an intermediate rate. When the soot in the DPF what Zhan et al. called “Type A uncontrolled regeneration”
drops to a specific quantity, fuel injection will be terminated. due to intensive exothermal reactions at the start of
However, the DPF wall temperature is still high enough to regeneration [9]. As a result, DPF may crack during the first
combust soot. That is the transient period of the regeneration, regeneration stage. Another reason a higher rate of fuel
which started from the end of the mild regeneration and injection is not allowed during the initial stage of
ended at the beginning of the aggressive regeneration. When regeneration is a high probability of a hydrocarbon slip due to
the DPF wall temperature drops below the soot ignition a comparatively low exhaust temperature.
temperature, the second stage of regeneration (aggressive
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

Figure 22. Soot loading and soot oxidation rate for two-
stage pulsed regeneration, single injection regeneration
at 575°C and 600°C

Figure 23. Fuel efficiency and DPF maximum wall

For single injection regeneration at 600°C, the maximum temperature at different split soot fractions (Test one at
DPF temperature has a concave shape around about 300s, as higher targeting regeneration temperatures)
shown in Figure 21. That is because the mass flow rate of
exhaust is suddenly increased, which helps to remove heat
from soot oxidation. The increased flow is due to a sudden Another parametric study of the split soot fraction for two-
increase in the soot oxidation products of H2O and CO2. stage pulsed regenerations at lower regeneration temperatures
is shown in Figure 24. In this test, the maximum wall
A parametric study of the split soot fraction is conducted. The temperature stays the same at the split soot fraction beyond
split soot fraction stands for how much of a percentage of around 0.63. That is because the regeneration taken at the
total soot is oxidized during the second stage regeneration. A second stage in this test is not as aggressive as test one due to
split soot fraction at zero means regenerations take place at a lower regeneration temperatures.
single temperature of 575°C, while a split soot fraction at
unity means regenerations take place at a single temperature
of 600°C. Two critical points, which are characterized by
maximum DPF wall temperature, are found to distinguish
regenerations into three phases.

In Figure 23, at a split soot fraction lower than 0.55, the

maximum DPF wall temperature is almost flat and mainly
controlled by the first stage of regeneration. Since about 45%
of the soot is combusted during the first stage of regeneration,
the maximum DPF wall temperature at the second stage
(aggressive regeneration) is suppressed. The fuel efficiency
slowly increases with split soot fractions as the regeneration
shifts from a mild regeneration to an aggressive regeneration.
At a split soot fraction between 0.55 and 0.9, the maximum
DPF wall temperature, as well as fuel efficiency, linearly Figure 24. Fuel efficiency and DPF maximum wall
increases with the split soot fractions. Regeneration temperature at different split soot fractions (Test two at
characteristics are determined by mild and aggressive lower targeting regeneration temperatures)
regenerations together. When the split soot fraction is higher
than 0.9, the maximum DPF wall temperature and fuel
efficiency show a highly non-linear relation with split soot SUMMARY
fractions. This characteristic is very similar to that of
Based on regeneration fuel economy and DPF durability,
regenerations at very high temperatures.
optimization studies on when to initiate and terminate active
regenerations were performed through multiple and single
regeneration cycle tests. Pulsed regeneration strategies were
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

proposed and investigated. Conclusions are summarized 5. Allansson, R., Goersmann, C., Lavenius, M., Phillips, P. et
below: al., “The Development and In-Field Performance of Highly
Durable Particulate Control Systems,” SAE Technical Paper
Less frequent regenerations (initiated at higher soot loading) 2004-01-0072, 2004, doi:10.4271/2004-01-0072.
result in lower fuel penalty but higher filter temperatures as
6. Masoudi, M., Konstandopoulos, A., Nikitidis, M.,
well, which can eventually lead to permanent DPF damage.
Skaperdas, E. et al., “Validation of a Model and Development
Furthermore, it is found that the optimal soot loading to
of a Simulator for Predicting the Pressure Drop of Diesel
initiate regenerations is close to the maximum soot loading
Particulate Filters,” SAE Technical Paper 2001-01-0911,
allowed in DPF but restricted by a temperature limit for
2001, doi:10.4271/2001-01-0911.
material integrity due to the durability requirement.
7. Flörchinger, P., Anderson, M., Hou, Z., Taubert, T. et al.,
The optimal soot loading to terminate regenerations is “Prediction and Validation of Pressure Drop for Catalyzed
dependent on regeneration temperatures. Even though Diesel Particulate Filters,” SAE Technical Paper
“Deep” regenerations lead to a lower engine back pressure, 2003-01-0843, 2003, doi:10.4271/2003-01-0843.
they are not recommended due to inefficiency of combusting 8. Charbonnel, S. and Opris, C., “Fundamental Diesel
small amounts of soot in DPFs. Considering the thermal Particulate Filter (DPF) Pressure Drop Model,” SAE
inertia effect, it is advisable to terminate regenerations a little Technical Paper 2009-01-1271, 2009, doi:
early before reaching the optimum soot loading for 10.4271/2009-01-1271.
terminating regenerations.
9. Kamimoto, T., Murayama, Y., Minagawa, T., and
Pulsed regeneration strategies were investigated and showed Minami, T., “Light scattering technique for estimating soot
promise to have higher fuel efficiency without sacrificing mass loading in diesel particulate filters,” International
thermal durability of DPFs during active regenerations. The Journal of Engine Research, 10, pp. 323-336, 2009.
effect of the number of injections was studied, and two to 10. Ochs, T., Schittenhelm, H., Genssle, A., and Kamp, B.,
three turned out to be optimal. The dwell time between each “Particulate Matter Sensor for On Board Diagnostics (OBD)
injection was investigated as well, and the soot ignition of Diesel Particulate Filters (DPF),” SAE Int. J. Fuels Lubr.
temperature was found to be an effective control parameter to 3(1):61-69, 2010, doi:10.4271/2010-01-0307.
determine the dwell time dynamically. Two-stage pulsed
regenerations were explored at different split soot fractions. 11. Kodama, K., Hiranuma, S., Doumeki, R., Takeda, Y. et
By using the two-stage pulsed regeneration strategy, the al., “Development of DPF System for Commercial Vehicles
regeneration process could be well-controlled by adjusting (Second Report) - Active Regenerating Function in Various
the split soot fraction, which will be very useful for advanced Driving Condition,” SAE Technical Paper 2005-01-3694,
DPF system development and regeneration control. 2005, doi:10.4271/2005-01-3694.
12. Bach, E., Zikoridse, G., Sandig, R., Lemaire, J. et al.,
REFERENCES “Combination of Different Regeneration Methods for Diesel
Particulate Traps,” SAE Technical Paper 980541, 1998, doi:
1. Kapparos, D., Brahma, I., Strzelec, A., Rutland, C. et al.,
10.4271/980541.
“Integration of Diesel Engine, Exhaust System, Engine
Emissions and Aftertreatment Device Models,” SAE 13. Hiranuma, S., Takeda, Y., Kawatani, T., Doumeki, R. et
Technical Paper 2005-01-0947, 2005, doi: al., “Development of DPF System for Commercial Vehicle -
10.4271/2005-01-0947. Basic Characteristic and Active Regenerating Performance
-,” SAE Technical Paper 2003-01-3182, 2003, doi:
2. England, S., Rutland, C., Foster, D., and He, Y.,
10.4271/2003-01-3182.
“Investigation of the Effect of DPF Loading and Passive
Regeneration on Engine Performance and Emissions Using 14. Alano, E., Amon, B., and Jean, E., “Fuel Vaporizer:
an Integrated System Simulation,” SAE Technical Paper Alternative Solution for DPF Regeneration,” SAE Technical
2006-01-0263, 2006, doi:10.4271/2006-01-0263. Paper 2010-01-0561, 2010, doi:10.4271/2010-01-0561.
3. Singh, N., Rutland, C., Foster, D., Narayanaswamy, K. et 15. Park, K., Song, S., and Chun, K., “Low Temperature
al., “Investigation into Different DPF Regeneration Strategies Active Regeneration of Soot Using Hydrogen in a Multi-
Based on Fuel Economy Using Integrated System Channel Catalyzed DPF,” SAE Technical Paper
Simulation,” SAE Technical Paper 2009-01-1275, 2009, doi: 2010-01-0562, 2010, doi:10.4271/2010-01-0562.
10.4271/2009-01-1275.
16. Huynh, C., Johnson, J., Yang, S., Bagley, S. et al., “A
4. Gong, J. and Rutland, C., “Study the DPF Regeneration at One-Dimensional Computational Model for Studying the
Transient Operating Conditions Using Integrated System- Filtration and Regeneration Characteristics of a Catalyzed
Level Model,” SAE Technical Paper 2010-01-0892, 2010, Wall-Flow Diesel Particulate Filter,” SAE Technical Paper
doi:10.4271/2010-01-0892. 2003-01-0841, 2003, doi:10.4271/2003-01-0841.
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

17. England, S. B., “Development of an Integrated Diesel Professor Christopher J. Rutland

Engine, Emissions, and Aftertreatment System Level Model,” rutland@engr.wisc.edu
Master Thesis, University of Wisconsin-Madison, 2007. Mechanical Engineering Department
University of Wisconsin-Madison
18. Hiranuma, S., Takeda, Y., Kawatani, T., Doumeki, R. et
1018B Engineering Research Building
al., “Development of DPF System for Commercial Vehicle -
1500 Engineering Drive, Madison, WI 53706
Basic Characteristic and Active Regenerating Performance
-,” SAE Technical Paper 2003-01-3182, 2003, doi:
10.4271/2003-01-3182. ACKNOWLEDGMENTS
19. Boger, T., Rose, D., Tilgner, I., and Heibel, A., The author would like to thank General Motors Research and
“Regeneration Strategies for an Enhanced Thermal Development for their continued funding and support of this
Management of Oxide Diesel Particulate Filters,” SAE Int. J. research through the GM-UW collaborative research
Fuels Lubr. 1(1): 162-172, 2009, doi:10.4271/2008-01-0328. laboratory program.
20. Mikulic, I., Zhan, R., and Eakle, S., “Dependence of Fuel
Consumption on Engine Backpressure Generated by a DPF,”
SAE Technical Paper 2010-01-0535, 2010, doi:
ABBREVIATIONS
10.4271/2010-01-0535. DOC
21. Prikhodko, V. and Parks, J., “Implications of Low Diesel Oxidation Catalyst Convertor
Particulate Matter Emissions on System Fuel Efficiency for
High Efficiency Clean Combustion,” SAE Technical Paper DPF
2009-01-2709, 2009, doi:10.4271/2009-01-2709. Diesel Particulate Filter
22. Michelin, J., Figueras, B., Bouly, C., and Maret, D.,
“Optimized Diesel Particulate Filter System for Diesel
m_dot_exh
Exhaust Aftertreatment,” SAE Technical Paper
exhaust mass flow rate
2000-01-0475, 2000, doi:10.4271/2000-01-0475.
23. Arasappa, R., Premchand, K., Chilumukuru, K., Johnson,
T_exh
J. et al., “The Filtration, Oxidation and Pressure Drop
exhaust temperature
Characteristics of a Catalyzed Particulate Filter during Active
Regeneration - A 1D Modeling Study,” SAE Technical Paper
2009-01-1274, 2009, doi:10.4271/2009-01-1274. msoot,max
24. Khan, M. R., and Shamim, T., “A numerical soot loading to initiate regenerations
investigation of the regeneration characteristics of diesel
particulate filters under simulated transient exhaust msoot,min
conditions,” Proceedings of the Institution of Mechanical soot loading to terminate regenerations
Engineers, Part D: Journal of Automobile Engineering,
222(12), pp. 2469-2485,2008.
Tinj, min
25. Hanamura, K., Suzuki, T., Tanaka, T., and Miyairi, Y.,
“Visualization of Combustion Phenomena in Regeneration of minimum exhaust temperature for injection fuel
Diesel Particulate Filter,” SAE Technical Paper
2003-01-0836, 2003, doi:10.4271/2003-01-0836.
minimum oxygen molar fraction in exhaust
CONTACT INFORMATION
Jian Gong Ttar,regen
gong3@wisc.edu target DPF regeneration temperature
Mechanical Engineering Department
University of Wisconsin-Madison
1008 Engineering Research Building Tregen,1
1500 Engineering Drive, Madison, WI 53706 target DPF regeneration temperature at stage one

Tregen,2
target DPF regeneration temperature at stage two
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

fsplit
split soot fraction

fthermal
thermal soot fraction
 Downloaded from SAE International by University of Minnesota, Monday, July 30, 2018

APPENDIX
CRITICAL VARIABLES FOR
REGENERATION CONTROL

Table A1. Critical variables for regeneration control

The Engineering Meetings Board has approved this paper for publication. It has Positions and opinions advanced in this paper are those of the author(s) and not
successfully completed SAE's peer review process under the supervision of the session necessarily those of SAE. The author is solely responsible for the content of the paper.
organizer. This process requires a minimum of three (3) reviews by industry experts. SAE Customer Service:
Tel: 877-606-7323 (inside USA and Canada)
All rights reserved. No part of this publication may be reproduced, stored in a
Tel: 724-776-4970 (outside USA)
retrieval system, or transmitted, in any form or by any means, electronic, mechanical, Fax: 724-776-0790
photocopying, recording, or otherwise, without the prior written permission of SAE. Email: CustomerService@sae.org
ISSN 0148-7191 SAE Web Address: http://www.sae.org
Printed in USA