Proceedings of ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition

GT2017
June 26-30, 2017, Charlotte, North Carolina, USA

GT2017-64960

IMPLEMENTING FULL ELECTRIC TURBOCHARGING SYSTEMS ON HIGHLY

BOOSTED GASOLINE ENGINES

Q Zhang P Lu P Dimitriou S Akehurst

University of Bath University of Bath University of Bath University of Bath
Bath, UK Bath, UK Bath, UK Bath, UK

C Copeland M Zangeneh B Richards G Fowler

University of Bath Advanced Design Technology Ltd Aeristech Ltd Jaguar Land Rover Ltd
Bath, UK London, UK Kenilworth, UK Whitley, UK

and the eTurbocharger plus a mechanical turbocharger

ABSTRACT configuration were evaluated and compared. The simulation
To secure the highly challenging 2°C climate change limit, results have revealed that the two stage eTurbocharging system
the automotive sector is expected to further improve the has the potential to reduce CO2 emission in the proximity of 1
efficiency of the internal combustion engines. Over the past percent in different drive cycles compared to conventional
decade, internal combustion engine downsizing through wastegate turbocharger and the benefit would be much higher for
turbocharging has become one of the major solutions that the future real world driving cycle. The single stage configuration
industry has offered to fulfil its carbon commitment. Although was shown to be impractical in that the power level of the turbine
the various new turbocharging technologies has changed the generator will not only limit the engine power output, but also
sluggish image of conventional turbocharged engines, the have negative impact on the system design, cooling and cost
turbocharger system is far from perfect. From the perspective of implied. Meanwhile, the two stage configuration where the
engine energy flow, the copious amount of waste energy is eCompressor acts as a supplementary boost device at low end
habitually harvested by the turbine with low efficiency, and transient device came out as a better solution with overall
subsequently the turbine power transmitted to the compressor is advantage in manageable power level, system efficiency,
used solely to charge the engine. When this power for charging transient response and implied cost.
is excessive for the set engine operating condition, it either is
consumed by throttling or is directly discharged through the INTRODUCTION
wastegate, both as a pure enthalpy loss. To more efficiently The electrification of the boosting system has been seen as
harness the waste energy without deteriorating other engine a crucial enabler technology to achieve optimal drivability and
performance parameters, a full electric turbocharging technology CO2 emission on highly boosted gasoline engines (1). There are
is proposed by Aeristech Ltd. The system is composed of an three main forms of boosting systems electrification: mechanical
electric turbo generator and an electric compressor connected turbocharger assisted by a small eCompressor, eTurbocharger
only through electrical system. Without the constraint of a with a shaft-mounted motor/generator and the mechanically
mechanical turboshaft, the compressor and the turbine can be decoupled eTurbocharger.
operated at different speeds. The electrically driven compressor Out of the three options, the conventional turbocharger
can be free floating when boost is not required and the motor can assisted by a small eCompressor has been the most popular
provide the boost promptly only when higher load is requested. arrangement and has recently seen commercial application by
Meanwhile, the electric turbine can be controlled by the Audi on Diesel engines with potential suppliers including all
generator to operate at any set speed, allowing maximum major turbocharger OEMs (2). A study presented by FEV (3) has
efficiency for energy harvesting. This paper presents a shown the benefit of such a system to both the emission control
simulation study of the capability of the decoupled and performance enhancement with the aid of a 48V hybrid
eTurbocharging system to charge a highly boosted 2 litre powertrain: increased EGR level enabled by eSupercharging can
gasoline engine. The single stage eTurbocharger configuration reduce the engine out NOx emission in the range of 50% during

1 Copyright © 2017 by ASME

transient; adaptation of the size of base turbocharger has the with dedicated physical model of the motor/generator and
potential to reduce fuel consumption up to 2% and downsizing battery. An AFR based control strategy was proposed in the
and downspeeding of an electrically boosted engine can lead to study and potential of positive energy balance from the
the further reduction of 4-5%; furthermore, the expected low end boosting/turbo-compounding system and engine transient
torque improvement and transient response supported by an performance was demonstrated. Nonetheless, the controller
eCompressor allows the mechanical turbocharger to charge the oscillation has revealed the potential control difficulty of the
engine to higher power density. system and the scope of the study did not include the practicality
The transient response improvement through the use of aspects such as engine full load condition, electrical system
eCompressor was discussed in more depth in a study presented efficiency and motor/generator power rating.
by University of Ljubljana (4). An eCompressor was shown to To summarise the existing studies, the eCompressor
be superior to the integrated starter generator (ISG) in heavy duty technology is a compact and cost-effective solution to support
Diesel vehicle application in terms of ratio of engine dynamics the turbocharged engines to improve the low end torque, rated
improvement to electric energy consumption. power and transient performance. Fuel economy is improved
Such comparison between eCompressor and ISG was also mainly through the eCompressor aided gas exchange process
confirmed by Continental (4), where a 2 litre TGDI engine was optimisation, turbo upsizing and engine downsizing. The
uprated from 220 to 250kW. While confirming the much technology will undoubtedly see popularity with the advent of
improved transient performance through electric boosting, it was universal hybridisation. On the other hand, the shaft-mounted
stated that a 1.7kW electric energy invested on eCompressor is motor/generator eTurbocharger technology adds in the
equivalent to 20kW of ISG power input within a P2 mild hybrid advantage of energy harvesting without increased control
architecture. complexity. However, it does not solve the problem of lack of
On the topic of extreme downsizing, Mahle and Aeristech low end torque in downsized engines due to the compressor
(1) have managed to charge a 1.2litre engine to 33 bar BMEP, surge. Finally, due to the complete independence of eCompressor
increasing the power rating of a demonstrator engine from and eTurbine speed, the decoupled eTurbocharger combines the
120kW to 193kW. The potential of turbocharging was greatly performance augmentation of eCompressor and the energy
enhanced thanks to the eCompressor which took over the engine harvesting capability of electric turbo-compounders. However,
boosting at low end. the potentials and problems of such a technology are yet to be
The eTurbocharger with a shaft-mounted motor/generator fully studied.
has long been a novel turbocharging option. Back in 2000, This study aims to find a viable solution of decoupled
Imperial College (5) has published theoretical study of a ‘hybrid eTurbocharger for a highly boosted 2litre gasoline engine
turbocharger’ for both steady state and transient performance through simulation. The engine equipped with the novel boosting
evaluation. Without considering emission due to the system should deliver a full load curve superior to the baseline
computational model limitation, the turbocharger can be resized setup. The potential of fuel consumption will be derived and the
and leading to a fuel economy improvement of 5-10%. Transient control strategy designed for the boosting system will be
response was improved with various level depending on the validated through improved fuel economy.
power rating of the motor generator.
The transient simulation methodology was seen much MODELLING AND SIMULATION
improved a few years later in a study by University of Ljubljana The engine model was implemented in the 1D wave
(6), where a 0D simulation code calibrated by experimental data dynamics modelling environment in GT Power. The baseline
was used. Transient response was observed relating to both the model has been calibrated for both full/part load steady state
power rating and inertia of the motor generator used; a typical conditions as well as the transient operation. The combustion
torque tip-in has seen a more than 50% improvement through the model was a test data calibrated Wiebe function matrix and could
use of electrically assisted turbocharger. potentially create predictive errors when extrapolating above the
A more recent study (7) on the shaft-mounted baseline limiting torque curve or simulating transient operations,
motor/generator turbocharger has further demonstrated the therefore care should be taken to use the exact numbers from
flexibility of such a system to be used either to assist engine such non-predictive combustion models.
boosting or to harness energy to support on-board use. Fuel The baseline engine was a 2.0litre production gasoline
consumption benefit of 4.6% improvement in NEDC was engine, charged by a conventional wastegated turbocharger,
reached on a highly downsized engine. achieving a knee point torque of 340Nm at 1750rpm, max torque
The mechanically decoupled eTurbocharger is gaining 350Nm at 3000rpm and max power 175kW at 5500rpm. The
attention only in recent years. With the maturity of electric baseline model favoured transient performance and therefore has
boosting and turbo-compounding, such arrangements face only employed a relatively small turbocharger, raising low end torque
the barriers of adequate control design and commercially viable at the expense of increased pumping work. With a novel boosting
cost-benefit ratio. A study by Clemson University (8) has system, the full load torque was slightly optimised by advancing
presented a theoretical study of a boosting system comprised of the knee point down to 1250rpm. Such optimisation will be the
an eCompressor and an electric turbo-compounder. The study result of extra boosting from the eCompressor.
was supported by a comprehensive model built in AMESim,

2 Copyright © 2017 by ASME

 Three eTurbocharger arrangements have been investigated The compressor design employed the inverse design
as shown in figure 1 below. An eTurbocharger as a single stage method. The engine operating line at full load was first crudely
boosting system has the benefit of easiest packaging and lowest generated using a scaled compressor map from the GT-Power
component number. In the two stage boosting system the library. Then the commercially developed mean-line design code
eTurbine can be placed either upstream or downstream of the TDpre was used to generate the meridional housing for the
mechanical turbine. An eCompressor can also be placed impeller wheel. Subsequently the design of impeller blade 3D
upstream or downstream of the mechanical compressor, however shape was conducted in 3D inviscid design code TD1. ANSYS
simulation has shown that the performance of neither engine nor CFX was used to predict the performance map of pressure ratio
the boosting system was sensitive to the eCompressor location. and efficiency which iteratively supported the TD1 design
practice until arriving at a satisfactory design when the full load
operating line lay within a compressor map. Such method has the
risk of predicting optimistic surge margin and stage efficiency
and will be further evaluated experimentally in the following
stage of the project.
The design of turbine wheel was a different process
compared with conventional application since the turbine speed
is independent of the compressor speed. Again, a mean-line
method is firstly used to find the combination of tip width and
tip radius that can fulfil the requirement of both the highest and
lowest mass flow conditions. The 1D results from TDpre was
then sent into TD1 for 3D blade and nozzle shape design, with
ANSYS CFX providing CFD calculation to support the design
iteration. Without the constraints of compressor inertia and the
shared turbomachinery speed, both turbine and compressor
design seemed slightly different from the usual shapes of
turbochargers for a similarly sized engine. A pair of compressor
and turbine for a single stage arrangement is shown below in
figure 2 and 3 (comparison not to scale).

Fig 1. Left: Single stage arrangement. Right: Two stage

arrangement (HP and LP mode).
The simulation will aim to first achieve the full load torque
Fig 2. Design of compressor wheel.
target regardless of the eTurbocharger arrangements. Then the
practical aspects such as the eTurbine power rating will be
discussed using the simulation results obtained. Following the
full load simulation, part load simulation will cover the full area
of the speed/torque map to evaluate the fuel economy and control
strategy. Drive cycle fuel economy improvement will be
calculated using minimap points with weighting for different
drive cycles. Transient simulation at the end will further confirm
the performance improvement.

TURBOMACHINERY Fig 3. Design of turbine wheel.

The various compressors and turbines used in the
simulations have been designed by Advanced Design CONTROL STRATEGY
Technology (ADT) Ltd specifically for the requirement of this The gas exchange system control strategy implemented in
study. this study has been focused on optimal fuel economy. Whenever

3 Copyright © 2017 by ASME

multiple solutions existed for the same target torque, a DoE The eTurbine bypass valve was the most complicated one.
simulation was always conducted to find the lowest BSFC. In a test environment the controller would require heavy
For consideration of transient performance, the calibration effort. The controller is described as follows:
conventional mechanical turbocharger routinely ‘overboosts’ - At low load, map based binary control: fully shut to
and throttles the engine at the same time in part load conditions, convert throttling loss to electric energy and fully open
so that a torque transient request can be met with instant mass when not economical to do so at higher engine speed
flow and a fast spinning turbocharger once the throttle opens up. - At low to mid load when throttle already fully opened,
An eCompressor gasoline engine, on the contrary, can afford to modulates engine back pressure through the
charge the engine with the exact amount of boost pressure involvement of eTurbine to target BMEP
needed for the torque target, knowing that any transient request - At mid to high load, the controller bypassed the
can be fulfilled by a responsive eCompressor. Engine de- eTurbine when wastegate starts to modulate BMEP.
throttling through a boost control regime as such is a huge - When eCompressor was active, the controller engaged
advantage for part load fuel economy. the eTurbine to generate just enough energy to supply
Following the control philosophy of engine de-throttling, the eCompressor.
when throttling is necessary for the low load condition, it was - Bypass valve controller also modulate the involvement
better to throttle the engine using the eTurbine than the throttle, of the eTurbine so that the power generated did not
in that part of the throttling loss can be converted to electric exceed the design limit of the motor generator (8kW in
energy. Depending on the eTurbocharger power system this study).
efficiency, such benefit disappeared at high mass flow, due to the
fact that the pumping loss exceeded the energy recuperation.
The details of each controller can be explained as below.
Throttle
The throttle controller was a PID controller with
feedforward and gain schedule targeting the BMEP at low to mid
load. In a test environment the BMEP target can be substituted
by a calibrated mass flow rate instead.
eCompressor
The eCompressor controller was a PID controller with
feedforward targeting the BMEP. The controller was only
activated at high load when extra boost was needed to achieve
the torque target or transient condition.
eCompressor bypass valve
The bypass valve opened by default and shuts completely
when eCompressor was active.
Turbocharger wastegate (two stage)
In the case of the two stage arrangements, a large wastegate
was necessary to bypass the mechanical turbocharger completely
when boost was not needed at low to mid load. At mid to high Fig 4. Speed control of eTurbine.
load, the wastegate controller was a PID controller with
feedforward targeting the BMEP.
SINGLE STAGE ARRANGEMENT
eTurbine
The single stage configuration required the least packaging
The eTurbine speed was controlled by changing the load on
and control effort and therefore was the first candidate for
the motor/generator it was attached to. The target of the
investigation. The eCompressor and the eTurbine have the full
controller was to operate the turbine at its highest isentropic
liberty of location. As would be expected, the eCompressor
efficiency region. A polynomial model with pressure ratio as
provided only the necessary boost during transient and at mid to
input was designed as shown in figure 4. A PID controller
high load steady state above the engine naturally aspirated line.
modulates the generator load to match eTurbine speed to the
Steady state simulation results divided the engine speed/torque
polynomial model output. A typical polynomial equation as
map into three regions as shown in figure 5.
shown in figure can be written as:
Under the naturally aspirated line, the eCompressor is not
−1.585
activated but the eTurbine is engaged to harness energy
= −4964 × + 5288 whenever economical to do so.
(1)
Above the NA line, the eCompressor provided the pressure
eTurbine bypass valve needed to achieve each speed/torque target point; while the
eTurbine aims to harness enough energy to supply the
eCompressor with a certain energy loss, raising the engine back

4 Copyright © 2017 by ASME

pressure in the meantime. As a result, the engine – eTurbine – compressor from ADT compared to the baseline compressor due
eCompressor became a closed loop system: the eTurbine to the inertia of the compressor no longer being a limitation.
generating power to meet eCompressor power requirement The eTurbine power plot revealed two major problems of
increased engine back pressure, which led to higher boost level the single stage arrangement: 1, at low end, the exhaust energy
required to meet the torque target, which subsequently led to was not sufficiently available to the eTurbine, creating an energy
eTurbine further increasing the engine back pressure to generate deficit of more than 6kW. The deficit can only be compensated
more power. This led to the split of the region above NA region by the battery of a hybrid powertrain. The inefficiency of such
in to two parts: in low mass flow region where the exhaust energy practice defeats the concept of eTurbocharging for fuel economy.
was not sufficient, the eCompressor consumed more power than A transient low end torque target would be more practical. 2, at
the eTurbine can recuperate; in high mass flow region where the high end, although the eCompressor was consuming a similar
exhaust energy was ample, eTurbine power was adequate to amount of power compared to the mechanical compressor, the
supply the eCompressor to achieve the set torque target. eTurbine power reaches to a prohibitive level of 46kW when a
The overall power balance depends on the drive cycle and power system efficiency of 65% was taken into consideration.
the eTurbine control strategy. In NA region, the eTurbine is For the above mentioned reasons, the single stage
continuously charging the battery while the eCompressor idles. configuration eTurbocharging system was not further
As the passenger car engines spend most of their steady state investigated.
operations in the NA region, the energy stored in the battery is
expected to supply the transient performance and the imbalanced
region. In the balanced region, excessive energy can be extracted
at the cost of inferior fuel economy and therefore should only be
considered if battery charge is below certain level.

Fig 5. Division of engine speed and torque for steady

state simulation.
The performance of the single stage arrangement can be
demonstrated through analysing the full load simulation as
plotted in figure 6.
The torque plot showed the single stage arrangement has no
problem reaching the torque target of the baseline engine and has
managed to optimise the low end torque, advancing the knee
point down to 1250rpm as planned.
The boost pressure and back pressure plot demonstrated the
aforementioned vicious cycle of engine – eTurbine –
eCompressor: the engine boost pressure was increased while the
engine back pressure was increased even more.
The eCompressor power was at a similar level compared to
the baseline mechanical compressor, although achieving a higher Fig 6. Performance of single stage arrangement.
boost. The reason for this was a more optimal design of

5 Copyright © 2017 by ASME

 In order to demonstrate the effects that eCompressor and Fig 8. Mechanical compressor lug lines of baseline and
eTurbine had on engine performance more clearly, the simulation eCompressor equipped engine.
of the two stage arrangement was carried out in three steps.
Firstly, bypasses were added parallel to the mechanical TWO STAGE ARRANGEMENT - ETURBINE HP MODE
compressor and turbine to eliminate the secondary influence that VS LP MODE
might be caused by additional pipes. Then, eCompressor and In order to decide the relative position of mechanical and
eTurbine would be installed to the model step by step. The results electrical turbine, simulations were carried out with the eTurbine
were introduced as following. being located upstream (HP mode) and downstream (LP mode)
the mechanical turbine. From figure 9, mechanical turbine speed
TWO STAGE ARRANGEMENT - ECOMPRESSOR was lower in LP mode, which indicated that the output from the
ADDED mechanical turbocharger was reduced. Therefore, more power
For the two stage boosting engine system, the target BMEP was consumed by eCompressor to make up for the deficit in
at 1750 rpm and below was enhanced the same as in the single mechanical boosting, as shown in figure 11. When the eTurbine
stage arrangement. As shown in figure 7, the mechanical was located downstream of the mechanical turbine, it generated
turbocharger provided very little boost at low engine speed. At more power than the HP mode because it shared more expansion
1250 rpm, for example, the inlet air pressure was less than 1.28 ratio. Due to the smaller size of the eTurbine, it brought about
bar for the baseline engine. By utilizing eCompressor, the inlet higher back pressure in LP arrangement. Consequently, in order
air was boosted up to 1.98 bar. As a result, the full load brake to achieve the full load target, higher boost pressure was required
torque was increased from 225 Nm to 340 Nm at 1250 rpm. for the LP mode as well. It was the self-amplifying circle such as
When the engine was operated above 1750 rpm, eCompressor this that magnified the small advantage of HP mode into a large
was shut down and bypassed, imposing no effects on the power difference in performance. As shown in figure 10, the full load
output from the engine. BSFC is lower (by up to 2.7% at 1500 rpm) when eTurbine was
located upstream the mechanical turbine (HP mode) because of
the lower back pressure and less power consumption by the
eCompressor as aforementioned. However, when considering
the exhaust temperature at the inlet port of eTurbine, the LP
mode was more preferable to retain greater margin to the limit,
which was beneficial to prolong the life cycle of the electric
turbine.
In addition, the eCompressor efficiency was nearly the same
for both configuration, while the operating points of mechanical
compressor in the LP model shifted away from the surge line to
the higher efficiency region. Based on synthesized
Fig 7. Brake torque and boost pressure for baseline and considerations, it was decided to locate the eTurbine downstream
eCompressor equipped engine. the mechanical turbine (LP configuration) in this model for
In figure 8, the operating points of the mechanical further studies.
turbocharger compressor moved away from the surge line even
though achieving higher torque at low end since the
eCompressor shared part of the boost and helped to increase the
inlet air mass flow. Such effect would allow a larger compressor
for the engine rated power, while avoiding low end surge. Above
1750 rpm, the bypass valve was fully opened to disengage the
eCompressor from the air path. The lugline in this part coincides
with the baseline setup.

Fig 9. Back pressure and mechanical turbocharger

speed of HP and LP mode.

6 Copyright © 2017 by ASME

 calculated for a set of pressure ratios according to the polynomial
equation (1).
In the control module, expansion ratio of the eTurbine was
used to calculate the desired eTurbine speed. A PI controller then
modulates the load to the eTurbine to achieve the target speed.
In figure 14, power output from the eTurbine was multiplied
by an overall conversion efficiency of 67% to indicate the power
availability to the eCompressor. The results showed that there
was a minor shortage at 1250 rpm. Such deficit would need to be
provide by the battery which was a manageable power level. The
power requirement of the eCompressor at other engine speeds
Fig 10. Full load BSFC and engine boost pressure of HP can be satisfied without imposing parasitic load to the engine.
and LP mode.

Figure 14. eTurbine power vs eCompressor power.

Fig 11. E-comp and E-turbine power and E-comp
pressure ratio of HP and LP mode. CONTROL STRATEGY EVALUATED IN PARTLOAD
SIMULATION
The fundamental principle of the control strategy for partial
load simulation is to explore as much potential of the eTurbine
as possible for all the engine loads to reduce throttle losses and
the exhaust energy expelled through the wastegate as long as it
was economical to do so. The simulation results were shown in
figure 15 to figure 20 respectively for a variety of partial load
conditions.
From the results, it was a general trend that the intervention
of eTurbine increased the back pressure and the boost demand
for the engine system. At the same time, since the engine
Fig 12. E-comp efficiency for HP and LP mode. breathing could be regulated by adjusting the eTurbine bypass
valve as an alternative, throttle angle was less closed. It was
effectively to move the throttle loss to the eTurbine where the
losses could be harvested. At 20% load, for example, back
pressure increased by up to 1 bar at 5800 rpm, while the pressure
drop across the throttle body was reduced from 0.7 bar to 0.3 bar.
And also, at lower engine speed when the exhaust mass flow
decreased, electric turbine imposed less effects on air scavenge.
The effect of de-throttle became less significant. The demand for
boost pressure at 20% load was higher in the case of the
eTurbocharged engine, which suggested the mechanical turbine
was pushed harder by the additional turbine. The similar trend
Fig 13. Mechanical compressor lug lines of HP and LP was seen in the results of 40% and 60% load. But, it is worth
mode. noting that it was possible to eliminate throttle loss at 40% load
and higher, and merely rely on electric turbine to regulate the air
ETURBINE PERFORMACE FOR LP MODE flow. At 80% load, the pressure drops across the throttle body
In order to locate the operation points in the region of were especially low for both baseline and eTurbocharged engine.
highest thermal efficiency, the eTurbine speed was regulated. The difference in boost demand between baseline and
Specifically, a favourable line of reduced turbine speed was eTurbocharged engine models stayed around 0.2 bar, as the

7 Copyright © 2017 by ASME

magnitudes of the increments in back pressure were about the
same, which was around 1 bar.
The control strategy for the eCompressor was relatively
simple, since it was only needed in a small region at upper left
corner on the engine map when the mechanical turbocharger
could not provide enough boost. In the partial load points
simulated, the electric compressor was completely bypassed.

Fig 18. Boost pressure and eTurbine power for baseline

and eTurbocharger equipped engine at 40% load.

Fig 15. Back pressure and throttle loss of baseline and

eTurbocharger equipped engine at 20% load.

Fig 19. Back pressure and throttle loss of baseline and

eTurbocharger equipped engine at 60% load.

Fig 16. Boost pressure and eTurbine power for baseline

and eTurbocharger equipped engine at 20% load.

Fig 20. Boost pressure and eTurbine power for baseline

and eTurbocharger equipped engine at 60% load.

Fig 17. Back pressure and throttle loss of baseline and

eTurbocharger equipped engine at 40% load.

Fig 21. Back pressure and throttle loss of baseline and

eTurbocharger equipped engine at 80% load.

8 Copyright © 2017 by ASME

 Fig 22. Boost pressure and eTurbine power for baseline
and eTurbocharger equipped engine at 80% load.

Fig 23. Delta BSFC for the whole system.

ENGINE PARTLOAD BSFC MAP
Although the eTurbine was able to generate useful work DRIVING CYCLE FUEL CONSUMPTION
from the otherwise wasted exhaust energy and reduce the throttle Minimap point simulations with weightings for different
losses, it increased the back pressure and thus the boost demand drive cycles were conducted using the points as in figure 23. The
for the same BMEP target. It was therefore not always advised results in Table 1 demonstrated the improved fuel economy for
to use the eTurbine harnessing mode. The BSFC map in figure the operation under a variety of driving cycles when comparing
23 has shown an ‘eTurbo enclosure’ line where inside the with the baseline turbocharged engine. However, the magnitudes
dashed-line region eTurbine operation can achieve a positive in BSFC reduction were relatively lower than that of the full and
balance. partial load simulation. It was due to the fact that the operation
Meanwhile, since the eCompressor is expected to support points of driving cycles mainly resided in the region of extremely
improvement in transient response and increased surge margin, low load when the eTurbine could not work effectively to recover
the mechanical compressor was allowed to be a 20% larger waste heat, as illustrated in figure 22. A few points having
device. The avoidance of over-boosting as in a conventional negative impact on fuel economy took up more than 50% of the
turbocharged engine was also showing fuel economy benefit as weighting. As a result, the HiWay driving cycle, which had larger
in the region outside of the ‘eTurbo enclosure’ line. weighting in the mid-high load region, showed better potential
In the part load simulations, the excessive eTurbine power of the novel technology, while fewer benefits were seen in
was converted into electricity and then fed back into to the NEDC. It was reasonable to expect the eTurbocharger will
crankshaft with a conversion efficiency of 85%. Practice as such provide larger benefits in real world driving cycle with a much
reduced the IMEP of the engine for the same torque generation, wider speed and load region covered.
which moved the engine to its less efficient region. In spite of
this, it was found that the integration of eTurbocharging system Table 1. Fuel saving in driving cycles.
brought about a promising improvement in fuel economy. The Driving cycles BSFC reduction
BSFC was reduced by up to 3.6% in the middle of the engine
speed/torque map. Nevertheless, under extremely low load NEDC -0.4%
conditions, a negative effect was seen because of the poor FTP -0.5%
efficiency and little power output of eTurbine. At the highest HiWay -0.8%
engine speed and WOT operation, on the contrary, even though WLTP -0.7%
the eTurbine was more efficient in harvesting exhaust energy, it
created too much back pressure to be remedied by the power
CONCLUSIONS
provided by itself. Therefore, the fuel economy of the whole
This paper compared the performance of a conventional
system deteriorated by about 0.3%.
turbocharged engine model with a two-stage boost engine model
equipped with a decoupled electric turbocharger. A carefully
designed control strategy was designed for the electric
turbocharger for a variety of engine speed/load conditions. The
steady state performance of both systems was evaluated by
examining the fuel consumption at full load, part load and
minimap based driving cycles. The total boost pressure, back
pressure of both the baseline and novel models were compared
to seek out the origin of the difference in fuel economy. Power

9 Copyright © 2017 by ASME

consumption of the eCompressor and power generation of the ACKNOWLEDGE
eTurbine was analysed to indicate the effects that they have on We are grateful for the funding support from Innovate UK
the performance of the overall system. and the technical support from Advanced Design Technology,
1. As for the location of eTurbine, LP mode was selected for Aeristech and Jaguar Land Rover.
the lower eTurbine temperature for the thermal protection
although HP mode would be a more economical solution. REFERENCES
2. At full load condition, eCompressor was expected to 1. Bassett, M., et al, “E-supercharging for Heavily Downsized
provide extra boost at 1750 rpm and below to enhance the low Gasoline Engines,” in SIA Powertrain: The low CO2 spark
end torque by up to 115 Nm. It was beneficial for improving the ignition engine of the future and its hybridisation, Versailles,
driveability of the engine system. Above 1750 rpm, when the 2015
mechanical turbocharger was able to provide sufficient boost on 2. Nick Gibbs, “Audi’s move to e-charger could boost Valeo,
its own, eCompressor was totally bypassed. At the same time, Honeywell, BorgWarner,” in Automotive News.
electric turbine was harvesting exhaust energy independently. http://www.autonews.com/article/20150515/COPY01/3051599
The power output from the eTurbine was sufficient to satisfy the 86/audis-move-to-e-charger-could-boost-valeo-honeywell-
power consumption of the eCompressor. borgwarner, 2015. Last accessed: Nov. 2016
3. At partial load condition, mass flow could be regulated by 3. Stapelbroek, M., et al, “Fuel Consumption Reduction and
adjusting the eTurbine bypass valve as an alternative to throttle, Performance Improvement by Electric Driven Supercharger,” in
so that throttle was less shut. It helped to reduce the throttle loss. Aachen Colloquium Automobile and Engine Technology, pp1-20,
At 40% load and higher, the pressure drop across the throttle Aachen, 2016
body was nearly eliminated. Besides, eTurbine was able to 4. Katrašnik, T., et al, “An Analysis of Turbocharged Diesel
harvest up to 8 kW power from the exhaust energy, which could Engine Dynamic Response Improvement by Electric Assisting
be converted into useful mechanical work. However, similar Systems,” in Transactions of the ASME, Vol. 127, pp918-926.
with the case in full load condition, back pressure and 2005
requirement of boost was increased because of the employment 4. Ehrhard, J., et al, “Future Emission Legislation
of eTurbine. The tendency to knock due to higher back pressure Requirements – Contribution of Electrically-Assisted Charging
should be further investigated in the experimental phase of the Systems,” in Aachen Colloquium Automobile and Engine
study. Technology, pp1137-1155, Aachen, 2016
4. Across the part load map the BSFC was reduced by up to 5. Panting, J., et al, “Turbocharger motor–generator for
3.6%. At the high engine load, fuel consumption will increase improvement of transient performance in an internal combustion
because of the high back pressure as a result of eTurbine trying engine,” in Proceedings of Institution of Mechanical
to harness energy to support eCompressor. Engineers Vol 215 Part D, pp368-383, 2001
5. The minimap points of driving cycles were located mostly 6. Katrašnik, T., et al, “Improvement of the Dynamic
in the lower left region of engine load. A few load points with Characteristic of an Automotive Engine by a Turbocharger
small or even negative impact on fuel economy took large Assisted by an Electric Motor,” in Transactions of the ASME, Vol
weighting in the calculation. That made the improvement to the 125, pp590-595, 2003
fuel economy less remarkable. From the results, the largest 7. Arsie, I., et al, “A Comprehensive Powertrain Model to
reduction in BSFC was seen in HiWay driving cycle, which is Evaluate the Benefits of Electric Turbo Compound (ETC) in
0.8%. In NEDC, the number decreased to 0.4%. The upcoming Reducing CO2 Emissions from Small Diesel Passenger Cars,”
RDE cycle will be investigated in future works and the eTurbo SAE Technical Paper 2014-01-1650, 2014
technology is expected to achieving larger fuel economy benefit 8. Divekar, P., et al, “Coordinated Electric Supercharging and
due to the higher load/speed of the real world drive cycle. Turbo-Generation for a Diesel Engine,” SAE Technical Paper
2010-01-1228, 201
NOMENCLATURE

AFR Air fuel ratio

BMEP Brake mean effective pressure
DoE Design of experiment
EGR Exhaust gas recirculation
HP High pressure stage
IMEP Indicated mean effective pressure
ISG Integrated starter generator
LP Low pressure stage
NA Naturally aspirated
TGDI Turbocharged gasoline direct injection engine
WOT Wide open throttle

10 Copyright © 2017 by ASME