Analysis of energy flows in engine coolant,
structure and lubricant during warm-up
R D BURKE1, C J BRACE1, A LEWIS1, A COX1 and I PEGG2
1. Department of Mechanical Engineering, University of Bath, UK
2. Ford Motor Company, UK
ABSTRACT
Improving engine warm up time is of great interest because cold start engines suffer
from increased fuel consumption. This study comprises of two phases: the first looking
at global benefits from a novel engine cooling system incorporating coolant flow
control valves and dual EGR setup. The new cooling system showed promising signs of
influencing energy flows during warm up, however only limited analysis was possible.
The second phase used an instrumented engine to show new ways of quantifying
performance enhancements through combustion chamber heat transfer, local bearing oil
film temperatures and bearing heat transfer. 1 dimensional heat transfer analysis
demonstrated the significant energy required to warm up the engine structure and
25~50kJ heat loss per main crank shaft bearing.
1
INTRODUCTION
Under cold starting conditions, internal combustion (IC) engines suffer from higher
friction and less efficient combustion resulting in higher fuel consumption and higher
harmful emissions. To improve fuel economy and meet more stringent emissions
standards there is a need to warm the engine as fast as possible. In addition, comfort and
safety aspects require that sufficient heat is available for cabin heating and demist
capabilities. However during warm-up available energy is scarce and demands are high,
meaning good energy management is essential for optimised operation.
2
BACKGROUND
Reducing warm-up times for IC engines has been a main focus for many years with a
number of concepts producing varying degrees of success. Active control of the coolant
circuit by replacing the traditional wax element thermostat with flow control valves
have been used in an attempt to match engine cooling to thermal loading. Electric
coolant pumps have also been considered, but their increased cost is probably the reason
that they have not been widely adopted in production vehicles. On the one hand, lower
degrees of throttling have had little effect on engine warm up rate but are useful for
controlling steady state coolant temperatures to avoid over cooling the engine when
fully warm [1, 2]. On the other hand, higher degrees of throttling representing over 80%
reduction compared to the production vehicle, have achieved faster warm up resulting in
improvement of fuel consumption [3, 4]. It was suspected in both cases that the
reduction in warm up time was the result of the valves isolating parts of the coolant
volume and effectively temporarily reducing the thermal inertia of the cooling system.
Complementing reduced thermal inertia, external heat addition is another effective way
to reduce engine warm up times. Although very much at concept stage, incorporating a
coolant to exhaust gas heat exchanger has yielded positive results, although this does
represent significant production and development costs [5, 6]. When considering these
designs it is important not to impact on exhaust gas heat that contributes to catalyst
light-off and ensures emissions regulations are met. Also the system must avoid
significant increases to coolant volume and offer a solution for fully warm up operation
without causing excessive coolant temperatures. Other concept designs include storing
energy from a previous hot running cycle such as hot coolant flask or two phase heat
batteries [7].
Since energy is the key issue during warm up, this study will focus on quantifying the
energy exchanges in a modern Diesel engine employing coolant control valves which
are considered a pragmatic approach to control engine thermal management. The study
is split into two sections: initially an engine with novel cooling circuit design is
considered to assess global changes in energy flows over the New European Drive
Cycle (NEDC). Detailed analysis of the behaviour and interaction of the novel system
with the engine control strategy has previously been published [4] and showed that
coolant flows could be well controlled to improve warm up rate, but that this did not
always result in improved fuel consumption due to engine fuelling strategies. The work
presented here will build upon previous findings and focus on energy flows. In a second
phase, an instrumented engine will be used to show the potential analysis when more
data is available.
3
METHODOLOGY
In both sections of the study all experiments were conducted on a 2.4L, 4 cylinder
common rail fuel injection, production diesel engine that meets EURO IV emissions
specifications. The engine was installed on a transient engine dynamometer in a cell
fitted with climatic control and good test to test repeatability, with coefficient of
variation for fuel consumption measurements from previous testing campaigns
estimated to be 0.5 %. All tests were a cold start (25oC) NEDC.
3.1
Novel coolant design
The modified thermal management circuit incorporates changes to the cooling, lubricant
and air circuits in an attempt to combine heat recovery with coolant flow control. The
principle modifications include the inclusion of a dual exhaust gas recirculation (EGR)
cooler setup which although both fitted in parallel were not calibrated for synchronised
running and only one of the two could be used at any one time. The coolant circuit also
incorporates three flow control valves in addition to the wax element thermostat.
Radiator
(d) Dual EGR
valve/cooler
PRT
Coolant
Pump
Oil
cooler
Degas
Bottle
Oil
Coolant
(a) Engine out
coolant throttle
(b) EGR loop
Coolant
throttle
(c) Oil cooler
control valve
Figure 1: Coolant circuit layout
A variable flow oil pump (VFOP) is also included to limit oil pressures over engine
operating range. These introduce five new degrees of control:
a)
b)
c)
d)
e)
Controlling engine-out coolant flow rate
Controlling coolant flow through EGR cooling loop
Allowing or inhibiting coolant/oil heat exchange
Cooling EGR gases using coolant or oil
Target main gallery oil pressure
As in the production vehicle, the pressure regulated wax element thermostat (PRT) was
used to control the ratio of flow flowing through or bypassing the radiator. A schematic
of the cooling system is shown in figure 1. Different calibrations were run over the
NEDC using a number of different combinations of the flow control valves. These have
been grouped in table 1 to examine the effects of the first four control parameters above.
The VFOP is not considered here because improvements in fuel consumption are
achieved by reduced parasitic loses rather than faster warm up and the device was
always beneficial, typically offering 2-3% fuel consumption benefit [4]. The standard
engine calibration was used for all tests which meant small changes in injection timing
may occur as a result of different warm up rates [8], however these were small in the
context of the drive cycle.
3.2
Instrumented engine
To complement the screening study, an identical engine was instrumented with over 100
thermocouples to estimate heat fluxes from the combustion chamber. Some
thermocouples were grouped in sets of three in a custom built heat flux sensor [9].
These sensors were positioned around each of the cylinders on both the intake and
exhaust side.
3.3
Theory
The majority of energy analysis was estimated using standard energy balance (see
equation 1). This is readily applied to the coolant circuit where ultrasonic flowmeters
 ), temperature differences (ΔT)
provided non intrusive measures of coolant flow ( m
Table 1: Details of calibrations tested for the advanced cooling system
Engine out
EGR cooler
Oil cooler coolant
EGR cooler
VFOP
coolant throttle
coolant flow
flow
Throttling Coolant flows with coolant cooled EGR gases
1a
Max flow
Max flow
Open
Coolant
Controlled*
#
1b
Mapped
Max flow
Open
Coolant
Controlled*
#
1c
Mapped
½ flow
Open
Coolant
Controlled*
Using oil or coolant to cool EGR gases
2a
Mapped#
Max flow
Open
Coolant
Controlled*
#
2b
Mapped
Max flow
Open
Oil
Controlled*
Opening or bypassing coolant oil heat exchanger
3a
Mapped#
1/6th flow
Bypassed
Oil
Controlled*
3b
Mapped#
1/6th flow
Open
Oil
Controlled*
Throttling Coolant flows with oil cooled EGR gases
4a
Max flow
Max flow
Open
Oil
Controlled*
4b
Mapped#
Max flow
Open
Oil
Controlled*
#
th
4c
Mapped
1/12 flow
Open
Oil
Controlled*
* Controlled oil flow is based on a target oil gallery pressure of 2bar
#
Mapped Engine out coolant flow maintains valve closed until a head metal temperature of 95oC.
Build
were easily measured and heat capacity (cp) well documented. The air path was
analysed in the same way, with cylinder air flow estimated including EGR rates. Care
must also be taken when fitting thermocouples to measure gas stream temperatures.
E  m  c p  T
(1)
For the instrumented engine only one build condition has been tested in which no
coolant throttles were installed. However, the results from this test point represent the
average of 9 different repeat tests to ensure good confidence in the results.
Heat fluxes from the combustion chamber were estimated using 1D heat transfer as
described by Lewis et al. [10]. Heat energy from the combustion chamber was then
obtained by integrating heat flux down the cylinder bore. Bearing heat transfer was
estimated in a similar manor but based on two single point temperature readings,
sufficiently distant to increase confidence in their respective readings.
4
RESULTS
4.1
Results from Non-instrumented engine
Figure 2 shows the cumulative energy flows from the fuel energy during a cold start
NEDC. The major contributors to energy consumption are the brake work, the waste
heat in the exhaust gases and heating up of the engine structure and fluids. It is
important to note the contribution of the last 400 seconds, the extra urban drive cycle
(EUDC), to overall energy flows. At this point the engine has almost reached operating
temperature and large energy flows develop from the combustion chamber, through the
coolant to the radiator. This stage also represents significantly higher engine loads,
increasing the brake work contribution. When conducting energy analysis these flows
can mask the small improvements that are achieved during the warm up period. As a
result, energy analysis will concentrate on effects over the first 780 seconds,
representing the 4 repeat urban drive cycles.
Energy (kJ)
100 Vehicle Speed (km/h)
50
0
4
x 10
4
3
Fuel Energy
2
Fuel Burn Efficiency
Oil
Radiator
Coolant
Friction +
Structure
1
0
0
200
400
600
Time (s)
800
1000
Exhaust
Brake Work
1200
Figure 2: Global energy flows during warm up over the NEDC cycle
Table 2 shows the global energy balances for each of the test runs after 780s of the
NEDC. The fuel energy is represented as a percentage change from the reference test
point and all other energies are quantified as percentage of fuel energy. Figure 3 details
the coolant energy exchanges over that same period. In each graph, the energy inputs to
the coolant (from engine and EGR cooler) are illustrated by the line graphs and the bar
charts show where this energy is transferred (radiator, oil cooler or coolant temperature
rise). Each of the cases will now be considered separately.
4.1.1 Throttling coolant with coolant cooled EGR:
Figure 3 (a) shows firstly that over the first part of the NEDC, approximately 1MJ of
heat is picked up by the coolant in the engine and 0.5MJ in the EGR cooler. As the
coolant is throttled, the proportion of energy picked up in the engine increases, however
the amount of heat picked up in the EGR cooler decreases. The increase in the engine is
Table 2: Global energy balance after 780s NEDC
Build
Fuel
1a
1b
1c
2a
2b
3a
3b
4a
4b
4c
Ref
-0.5%
-0.6%
Ref
-0.4%
Ref
+0.7%
Ref
-1.5%
-0.7%
Unburned
Fuel# (%)
2.5
2.5
2.3
2.5
2.6
2.6
2.5
2.6
2.4
2.2
Brake
(%)
10.9
10.8
11.1
10.8
10.8
10.8
10.8
10.5
11
10.9
Exhaust
(%)
30.3
30.1
30
30.1
30.5
30.9
30.8
30.1
30.6
30.9
Coolant
(%)
6.5
8.5
9.1
8.5
10.4
8.5
7.9
5.7
9.3
5.4
Other*
(%)
49.9
48.2
47.5
48.2
45.8
47.2
48.0
51.1
46.7
50.6
*Other energy dissipation includes heat warming up the engine structure, friction and auxiliary power
consumption.
#
Unburned hydrocarbons and carbon monoxide emissions
interesting as reduced flow rates and hence coolant velocities would be expected to
reduce heat transfer. Whilst this is probably the case, the coolant throttling has a
secondary effect of reducing the coolant inlet temperature. This is due to the lower mass
flow rate through the external circuit and the significantly reduced flow through the
Degas bottle [4].
It is interesting to note that the effect of coolant temperature is dominant over the effect
of coolant flow with respect to the heat transfer and also shows the importance of
considering all aspects when assessing modifications to the thermal management
strategy. The reduction in the EGR cooler is a double effect of reduced coolant
velocities and increased engine out coolant temperature. Conversely, the increased
coolant temperature is beneficial in the oil cooler and heat transfer to oil is increased.
Referring back to table 2, the increased proportion of energy transferring to coolant is
accompanied with a reduction in other energy flows. These include engine structure
warm up but also friction losses which would be expected to reduce with higher
temperatures. Finally, it is also interesting to note that engine out coolant throttling
suppresses all losses in the radiator over the warm up period, despite hotter coolant. It
appears that in this case the throttled flow reduces any leakage in the thermostat,
probably through lower coolant pressures in that part of the circuit.
4.1.2 Switching oil or coolant cooled EGR gases
Figure 3 (b) shows the effect of switching from coolant to oil cooled EGR gases: as a
result, 0.4MJ heat input in the cooler becomes a 0.2MJ heat loss as the coolant warms
up heat exchanger structure. The consequence is that heat transfer in the engine is
increased which, in lieu of results from section 4.1.1, is most probably the result of
colder coolant. The switch is detrimental to oil/coolant heat exchange because locally
the coolant temperatures in the oil cooler will be significantly lower, however overall
heat transfer to the oil should be greater as the heat from EGR gases is transferred
directly to it.
4.1.3 Enabling/Disabling oil/coolant heat exchange during warm up
Table 1 and figure 3(c) show that bypassing the oil/coolant heat exchanger results in
higher fuel usage and lower heat transfer to coolant in the engine. The lower transfer to
coolant could be the result of higher coolant temperatures, as heat transfer to the oil is
inhibited. The higher fuel consumption is thought to be the result of higher friction
levels.
4.1.4 Throttling coolant with oil cooled EGR
Figure 3(d) shows the results from coolant throttling when oil was used to cool EGR
gases. In this case, the throttling of coolant in the EGR loop was much greater than for
coolant cooled EGR as this was not critical to avoid thermal throttling and increased
NOx. Closing the engine out coolant throttle has a similar effect which is seen when
using coolant cooled EGR. However the severe throttling combining engine out and
EGR loop throttles significantly reduced the amount of heat transferred from coolant to
oil. This shows that with significantly reduced coolant flows the heat transfer in the
engine can be reduced despite the lower coolant inlet temperatures from external circuit
effects.
Throttling Coolant with coolant cooled EGR
Throttling Coolant with coolant cooled EGR
Coolant Heating
Oil cooler
Radiator
EGR+Engine
Engine
Coolant
Heating
2000
2000
Oil cooler
779
715
1500
Radiator
609
715
Energy (kJ)
Energy (kJ)
1500
1000
821
EGR+Engin
1000
e
500
1083
1018
864
Engine
500
57
1018
610
0
Non
Eng out
Eng out +
Controlled Coolant
Throttle with
EGR
Cool cooled EGR
Throttling
coolant
Throttle
Throttling Coolant with coolant cooled EGR
0
Coolant
Coolant Heating
Oil cooler
Radiator 2000
EGR+Engine
Engine
(a)
2000
Oil
(b)
1500
1000
Energy (kJ)
Energy (kJ)
1500
599
673
1000
312
310
500
1125
409
500
621
423
59
0
628
0
Non
Controlled
0
On
Eng out
Throttle
Off
(c)
Eng out +
EGR Cool
Throttle
(d)
Throttling Coolant with coolant cooled EGR
Coolant Energy Exchanges
Energy (kJ)
2000
1500
1000
500
0
715
779
609
864
Coolant to Ambient (Radiator)
Non Controlled
1018
1083
Eng out Throttle
Eng out + EGR Cool Throttle
Coolant internal energy
Coolant to Oil (Oil/coolant H/X)
Engine to Coolant (Coolant jacket)
EGR to Coolant (EGR Cooler)
Figure 3: Coolant energy analysis after 780s NEDC for (a) coolant throttling with
coolant cooled EGR, (b) coolant or oil cooled EGR, (c) coolant/oil heat exchange
during warm up and (d) coolant throttling with oil cooled EGR
4.2
Results from the instrumented engine
The estimated heat loss from all cylinders is compared to the heat input to coolant over
the engine in figure 4. For the majority of the drive cycle the measured heat leaving the
cylinders is significantly greater than the total energy transferred to coolant over the
engine. It should be noted that this does not include heat transfer in the cylinder head
which clearly also contributes to coolant heat input. This suggests that despite a large
heat transfer to the coolant around the combustion chambers, this heat is distributed
around colder parts of the structure in the proximity of the coolant jacket. It is only over
the final 100 seconds, when the thermostat is open and the structure is warm, that the
coolant energy input is similar to the loss from the cylinders.
Figure 5 shows the heat dissipation from the main crank shaft bearings. Over the drive
cycle the heat transfer rate reduces as the structure warms up and the thermal gradient
between metal and hot oil film reduces. However, there is still a significant heat transfer
rate at the end of the 20 minute cycle as the mass of the lower part of the engine
represents a significant thermal inertia [11]. Over the drive cycle, around 25-50kJ of
heat energy is lost at each bearing.
Coolant Heating
Oil cooler
Radiator
EGR+Engine
Engine
RPM
Heat transfer (kJ/s)
3000
1000
Direct energy transfer from cylinders to coolant
15
Energy transfer to coolant over engine
10
5
0
0
200
400
600
800
1000
1200
Time (s)
Crank cap 5
Crank cap 1-4
100
50
0
Crank cap 2
Crank cap 5
Crank cap 3
50
25
Crank cap 1
0
200
400
600
Time (s)
800
Crank cap 4
1000
Heat energy (kJ)
Heat transfer (J/s)
Figure 4: Heat flow from cylinders compared to heat pick up by coolant
1200
Figure 5: Heat transfer from main crank shaft bearings
Figure 6 shows the temperature measurement distributions for all metal, coolant and oil
thermocouples. Following a homogeneous starting temperature the structure
temperatures lead the coolant and oil temperatures. This would be expected because the
majority of the temperature readings are taken in the proximity of heat sources and other
block temperatures would be expected to be below that of oil and coolant. The oil in
particular would be expected to be closely linked to metal temperatures, which becomes
obvious under warmed up conditions where they have overtaken coolant temperatures.
5
DISCUSSION
Controlling coolant, lubricant and oil flows within the engine has clearly demonstrated
the ability to influence the energy flows within the system. Throttling coolant to reduce
flow rates in different legs of the circuit was shown to affect heat transfer in the engine
both directly through control of the convection process and indirectly by influencing
coolant inlet temperatures. The initial analysis has left a number of questions regarding
the heat transfer process that will be answered using the instrumented engine. Also,
modified energy flows are only of interest if they result in measurable benefits in engine
behaviour. Fuel consumption and engine emissions are obvious measures or
performance improvements, but the nature of these measurements is that small benefits
Start NEDC
Metal
Oil
Coolant
End 1st UDC (195s)
Frerquency
End 2nd UDC (390s)
End 3rd UDC (585s)
End 4th UDC (780s)
End NEDC (1200s)
30
50
70
Temperature (oC)
90
110
Figure 6: Metal, coolant and oil temperatures evolution through cold NEDC
may not be detected with sufficient confidence using only this measure. The need exists
for more detailed measures of events within the engine when operating with these
altered energy flows. The instrumented engine offers a number of additional measures
that attempt to answer this need.
Heat flux measurements and estimated heat transfer through the cylinder liners can be
compared to the enthalpy increase of coolant over the engine. This comparison gives an
insight into the amount of energy required to increase the engine structure temperature,
which has a strong influence on oil temperatures during warm up. Heat dissipation in
the main bearings gave an idea of the local friction levels, but also includes any heat
losses from oil in that vicinity. Local oil film temperatures at these bearings will give a
measure of the benefits of the oil cooled EGR system. By channelling energy directly to
the oil, it is hoped that a measured increase in temperature will be present not only at the
exit of the heat exchanger, but in the key friction locations within the engine.
The 1D analysis used in this work clearly will not accurately represent the complex 3D
heat flows that will occur in reality within the structure during warm up. However, they
will offer an interesting insight into the behaviour under the different thermal
management system calibrations that will now be run on the instrumented block.
6
CONCLUSIONS
A novel coolant circuit has been fitted to a modern production Diesel engine and
demonstrated the ability to influence energy flows during engine warm up. Notably the
1MJ of heat from the engine was reduced and 0.5MJ from cooling EGR gases was
redirected directly to engine lubricant. However, further information on energy
exchanges and local oil temperatures were desired as an indicator of performance
benefits. An instrumented engine has been developed and demonstrated additional ways
of measuring benefits through heat fluxes from the combustion chamber, local oil film
temperatures and heat loss from main bearings. These new measures should now be
used to quantify the benefits from the novel thermal management system.
7
ACKNOWLEDGMENTS
The authors would like to acknowledge the funding support for this project received
from the Technology Strategy Board and the Ford Motor Company and their permission
to publish this paper.
REFERENCE LIST
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
CJ Brace, JG Hawley, S Akehurst, M Piddock & I Pegg, Cooling system
improvements - Assessing the effects on emissions and fuel economy,
Proceedings of the Institution of Mechanical Engineers, Part D: Journal of
Automobile Engineering, vol. 222, pp. 579-591, 2008.
RD Burke, CJ Brace, JG Hawley & I Pegg, Review of the systems analysis of
the interactions of thermal, lubricant and combustion processes of Diesel
engines, Proceedings of the Institution of Mechanical Engineers Part D-Journal
of Automobile Engineering, vol. 224, pp. 681-704, 2010.
W. Krause and K. H. Spies, Dynamic Control of the coolant temperature for
reduction of fuel consumption and hydrocarbon emission, SAE paper number
960271, 1996.
RD Burke, CJ Brace, JG Hawley & I Pegg, "Systems approach to the
improvement of engine warm-up behaviour," Proceedings of the Institution of
Mechanical Engineers Part D-Journal of Automobile Engineering, vol. 225(2),
pp.190-205, 2011.
D Kay, M Davies, J Caine, R McCabe, J Theis, "Application of a gas-to-coolant
exhaust gas heat exchanger: Fuel economy and emissions benefits", VTMS 8,
Nottingham, United Kingdom, 2007, pp. 379-388.
GE Andrews, AM Ounzain, H Li, M Bell, J Tate & K Ropkins, The use of a
water/Lube oil heat exchanger and enhanced cooling water heating to increase
water and lube oil heating rates in passenger cars for reduced fuel consumption
and CO2 emissions during cold start, SAE Paper Number 2007-01-2067, 2007.
P Revereault, C Rouaud & A Marchl, Fuel Economy and Cabin Heating
Improvements Thanks to Thermal Management Solutions Installed in a Diesel
Hybrid Electric Vehicle, SAE paper number 2010-01-0800, 2010.
RD Burke and CJ Brace, The effects of Engine thermal conditions on
performance, emissions and fuel consumption, SAE paper number 2010-010802, 2010.
A. Lewis, C.J. Brace, S. Akehurst and I. Pegg, "Spatially resolved heat flux
measurements from a HSDI engine over NEDC," presented at the VTMS 11,
Gaydon, UK, 2011.
A. Lewis, S Akehurst, CJ Brace, CA Finol & K Robinson, Dynamic
Measurement of Heat Flux Through the Cylinder Wall of a Modern HSDI
Engine Over a New European Drive Cycle, SAE Paper Number 2010-01-0322,
2010.
J.-P. Zammit, PJ Shayler & I Pegg, Thermal coupling and energy flows between
coolant, engine structure and lubricating oil during engine warm up, presented at
the VTMS 11, Gaydon, UK, 2011.