The Effect of Forced Cool Down on Cold Start Test Repeatability

ABSTRACT

Increasing the number of cold-start engine cycles which could be run in any one day would greatly improve the productivity of an engine test facility. However with the introduction of forced cooling procedures there is the inherent risk that test-to-test repeatability will be affected. Therefore an investigation into the effects caused by forced cooling on fuel consumption and the temperature distribution through the engine and fluids is essential. Testing was completed on a 2.4 litre diesel engine running a cold NEDC. The test facility utilises a basic ventilation system, which draws in external ambient air, which is forced past the engine and then drawn out of the cell. This can be supplemented with the use of a spot cooling fan. The forced cool down resulted in a much quicker cool down which was further reduced with spot cooling, in the region of 25% reduction. The implementation of forced cooling was found to have a small but repeatable effect on the fuel consumption, around 1%, and the fluid temperature distribution through the engine. Using this data it is possible to increase the number of cold start tests performed per day from one to three. NOTATION NEDC – New European Drive Cycle PRT – Pressure Regulated Thermostat ECU – Engine Control Unit CWF – Carbon Weight Factor THC – Total HydroCarbons CO – Carbon Monoxide CO2 – Carbon Dioxide

INTRODUCTION

The requirement for reduced fuel consumption and improved emissions is becoming increasingly demanding; this in turn is creating a demand for test procedures which are more reliable and repeatable. This need combined with the requirement for more test cycles to be run each day, to increase the productivity of a test facility. Currently, before a cold test cycle is performed an engine is often left overnight to achieve stable temperatures throughout; this means that only one cold test can be carried out each working day. The engine can be cooled rapidly to allow more tests to be performed each day; however there is an inherent risk of a loss in test-to-test repeatability. A small change in experimental conditions can have a dramatic effect on the outcome for each test. This is even more significant when it is taken into account that the improvements hoping to be achieved are in the region of 1 or 2%. Evidence shows that there can be in the region of a 10% saving in fuel consumption over a New European Drive Cycle (NEDC) when completed with a higher engine temperature. (1) This reiterates the need for a consistent temperature between cold start tests. Forced cooling involves forcing air and/or operating fluids around the engine to remove heat at a much faster rate than achieved by natural means. The work presented here was part of a thermal management study and as such is was desirable to leave the water and oil circuits unchanged. As a result the efficacy of forced air cooling was assessed. This can be carried out using dedicated fans or by utilising the test cell ventilation system. When an engine is to be forced cooled it is important to take into account the temperature distribution through the engine. One of the main concerns when it comes to force cooling an engine is that the temperatures do not return to their

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

Printed in USA

2009-01-1976

The Effect of Forced Cool Down on Cold Start Test Repeatability

A. Lewis, C. J. Brace and A. Cox University of Bath

Copyright © 2009 SAE International

Downloaded from SAE International by Automotive Research Association of India, Friday, August 01, 2014

ambient conditions leading to an effect on the fuel consumption values for the drive cycle. It is desirable that areas are not overcooled past their ambient conditions; this is known as cold spotting. This in turn can give inaccurate test results. The mechanism for the erroneous result may be due to a change in the heat physical environment (combustion, heat transfer, viscosity etc) or a change in the controller behaviour due to the measurement system sensing an unrepresentative temperature. Often a combination of both will be apparent. EXPERIMENTAL SET-UP The experimentation carried out in this paper was to investigate the effects a forced cool down procedure had on the fuel consumption measured for a cold NEDC. The NEDC simulates a 20 min drive with a 14 min urban and a 6 min interurban part, for a cold start the vehicle temperature should be completely uniform and in a range between 20°C and 30°C. The work was conducted on a 2.4 litre production specification diesel engine coupled to a transient dynamometer. The engine had an experimental cooling circuit which incorporated a PRT (Pressure Regulated Thermostat), throttled outlet to the water pump and a VFOP (Variable Flow Oil Pump). However the VFOP was switched off to allow full flow. The coolant circuit utilised a radiator and road speed fan positioned to the front of the engine rather than a heat exchanger; this was to best replicate road conditions.

Figure 1 - Cell Ventilation Schematic

The test cell utilises a very basic ventilation system shown in Figure 1. In this system the ventilation fan and road speed fan can be either automatically or manually controlled between 0 and 100%, whereas the recirculation flaps work on a closed loop control system based on the cell ambient temperature which was set at 23°C. To aid with the forced cooling process the option of a free standing fan to provide spot cooling was included if necessary. This fan would be positioned on the gearbox side of the engine directed towards the cylinder head and block; this is due to there being no air incident on this area from the road speed fan. There were to be two principal methods of fuel consumption measurement used in this experimental procedure, Gravimetric, carbon balance according to the

method specified in ECE. The gravimetric consumption is simply the difference in mass of fuel in a beaker between the start and finish of the drive cycle. The ECE fuel consumption is calculated from the emissions data collected using a Horiba MEXA 7000 series gas analyser. This is processed using a simple carbon balance equation:

( )( )( )⎪⎭

⎪⎬

⎫

⎪⎩

⎪⎨

⎧

×+×+

××⎟⎠⎞

⎜⎝⎛=

2273.0429.01

COCO

THCCWF

CWFFC

Equation 1 - ECE Formula where

FC = Fuel Consumption THC = Total HydroCarbons CO = Carbon Monoxide CO2 = Carbon Dioxide

This calculation is dependant on the correct value used for the carbon weight factor (CWF) as a change in this value by 0.01 can cause a difference in fuel consumption of over 1%. A CWF of 0.865 was used for the ECE fuel consumption calculations; this was obtained from the fuel test certificate. ECE fuel consumption data must also be synchronised so that the emission measurement corresponds to the correct ‘slug’ of exhaust gas. The total delay is due to the analyser response, sample line transport time and the exhaust transport. (2) In addition ECU fuel demand was recorded as a diagnostic aid. The ECU fuel demand is calculated using the ECU variables which specify the volumetric fuel demand and the engine speed. This value is more indicative and can not be considered as an accurate measurement device. Gravimetric measurement of fuel consumption is generally regarded as the most accurate of the three methods. EXPERIMENTAL APPROACH The aim was to achieve satisfactory consistency between the fuel consumption figures for a drive cycle after an overnight soak and for a drive cycle after the engine had been forced cooled to within 1°C of the overnight temperature. The cell ambient demand temperature was set at 23°C with the average temperature distribution through the engine being shown in Table 1.


Table 1 - Average Overnight Soak Temperature Distribution (based on 8 repeat tests)

Location

Average Starting

Temp (°C )

Oil – Sump 21.7 Coolant - Engine In 23.6 Coolant - Engine Out 22.8 Fuel - High Pressure Pump In 22.4

Engine Temperature 24.5 These average temperatures were calculated using 8 different past data sets for the current engine setup. A structured experimental approach to the testing was required as very small effects on fuel consumption had to be measured. The three following variables were to be considered:

Table 2 - Test Variables Variable Range Ventilation Fan 0 – 100% Road Speed Fan 0 – 100% Spot Cooling Fan

ON / OFF

The baseline test for this experimental procedure would be a natural cool down after a cold drive cycle.

Three experiments were performed to evaluate the effect that the forced cool down process had on the fuel consumption of the second drive cycle. The first test of each experiment was performed after an overnight soak of the engine; this allowed all the temperatures to reach equilibrium. The forced cool down process took in the region of 11000 seconds (3 hours), after which the second test was carried out. The target temperature for the forced cool down was an ECU engine temperature of 24°C.

Figure 2 - Experimental Flow Diagram

RESULTS AND DISCUSSION FUEL CONSUMPTION EFFECT Figure 3 shows the cumulative fuel consumption by all three measurement methods for a single drive cycle. Ideally these three methods should be the same.(3) The values of fuel consumption at the end of each test (i.e. 1180 seconds), pre and post forced cool down, for each measurement method are shown in Table 3.

Figure 3 - Cumulative Fuel Consumption

-450

-400

-350

-300

-250

-200

-150

-100

-50

0

50

100

150

0 100 200 300 400 500 600 700 800 900 1000 1100 1200Corrected UoB Test Time (sec)

0

100

200

300

400

500

600

700

800

900

1000

Cum

ulat

ive

Fuel

Con

sum

ptio

n (g

/test

)

VSIM Vehicle ECE Fuel ECU Fuel Dmd Grav. Fuel

1a Forced Cool down

1bOvernight Soak


Table 3- Fuel Consumption Results Pre and Post Forced Cool down

Exp

erim

ent

Num

ber

Spo

t Coo

ling

Inta

ke V

ent

Roa

d S

peed

Fan

(%

) V

entil

atio

n Fa

n (%

)

Fuel

M

easu

rem

ent

Type

‘a’ P

re F

orce

d C

ool D

own

(g)

‘b’ P

ost F

orce

d C

ool D

own

(g)

Diff

eren

ce (g

)

Diff

eren

ce (%

)

1 ON OPEN 100 100 Gravimetric 837.2 829.9 7.3 0.9 ECU 724.0 715.1 8.9 1.2 ECE 920.1 914.6 5.5 0.6

2 ON OPEN 100 100 Gravimetric 834.5 830.0 4.5 0.5 ECU 718.3 712.6 5.7 0.8 ECE - - - -

3 ON OPEN 100 100 Gravimetric 834.9 830.5 4.4 0.5 ECU 710.3 713.9 -3.6 -0.5 ECE 892.9 925.9 -33.0 -3.7

The normal overnight variability in gravimetric fuel consumption for test ‘a’ of each experiment is shown in Table 4. Table 4 - Difference in Gravimetric Fuel Consumption for

Test 'a'

Exp

erim

ent

Num

ber

Gra

vim

etric

Fue

l C

onsu

mpt

ion

(g)

Diff

eren

ce (%

)

1 837.2

2 834.5

0.32

3 834.9

0.05

This was calculated to illustrate that there is a small difference in the gravimetric fuel consumption from test-to-test when completed under similar conditions. This must be taken into account when comparing ‘pre’ and ‘post’ forced cool down results. There is a large range of 200g per test between the ECE and the ECU method of fuel consumption measurement although the ECU indication is used only for diagnostic purposes. Of more concern is the offset between ECE and gravimetric measurements, despite the normal precautions having been taken to ensure correct function of the equipment. A study to determine the causes of this discrepancy will be reported separately. In experiment 2 there was an error in the ECE fuel measurement method resulting in a missing data set.

The table shows that the forced cool down process has a very small affect on the fuel consumption, in the region of a 1% decrease. An error of 1% is too great when it is taken into account that an improvement of 0.5% fuel consumption is regarded as a useful improvement to the engine operation. The reduced fuel consumption may be due to the temperature distribution in the different fluids being different to that achieved through a natural cool down; an investigation into that phenomenon is performed later in this paper. There is a large error of 3.7% in the ECE calculated fuel consumption in experiment 3. This is far outside the acceptable range of error in fuel consumption, however this marked increase in fuel consumption is not conveyed in either of the other measurement methods. Also it would be expected that the fuel consumption would be reduced in the second test, due to residual heat rather than increased. Through further investigation into this anomalous result it was found that there was a fault in the values recorded by the MEXA analyser in the first 11 seconds of the test, in which time they reported the default value. Therefore the ECE data for this test should be discounted.

600

650

700

750

800

850

Baseline 1a 1b 2a 2b 3a 3b

Test

Fuel

Con

sum

ptio

n (g

ram

s)

ECU Fuel Demand Gravimetric Fuel Consumption Figure 4 - Fuel Consumption Comparison

Figure 4 shows the fuel consumption data for the 6 tests performed compared with the result from the baseline test. Each test ‘a’ was performed after an overnight cool


down, and test ‘b’ was performed after a forced cool down. The gravimetric fuel consumption trace shows that there is a reduction in the fuel consumed after a forced cool down. The average reduction in fuel was found to be 5.4 grams (0.65%). This was due to there still being some residual heat left in the engine metal or the fluids. The ECU fuel demand for the first test (1a) is 10.7 grams more than the baseline test, this increase is not conveyed in the gravimetric fuel consumption, which showed a decrease of 0.5 grams. This may be due to an error in the value calculated for ECU fuel demand. This is due to a small variation in fuel temperature which directly affects the calculated fuel density and therefore the amount of fuel injected. This proves how vital stable temperatures throughout the engine fluids are to successful and reliable test results. Other than this result for test ‘1a’ a clear trend is visible for both the ECU fuel demand and gravimetric, with only a deviation in fuel consumption of 1%. EFFECT OF SPOT COOLING An investigation into the effect of spot cooling on five key temperatures around the engine was performed. Figure 5 shows the effect of the spot cooling fan on the oil sump temperature with respect to time. Even though the fan is directed towards the cylinder head and block, it still has a small effect, in the region of 2°C increased cooling, reaching the desired temperature in 72% of the time. This may be due to the increased heat transfer from this region to the side which is being cooled faster by the fan.

0

10

20

30

40

50

60

70

80

90

100

0 600 1200 1800 2400 3000 3600 4200 4800 5400 6000 6600 7200 7800 8400 9000 9600 10200 10800

Time (secs)

Tem

pera

ture

(C)

SpotOFF - Oil - ShallowSump SpotON - Oil - ShallowSump Figure 5 - Spot cooling effect on Oil Sump

When these results are compared to Figure 6, the ECU engine temperature, it is noticeable how much of an effect the free standing fan has. The engine temperature has reached 24°C in less than 50% of the time taken without the spot cooling. This is a dramatic reduction in cooling time. The other temperatures looked at were the engine coolant in and out, and the fuel temperature, all of which showed an increased rate of cooling when the spot cooling fan was included in the cool down process. This led to the target temperature of 24°C being achieved in 60% of the time taken without forced cooling.

0

10

20

30

40

50

60

70

80

90

100

0 600 1200 1800 2400 3000 3600 4200 4800 5400 6000 6600 7200 7800 8400 9000 9600 10200 10800

Time (secs)

Tem

pera

ture

(C)

SpotOFF - IPROV_IN_ENG_TEMP SpotON - IPROV_IN_ENG_TEMP Figure 6 - Spot Cooling effect on ECU Engine

Temperature

TEMPERATURE DISTRIBUTION AFTER FORCED COOL DOWN It was important to evaluate the effect the forced cool down process had on the temperature distribution through the engine. If certain areas achieved their target temperature faster than others then this would affect the fuel consumption figures. The same key temperatures were used as in the spot cooling investigation. Figure 7 shows the temperatures at the start of each test whether it is after an overnight soak or a forced cool down process. This is also compared with the average temperature after an overnight soak shown in Table 1. The forced cool down used in experiment 2 was insufficient as the starting temperatures for each of the fluids was in the region of 5°C higher than after an overnight cool down. This occurred because the forced cool down procedure was stopped at an ECU engine temperature of 28°C, which is hotter than the demanded temperature. This was due to the fuel temperature coming to within 1°C of its target. When this same procedure was carried out in experiment 3 the effect of the forced cool down on the fuel temperature was closely monitored. The ECU engine temperature was forced down to the demand temperature of 24°C with no adverse effects on the fuel temperature. The baseline test achieved the lowest temperatures for all the monitored temperatures; in the region of 2°C less than the average. There is a common trend of the oil temperatures for each test being about 1°C lower than the other fluid temperatures in the same test except for experiment 1, where the oil temperature is greater than the coolant and fuel temperatures. This is undesirable as it will have an effect on the oil viscosity and other properties. (4) The forced cool down process in future should be controlled to bring the fluid temperatures to within 1°C of their original. Experiment 3 would have achieved the desired temperature distribution if its method of cooling had been maintained for a longer period of time.


18

20

22

24

26

28

30

Oil - ShallowSump Cool - Engine In Cool - Engine Out Fuel - HP Pump In IPROV_IN_ENG_TEMP

Tem

pera

ture

(C)

1a 1b 2a 2b 3a 3b Baseline Average Figure 7 - Temperature Distribution around Engine

It is interesting to note that using both ventilation and road speed fans at 100%, and the inclusion of the free standing fan, did not lead to an uneven temperature distribution through the fluids in the engine. Therefore further testing varying the duty of each fan required in the forced cool down process was not necessary. However it did prove that the once the cooling process had been stopped there was remaining residual heat which caused the fluid and metal temperatures to rise slightly, in the region of 0.5°C, therefore in future the period of time which the cooling process was performed over should be increased. DISCUSSION Overall there was a small but repeatable effect on the fuel consumption when forced cooling was implemented. The difference in fuel consumption between when the engine experienced an overnight soak and when a forced cool down was applied is in the region of 9.0 grams, which is a 0.9% error. This is an acceptable error when using a gravimetric method to measure fuel consumption. Close monitoring of the measuring equipment is vital as a small change can cause large errors in the measured overall fuel consumption figures as shown in experiment 3 where an increase in fuel consumption of 3.7% was evident. The spot cooling fan on the reverse side of the engine reduced the time needed for the forced cooling process by over 25%; this may be improved further by better control of the path which the air takes around the engine. Currently the majority of air from the road speed fan passes under the engine or out past either side as this is its natural path. The cooling time may be further decreased by motoring the engine with the dynamometer, this will move the fluids around there respective systems. This would also benefit the control of the temperature distribution through the engine. Further investigation into this is required. The use of a forced cooling process was also seen to not have adverse effects on the temperature distribution through the engine. However it was shown that elevated fluid temperatures of 2°C does produce a reduction in fuel consumption of 0.65%. This proves that the starting temperatures for every test must be within 0.5°C to ensure repeatability.

The air in the intake duct of the ventilation system ranged between 11°C and 23°C depending on the amount of air being re-circulated. This range would depend on the time of year; therefore the temperature of the air used to cool the engine would also vary, affecting the forced cooling time required. For repeatability it would be desirable to control the temperature in the cell to a set value throughout the year, therefore a closed loop system could be designed with a chiller unit to reduce the in-cell temperature to the desired value. CONCLUSION The work undertaken and reported in this paper was specifically undertaken to quantify the effects that the implementation of a forced cool down procedure has on the fuel consumption of the following test. The effects of the procedure on the temperature distribution through the engine were also investigated. The following conclusions can be made: 1) The implementation of a forced cool down

procedure has a small but repeatable effect on the fuel consumption, in the region of 0.9%.

2) The effect on fuel consumption is compounded by unrealistic starting temperatures through the engine, increasing the error to over 1.5%.

3) The inclusion of a spot cooling fan reduces the required cooling time by 25% and improves the temperature distribution through the engine by supplying cooling effects to the reverse side.

A possible step for the future that could be taken in this research would be to motor the engine during the cool down process and to maintain a constant air inlet temperature. This would much improve the predictability of the cool down time.

ACKNOWLEDGEMENTS The authors would like to acknowledge the funding support of the Ford Motor Company Limited to carry out this work as well as the Powertrain and Vehicle Research Unit (PVRU) at the University of Bath for assisting in this work.

REFERENCES

1. A Systematic Analysis of CO2-Reduction be an Optimized Heat Supply during Vehicle Warm-up Klaas Kunze, Stefan Wolff, Irina Lade and Johann Tonhauser (BMW Group) SAE Paper 2006-01-1450

2. Further Investigations on Time-Alignment CD Bannister, JG Hawley, CJ Brace, A Cox, SAE Paper 2004-01-1441

3. Fuel Consumption Measurement – Techniques and Issues CJ Brace, JG Hawley, invited paper for UnICEG winter 2006 meeting, December 2005.

4. Determining the effect of lubricating oil properties on the diesel engine fuel economy - CD Bannister, JG Hawley, CJ Brace, I Pegg, JC Dumenil, A Brown ‘Internal Combustion Engines: Performance, fuel economy and Emissions’, 11 – 12 December 2007


Documents

The Effect of Forced Cool Down on Cold Start Test Repeatability