Data Collection for Improved Cold Temperature Thermal Modeling and Strategy Development

2011 DOE Hydrogen and Vehicle TechnologiesAnnual Merit Review

May 10, 2011

Forrest Jehlik, Eric RaskArgonne National Laboratory

Sponsored by Lee Slezak

This presentation does not contain any proprietary, confidential, or otherwise restricted information

Project ID # VSS050

1Q 2010-current 50% complete

– Complete evaluation of Gen 2 Prius– Begun evaluation of 2010MY Gen 3

Prius– Leverages APRF benchmarking and

Environment Canada vehicle testing

Barriers addressed– Cold ambient testing shows

significant fuel consumption increase (w/o creature comforts)

– Engineered solutions could have potential drastic petroleum use reduction

Timeline

Budget

Barriers

Partners Environment Canada Total project funding

– FY10 funding: $200k

– FY11 funding: $300k

2

Overview

Majority of fuel economy testing does not reflect real world ambient conditions

Four back-to-back UDDS cycles

Relevance: Cold ambient temp greatly reduces hybrid fuel economy

3

30%60%

-10

0

10

20

30

40

50

60

70

80

0 200 400 600 800 1000 1200 1400

MPH

/Toi

l

Toil (Tamb = -5oC) Toil (Tamb = 35oC)

UDDS engine-off operation changes

Engine-off operation changes dramatically under cold ambient conditions and during vehicle warm-up

Engine efficiency differs throughout the entire warm-up period as a function of ambient temperature

Relevance: Cold operation impacts vehicle operation and efficiency

Steady state temperature reduced

4

Relevance: Cold ambient operation still affects vehicle controls after initial warm-up

Usage Oscillation Due to Engine-off Cool-down to Ambient

Coolant Temperature Oscillation Due to Cool-down

Eng. Off disabled for “hot” run

Coolant temp. lower than other cycles

* Cool-down issue is even larger for PHEVs5

Simple methodology to estimate the impact of cold engine temperature on engine efficiency and estimate engine thermal state given usage and ambient temperature

– Current methodology uses coolant temperature, but can be generalized to other temperatures

– Methodology uses response surface and empirical data-fitting techniques

– Techniques result in simplified general models

Cold Ambient Testing

Fueling vs. Engine Temperature, Speed, Load

Coolant Temperature vs. Engine Usage, Ambient Temp.

Engine Usage and Starting Conditions

Estimated Temperature

Profile

Temperature based Engine Fueling and Fuel

Economy Estimate

Response Surface Methodology Model (RSM)

Approach: Combine RSM fueling map with temp prediction model

Lumped capacitance model

6

Hymotion Prius was used for vehicle testing and methodology development

Vehicle testing focused on the UDDS and US06 cycles to obtain a cross section of usage

UDDS

020406080

100120

0 200 400 600 800 1000 1200 1400

Time (s)

Veh

icle

Spe

ed (k

m/h

)

US06

020406080

100120140

0 100 200 300 400 500 600

Time (s)

Veh

icle

Spe

ed (k

m/h

)

Signal Description NotesVehicle Speed Dynamometer/

CAN measuredDrive trace

measurementEngine RPM CAN/ spark

frequencyModel input

Brake Torque CAN Flywheel torque, model

inputT_oil Dipstick

thermocoupleModel input

T_coolant CAN Model inputFuel Flow Emissions

Bench carbon count

Model input

Approach: Experimental setup

7

Project will develop a methodology for estimating the impact of engine thermal state on engine efficiency during cold ambient operation

– Quantifying/qualifying magnitude of losses highlight benefit for engineered solutions• Opportunity to greatly reduce all season, real world fuel consumption

Approach: Data from multiple cycles at varied temperatures

8

Estimated fueling versus temperature and load works well (within 1%)

For more robust estimations, additional low temperature testing required

UDDS Estimated vs. Actual Fuel Used Fuel Rate During Warm-up

Accomplishments: Fueling rate sensitive to temperature

Estimated fueling map shows a strong sensitivity to coolant temperature

UDDS cycle coolant temperature UDDS engine speed/load points

25°C 50°C

10

Using a single temperature does not provide adequate estimation during warm-up– Necessary to determine the number of points required to adequately reduce error

In this example, at least four temperatures are needed to evaluate the previous UDDS comparison points within 1%

Accomplishments: Temperature resolution sensitivity

UDDS Fuel Rate Error: Single Temp. vs. Time-series Total Fuel Comparison: Single Temp. vs. Time-series

While only 4 points are required, minimal points would not allow for control strategy optimization/ characterization of where losses are most prevalent

11

Accomplishments: Estimating thermal state

Max Error ~6 deg.

12

Developed technique works well for estimating UDDS coolant temperature

During US06 cycle operation, coolant temperature is over estimated due to significant thermostat operation

Modeled UDDS and US06 cycles when operating in cold ambient conditions

Cycle Percent Error of Actual versus Estimated Grams Fuel Used (%)

UDDS #1 1.3UDDS #2 0.7US06 #1 3.0US06 #2 4.4

Accomplishments: Model works well predicting UDDS/US06

Back-to-Back UDDS Runs

13

-5oC

15oC

35oC

3

4

5

6

7

8

9

Fuel

con

sum

ptio

n (L

/100

km)

14

Accomplishments: Early results clearly show large petroleum reduction potential*

-10

-5

0

5

10

15

20

25

30

Am

bien

t tem

pera

ture

[oC]

Chicago Washington DC Los Angeles

20

22

24

26

28

30

32

Gal

lons

con

sum

ed /

mon

th

Chicago Washington DC Los Angeles

*Results shown are trends from experimental, not modeling data

37%

25%Predominant EPA test temperature range

1. Testing at multiple temps… 2. Greater variation in real world…

3. Significantly higher fuel consumption

Real world

There are still many open issues to investigate

– Improve procedure/technique for model development

– Define potential for engineered solutions to reduce real world fuel consumption

– Incorporate creature comfort features into modeling effort (NREL)

– Incorporate catalyst light off features into modeling effort (ORNL)

– Further investigate thermal linkages between components

– Ensure robustness with additional vehicle testing leveraging APRF thermal capability upgrade and continuing Environment Canada collaboration

Conclusions/Proposed future work

Alternative Temperature Signals and Cool-down

15

Automonie•Development of thermal capability in models

16

J1711 HEV & PHEV test procedures• Early thermal worked

used in guidance

DOE technology evaluation• Future collaborative potential with ORNL/NREL

Environment Canada•Testing and tech support

Collaborations and coordination with other institutions

APRF• Warm testing data collected at APRF• APRF under construction to begin

cold testing

®

http://www.sae.org/�http://images.google.com/imgres?imgurl=http://www.pppl.gov/ncsx/ORNLlogo.jpg&imgrefurl=http://www.pppl.gov/ncsx/&h=196&w=358&sz=17&hl=en&start=1&tbnid=x8Wk1C7Sl4DZQM:&tbnh=66&tbnw=121&prev=/images?q=ORNL+logo&svnum=10&hl=en&lr=�

Summary Current vehicle testing protocols under represent real world ambient conditions

Methodology developed to predict real season fuel consumption in HEV’s• Response surface fueling rate modeling technique finalized• Lumped capacitance temperature modeling technique developed

Modeling shown to be relatively accurate relative to experimental data• Model vs. experimental fuel consumption data within a few %

Results demonstrate significant engineering potential to reduce petroleum consumption

Further work needs to be complete to assess how much efficiency can be gained and the most effective pathways

17

PublicationsI. “PHEV Energy Management Strategies at Cold Temperatures with Battery Temperature Rise and Engine Efficiency

Improvement Considerations”, Shidore, Neeraj, S., Jehlik, F., Rask, E., SAE World Congress, Detroit, SAE 2011-01-0872

II. “Development of Variable Temperature Brake Specific Fuel Consumption Engine Maps”, Jehlik, F., Rask, E., SAE Powertrain Fuels and Lube Conference, San Diego, SAE 2010-01-2181

III. “Simplified Methodology for Modeling Cold Temperature Effects on Engine Efficiency for Hybrid and Plug-in Hybrid Vehicles”, Jehlik, F., Rask, E., Christenson, M., SAE Powertrain Fuels and Lube Conference, San Diego, SAE 2010-01-2213

IV. “Methodology and Analysis of Determining Plug-In Hybrid Engine Thermal State and Resulting Efficiency”, Jehlik, F., SAE World Congress, Detroit, Mi., SAE 2009-01-13081

Slide Number 1Slide Number 2Slide Number 3Slide Number 4Slide Number 5Slide Number 6Slide Number 7Slide Number 8Slide Number 9Slide Number 10Slide Number 11Slide Number 12Slide Number 13Slide Number 14Slide Number 15Slide Number 16Summary