arc
AUTOMOTIVE RESEARCH CENTER
NAC
MODELING AND INTEGRATION OF
CONVENTIONAL AND HYBRID TRUCK SYSTEMS
Zoran Filipi, Loucas Louca,
Yongsheng Wang, Chan-Chiao (Joe) Lin, Erik Simmons,
Dennis Assanis, Huei Peng, Jeffrey Stein
6th Annual ARC Conference
Ann Arbor, May 24, 2000
AUTOMOTIVE RESEARCH CENTER
arc
Outline
y Introduction
y Variable fidelity and complexity
y New component and control modules - VNT
y Hybrid Electric Truck Study
y Sub-system and component models
y Performance and fuel economy results for the Class VI truck
y Conclusions
NAC
arc
AUTOMOTIVE RESEARCH CENTER
Objectives
y 2X-3X fuel economy:
y Optimize advanced diesel powerplants
y Explore where hybridization makes sense
y Improve the match between the powertrain and the vehicle
y Improved accessories, e.g. electric pump
y Reduce emissions:
y New technologies for improved NOx/PM trade-off, e.g.
VNT+EGR, HCCI …
y Electric powertrain components
y Fuel cell
y Active Safety:
y Rollover warning and prevention
y Differential braking, engine braking
y Intelligent Transportation – truck “platooning”
NAC
AUTOMOTIVE RESEARCH CENTER
arc
Key Modeling Issues
y Variable complexity and fidelity of engine, driveline
and vehicle dynamics modules, depending on overall
vehicle simulation goal
y Development of advanced component models
y Feed-forward system for realistic transient response
and control
y Integration in different computing environments:
flexibility vs. computational speed trade-off
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Integrated Simulation in SIMULINK –
Vehicle-Engine SIMulation
1-2 Band Clutch
Second Clutch
Shift duration
Input Planetary
800
600
ReactionPlanetary
Speed ratio
First Clutch
To Final Drive
400
Torque ratio
AUTOMOTIVE RESEARCH CENTER
arc
200
NAC
Time [sec]
0
300 0
Longitudinal
Heave
200 0
•Total Vehicle Mass
•Aerodynamic Drag
8
100 0
6
4
0
x 10
-5
2
Sprung
Mass
Suspension
Unsprung
Mass
Coupling
C
•Wheel Inertia
•Wheel Slip
•Rolling Resistance
T
Flexible Tire
Road Excitation
InterCooler
IM
EM
Sprung
Mass
Sprung Mass
Front
Suspension
Front
Tire
Front
Unsprung
Mass
Front
Wheel
Inertia
Front
Suspension
Front
Tire
Front
Unsprung
Mass
Front
Wheel
Inertia
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Previous and Ongoing VESIM Applications
y Extensively tested under variety of very dynamic conditions
y Validated against experimental data
y High fidelity version used by International and UM for mobility and
driveability studies
y Reduced complexity version developed for target cascading
Vehicle Specifications:
International 4700 Series
Engine Speed
Validation: 0 - 60 mph
GVWR: 7950 Kg
Wheelbase: 3.7 m
CG Location: 2.2 m from front
Frontal Area: 5 m2
Air Drag Coefficient (CD): 0.8
4 Speed Automatic Transmission
Rear Wheel Drive - 4x2
y V8 DI Diesel
70
y Turbocharged, Intercooled
60
y Displacement: 7.3 liters
y Bore: 10.44 cm
y Stroke: 10.62 cm
y Compression Ratio: 17.4
y Rated Power: 155 kW@2400 rpm
Vehicle Speed
•
•
•
•
•
•
•
2750
2500
2250
2000
1750
1500
1250
1000
750
500
Simulation
Measurement
0
5
10
15
20
25
30
35
50
40
30
20
Simulation
Measurement
10
0
5
10
15
Time (s)
20
25
30
35
arc
AUTOMOTIVE RESEARCH CENTER
Current Emphasis
y Increase the fidelity and add new components for
studies of critical driveability and emissions issues
y New system engine configurations for advanced diesels
y Novel turbomachinery models
y Exploring matching and control strategy
y Develop integrated HEV simulation:
y Evaluate modeling options
y First generation models of electric components
y Basic power controller
y Initial fuel economy study
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Model Complexity = f (Simulation Goals)
HIGH-SPEED
Conventional
Hybrid
Mobility, fuel economy
over driving cycles,
subsystem
matching,optimization
Mobility, fuel economy,
energy management,
subsystem matching,
optimization
HIGH-FIDELITY
Critical transients (e.g.
launch), driveability,
control, component
matching,emissions
Critical transients (e.g.
starting-stopping),
emissions
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Engine System Models in SIMULINK
High-Fidelity
ENGINE
DYNAMICS
High-Speed
CYLINDER PHASING
ENGINE
DYNAMICS
INTAKE
MANIFOLD
TURBOCHARGER
INTERCOOLER
FUEL
CONTROLLER
EXHAUST
MANIFOLD
FUEL
CONTROLLER
ENGINE
CYLINDERS
TORQUE
LOOK-UP TABLE
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Motivation for the Application of VNT
to a Diesel Powertrain
y Mobility:
y Improved response and acceleration performance
y Emissions:
y Control pressure difference across the engine for EGR
application to heavy duty diesels
Standard
Emissions Limits [g/bhp.h]
Test Cycle
NOx
US'98
1998
US-FTP
US'02
2002
US-FTP
NMHC
4.0
-
THC
CO
1.3
15.5
PM
0.10
(0.05-bus)
2.5
1.3
15.5
0.10
(0.05-bus)
AUTOMOTIVE RESEARCH CENTER
arc
EGR System - Options and Implications
y EGR – external:
y short route (manifold to manifold)
y long route
y pumped EGR
y EGR – internal:
y variable valve timing (camless)
y In every case, negative pressure drop needed across
engine
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Neural Network for Compressor Modeling
3.5
3.4
3.2
Predicted
Measured
GTP
38
3
3
72%
Pressure Ratio
Pressure Ratio
2.8
68%
74%
125921
2.6
2.4
60%
76%
2.2
77%
2
113647
1.8
2.5
2
105323
1.6
96257
1.5
1.4
83854
69773
1.2
1
0
45953
10
20
30
40
50
60
1
0
70
Corrected
Flow
Corrected
Mass Air
Flow
Rate(lbs/min)
10
20
30
40
50
Corrected Mass Flow Rate
Speed
Mass Flow Rate
Train NN
Pressure Ratio
.
.
.
.
.
.
.
.
.
ANN model
Efficiency
60
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Variable Nozzle Turbine (VNT)
VNT allows modification of turbine flow characteristics for best
performance:
y “small” flow area (nozzles partially closed) for low engine speed and load
y “large” flow area (nozzles open) for high speed and high load
rbin
e
Sm
all T
u
Compressor pressure ratio
Turbocharger Performance
Target Boost Level
m
Co
pr
om
ise
Engine Speed
L
ge
ar
Tu
e
in
b
r
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Neural Network Model of VNT Turbine
0 .8
0 .7
M e a s u re d
P re d ic t e d
0 .6
0 .5
0 .4
0 .3
0 .2
0 .1
0 .8
0 .7
0 .6
M e a s u re d
P re d ic t e d
0 .5
0 .4
0 .3
0 .2
0 .1
0
0
0 .2
0 .4
0 .6
0 .8
T u rb in e E x p a n s io n R a t io (N o rm a liz e d )
1/4
1
0 .2
0 .4
0 .6
0 .8
T u rb in e E x p a n s io n R a t io (N o rm a liz e d )
V a n e P o s it io n : 2 5 %
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
0 .3
M e a s u re d
P re d ic t e d
0 .2
V a n e P o s it io n : 5 0 %
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
M e a s u re d
P re d ic t e d
0 .3
0 .2
0 .1
0
0
C o rre c t e d T u rb in e M a s s F lo w R a t e (N o rm a liz
0
V a n e P o s it io n : 7 5 %
1
0 .9
Corrected Turbine M as s Flow Rate (Norm alized)
V a n e P o s it io n : 1 0 0 %
1
1
1/2
C o rre c t e d T u rb in e M a s s F lo w R a t e (N o rm a liz
C o rre c t e d T u rb in e M a s s F lo w R a t e (N o rm a liz
3/4
0 .9
C o rre c t e d T u rb in e M a s s F lo w R a t e (N o rm a liz
C o rre c t e d T u rb in e M a s s F lo w R a t e (N o rm a liz
Blade (nozzle)
Position =1
0 .1
0
1
0
0 .2
0 .4
0 .6
0 .8
T u rb in e E x p a n s io n R a t io (N o rm a liz e d )
1
V a n e P o s it io n : 0 %
1
0 .9
0 .8
0 .7
0
0 .6
0 .5
0 .4
0 .3
0 .2
M e a s u re d
P re d ic t e d
0 .1
0
0
0 .2
0 .4
0 .6
0 .8
T u rb in e E x p a n s io n R a t io (N o rm a liz e d )
1
0
0 .2
0 .4
0 .6
0 .8
T u rb in e E x p a n s io n R a t io (N o rm a liz e d )
1
Discrete number of maps
for fixed blade (nozzle) positions
0.9
100%
0.8
75%
0.7
50%
0.6
0.5
25%
0.4
0.%
0.3
Predictions for
any blade position
0.2
0.1
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Turbine E x pansion Ratio (Norm alized)
0.8
Nozzle Position
Mass Flow Rate
Speed
Train ANN
Pressure Ratio
.
.
.
.
.
.
.
.
.
ANN model
Efficiency
0.9
1
arc
AUTOMOTIVE RESEARCH CENTER
NAC
Application of VNT:
Control Strategy for Maximum Mobility
y Nozzle Position signal varies from 0 – 1:
y If boost pressure below corresponding steady-state values
obtained with the conventional turbocharger NP=0
y If boost pressure is above corresponding steady-state values
obtained with the conventional turbocharger and below a set
target level NP varies between 0 and 0.7
y If boost pressure is above the target value NP between 0.7 and 1
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: The Effect on Engine and Vehicle
Response
Vehicle launch and acceleration from 0 - 60 mph
3000
45
VGT
CT
40
2500
30
E ngine S peed (rpm )
V ehic le S peed (m ph)
35
25
20
15
10
2000
V GT
CT
1500
1000
5
500
0
0
2
4
6
8
10
12
Tim e (sec )
14
16
18
20
0
5
10
15
Tim e (s ec )
20
25
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: Nozzle Position Signal and
Turbocharger Speed
0.9
110
0.8
100
90
0.7
80
Turbo Speed (krpm)
nozzle position
0.6
0.5
0.4
0.3
70
60
50
40
0.2
VGT
CT
30
0.1
20
0
0
2
4
6
8
10
12
Time (sec)
Start of the transient
14
16
18
20
10
0
5
10
15
20
25
Time (sec)
30
35
40
45
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: Turbine Operating Points
During Vehicle Acceleration
Correc ted Turbine M as s F low Rate (Norm aliz ed)
1
0.9
1.00
0.8
0.75
0.7
0.50
0.6
0.5
0.25
0.4
0.00
0.3
0.2
0~ 60 ac c eleration w ith V G T
0.1
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Turbine E x pans ion Ratio (Norm aliz ed)
0.8
0.9
1
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: Compressor Operating Points
During Vehicle Acceleration
1
0.9
0~ 60 ac c eleration w ith V G T
0.8
P res s ure Ratio (Norm aliz ed)
" *" : S tarting P oint
0.7
" o": E nding P oint
0.6
0.5
N9
0.4
N8
0.3
N7
N6
0.2
N5
0.1
N2
N1
0
0
0.1
0.2
0.3
0.4
N4
N3
0.5
0.6
0.7
0.8
Correc ted Com pres s or M as s F low Rate (Norm aliz ed)
0.9
1
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: Manifold Pressure and the Response
of the Fuel Controller
Intake Manifold Pressure (Pa)
5
2.5
x 10
2
VGT
CT
1.5
1
0.5
0
2
4
6
8
10
12
Time (sec)
14
16
18
20
2
4
6
8
10
12
Time (sec)
14
16
18
20
Fuel Injected (kg/cycle)
-5
8
x 10
6
4
2
0
0
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VNT: Changing the Control Strategy to
Affect Pressure Ratio Across the Engine
5
2.4
5
Target: 1.8
x 10
Increased
Target
2.2
2.4
2.2
2
2
1.6
1.4
1.8
Presure (Pa)
Value
1.8
Presure (Pa)
Target: 2.0
x 10
1.6
1.4
Exhaust Manifold
Intake Manifold
Exhaust Manifold
Intake Manifold
1.2
1.2
1
1
0.8
0
2
4
6
8
10
12
Time (sec)
14
16
18
20
0.8
0
2
4
6
8
10
12
Time (sec)
14
16
By increasing the target value we increased not only the
average boost level at high engine speed, but also the negative
pressure difference across the engine (Pexhaust – Pintake)
18
20
arc
AUTOMOTIVE RESEARCH CENTER
Future Directions – VNT and High
Fidelity Engine System
y Improved component modules for advanced diesel
and hybrid electric systems, e.g. EGR sub-system
components, camless valvetrain
y Explore various system configurations for EGR
application
y Explore engine control strategies for improved
response and reduced emissions – utilize predictive
spray, combustion and emissions models
y Validate component and system predictions
NAC
arc
AUTOMOTIVE RESEARCH CENTER
VESIM-HEV Case Study
y HEV Class VI Truck:
y Proper modeling
y Performance
y Fuel economy over a driving cycle
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Parallel Hybrid Truck Configuration
Vehicle Dynamics
Traction Force
Engine
Exhaust
Gas
T
TC
ICM
DS
Drivetrain
EM
PS
Trns
D
C
Air
Inter
cooler
Motor
Power
Control
Module
IM
DS
Battery
arc
AUTOMOTIVE RESEARCH CENTER
Drivetrain
TC
Trns
Propshafts
Driveshaft
D-R
NAC
arc
AUTOMOTIVE RESEARCH CENTER
NAC
Vehicle Dynamics
Sprung
Mass
Sprung Mass
Rear
Suspension
Rear
Tire
Rear
Unsprung
Mass
Rear
Wheel
Inertia
Front
Suspension
Front
Tire
• Nonlinear 3D kinematics
• Aerodynamic and rolling resistance losses
• Nonlinear tire traction forces
Front
Unsprung
Mass
Front
Wheel
Inertia
AUTOMOTIVE RESEARCH CENTER
arc
Proper Model: Definition
Characteristics
y Minimum complexity
Advantages
y Computational efficiency
y Design insight
y Physical parameters
y Geometry
y Physically meaningful
y Direct design decisions
y Material
y Clear interpretation of model
y Specified level of accuracy
predictions
Can meet model objectives
NAC
AUTOMOTIVE RESEARCH CENTER
arc
Proper Models
y Need proper models for:
y System evaluation
y Design/Sensitivity studies
y Ad-hoc modeling approach:
y Require experience to build and interpret models.
y Require excessive time to:
¡ Develop
model
¡ Compute model
y Lack systematic approaches to generate models of
appropriate:
¡ Complexity
¡ Accuracy
NAC
AUTOMOTIVE RESEARCH CENTER
arc
Automated Modeling Algorithms
y Overly-complicated models are:
y Difficult to compute
y Difficult to:
¡ Do
design optimization, controller design, etc.
¡ Gain physical insight into the system behavior
y Automated modeling algorithms addresses the model
complexity in a systematic and quantitative approach:
y Frequency-based MODA - Linear systems
y Energy-based MORA - Nonlinear systems
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Proper Models for Fuel Economy
Drivetrain
Vehicle Dynamics
TC
arc
Trns
• Quasi-Static
• Lookup Tables
AUTOMOTIVE RESEARCH CENTER
Longitudinal
• Gear Ratios
• Blending Functions
Heave
•Total Vehicle Mass
•Aerodynamic Drag
Rigid Propshafts
Suspension
Coupling
•Wheel Inertia
•Wheel Slip
•Rolling Resistance
Rigid Driveshaft
D-R
NAC
Flexible Tire
Sprung
Mass
Unsprung
Mass
Road Excitation
AUTOMOTIVE RESEARCH CENTER
arc
Engine Model
y Look-up table for engine output torque as a function of
fuel input and speed
y Includes fuel injection controller and engine dynamics
Torque (N-m)
800
600
400
200
0
3000
2000
Engine
Speed
8
1000
6
0
4
2
x 10
Fuel Injected
-5
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Electric Motor Model
y Permanent magnet DC
200
7
0.7 5
0.
Motoring
Torque
100
0. 9
0 .8
0 .8 5
0 .9 3
50
0.9
0.8 5
0
8
0. 85
0.
M otor Torque (Nm )
150
0.85
0. 9
-50
200
0.93
100
3
0
0.9
-150
9
-100
0.
to 117 kW
250
5
0.8
0.8
.7
0.7 5 0
y Reduce engine size form 157 kW
M otor E ffic ienc y
0 .9 3
motor/controller
y 49 kW Max
y Max Propulsive Torque :
274.4 N-m
y Max Reg. Torque : 170 N-m
y Mass : 60 kg
y (Source : Advisor 2.2.1)
300
400
500
600
M otor S peed (rad/s )
700
800
Regenerating
Torque
AUTOMOTIVE RESEARCH CENTER
arc
Battery Model
y Lead Acid Battery
y Max Capacity : 18 Ah
y Mass : 165 kg
y Max Voltage : 412.5 volts
y Battery efficiency, open circuit voltage and internal resistance are
function of battery state of charge (SOC)
SOC =
Qmax − Qused
Qmax
equivalent circuit model (Powell and Pilutti, 1993)
~
R
Voc
Rb
ia
Cp
~
etb
ib
Ci
tb
Rb Ci etb = e~tb − etb − Rb itb
e tb
~
~
 Rb + R  ~
R
~ ~
 etb +
RC p etb = Voc − 
e
Rb tb
 Rb 
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
HEV Power Control Function
Shift Logic Control
HEV Power Control Module
T_eng
P_eng
Driver
Model
P_dem
Engine/Trans.
Control Module
Power Splitting
control module
P_mot
Motor Only Mode
Engine Only Mode
Hybrid Mode
Charge Sustaining Strategy
Motor
Control Module
Gear #
Clutch
Vehicle
Model
T_mot
Battery Power Limit
Regenerative Braking
Control
AUTOMOTIVE RESEARCH CENTER
arc
NAC
HEV Power Splitting Strategy
Î
“Engine on” power level
Î
“Engine high” power level
High power
Î
Æ Both engine & motor
24
26
0.
0.
0.
7
0 .2
5
0 .2
400
0 .2 1 2
Æ Use
450
23
350
300
250
200
150
100
6
Medium power & recharge
engine
500
4
Ê
Prevent the engine from operating
at an inefficient region
0 .2 1
Ê
F uel Cons um pt ion (k g/k W hr)
0 .2
21
Î
Æ Motor only
2
0.2 4
0.26
0. 27
0.
Low power
E ngine Torque (Nm )
Ê
Power assist
0.23
0.25
50
800
1000
1200
1400
1600
1800
E ngine S peed (rpm )
The motor is used for power assist
Motor only
2000
2200
2400
arc
AUTOMOTIVE RESEARCH CENTER
VESIM-HEV in SIMULINK
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
HEV Specifications: Class VI Truck
GVWR: 7950 Kg
Frontal Area: 5 m2
Air Drag Coeff.: 0.8
ENGINE size reduced to:
ELECTRIC MOTOR:
BATTERY:
y V6 DI Diesel
y Power: 49 kW
y Lead-Acid
y Turbocharged, Intercooled
y Max Propulsive
y Max Capacity: 18 A-h
y Displacement: 5.47 liters
(same Bore and Stroke as
V8)
y Rated Power: 117 kW
@2400 rpm
Torque: 274.4 N-m
y Max Reg. Torque:
170 N-m
y Mass: 60 kg
y Max Voltage: 412.5 V
y Mass: 165 kg
arc
AUTOMOTIVE RESEARCH CENTER
VESIM – HEV: Performance Test,
Acceleration: 0-60mph
NAC
AUTOMOTIVE RESEARCH CENTER
arc
VESIM – HEV: FUDS Cycle
Fuel Economy
(mpg)
Conventional
HEV (Charge
Sustaining)
HEV (Charge
Depleting)
9.59
12.63
17.75
NAC
AUTOMOTIVE RESEARCH CENTER
arc
NAC
VESIM – HEV: FUDS Cycle,
Time Zoom In: 400 ~ 550 sec
Engine off at
cruising
Motor starts up
Regenerative Braking
Engine drives
vehicle and
charges battery
SOC hits low limit
Engine recharges battery
AUTOMOTIVE RESEARCH CENTER
arc
NAC
Engine Operation Points
VESIM - HEV
VESIM - Conventional
F uel Cons um ption (k g/k W hr)
F uel C ons um ption (k g/k W hr)
700
500
200
30000
800
1000
1200
1400
1600
25
0.
00
00
00
6
00
7 00
00
5 00 0
0 .2
3
0.2 4
0.25
0.26
0. 27
0
3 00 00
100
10 00 0
50
00
21
6
21
0.
0.2
3
0. 24
0.25
0.26
0. 27
300
15
00
9 00
12
100
00
4
150
500
00
11
400
0 .2 2
200
13
0 .2
4
250
7
6
0 .2
4
2
0.
3
0 .2
500
0
00
2
0.
0 .2 1
700
00
600
00
0.
90
0.21
0
.23
300
0 .2 2
E ngine torque (N m )
350
0 .2 1 2
7
0.26
0.2
4
5
0.2 0.2
400
00
E ngine torque (N m )
11
450
10 00 0
1800
2000
2200
2400
800
1000
1200
1400
1600
1800
2000
E ngine s peed (rpm )
E ngine s peed (rpm )
Shift logic control toward more efficient region
Use motor to avoid engine inefficient operation
2200
2400
arc
AUTOMOTIVE RESEARCH CENTER
NAC
Future Directions - HEV
y
Develop improved electrical component models
y
Apply proper modeling methodology to other systems
y
Use worst-case analysis to generate driving cycle able
to reveal possible weaknesses of overall HEV system
y
Dynamic energy HEV management strategies
y
Support target cascading for HEV vehicle system
optimization
y
Explore synergisms with other HEV system level
codes (e.g. PSAT)