DESIGN AND DEVELOPMENT PROCESS FOR A RANGE …...For the EcoCAR competition, HEVT was supplied...

1 Copyright © 2010 by ASME

Proceedings of the ASME 2010 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference

IDETC/CIE 2010 August 15-18, 2010, Montreal, Quebec, Canada

DETC2010-28576

DESIGN AND DEVELOPMENT PROCESS FOR A RANGE EXTENDED SPLIT PARALLEL HYBRID ELECTRIC VEHICLE

Lynn R. Gantt Graduate Research Assistant

Blacksburg, Virginia, USA

Patrick M. Walsh Graduate Research Assistant


Douglas J. Nelson Professor, Virginia Tech


ABSTRACT The Hybrid Electric Vehicle Team of Virginia Tech

(HEVT) is participating in the 2009 – 2011 EcoCAR: The

NeXt Challenge Advanced Vehicle Technology Competition

series organized by Argonne National Lab (ANL), and

sponsored by General Motors Corporation (GM) and the U.S.

Department of Energy (DOE). The goal of EcoCAR is for

student engineers to take a GM-donated crossover SUV and re-

engineer it to reduce greenhouse gas emissions and petroleum

energy use, while maintaining performance, safety and

consumer appeal. Following GM's Vehicle Development

Process (VDP), HEVT established team goals that meet or

exceed the competition requirements for EcoCAR in the design

of a plug-in range-extended hybrid electric vehicle. HEVT is

split up into three subteams to complete the competition and

meet the requirements of the vehicle development process. The

Mechanical subteam is tasked with modifying and refining the

Year 1 component specifications and designs for packaging in

the vehicle. The Electrical subteam is tasked with

implementing a safe high voltage system on the vehicle

including the design and development of a Lithium Iron

Phosphate (LiFePO4) energy storage subsystem (ESS) donated

by A123 Systems. The Controls subteam is tasked with

modeling the Vehicle Technical Specifications (VTS) so that

the subteams can make intelligent design decisions. The

Controls subteam also used a controller Hardware-In-the-Loop

(HIL) simulation setup running a real-time vehicle model

against the controller hardware to test the HEVT-designed

Hybrid Vehicle Supervisory Controller (HVSC). The result of

this design process is an Extended-Range Electric Vehicle (E-

REV) that uses grid electric energy and E85 fuel for propulsion.

The vehicle design is predicted to achieve an SAE J1711 utility

factor-corrected fuel consumption of 2.9 l(ge)/100 km (82

mpgge) with an estimated all-electric range of 69 km (43

miles). Using corn-based E85 fuel in North America for the

2015 timeframe and an average North American electricity

mix, the well-to-wheels petroleum energy use and greenhouse

gas emissions are reduced by 90 % and 30 % respectively when

compared to the stock vehicle: a 4-cylinder, gasoline-fueled

Vue XE.

INTRODUCTION HEVT is in the process of developing and validating a

hybrid architecture that was selected by the Year 1 team. This

paper will first show how HEVT used the VDP to select the

hybrid powertrain architecture in order to reduce petroleum

energy use and greenhouse gas emissions. The paper will then

explain the physical integration difficulties that were solved in

software using computer aided design (CAD) with Unigraphics

NX6 software, along with the limitations of the modified

HEVT vehicle CAD model. In order to interface with and

control the vehicle, HEVT added a National Instruments

CompactRIO (cRIO) to act as a hybrid vehicle supervisory

controller (HVSC). An explanation of the control strategy and

the vehicle controller goals are shown, as well as the

development of the vehicle control logic using HIL. The

"VTREX" as the vehicle is known to the team (Virginia Tech

Range-Extended Crossover), is compared to the baseline

competition requirements and specifications as well as against

the team-predicted VTS. Preliminary testing and validation of

the vehicle are used to show safe operation and explain any

discrepancies between the simulated model and actual vehicle

results. HEVT plans to use Year 3 of EcoCAR to refine the

HVSC while light-weighting components added to the vehicle

to further reduce vehicle energy consumption and improve

drivability.

NOMENCLATURE HEVT: Hybrid Electric Vehicle Team of Virginia Tech


GM: General Motors Corporation

DOE: United States Department of Energy

ANL: Argonne National Labs

EREV: Extended Range Electric Vehicle

UDDS: Urban Dynamometer Drive Schedule (City)

HWFET: Highway Fuel Economy test (Highway)

CS: Charge Sustaining (engine dominant hybrid)

CD: Charge Depleting (electric only mode)

MPGGE: miles per gallon gasoline equivalent

HIL: Hardware in the Loop

TEAM GOALS For the EcoCAR competition, HEVT was supplied

minimum requirements for vehicle efficiency, vehicle utility,

and performance. The team decided to surpass the goals set

forth by the competition by adding goals for petroleum

reduction, all-electric operation, passenger space, and towing,

as shown in Table 1. The goal of having a large all-electric

range, in order to displace petroleum energy and provide a

consumer delighter, is met by a battery pack with an energy

capacity of 15 kWh useable (20 kWh total). The HVSC is

simpler to develop for all-electric driving than an engine-

blended charge depleting strategy. Further dependence on

petroleum is reduced through the use of a FlexFuel engine from

GM, which allows the vehicle to be fueled with E85.

Table 1: HEVT Goals

Goal Description

Petroleum Reduction

Reduce petroleum consumption by

> 80 %

Plug-in Range

> 56 km (35 mi) range as a pure

all-electric vehicle

Cargo and Passenger Space

Retain stock cargo space and 5

passenger capacity

Towing

Increase mass, speed and grade

requirements to 5% at 89 km/hr

The petroleum reduction goal is directly related to the

goal of a large all-electric range. By storing energy in a battery

pack from the electric grid for propulsion, less petroleum is

used to propel the vehicle during the initial charge depleting

(CD) mode of vehicle operation. Consumer acceptability

features are important design criteria for vehicle performance

metrics as well, and include meeting the 0-60 mph acceleration

test, the 50 - 70 mph passing acceleration test, and the

competition drive cycles. HEVT designed the VTREX to meet a

0-60 mph time of less than 9 seconds in charge sustaining (CS)

mode and less than 14 seconds in all-electric, CD mode. The

final goal set by HEVT is high fuel economy. The team has

predicted a combined fuel economy of 84 mpgge based on the

SAE J1711 utility factor-weighted standard and a 30 mpgge

combined fuel economy for CS mode. These goals will help

ensure that HEVT creates a vehicle that meets the competition

requirements while safely completing all events.

VEHICLE TECHNICAL SPECIFICATIONS

The critical vehicle technical specifications (VTS) can be

seen in Table 2 while the full competition VTS can be seen in

Appendix A. Powertrain Systems Analysis Toolkit (PSAT),

developed by Argonne National Lab (ANL), is used to translate

the competition goals into engineering metrics for vehicle

component selection [1]. The vehicle powertrain architecture is

designed around the general HEVT goals. The overall

specifications for the vehicle come from a combination of

PSAT modeling and the components available for vehicle

integration. The specifications are updated as more information

is supplied about the components so that the VTS remain an

accurate representation of the vehicle. Maintaining current and

accurate VTS is important because later in the EcoCAR

competition, the teams will be judged on how well the as-built

vehicle performs against the predicted VTS. For towing

capability, the team meets the goal of increased speed but is

only reporting an increased mass at the required EcoCAR speed

as seen in the supplemental table in Appendix A. The

calculated combined fleet utility factor is 0.64 (64% of time as

EV) based on the proposed SAE J1711 standard.

Table 2: HEVT Vehicle Technical Specifications

HEVT Specifications Metric

Petroleum Reduction in

comparison to the stock Vue

> 80 %

Electric Only Range > 56 km

Towing speed > 89 km/hr

Towing Grade > 5 %

Mass additions < 430 kg

UDDS trace misses < 2 seconds

HWFET trace misses < 2 seconds

US06 trace misses < 12 seconds

IVM – 60 mph, CS mode 8.8 + 0.5 s*

50 – 70 mph, CS mode 4.8 + 0.5 s*

Fuel economy, CAFE unadjusted,

combined with UF correction

2.9 l(ge)/100 km

(84 mpgge)*

Fuel economy, CAFE unadjusted,

combined in CS operation

7.2 l(ge)/100 km

30 mpgge*

* Depending on testing conditions

LITERATURE REVIEW AND DESIGN PARAMETERS

This section focuses on vehicle fuel and powertrain

architecture technology advancements. There are several

existing studies on improvements in automotive powertrain

technologies and their overall impact on energy use and GHG

emissions. Background information will be provided via this

literature review and sources are used to support each category

below.


There are multiple operation modes that a plug-in hybrid

electric vehicle can utilize. These modes are described in the

Boyd and Nelson paper “Hybrid Electric Vehicle Control

Strategy Based on Power Loss Calculations” [2]. The four

modes are as follows: engine only, engine generate, engine

blended, and electric only. HEVT plans to take advantage of

three of the modes but will only use the engine blended mode

in CS operation.

One of the main sources of power for the vehicle and the

power source for the all-electric mode will be a Lithium Ion

battery pack. Lithium Ion batteries are not a brand new

technology, but are gaining popularity in the automotive

industry today. A Lithium Ion battery pack has greater energy

density than a Nickel Metal Hydride pack, allowing for

comparatively lighter and a more energy dense system [3]. One

main disadvantage of Lithium Ion batteries is that there are

more safety considerations than other battery technologies.

Therefore, procedures have been developed by HEVT to ensure

the safe operation of the vehicle and for the protection of the

passengers.

Hybridization combines two power sources to improve

overall powertrain efficiency. Hybrid electric systems are well-

suited for light-duty vehicles, while hydraulic hybrids may have

advantages for larger vehicles. A spark-ignited (SI) or

compression-ignited (CI) engine can be used in a hybrid

system, which pairs an engine with electric motors and an

energy storage system (ESS). One of the most basic hybrid

systems is a 42 V system (36 V battery) that has one electric

motor coupled to the engine. Commonly referred to as an

integrated starter alternator (ISA) or belted alternator starter

(BAS) system, this system can assist the engine, allow engine

start/stop to prevent unnecessary idling, and charge the battery.

Studies and testing have shown that this type of mild hybrid

system can produce fuel economy gains of around 4-7 % [4,5].

Typically the ISA or BAS does not have enough power to allow

electric-only operation.

Engine downsizing is another consideration for

hybridization; where a full hybrid vehicle incorporates one or

two high power electric motors into the powertrain to make up

for the loss of propulsive power due to downsizing. Since these

motors are larger than an ISA or BAS, they may be

mechanically coupled elsewhere in the vehicle powertrain.

Downsizing either an SI or CI engine may raise efficiency by

providing relatively higher loading of the engine, possibly at

the expense of emissions increases.

VEHICLE MODELING AND FUEL SELECTION

In order to select a hybrid powertrain architecture,

HEVT first looked at the fuel and energy that would be used to

propel the vehicle. Basic modeling of both a Vue XE and the

predicted VTS of the VTREX were performed to make the final

architecture decisions. The selection of fuel was also

influenced by the components that would be used to convert the

energy into propulsion on the vehicle. Vehicle fuel has a larger

impact than tailpipe emissions and fuel economy at the vehicle.

Both greenhouse gases and criteria emissions should be

measured over the entire fuel production and consumption

cycle, from well-to-wheels (WTW). Therefore, ANL’s

Greenhouse Gases, Regulated Emissions, and Energy Use in

Transportation (GREET) model was used to calculate total

well-to-pump (WTP) energy, greenhouse gas (GHG) emissions

and criteria emissions released for the fuel as well [2].

EcoCAR teams are given WTP numbers for North America for

different candidate fuels: gasoline (E10), E85, biodiesel (B20),

electricity and hydrogen (H2) [3].

A primary goal of HEVT is the reduction of WTW

petroleum consumption of the VTREX. To meet this goal, and

reduce petroleum consumption by upwards of 80 %, a review

of the candidate fuels was performed. Figure 1 presents the

total WTW petroleum energy content per unit of fuel energy

used at the vehicle for all of the fuels available for EcoCAR on

the x-axis. Tracing the intersection points of the sloped lines of

vehicle efficiency to the columns representing fuel types, the

team was able to see how each fuel will perform for a given

vehicle powertrain efficiency. HEVT used this plot as a design

tool with the team goals to make a comparison between the

fuels. The stock vehicle petroleum consumption, based on

combined CAFE ratings, is shown by the black dot in the top

right; the final estimates of the VTREX in both modes of

operation are shown by the orange dots. The goal of the team

was to reduce vehicle WTW petroleum energy use, displayed

on the y-axis of Figure 1.

Hydrogen, while attractive in terms of petroleum

reduction, would have led to packaging large storage tanks and

adding vehicle mass; in addition, since no current fueling

infrastructure is present in North America, hydrogen was not

selected. HEVT decided to use stored on-board grid energy to

reduce petroleum consumption. The stored grid energy will not

be able to propel the vehicle for the competition required 322

km (200 mile) range, so an engine capable of running on E85

was selected to further reduce consumption of petroleum

energy and extend the range of the vehicle.

Figure 2 is a representation of total GHG emissions.

One way to lower the total emissions associated with the

vehicle is to select a fuel that has low or zero vehicle emissions,

such as electricity. HEVT decided to use stored on-board grid

energy to utilize the zero vehicle emissions benefit of

electricity. This decision was based primarily on the ability to

meet the large electric range goal set forth by HEVT, as well as

the knowledge that the energy from electricity can be used

much more efficiently on the vehicle than energy from a typical

liquid fuel like E85. Even though the upstream GHG emissions

associated with electricity are the highest of the candidate fuels,

the high efficiency of an electric vehicle helps to reduce the

overall GHG impact. The amount of GHG reduction achieved

is about 30 %, but this number will only increase as the electric

grid becomes more renewable and cellulosic ethanol becomes

more available.


Figure 1: WTW Petroleum Energy Use for the VTREX in both Charge Sustaining (CS) and Charge Depleting (EV) modes

Figure 2: WTW greenhouse gas emissions per kWh of fuel compared to per km at the vehicle for CS and CD (EV) operation

COMPONENT SIZING AND SELECTION

The next step in the vehicle design process for HEVT was

the selection and sizing of the components based on the

analysis performed in the previous sections. The goals

established by HEVT are meant to meet and exceed the

competition goals established by EcoCAR. These goals and

modeling of the stock vehicle are the driving factors towards

the final design of the vehicle. To meet the goal of a large

electric-only range, HEVT chose to design and build a 20 kWh

(15 kWh useable) battery pack with materials and modules

donated by A123 Systems. The team originally partnered with

Danaher Motion, a local company, to build a custom permanent

magnet motor capable of 90 kW to propel the vehicle

electrically. At the beginning of Year 2, Danaher project plans

changed, so HEVT adapted a Siemens AC induction motor into

the vehicle, due to past team experience and availability of the

motor. The Siemens motor has a rating of 67 kW peak but

using a newer inverter and the higher voltage A123 ESS, the

motor is predicted to reach 80 kW peak.

This peak power does not meet the full performance needs

for the aggressive US06 drive cycle in CD mode, but is more

than enough to meet the UDDS (city) and HWFET (highway)

drive cycles. The electric rear traction motor (RTM) has a gear

reduction of 11.585:1 with a top powered speed of 12,500 RPM

(14,000 RPM unpowered). This gearing will still give the

vehicle an estimated top speed of 100 mph before over-

speeding the RTM, and will provide sufficient torque to launch

the vehicle from a stop. HEVT predicts a 0-60 mph time of

about 14 seconds while in CD mode. The ESS will be charged

using an on-board Brusa 3.3 kW charger that can operate on


either 120 or 240 V. A complete recharge of a 75 % state of

charge (SOC) swing (100 – 25 %) will take 5 hours from a 240

V 20 A household circuit. Figure 3 shows a basic energy flow

diagram of the components that HEVT integrated into the

vehicle.

Figure 3: Energy flow diagram of the VTREX

The stock engine was replaced with a 2009 GM LE9

(4-cylinder 2.4 L) FlexFuel engine to power the front axle. By

using a FlexFuel engine burning E85, the team will further

meet the goal of petroleum reduction. The engine will act as

the primary source of propulsion in CS mode, while a 15 kW

peak MES belted alternator starter (BAS) motor load levels the

engine when possible to keep the average engine efficiency

high and directly charges the ESS. Since the powertrain can do

electric launch with the RTM, load level the engine with the

BAS, and restart the engine quickly from an off state with the

BAS, the integration difficulty of installing an available GM 6-

speed transmission was avoided. The stock transmission was

replaced with the GM ME7 4-speed automatic transmission

from a Vue Green Line (36 V hybrid). The transmission has an

auxiliary fluid pump that keeps line pressure high inside the

transmission during engine idle stop to allow smooth, quick

takeoff as soon as the engine is restarted. These components

acting together produce a zero to sixty time of 8.8 s in CS mode

and a fleet utility factor-weighted combined fuel economy of

2.9 l(ge)/ 100 km (82 mpgge). Table 3 shows an overview of

the components on board the VTREX.

Since there are large amounts of time when the vehicle

operates with the engine off, a high voltage electric air

conditioning (A/C) compressor is installed on the vehicle to

keep passengers comfortable regardless of propulsion mode.

HEVT has a plan to integrate an electric resistance heater in

Year 3to allow for heating the passenger compartment during

all-electric operation. The DC/DC converter that came with the

GM-donated vehicle provides power for the 12V system

directly from the high voltage A123 battery pack. This

converter is a requirement for CD mode and CS mode as there

is no traditional 13.8V alternator driven by the engine.

Table 3: VTREX Component Specifications

Architecture E85 Range Extended Split

Parallel Architecture

Components Size Type

Engine 130

kW

peak

GM LE9 2.4 L

ECOTEC VVT DOHC

16V I4 FlexFuel SI

Transmission - ME7 4-speed automatic

FWD

Belted

Alternator

Starter

8 kW

cont.

MES 150-125W

Rear Traction

Motor

80 kW

peak

Ballard/Siemens Ranger

EV Transaxle

Energy

Storage

360 V

A123 custom prismatic

built pack, 15 kWh

useable

12 V Supply 1 kW

cont.

DC/DC Converter from

360 V to 13.8 V

A/C System 10 kW HV electric drive, high

efficiency, variable

speed

ESS Charger 3.3 kW

cont.

Brusa 120/240 V 50/60

Hz AC, integrated on-

board

Controls - NI CompactRIO with

FPGA

POWERTRAIN INTEGRATION

This section focuses on the integration progress and

challenges which HEVT ran into with the energy storage

system, the rear powertrain, and the front powertrain. A full

CAD model of the vehicle can be seen in Appendix B.

Seven waivers were submitted by the team to the

EcoCAR technical steering committee for vehicle revisions to

the VTREX. All of these have been approved and the designs

implemented into the vehicle. The first waiver was for the rear

subframe, which was needed for mounting the RTM. The

RTM also intersected the factory rear floor pan, so a waiver

was written to install a cross member to keep the structural

integrity of the vehicle frame with a portion of the floor pan

removed. Since the team built its own ESS case, it was needed

to prove that the modules would remain in place during a crash.

The load floor and the stock D-ring cargo tie down locations

had to change so that the ESS could fit between the rear fender

wells. A waiver was submitted to prove that mounting the

emission test equipment to the revised location would still be

possible.

HEVT “Split Parallel” Plug-In Hybrid Vehicle

2.4 L Engine

E85

FuelMechanicalElectrical

~Charger

80 kW Rear Traction Motor

(RTM)

BAS

20 kWh Lithium Ion

Battery Pack

Grid Energy

Front


Other electrical waivers were submitted as well, in

addition to the mechanical ones listed above. The BAS motor

came from the factory with one external wiring conduit for both

high voltage and low voltage wiring to drive the AC induction

motor. Running high and low voltage in the same conduit is a

violation of the competition rules, but was allowed as the motor

came this way from the factory. The LE9 ECM and the ME7

TCM require a keep alive memory circuit to remember the

ethanol content of the fuel currently in the gas tank. The entire

12 V system is supposed to be disabled during shipping, but a

memory circuit was allowed by the EcoCAR organizers. The

hydraulic braking system in the vehicle was changed for

integration and pedal feel reasons, and a waiver for replacing

the yaw sensor was submitted, in order to have a functional

anti-lock braking system (ABS).

ENERGY STORAGE SUBSYSTEM

The VTREX uses a custom built battery pack consisting

of 5 prismatic modules from A123 Systems as shown in Figure

4. CD mode, the main mode of operation, will use the battery

to supply energy to propel the vehicle. The HVSC allows the

VTREX to capture regenerative braking energy from the RTM

and uses the BAS to load level the engine and increase the

average engine efficiency over a drive cycle. These functions

all require use of the high voltage ESS (battery pack) and thus

generate heat from losses associated with energy conversion.

All control decisions for requesting power are made by the

HVSC and power requests can be limited based on component

temperature and the current operational safety level (OSL) that

each component is reporting to the cRIO. More information

about the control logic and design considerations will be

explained in later sections.

Because the team's hybrid design is based on an E-REV

architecture, there are two distinct operational modes that must

be considered: CS and CD. CD mode will be the most

thermally demanding of the two modes since the battery will

provide all propulsive energy for the vehicle. To design a

simple lightweight pack, passive cooling was selected after an

analysis of the battery modules and their operational

temperature range. The prismatic modules are cooled by

conducting heat through the base plate of the battery pack and

convectively removing the heat to the ambient air passing under

the vehicle. Based on information from A123 Systems, the

module temperatures will only increase 15 0C for a full

discharge test at rated current of 180 A. The average current

demand of less than 60 A RMS for propelling the VTREX in CD

mode in combined cycle driving is much lower, resulting in

only a 6 0C rise in the thermal mass of the modules. CS mode

average current demand of less than 20 A RMS will still

generate a small amount of heat in the pack due to regenerative

braking and engine load leveling. While this design would not

be chosen for a production-intent vehicle, it will meet the

requirements of the EcoCAR competition to be held in Yuma,

Arizona in May of 2010. Steps are taken in the control strategy

to reduce the use of the ESS if the module temperatures start to

reach temperatures above 50 0C and will open the contactors at

60 0C for safety.

Figure 4: HEVT ESS installed into the vehicle

REAR POWERTRAIN INTEGRATION

HEVT packaged an electric rear traction motor (RTM)

with integrated transaxle which is isolation mounted directly to

a custom rear subframe. The RTM is positioned to sit as low in

the subframe as possible to decrease the angle in the CV joints

and clear the battery pack which is mounted above the RTM.

Packaging the RTM in the subframe caused a spatial conflict

with the exhaust pipe, which runs beneath the rear subframe

and the fuel tank. HEVT changed the routing of the exhaust to

alleviate this issue and provide ample clearance underneath the

vehicle, as shown in Figure 4. The team had a local muffler

shop re-route the exhaust so that the entire system is heat

shielded, has sufficient ground clearance, and has solid welds.

The front cross member of the revised subframe intersects the

stock fuel tank. TriFab Inc., a fabrication company in

Pennsylvania, was contracted to build a revised fuel tank,

CUSTOM FUEL TANK

RANGER EV REAR TRACTION MOTOR

RINEHART INVERTER FOR RTM

CUSTOM REAR SUBFRAME

EXHAUST LINE

FRAMERAIL

COMPACT RIO

BATTERYCHARGER

BATTERYPACK

CUSTOM CARGO FLOOR

STOCK FLOOR LEVEL

7” ABOVE STOCK FLOOR LEVEL


designed by HEVT, which has a reduced capacity of 11 gallons

and allows the tank to be removed during competition without

removing the exhaust.

Figure 5: View of the rear powertrain

A large amount of design and engineering went into

the redesign of the custom subframe for packaging the RTM.

To accommodate the motor, the subframe was modified in

several respects according to a waiver that was approved by the

EcoCAR steering committee. The new subframe design is

shown in Figure 6, with the RTM mounted in the subframe.

The front cross member was cut vertically, parallel to the

vertical portion of the front cross member, and was replaced by

tubular steel with a cross section of 4" x 2" and a wall thickness

of 1/8" (3.175 mm). Similarly, the rear cross member was also

cut and replaced by tube of the same size. These suggestions

for manufacturability came from TriFab, which was also

contracted to build the subframe. The key benefit of the

redesigned subframe is that it successfully integrates the motor

onto the rear axle while maintaining all stock suspension

geometry and mounting points.

Figure 6: Rear traction motor subframe design and implementation

As per the EcoCAR competition rules, HEVT

performed finite element analysis on both the stock GM

subframe and the redesigned subframe. HEVT designed a

subframe that has a factor of safety of 1.51 in comparison to the

GM subframe, based on a comparative analysis of matching

loads and material strength. The image on to in Figure 7 shows

the direction of the loads applied to the subframe. Then the

HEVT subframe design was modeled under the same loads,

with the criteria that the maximum stress in the entire frame

must be two thirds of the maximum stress in the GM subframe.

In order to keep the number of elements and nodes to a

minimum, a symmetrical half model was used. The critical

section was found to be the A-arm mounts as seen in the lower

picture in Figure 6 by the location shown in red.

CUSTOM FUEL TANK

EXHAUST LINE

RANGER EV REAR TRACTION MOTOR

RINEHART INVERTER FOR RTM

CUSTOM REAR SUBFRAMEEVAPORATIVE EMISSIONS CANISTER


Figure 7: Load case and results of a finite element ½ model for a

rear subframe

FRONT POWERTRAIN INTEGRATION

As Figure 8 shows, the stock V6 engine and 2-Mode

Hybrid transmission were removed and replaced with a 4-

cylinder LE9 FlexFuel engine and 4-speed automatic

transmission. HEVT purchased and installed a front cradle from

a Vue Green Line which was designed for the same family of

engine/transmission used in the vehicle design. A belted

alternator starter (BAS) is mounted to the engine in the space

normally occupied by the belted A/C compressor. The 2-Mode

traction power inverter module (TPIM) box, located in its stock

position, houses the BAS inverter along with the fuses for HV

distribution in the front of the vehicle. The DC/DC converter

from the 2-Mode is in the factory original position below the

TPIM box. The HVAC compressor from the 2-Mode system

replaces the LE9 belted A/C compressor and is mounted above

the electric vacuum pump motor in between the engine and the

passenger side bulkhead. The vacuum pump is needed to

supply vacuum assist to the brake booster since the 2-Mode

system was replaced with a conventional XE system to improve

pedal feel. A spatial issue with the exhaust routing and ground

clearance also helped in the decision to change the brake

system.

Figure 8: Hybrid components in the front of the VTREX.

CONTROL SYSTEM INTEGRATION

The hybrid vehicle supervisory controller (HVSC)

interfaces with all control modules on the high-speed GMLAN

network as well as all control modules added by HEVT. HEVT

is using four different high-speed CAN buses on the vehicle. A

schematic of the VTREX CAN bus is shown in Figure 9. HEVT

ran shielded, twisted-pair wire for the additional CAN buses in

the vehicle and has grounded the shielding only at the

CompactRIO controller (cRIO) to prevent ground loops and

suppress noise.

The first CAN bus interfaces with GMLAN, so that the

controller can interface with the front powertrain and body

control systems. HEVT replaced several control modules on

GMLAN so that the vehicle would recognize the front

powertrain in the VTREX. The LE9 engine sends out some CAN

messages which are understood by the BAS control modules

and some which are not. The RIO therefore acts as a gateway

for CAN message interpretation as well as an HVSC. An

example of a message interpreted by the cRIO is vehicle speed,

HIGH VOLTAGE A/C

EXHAUST

HIGH VOLTAGE DISTRIBUTION

BOX

HIGHVOLTAGE

LINES

BAS


which is needed to drive the speedometer on the vehicle

instrument panel and to make control strategy decisions.

Additionally, in order to have power steering and speedometer

functionality in EV mode, the cRIO sends the appropriate CAN

message to GMLAN to fake the “Engine On” signal; the engine

must be seen as “on” by the PSCM and BCM for power

steering and speedometer functionality. The LE9 ECM was

therefore taken off the GMLAN network, so that conflicting

signals were not sent during EV mode. The second CAN bus

interfaces with the DC/DC converter out of the 2-Mode vehicle

and the LE9 ECM. The RTM and BAS inverters were placed

on one bus to minimize the effects of inverter noise on other

CAN buses. When high voltage is enabled, roughly 5% noise

penetration is seen on the RTM inverter CAN bus by

monitoring error frames in Vector CANoe. The fourth bus is

used by the Brusa Charger and the A123 BCM because the

A123 wiring convention calls for a CAN ground; they must

also be on the same CAN bus in order to perform CAN-

controlled ESS charging.

Figure 9 VTREX CAN Network Diagram.

CONTROL DEVELOPMENT PROCESS

The control strategy responds to component failures in a

way that ensures safety and minimizes damage to the vehicle.

The team is using Design for Failure Modes and Effects

Analysis (DFMEA) and Fault Tree Analysis (FTA) to

investigate methods and causes of failure of all components

controlled through the control code. The real-time hybrid

vehicle model, originally supplied by GM and ANL was

modified by HEVT to represent the VTREX, runs on an NI PXI

chassis against the control code on the cRIO. New iterations of

the control code are tested for functional operation and safety

before being used on the vehicle. Using both software and

hardware fault insertion in NI's Veristand software, the code is

verified to operate as intended, including but not limited to

proper selection of state upon start up, proper transitions of

state, infinite loop checking, and correct mapping of CAN

messages. Shown in Figure 10 on top is the HIL setup at

EcoCAR Winter Workshop, where VT placed first amongst NI

HIL teams. The lower picture is the screen of the vehicle user

interface that gives the driver information about the controller

and the vehicle.

Figure 10: HIL setup (top) and driver user interface (bottom)

After analysis of possible failures and determination of

proper responses, a robust fault handling strategy was

incorporated into the control code to mitigate any faults that

occur and maximize the safety of the vehicle. The HVSC will

respond by warning the driver and, when necessary, shutting

down or limiting the functionality of vehicle components. The

team has developed a two-layer approach to fault mitigation

within the control strategy. The first level of the structure is a

component-based Operational Safety Level (OSL) that is

assigned to each component based on its operability as can be

seen visually in Figure 11. The engine, RTM, BAS, and ESS

have an ideal defined OSL of three, indicating that each

component is fully functional and can be used as intended. If a

failure occurs, each OSL changes to reflect the operability of

each component, based on the severity of the failure. An OSL

of 2 indicates reduced functionality, and an OSL of 1 indicates

inoperability. The control strategy can then make decisions,

based on the OSL of components, for states of operation and

overall vehicle functionality.


1 Component no longer operates

2 Reduced component

performance/functionality

3 No Faults Detected

Figure 11: The three OSL of the fault mitigation strategy

implemented on the HVSC.

The second layer of fault mitigation is based on overall

vehicle functionality and safety. Based on the lowest (worst)

OSL and its effect on the other levels, an overall DEFCON

level is defined for the vehicle. If all functionality is present

and an OSL of three is seen for all components, then the vehicle

is in DEFCON five, indicating all systems are operating as

intended which can be seen in Figure 12. A level of one means

that the vehicle is unsafe and the control strategy will initiate a

vehicle shutdown by opening the high voltage contactor, setting

torque and voltage demands to zero, and removing power to the

engine via the powertrain relay. All other 12V components,

such as the electric12V brake vacuum pump, power steering,

and lights, will remain active to allow the driver to come to a

safe, controlled stop before exiting the vehicle.

1 Walk Home Event

2 Limp Home Event

3 Reduced Vehicle Performance/Functionality

4 Minor Faults – Monitor but continue normal

operation

5 No Faults Detected

Figure 12: The five DEFCON levels of the fault mitigation strategy

implemented on the HVSC.

As the vehicle completed the construction phase,

implementing the controller into the vehicle was the next step.

While functional and safety-critical testing is performed

thoroughly with HIL, safe implementation the controller into

the vehicle is still important to prevent damage from unforeseen

problems. HEVT has a documented procedure of controller

integration with basic logic checks on individual components

before running the entire control strategy. This process

includes ensuring that all signals are read, understood, and

scaled properly. The plan then calls for manually sending

signals to the components to establish two-way communication

via CAN. Testing of communication independent of the

control strategy is important as the control code expects all

signals to be read and scaled correctly. Once communication

with all components is verified, the control strategy can be

tested on the vehicle, starting will each individual mode and

eventually moving on to testing proper transitions between

modes. Currently the vehicle is at Level 1 safety certification

with a push to be Level 2 by competition. Achieving Level 2

means that the control code running in the vehicle is heavily

documented, tested, and has been without uncorrected failure

(i.e. mashing the emergency disconnect switch) for greater than

40 hours.

CONTROL ALGORITHM AND STRATEGY

Choosing the Split-Parallel Architecture design gave

HEVT six operating states to work with in control system

design. From these states, the Controls Subteam chose to

create three modes: charge sustaining (CS), charge depleting

(CD), and normal. The driver will be able to select between

these modes, and normal mode will allow the control system to

decide which mode is best for proper vehicle performance.

Table 4 shows the states possible with HEVT’s selected

architecture. Each state has an explanation of the torque

demand on the major components of the vehicle. Utilizing

these states will enable the performance of the vehicle to meet

the VTS.

Table 4: Vehicle State Explanation

States Engine

Torque

BAS

Torque

RTM

Torque

Engine Only + 0 0

Engine Idle Stop 0 0 0 Engine Generate + - 0

Engine Assist + 0 +

Electric Only 0 0 +

Regen Braking - / 0 - / 0 -

The HVSC interacts with CAN on the vehicle with the

rear traction motor system, belted alternator starter system, the

energy storage system, driver inputs (accelerator pedal, brake

pedal, shifter, and key position) and the DC/DC converter to

select the proper state and mode for the vehicle. The controller

chooses between charge sustaining and charge depleting

modes, and monitors the vehicle to mitigate any faults that may

occur with components on the vehicle. The control strategy is

tuned to meet the VTS set forth by the team. An example of

how HEVT validated the strategy for acceleration is in the next

section. The Controls subteam is using LabVIEW 2009 to

create different operating states for the vehicle. The states

shown in Table 4 are written into LabVIEW StateChart. For

example, in CS mode, as braking occurs during driving,

regenerative braking is activated and the battery is charged if

vehicle conditions permit.


TESTING AND VALIDATION

Currently HEVT is in the process of final integration

and is constantly refining the vehicle control strategy. Further

efforts will be made to compare the predicted VTS to the actual

vehicle VTS for the Year 2 vehicle as the year concludes.

HEVT is able to propel the vehicle with both the RTM and the

FlexFuel engine. At this time the vehicle is still at Level 1 and

is in closed course testing. Figure 13 shows a plot of a recent

drive in the VTREX, and shows vehicle speed and the small dip

in voltage of the ESS during acceleration. The team will soon

validate the acceleration VTS once the vehicle has been

deemed road worthy and eligible for high speed propulsion.

The testing and validation performed so far on the vehicle has

been safety focused to prove that the vehicle is safe to drive.

Areas of interest include testing the operation of the emergency

disconnect switches, proper accelerator pedal response, and

intended direction of propulsion. The Controls subteam is

developing a series of HIL testing stimulus profiles and a relay

simulator module that will automate hardware fault testing. As

competition nears, the team is in the process of documenting

that all of the safety systems are present and functional.

Figure 13: Vehicle speed and battery voltage for a test drive of

the vehicle

YEAR 3, 99% BUY OFF PLANS

All of the major components for vehicle propulsion are

installed in the vehicle which is key for the development of a

mule vehicle (65 % complete) during Year 2 of the EcoCAR

competition. This year the focus on vehicle integration has

meant that brackets for mounting have been made so that the

vehicle can be operational and available for testing. These

brackets are not quite production quality and are a source of

extra mass on the vehicle. Efforts to light-weight the vehicle,

along with better integrating a few subsystems, will help

improve the fuel economy and performance. HEVT will focus

Year 3 primarily on control strategy refinement. An example

of a refinement to be made in Year 3 is to modify the

accelerator pedal request to the engine throttle. Currently this

is a 1 to 1 ratio request but for optimal engine efficiency HEVT

would like to blend in electric energy requests from the RTM to

match the driver intended torque. Another area of interest is

the BAS control strategy; due to component issues, the BAS

will be the last system to become functional. With these

improvements to the vehicle, HEVT hopes to increase the

electric-only range and the charge sustaining fuel economy of

the VTREX.

CONCLUSIONS

HEVT is at the end of Year 2 of the EcoCAR competition

and has built the proposed vehicle design from Year 1 with a

few small component upgrades. Following the VDP as

explained by GM and the EcoCAR organizers, the team

developed a plan in Year 1 for a vehicle architecture that was

implemented in the creation of a mule vehicle. HEVT

established individual team goals that met or exceeded the

competition targets and requirements. After the establishment

of VTS, modeling was performed in order to select

components. The field of components that met HEVT goals

and VTS was narrowed by performance requirements, fuel

selection, and component availability from suppliers and

sponsors. With the components selected, HEVT has moved

into the phase of component integration and vehicle

model/VTS validation. The Controls subteam has created an

HVSC and control strategy that will work towards maximizing

the efficiency of the vehicle while maintaining performance

and drivability.

Year 2 saw a variety of issues arise that were great

learning experiences for the students involved in the project.

This year the team as a whole has excelled in the use of HIL,

working on vehicle integration, and keeping the safety of team

members and observers paramount. Several of the team

members have viewed their experience in EcoCAR as well

worth the extra effort required; close to one third of the team

has chosen to go work in the automotive industry after

competition because of their involvement with EcoCAR. Year

3 will see an increased emphasis on refinement, but will still

focus on safety and the use of HIL for controller development.

HEVT is well on the way to finishing all of the events in Yuma,

Arizona in May 2010. The current state of the vehicle can be

attributed to HEVT following the VDP and advice from other

universities, and working with the EcoCAR organizers. HEVT

feels that the VTREX is a viable solution to sustainable mobility

by using an efficient powertrain and a renewable fuel with an

established fueling infrastructure.

ACKNOWLEDGMENTS The contributions of the research, modelling, and

results from the HEVT design team are gratefully

acknowledged. We would like to thank General Motors, the

350

352

354

356

358

360

362

364

366

368

370

0

5

10

15

20

25

30

35

850 950 1050 1150 1250

ESS

Vo

ltag

e (

V)

Ve

hic

le S

pe

ed

(kM

h)

Time (s)

Veh. Spd (kMh)

Ess Volts


U.S. Department of Energy, ANL, and the rest of the sponsors

of the EcoCAR competition. Finally we would like to thank all

the team members of the 2009-2010 Hybrid Electric Vehicle

Team and local sponsors.

APPENDIX

Appendix A: Complete Vehicle Technical Specifications

Appendix B: CAD isometric view of the VTREX as design by

HEVT

REFERENCES [1] Powertrain Systems Analysis Toolkit (PSAT),

http://www.transportation.anl.gov

/software/PSAT/index.html

[2] Steven Boyd, and Douglas J. Nelson (2008), "Hybrid

Electric Vehicle Control Strategy Based on Power

Loss Calculations”, Paper 2008-01-0084, 2008 SAE

International World Congress, Detroit, Mi, April 14-

17, 2008, 15 pgs.

[3] Johnson, Kurt. “A Plug in Hybrid Electric Vehicle

Loss Model to Compare Well to Wheel Energy Use

From Multiple Sources.” MS Thesis Virginia

Polytechnic Institute and State University, 2008.

[4] Friedman, D., 2003, “A New Road: The

Technology and Potential of Hybrid Electric

Vehicles”, Union of Concerned Scientists,

Cambridge, Mass.

[5] Husted, Harry L., 2003, “A Comparative Study of

the Production Applications of Hybrid Electric

Powertrains,” SAE paper 2003-01-2307.

[6] N. Brinkman, T. Weber, M. Wang, T. Darlington,,

2005, “Well-to-Wheels Analysis of Advanced

Fuel/Vehicle Systems - A North American Study of

Energy Use, Greenhouse Gas Emissions, and Criteria

Pollutant Emissions”, available from ANL.

[7] EcoCAR Rules, Draft 4/1/2010, available from

ANL.


Appendix A: Complete Vehicle Technical Specifications

Table A-1: EcoCAR Competition Requirements

Specification Requirement

EcoCAR Production VUE Competition Team VTS

Accel 0-60 (s) 10.6 s ≤ 14 s 8.8 + 0.5 s*

Accel 50-70 (s) 7.2 s ≤ 10 s 4.8 + 0.5 s*

Towing Capacity

(kg, [lb])**

680 kg (1,500 lb) ≥ 680 kg @ 3.5%, 20

min @ 72 kph (45

mph)**

1588 kg (3500 lbs) @

3.5%, 20 min @ 72 kph

(45 mph)**

Cargo Capacity

(m3, [f

3])

.83 m3 Height: 457 mm (18”)

Depth: 686 mm (27”)

Width: 762 mm (30’”)

Height: 508 mm (20”)

Depth: 885 mm (34")

Width: 965 mm (38")

Passenger

Capacity

5 ≥ 4 5

Braking 60 – 0

(m, [ft])

38 m- 43 m

(123 -140 ft)

< 51.8 m

(170 ft)

43 + 5 m Δ,*

(140 + 15 ft)

Mass (kg, [lb]) 1,758 kg

(3,875 lb)

≤ 2,268 kg*

(5,000 lb)

2200 + 50 kg

(4850 + 110 lb)

Starting Time (s) ≤ 2 s ≤ 15 s 10 + 3 s

Ground Clearance

(mm, [in])

198 mm

(7.8 in)

≥ 178 mm

(7 in)

165 mm

(6.5 in)

Range (km, [mi]) > 580 km

(360 mi)

≥ 320 km

(200 mi)

340 + 20 km Δ,*

(210 + 10 mi) * Depending on specific test conditions** State maximum towed mass at 3.5% grade, 45 mph, 20 min (i.e. 750 kg). Vehicles must reach the target 45 mph speed in 30 seconds or less. ***Test weight is

nominally 2340 kg Δ E&EC test weight is 2450 kg (this is 2 people and Semtech equipment)

Table A-2: EcoCAR Competition Targets

EcoCAR Production

VUE XE

Competition

Target

Team VTS

Fuel Consumption, CAFE

Unadjusted, Combined, Team:

U.F Weighted (l(ge)/100 km)

8.3 l/100 km

(28.3 mpgge)

7.4 l/100 km

(32 mpgge)

2.9 + .4 l(ge)/100 km Δ,*

(82 + 2 mpgge)

Charge Depleting Fuel

Consumption (l(ge)/100 km)

N/A - 0 l(ge)/100 km

(0 mpgge, all-electric)

Charge Sustaining Fuel

Consumption (l(ge)/100 km)

N/A - 7.8 + .4 l(ge)/100 km Δ,*

(30 + 2 mpgge)

Charge Depleting Range (km) N/A - 68 + 5 km Δ,* (42 + 3 mi)

Petroleum Use U.F Weighted

(kWh/km)

0.85 kWh/km 0.77 kWh/km 0.108 + .02 kWh/km

(0.174 + .01 kWh/mi)

Emissions Tier II Bin 5 Tier II Bin 5 Tier II Bin 5

WTW GHG Emissions U.F

Weighted (g/km)

250 g/km 224 g/km 175 + 5 g/km

(282 + 8 g/mi)


Appendix B: CAD isometric view of the VTREX as design by HEVT

CUSTOM REAR SUBFRAME

CUSTOM FUEL TANK

BATTERY PACKEXHAUST LINE

HIGH VOLTAGE A/C

HIGH VOLTAGE DISTRIBUTION

BOX

RINEHART INVERTER FOR

BAS

BAS

4 CYLINDER 2.4 L LE9 ENGINE

4 SPEED AUTOMATIC TRANSMISSION

EVAPORATIVE EMISSIONS CANISTER

HIGH VOLTAGE WIRING HARNESS

DESIGN AND DEVELOPMENT PROCESS FOR A RANGE …...For the EcoCAR competition, HEVT was supplied...

Documents

Transcript of DESIGN AND DEVELOPMENT PROCESS FOR A RANGE …...For the EcoCAR competition, HEVT was supplied...