02 cycles ice handout

7/29/2019 02 Cycles Ice Handout

1/7

Vehicle Propulsion Systems

Lecture 2

Fuel Consupmtion Estimation & ICE Powertrains

Lars ErikssonAssociate Professor (Docent)

Vehicular SystemsLinkoping University

October 30, 2012

2 / 5 1

OutlineRepetition

Energy demand cont.Energy demand and recuperationSensitivity Analysis

Forward and Inverse (QSS) Models

IC Engine ModelsNormalized Engine VariablesEngine Efficiency

Gear-Box and Clutch ModelsSelection of Gear RatioGear-Box EfficiencyClutches and Torque Converters

Analysis of IC PowertrainsAverage Operating PointQuasistatic AnalysisSoftware tools

3 / 5 1

W2M Energy Paths

4 / 5 1

Energy System Overview

Primary sources

Different options for on-board energy storage

Powertrain energy conver-sion during driving

Cut at the wheel!

Driving mission has a mini-mum energy requirement.

5 / 5 1

Mechanical Energy Demand of a Cycle

Only the demand from the cycle

The mean tractive force during a cycle

Ftrac =1

xtot

xtot0

max(F(x), 0) dx =1

xtot

ttrac

F(t)v(t)dt

where xtot =tmax

0 v(t)dt.

Note t trac in definition.

Only traction.

Idling not a demand from the cycle.

6 / 5 1

Evaluating the integral

Tractive force from The Vehicle Motion Equation

Ftrac =1

2a Afcd v

2(t) + mv g cr + mv a(t)

Ftrac = Ftrac,a + Ftrac,r + Ftrac,m

Resulting in these sums

Ftrac,a =1

xtot

1

2a Afcd

itrac

v3i h

Ftrac,r =1

xtotmv g cr

itrac

vi h

Ftrac,m =1

xtot

mv itrac

ai vi h

7 / 5 1

Values for cycles

Numerical values for the cycles: {MVEG-95, ECE, EUDC}

Xtrac,a =1

xtotitrac

v3i h = {319, 82.9, 455}

Xtrac,r =1

xtot

itrac

vi h = {0.856, 0.81, 0.88}

Xtrac,m =1

xtot

itrac

ai vi h = {0.101, 0.126, 0.086}

EMVEG-95 Afcd 1.9 104 + mv cr 8.4 10

2 + mv 10 kJ/100km

Tasks in Hand-in assignment8 / 5 1

OutlineRepetition






9 / 5 1


2/7

Energy demand again Recuperation

Previously: Considered energy demand from the cycle.

Now: The cycle can give energy to the vehicle.

Recover the vehicles kinetic energy during driving.

10/51

Perfect recuperation

Mean required forceF = Fa + Fr

Sum over all points

Fa =1

xtot

1

2a Afcd

Ni=1

v3i h

Fr =1

xtotmv g cr

N

i=1

vi h

11/51

Perfect recuperation Numerical values for cycles

Numerical values for MVEG-95, ECE, EUDC

Xa =1

xtot

i

v3i h = {363, 100, 515}

Xr =1

xtot

i

vi h = {1, 1, 1}

EMVEG-95 Afcd 2.2 104 + mv cr 9.81 10

2 kJ/100km

12/51

Comparison of numerical values for cycles

Without recuperation.

Xtrac,a =1

xtot

itrac

v3i h = {319, 82.9, 455}

Xtrac,r =1

xtotitrac

vi h = {0.856, 0.81, 0.88}

Xtrac,m =1

xtot

itrac

ai vi h = {0.101, 0.126, 0.086}

With perfect recuperation

Xa =1

xtot

i

v3i h = {363, 100, 515}

Xr =1

xtot

i

vi h = {1, 1, 1}

13/51

Perfect and no recuperation

Mean force represented as liter Diesel / 100 km.

14/51

Sensitivity Analysis

Cycle energy reqirement (no recuperation)

EMVEG-95 Afcd1.9104+mv cr 8.410

2+mv 10 kJ/100km

Sensitivity analysis

Sp = limp0

EMVEG-95(p+ p) EMVEG-95(p)

/EMVEG-95(p)

p/p

Sp = limp0

EMVEG-95(p+ p) EMVEG-95(p)

p

p

EMVEG-95(p)

Vehicle parameters: Afcd cr

mv

15/51

Sensitivity Analysis

Vehicle mass is the most important parameter.

16/51

Vehicle mass and fuel consumption

17/51


3/7

Realistic Recuperation Devices

18/51

Vehicle Mass and Cycle-Avearged Efficiency

19/51

OutlineRepetition






20/51

Two Approaches for Powertrain Simulation

Dynamic simulation (forward simulation)

Driver Engine Tra nsm. WheelCycle Vehicle

Normal system modeling directionRequires driver model

Quasistatic simulation (inverse simulation)

T ra ns m. E ng in eWheelCycle Vehicle

Reverse system modeling directionFollows driving cycle exactly

Model causality

21/51

Dynamic approach

Drivers input u propagates to the vehicle and the cycle

Drivers input . . . Driving force Losses Vehiclevelocity Feedback to driver model

Available tools (= Standard simulation) can deal witharbitrary powertrain complexity.

22/51

Quasistatic approach

Backward simulation

Driving cycle Losses Driving force Wheel torque Engine (powertrain) torque . . . Fuel consumtion.

Available tools are limited with respect to the powertraincomponents that they can handle. Considering new tools suchas Modelica opens up new possibilities.

See also: Efficient Drive Cycle Simulation, Anders Frobergand Lars Nielsen (2008) . . .

23/51

OutlineRepetition






24/51

Causality and Basic Equations

High level modeling Inputs and outputs

Causalities for Engine Models

ICEICE

Dynamic Approach

e

Te

Pc

e

Te

Pc

Quasistatic Approach

Engine efficiencye =

eTePc

Enthalpy flow of fuel (Power Hfuel = Pc)

Pc = mfqLHV

25/51


4/7

Engine Efficiency Maps

Measured engine efficiency map Used very often

Engine Speed [rpm]

EngineTorque[Nm]

Engine efficiency map

0 1000 2000 3000 4000 5000 60000

50

100

150

200

250

300

350

What to do when map-data isnt available?

26/51

Engine Geometry Definitions

B

l

c

BDC

TDC

L

a

theta

Cylinder, Piston, Connecting rod, Crankshaft

Bore, B

Stroke, S = 2 a

Number of cylinders z

Cylinder swept volume, Vd =B2 S

4

Engine swept volume, Vd = zB2 S

4

Compression ratio rc =VmaxVmin

= Vd+VcVc

27/51

Definition of MEP

See whiteboard.

28/51

Normalized Engine Variables

Mean Piston Speed (Sp = mps = cm):

cm =eS

Mean Effective Pressure (MEP=pme (

N=

nr 2)):

pme =N Te

Vd

Used to: Compare performance for engines of different size Design rules for engine sizing.

At max engine power: cm 17 m/s, pme 1e6 Pa (no turbo) engine size

Connection:Pe = z

16B2 pmecm

29/51

Torque modeling through Willans Line

Measurement data: x: pmf y: pme = BMEP

0 5 10 15 20 25 30 35 405

0

5

10

15

Fuel MEP [bar]

EngineBMEP[bar]

Torque and fuel connection (=1)

Linear (affine) relationship Willans line

pme = e(e) pmf pme,0(e)

Engine efficiency: e = pmepmf

30/51

Engine Efficiency Map Representation

Willans line parameters: e(e) pme,0(e)

31/51

OutlineRepetition






32/51

Causality and Basic Equations

Causalities for Gear-Box Models

GB GB

Dynamic Approach

1 2

T2 T2

21

T1 T1

Quasistatic Approach

Power balance Loss free model

1 = 2, T1 =T2

33/51


5/7

Different Types of Gearboxes

Manual Gear Box

Automatic Gear Box, with torque converter

Automatic Gear Box, with automated clutch

Automatic Gear Box, with dual clutches (DCT)

Continuously variable transmission

34/51

Connections of Importance for Gear Ratio Selection

Vehicle motion equation:

mvd

dtv(t) = Ft

1

2a Afcd v

2(t) mv g cr mv g sin()

Constant speed ddt

v(t) = 0:

Ft =1

2a Afcdv

2(t) + mv g cr + mv g sin()

A given speed v will require power Ftv from the powertrain.

This translates to power at the engine Tee.Changing/selecting gears decouples e and v.

Required tractive force increases with speed.For a fixed gear ratio there is also an increase in requiredengine torque.

35/51

Selection of Gear Ratio

Gear ratio selection connected to the engine map.

Additionally: Also geometric ratio between gears.ig,1ig,2

ig,2ig,3

ig,3ig,4

ig,4ig,5

36/51

Selection of Gear Ratio

Optimizing gear ratio for a certain cycle. Potential to save fuel.

Case study 8.1 (well look at it later).

37/51

Gear-box Efficiency

In traction mode

T2 w = egbT1 e P0,gb(e), T1 e > 0

In engine braking mode (fuel cut)

T1 e = egbT2 w P0,gb(e), , T1 e < 0

38/51

Clutch and Torque Converter Efficiency

Friction clutch torque:

T1,e(t) = T1,gb(t) = T1(t) t

Action and reaction torque in the clutch, no mass.

39/51

Torque Characteristics of a Friction Clutch

Approximation of the maximum torque in a friction clutch

T1,max = sign()

Tb (Tb Ta) e||/0

40/51

Main parameters in a Torque Converter

Input torque at the converter:

T1,e(t) = ((t))h d5p

2e(t)

Converter output torque

T1,gb(t) = ((t)) T1,e(t)

Graph for the speed ratio (t) =gbe

, and

the experimentally determined ((t))

The efficiency intraction modebecomes

tc =gbT1,gbeT1,e

= ()

41/51


6/7

OutlineRepetition






42/51

Average Operating Point Method

Average operating point method

Good agreement for conventional powertrains. Hand-in assignment.

43/51

Quasistatic analysis Layout

More details and better agreement (depends on model quality)Good agreement for general powertrains

Hand-in assignment.

44/51

Quasistatic analysis IC Engine Structure

45/51

Quasistatic analysis Engine Operating Points

46/51

Software tools

Different tools for studying energy consumption in vehiclepropulsion systems

Quasi static Dynamic

QSS (ETH) XAdvisor, AVL X (X)PSAT XCAPSim (VSim) X

Inhouse tools (X) (X)

47/51

PSAT

Argonne national laboratory

48/51

Advisor

49/51


7/7

Advisor

Information from AVL:

The U.S. Department of Energys National Renewable EnergyLaboratory (NREL) first developed ADVISOR in 1994.

Between 1998 and 2003 it was downloaded by more than7,000 individuals, corporations, and universities world-wide.

In early 2003 NREL initiated the commercialisation ofADVISOR through a public solicitation.

AVL responded and was awarded the exclusive rights tolicense and distribute ADVISOR world-wide.

50/51

02 cycles ice handout

Documents