Piston Bowl Optimization for a Diesel Engine with Variable ... · Piston Bowl Optimization for a...

Piston Bowl Optimization for a Diesel Engine with Variable Compression Ratio

STAR Global Conference 2016

Prague, March 2016

Christian Schramm, Carolus Gruenig, Maximilian Brauer, Matthias Diezemann

Why VCR?

© IAV · 03/2016 · Star Global Conference in Prague · C. Schramm · Piston Bowl Optimization for a VCR Diesel Engine

95 g/km CO2 in 2020

Thermal efficiency

CR

Specific power / CR

downsizing

Low friction CR

powertrain

Real Driving Emissions

• Soot / NOx at high CR

high

• EGR at full load CR

with constant PFP

• HC / CO at low load CR

Cold Test -7°C for

Diesel

• Cold start ability CR

• HC, CO emission CR

• Fuel conversion CR

efficiency

Why Variable Compression Ratio (VCR)?

2

Content


1. Introduction

2. Simulation Approach

3. Simulation Results

4. Testing Results & Validation

5. Summary

3

Content


1. Introduction




5. Summary

4

Articulated

con-rod

Swinging

crank

Linking

gear rack

Linking

con-rod

Crankpin

eccentric

Piston pin

eccentric

Piston

height

adjustment

Con-rod

length

adjustment

Additional

volume

Tilting

cylinder-

block

Lifting

cylinder-

block

Shifting

crankshaft

axis

Introduction / Technical Solutions


Multiple link mechanisms

Alteration of kinematic lengths in the crank train

Repositioning of unmoving parts

All suitable concepts have a variable squish gap height

5

Reduced piston bowl volume for CR = 20 reduces free spray penetration length

Changed top dead center position for CR = 11 leads to different spray targeting

Piston bowl shape needs to be developed for VCR requirements

Introduction / Challenges with VCR

© IAV · 03/2016 · STAR Global Conference in Prague · C. Schramm · Piston Bowl Optimization for a VCR Diesel Engine

Variation Range for VCR

Determined based on GT-Power studies

Selected variation range: CR = 11 … 20

Impact on Combustion Chamber Geometry CR = 20.0

CR = 11.0

Ds Quench = 4.7 mm

Ds Quench = 0.0 mm

CR = 20.0

CR = 11.0

Ds Quench = 4.7 mm

Ds Quench = 0.0 mm

CR = 20 (Part Load) CR = 11 (Full Load)

6

Content


1. Introduction




5. Summary

7

Automated CFD meshing

and model generation of

segment models

Define geometry based on

a parametric piston bowl

shape

Piston bowl shape or

selected parameters are

scaled to fit a predefined

compression ratio

Call pro-STAR / STAR-CD®

Automated Post-

processing: Collecting

simulation results in one txt-

/pdf-file

Simulation Approach


Results File

txt/pdf

Input File

txt

STAR-CD Sector Model

Parametric Bowl

Analyze Results and

Decide new Variants

Manually

(by Engineer)

Automatically (by

Optimization Tool)

CR = const.

8

IAV Tool

Simulation Approach


Automated vs. Manual Approach

Automated Approach Manual Approach

+ Requires less man power

+ Very suitable to find a global optimum

+ Effective usage of computational

resources

- Too much parameters if all geometric

parameters have to be considered

- Useful only for ’fine tuning’ of a

geometry (global optimum)

- Need of more simulation runs to find an

optimum

+ Very suitable to investigate different

bowl layouts

+ Separation of secondary effects

possible / plausibility checks

- Normally the optimum cannot be

detected

- More time-consuming

For the given task the manual approach has been selected

9

Simulation Approach

© IAV · 03/2016 · STAR Global Conference in Prague · C. Schramm · Piston Bowl Optimization for a VCR Diesel Engine 10

1. Step:

Definition of different

piston bowl base

concepts (3)

CR = 20.0

CR = 11.0

Ds Quench = 4.7 mm

Ds Quench = 0.0 mm

e.g. open w-bowl omega

flat dish

open w

r in (mm)

z in

(m

m)

0 10 20 30 40 50 -20

-15

-10

-5

0

5

2. Step:

Optimization of

Piston Bowl Main

Dimensions

3. Step:

Optimization of

Piston bowl features

(collar, cone, …)

Simulation Approach


Operating Conditions

Consideration of 2 operating conditions

12

BM

EP

in b

ar

0

5

10

15

20

25

30

Engine Speed in mm-1

1000 2000 3000 4000 5000

Part Load:

n = 1200 min-1

IMEP ~8.5 bar, CR=20

Part Load:

n = 1200 min-1

IMEP ~8.5 bar, CR=20

Full Load:

n = 4000 min-1

Full Load, CO2-Reduction

Power Increase or

Reduction of Friction

CR=11

Model Setup


Turbulence / Heat Transfer

Turbulence Model k-ε-model, high

Reynolds

Wall Function Kader

Spray

Nozzle model MPI2

Atomization model Huh

Break-up model Reitz +

Submodels

Droplet-Wall-Interaction Bai

Liquid Fuel C12H26 (DF2)

Combustion/Ignition

Ignition model Double Delay

Ignition Model

Combustion model ECFM-3Z

Evaporated Fuel C12H26 (DF2)

Emissions

NO Zeldovich

Soot ERC

Cyclic Symmetry

Boundaries

Injector

BDC TDC

13

CFD, Liquid

Measurement

Maximum (95%)

Average (50%)

Minimum (5%)

Rail Pressure: 800 barInjected Mass: 25 mg

Chamber Pressure: 50 barChb. Temperature: 320 °C

CFD, Vapour

CFD, Liquid Fuel

CFD, Evaporated Fuel

CFD Spray ChamberConti PCR NG Q355, 8 Holes

Ma

ss F

ractio

n [

1]

0.00

0.25

0.50

0.75

1.00

Dro

ple

t D

iam

ete

r [µ

m]

0

50

100

150

200

Time [ms]

0.00 0.25 0.50 0.75 1.00 1.25 1.50

(all CFD)

Inje

ctio

n R

ate

[kg

/s]

0.00

0.01

0.02

0.03

0.04

Pe

ne

tra

tio

n [

mm

]

0

10

20

30

40

50

Co

ne

An

gle

[d

eg

]

0

10

20

30

40

Time [ms]

0.00 0.25 0.50 0.75 1.00 1.25 1.50

CFD, Liquid

Measurement

Maximum (95%)

Average (50%)

Minimum (5%)

Rail Pressure: 800 barInjected Mass: 25 mg

Chamber Pressure: 50 barChb. Temperature: 320 °C

CFD, Vapour

CFD, Liquid Fuel

CFD, Evaporated Fuel

CFD Spray ChamberConti PCR NG Q355, 8 Holes

Ma

ss F

ractio

n [

1]

0.00

0.25

0.50

0.75

1.00

Dro

ple

t D

iam

ete

r [µ

m]

0

50

100

150

200

Time [ms]

0.00 0.25 0.50 0.75 1.00 1.25 1.50

(all CFD)

Inje

ctio

n R

ate

[kg

/s]

0.00

0.01

0.02

0.03

0.04

Pe

ne

tra

tio

n [

mm

]

0

10

20

30

40

50

Co

ne

An

gle

[d

eg

]

0

10

20

30

40

Time [ms]

0.00 0.25 0.50 0.75 1.00 1.25 1.50

Test-rig

CFD

CFD

Measurement / Reference

Engine Speed: 3700 1/minBMEP: max.

CFD CombustionMB OM651 Q355

EKAS: 5%Rail Pressure: 1600 bar, 8 Holes

Crank Angle [°CA]

225.0 270.0 315.0 360.0 405.0 450.0 495.0

Cyl. P

ressu

re [

ba

r]

0

30

60

90

120

150

180

Bu

rn

Ra

te [

J/°

CA

]

0

15

30

45

60

75

90

CFD

Measurement / Reference

Engine Speed: 1200 1/minBMEP: 7.0 bar

CFD CombustionMB OM651 Q355

EKAS: 95% , 8 Holes

Crank Angle [°CA]

225.0 270.0 315.0 360.0 405.0 450.0 495.0

Cyl. P

ressu

re [

ba

r]

0

20

40

60

80

100

120

Bu

rn

Ra

te [

J/°

CA

]

0

20

40

60

80

100

120

Model Calibration


Spray Model Calibration

Combustion Model Calibration

n = 1200 min-1, IMEP ~8.5 bar n = 4000 min-1, Full Load

Test-rig

CFD

Test-rig

CFD

14

Content


1. Introduction




5. Summary

15

Boost Pressure, EGR-Rate,Swirl at IVC, Inj. Fuel Mass,Injection Timing (PI, MI)

base (CR = 16.2)

open wflat dish final w variant

n = 1200 min-1, IMEP ~8.5 barCR = 20

Constant: Variable:

CA50 %IMEP,pCyl,max

omega

n = 4000 min-1, Full LoadCR = 11

Bowl Concepts: Single Variants:

Wall

Heat

Lo

ss i

n (

%)

85

90

95

100

105

110

115

120

125

Wall

Heat

Lo

ss i

n (

%)

65

70

75

80

85

90

95

100

105

110

115

Simulation Results


Piston geometry selected for max. ISFC benefit

Increase of Soot and NO unavoidable with CR=20

CR = 20.0

CR = 11.0

Ds Quench = 4.7 mm

Ds Quench = 0.0 mm

Soot – NO – ISFC – Trade-off

base (CR = 16.2) final w variant

omegaflat dishopen w

Variable:CA50 %,IMEP, pCyl,max

Constant:Boost Pressure, EGR-Rate, Swirl at IVC, Inj. Fuel Mass, Injection Timing (PI, MI)

3D Simulation

Pa

rtic

ula

tes

in %

10

40

70

100

130

160n = 4000 min-1, Full Load, CR = 11

ISFC

in %

90

95

100

105

110

NOx in Comparison to CR = 16.2 in %

90 100 110 120 130 140 150 160

ISFC - 5.8 %

ISFC

in %

90

95

100

105

110


75 80 85 90 95 100 105 110

Pa

rtic

ula

tes

in %

0

100

200

300

400

500

NOx + 37 %; PM + 34 %

PM - 65 %

n = 1200 min-1, IMEP ~8.5 bar, CR = 20

omega - reduced bowl depth

omega - scaled

flat dish

open w

16

Part Load, CR = 20 Full Load, CR = 11





3D Simulation

Par

ticu

late

s in

%

10

40

70

100

130

160n = 4000 min-1, Full Load, CR = 11

ISFC

in %

90

95

100

105

110


90 100 110 120 130 140 150 160

ISFC - 5.8 %

ISFC

in %

90

95

100

105

110


75 80 85 90 95 100 105 110

Par

ticu

late

s in

%

0

100

200

300

400

500

NOx + 37 %; PM + 34 %

PM - 65 %

n = 1200 min-1, IMEP ~8.5 bar, CR = 20

PM -65%

NOx +37%; PM +34%

ISFC -5.8% ISFC -1.1%





3D Simulation

Par

ticu

late

s in

%

10

40

70

100

130

160n = 4000 min-1, Full Load, CR = 11

ISFC

in %

90

95

100

105

110


90 100 110 120 130 140 150 160

ISFC - 5.8 %

ISFC

in %

90

95

100

105

110


75 80 85 90 95 100 105 110

Par

ticu

late

s in

%

0

100

200

300

400

500

NOx + 37 %; PM + 34 %

PM - 65 %

n = 1200 min-1, IMEP ~8.5 bar, CR = 20

Simulation Results


O2-Iso-Surface (kO2 = 5%) min max

Temperature [K]

Part Load (n = 1200 min-1, IMEP ~8.5 bar, 365 deg CA)

Omega-Bowl CR = 20, sSQUISH = 0.9 mm

w-Bowl CR = 20, sSQUISH = 0.9 mm

Flat Bowl CR = 20, sSQUISH = 0.9 mm

Basis CR = 16, sSQUISH = 0.9 mm

17

Omega-Bowl CR = 11, sSQUISH = 5.6 mm

w-Bowl CR = 11, sSQUISH = 5.6 mm

Flat Bowl CR = 11, sSQUISH = 5.6 mm

Basis CR = 16, sSQUISH = 0.9 mm

Full Load (n = 4000 min-1, Full Load, 375 deg CA)

Boost Pressure, EGR-Rate,Swirl at IVC, Inj. Fuel Mass,Injection Timing (PI, MI)

base (CR = 16.2)

open wflat dish final w variant

n = 1200 min-1, IMEP ~8.5 barCR = 20

Constant: Variable:

CA50 %IMEP,pCyl,max

omega

n = 4000 min-1, Full LoadCR = 11

Bowl Concepts: Single Variants:

Wall

Heat

Lo

ss i

n (

%)

85

90

95

100

105

110

115

120

125

Wall

Heat

Lo

ss i

n (

%)

65

70

75

80

85

90

95

100

105

110

115

Simulation Results


Smaller Swirl Number @ SOI due to smaller piston bowl volume

Smaller piston bowl surface potential for decreased wall heat losses

CR = 20.0

CR = 11.0

Ds Quench = 4.7 mm

Ds Quench = 0.0 mm

Swirl and Wall Heat Losses

omega - reduced bowl depth

omega - scaled

flat dish

open w

18


Content


1. Introduction




5. Summary

19

Testing Results & Validation


Test-bench Environment

Single Cylinder Engine (SCE)

Replacement of piston

Variation of CR by inserting washers between

crank train and cylinder block

20

Base piston Optimized Piston

n = 1200 min-1, IMEP ~8.5 bar, CR = 20 n = 4000 min-1, Full Load, CR = 11

Injection

Pilot injection controlw.r.t. heat release

const. Injection Rate

SCE

CFD

EGR

Ext. EGR controled w.r.t.Iso-NOx / Iso-Soot

constant

Fuel Mass

Fuel mass controled w.r.t. IMEP

constant

Change w

.r.t

. B

ase in (

%)

-80

-60

-40

-20

0

20

ISFC Soot NOx

Change w

.r.t

. B

ase in (

%)

-20

0

20

40

60

80

100

ISFC

Soot NOx

Testing Results & Validation


Comparison Test-rig / CFD Results

Testing results confirm CFD predictions

Slight differences in predicted fuel consumption due to different conditions

(testing)

21

-5.8% -4.2%

+34%

+77%

+37% +32%

-1.3% -1.1%

-64% -65%

±0% -5%


Content


1. Introduction




5. Summary

22


Injection



SCE

CFD

EGR


constant

Fuel Mass


constant

Change w

.r.t

. B

ase in (

%)

-80

-60

-40

-20

0

20

ISFC Soot NOx

Change w

.r.t

. B

ase in (

%)

-20

0

20

40

60

80

100

ISFC

Soot NOx

(testing)

-5.8%-4.2%

+34%

+77%

+37%+32%

-1.3%-1.1%

-64% -65%

±0%-5%



Injection



SCE

CFD

EGR


constant

Fuel Mass


constant

Change w

.r.t

. B

ase in (

%)

-80

-60

-40

-20

0

20

ISFC Soot NOx

Change w

.r.t

. B

ase in (

%)

-20

0

20

40

60

80

100

ISFC

Soot NOx

(testing)

-5.8%-4.2%

+34%

+77%

+37%+32%

-1.3%-1.1%

-64% -65%

±0%-5%


Summary


• Almost all technical solutions for a variable

compression ratio (CR) have a variable squish gap

height

• For a CR variation range from 11 to 20 the squish

gap height must be changed by 4.7 mm

Big differences in spray-wall-interaction

• Manual, step-wise optimization of piston bowl design

at 2 operating points

• Piston bowl optimization results:

~5% reduction of ISFC/CO2 at part load

The increase of NO/Soot emissions at part

load caused by the increase of CR could not

be prevented

At rated power slight reduction of ISFC/CO2

(~1.2%) and strong decrease of soot emissions

• In general the simulation results were confirmed by

single cylinder engine

23

n = 1200 min-1, IMEP ~8.5 bar, CR = 20.0 n = 4000 min-1, Full Load, CR = 11.0

SCE (Testing) CFD

Change w

.r.t

. B

ase in (

%)

-70

-60

-50

-40

-30

-20

-10

0ISFC

Soot

NOx

Change w

.r.t

. B

ase in (

%)

-10

0

10

20

30

40

50

60

70

80

ISFC

Soot NOx

Thank You

Christian Schramm

IAV GmbH

Auer Straße 54, 09366 Stollberg (GERMANY)

Phone +49 371 237-34570

[email protected]

www.iav.com

© IAV · 03/2016 · Star Global Conference in Prague · C. Schramm · Piston Bowl Optimization for a VCR Diesel Engine 24

Piston Bowl Optimization for a Diesel Engine with Variable ... · Piston Bowl Optimization for a...

Documents

Transcript of Piston Bowl Optimization for a Diesel Engine with Variable ... · Piston Bowl Optimization for a...