Application of DARS 1D on in-cylinder

combustion and aftertreatment

Karin Fröjd

STAR Global Conference 2012

Netherlands 20 March 2012

Introduction

Aim: To study the effect of 𝜆 oscillation frequency in an

SI engine on emission conversion in TWC.

Means: 1D reactor network:

In-cylinder combustion Three-way-catalyst (TWC) Piping

CO conversion?

NOx conversion?

HC conversion?

2

All engine

powertrain parts

simulated with

chemical kinetics

Models are stochastic (“SRM”)

(detailed chemistry &

short simulation times)

Species treated

consistently from

in-cylinder

combustion to

tail-pipe emission

Available engine

models: • SI

• DICI

• HCCI

• PPC

DARS 1D

3

Methodology

DARS Basic

Reactors

Homogeneous

Theoretical

CV

CP

PSR

RCM

Engine Reactors

SI (Two-Zones)

HCCI

SRM

(Stochastic Reactor Models)

PSR

SRM-HCCI

SRM-SI

SRM-DICI

SRM PPC

DARS

1D Models

SRM Pipe

Coolers After-

treatment

Catalytic Converter

Diesel Particulate Filter

(DPF)

Turbo charging

DARS-ESM Flames

Premixed

Burner stabilized

Freely propagatin

g

Counter Flow

Diffusion

Back to Back

Flamelet

Single

Library

Chemical Mechanisms

Development

Analysis

Reduction

In-cylinder combustion Three-way-

catalyst (TWC)

Piping

DARS 1D

4

Methodology

In-cylinder combustion Three-way-catalyst (TWC) Piping

1. Run SI SRM in-cylinder combustion model

at 𝜙 = 0.98, 1.00 and 1.02. Get outlet

composition.

2. Make a step function of the outlet

composition from fuel rich, fuel lean and

stoichiometric composition. Impose the

frequencies 1 Hz and 2 Hz, and use this as

inlet to the pipe model.

3. Run the catalyst model with the outlet from

the pipe model as input.

0.975

0.98

0.985

0.99

0.995

1

1.005

1.01

1.015

1.02

1.025

0 0.5 1

Ph

i

Time [sec]

2 Hz

1 Hz

5

Bjerkborn, S. et al., Predictive Flame Propagation Model for Stochastic

Reactor Model Based Engine Simulations, ICDERS 2011, USA

Methodology – in-cylinder combustion

SI SRM:

• Probability function based stochastic

reactor model

• Relevant variables are spread over a

number of virtual particles

500000

1e+06

1.5e+06

2e+06

2.5e+06

3e+06

3.5e+06

-20 -10 0 10 20 30

Pre

ssure

(P

a)

CAD

Phi = 0.77

Phi = 0.77

Phi = 0.86

Phi = 0.86

Phi = 1.00

Phi = 1.00

Phi = 1.20

Phi = 1.20

Phi = 0.77

Phi = 0.86

Phi = 1.00

Phi = 1.20

6

Methodology – in-cylinder combustion

• Chemistry: Skeletal mechanism for PRF (iso-octane / n-heptane)

blends [1] (238 species, 2197 reactions)

• Calculations at 𝜙 = 0.98, 1.00 and 1.02.

• Get gas composition at EVO, use as input to pipe model

[1] Seidel, L., Diplomarbeit, Lehrstuhls Thermodynamik, BTU Cottbus (2009) 7

Results – in-cylinder combustion

Pressure trace for ϕ 0.98, 1.00 and 1.02.

𝝓 = 0.98 𝝓 = 1.00 𝝓 = 1.02

CO 5.35E-4 2.32E-3 6.90E-3

OH 4.48E-5 1.97E-5 6.87E-6

NO 3.91E-3 3.19E-3 2.49E-3

NO2 1.71E-05 1.10E-6 1.75E-7

Outlet mass fractions for ϕ 0.98, 1.00

and 1.02.

118 s (9000 times real time)

for one cycle on 8 CPU’s

8

The pipe is discretized into a number of cells. Each cell is a

stochastic, constant volume PSR.

Chemistry calculated in each stochastic particle

Stochastic mixing and heat transfer to the walls

n-1 n n+1 n-2

k-1 k k+1 k+2

p, v, Yi, hg

Methodology – exhaust piping

9

Methodology – exhaust piping

0.975

0.98

0.985

0.99

0.995

1

1.005

1.01

1.015

1.02

1.025

0 0.2 0.4 0.6 0.8 1

Ph

i

Time [sec]

2 Hz

1 Hz

[1] Seidel, L., Diplomarbeit, Lehrstuhls Thermodynamik, BTU Cottbus (2009)

Applying a step function, using the outlet composition from SI SRM

as inlet composition

Chemistry: Skeletal mechanism for PRF (iso-octane / n-heptane)

blends [1] (238 species, 2197 reactions) - same as for in-cylinder

combustion calculation

Pipe length: 1 m

Pipe diameter: 5 cm

Cell length = 10 cm

20 particles per cell

10

Emissions are not converted

Radicals are converted

Under fuel lean conditions NO is

partially oxidized to NO2

Results – exhaust piping

ϕ oscillation (green line) and mass fraction CO, inlet and

exhaust flow to/from the pipe, for ϕ oscillation of 2 Hz.

Inlet Exhaust Conversion

CO 6.90E-03 6.88E-03 2.4 %

OH 4.48E-05 1.67E-08 99.96 %

NO 3.91E-03 3.89E-03 0.51 %

NO2 1.71E-05 4.75E-05 -177 %

Maximum value of species mass fraction

and conversion in pipe, 2Hz ϕ oscillation.

1000 times slower than

real time on 8 CPU’s

11

Transient 1D catalyst model

Three level solution:

Reactor – conductive heat transfer

Channel – flow and gas phase

chemistry

Washcoat – surface chemistry

The effect of C3H6 inhibition on NO reduction for lean

exhaust gases in a Pt-ϒ-Alumina catalyst, 250 ˚C

Methodology – Three way catalyst

12

Detailed and/or

global reaction

kinetics can be

applied

Chemistry: A global TWC chemistry with oxygen storage [1] (9

gas phase species, 2 surface species, 9 global reactions + 6

surface reactions for oxygen storage)

Unstable unburned hydrocarbons are lumped to the reference

species C3H6

Stable hydrocarbons are lumped to the reference species C3H8.

Catalyst temperature = 850 K

Exhaust gas temperature = 700 K

Catalyst length = 10 cm

Cell length = 0.5 cm

Methodology – Three way catalyst

[1] Rao, S. K. , Imam, R., Ramanathan, K., Pushpavanam, S.,

Ind. Eng. Chem. Res. 2009, 48, 3779–3790 13

TWC inlet & outlet CO and NO mass fractions, at

Φ oscillation of 1 and 2 Hz

Significant difference between

conversion at 1 and 2 Hz

Overall conversion rate for

CO is 54 % higher at 2Hz -

oxygen storage

Results – Three way catalyst

1 Hz 2 Hz

CO 44 % 68 %

NO 49 % 49 %

HC 85 % 84 %

Emission conversion in TWC

240 times slower than

real time on 1 CPU

14

Conclusions

DARS 1D can be used to study the effect of 𝜆 variations

on TWC conversion performance

DARS 1D can be used to in a consistent way study

chemical effects in in-cylinder combustion, piping and

aftertreatment systems

Radicals are converted and NO is partially oxidized in

the exhaust piping

TWC performance is highly dependent on 𝜆 oscillation

frequency

Total time for the project: ~1-2 man weeks

15

Thank you for your attention!

kfrojd@loge.se

www.loge.se