Large Engines Competence Center · Lucas Eder, M.Sc. CFD Research Engineer Email: [email protected] LEC GmbH Inffeldgasse 19 A-8010 Graz, Austria Phone: +43 (316) 873-30141

LARGE ENGINES COMPETENCE CENTER © LEC GmbH

Dual Fuel Workshop Rostock 3D-CFD model development and validation for dual-fuel combustion 2018-04-26 Slide 1

Large Engines Competence Center

3D-CFD model development and validation for dual-fuel

combustion in large bore engines

L. Eder1, M. Lackner1, H. Winter1, G. Pirker1, A. Wimmer1,2

1LEC GmbH 2Institute of Internal Combustion Engines and Thermodynamics, TU Graz



Agenda

• Introduction

• Dual fuel combustion simulation: challenges

• Description of modeling framework

• Validation of DF combustion simulation with SCE measurement

data

• Conclusions and further steps



Agenda

• Introduction




data




Introduction

• Working principle

• Ignition of a homogeneous natural gas/air

mixture with diesel pilot injection

• Ignition energy provided only by diesel

• Premixed gas/air flame front propagation after

autoignition of the diesel

• Advantages

• Fuel flexibility [7]

• Possibility to operate with different gas

compositions

• Possibility to operate in gas mode but also in pure

diesel mode

• Great reduction in engine-out NOx compared to

monovalent diesel engines [8]

Diesel ignited gas engine concept

Fig.1.: Intake and high pressure phase of the diesel ignited gas engine concept [2]

Gas (Central mixture formation)

Gas (Port fuel injection)

or

Air



Agenda

• Introduction




data




Dual fuel combustion simulation: challenges

Overview

Air/gas induction:➢ Provision of a lean, homogeneous

background mixture➢ Proper representation of the flow field

Diesel pilot injection:➢ Operation of the diesel injector in the

ballistic range➢ Very different behavior than with a fully

open needle

Combustion:➢ All combustion regimes are present➢ Interaction of the two fuels has to be

considered regarding ignition delay andlaminar flame speed

Emissions:

Nitric oxides (NOx)

Soot

Unburned hydrocarbons

Knocking

Premixed

DiffusionAutoignition

Dual Fuel

Diesel ignited gas engine

Air

Gas

Fig.2.: Main challenges of dual fuelcombustion for CFD



Agenda

• Introduction




data




Description of modeling framework

• Model is able to depict the three different

combustion regimes simultaneously

• Multiple fuel option implemented in the

standard ECFM-3Z model

• Several shortcomings had to be addressed

• Ignition delay prediction for dual fuel applications

• Flame surface density deposition

• Flame front propagation

Combustion models – ECFM-3Z

Fig.3.: Standard ECFM-3Z description , taken from [2,3]

Fig.4.: ECFM-3Z mixing model after diesel pilot injection [4]




• Physical effect

• The ignition delay varies with the fuel

mixture fraction [9, 10].

• Strong non-linear behavior of the fuel

mixture fraction has been found with

0D reactor simulations [1,4].

• Numerical approach

• Dedicated tabulation with a dual fuel

mechanism

Improvements for the dual fuel combustion modelling

Ignition delay

Fig.5.: 0D simulations for different fuel fractions (top)Selected ignition delay for one ratio with tabulation points (bottom)




• Physical effect

• After the diesel pilot auto ignites, flame

front propagation starts.

• A flame surface density has to be

provided to the ECFM model as a starting

value


• A simple relation for thermal gas

expansion is taken from [6,7], and

modified:

Σ𝑖𝑛𝑖 = 𝐶 ∗ 𝛻 ǁ𝑐 ∗ 1 + 𝑡𝑖

𝑡𝑖 … turbulent intensity

Necessary modifications to the ECFM-3Z combustion model

Flame surface density deposition

Ignitable mixture

Homogeneous background mixture

Developed flame kernel

Diesel (liquid)

Diesel (gas)

Gas expansion Turbulence

Fig.6.: Initial flame surface density deposition fordual fuel modifications



• Physical effect

• Laminar flame speed varies with the fuel

mixture fraction [6].

• Linear behavior was found from

numerical simulations

[4,5].aaaaaaaaaaa


• Interpolation function:


Improvements for the dual fuel combustion modelling

Laminar flame speed

Fig.7.: 1D simulations carried out for the laminar flamespeed of n-heptane/methane/air mixtures

50

100

150

200

250

300

0 0,2 0,4 0,6 0,8 1

Lam

inar

fla

me

sp

ee

d [

cm/s

]Methane fraction of the dual fuel blend [-]𝝋𝒅𝒇 =

𝝂𝑪𝟕𝑯𝟏𝟔

𝝂𝑪𝑯𝟒+ 𝝂𝑪𝟕𝑯𝟏𝟔

𝒔𝑳𝒅𝒇𝟎 = 𝒔𝑳𝑪𝟕𝑯𝟏𝟔

𝟎 ∗ 𝝋𝒅𝒇 + 𝒔𝑳𝑪𝑯𝟒𝟎 ∗ (𝟏 − 𝝋𝒅𝒇)



Agenda

• Introduction




data




Validation with SCE measurement data

Measurement setup

Engine Conditions

Point Diesel share (energetic) [%] Global Lambda [-] SOC

1 1.5 Lean Early

2 1.5 Lean Middle

3 1.5 Lean Late

4 1.5 Lean Late

5 1.0 Lean Late

6 0.5 Lean Late

Hardware Summary

Rated speed 1500 rpm

Displacement ~ 6 liter

Swirl/tumble ~ 0/0

Charge air Provided by external compressors

Gas fuel supply External mixture formation via venturi mixer

Diesel injector nozzle 4 x 140° Fig.8.: Single Cylinder Research Engine




Simulation setup

Meshing

Type ESE diesel sector mesh (90°) with refinement

No. of cells @ TDC ~ 200.000 (0.6 mm in Spray region)

No. of cells @ BDC ~ 380.000Fig.9.: Sector mesh for engine simulations

General setup

Solver AVL FIRE

Temporal discretization Compression 50 us (0.5 deg @ 1500rpm)

Spray / combustion 10 us (0.1 deg @ 1500rpm)

Expansion 20 us (0.2 deg @ 1500rpm)

Turbulence modeling RANS approach, k-ζ-f turbulence model

Spray Lagrangian approach• WAVE breakup• Ducowicz evaporation• O‘Rourke turbulent dispersion

Model calibration according to spray box tests

Combustion ECFM-3Z modelDual fuel adjustments




Measurement setup

Engine Conditions


1 1.5 Lean Early

2 1.5 Lean Middle

3 1.5 Lean Late

4 1.5 Lean Late

5 1.0 Lean Late

6 0.5 Lean Late

Hardware Summary



Swirl/tumble ~ 0/0







Results of diesel mass variation

Fig.11.: Pressure trace and normalized heat releaserate for the selected diesel mass variations




Measurement setup

Engine Conditions


1 1.5 Lean Early

2 1.5 Lean Middle

3 1.5 Lean Late

4 1.5 Lean Late

5 1.0 Lean Late

6 0.5 Lean Late

Hardware Summary



Swirl/tumble ~ 0/0







Results of early SOC

Fig.13.: ROHR and pressure trace comparison (left)Topdown view into the cylinder with an isosurface of progress variable (right)




Results of middle SOC





Results of late SOC




Agenda

• Introduction




data




Conclusions and further steps

• Conclusions

• Workflow for DF combustion modeling defined

• Dual fuel combustion modeling improvements tested and validated on a single

cylinder research engine

• Optical validation showed good agreement with the simulation

• Further steps

• Generate a tabulated kinetics of ignition (TKI) database [11] for dual fuel mixtures

based on a dedicated dual fuel mechanism

• Generate laminar flame speeds for dual fuel mixtures based on a dedicated dual

fuel mechanism

• Compare ECFM-3Z results with detailed chemistry simulations



Funded by comet

CONTACT:

Lucas Eder, M.Sc. CFD Research Engineer Email: [email protected]

LEC GmbH Inffeldgasse 19 A-8010 Graz, Austria Phone: +43 (316) 873-30141 www.lec.at

Thank you for

your attention!

ACKNOWLEDGEMENT:

The authors would like to acknowledge the financial support of the "COMET - Competence Centres for Excellent Technologies

Programme" of the Austrian Federal Ministry for Transport, Innovation and Technology (BMVIT), the Austrian Federal Ministry of

Science, Research and Economy (BMWFW) and the Provinces of Styria, Tyrol and Vienna for the K1- Centre LEC EvoLET. The

COMET Programme is managed by the Austrian Research Promotion Agency (FFG).

The authors would like to express their gratitude for the measurement data provided by CMT-Motores Térmicos (Universitat

Politècnica de València).

Documents

Large Engines Competence Center · Lucas Eder, M.Sc. CFD Research Engineer Email: [email protected] LEC GmbH Inffeldgasse 19 A-8010 Graz, Austria Phone: +43 (316) 873-30141