(735g) System-Wide Modeling of HCCI Engine With After-Treatment

System-Wide Modeling of HCCI Engine
with After-Treatment

C N Pratheeba[a] and Preeti
Aghalayam[b]

Department of Chemical Engineering,

Indian Institute of Technology
Madras, Chennai - 600036, India

Automotive engines are synonymous
with power and emissions. The stringent norms imposed by legislations drive the
design of IC engines. Typical engine-out emissions contain unburned fuel, CO,
CO₂, particulate matter and NOx. Many after treatment methods help
in curtailing emissions and adhering to emission norms. Engine design also has
a major role in reducing emissions. With more stringent emission norms, optimal
operation can only be envisaged when the engine and the after-treatment device
are considered in conjunction. This paper focuses on system-wide transient
models for HCCI engines with the after-treatment options being Exhaust Gas
Recirculation (EGR) and Selective Catalytic Reduction (SCR) unit.

EGR refers a part of the exhaust
stream is recycled to the engine along with air-fuel intake¹. EGR
helps in NOx reduction by reducing the peak in-cylinder temperature and air
required for combustion.  The engine-out NOx is then selectively reduced using
NH₃ in a catalytic converter to yield N₂, this is known
as Selective Catalytic Reduction (SCR). This operates at lower temperatures
than its SCNR (Selective Non Catalytic Reduction) counterpart, and thus
presents many advantages.  

In this work, the Engine +
Aftertreatment Device units are to be simulated using the commercial reaction analysis
software called DARS. The engine model chosen to represent HCCI is a Stochastic
Reactor Model in DARS.

A zero dimensional stochastic model
is used for modelling the HCCI engine. The Stochastic model is conceived as a
probability density function (PDF) consisting of a discrete number of
particles. The number of particles is a measure of the accuracy and
computational load.

The mass density function (MDF) is
the discretized form of PDF, contains the user-defined number of particles. The
number of particles is a measure of accuracy and computational load. The
transient MDF is assumed to be a partially stirred tank reactor with the
following equation containing the flow, source and mixing terms²:

The terms in the above equation represent
change due to piston movement, mixing, chemical reaction and heat transfer, respectively.
           

The conservation equations for the SCR³ (including
intrinsic catalytic reaction and external mass transfer resistance) are as
follows:

The above equations are Partial
Differential Equations (PDEs). They are solved in DARS using the operator
splitting technique. The initial conditions that are to be specified for the
engine module are the fuel and the oxidant concentration. The engine physical
parameters and its rpm, Woschni's heat transfer coefficients, the number of
particles for MDF and percentage of EGR are specified. The kinetics used here
is the microkinetic combustion model for           n-heptane developed by Zeuch
et al ⁴ ? consisting of 121 species and 381 reversible reactions.

The SCR model requires the engine-out
exhaust and ammonia dosage to be specified as inlet condtions. The reduced
kinetics for the intrinsic reaction, developed by Tsinoglou and Koltsakis⁵
will be used here.

The solutions of the above HCCI equations
yield the concentration profiles, temperature and pressure profiles as a
function of Crank Angle Displacement (CAD), at engine-out. The profiles are
validated with experimental data with similar engine parameters. The exhaust
profiles with and without EGR are compared, first.

Presented in Figure (1) are the NOx
profiles from the engine out emissions, obtained in our preliminary simulations.
The mole fractions of NOx in the engine-out exhaust are plotted for different
EGR namely 0%, 5% and 10%. The case of No-EGR has significant NOx in the
exhaust stream. Even the 5% EGR reduces NOx considerably. This is due to the
reduction in In-cylinder peak temperature as shown in Table (1). In-cylinder
Peak Pressure variation is insignificant, though a boost compressor can be
added for making up the lost pressure head, if required.⁶

Fig 1. Comparison
of NOx Profiles Vs CAD at various EGR fractions

Table 1: In-cylinder Peak Temperature
and Pressure Vs EGR

In the next step, the transient
engine-out profiles will be used as inlet profiles for the SCR device, and by solving
the SCR equations, profiles of various species of relevance in the tail-pipe
will be evaluated. Independent validations of the Engine & SCR units will
be undertaken as well.

The integrated approach proposed
above has the potential to provide clues for better engine operation and
optimal control of tail-pipe emissions. Since the HCCI engine can be a possible
candidate for achieving very low concentrations of NOx when coupled with after
treatment options, such modeling efforts will have good benefits for the future.

Keywords:

Emissions, Exhaust Gas Recirculation, Selective Catalytic
reduction.

References:

1.      Heywood JB. Internal Combustion
Engine Fundamentals, Mc Graw Hill; 1988.

2.      DARS Manual Book 3 Engine In-cylinder
Models. 2012

3.      Kaisare NS, Lee JH, Fedorov AG. Hydrogen
generation in a reverse flow microreactor: 1. Model formulation and Scaling. AIChEJ.
2005; 51, 8: 2254-2264.

4.      Zeuch T, Moreac G, Ahmed SS, Mauss FA.
Comprehensive skeletal mechanism for the oxidation of n-heptane generated by
chemistry-guided reduction.  Combust Flame 2008; 1555: 651-674.

5.      Tsinoglou D, Koltsakis G. Modelling
of the selective catalytic NOx reduction in diesel exhaust including ammonia
storage. P I Mech Eng D-J Aut. 2007; 221: 117 - 133.

6.      http://www.dieselnet.com/tech/engine_egr_control.php