CIMAC Congress - Schiff & Hafen
CIMAC Congress - Schiff & Hafen
CIMAC Congress - Schiff & Hafen
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>CIMAC</strong> CONGRESS | BERGEN 2010<br />
are tested combined with both high and low engine loads. The<br />
same setup with the highest rail pressure is used for all the run<br />
loads. The results show that there is a clear dependency between<br />
the injection parameters and the engine performance. An<br />
optimization may be possible but the overall view of all the main<br />
engine outcomes has to be taken into account.<br />
Predictive simulation and experimental<br />
validation of phenomenological<br />
combustion and pollutant models for<br />
medium-speed common rail diesel engines<br />
at varying inlet conditions<br />
P. Kyrtatos, P. Obrecht, K. Boulouchos, ETH Zürich,<br />
Switzerland,<br />
K. Hoyer, Paul Scherrer Institut, Switzerland<br />
As internal combustion engines are becoming ever more complex,<br />
there is increasing need for engine parameter optimization through<br />
simulation, to avoid numerous timely and costly test-bed<br />
measurements. When performing simulations for engine<br />
performance and emission optimization, the capability of the<br />
combustion model used to accurately predict NOx emission<br />
formation as well as heat release rates at varying engine conditions<br />
becomes increasingly important. Considering the trade-off between<br />
computational cost and accuracy of predictions of diesel engine<br />
combustion and pollutant models, phenomenological models have<br />
a clear advantage compared to their CFD and simple mathematical<br />
approximation alternatives. The detailed phenomenological model<br />
used in this study is able to capture changes in fuel injection system<br />
and charge-air thermal and chemical properties for direct injection<br />
diesel motors, while being computationally efficient. This paper<br />
aims to show the ability of these models to predict diesel combustion<br />
and emission formation during significantly varying inlet charge<br />
and injection conditions, in common rail medium-speed diesel<br />
engines. Initially the phenomenological models are calibrated using<br />
measurement data from a production common rail medium-speed<br />
Wärtsilä 6L20CR diesel engine, employing a state-of-the-art<br />
turbocharging system. The model calibration includes data from<br />
experiments where injection timing and pressure as well as engine<br />
load were varied, to determine their influence on combustion and<br />
NOx emissions. The models are then used to predict the heat release<br />
rate and NOx formation when the inlet valve timing is changed to<br />
earlier Miller timing and the charge air pressure is raised using twostage<br />
turbocharging. Additionally, the models are embedded in a<br />
1-D simulation model of the engine to predict the resulting engine<br />
performance. The simulation results are compared with experimental<br />
results obtained from the test engine with matching hardware<br />
changes, giving an indication of the models’ ability to capture the<br />
most important combustion and emission formation characteristics.<br />
Results from the study show very good performance of the<br />
combustion and emission models, when used to perform operating<br />
map-wide simulations with varying fuel injection conditions. When<br />
the models are used to predict heat release rate in the two-stage<br />
turbocharged engine with Miller timing, the combustion rate is<br />
predicted well, with small discrepancies in ignition delay calculation.<br />
The emission model correctly forecasts the reduction in NOx<br />
emissions as a result of the advanced Miller valve timing, but<br />
underestimates the true level of NOx produced. Overall, the<br />
combustion and emission models show good performance, and<br />
their short calculation time allows them to be used for multi-variable<br />
engine optimization within the calibration ranges. With<br />
improvements in the ignition delay and NOx calculation, the<br />
models can additionally be used for preliminary engine concept<br />
design studies and turbocharger matching through simulation.<br />
Emission reduction potential of 3000 bar<br />
common rail injection and development<br />
trends<br />
S. Pflaum, J. Wloka, G. Wachtmeister, Technical<br />
University of Munich, Germany<br />
Due to the introduction of new emission limits, engine developers<br />
are forced to optimize both, combustion process and peripheral<br />
equipment of diesel engines. For this purpose peripherical systems<br />
like the cooling system, the exhaust-gas-recirculationsystem (EGR)<br />
and the injection system play a major role. Beside cooling- and EGRsystem<br />
highly affecting the generation of nitrogen oxide (NOx)<br />
emissions, the soot production is primarily influenced by the<br />
injection system. To investigate and develop new combustion<br />
processes the Chair of Internal Combustion Engines (LVK) at the<br />
Technische Universitaet Muenchen (TUM) developed a novel singlecylinder-research-engine,<br />
equipped with a special EGR system and a<br />
3000 bar common rail system. This common rail system based on a<br />
standard 1800 bar system had to be adapted and redeveloped for<br />
the extremely high injection pressures. The LVK-Research-Engine<br />
itself was build for combustion pressures up to 300 bar, which is<br />
high above series. The first engine tests with the new 3000 bar<br />
injection system showed a great correlation between the injection<br />
pressure and the emission. As the extremely high injection pressures<br />
in combination with standard injector nozzles (designed for<br />
1800 bar) did not yet produce satisfying emission results, the LVK<br />
started to adapt and design new nozzles for this high injection<br />
pressures. The development process of the new highpressure-nozzles<br />
is based on Computional Fluid Dynamic (CFD) calculations, which<br />
show the diesel flow in the injector nozzle holes. After adapting the<br />
calculations to the new high-pressure range, the influence of the<br />
different geometry parameters, like nozzle-hole number, diameter,<br />
conicity and degree of hydro erosive (HE)-rounding of the nozzle<br />
holes was studied. With consideration of the needs of a low-emission<br />
combustion process, the new, adapted nozzles for high pressures<br />
were designed by CFD and manufactured by drilling and HErounding.<br />
The high-pressure-nozzles were mounted in the LVK-<br />
Research-Engine for further investigations. The emission behaviour<br />
of the new nozzles was tested and validated in the research engine.<br />
With the new high-pressure-nozzles remarkable good emissionresults<br />
could be achieved. The 3000 bar common rail system with<br />
the new CFD-optimized nozzles showes big potential to comply<br />
with EURO VI<br />
in a distinctive area of the engine map. Beside the described<br />
development process the paper will discuss engine development<br />
trends concerning also costs, lifetime and potentials, like reduction<br />
of emission and fuel consumption by applying new, adapted highpressure-injection-systems<br />
(up to 3000 bar and above). In addition<br />
some future visions will be presented.<br />
NOx emission reduction by use of N 2<br />
diluted charge air<br />
O. Melhus, I. J. Garasen, B. Haukebo, K. K. Langnes,<br />
Ecoxy AS, Norway,<br />
D. J. Stookey, Compact Membrane Systems, Inc.,<br />
USA,<br />
J. E. Hustad, Norwegian University of Science and<br />
Technology (NTNU)<br />
In the years from 2004 to 2009 Ecoxy has tested three different ways<br />
to dilute the charge air for NOx reducing purposes. EGR is a wellknown<br />
measure to reduce NOx from diesel engines and has been<br />
extensively used for automotive diesel engines for a number of years.<br />
For marine diesel engines, which run on totally different types of<br />
52<br />
Ship & Offshore | 2010 | No. 3