Search Results

The combustion process and NOx emission of the world's largest medium speed diesel engine, the Wärtsilä W64L (W64L), were simulated with the modified Kiva2-CFD code. The following submodels were added to the standard Kiva2 code: The Magnussen EDC (Eddy Dissipation Concept) turbulent mixing combustion model, the extended Zeldovich NO model and a simple rebounding model for the droplet wall impingement. The Tesner&Magnussen soot formation and combustion model was implemented in the code, but it has not yet been used in this stage. The modified TAB model was used for the spray. The simulation results, such as the cylinder pressure, the heat release rate and the cumulative heat release were compared to the measured values of the W64L engine. The spray model correctness was tested by comparing the predicted spray tip penetration with Hiroyasu's correlation at room temperature conditions.

With the tightening requirements on engine emission and performance, pressure based combustion controls are becoming common in medium speed large bore reciprocating internal combustion engines. The accuracy of the cylinder pressure data including the raw pressure value at its corresponding crank angle, has a vital impact on engine controllability. For instance, this work shows that a 1-bar pressure offset leads to a 0.6% variation in the total heat release (THR) while the 50% heat release crank angle (CA50) can be shifted by 1.5 degrees. Similarly, with a single degree error in the crank position, the indicated mean effective pressure (IMEP) gets a 1.8 bar error. Thus, in this work the typical errors for cylinder pressure measurement are reviewed and analyzed for large bore four stroke marine and power plant production engines. The main sources of error for pressure measurement are thermal shock and installation defects.

The purpose of this study is to simulate combustion, soot and NOx-emissions in the world's largest medium speed diesel V-engine, the W64V, based on simulation results of in line-engine W64L. The simulation tool was the modified KIVA2-CFD code to which the following submodels were added: The Magnussen EDC ( Eddy Dissipation Concept ) turbulent mixing combustion model, the extended Zeldovich NO model and a simple rebounding model for the droplet wall impingement. Further the Tesner&Magnussen soot formation and combustion model was implemented in the code and the modified TAB model was used for the spray. The simulation results, such as the cylinder pressure, the heat release rate and the cumulative heat release were compared with the results of a thermodynamic two zone model because the V-engine is under design and therefore measured quantities are not available.