Assessment of the predictive capabilities of a multizone predictive combustion model for a prechamber-ignited large bore gas engine

Lean burn gas engines are gaining large attention in the marine sector thanks to the low emissions levels and high efficiencies. However, as the ignitability and combustion stability of leaner mixtures are reduced, ignition systems able to deliver high energy content are necessary. One the most effective approach for large bore marine engines involves the use of an active prechamber in which a small quantity of quasi-stoichiometric air/fuel mixture is obtained and then ignited through a spark plug. The corresponding jet, flowing from prechamber to main chamber, contains many wrinkled turbulent flames and active radicals enhancing the probability of successful ignition. The ability to describe and to model this combustion process has acquired great importance in recent times. In fact, simulation tools have the potential to reduce testing expenses, especially for large bore marine engines, and to support the development of new technologies. In this master thesis, the potential and reliability of the new predictive combustion models available in GT-SUITE were investigated for large bore prechamber-ignited gas engines. Their calibrations were performed against measurement data, which belonged to the Wӓrtsilӓ 6L46-TS SG engine and included pressure recordings from the prechambers, sweeps for load, spark timing, boost pressure and prechamber fuel injection timing and fuel injected quantity. A comparison between different approaches for prechamber and main chamber combustion simulation was performed, in order to assess their predictive capabilities and to highlight advantages and disadvantages of each approach. As a first step, a predictive combustion model was exploited for the main chamber while an optimized Wiebe profile was adopted to describe the combustion process within the prechamber. As a second step, a predictive combustion model was considered also in the prechamber, obtaining a fully predictive simulation approach. Moreover, once a good correspondence between numerical and experimental results was achieved, a NOx emission model was considered and calibrated for both main and prechamber in order to evaluate the contribution of each chamber to the total NOx emissions. Promising results were achieved by exploiting a predictive combustion model, both in main chamber and prechamber. In fact, a good match was obtained between measured and predicted values allowing accurate performance predictions. For instance, combustion parameters such as combustion anchor angle and combustion duration average errors were lower than 2 degrees. With the numerical tools analyzed in the thesis, it was also possible to estimate prechamber’s heat release and pressure profiles, its mixture composition prior and after combustion and the mass flow through its tip holes for a wide variety of operating conditions.