Politecnico di Torino (logo)

Calibration and validation of the GT-Power SITurb predictive combustion model for a spark-ignition CNG heavy duty engine

Carlo Alessandro

Calibration and validation of the GT-Power SITurb predictive combustion model for a spark-ignition CNG heavy duty engine.

Rel. Alessandro Ferrari, Oscar Vento, Omar Marello. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Meccanica, 2022


This thesis is the result of a work carried by the candidate within a collaboration project between the Energy department of Politecnico di Torino, and FPT Industrial. The focus of the project is on the renewal of a FPT Heavy Duty Spark Ignition engine concept, evaluating its performance as the fuel whereby it is supplied is switched from compressed natural gas to alternative fuels. Developed in the preliminary phase of the project, the present work deals with the calibration of a 0D entrainment-based combustion model, directly implemented on GT-Power commercial software. For such a purpose, FPT industrial provided a detailed engine model that initially implemented a Wiebe function to model combustion process, calibrated upon 42 engine points spacing 9 ranges of engine revolution speed, and featuring Break Mean Effective Pressure and Spark Advance sweeps for each speed. In-cylinder pressure traces have also been provided for 696 cases of study, together with an i-file that collected other useful measured data. The implemented combustion model, namely EngCylCombSITurb combustion, is considered predictive since it allows to determine in-cylinder pressure and temperature histories from mathematical analysis, in the range of operating conditions upon which it has been calibrated: starting from the mechanism of flame propagation within the turbulent combustion regime encountered in the majority of SI engines, a "burning law" correlation based upon the governing combustion equations is obtained. On the one hand, to accurately capture the physics of the intake and combustion processes, and the influence that turbulence has on these aspects of the engine cycle, a 3D Computational Fluid Dynamics (CFD) analysis should be carried; this approach is certainly the most accurate. On the other hand, 3D CFD analyses are restricted to single components, and they suffer due to increased setup time of the model, in addition to the increased computational time to run the simulation. Moreover, SITurb represents a better trade-off between these aspects than the already implemented Wiebe model, and it gives improvements in terms of prediction capabilities with respect to imposed combustion law model: for the latter, a set of 4 coefficients must be specified for each operating point of the engine to reproduce the thermodynamic aspects of the system during its evolution; the former requires instead a more considerable effort due to calibration and setup, but the tuning constants are fixed for all cases. The initial step of the calibration process consists in a cylinder pressure analysis: a thermodynamic approach is used to obtain a mass burn rate profile for the cylinder charge (air and fuel mixture), starting from the measured pressure traces. Thanks to the obtained combustion profiles, consistency has been studied to determine whether the engine points showed a correspondence between the simulated data and the experimental data or not: consistency criteria has allowed to manually tune the heat transfer model and to evaluate those cases that cannot be used for further calibration. Finally, the combustion model has been calibrated and validated thanks to the GT-Power "design optimizer" with an optimization tool based on a genetic algorithm, that resulted in the choice of an optimal value for the model constants. Further studies will be carried by the research group to investigate the combustion process more deeply, and the effects that the gas intake process and turbulence have on engine performance.

Relators: Alessandro Ferrari, Oscar Vento, Omar Marello
Academic year: 2022/23
Publication type: Electronic
Number of Pages: 125
Additional Information: Tesi secretata. Fulltext non presente
Corso di laurea: Corso di laurea magistrale in Ingegneria Meccanica
Classe di laurea: New organization > Master science > LM-33 - MECHANICAL ENGINEERING
Aziende collaboratrici: UNSPECIFIED
URI: http://webthesis.biblio.polito.it/id/eprint/25698
Modify record (reserved for operators) Modify record (reserved for operators)