Josefsson, G.; Magnusson, I.; Hildenbrand, F.; Schulz, Christof; Sick, V.:
Multidimensional laser diagnostic and numerical analysis of NO formation in a gasoline engine
1998
In: Proc. Combust. Inst., Jg. 27th (1998), Heft Vol. 2, S. 2085 - 2092
Artikel/Aufsatz in Zeitschrift / Fach: Maschinenbau
Titel:
Multidimensional laser diagnostic and numerical analysis of NO formation in a gasoline engine
Autor(in):
Josefsson, G.; Magnusson, I.; Hildenbrand, F.; Schulz, Christof im Online-Personal- und -Vorlesungsverzeichnis LSF anzeigen; Sick, V.
Erscheinungsjahr:
1998
Erschienen in:
Proc. Combust. Inst., Jg. 27th (1998), Heft Vol. 2, S. 2085 - 2092

Abstract:

Laser diagnostic imaging techniques were used to obtain detailed in-cylinder data from a com. gasoline engine. Mean flowfields and turbulence intensities were acquired using particle imaging velocimetry (PIV). Instantaneous quant. NO-concn. fields were measured using planar laser-induced fluorescence (LIF). From exptl. images of NO, flame propagation could also be deduced. Combustion and NO formation were simulated with a 3-D computer code SPEEDSTAR. The overall agreement between exptl. data and computational results is encouraging in general, with remaining issues to be resolved. The mean flow is well predicted, whereas the prediction of turbulence quantities is less satisfactory. Calcd. results for flame propagation are in good agreement with measurements. NO concns. resulting from calcns. are close to those measured, both in respect to their spatial distribution and abs. no. densities. As could be expected, the highest NO concns. are found in regions where combustion started earliest. Local concns. of NO are found be up to 4 times higher than those in the exhaust. The comparison of exptl. results with calcns. clearly shows that, although the 3-D computer model can predict major features of the in-cylinder processes in agreement with measurements, details such as the exact flow pattern and flame development are difficult to capture and depend critically on some of the models parameters used.