To understand the influence of the orifice design on the spray spreading angle in conventional direct injection diesel engines, measurements were carried out on a high-pressure high-temperature chamber using three different nozzles and diesel fuel (DIN-EN-590). The first nozzle is a cylindrical one, the other two are conical (K-factor 2.7 and 5.7). Unexpectedly, the nozzle with the highest conicity factor exhibits the largest cone angle. The other two nozzles have nearly a similar value of the angle despite the different geometry.

A CFD-study showed that the turbulent kinetic energy of the conical holes internal flow is smaller than that of the cylindrical one. Furthermore, cavitation occurs within the cylindrical hole, which induces the break-up and enlarges the cone angle. Further investigations were carried out using seven different cone angle models from the literature. The comparison with the experimental values indicated that none of the used models was able to capture the measured trend of the spray dispersion because they do not consider the impact of the modern nozzle geometry parameters on the jet spreading angle.

From this background, a new model was developed to determine the spray cone angle of direct injection diesel engines, taking into account the actual nozzle design. The model distinguishes between three different effects of the nozzle geometry on the spray dispersion. Considering a cylindrical orifice with sharp edges, tapering the hole and rounding the inlet edges will lead to a decrease in turbulence intensity and cavitation probability and hence the spray cone angle, and on the other hand, to an increase in the nozzle discharge coefficient and the mean velocity at the orifice exit. The last one induces the aerodynamic interaction between the jet and the surrounding gas and hence the spray dispersion. Therefore, a relationship between the change of spray dispersion and discharge coefficient was assumed. The third effect is caused by the velocity profile relaxation which occurs under the influence of the viscous forces i.e. the velocity gradient after leaving the orifice and its wall boundary conditions. This enhances the break-up in the outer boundary of the jet. Reducing the flow losses causes a steeper velocity gradient in the near wall zone and thus increases the spray dispersion. This increase in the cone angle was assumed to be a function of the orifice conicity factor. The validation of the new model with experiments showed a good agreement with the measurements.