The use of schlieren imaging at high acquisition rate has been adopted as a standard optical technique for the analysis of vaporizing diesel sprays under engine-like conditions. A single-pass schlieren arrangement is typically used for the study of axially drilled single-orifice nozzles, as vessels with multiple optical accesses regularly allow line of sight visualization. Contrarily, for multi-spray nozzles, measurements are commonly performed through a single optical access, in which case a double-pass arrangement is employed. As a consequence, the light beams pass through the test section twice, increasing the optical sensitivity of the schlieren setup. However, the influence this has on the macroscopic spray characteristics is still unclear. The scope of this study is to analyze the differences in vapor phase penetration and spreading angle measured for the same injection event, through high-speed imaging, for both single and double-pass schlieren configurations. Experiments were carried out with a three hole nozzle with a nominal orifice diameter of 90μm, named Spray B from the Engine Combustion Network, using commercially available diesel fuel and in non-reactive conditions. The impact of different injection pressures, chamber temperatures and densities on the spray captured by each setup was assessed. On the results, vapor phase penetration and spreading angle followed the expected trend found in the literature, for the different boundary conditions tested. Comparing the optical setups, vapor phase penetration and spreading angle results obtained with the double-pass arrangement were marginally higher than those from the single-pass. The deviation was observed throughout all tested conditions. For spray tip penetration, although the discrepancy was approximately constant for different injection pressures and chamber temperature, it increased with increasing density. These results highlight the importance of a proper understanding regarding the limitations of optical diagnostics, in particular for results used in calibration of computational models.
Read full abstract