Quantification of diesel injector dribble using 3D reconstruction from x-ray and DBI imaging

[EN] Post-injection dribble is known to lead to incomplete atomisation and combustion due to the release of slow moving, and often surface-bound, liquid fuel after the end of the injection event. This can have a negative effect on engine emissions, performance, and injector durability. To better quantify this phenomenon we present a new image processing approach to quantify the volume and surface area of ligaments produced during the end of injection, for an ECN ‘Spray B’ 3-hole injector. Circular approximation for cross-sections was used to estimate three-dimensional parameters of droplets and ligaments. The image processing consisted in three stages: edge detection, morphological reconstruction, and 3D reconstruction. For the last stage of 3D reconstruction, smooth surfaces were obtained by computation of the alpha shape which represents a bounding volume enveloping a set of 3D points. The object model was verified by calculation of surface area and volume from 2D images of figures with well-known shapes. We show that the object model fits non-spherical droplets and pseudo-cylindrical ligaments reasonably well. We applied our processing approach to datasets generated by different research groups to decouple the effect of gas temperature and pressure on the fuel dribble process. High-speed X-ray phase-contrast images obtained at room temperature conditions (297 K) at the 7-ID beamline of the Advanced Photon Source at Argonne National Laboratory, together with diffused back-illumination (DBI) images captured at a wide range of temperature conditions (293-900 K) by CMT Motores Térmicos, were analysed and compared quantitatively.
This work was supported by the UK’s Engineering and Physical Science Research Council [grants EP/K020528/1 and EP/M009424/1], and BP Formulated Products Technology. The authors acknowledge the support of this work from CMT Motores Térmicos (Universitat Politècnica de València, Spain). Parts of this research were performed at the 7-ID beam line of the Advanced Photon Source at Argonne National Laboratory. Use of the APS is supported by the U.S. Department of Energy (DOE) under Contract No. DEAC02-06CH11357. This research was partially funded by DOE's Vehicle Technologies Program, Office of Energy Efficiency and Renewable Energy. The authors would like to thank Team Leaders Gurpreet Singh and Leo Breton for their support of this work