Experimental Investigation of Roughness Effects on Cavitating Flow in Diesel Injectors blank Authors names: C. Mauger1 , L. M´ee` s1 , S. Valette2 and A. Azouzi1 blank

1: Laboratoire de M´ecanique des Fluides et d’Acoustique (LMFA), CNRS UMR5509 – Ecole Centrale de Lyon – INSA de Lyon – Universit´e Claude Bernard - Lyon 1, 36 avenue Guy de Collongue, Ecully, France 2: Laboratoire de Tribologie et Dynamique des Syst`emes (LTDS), CNRS UMR5513 – Ecole Centrale de Lyon – Ecole Nationale d’Ingnieurs de Saint Etienne - Saint Etienne, 36 avenue Guy de Collongue, Ecully, France blank

Corresponding authors: [email protected], [email protected] blank ABSTRACT In diesel injectors, the wall roughness of the nozzle holes is thought to influence cavitation formation during the injection cycle. In order to highlight the roughness effect, an experimental set-up has been designed. This set-up consists of a 2D-orifice that confines a cavitating flow. Specific tools are used to control – by means of femtosecond laser – and finely characterise wall roughness, and the cavitating flow is visualised using optical techniques. The different devices are introduced and the first results are presented.

1. Introduction In modern diesel injectors, cavitation plays an important role in spray formation – in particular, in primary break-up mechanism. The primary break-up mechanism is essential as it determines the quality of the spray. Many studies show how cavitation affects spray formation. In 1959, Bergwerk [1] was a pioneer in showing that cavitation occurs in the nozzle of high pressure atomizers, and impacts spray formation and discharged liquid jet. He assumed that the onset of cavitation in the injector hole may affect the break-up length and the liquid flow at the injector outlet. This assumption has been later confirmed by other authors such as Hiroyasu [2], Soteriou et al. [3] and Tamaki et al. [4, 5]. Over the last decades, several studies have highlighted the influence of various parameters – such as inlet radius [6, 4, 7], hole taper [8, 9, 10, 11], hole roughness [12, 7] and fluid properties [13] – on cavitation inception. Yet, our knowledge about cavitation phenomena in injectors is incomplete and further investigations are needed. a

Pressure sensor

effectiveness must be improved, in particular to predict more precisely how cavitation occurs in the nozzle and how bubbles collapse inside it. The study aims to investigate the roughness effects on cavitating flow in diesel injectors. Such a study is difficult to carry out in industrial configuration, especially because of the high-speed flow in the nozzle, the small measured volume and the lack of optical access inside the nozzle. A transparent 2D-orifice was designed to experimentally investigate how surface roughness affects cavitation behavior. This paper presents the experimental tools used in this study, namely the 2D-orifice set-up, and the devices dedicated to the characterisation and texturation of the wall surface in the orifice. The first data acquisitions are presented.

Flow direction

Window Flat plate sheets

Defect

Window b

Figure 2: SEM visualisations of 2D-orifice geometry Outlet Inlet

2D-orifice

Figure 1: Exploded view of the experimental 2D-orifice set-up The present paper introduces the first steps of a study carried out as part of the NADIA bio (New Advanced Diesel Injection Analysis for bio fuels) project to obtain experimental data to be compared to numerical results. CFD codes are useful tools that enable vehicle manufacturers to design future injectors but their

2. 2D-orifice set-up Orifice design has been thought to allow visualization of cavitating flow (Figure 1a). A transparent material and a 2D flow configuration have been chosen. The 2Dorifice is based on Winklhofer et al. [7] studies with modifications allowing wall roughness characterization and texturation. ISO 4113 oil flows through a rectangular hole (Figure 1b) made of a two-part steel sheet sandwiched between two glass windows. At the 2Dorifice inlet, an abrupt change in cross-section forces the flow to accelerate. This causes a major pressure drop

with high flow stress and cavitation may occur. At the outlet of the 2D-orifice, the flow is discharged in a liquid volume at variable back-pressure. The 2D-orifice is supplied with test oil (ISO 4113) pressurized up to 5 MPa. The flow through the orifice is continuous. Oil pressure levels are measured 40 mm upstream and downstream of the 2D-orifice using metal thin film sensors. A variable area meter measures the flow rate. Oil temperature in the supply duct of the orifice is controlled by a T-type thermocouple. 3. Geometry and roughness characterization A scanning electronic microscope (SEM) is used to assess the 2D-orifice geometry (Figure 2). A 2 mm thick 2D-orifice is used in this study. The orifice is quite straight. It is approximatively 213 µm heigh and the length/hydraulic diameter ratio is above 3.8. Inlet radii are estimated to be about 7 µm and 9 µm. An offset of 17 µm has been observed between the two sheets at the entrance along the flow direction. The surface finish of the rectangular orifice has been grinded. Grinding process leads to little grooves which are perpendicular to the fluid flow (Figure 2). Wall roughness was characterized using a 3D-interferometer. A roughness arithmetic average Ra of 0.71 µm and 0.70 µm have been measured on each steel orifice wall. 4. Wall texturation In order to fine-control the roughness of the 2Dorifice wall, micro-machining by femtosecond laser is used. Ultra-short pulsed lasers are able to produce highquality microstructures in metals. Choosing ad hoc parameters, it is possible to texture micro-patterns with various geometries and sizes [14, 15, 16]. A surface can be totally textured by a repetitive and regular micropattern (Figure 3). The effects of such a micro-patterned surface on cavitating flow are to be investigated. Textured surface

Non textured surface

Figure 3: Example of texturation by femtosecond laser 5. First experimental results Visualizations in the 2D-orifice are made using a Optem zoom 125C optical system coupled with a 10-bit digital CCD camera of 2 048 x 2 048 px2 to obtain high magnification and resolution (1.15 µm.px−1 ). Due to the small aperture angle of such an optical device, a powerful illumination is required. In addition, the lighting duration must be short enough to freeze the flow. An original lighting method is used to visualize the flow inside the

transparent orifice. It consists in producing a whitelight continuum by means of a short (5 ns) and intensive (50 mJ) laser pulse that focuses in ambient air. Artificially colored walls

Cavitation sheets Flow direction Cloud cavitation

Figure 4: Example of cavitating flow in the 2D-orifice with a 4 MPa injection pressure and a 1.47 MPa back-pressure

Figure 4 is an example of instantaneous shadowgraph image. Light areas correspond to liquid phases whereas dark areas correspond to vapor phases. Cavitation sheets appear at the inlet corners and desintegrate into cloud cavitation and bubbles. References [1] W. Bergwerk. Proceedings of the Institution of Mechanical Engineers, 173, (1959). [2] H. Hiroyasu, M. Arai, and M. Shimizu. In ICLASS91, (1991). [3] C. Soteriou, R. Andrews, and M. Smith. In SAE, (1995). [4] N. Tamaki, M. Shimizy, K. Nishida, and H. Hiroyasu. Atomization and Sprays, 8, (1998). [5] N. Tamaki, M. Shimizu, and H. Hiroyasu. Atomization and Sprays, 11, (2001). [6] D. P. Schmidt, C. J. Rutland, and M. L. Corradini. In SAE, (1997). [7] E. Winklhofer, E. Kelz, and A. Morozov. In ICLASS, (2003). [8] E. Winklhofer, E. Kull, E. Kelz, and A. Morozov. In ILASS, (2001). [9] J. Benajes, J. V. Pastor, R. Payri, and A. H. Plazas. Journal of Fluids Engineering, 126, (2004). [10] R. Payri, J. M. Garca, F. J. Salvador, and J. Gimeno. Fuel, 84, (2005). [11] V. Macian, R. Payri, X. Margot, and F. J. Salvador. Atomization and Sprays, 13, (2003). [12] C. Badock, R. Wirth, and C. Tropea. In ILASS’99, Toulouse, France, (1999). [13] H. K. Suh, S. H. Park, and C. S. Lee. International Journal of Automotive Technology, 9, (2008). [14] N. Sanner, N. Huota, E. Audouard, C. Larat, and J.-P. Huignard. Optics and Lasers in Engineering, 45, (2007). [15] W. Wang, X. Mei, G. Jiang, S. Lei, and C. Yang. Applied Surface Science, 255, (2008). [16] D. Bruneel, G. Matras, R. Le Harzic, N. Huot, K. Knig, and E. Audouard. Optics and Lasers in Engineering, (2010).