EXP 1-1 Measurement Quantities and Techniques

Pressure-swirl spray in a low-turbulence cross-flow

Front Page

Introduction

Review of experimental studies

Description

Experimental Set Up

Measurement Quantities and Techniques

Data Quality and Accuracy

Measurement Data and Results

Measurement quantities and techniques

Setup of PDA for measuring size and velocity of droplets in spray

The size and velocity of droplets in the spray were determined using a two-component fibre-based PDA measurement system (Dantec Dynamics A/S Skovlunde, Denmark). The arrangement of the measurement setup is shown in Figure 9.
A multiline Ar-Ion$^{+}$ laser produced a horizontally polarised light beam with 0.35 W output power. The green (514.5 nm) and blue (488 nm) wavelength components were extracted and used to provide two-component velocity measurements (in the Y and Z-axis) in the coincidence mode simultaneously with droplet sizing. Both the beams were split into a pair of parallel beams with a separation of 38 mm which were consequently expanded by a 1.98× beam expander and symmetrically intersected using transmitting optics. The frequency of one beam from each pair was shifted by 40 MHz. The intersected beams formed an elongated ellipsoidal measurement volume with the axes length of 0.123 × 0.123 × 1.63 mm. The measurement volume length was truncated by a 0.1-mm wide spatial filter. The positioning of the receiving optics at 48° from the forward direction was used to collect the light scattered from droplets dominated by the first order of refraction and to minimise reflections from windows. Both transmitting and receiving optics used lenses with 500 mm focal lengths. The principle of phase Doppler anemometry, data processing, measurement precision and uncertainties and other features are explained in .
Three planes in the spray were probed at axial positions Z = 10, 15 and 20 mm from the nozzle exit. For each plane, seven lines perpendicular to the flow direction and one line parallel with the flow direction in the atomizer axis were measured as seen in Figure 10. At each point, either 50,000 droplet samples were acquired or a 10-second acquisition duration was achieved. The Dantec BSA software 5.2 was used to control the measurement. The PDA configuration is described in Table 3. Each sample contains information on diameter and velocity components in the Y and Z-axis (the third X-component could not be measured with the two-component system) as detailed in the section Measurement data/results.

Figure 9: PDA setup, top view on the test section (7)

• 

  Figure 10: Measurement scheme with the atomizer placing and arrangement of the measurement points; top left: side view on the test section, top right: detail A showing the atomizer (axial cut-out) with its holder and measurement points in the YZ-plane, bottom left: top view on the test section, bottom right: detail B showing the measurement points in the XY-plane

Table 3 PDA Setup

Parameter	Value
Laser power output	0.35 W
Scattering angle	48°
Receiver mask	B
Receiver spatial filter	0.1 mm
The focal length of transmitting/receiving optics	500/500 mm
Wavelength	488 nm	514.5 nm
Velocity component	Axial	Radial
Velocity centre	19 m/s	0 m/s
Velocity span	51 m/s	73 m/s
Sensitivity	800 V	1050 V
SNR	0 dB	0 dB
Signal gain	8 dB	10 dB
Level validation ratio	8	2

High-speed visualisation

An HSC FASTCAM SA-Z type 2100K-M-16GB (Photron, Japan) was used to capture the instantaneous images of the spray under cross-flow. LED light model HPL3-36DD18B (Lightspeed Technologies, USA) provided spray illumination with a light pulse duration of 400 ns. The camera was equipped with a long-distance microscope 12 X Zoom lens (NAVITAR, New York, USA) composed of a 2X F -mount adapter (type 1 –62922), a 12 mm F.F zoom lens (type 1 –50486) together with 0.25 X lens (type 1 –50011).
Two different high-speed measurements were carried out. First, the visualisation of the spray trajectory was performed with the aim of recording the maximum possible area with sufficient pixel size for resolving important flow features. The second liquid sheet was visualised in detail to extract the sheet break-up length and surface wave structure.
The camera recorded for each measurement and flow regime a sequence of 4000 instantaneous images at a frame rate of 60,000 frames per second (fps) with an image resolution of 512 × 512 pixels (width × height) and at a shutter speed of 1 µs. The image area was 23.58 × 23.58 mm$^{2}$ and 11.44 × 11.44 mm$^{2}$ for the first and second measurement, respectively. The depth of field was approximately 6 and 13 mm respectively.
The spatial resolution is hence in the first case 46 μm/px × 46 μm/px, allowing the detection of large liquid structures to determine the spray trajectory in cross-flow. In the second case, it is 22.3 μm/px × 22.3 µm/px for the precise extraction of the liquid sheet break-up position. The camera axis was aligned perpendicularly to the flow and the main atomizer axis with backlit illumination, see Figure 11.

• 

  Figure 11: Configuration of the HSV experiment

• 

  Figure 12: Spray boundary for $p_{in}=$ 0.5 MPa, $u_{cf}=$ 0 m/s

The image sequence in the dataset documents the stages of the liquid discharge, formation of the conical liquid film, its destabilisation, primary break-up into bag and ligament structures and the secondary break-up of the ligaments into droplets and the final spray. It gives qualitative information on these phenomena as well as allows for photogrammetric estimation of the liquid sheet geometry, sheet fluctuations, wavy structures and corrugations, and break-up position. Sizes of ligaments, liquid bags and large droplets can be estimated, and their deformation can be tracked. The morphology and shape of the spray can be acquired. Spray boundaries were derived from full sets of 4000 instantaneous images using a function in MATLAB®. Boundaries for the windward and leeward sides were evaluated separately for each image. The average spray trajectory (green line) was derived afterwards and also the standard mean deviation was determined (red dashed line) as illustrated in Figure 12.

Boundary conditions

The geometrical domain, which determines the boundaries for CFD calculations of the spray-gas interaction, is represented by the test section, including the atomizer. If also the water flow inside the atomizer is of interest, then the internal atomizer cavity must be simulated apart from the space outside the atomizer.
The first domain contains the test section (7), with the atomizer holder and the atomizer body (17) protruding into the test section, see Figure 10. Their dimensions, positioning and other features are described here and in the section Experimental Setup, subsection Wind Tunnel Facility and drawn in Figure 10. The inlet of the test section (7) is connected to the laminarisation piece - channel (6). The test section outlet is connected to the exit diffuser (8). The pressure difference between the tunnel inlet and the test section is less than 200 Pa for the maximum set velocity of 32 m/s, so that the total pressure can be considered equal to the ambient air pressure. The working fluids are water and air with their properties described in the section Experimental Setup, subsection "Atomizer under test and its supply system. The water was sprayed into the air flowing in the test section. It produced a two-phase gas-liquid mixture where the liquid was dispersed in the gas phase.
The second domain, which is represented by the internal cavity in the atomizer (17), is documented in Figure 7. A flat velocity profile with zero $Tu$ is usually prescribed at the inlet of the swirling ports . The velocity results from the flow rate given in Table 2. The numerical setup used for such flows and geometrical models is studied in .

References

---

Contributed by: Ondrej Cejpek, Milan Maly, Ondrej Hajek, Jan Jedelsky — Brno University of Technology

Front Page

Introduction

Review of experimental studies

Description

Experimental Set Up

Measurement Quantities and Techniques

Data Quality and Accuracy

Measurement Data and Results

© copyright ERCOFTAC 2024