The Investigation of the Industrial Spray Systems Using Interferometry Particle Imaging Method

Veiw figure View Figure

The IPI method employs two cameras alignment. This alignment consists of one focused and one defocused camera. The first camera acquires focused images, which are used for particle detection and for velocity measurements. The second camera records defocused images, which are used for size measurements and for validation. By knowing reflection and first-order refraction of the particles this method provides measurements of the particle diameter. Once the particle is out of focus, the interference fringes appear, arising from the two scattered rays. The shape of the defocused image of each glare point depends only on the shape of the aperture, whereas the size of the defocused images depends on the degree of defocusing. As the degree of defocus increases and the spatial resolution of the optical system decreases, the two points merge into one single image with interference fringes. Large particles exhibit more fringes in the defocused image than small particles.

The particle size limits of the IPI technique are fixed by the aperture size and the working distance of the camera. Particle size is determined from the spatial fringe information. The number of fringes in the particle images is linearly related to the particle diameter through:

(1)

For droplets where the relative refractive index (m > 1), liquid droplets, glare particles:

(2)

For (m < 1 ), bubbles, glass spheres:

(3)

Where:d_p particle diameter, κgeometric factor, N_fr number of fringes, z distance from light sheet to camera lens, d_a aperture diameter, m relative index of refraction, ϕ observation angle.

The minimum particle size is calculated from size that represents one fringe:

(4)

The maximum size is determined from Nyquist criteria in meaning that at least two pixels define the fringe:

(5)

Where: n_x the number of pixels in the x-direction, Δx dimension of the CCD array in the x-direction.

The IPI measurement system consists of NewWave Gemini Nd:YAG pulsed laser of energy in each pulse 120mJ. The duration of one pulse was 10ns. The light sheet used in these studies was 3mm thick. The image records were captured by two HiSense Neo, the sCMOS cameras of the spatial resolution (2560×2160) px with pixel size 6.5 µm. This new camera is unique in its ability to simultaneously offer ultra-low noise (1.4 e-), extremely fast frame rates (50 Hz in global shutter mode), wide dynamic range (>20,000), high resolution (5.5 MP) and a large field of view. This unique equipment allows obtaining high quality image and high contrast that enables the mathematical analysis.

The cameras were aligned to observe the same area in the same optical axis, for this arrangement we used special IPI lens fitting adapter. Each camera was fitted with the lens of focal length 60mm; the one for focused image and one for de-focused. The cameras setting was placed under angle ϕ = 90 ° in the distance 670mm of the camera lens to the laser sheet. The aperture diameter was 15mm. The datasets of double images was recorded by IPI system of focused and defocused images were taken. The focused images were cross correlated and validated using the set of validation methods. Defocused images were analysed by Interferometry calculating algorithm. The validation of defocused images is dependent on the saturation density of particles, distance between particles and level of background noise. The optical setup enables to measure particles in the range from 20 up to 280 micrometres. The Figure 1 shows the IPI measurement configuration, synchronization between cameras and laser that is provided by the synchronization hub unit. Here it is also seen the angle ϕ between camera alignment and laser sheet that is dependent on the drop reflection index.

2.2. Atomization Setup
2.2.1. The pneumatic nozzle

The analysed pneumatic nozzle system that we used here enables huge range of modification for the purpose of exact industrial application. The nozzle attributes can be influenced by the nozzle diameter, shape, design, pressure adjustment, working liquid characteristics or the inner chamber construction and the external add-ons. The basic adjustment has potential for application, where the droplets of relatively small size are required and the drop impact velocity is small.

Here we use the full cone spray nozzle. This kind of nozzle produces full cone sprays round pattern that is completely filled with spray drops. The spray is formed by the vane less chamber, which imparts the liquid to the orifice due the controlled turbulence. The turbulence is induced by inner structures and proportions of the chamber. The inner geometry of mixing chamber is shown in Figure 2. The spray pattern is taken in the working distance of the atomization process. The working distance depends on the applied pressure or liquid feed rate. In this study we used the external mix, so the liquid is aspirated due the gravity force or the ejector effect to the nozzle. According to Figure 2 the spatial distribution of spray pattern is homogeneous and well sharped.

Figure 2. Simplified sketch of the spraying nozzle and the full cone spray pattern

Download as

Veiw figure View Figure

The full cone spray nozzle consists of 1/8J stainless steel nozzle body and round spray air cap for external mix to produce full cone found spray pattern. This type of nozzle works with maximum air pressure 4bar, feed rate 0.68l/hour and maximum spraying distance 1.8m.

The spray nozzle was joined into the system as an external mix configuration. The external mix system supplied the liquid into the nozzle by liquid siphon. This setup is designed to draw the liquid from a container through the feed line into the air flow, where it is atomized. The second exit of the nozzle was fed by compressed air. The pressure of the incoming air was measured by digital barometer Almemo and controlled by the reduction valve.

Figure 3. The investigated areas A, B, C of the spray plume

Download as

Veiw figure View Figure

2.2.2. The EHDA Nozzle

The second kind of nozzle that is mostly used for colouring bodies in car industry is based on uniform electrostatic field.

Electrohydrodynamic atomization (EHDA) is the disruption of a liquid surface by electrical forces resulting in the formation of charged droplets. The physics of electrostatic spraying was first systematically examined by Lord Rayleigh in the late 1800’s. He also calculated the critical amount of charge that is necessary to destabilize a spherical droplet and observed the resulting instability that leads to the liquid jets production. These jets break up forming small stable charged droplets. Further motion of these droplets and their velocity is influenced by the electrostatic field. According to the applied high voltage, electrode distance and of course the chemical and physical properties of used solution, the droplets of micrometer or nanometer diameter can be produced.

The basic principle of the electrostatic spray technique can be briefly described as a linear pumping system with variable feed rate that is used to eject a liquid solution from a reservoir and fill the capillary. The electrical system consists of two electrodes axially placed in a certain distance, the gap. When a high voltage is applied on one of the electrodes, mostly the nozzle and the electrostatic force acting on the ejected droplet overcomes the surface tension of the solution and a straight jet erupts from the apex of the Taylor cone at the nozzle and travels toward an electrically oppositely charged electrode, generally named as a collector. The collecting mechanism at the target electrode can be explained using a dielectrophoresis force. The force directs the sprayed droplets toward the highest voltage gradient and is closely dependent on the droplet diameter, conductivity and applied field.

The attracting force acting on a droplet in a non-uniform electric field， is given by, where µ is a dipole moment. This dipole moment can be written as where β are a function of the dielectric constant and the electrical conductivity of the sprayed droplet, and ε0 is a permittivity of vacuum.

The prediction of optimal feed rate for particular solution can be calculated from the quotation (6) ^[4] A critical voltage for needle spraying apparatus that is needed to disintegrate the droplets can be calculated correspondingly with Rayleigh theory about a disintegration of liquid bodies and is set along the equitation (7) Finally we can set the optimal feed rate, that corresponds directly with critical charge, desired droplet size and vaporization rate of the sprayed solution, counting with the humidity of ambient space and minimum sphere segments height needed for good start of electrostatic spraying.^{[5, 6, 7]}

(6)

(7)

where f is the optimal feed rate of solution, d is the prediction of droplet diameter, q is the suitable voltage that should be applied, γ is liquid–gas surface pension and ρ the density of the solution, ε is the permittivity of the solution, h is the distance from the needle tip to the collector in centimeters, R is the needle outer radius.

2.2.3. The Liquid Flow behind the Blade Profile

The third application of the IPI method was the visualization of the liquid flow behind the blade profile. This situation is common in power industry and the effect of water condensationon the blades of turbine is causing a lot of damages on the equipment itself and the others in the row. The measurements for flow profile in the wind tunnel were implemented. Here we used this non-invasive technique for the size distribution of aerosols that was monitored withheld from the profile at three basic flows (0.1 l / min, 1l/min,and 5l/min). The air velocity in the wind tunnel was maintained at all times at 30 m / s

Method IPI evaluated the droplet size sprays using interference structures drop illuminated laser cut. In principle it is possible to measure only single drops falling apart and not departing jets of water. These non-segregated streams primarily occurred when water 5l/min and could not be evaluated. A method IPI was also recorded in this mode each breakaway drops at a greater distance from the profile.

Here we were tested different distances of the camera from the laser cut (190mm, 700 mm and 1080 mm). Distance from lens to laser cut is affected by the size of the scan area, the spatial resolution of the current field and the range and quality of measurement of droplet size. At baseline measurement the optimum arrangement in terms of the given parameters was selected for placement of the camera lens to a distance of 1080 mm from the laser cut.

Depending on the velocity of the fluid flow on the surface profile is also changing the character and appearance of the spray.

Figure 4. The spray of water just behind the wing profile

Download as

Veiw figure View Figure

Just beyond the edge of the profile were tracked the individual droplets. We could be very accurate in this mode as well as in determining the dynamic properties of the spray. Using the basic PIV method we measured the velocity of the droplets at each location for the profile.

3. Results and Discussion

3.1. The Results of the Pneumatic Nozzle

The effect on the spray geometry has several liquid properties: viscosity, surface tension and density. The surface tension and viscosity represent the forces acting at the interface between two fluids (water, air). During the atomization the breakup of continuous liquid body into many small one increase the surface area over the whole liquid, thus the potential energy increases. The potential energy of the system is supplied by shear force acting on the liquid flow. The liquid flow is expressed as a pressure driven liquid flow through an orifice.

In this section is presented a visualization of spray geometry investigated in three areas. The first area was close to the nozzle exit, the second in the distance 0.7m far from the nozzle exit and the third one in 1.7m. In all investigated areas the spray was analysed on the drop size distribution. The applied air pressure was chosen according to the optimal nozzle range.

Figure 5. The velocity profiles in the distance 0.1m from the nozzle exit

Download as

Veiw figure View Figure

The spray nozzle was investigated in the pressure air range from 0.7 to 5.0bar. For industrial application the pressure 1.3, 2.7 and 3.5bar is most interesting so the results we bring here were taken, measured and evaluated by these pressures.

With higher applied pressure, the stream of droplets starts to be oversaturated and the IPI measurements is disabled. This problem was partly solved by changes in optical setup, decreasing of aperture diameter, so we could partly observe the disintegration of particles in this area. The average size of the liquid a was in order of submillimeters.

With increasing pressure there were only very small changes in the spray angle. That is achieved by the optimal shape of nozzle’s orifice. As it is seen on visualizations in Figure 5, the divergence of main droplet stream is in the limit of 17 degrees. If the stream starts to shift its parameters it is the sign of the nozzle damage or malfunction.

With higher pressure the instabilities in the stream occur. The instabilities are caused by local vorticities in the flow. The PIV method was used to determine the flow field of the carrying air and recognize the turbulences. The air flow field has the typical shape of submerged jet. About 30mm above the nozzle exit the stream of the liquid disrupt into the filaments. Upper from this area continues the secondary part of liquid bodies’ disintegration into the larger and smaller droplets. Larger bodies are carried close to the central line of the main air stream, where the velocities are higher and the local depression effects.

Figure 5 shows the calculated scalar maps of droplets motion velocity for applied pressure 1.3, 2.7 and 3.5bar. It is significant velocity rise. The maximum velocity of the spray was 75m/s under pressure 3.5bar. In industrial process the common pressure is set on 2.5bar. The velocity of droplets in this regime and area reach the value of 30m/s.

The results of the droplet diameter just above the nozzle orifice are influenced by the measurement error. The droplets in this region are joined into one massive bulk of liquid and the motion of this jet cannot be recorded and analysed with PIV procedures. In the distance 15mm above the nozzle exit the jet disrupts and separated liquid blobs are measurable. These initial blobs have 1mm in diameter. According to this diameter, the dynamics of the blob is considerable and the velocity doesn’t correspond to the velocity of air flow.

Figure 6. The histogram of droplet size distribution taken in the distance 1.7 m far from the nozzle exit. The results are compared according to pressure for 1.3 bar and 3.5 bar

Download as

Veiw figure View Figure

The recorded flow field was evaluated and compared on velocity profiles in 0.1m and 1.7m distance from the nozzle exit. These distances conform to the selected investigated areas A and C Figure 5. The highest velocities can be measured close to the nozzle in core of the central stream. During the atomization process the droplets lose their kinetic energy on disintegration,friction force and the part of the energy converts to potential energy according to Navier-Stokes equation.

3.2. The Results of the EHDA Nozzle

The EHDA nozzle was connected to the positive high voltage power supply (FX50P06, Glassman High Voltage) with an applied the voltage from 10 kV to 20kV. A grounded steel collector was placed at 250 mm from the nozzle orifice. The nozzle was continuously fed with 20ml/min with linear pump and pressure of 1.3bar. Here we were focused on the resulting impact velocity that strongly corresponds to the droplet diameter. For this purpose the histogram of spray distribution was analysed in the distance 250mm from the nozzle orifice close to the surface of the grounded electrode ^[8].

The final size distribution is dependent not only on the applied pressure but also on applied voltage. This dependence is shown on Figure 7. The resulting impact velocity is not significantly influenced, but using optimal high DC voltage can be reduced the droplet diameter. The histogram is uncovering the effect of primary and secondary atomization. The secondary atomization is caused due the surface charge on the droplets and effective optimization can lead to micro scale diameters.

Figure 7. The histogram of droplet size distribution effected by high voltage applied directly to the nozzle orifice

Download as

Veiw figure View Figure

3.3. The Results of the Liquid Flow behind the Blade Profile

For the profile in the regime forms part of the water stream that couldn’t be pulled by air apart into individual droplets. IPI measurement in this area can be considered correct. The nature of these non-disruptiveligaments close to the blade surface could be monitored by other visualization techniques.

At maximum liquid flow speed - 5l/min, reached the compactflowbefore spreading to spray a few inches. The stream changed its geometry depending on the air speed from the blade profile and finally shaped into the form expected Karman vortex trail.

As the first stage there was measured the flow fields of the droplet spray. These results demonstrate the dynamics of particles and illustrate the extent to spray that follows the movement of the air stream. Beyond the basic vector maps were plotted stream profiles in the different sections of the monitored area.

Compactness of the stream flow 5l/min is obvious from visualization. The graph in Figure 8 compares the streamwhich is less accelerated with the ambient air with the individual drops in previous regimes. The comparison of histograms (see Figure 9) shows the changes in the droplet size diameter and distribution that is non-affected by the liquid flow. From the Kelvin-Helmholz theory of disintegration the liquid into the spray results that the spray is mostly influenced by the surrounding air flow of higher velocity. This condition is also relevant in this case.

Figure 8. The vector map of the spray of droplet motion in the 0.1l/min, 1l/min and 5l/min liquid flow in the wind tunnel of constant air flow speed 30m/sec

Download as

Veiw figure View Figure

Figure 9. The histogram of droplet size distribution taken for 0.1l/min and 5l/min liquid flow behind the blade profile

Download as

Veiw figure View Figure