EXPERIMENTS ON SPRAY NOZZLES USING THE FLUORESCENCE IMAGING METHOD

Transcription

1 2 nd International Conference on Experiments/Process/System Modelling/Simulation & Optimization 2 nd IC-EpsMsO Athens, 4-7 July, 2007 IC-EpsMsO EXPERIMENTS ON SPRAY NOZZLES USING THE FLUORESCENCE IMAGING METHOD M.Vlachogiannis 1 * and T. Tsirikoglou 1 1 Laboratory of Fluid Mechanics Technological Institute of Larisa Dept. of Mechanical Engineering 411 Larisa, Greece mvlach@teilar.gr Keywords: Water Jet, Spray Nozzles, Plant Irrigation Abstract. Experimental results are reported on the structure of water jet/atomization, which is created by spraying nozzles. The fluorescence imaging method is applied, for a first time, to capture the evolution of the water jet structure in space and time. Preliminary results indicate that the fluorescence imaging (FI) method is effective to measure the geometrical characteristics of the water jet. Currently, by using the FI method, the atomization of the water droplets is clearly demonstrated only qualitatively. Secondary jet-like structures, which are not anticipated according to the technical characteristics of the nozzle, are created throughout the evolution process of the primary jet. We have observed coalescence events of small droplets and trajectories of larger droplets in the outer region of the water jet. 1 INTRODUCTION A large number of studies have been carried out, demonstrating many characteristics of a water spraying system. In this work we have concentrated on the primary fluid structure of commercial water spraying nozzles, which are employed in many agriculture applications. Our primary objective is to build a testing facility of water spraying nozzles in order to examine the geometrical characteristics of the fluid flow as well as some possible flow imperfections, which are originated from structural defects. Sudheera &Panda [1] studied the feasibility of an image processing technique for determination of the drop sizes from an irrigation spray nozzle. Their technique performed reasonably well in determining the drop size distribution of water spray from irrigation nozzles and may be useful to ascertain a quantitative relationship between the drop size distribution, operating pressure of sprinkler and its nozzle diameter. The drop size distributions for irrigation spray nozzles [2] were measured with a calibrated stain technique and with photographic or pellet techniques, a simple regression model used to explore the influence of nozzle size and pressure on drop size distributions for the two types of nozzles. Most of the current research activities are concentrated on the drop size rather on the resulting flow field of the water spraying system. Another important work by Tarjuelo et. al. [3] shows that the larger the spacing between sprinklers, the lower irrigation uniformity. In addition, the aforementioned researchers show that the uniformity increases by using jetstraightening vanes in the main nozzle, double nozzles under low wind speeds and single nozzles under high wind speeds, and by locating the sprinkler at 2 m aboveground. Results show that wind speed has a clear negative effect on irrigation uniformity. In our work the aboveground location of the nozzle is related with the uniformity of the flow field. 2 THE IMPLEMENTATION OF THE FLUORESCENCE IMAGING METHOD In this section, we describe the experimental apparatus for producing the water spraying.. We then describe the fluorescence imaging system, which is used to obtain qualitative measurements of the flow formation as a function of time and space. Finally, we discuss the image processing analysis, which involves extensive use of appropriate software. The experimental apparatus is shown in Figure 1. A transverse moving system is located above the nozzle to ensure that the fluid flow remains unaffected by the sidewalls. The UV light (two lamps) is placed, symmetrically, alongside the nozzle to ensure that the power of the light is sufficient for the fluorescence of the liquid. A piston-bearing pump (STAR, POWER SPRAYING, 3PS) is used for the creation of the water jet. A

2 worm screw is used to modify the location of the camera (optical field) across the flow downstream distance Figure 1. The experimental apparatus: 1) The UV light source and the camera 2) The piston pump 3) The transverse system 4) The real time experiment In order to describe the spatial and temporal dynamics for the spraying capability of the nozzle, it is necessary to obtain space-time measurements. To achieve this, we use the fluorescence imaging method described by Vlachogiannis et al. [4,5,6]. In our experiments, we dope the fluid with a small concentration, about p.p.m. of dye (sodium salt of fluorescein; C 20 H O 5 Na 2 ), which fluoresces under ultraviolet light and which has been proved to leave the relevant physical properties of the liquid unaffected. The ultraviolet (UV) source consists of two high-intensity lamps (Philips, TL40/05). The above parameters (dye, concentration of dye, ultraviolet source, distance from the nozzle plane) are kept as close to constant as possible, with the calibration taking care of small variations. A high-resolution CCD camera (Sony XC-77/77CE) and a monochrome frame grabber board (Matrox) are used to acquire and digitize the flow images up to a maximum speed of 20 frames per second. A combination of two optical filters (green corrective, yellow subtractive) mounted on the camera guarantees that only light in the fluorescence wavelength range is recorded. Acquired images are digitized at 576 x 768 pixels with 8-bit of resolution and each frame corresponds to an variable optical of spraying area. To increase the grabbing speed, we store the images in the random access memory (RAM) by using in-house software, and then we save them on the hard disk. The light intensity in the image plane is found to vary linearly with the local liquid thickness. The relation between the fluorescence intensity I(x; y; t) and the liquid thickness h(x; y; t) is modelled by the expression: I(x,y, t) = a(x, y) h(x, y, t) + b(x, y) (1) Image processing is accomplished by using the MATLAB software. Each image, which corresponds to a snapshot of the field of view at a specific time instant, is converted into a two-dimensional matrix (576 x 768 elements). By treating images as matrices, we add flexibility in analysing and displaying data. The usual forms of presentation are instantaneous profile scans in the streamwise or the transverse direction and iso-contours of the liquid thickness. In figure 2, an example of the fluorescence imaging method is depicted with two consecutive images and their transverse line profiles. All the profiles are taken at various distances from the nozzle edge. As it is clearly seen, the geometrical characteristics of the produced water jet could be estimated with great accuracy. The line profile at a specific downstream distance shows the time evolution of the water jet size as well as the development of secondary jet-beam structures. At the same figure, the iso-contours of the light intensity enrich our observations on the spray structure, providing useful information about the critical length, from which the liquid flows uniformly downstream. We believe that plant irrigations close to the critical length might evade the water loss due to air humidification. Simultaneously, will provide an optimum and homogeneous irrigation across the plantation

3 Liquid Intensity Liquid Intensity a t=0.1s t=0.15s Line Profile Line Profile 90 b 80 t=0.1s t=0.15s Transverse Distance (mm) Transverse Distance (mm) c t=0.1s t=0.15s Figure 2: The implementation of the fluorescence imaging method and the image processing. a) Fluorescence images at 2 consecutive time instants (t=0.1s & t=0.15s) b) The corresponding line profiles at specific distance from the nozzle edge & c) The iso-contours depict with great detail the plumes of the water spraying nozzle 2 EXPERIMENTAL RESULTS In the present work, experiments were carried out for four different types of nozzles. The first typical water jet nozzle was used to study the fluorescence imaging method. The working pressure varies between 5 and 15 bars. Figure 3 depicts the line profile of the light intensity (liquid thickness) as a function of the transverse distance for three different downstream planes (P= 5 bar). As the main water jet flows downstream, the intensity of the emitted light is reduced, proving that the sprinkling effect occur at very high distances from the nozzle edge. The light intensity of the main water jet is almost time independent for all the downstream planes measured. Figure 3 also depicts the appearance of a secondary water jet due to possible structural imperfections of the

4 Light Intensity Light Intensity Light Intensity nozzle. This secondary water jet is of less light intensity, showing that the entire water mass flow rate is divided, asymmetrically, in two main streams. Integrated the line profiles in time and space, we have found that 70% of the mass flow rate per width is leaded to the main water jet, 25% to the secondary water jet and only 5% is sprayed to the ground. Another important observation is the time evolution of the light intensity of the secondary water jet, which is related to the mass flow per width at the specific plane. As it is clearly seen, the distribution of the light intensity varies with time at two downstream distances (y=15mm, y=55mm). This implies that the operation of the piston-like pump affects the flow structure of the entire spraying system. However, further downstream (y=120mm), the time evolution of the light intensity distribution remains constant, indicating that possible drop coalescence events as well as mass dispersion of the main water jet are the dominant mechanisms. 2 2 t=0.1sec t=0.2sec t=0.3sec 2 2 t=0.1sec t=0.2sec t=0.3sec Transverse Distance Transverse distance Pixel Number t=0.1sec t=0.2sec t=0.3sec Transverse distance Pixel Number Transverse distance Figure 3: The line profile of the light intensity (liquid thickness) as a function of the transverse distance for three different downstream planes a) At y=15mm, b) at y=55mm and c) at y=120mm Measurements of the iso-contours of the light intensity are presented in figure 4. The image processing and the high resolution of the images provide valuable information about the flow structure. In this experiment we used the second nozzle, which is typical for improved water irrigation systems. Figure 4a & 4b depicts the flow structure at P=5bar and P=15bar. An umbrella like structure is clearly seen for both conditions, in which two main water jets are produced close to the nozzle edge. The intensity of the two main water streams is not affected by the pressure increase. However, the formed cone angle is increased slightly, the generation of a weak third water jet is observed and the plumes like structures in front of each jet become more intensive. By using the fluorescence imaging method and the iso-contour illustration, which is a sensitive indicator of the created flow structure, we are able to identify the optimum length of uniformity for irrigation systems. The optimum length is located in the plane where the plumes like structures are diminished. As a consequence of resolution enhancement, the optical field in figure 4 does not provide the optimum length of uniformity.

5 Light intensity a b Figure 4: The iso-contours of the light intensity, using the same nozzle but different pressures. a) P=5bar b) P=15bar The next issue on this work is to employ the fluorescence imaging method in order to investigate the spraying efficiency of two nozzles, which are applied in weed control and crop development. Nozzle #1 contains a metallic spinning disk, giving the optimum spraying effect. Nozzle #2 is a commercial, fully plastic item. Figure 5 depicts the line profile of the light intensity as a function of the transverse distance for Nozzle #1. Very close to the nozzle edge (y=15mm) the water jet forms a Gaussian distribution. Further downstream (y=55mm) the water jet is dispersed, very small drops are created and the light intensity of the main water stream is reduced by 35%. At distance y=120mm downstream, the water jet vanishes and a uniform spraying irrigation is formed y=15mm y=55mm y=85mm y=120mm Transverse distance (mm) Figure 5: The line profile of the light intensity (liquid thickness) as a function of the transverse distance for Nozzle #1

6 Light intensity The differences between the aforementioned nozzles are illustrated in figure 6. Nozzle #2 creates a concentrated water jet with several secondary water streams, which are responsible for the lost of the spraying efficiency. At the same downstream distance nozzle #1 creates a fully dispersed flow field, increasing the irrigation uniformity. These types of nozzles should be installed very close to the ground. As a consequence, the harmful effect of the wind velocity could be minimized since natural protection from the plantation may occur at such low distances from the ground. On the contrary, the commercial plastic nozzle must be installed at least 1m aboveground in order to generate a uniform spraying irrigation. As a consequence, the water is exposed to the wind velocity giving unexpected results nozzle #1 Nozzle # Figure 6: The line profile of the light intensity (liquid thickness) as a function of the transverse distance for two nozzles. Here the downstream distance is y= 120mm. The comparison of the iso-contours of the light intensity is depicted in figure 7. The first six images are consecutive time instants of the resulting iso-contour profile for the nozzle #2. The asymmetric distribution of the spray and the plume like structures are evident in the entire field of view. Of particular interest are water masses that flowing in the outer surface of the spraying cone, which are clearly illustrated in the series of images presented in figure 7. Those masses are drop aggregates that initiate at the nozzle edge. The generation of these large drops is a time dependent phenomenon due to structural imperfections of the nozzle. The appearance of this malfunction occurs periodically and caused by interfacial instabilities in the interface between the core water jet and the concentrated mass in the imperfection. The last iso-contour illustration corresponds to the spray structure of the nozzle #1. Two symmetrically water streams are produced, which are diminished further downstream. An umbrella like structure is created resulting in an improved spraying uniformity. 3 CONCLUSIONS Transverse distance (mm) The major objective of the present work is the conception of a nozzle testing facility. We have implemented the fluorescence imaging method in order to study the resulting flow structure in front of a spraying nozzle. Secondary water streams are detected, which are the result of structural nozzle imperfections. The uniformity of the spraying system is associated with the working pressure and the nozzle structural characteristics. Thus, the effect of the whether conditions, such as high wind speed, in connection with the nozzle type are of great importance for the optimum spraying system. The distance between the nozzle and the ground depends on the spray formation. In our future research we intend to carry out experiments with 2 or 3 nozzles in parallel in order to examine the interactions between the flow fields ACKNOWLEDGEMENTS This work was partly supported by the European Union, the Ministry of Education and Religious Affairs of Greece through Programme ARCHIMEDES I.

7 t=0.05s t=0.s t=0.15s t=0.20s t=0.25s t=0.30s Figure 7: Six consecutive iso-contours of the light intensity for nozzle #2 and one for nozzle #1. Here the working pressure is 2-3 bars. REFERENCES [1] Sudheera, K.P. & Pandab, R.K. (2000) Digital image processing for determining drop sizes from irrigation spray nozzles, Agricultural Water Management Vol. pp [2] Solomon, K.H.; Kincaid, D.C., Bezdek, J.C. (1985) Drop Size Distributions for Irrigation Spray Nozzles, Transactions of the ASAE TAAEAJ, Vol. 28, No. 6, pp [3] Tarjuelo, J. M., Montero, J., Carrión, P.A., Honrubia, F. T. & Calvo, M. A. (1999) Irrigation uniformity with medium size sprinklers part ii:influence ofwind and other factors onwater distribution, Transactions of the ASAE, Vol. 42 (3), pp [4] Vlachogiannis M. & V. Bontozoglou (2001) Observations of solitary wave dynamics of film flows. J. Fluid Mech., Vol 435, pp [5] M. Vlachogiannis & V. Bontozoglou (2002) Experiments on laminar film flow along a corrugated wall. J. Fluid Mech., Vol 457, pp [6] Argyriadi, K., Vlachogiannis, M. & V. Bontozoglou (2006) Experimental study of inclined film flow along periodic corrugations: The effect of wall steepness Vol. 18 (1): art. no