INVESTIGATION OF SUPERHEATED LIQUID CARBON DIOXIDE JETS FOR CUTTING APPLICATIONS L. Engelmeier, S. Pollak*, E. Weidner Chair of Particle Technology, Ruhr-University Bochum Universitätsstraße 150, 44801 Bochum, Germany Pollak@fvt.rub.de INTRODUCTION High pressure water jets are commonly used in cutting applications. If carbon dioxide is used instead of water, a residue free and dry processing is possible. Furthermore, improved cutting properties are expected, in comparison to the water jet technology. The premise for using CO 2 as a cutting fluid is the formation of liquid and coherent jets with a high specific energy. [1, 2] Coherent jets are characterized by an intact column with a constant diameter for a certain distance from the nozzle exit. In thermodynamic equilibrium, CO 2 does not exist in a liquid state at atmospheric pressure (figure 1). Below its triple point pressure of p T = 5,18 bar CO 2 exists only in a gaseous or solid state. Therefore, during the isentropic expansion through a nozzle from a liquid state to atmospheric pressure the saturation curve is crossed and a phase transition to gaseous/solid expected. 40 critical point Temperature in C 20 0-20 -40-60 -80 solid 3000 bar 2000 bar 1000 bar solid/liquid gaseous 50 bar liquid liquid/ gaseous triple point pressure = 5,18 bar solid gaseous 1 bar -20 0 20 40 60 80 100 Entropy in J/(mol K) Figure 1: Ts-diagram of carbon dioxide, calculated with [3]. The aim of this study is to investigate the possibilities for generating liquid and coherent CO 2 - jets. For this purpose CO 2 is compressed up to 350 MPa, cooled and expanded via a cylindrical, sharp edged nozzle. A particle image velocimetry system is used to investigate the influence of the injection temperature, injection pressure and post-expansion pressure on phase state and jet structure.
MATERIALS AND METHODS Experimental Setup Figure 2 shows a photograph of the experimental setup. Liquid carbon dioxide is compressed in a high pressure pump up to 3500 bar, is led via high pressure tubing to a heat exchanger, cooled, and is expanded through a nozzle. The nozzle is made of sapphire as commonly used in water jet cutting applications. These nozzles have a sharp edged entrance enabling the formation of long and coherent jets with a high specific energy density. The experiments were conducted under variation of injection pressure (50 bar < p < 3500 bar), injection temperature (minus 30 C < T < 20 C) and nozzle diameter (0,08 mm < d < 0,17 mm). nozzle high pressure intensifier thermostat II heat exchanger II (on the back of the front panel) heat exchanger I thermostat I Visualisation of the carbon dioxide jets Figure 2: Experimental setup. A particle-image-velocimetry system is used to analyse the jet phase composition and jet geometry. A laser light source, the camera and the jet are arranged along an optical axis as shown in figure 3. The laser light is expanded by a diffusor unit and illuminates a dye plate which in turn emits diffuse light with a wave length of 574 to 580 nm and a pulse duration of 20 ns. Thereby sharp images of the jets in spite of high jet velocities can be taken. nozzle lens jet dye plate camera object plane diffusor unit laser Figure 3: Arrangement of the particle-image-velocimetry system for flow visualisation.
RESULTS Influence of injection temperature The influence of injection temperature is shown in figure 4. The jets are formed with a nozzle diameter d = 0,08 mm, the injection pressure is increased from 200 bar to 1000 bar in steps of 200 bar. The injection temperature of the jets shown in the upper row is 20 C, the temperature of the jets in the second row is minus 14 C. At 20 C the jets fan out at the nozzle exit, the angle increases with increased injection pressure. The phase behaviour cannot be clearly identified by the photographs, but it is presumed that the continuous phase is in a gaseous state and the discontinuous phase consists of dry ice particles. With decreasing the injection temperature the jet structure changes considerably. The jets have a constant diameter over a certain distance from the nozzle exit. Magnifications show a wavy but closed, liquid column in the coherent portion of the jet. As CO 2 does not exist in a liquid phase at atmospheric pressure in thermodynamic equilibrium, the carbon dioxide must be in a metastable, superheated state. That implies that when crossing the saturation curve the phase change does not happen instantaneously and liquid, superheated jets can be formed. Within an injection pressure range of 50 to 3500 bar and with nozzle diameter of d = 0,08 mm and d = 0,1 mm liquid jets can be formed if the injection temperature is sufficiently low. Measurements show that the formation of liquid jets works well with injection temperatures close to the melting temperature. Figure 4: Influence of injection temperature on jet appearances.
Mechanical jet breakup of superheated jets Compared to water jets formed under similar pressure and nozzle conditions the coherent jet length is considerably shorter with CO 2. Water jets have a smooth and glassy appearance, whereas the CO 2 -jets have a wavy and ruffled surface. The shorter jet length and thus the jet breakup are caused by either thermodynamic or fluid-dynamic instabilities or a combination of both. The jet breakup mechanism of thermodynamic stable liquids is well known and dates back to the beginning of the 20 th century. Haenlein [4] identified four breakup regimes depending on jet velocity (figure 5a). The Rayleigh-regime is characterised by axisymmetric surface waves which are amplified by surface tension forces. The jet disintegrates into droplets with a diameter greater than the jet diameter. With increased jet velocity the surface waves become sinusoidal and are enhanced by aerodynamic forces resulting in a breakup of the jet into droplets with a diameter comparable to or smaller than the jet diameter (first and second wind induced jet breakup). With further increase in jet velocity the jet disintegrates into small droplets at the nozzle exit, also known as atomization. Ohnesorge [5] has shown that these four regimes can be depicted in the so called Oh-diagram, where the dimensionless Ohnumber is plotted against the Re-number (figure 5b). The Oh-number is a function of fluid properties (dynamic viscosity η L, density ρ L, surface tension σ) and the nozzle diameter. Oh = η L ρ l σ d d This classification predicts an atomization of high pressure water jets as well as of the liquid carbon dioxide jets that are shown in figure 4. These jets do not disintegrate at the nozzle exit. Thus the Oh-diagram is not suitable for the classification of these jets. The reason is that the influence of the nozzle geometry is not considered in the Oh-diagram. In sharped edged nozzles (as used in water jet cutting and in our experiments), the fluid is constricted at the nozzle entrance and emanates from the nozzle without any contact to the nozzle wall (figure 5c). Thereby friction is almost negligible and the jets are characterised by a smooth and glassy appearance. It is conceivable that due to the nozzle flow the boundaries in the Oh-diagram are shifted to higher Re-numbers. 10 1 1st and 2nd wind induced L Rayleighregime L 1st and 2nd wind Atomization induced jet breakup regime regime (a) Ohnesorge number 10 0 10-1 10-2 jet breakup Rayleigh-Regime Atomization HO 2 CO 2 10-3 10 1 10 2 10 3 10 4 10 5 10 6 Reynolds number (b) (c) Figure 5: (a) Jet breakup regimes according to [4], (b) Oh-diagram according to [5], (c) Constricted jets according to [6].
Characteristic appearances of the jet breakup that strongly depend on the jet velocity are shown in figure 6. The nozzle diameter is d = 0,08 mm and the injection temperature is T = minus 14 C. At relatively low jet velocity of 70 m/s the jet shows axisymmetric as well as sinusoidal waves, which are enhanced with increasing distance from the nozzle exit. Further down the jet breaks up into droplets with a diameter twice the jet diameter. This behaviour shows similarities to the Rayleigh- and first wind induced breakup regime. Therefore it can be assumed that the jet breakup is dominated by surface tension forces. With increasing jet velocity (170 m/s) the wavelength of the surface waves becomes shorter. With increasing distance to the nozzle exit the displacement of the jet increases and the jet breaks up into droplets smaller that the jet diameter. With further increase in jet velocity the surface waves become more ruffled. Both jets (figure 6b and c) show similarities to the second wind induced breakup regime and aerodynamic forces seem to dominate the jet breakup. Figure 6: Jet breakup regimes of superheated, liquid carbon dioxide jets depending on jet velocity. The results show that similar breakup regimes to the ones shown in figure 5a can be observed, and the boundaries in the Oh-diagram are shifted to higher Re-numbers. That fact that the breakup length is considerably shorter than that of water jets at same pressure and nozzle conditions may be caused by properties of the liquids. The surface tension and the kinematic
viscosity of water are around 5 times as big as the ones of liquid carbon dioxide. In consequence carbon dioxide jets are more sensitive to aerodynamic forces and disintegrate in a shorter distance to the nozzle exit. Jet breakup due to flashing phenomena With increasing nozzle diameter the appearance of the liquid jets changes significantly. Characteristic jet images using the example of a nozzle diameter d = 0,15 mm, an injection pressure p = 80 bar and an injection temperature T = 14 C are shown in figure 7, but similar observations are made for d = 0,12 mm and d = 0,15 mm in a pressure range of 50 to 1000 bar. The images show disturbances in the otherwise coherent portion of the jet. Magnifications clarify that the disturbances result from the formation of bubbles. The bubbles can be observed as scratches or round bubbles with a diameter bigger than the jet diameter. It can be assumed that these bubbles are formed due to nucleation phenomena and result in a disintegration of the jet, also known as flashing. Within a pressure range of 50 to 1000 bar jet patterns change randomly between all patterns shown in figure 7, whereby the bubble frequency increases with nozzle diameter and injection pressure. Above 1000 bar the jets become completely atomized. Figure 7: Thermodynamic jet instabilities.
SUMMARY AND CONCLUSION Up to a certain nozzle diameter and injection pressure liquid CO 2 -jets can be formed. The observed jet breakup regimes are similar to the jet breakup regimes of cold jets (saturation curve is not crossed during the expansion through a nozzle) and hence are dominated by fluiddynamic instabilities. To increase the breakup length of these jets one can decrease the ambient density to decrease the influence of aerodynamic forces acting on the jet surface. Above a certain nozzle diameter and injection pressure nucleation phenomena in the coherent portion of the jet dominate jet breakup. The relation between increased nucleation rate and both increased nozzle diameter and increased pressure are not known until now. The use of a superheated liquid jet enables residue free and precise cutting, which so far was successfully carried out for relatively soft materials. This gives rise to a number of interesting applications that require a gentle, precise and safe treatment of workpieces and their residues. Further investigations should lead to stabilization of the jet and subsequent increase in coherent jet length. Consequently, the cutting application of liquid superheated jets can be extended to harder and thicker materials. REFERENCES [1] WEIDNER, E., POLLAK, S., CO 2 : Abtrennung, Speicherung, Nutzung., 2015, p. 93 [2] ENGELMEIER, L., POLLAK, S. KRETZSCHMAR, M., WEIDNER; E., Chemie Ingenieur Technik, Vol. 88, 2016, p. 672 [3] SPAN, R., ECKERMANN, T., HERRIG, S., HIELSCHER, S., JÄGER, A., THOL, M., Thermodynamic Reference and Engineering Data, Version 2.0.1, 2015 [4] HAENLEIN, A., Forschung auf dem Gebiet des Ingenieurwesens, Vol. 2, 1931, p. 139 [5] OHNESORGE, W., Zeitschrift für angewandte Mathematik und Mechanik, Vol. 16, 1936, p. 355 [6] HIROYASU, H., ARAI, M., SHIMIZU, M., Recent advances in spray combustion, 1996, p. 173