Mean droplet size parameters

Both effects can produce coarser atomization. However, the influence of Hquid viscosity on atomization appears to diminish for high Reynolds or Weber numbers. Liquid surface tension appears to be the only parameter independent of the mode of atomization. Mean droplet size increases with increasing surface tension in twin-fluid atomizers (34). is proportional to CJ, where the exponent n varies between 0.25 and 0.5. At high values of Weber number, however, drop size is nearly proportional to surface tension. 

To characterize a droplet size distribution, at least two parameters are typically necessary, i.e., a representative droplet diameter, (for example, mean droplet size) and a measure of droplet size range (for example, standard deviation or q). Many representative droplet diameters have been used in specifying distribution functions. The definitions of these diameters and the relevant relationships are summarized in Table 4.2. These relationships are derived on the basis of the Rosin-Rammler distribution function (Eq. 14), and the diameters are uniquely related to each other via the distribution parameter q in the Rosin-Rammler distribution function. Lefebvre 1 calculated the values of these diameters for q ranging from 1.2 to 4.0. The calculated results showed that Dpeak is always larger than SMD, and SMD is between 80% and 84% of Dpeak for many droplet generation processes for which 2left-hand side of Dpeak. The ratio MMD/SMD is... 

In many atomization processes, physical phenomena involved have not yet been understood to such an extent that mean droplet size could be expressed with equations derived directly from first principles, although some attempts have been made to predict droplet size and velocity distributions in sprays through maximum entropy principle.I252 432] Therefore, the correlations proposed by numerous studies on droplet size distributions are mainly empirical in nature. However, the empirical correlations prove to be a practical way to determine droplet sizes from process parameters and relevant physical properties of liquid and gas involved. In addition, these previous studies have provided insightful information about the effects of process parameters and material properties on droplet sizes. 

In fan spray atomization, the effects of process parameters on the mean droplet size are similar to those in pressure-swirl atomization. In general, the mean droplet size increases with an increase in liquid viscosity, surface tension, and/or liquid sheet thickness and length. It decreases with increasing liquid velocity, liquid density, gas density, spray angle, and/or relative velocity between liquid and surrounding air. 

Various correlations for mean droplet size generated by plain-jet, prefilming, and miscellaneous air-blast atomizers using air as atomization gas are listed in Tables 4.7, 4.8, 4.9, and 4.10, respectively. In these correlations, ALR is the mass flow rate ratio of air to liquid, ALR = mAlmL, Dp is the prefilmer diameter, Dh is the hydraulic mean diameter of air exit duct, vr is the kinematic viscosity ratio relative to water, a is the radial distance from cup lip, DL is the diameter of cup at lip, Up is the cup peripheral velocity, Ur is the air to liquid velocity ratio defined as U=UAIUp, Lw is the diameter of wetted periphery between air and liquid streams, Aa is the flow area of atomizing air stream, m is a power index, PA is the pressure of air, and B is a composite numerical factor. The important parameters influencing the mean droplet size include relative velocity between atomization air/gas and liquid, mass flow rate ratio of air to liquid, physical properties of liquid (viscosity, density, surface tension) and air (density), and atomizer geometry as described by nozzle diameter, prefilmer diameter, etc. 

The studies on the performance of effervescent atomizer have been very limited as compared to those described above. However, the results of droplet size measurements made by Lefebvre et al.t87] for the effervescent atomizer provided insightful information about the effects of process parameters on droplet size. Their analysis of the experimental data suggested that the atomization quality by the effervescent atomizer is generally quite high. Better atomization may be achieved by generating small bubbles. Droplet size distribution may follow the Rosin-Rammler distribution pattern with the parameter q ranging from 1 to 2 for a gas to liquid ratio up to 0.2, and a liquid injection pressure from 34.5 to 345 kPa. The mean droplet size decreases with an increase in the gas to liquid ratio and/or liquid injection pressure. Any factor that tends to impair atomization quality, and increase the mean droplet size (for example, decreasing gas to liquid ratio and/or injection pressure) also leads to a more mono-disperse spray. 

Generally, the mean droplet size is proportional to liquid surface tension, and inversely proportional to liquid density and vibration frequency. The proportional power index is —1/3 for the surface tension, about -1/3 for the liquid density, and -2/3 for the vibration frequency. The mean droplet size may be influenced by two additional parameters, i.e., liquid viscosity and flow rate. As expected, increasing liquid viscosity, and/or flow rate leads to an increase in the mean droplet size,[13°h482] while the spray becomes more polydisperse at high flow rates.[482] The spray angle is also affected by the liquid flow rate, vibration frequency and amplitude. Moreover, the spray shape is greatly influenced by the direction of liquid flow (upwards, downwards, or horizontally).[482]... 

Thus, both the mean droplet size and the size distribution may be predicted using these correlations [Eqs. (26), (27), (28), or (29) and Eqs. (30), (31)] for given process parameters and material properties. For a given atomizer design, the standard deviation of droplet size distribution has been found to increase with the melt flow rate, but appears to be less sensitive to the gas flow rated5 Moreover, the variation of the standard deviation is very atomizer- and melt-specific. An empirical correlation which fits with a wide range of atomization data has the following form ... 

HIPE stability depends greatly on a number of parameters, including the nature and concentration of the surfactant, the nature and viscosity of each liquid phase, system temperature, mean droplet size, interfacial tension between the phases, strength of the interfacial film and the presence of added electrolyte in the aqueous phase. The formation of a rigid interfacial film is thought to be of paramount importance to the stability of HIPEs. 

Using a multivariate regression approach based upon a modified El-Shanawany and Lefebvre equation (32) yields for following predictive relation of Sauter mean droplet size (SMD) for a twin-fluid type atomizer (33) utilizing the non-dimensional parameters of ALR, Weop and Ohj>p ... 

It is probable that numerous interfacial parameters are involved (surface tension, spontaneous curvature, Gibbs elasticity, surface forces) and differ from one system to the other, according the nature of the surfactants and of the dispersed phase. Only systematic measurements of > will allow going beyond empirics. Besides the numerous fundamental questions, it is also necessary to measure practical reason, which is predicting the emulsion lifetime. This remains a serious challenge for anyone working in the field of emulsions because of the polydisperse and complex evolution of the droplet size distribution. Finally, it is clear that the mean-field approaches adopted to measure > are acceptable as long as the droplet polydispersity remains quite low (P < 50%) and that more elaborate models are required for very polydisperse systems to account for the spatial fiuctuations in the droplet distribution. 

In (5) D00 is the median diameter and a is the standard deviation of the distribution. By fitting the experimental R-values, the parameters D0 0 and a can be determined and hence the size distribution of the droplets in the emulsion can be obtained. For microbiological safety aspects Dj 3 is more important. D3>3 is the volume weighted mean droplet diameter and a is the standard deviation of the logarithm of the droplet diameter. The parameter D3 3 is related to the parameter D00 according to ... 

Eq. 1 showed that in the case of unrestricted diffusion the echo attenuation value R depends upon the durations 8 and A. This is also true in the case of restricted diffusion, although in a different manner. The dependence of the R-value upon these two parameters is shown in Fig. 5. This figure clearly shows that the echo attenuation factor R steadily decreases with increasing A in the case of unrestricted diffusion, but becomes independent of this parameter in the case of restricted diffusion. It may be deduced from this figure that it is necessary to determine the parameters of the log-normal droplet size distribution R as a function of A or by measuring R as a function of 8 for a fixed large value of A. Measurement of only one R-value, at a chosen 8 or A, is not sufficient for a careful determination of the droplet size distribution in Fig. 5 a given In R-value can be found on more than one In R versus A-5/3 curve. This means that the In R-values have to be determined for different values of A and/or 8. 

When two liquids are immiscible, the design parameters include droplet size distribution of the disperse phase, coalescence rate, power consumption for complete dispersion, and the mass-transfer coefficient at the liquid-liquid interface. The Sauter mean diameter, dsy, of the dispersed phase depends on the Reynolds, Froudes and Weber numbers, the ratios of density and viscosity of the dispersed and continuous phases, and the volume fraction of the dispersed phase. The most important parameters are the Weber number and the volume fraction of the dispersed phase. Specifically, dsy oc We 06(l + hip ), where b is a constant that depends on the stirrer and vessel geometry and the physical properties of the system. Both dsy and the interfacial area aL remain unaltered, if the same power per unit volume (P/V) is used in the scale-up. 

---