| Description: |
Motivation/ Aim
The principle of airblast atomization is used in many different
applications, e.g. gas turbines or gasification of slurries, to transfer
the liquid fuel into the gas phase. The first part of this process
which produce liquid ligaments is called primary atomization. The
primary ligaments and big droplets are further decomposed by the
secondary atomization and form the spray.
Recent experiments [1] on the prefilmer surface show that
Weber numbers greater than 100 cause an intensification of the wave
formation at the liquid film surface. In this case the Weber number is
calculated in the following way:
Here ρ is the air density, u is the relative velocity, D is the film thickness and σ is the surface tension of the liquid.
For this case ligaments are separated from the wave crest.
Therefore the atomization is nearly completed before the liquid film
reaches the atomizing lip (Figure 1). This phenomenon is called "surface
stripping".
Figure 1: Formation of ligaments at high Weber number [1]
By future aimed processes with pressure
levels up to 60 bars the air density will increase drastically.
Therefore, the We-number which is proportional to the air density will
reach values greater than 100. Therefore the investigation of the
surface stripping is very important for future development.
Method
As a first step an experiment similar to the experiment described in
literature [1] will be designed and numerically calculated. The “volume
of fluid”- method (VOF) will be used to compute the two-phase flow [2].
Within this method, the phase interface is characterized by the volume
fraction in a control volume (= a volume of the grid cell). This
constitutes the main difference to the so called surface methods.
The VOF-method uses an indicator function (γ)
for the characterization of the phases, which marks the volume fraction
of a specified phase. Therefore the exact position of the interface is
unknown and has to be reconstructed by a special interpolation
technique. The balance equations of the flow are solved for both phases.
They are weighted with the corresponding volume fraction 0<γ<1
of the control volumes. Furthermore, an extra term that describes the
link of the momentum equations is considered. The volume fraction itself
is described by a transport equation without diffusive term. At this
point, the main problem of the VOF-method arises. Numerical inaccuracies
appear due to the lack of diffusive terms and can influence the
stability of the numerical method seriously. Within the currently
proposed project the software package OpenFOAM is used for the numerical
simulation. The indicator function is solved with a level-set method
[3, 4].
In the second part the described phenomenon will be investigated
by a basic experiment which has to be designed and manufactured. The
development of ligament formation and propagation will be recorded by a
high-speed camera.
In the third step a comparison between the simulation and the
experiment will be performed in order to derive empirical correlations.
These correlations will enable the calculation of the primary ligament
sizes as a function of thermodynamic parameters (pressure, temperature),
flow conditions (velocity, turbulence), material properties of the
atomized fluid (surface tension, viscosity) and geometry.
Literatur:
[1] U. Bhayaraju, C. Hassa, ICLASS06-073, (conference article). 2006. Surface wave propagation and breakup in planar liquid sheets of prefilming airblast atomisers.
[2] C. W. Hirt, B. D. Nichols, J. Comput. Phys. 1981, 39, 201–225. Volume of fluid (VOF) method for the dynamics of free boundaries.
[3] E. Olsson, G. Kreiss, J. Comput. Phys. 2005, 210, 225–246. A conservative level set method for two phase flow.
[4] E. Olsson, G. Kreiss, S. Zahedi, J. Comput. Phys. 2007, 225, 785–807. A conservative level set method for two phase flow II. |
