Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY
Edited by: B.H.V. Topping and C.A. Mota Soares
The Creation of a Gas-Liquid Nozzle with Predictable Behaviour
A.V. Karpyshev, A.V. Tsipenko and A.A. Yakovlev
Scientific Research Center of Advanced Technologies, Moscow State Aviation Institute - MAI, Moscow, Russia
A.V. Karpyshev, A.V. Tsipenko, A.A. Yakovlev, "The Creation of a Gas-Liquid Nozzle with Predictable Behaviour", in B.H.V. Topping, C.A. Mota Soares, (Editors), "Proceedings of the Fourth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 67, 2004. doi:10.4203/ccp.80.67
Keywords: gas-liquid flow, convergent-divergent nozzle, Euler-Euler model, experimental investigations.
Water mist application dramatically increases fire-fighting efficiency. The efficient atomization may be achieved with gas-liquid nozzle application, when the liquid is broken into tiny droplets and accelerated in a gas flow. A method of designing these nozzles and nozzle flow description are given in this paper.
First of all, it is important to know discrete phase parameters which determine a gas-drop jet range. The jet was numerically investigated with the use of the original model . On the basis of the outcome of the calculation analysis it is possible to make the following conclusion: there is a particle flow rate to gas flow rate ratio when the gas-drop range is greater than the hose barrel jet range with an equal inlet pressure, and we can assess optimal quantity. The conclusion is verified by experimental investigation.
To determine parameters at the initial jet cross-section, at the nozzle exit, a monodimensional and monodispersal model of nozzle section design by a pre-set pressure difference was used. The same model is used to obtain the first nozzle section version.
Due to the difficulty of simulating the gas-and-liquid-mixing processes constructive decisions are made mostly after the tests. With this approach it is very difficult to find out the reason when a system fails to give a design jet. It is a dilemma to guess either a "bad" nozzle would not accelerate a "good" flow of droplets or it is the other way round or the operating mode is wrong. However, an optimum operating mode is urgent for a particular design and it should be done with the least costs.
The analysis of different-configuration nozzles shows  that a test and hypothetic (for "equilibrium" mixture) phase flow rate interposition characterizes a transverse liquid distribution, namely: with a hypothetic curve located higher in the graph an experimental flow is sufficiently homogeneous, the dependence difference caused by phase velocity difference; with a contrary disposition the difference is caused by the fact that there is a lot of film and quite a solid liquid nucleus (stream) inside the nozzle or, which is the same, there are areas with little liquid mass compared with the core. A supposition set forward allows to select an operating mode and system improvement direction quite easily.
To make a physical flow model more definite and receive the data necessary for the spatial numerical simulation the local parameters of the nozzle exit section and some nozzle inlet flow data were obtained experimentally. Photography in a flash light (duration from 10) was used. The gas and liquid mass flow rate, pressure distribution along a convergent-divergent nozzle wall were measured.
Since the flow is optically dense, a probe-sample selector, a probe-static pressure meter, a probe-meter of a total incoming flow pressure were used to obtain the nozzle exit flow parameters. The results of probe measurements were evidence of an annular flow shape and film.
A numerical nozzle flow simulation was performed with experimental data of droplet-to-droplet and droplet-to-film interaction applied . A method of "large particles" , which is Harlow's "particles in cells" method development, was also used. Euler-Euler non-stationary models were employed for gas and droplets. The numerical outcomes show that there is an area in the diverging part of the nozzle near the wall. An experimental check with application of an electric probe has shown that a narrow area does exist where liquid is actually absent.
You can see from the computation experience that the scope and complexity of a calculation model are greater than the abilities of single-processor computers, when you can have the results showing only tendencies and qualitative picture of the flow. By virtue of the above fact the nozzle development procedures combine computational modelling and tests.
purchase the full-text of this paper (price £20)