Technical field
[0001] The subject of this invention is a unit forming a part of a generator which produces
in a liquid small bubbles of gas brought into the unit by a pipe or other cavities.
Small bubbles are required in a wide range of production processes, in particular
those in which the supplied gas has to diffuse across the phase interface into the
liquid. An example is aeration of water to increase its oxygen content. In this as
well and in majority of other cases it is desirable for the generated bubbles to have
the least possible dimensions, because the gas diffusion transport increases with
increasing interface area - and the total area is much larger if a given gas volume
is divided into a large number of small bubbles. Moreover, the small bubbles rise
more slowly towards the liquid surface and this increases the total time over which
the transport from the bubbles into the liquid takes place.
[0002] In many industrial processes is demanded rather high overall transport rate from
the gas into the liquid. In such cases it is expected that for achieving the high
total transport rate will be used a larger number of the units as described in this
invention, all operating in parallel.
[0003] Use of gas bubbles in a liquid is common in a whole range of engineering areas and
in practically all is found a possible application for a method producing the bubbles
extraordinarily small, as can be done in the unit according to this invention. To
such areas belong in particular waste water processing, production and recycling of
paper, separation of various materials by flotation, producing such organic substances
as yeast and, in particular, potentially very important growing of unicellular (as
well as even more complex) plants such as the well known algae.
Background art
[0004] The simplest and most often used method of generating bubbles is bringing the gas
into a device known as aerator. This device is characterised by a large number of
small orifices connected at one end to the common gas supply and at the other end
open into the liquid. The liquid is usually inside a vessel or tank and the aerator
is submerged under the liquid level. The bubbles are then formed by outflow of gas
from the small orifices of the aerator. The main problem is instability of parallel
bubble formation. This is a direct consequence of the basic law governing bubble behaviour.
According to this law, the pressure difference between the gas inside the bubble and
the surrounding liquid is inversely proportional to the curvature radius of the bubble
surface (the Young-Laplace law). If the bubbles are generated simultaneously at the
exits from the small parallel orifices of the aerator, and due to some chance effect
one of these bubbles increases its size - and thus also increasing its curvature radius
- then the gas pressure inside this larger bubble will be lower than in the neighbour
bubbles supplied by the gas from the same gas source. Of course, the gas then will
flow, driven by the larger pressure difference, into the bubble that is already larger,
at the expense of decreased or even stopped gas flow into the neighbour bubbles. They
cease to grow, while the large bubble will reach an extremely large size, out of proportion
to the size of the aerator orifice exits. Despite all the effort directed to making
the aerator orifices as small as is allowed by the orifice manufacturing method, the
growth of very large bubbles is not influenced. What is mostly generated are big bubbles.
[0005] Way and means towards suppression of this process were sought and one of the promising
possibilities suggested as a solution is disclosed in the European Patent
EP2081666 (inventors: Tesa
and Zimmerman). The idea is to act on the gas supplied into the aerator orifices
by flow pulsation produced by a fluidic oscillator.
[0006] The fluidic oscillator to be used for the purpose was already earlier described in
an article by authors Tesa
V., Hung C.-H., and Zimmerman W.: "No-Moving-Part Hybrid-Synthetic Jet Actuator",
Sensors and Actuators A, Vol. 125, pp. 159-169, 2006. This oscillator, as it is also described in the Patent
EP2081666, consists of a fluidic diverter amplifier having no components that are moved or
deformed in the course of amplifier operation - and from a feedback loop channel.
The channel connects the two control terminals of the amplifier. The gas is supplied
into the supply nozzle of the amplifier and leaves it as a gas jet into a space between
two mutually opposed attachment walls. The attachment of the jet to one of them is
by means of the well-known Coanda phenomenon of fluid jet clinging to a solid wall.
The jet is actually held deflected at the wall by the low pressure which is acting
also in the control nozzle on the same side of the amplifier. Since there is the feedback
loop channel connected to the control terminals, the pressure difference arises between
the ends of the feedback channel. This difference generates a flow in the channel.
Gas leaving the low-pressure end of the feedback channel forms a control jet that
acts on the supply jet and causes its separation from the attachment wall. The jet
cannot remain straight and after the separation therefore attaches to the opposite
attachment wall. These changes in the control nozzles generate in the feedback loop
channel alternating-direction flow. The control actions in the amplifier produce alternating
output flow in the two exit terminals of the diverter oscillator. Gas flow leaves
one output terminal in first half of the cycle and the other terminal in the rest
half of the cycle. According to
EP2081666, there is an aerator under the liquid surface connected - usually by means of a hose
- to each output terminal of the oscillator.
[0007] In existing patent literature are described inventions using the effect of oscillation
on the generated bubbles. Characteristic for these earlier inventions is the oscillator
action achieved by moving or deformed mechanical components. Even though this leads
to small size of generated bubbles, there are several disadvantages associated with
the mechanical movements or deformations. Typically, it is necessary to lubricate
the bearings holding the mechanical components, the contact or sealing surfaces, or
it is possible that the deformed component breaks as a consequence of material fatigue.
All these problems are removed with the fluidic oscillator. In particular, with the
use of the fluidic oscillator according to
EP2081666 is obtained a long life, reliability, and the absence of maintenance.
[0008] On the other hand, the solution according to the European Patent
EP2081666 is also not without some disadvantages.
[0009] These disadvantages follow from the recently discovered fact that the desirable small
size of generated bubbles is obtained in particular at low oscillation frequency.
The low frequency is also desirable also in those situations where the bubble size
is not significantly dependent of frequency. This is because the energy spent on generation
of each is roughly the same so that if the oscillation cycles are repeated at a low
frequency the power spent on them during a unit of time is lower. However, the fluidic
oscillator described in the Patent
EP2081666 is suited for high frequency oscillation. The frequency in that oscillator is determined
by two factors: by the magnitude of the air (or another gas) flow rate through it,
and by the length of the feedback loop channel.
[0010] This is associated with the following disadvantages:
- a) If the flow rates are small, the number of generated bubbles from a particular
oscillator is small; thus the economy of bubble generation is decreased.
- b) A more important factor are the friction losses in fluids which increase in relative
importance as the flow rate - and the corresponding Reynolds number values - are decreased.
The friction can increase to such a degree that at very low frequency the processes
inside the fluidic amplifier are damped and the amplifier loses its functional capability.
- c) At low oscillation frequencies the lengths of the feedback loop channels are excessive.
In the Patent EP2081666 is shown an example of the diagram of dependence of the frequency on the loop length
where the necessary length of the feedback channel was 50 metres and more. Cavities
so long cannot be manufactured together in a single production operation with the
cavities of the amplifier by the manufacturing methods usually used for the amplifiers,
like laser cutting or photochemical processes with etching the material of the cavities.
What come practically in question is to make the feedback loop channel in the form
of a long hose or tube fixed by its ends to the amplifier body. The necessary lengths
of the hoses or tubes are, however, extraordinarily impractical. The oscillator is
then no more - as it is at high frequencies - a compact entity but it is difficult
to stow. It is necessary to make or find suitable spaces into which the long lengths
of the hoses may be placed. Manufacturing of the fluidic oscillator is then more expensive
due to the manual work necessary to connect the hose and the ferrules on the amplifier
body. No connection of disparate entities like the hose and a solid body of the amplifier
is one hundred percent reliable, the oscillator with the hose is less robust and more
sensitive to various mechanical action: the connected hose may be accidentally removed
in any manipulation with the oscillator.
- d) Current soft materials of hosed and tubing, such as, e.g., rubber, lose their properties
with time. They exhibit a lower mechanical strength. It is necessary to incorporate
into the procedures some maintenance operations associated with replacement of weathered
or aged hoses - or it is necessary to choose higher quality hose material, which is,
of course, more expensive.
- e) In very long hoses used in the role of the feedback loop channel, sometime with
lengths of the order of metres and more, it is inevitable that the energy of the transported
fluid flow is decreased by friction on the hose walls - especially if the hoses are
not straight but have to be coiled for stowage. A considerable percentage of the transferred
feedback signals become lost. It is necessary, to incorporate these losses, to select
a higher working pressure of the air (or gas in general) supplied into the oscillator.
This increases the operation expenses, because compressed air is a quite expensive
commodity.
Disclosure of the invention
[0011] The disadvantages named above are removed by the unit of a generator of gas bubbles
in a liquid connected to the inlet of the gas into the vessel containing the liquid
according to this invention.
[0012] The subject of this invention is a unit of a generator of gas bubbles in a liquid
connected to the inlet of the gas into the vessel containing the liquid that has the
gas supply channel for the flow of the gas branched into at least two concurrent flowpaths
each of which contains six components connected in series, namely a nozzle, pre-chamber,
vortex chamber, central exit, distribution cavity, and a porous wall open into the
vessel with the liquid where the nozzle is directed by its mouth into the pre-chamber
and downstream from the nozzle is each flowpath inside the pre-chamber bifurcated
into two alternative routes, namely a tangential and a radial route, both entering
the vortex chamber, where the radial route is adjacent to an attachment wall directed
towards the central exit while in the diverting location at the beginning of the two
routes between the nozzle and the beginning of the attachment wall contains mouth
of the connection channel leading between the first flowpath and the second flowpath.
[0013] According to this invention the unit of a generator of gas bubbles in a liquid connected
to the supply of the gas into the vessel containing the liquid may be also characterised
by the fact that both the tangential route as well as the radial route lead from the
pre-chamber to the axisymmetric vortex chamber through a single common orifice and
the pre-chamber has opposite to the attachment wall a antipodal wall that is inclined
to it by an angle α larger than 16 angular degrees and further that between the antipodal
wall and the vortex chamber there is a protruding nose.
[0014] The purpose of this arrangement is simplification of manufacturing of the unit, especially
if it is made from a stack of plates with cavities made by removal of the plate material.
[0015] The unit of a generator of gas bubbles in a liquid connected to the supply of the
gas into the vessel containing the liquid may also have two orifices leading into
the same vortex chamber, namely one orifice for the tangential route and another orifice
for the radial route the said routes being formed each of them at a different side
of the splitter.
[0016] This alternative layout may cause difficulties in the manufacturing process, but
if the manufacturing problems are solved then this layout has the advantage of lower
energetic losses, in particular for the radial route flow, because the flow through
a closed conduit leads to lower pressure loss than a flow into which may be entrained
the surrounding fluid.
[0017] The unit of a generator of gas bubbles in a liquid according to this invention may
have inside the vortex chamber positioned a guiding blade shaped into an arch. It
was demonstrated that the presence of the guiding blade has a favourable influence
on the radial flow through the vortex chamber.
[0018] The unit of a generator of gas bubbles in a liquid connected to the supply of the
gas into the vessel containing the liquid according to this invention may have, if
the manufacturing aspects are solved, the upper wall of the vortex chamber and/or
the bottom wall of the vortex chamber the shape of a cone.
[0019] Unit of a generator of gas bubbles in a liquid according to this invention may be
advantageously made as a set of plates stacked on each other, containing the top plate
provided with exit holes and under it a woven metal textile the part of which under
the exit holes forms the porous wall of the distribution cavities where fastened to
the top plate under the metal textile is the distribution plate to which is fastened
partition containing central exits, the latter connected to the main plate to which
is fastened the bottom plate and with advantage the carrying pipe connected to the
top plate while the pre-chamber and the vortex chamber are mead in the main plate
and the distribution cavities are made in the distribution plate.
[0020] Making the cavities in a body by the method of stacking the body from plates is the
simplest way of manufacturing complex shaped cavities inside a solid body.
[0021] Conveniently the supporting ribs are placed between the exit holes in order to decrease
the mechanical stressing of the metal textile caused by the gas pressure.
[0022] It is possible to apply for manufacturing of the unit a generator of gas bubbles
in a liquid connected to the supply of the gas into the vessel containing the liquid
according to this invention various materials and manufacturing techniques, some of
which are mentioned in the following discussion of the examples of the unit.
[0023] The basic manufacturing problem is the rather complex shape of internal cavities.
[0024] It is advisable to make them by applying some computer controlled production methods
as is cutting by laser or controlled polymerisation. As for the materials suitable
for making the unit, there is a quite wide choice. The selection criterion is resistance
to corrosion and to mechanical stressing. Suitable from this point of view are stainless
steels or polymer materials. A stainless steel is the choice for material removal,
e.g. by laser cutting the cavities in plates, the "rapid prototyping" with computer
controlled polymerisation will be the choice method for manufacturing from a monomer
liquid. The "rapid prototyping" offers the advantage of making the whole unit in a
single manufacturing operation - while its disadvantage is a higher cost. Very advantageous
is the computer-controlled laser cutting from flat metal plates.
[0025] The unit according to this invention achieves new and higher effects than the previously
known versions due to the fact that it is particularly suitable for being made in
a very compact layout and with minimum spatial requirements, because it unites in
a single solid body, submerged under the liquid level, the oscillator as well as two
aerators immediately connected to the oscillator exits. The most important fact, however,
is easy achieving of low oscillation frequency of the oscillation due to the time
delays which are due to gradual spin-up of the fluid motion in the vortex chambers
during each cycle - and then equally slow stopping of the rotation in the subsequent
part of the cycle. The disadvantages are removed that are listed above as arising
in known oscillators from the necessity to have the very long hoses for the feedback
loops.
[0026] The most frequent envisaged application of the unit is producing tiny air bubbles
in processed waste water, where the actual processing is done by bacteria that in
present waste water processing plants die due to the lack of oxygen and cannot fulfil
their task completely. With the aeration by very tiny air bubbles the life of the
bacteria is longer and at the same time the smaller financial expenditure is needed
for supplying the compressed air. Similar advantages for providing more gas transfer
surface into the liquid are there in oxidative leaching of plutonium, photoresist
removal from silicon wafers, separation of various materials by froth flotation, yeast
production, sonochemical synthesis, salvaging crude oil from exhausted oil wells,
and growing unicellular organisms and algae as the basis of food chain.
Description of the drawings:
[0027]
Fig. 1 - The top plate of the unit in the first of the discussed examples.
Fig. 2 - View of the fist example of the unit without the top plate and without the
metal textile which in the fully assembled unit is under the top plate.
Fig. 3 - View of the first example of the unit with some of the parts removed so that
it is possible to se the part called partition.
Fig. 4 - View of the most important plate with cavities of the first example.
Fig. 5 - Schematic representation of the units valid for all discussed examples.
Figs. 6 and 7 - Trajectories of fluid flow in the cavities of the first example. Two
illustrations show two different basic function regimes.
Fig. 8 - The most important details of the component of the first example.
Fig. 9 - Flow in the cavities of the first example obtained by computer flowfield
solution.
Fig. 10 - Schematic representation of the pressure distribution along the two flowpaths
of the gas passing through the unit.
Fig. 11 - A part of the alternative example differing in the splitter between the
two routes leading into the vortex chamber.
Fig. 12 - Another of a part of the unit made by stereolithography.
Fig. 13 - Yet another example of the unit made by stereolithography.
Fig. 14 - An example of a part of the unit made by selective laser sintering.
Fig. 15 - An example of a part of the unit containing the guiding blade.
Fig. 16 - Computed trajectories of the gas flow in the pre-chamber as well as vortex
chamber of the example with the guiding blade.
Fig. 17 - Another example of the unit with easier flow through the vortex chamber
in the regime without rotation.
Fig. 18 - Another alternative example of the vortex chamber.
Fig. 19 - Yet another example of the vortex chamber in the unit according to this
invention.
Examples
Example 1
[0028] The unit according to this invention as shown in Figs. 1 to 4, is an example manufactured
by the plate stack method. The necessary cavities inside the unit body are manufactured
by material removal separately in the plates. Operation of the unit - as is the case
in all the examples of the unit discussed below - depends on oscillatory gas flow
with the oscillation generated in cavities inside a solid body. There are several
methods how to make the cavities and the stack method is one of them. The cavities
are made separately in each flat plate of the stack by the known methods like laser
cutting or electric discharge machining. Then the finished plates are stacked and
held together. They may be welded together to form a single solid body - or they may
be held by screws or similar fasteners allowing disassembling the unit.
[0029] The particular version presented the first four pictures is intended for producing
air bubbles in waste water, such aeration being an important step in the wastewater
processing. Without significant changes this unit may be used in other cases of aeration
of a liquid. The unit consists of five stainless-steel plates of equal outer shape
and, in addition, there is also a thin mesh made by weaving from very thin stainless
steel wires (in this particular case the wires are of 40 µm diameter). This woven
steel textile layer is of the same external shape as the other plates in the stack,
inside which it is clamped between the uppermost plate and the plate immediately below.
The unit is in operation wholly submerged in the liquid, held from above by the carrying
pipe
11. The plates forming the unit are held in a horizontal position horizontal. The internal
cavity of the carrying pipe
11 serves as the supply channel
1 bringing compressed air into the unit. In the form of bubbles, the air then leaves
the unit into water through the above mentioned mesh.
[0030] Mutual position of all plates in the unit is secured by two dowels inserted into
the dowel holes
902 made at the same position in each plate. There are also nine screw holes 901, also
at the same position in each plate. Stainless-steel crews passing through the screw
holes
901 keep the stack together and secure the clamping force holding the mesh.
[0031] The top plate
95 removed from the unit is shown in Fig. 1. The carrying pipe
11 (only a short segment of which is shown in Fig. 1) is welded to the top plate 95.
Apart from the nine screw holes
901 and two dowel holes
902 there are in the top plate 95 also fourteen exit holes
950 of roughly hexagonal shape. These exit holes
950 expose the metal-textile layer, which is held under the top plate
95. Some areas of this layer, located under the exit holes
950, thus form the porous walls
9a,
9b of the distribution cavities
8a,
8b made in the distribution plate
94. Between the exit holes
950 there are supporting ribs 951. They support the thin metal textile layer and prevent
its damage by the force action of compressed air, which is brought to the textile
from below.
[0032] In the picture of the unit presented in Fig 2, both the top plate
95 and the metal textile layer are removed, which allows seeing the distribution plate
94. Also the screws and dowels are removed in Fig. 2. Apart from the holes for these
fastening components, there are in the distribution plate
94 two holes, made together with other material removal in a single manufacturing operation
by programmed laser cutting. These two holes are star-shaped, with blunt tips of the
star. At left this hole forms the first distribution cavity
8a and at right it forms the second distribution cavity
8b. By comparison with Fig. 1 it is apparent that the distribution cavities
8a,
8b are each under the seven exit holes 950. The air from the distribution cavities
8a,b thus can escape upwards through the metal textile supported by the supporting ribs
951. In addition, in the top part of Fig. 2 is seen a round hole which is a part of
the supply channel
1. In the bottom part of Fig. 2 there is an arc-shaped interconnection channel 941.
[0033] Shown on the next Fig. 3 is the unit with also the distribution plate
94 removed. This makes possible to see well the partition
93. It is thinner than other plates, also made of stainless steel and with the same
outer shape as the shape of other plates. Apart from the nine screw holes
901 and two dowel holes
902 there are in the partition
93 altogether five round holes. On top is again visible the hole that forms a part of
the supply channel
1. On the right-hand side there is the first central exit
7a and on the opposite left-hand side there is the second central exit
7b. Then there are also the interconnection holes
931 which in the assembled unit are each under one end of the interconnection channel
941 in the distribution plate
94.
[0034] It remains only to remove the partition
93, as is shown in the following Fig. 4, and this makes possible to see well the most
important plate from the stack - the main plate
92. Under this main plate
92 is only the bottom plate
91 which is not shown in a separate illustration because it does not contain anything
of particular interest: its outer shape is the same as that of other plates and there
are only nine screw holes
901 and two dowel holes 902.
[0035] In the top part of Fig. 4 is in the main plate
92 a hole which is a part of the supply channel
1. Two concurrent flowpaths are branching from it, one at left and the other on the
right-hand side. The right-hand side air flow path leads through the first pre-chamber
4a into the first vortex chamber
6a - while, almost symmetrically, on the left-hand side there is connected to the supply
channel
1 the second pre-chamber
4b and the second vortex chamber
6b. By comparison with Fig. 3 it is apparent that air may leave the first vortex chamber
6a through the first central exit
7a made in the partition
93 and, on the other side, air may leave the second vortex chamber
6b through the second central exit
7b in the partition
93. To the inlet parts of the pre-chambers
4a, 4b leads the connection channel
20 which in the main plate
92 consist of two parts which both end, at the bottom of Fig. 4, in those locations
where the partition
93 (as seen in Fig. 3) has its interconnection holes 931. These holes make possible
- together with the interconnection channel
941 in the distribution plate
94 - union of both parts of the connection channel
20 into a single continuous channel. This connection through other plates is useful
because it evades formation in the plates during their manufacturing of " islands"
- the unsupported parts of the original plate that fall out in the material removal
process and would necessitate expensive repositioning and fixing, a manual operation.
[0036] The gas flow is divided in pre-chambers
4a, 4b into two alternative flow routes, the tangential route
5a and radial route
5b. In Fig. 4 there are in the two flowpaths
2a, 2b two different flow conditions. In the first pre-chamber
4a the gas follows the first tangential route
5aa while in the second pre-chamber
4b the gas takes the second radial route.
[0037] For description and explanation of the processes that take place inside the unit
it is useful to note the schematic representation of the air flows in the following
Fig.5. In principle this is a diagram of topological structure which is the same in
all cases of the unit according to this invebntion. The vertical straight lines indicate
some important locations. At the left-hand side of the picture Fig. 5 there is the
supply channel
1 where gas (air) enters the unit. The illustration is oriented so that the gas (air)
passes through the schematic representation from left to right. Represented schematically
at the right-hand side are thus the two location where the gas leaves the unit and
enters the vessel containing the liquid in which the bubbles are produced. This departure
of the gas takes place through the first porous wall
9a of the first distribution cavity
8a on one side - and through the second porous wall
9b of the second distribution cavity
8b on the other side. So that they can mutually exchange their role during each operating
cycle, the porous walls
9a,
9b are placed at the ends of two concurrent paths, the first flowpath
2a and second flowpath
2b into which the supply channel
1 bifurcates. In Fig. 5 the first flowpath
2a is in the top part of the picture while in the bottom half is the second flowpath
2b. Both flowpath pass through nozzles. The first flowpath
2a passes through the first nozzle
3a and the second flowpath
2b through the second nozzle
3b. Air issued through both nozzles
3a, 3b into the respective pre-chambers: the first nozzle
3a into the first pre-chamber
4a and the second nozzle
3b into the second pre-chamber
4b. Immediately downstream from the pre-chambers
4a, 4b there are diverting locations
10a,
10b where secondary branching takes place. The two alternative routes downstream from
this diverting locations
10a,
10b have different purposes and essentially should not be active simultaneously. Shown
in Fig. 5 is a particular phase of the oscillation in which the gas passing through
the first flowpath
2a follows after leaving the first nozzle
3a the first tangential route
5aa through the pre-chamber
4a and into the first vortex chamber
6a. It does not use the available first radial route
5ba. In the second flowpath
2b the reverse is true at this phase. The gas leaving the second nozzle
3b attaches to the second attachment wall 14a which leads it through the second radial
route
5bb into the second vortex chamber
6b. Tangential routes
5aa,
5ab and radial routes
5ba,
5bb meet again in the vortex chambers
6a,
6b. For example, the gas following the first flowpath
2a leaves the first vortex chamber
6a through the first central exit
7a into the directly to it connected first distribution cavity 8a and through the first
porous wall
8a into the vessel containing the liquid. The attachment of the gas to the second attachment
wall
14b shown in the second flowpath
2b - and, on the other hand, the absence of such attachment to the first attachment
wall
14a - is controlled by alternating flow in the connection channel
20 and its flow out from the second mouth
13b. Decisive role is played by pressure distribution in the cavities of the unit.
[0038] On the two following illustrations, Fig.6 and Fig. 7 are shown in identical perspective
views the results of numerical flowfield computations showing trajectories of gas
particles in the pre-chambers
4a, 4b and vortex chambers
6a,
6b. The geometry of the chambers corresponds exactly to that shown in Fig. 4. The computations
were made for everywhere the same depth of the cavities, equal to the thickness 4
mm of the main plate
92 and for the 70 mm diameter of the vortex chambers. The width of the nozzle
3a, 3b in this case was 2.56 mm. In the situation shown in Fig. 6 the gas, leaving the first
nozzle
3a follows the tangential route
5a while in Fig. 7 the gas follows the radial route. In the original presentation of
the computation results, the gas trajectories were colour coded so that the local
colour corresponded to the local pressure according to the scale at left of the pictures.
In the conversion into the grayscale this information is not easily recognised, nevertheless
the most important are the pressure values in the two mouths
13a,
13b, shown in these illustrations. In both cases, the computations were performed for
the same value
of the pressure in the supply channel
1 relative to the atmosphere which in the tests were outside of the porous walls
9a,
9b. In the case presented in Fig. 6 air enters into the first vortex chamber
6a by the tangential route
5a. It is directed by the nose
46 into the tangential direction and the moment of momentum thus gained generates in
the first vortex chamber
6a intensive gas rotation. The flow towards the first central exit
7a is difficult. It may be noted in Fig. 6 that this situation requires additional flow
from the first mouth
13a (the corresponding trajectory is recognisable in Fig. 6) joining the flow from the
first nozzle
3a and actually deflecting it from its original direction of flow from the first nozzle
3a.
[0039] If, however, the gas leaving the second nozzle
3b attaches - as shown in Fig. 7 - to the second attachment wall
14b, this wall direct it without rotation towards the second central exit
7b. The computed results actually show somewhat more complex situation - it is not easy
for the gas to change its direction of flow by 90 degrees to enter the second central
exit
7b. Nevertheless, even with this complication the pressure in the second mouth
13b is low and in fact the difference relative to the atmosphere is negative. This is
possible due to the fact that a large part of the gas energy is converted into kinetic
energy.
[0040] In the following Fig. 8 are presented parts of the first flowpath
2a consisting of_the first nozzle
3a , the first pre-chamber
4a as well as the first vortex chamber
6a. The illustration shows some most important aspects of the planar geometry, which
were found by computations with alternative geometric features. It is shown here,
that if the first diverting location
10a into the first tangential route
5aa and the first radial route
5ba is to be as requested in the first pre-chamber
4a, then it is necessary do select a large angle α between the first attachment wall
14a and the first antipodal wall
15a. As is known from the behaviour of diffusers, if the divergence angle between walls
is smaller than α = 16 deg, than the floe leaving the first nozzle
3a can follow both walls simultaneously. The successful behaviour as shown above in
Figs. 6 and 7 was achieved with the angle twice as high than this generally accepted
limit: the angle there is α = 32 deg.
[0041] The next important fact is the necessity of the first attachment wall
14a oriented so that the extension 214 of its end is directed exactly into the centre
of the first vortex chamber
6a. Also of importance is shaping of the nose
46 and a proper width of the first orifice
146a, which in the case of Figs. 6 and 7 was equal to five widths of the exit from the
first nozzle
3a. For proper directing of the first radial route
5ba is was also important to form the shallow recession
206 in the wall of the first vortex chamber
6a.
[0042] In the following Fig. 9 is demonstrated the asymmetry of the flowfield in the cavities
arising despite the fact that their shape is symmetric relative to the symmetry axis
100. The picture presents computed flow in the cavities of the main plate
92 in both flowpaths
2a, 2b at the extreme situation arising once in each oscillation cycle. In a manner similar
to the previous illustrations, also here the computed trajectories of gas are coloured
by the local pressure according to the scale at the left-hand side of the picture.
Unfortunately, the monochromatic greyscale representation makes the pressure distribution
less apparent. Both flowpaths
2a, 2b are connected to the same supply channel
1 so that the driving pressure is the same driving pressure
relative to the atmosphere.
[0043] The following Fig. 10 shows schematically the pressure conditions along the two flowpaths
2a,
2b. In the horizontal direction are indicated individual locations while on the vertical
axis is plotted the pressure difference ΔP relative to the atmosphere. The situation
represented corresponds to the case in Fig. 9: full rotation in the first vortex chamber
6a and the easy, essentially non-rotational flow in the second flowpath
2b. Different pressure values in the two flowpaths
2a, 2b in the same locations are due to two factors. The first one is the intensity of the
gas flow. The second is the effect of the large pressure drop on the first vortex
chamber
6a with the rotation in it. Even though both nozzles
3a, 3b are geometrically the same, due to the easy character the flow rate in the second
flowpath
2b is much larger - and, consequently, the pressure drop on the second nozzle
3b is much larger. On the other hand, due to the large pressure drop in the first vortex
chamber
6a the flow rate through the first nozzle
3a is much smaller. In both flowpaths
2a, 2b a small pressure increase takes place in the upstream part of both pre-chambers
4a,
4b. This is an effect of pressure recovery in the slightly decelerating flow downstream
from the exit from the nozzles
3a,
3b.
[0044] The atmospheric pressure level is marked by "0" at the right-hand side of the picture
Fig. 9. While in the first mouth
13a of the connection channel
20 the pressure difference relative to the atmosphere is
- in the second mouth
13b the pressure difference is negative (cf. Fig. 10)
[0045] The total pressure difference between the two ends of the connection channel
20 is
[0046] It is this difference that causes the flow in the connection channel
20. An important fact is the difference ΔP
J is actually larger than the supplied ΔP
S. This is a welcome factor, increasing the intensity of the flow rate in the connecting
channel
20.
[0047] Despite the fact that the geometry of the two flowpaths
2a, 2b is completely symmetric, the flow in them are asymmetric. The different conditions
in the vortex chambers
6a, 6b are the usual situation. Different direction of the entrance flows into the vortex
chambers
6a, 6b generated different pressure conditions and these lead to air flow in the connection
channel
20. In the situation shown in Figs. 9 and 10 the flow in the connection channel
20 will be directed from its first mouth
13a, with higher pressure, to the second mouth
13b, where in this flow configuration the pressure is lower. This stops in the first
pre-chamber the deflection effect that was seen in Fig. 6. The air flow leaving the
first nozzle
3a stops following the tangential route
5a and sooner or later attaches to the first attachment wall
14a. The entrance flow into the first vortex chamber
6a will cease to be tangential. At the other end of the connection channel 20 the outflow
from the second mouth
13b will cause the flow deflection away from the second attachment wall
14b. The flow into the second vortex chamber
6b will cease to be radial. The first flowpath
2a and the second flowpath
2b will exchange their respective roles. This, however, does not take place immediately.
The vortical motion in the first vortex chamber
6a will tend to keep its moment of momentum. On the other hand, spinning up the rotation
in the second vortex chamber
6b also takes some time. The exchange of the radial route
5b and the tangential route
5a will be not instantaneous but will proceed with a time delay - despite the relative
shortness of the connection channel
20. When the exchange finally takes place, the conditions will be a mirror image relative
to the symmetry axis 100. The air will rotate in the second vortex chamber
6a and will find an easy way out from the first vortex chamber
6b. This stops the previous flow in the connection channel
20 and then reverses the flow in it. As a consequence of the flow reversal, the roles
of the vortex chambers
6a, 6b will exchange again - coming, after a certain delay, to the initial regime. The process
then can start another cycle. Each time the air rotates in the vortex chamber
6a, 6b of a particular flowpath
2a, 2b the air flow through the respective porous wall
9a, 9b decreases. The consequent flow and pressure pulsation influence the formation of
the air bubbles in the exits from the pores of the porous wall
9a, 9b. In contrast to the known bubble generator disclosed in
EP2081666 there are no feedback tubes of the order of metres and the unit according to the
present invention is thus very compact.
Example 2
[0048] In the next Fig. 11 is shown an alternative layout of the pre-chamber
4a, 4b. The picture shows the first flowpath
2a and a part of the second flowpath
2b, which is of similar layout (it is a mirror image with respect to the symmetry axis
100, Fig. 9). The geometry of cavities is very similar to the example presented in Fig.
4. As there, the unit is a planar layout made from a stack of plates and what is shown
in Fig. 11 is the main plate
92, made in a constant-thickness plate with laser-cut cavities. The layout is very similar
to Fig. 4. There is the supply channel
1 at left bifurcating into two flowpaths
2a, 2b - of the latter is drawn in Fig. 11 only a part. The first flowpath
2a leads through the first nozzle
3a and the second flowpath
2b leads through the second nozzle
3b. Nozzles
3a, 3b have their exits open into the pre-chambers
4a, 4b. One of the pre-chamber walls is the attachment wall
14a, 14b where there are the secondary bifurcations into the radial route
5b and the tangential route
5a, both leading into its vortex chamber
6a, 6b. The radial route
5b in the first flowpath
2a is a direct continuation of the first attachment wall 14a of the first pre-chamber
4a and like in the layout shown in Fig. 4 the attachment of the air jet issuing from
the first nozzle
3a is facilitated by the first nozzle
3a exit direction coinciding with the direction of the first attachment wall
14a. The basic difference between the layouts in Fig. 4 and Fig. 11 is there is only one
orifice between the first pre-chamber
4a and the first vortex chamber
6a in Fig. 4 while here in Fig. 11 there are two such orifices: the first orifice
146aa and the second orifice
146ba. In other words, the tangential route
5a has its separate orifice for entry into the first vortex chamber
6a. If there should be a rotation in the first vortex chamber
6a, the gas jet leaving the first nozzle
3a is to be directed into the tangential route
5a - and this means, equally as in the configuration from Fig. 4, that this jet must be
separated from the first attachment wall 14a. Here in Fig. 11 the sense of the rotation
generated in the first vortex chamber
6a after this separation is opposite to the rotation sense in the case of Fig. 4. The
tangential route
5a is separated from the radial route
5b by the splitter
50. The straight flow through the orifice for the tangential route
5a is associated with lower hydraulic losses than in the case with the turning of the
flow direction in Fig. 4. In the present case, however, the higher losses are actually
welcome in the regime with air rotation in the vortex chamber
6a, 6b since it is desirable to have high pressure in the mouth
13a, 13b. Also, there is yet another advantage of the layout with the nose
64 and absence of the splitter
50: in manufacturing by the method of stacked plates the absence of the splitter
50 means there are no "islands" - those parts that fall out from the machined plate,
which it is later necessary to put back and fix them in their proper position - a
manual manufacturing operation which increases the cost of the unit.
Example 3
[0049] Recently introduced manufacturing methods, know collectively as "rapid prototyping"
or also "three-dimensional printing", offer an alternative to the manufacturing by
the above described method of material removal from flat plates that are then stacked.
As opposed to the material removal, these alternative methods are based on three-dimensional
computer-controlled addition of the material that forms cavity walls. A typical case
is, e.g., the stereolithography in which the walls are grown from a monomer liquid
that solidifies - polymerises - upon laser light activation. Another possibility is
selective laser sintering in which the body of the unit is formed from powder added
in thin layers to the wall locations after which the powder particles are fixed by
heating them, by laser, up to the partial melting temperature.
[0050] In this third example of the bubble generating unit, presented on Fig. 12, is applied
the stereolithography method. The advantage gained is more freedom in the choice of
shapes of the internal cavities. Otherwise the configuration corresponds exactly to
the schematic representation shown in Fig. 5. The particular problem solved by this
manufacturing approach is achieving very low pressure drop across the vortex chamber
6a, 6b if the gas enters it by the radial route
5ba, 5bb. As demonstrated by the computational results presented in Figs. 7, and 9, fluid inertia
makes it difficult for the radial inflow into the vortex chamber
6a, 6b upon reaching its central exit
7a, 7b to change suddenly its flow direction by 90°. This complicates the exit from the
vortex chambers
6a, 6b and increases the pressure drop across the second flowpath
2b. The freedom offered by the stereolithography is used in this example to shape the
vortex chambers
6a, 6b so that the requested change in the flow direction is by an angle less than by 90°
.
[0051] Figure 12 shows a part of the first flowpath
2a involving the first pre-chamber
4a and the first vortex chamber
6a. The cavities are shown in imagined meridian-plane section through the body of the
unit. It is immediately apparent that - instead of the flat disk geometry in the previous
examples Fig. 4 and Fig. 11 - here the first vortex chamber
6a is of conical shape, decreasing in diameter towards the apex of the cone. In the
regimes without rotation in the first vortex chamber 6a the necessary turning angle
for entry into the first central exit
7a is only 45° rather than the full 90°. Such a flow is easier, i.e. producing a lower
hydraulic loss, which leads to a higher desirable pressure drop ΔP
J that drives the flow in the connection channel
20.
[0052] Of course, the manufacturing methods known as "rapid prototyping" need not be generally
accessible. They may also be costly and their special requirements on the character
of the material easily polymerised from a liquid may be not compatible with the mechanical
requirements of the body of the unit. After all, at present the methods are really
used just to produce a prototype verifying the geometrical spatial conditions. In
the cases where the "rapid prototyping" methods are not suitable, the complex internal
cavities of the unit may be made by the classical method of "lost wax" casting.
Example 4
[0053] In this next example, the freedom offered by the stereolithographic manufacture is
used even more. Again, the overall configuration of the unit corresponds to the schematic
representation in Fig. 5 and the part that is shown in Fig. 13 shows a section through
the body of the unit involving the first pre-chamber
4a and the first vortex chamber
6a. The difference is in the shape of the first central exit
7a. It is here not coaxial with the vortex chamber axis
101 as was the case in Fig. 12 but it is here inclined so that its axis is in line with
the first radial route
5ba flow. In the regime of gas flow following the first tangential route
5aa, the nose
46 directs the gas to enter the first vortex chamber
6a tangentially, as is shown by the gas trajectory with arrows. The conditions with
the rotation do not differ very much from the conditions inside the cylindrical flat
disk geometry in the previous examples Fig. 4 and Fig. 9. The centrifugal action makes
the rotating flow difficult - and the inclined first central exit
7a makes the pressure drop even higher. On the other hand, the straight flow following
the first radial route
5ba causes much lower overall pressure drop.
Example 5
[0054] The manufacturing method of selective laser sintering from powder makes possible
making the manufactured product not only with solid walls, but also in a selected
part of the manufactured object to make the cavity walls with tiny pores. In the example
presented in Fig. 14, in a section by a meridian plane of the first vortex chamber
6a this is used for producing in a unit body simultaneously with making the solid walls
also the porous walls
9a, 9b of the distribution cavity
8a,
8b, needed for generation of the gas bubbles. In Fig. 14 the first porous wall
9a is shown arched inwards into the first distribution cavity
8a. This allows making the first porous wall
9a thinner and easier for the air to pass through it without unduly stressing the sintered
first porous wall
9a in tension by the force action of the compressed air. The dome-shaped first porous
wall
9a is stressed by the acting pressure difference in compression, which it can better
withstand.
Example 6
[0055] Easier the entry from the vortex chamber
6a, 6b into the central exit
7a,
7b by the air in the regime with the flow following the radial route
5ba,
5bb may be achieved by alternative shapes of the cavities not necessary requesting unusual
manufacturing methods. The example presented in Fig. 15 achieves the desirable effect
by the presence of the guiding blade
16 positioned inside the vortex chamber
6a, 6b not far from the central exit
7a,
7b. The shape of the guiding blade
16 is derived from the shape of the air trajectories in the particular location inside
the vortex chamber
6a, 6b in the regime with air rotation as shown in Fig. 6. The guiding blade
16 is very thin so that it does not in that regime produce a significantly large wake
downstream from it. Thus its presence plays practically no role in the rotational
regime. On the other hand, it becomes important in the regime with the air flow following
the radial route
5ba,
5bb. In that regime the guiding blade
16 prevents the air flow from reaching the vortex chamber
6a, 6b wall opposite to the radial entrance, as it is seen in Figs, 7 or 9. The air is obviously
forced by the blade
16 to enter the central exit
7a,
7b. This is proved by computations the result of which are the air flow trajectories
shown in Fig. 16. Even though the view in this picture is a perspective view from
another position than in Fig. 7 (the view angle was chosen to make more apparent the
guiding blade
16), it is essentially the same regime as in Fig. 7. The trajectories in Fig. 16 show
how the air has no other choice but to enter the central exit
7a,
7b - which leads to lower pressure drop across the vortex chamber
6a, 6b, as demonstrated by the computed pressure difference value included into Fig. 16:
- significantly lower than ΔP
b = -1 790 Pa obtained under otherwise the same conditions without the guiding blade
16 in Fig. 7.
Example 7
[0056] Another possibility how to force the gas to change its flow direction in the centre
of the first vortex chamber
6a and induce it to enter the first central exit
7a is shown in Fig. 17. In place of a single first pre-camber
4a considered in all alternatives discussed above, there are here many pre-chambers
distributed evenly on the perimeter of the first vortex chamber
6a so that first radial routes
5ba leaving these pre-chambers collide. There is, in Fig. 17, the first perimeter pre-chamber
4a1 on top at left, the second perimeter pre-chamber
4a2 on the left side below, the third perimeter pre-chamber
4a3 at right on top, and finally the fourth perimeter pre-chamber
4a4 bottom right. All of them are the same so that the stagnation point of their collision
is exactly in the centre of the first vortex chamber
6a. The air flow stops in this point or slows down in its vicinity and thus the pressure
rises there (by conversion from the kinetic energy). The problem with turning the
flow direction of a fast flow disappears completely, air is driven through the first
central exit
7a.
Example 8
[0057] In the next example Fig. 18 presents the layout of the first vortex chamber
6a in principle similar to the previous example in Fig. 17 where, however, the four
perimeter pre-chambers
4a1, 4a2, 4a3, 4a4 are replaced by altogether sixteen inlets. All the air supplied by way of the first
flowpath
2a into the sixteen first nozzles
3a is led into a single annular space so that the body in which in Fig. 17 are the pre-chambers
is here in Fig. 18 disintegrated into sixteen bodies
1000. Each
body 1000 has its first attachment wall
14a to which after leaving the first nozzles
3a attach the first radial routes
5ba directed towards the first central exit
7a. Also, each
body 1000 has its first antipodal wall 15a the end of which is directed tangentially into the space
that forms the first vortex chamber
6a. Since in the centre of the first vortex chamber
6a collide altogether sixteen radial inflows (of which only a single representative
is shown in Fig. 18) the symmetry of flows is secured and with it also the low pressure
drop across the first vortex chamber
6a in the regime with absence of rotation. As soon as air from the connection channel
20 starts to flow from the first mouths 13a, the air flow through the first nozzles
3a is separated from the first attachment walls 14a and follows the first antipodal
wall 15a. These walls guide it - in a manner similar to the first tangential route
5aa presented in Fig. 11 - tangentially into the first vortex chamber
6a. Due to the large number of tangential inflows also this regime is characterised by
welcome symmetry of the flowfield.
[0058] Obviously, the first vortex chamber
6a arranged according to Fig. 18 has all prerequisites for high efficiency, i.e. on
one hand the very low pressure drop in the radial flow regime and on the other hand
very high pressure drop in the tangential flow regime with rotation. The only disadvantage
of this layout is the large diameter, especially with the system of the bodies
1000 on its outer circumference an, in addition, the annular space for air distribution
into the first nozzle
3a.
Example 9
[0059] In some applications calling for compactness of the unit this large diameter may
be a disadvantage and a more compact layout may be in demand. One solution is presented
in Fig. 19. Essentially, the principle of operation is the same as in Fig. 18. The
radial layout of the bodies
1000 in Fig. 18 is here in Fig. 19 replaced by an axial layout. The picture shows one
from the two identical vortex devices in a unit. At the left-hand side in Fig. 19
there is the entrance by axial first flowpath
2a which leaves, also axially, on the right-hand side. The long conical diffuser (that
improves the ratio of the pressure drops in the two regimes) is the first central
exit
7a from the first vortex chamber
6a. The first vortex chamber
6a is of flat cylindrical shape. The first distribution cavity
8a with its first porous wall
9a connected directly to the first central exit
7a are not shown here.
[0060] The first vortex chamber
6a is here formed between the outer shell (only one half of which is shown, the other
half being removed by imagined meridian section) and the axially symmetric central
body
2000 similarly shown with one half removed. Coming from the left-hand side, the first
flowpath
2a passes through the annular space between the outer shell and the central body
2000. Positioned in this annular space are bodies
1000 with the first mouth 13a in each of them. On the inflow side, the spaces between
the bodies
1000 form the first nozzles
3a. Exit from the first nozzles
3a is directed axially, i.e. in parallel with the axis of the device. Further downstream,
in the direction of the exit from the first nozzle
3a, one of the sides of the bodies
1000 forms the first attachment wall
14a. The opposite first antipodal wall
15a wall of bodies
1000 is inclined and curved so as to lead into the first vortex chamber
6a tangentially. Between this wall and the attachment wall 14a is inserted the splitter
50 (similar as shown in Fig. 11).
[0061] If a small pressure drop across the first vortex chamber
6a is requested, there is to be no air flow issuing from the first mouth 13a. The air
of the first flowpath
2a enters the first nozzle
3a , is accelerated there and directed to the first attachment wall
14a. This guides it axially and then, following the downstream side of the central body
2000 it is turned into the radial inflow into the first vortex chamber
6a where there is no rotation. The pressure in the first mouth
13a is very low - and it is further decreased by the conversion taking place in the diffuser
of the first central exit
7a.
[0062] This low pressure would induce an air flow into the connecting channel
20 (not drawn in Fig. 19) leading towards the first mouth 13a. from the second mouth
13b where the pressure is higher. When the flow in the connecting channel
20 gains momentum, it will assume a magnitude sufficient for separating the axial air
flow from the first attachment wall
14a and switching it towards the first antipodal wall 15a thatg guides it into the tangential
route
5a past the splitter
50. This tangential inflow starts the rotation in the first vortex chamber
6a. Pressure difference across it increases. This is leads to flow direction reversal
in the connecting channel
20 and subsequently to the next half of the pulsation cycle.
Industrial applicability
[0063] Small gas bubbles in a liquid, that may be economically generated by the unit according
to this invention, us desirable in a large number of industrial processes where already
now air (or gas in general) bubbles are already made and used. The decrease of the
bubble size means that for a given volume of air is much larger the overall surface
area across which gas diffused into the liquid - and the smaller size also decreases
the velocity of bubble rising up to the liquid surface. As a result the gas diffusion
is intensified and this is done by compact units characterised by economical operation.
The most important envisaged application of the unit is producing tiny air bubbles
in processed waste water, where the actual processing is done by bacteria that in
present waste water processing plants die due to the lack of oxygen and cannot fulfil
their task completely. With the aeration by very tiny air bubbles the life of the
bacteria is longer and at the same time the smaller financial expenditure is needed
for supplying the compressed air. Similar advantages for providing more gas transfer
surface into the liquid are there in oxidative leaching of plutonium, photoresist
removal from silicon wafers, separation of various materials by froth flotation, yiest
production, sonochemical synthesis, salvaging crude oil from exhausted oil wells,
and growing unicellular organisms and algae as the basis of food chain.
List of identification numbers :
[0064]
1 |
supply channel |
2a |
first flowpath |
2b |
second flowpath |
3a |
first nozzle |
3b |
second nozzle |
4a |
first pre-chamber |
4a1 |
first perimeter pre-chamber |
4a2 |
second perimeter pre-chamber |
4a3 |
third perimeter pre-chamber |
4a4 |
fourth perimeter pre-chamber |
4b |
second pre-chamber |
5a |
tangential route |
5b |
radial route |
5aa |
first tangential route |
5ba |
first radial route |
5ab |
second tangential route |
5bb |
second radial route |
6a |
first vortex chamber |
6b |
second vortex chamber |
7a |
first central exit |
7b |
secondfirst central exit |
8a |
first distribution cavity |
8b |
second distribution cavity |
9a |
first porous wall |
9b |
second porous wall |
10a |
first diverting location |
10b |
second diverting location |
11 |
carrying pipe |
13a |
first mouth |
13b |
second mouth |
14a |
first attachment wall |
14b |
second attachment wall |
15a |
first antipodal wall |
15b |
second antipodal wall |
16 |
guiding blade |
20 |
connection channel |
46a |
first nose |
46b |
second nose |
50 |
splitter |
91 |
bottom plate |
92 |
main plate |
93 |
partition |
94 |
distribution plate |
95 |
top plate |
100 |
symmetry axis |
101 |
vortex chamber axis |
146a |
first orifice |
146b |
second orifice |
214 |
extension |
206 |
recession |
901 |
screw holes |
902 |
dowel holes |
931 |
interconnection holes |
941 |
interconnection channel |
950 |
exit holes |
951 |
supporting rib |
1000 |
bodies |
2000 |
central body |