Acknowledgement: Research is subsidized by
J04/98: 212200008
15 x 21
1.
INTRODUCTION:
Parallel flows are typical for many apparatuses e.g.
shell&tube or plate heat exchangers, heaters,
reactors. Sometimes instabilities or non-uniform
distribution of flow in parallel channels occur if the
apparatus operates at non-isothermal conditions.
Parallel flow instabilities have been observed also in
lateral channels of direct ohmic heater.
2.
R.Žitný, J.Thýn
Department of Process Engineering
CTU in Prague, Faculty of Mechanical Engineering
DIRECT OHMIC HEATER:
Volumetric heat source enables in principle more
uniform heating, important e.g. for food processing
(sterilization). Problems are with overheating at
walls, e.g. at the surface of electrodes which must
be cooled. The current design makes use processed
liquid flowing in lateral channels for the electrodes
cooling. Electrodes of heater are perforated - cross
flow from lateral channels should improve
uniformity of temperatures in central channel and
uniformity of residence times. This design is
suitable for very viscous liquids (sauces, juices,....),
i.e. for creeping flow (Re<<1) when natural
convection can be neglected.
E-mail: zitny@fsid.cvut.cz
5. CONTROL VOLUME MODEL
Cross-flow through perforation has been respected in a simple
integral model, based upon
•Mass balances
•Heat transfer
•Momentum
•Tracer balances
of control volumes characterised by:
•parabolic velocity profile (experimentally evaluated u /u =1.73 at 30 ml/s, theory 1.744)
•linear increase of temperature
•residence times - serie of mixers
Responses to tracer injection will be compared with experiments.
max
mean
6. EXPERIMENTS
Stimulus - response technique (injection of a tracer and
measurement responses) has proved to be useful for detection of
cross-flow. As tracers KCl (conductivity method), KMnO4
(visualisation), Tc99 (radioisotope) were used. Asymmetries of
flow are better observed by thermometers arranged along the lateral
channels.. T0,p0,Q
Flow-rate 32 ml/s, H=8 mm, full electrode
There are problems with heating of low viscous liquids (e.g.
water, milk), associated with influence of buoyancy and
natural convection.
• ASYMMETRY of parallel flows in lateral channels. If
one of parallel flows slows down its temperature increases,
and liquid in this channel becomes lighter. This creates
driving force (pressure difference) promoting circulation
from “fast“ channels to the “slow“ channel thus increasing
initial small disturbance. One parallel stream is delayed or
even stopped if the temperature increase is too high.
• CROSS-FLOW suppression. Warmer liquid in the
central channel is lighter than liquid in lateral channels and
corresponding pressure profile forces liquid to flow
through perforation in opposite direction (out of the
heating zone).
Warm liquid (Te) stands
in the right channel if
p1-p0=0gL[1-(Te-T0)]
40
35
T [C]
3. PROBLEMS - BUOYANCY
L
R1(left up)
R2 (right up)
R3 (left middle)
R4 (right middle)
T3 (left electrode)
T4 (right electrode)
30
25
20
15
0
2
4
6
Q [kW]
H
Narrow channel (H=8mm)
without heating. Very large
cross flow.
h
Narrow channel (H=8mm)
5kW. Cross-flow
suppressed.
4. MATHEMATICAL DESCRIPTION
Theoretical analysis which does not take into account crossflow through perforation predicts two solutions of
temperature and flow-fields:
Symmetric solution (flow-rates and temperatures in
lateral channels are equal)
Asymmetric solution exists within a certain range of
flow-rates and heating power.
Asymmetric solution can be interpreted as a magnitude of
disturbance causing instability of flow. Mathematical model
identifies parameters having significant influence upon the
stability limits, e.g. width of lateral channels.
More details can be found in
http://www.fsid.cvut.cz/en/u218/peoples/zitny/imagohm/instabil/instabil.doc
Wide channel
(H=18mm)
without heating
Wide channel (H=18mm)
5kW-cross flow suppressed
CONCLUSIONS:
•Asymmetry and instability of parallel flows can be explained by natural convection.
•Stability of flow can be improved by increasing friction losses, e.g. by narrowing lateral channels.
•Model and experiments predict that the cross-flow is suppressed (or even reversed) at non-isothermal flow.