of the liquid film and the resulting flows within the film
caused by the Marangoni effect are shown schematically
in Fig. 1.
Recent research has identified that the Marangoni
effect will generate non-smooth condensate films thatrange from ‘pseudo droplet’ [6,8] to ‘ringwise’ [9] mor-
phologies. This non-smooth behaviour has been shown
to significantly reduce the condensate film heat transfer
resistance; however this effect has been outweighed by
the increased resistance of the vapour diffusion layer. It
was in this context that Morrison and Deans [10] pub-
lished research showing that at very low concentrations
of ammonia, steam condensation heat transfer on a
short, horizontal tube was actually enhanced by up to
13%. This enhancement was attributed to the Marang-
oni effect which produced a disturbed condensate film at
ammonia concentrations where the diffusion layer
resistance remained low. This paper extends the work of
Morrison and Deans [10] by investigating the conden-
sation of weak ammonia–water mixtures in a simple,
horizontal shell and tube condenser.
2. Experimental method
The test condenser consisted of a 700 mm long, 150
mm diameter, stainless steel shell and a horizontal, 20
mm diameter, stainless steel tube arranged in a counter-
flow, single tube and shell pass configuration. The shell
(vapour) side was pided into six sections by a series of
baffles, which produced an approximately cross-flow
vapour flow pattern. Vapour composition, temperature
and pressure could be measured at several points along
the condenser shell and sight glasses allowed the con-
densation process to be observed visually. The tube was
instrumented with three thermocouples embedded in the
tube wall along the length of the tube and could be ro-
tated through 360, facilitating radial tube surface tem-
perature measurements. Thermistors positioned along
the length of a twisted tape insert inside the tube mea-
sured the temperature of the cooling water, which
combined with the cooling water flow rate provided an
estimate of the energy transferred to each section of the
condenser tube. A schematic diagram of the test con-
denser showing its major components and general
dimensions together with the flow path of the vapour
and condensate is shown in Fig. 2. It should also be
noted that except for small amounts of vapour that were
purged prior to each test to remove any non-condens-
able gas, all of the vapour entering the test condenser
was condensed, i.e. the condenser operated as a total
condenser.
The tube wall temperature in condenser sections 2, 4
and 6 (see Fig. 2) was measured at 30 degree intervals
around the circumference of the tube. The average
outside tube surface temperature (Tw;o) for each of
these sections was then calculated using the corre-
sponding average tube wall temperature (Tw), the
cooling water energy balance (Qcw) and an estimate of
the thermal resistance (Rtc;w,) between the tube wall
thermocouple and the tube surface (Eq. (3a)). Finally,
the average condensation heat transfer coefficients
( hw;o) for sections 2, 4 and 6 of the condenser tube were
calculated using Eq. (4). This method was verified by
the excellent agreement of the experimental results for
pure steam condensation with Nusselt’s theory [11].
Furthermore, a worst case error analysis, based on
the errors presented in Table 1, suggested that the
maximum experimental error associated with the con-
densation heat transfer coefficient results was less than
The experimental condensation heat transfer coefficients
were compared to both Nusselt’s theory of condensation 卧式壳管式冷凝器英文文献和中文翻译(2):http://www.youerw.com/fanyi/lunwen_7995.html