Summary: | This thesis is concerned with factors affecting the thermal design of a
compact recuperative type crossflowing heat exchanger for the primary
heater of a Stirling engine. The exchanger is constructed of small
diameter metal tubes (in the range of 3.0 mm to 6.0 mm) and close spacings
are maintained between the tubes (i. e. in the range of 0.30 mm to 1.80 mm).
These small slender tubes are usually arranged in single or double rows
and are mounted around the combustion chamber. The exchanger is used to
heat hydrogen or helium which act as a working fluid for the Stirling
cycle.
A survey of the published literature indicated that the available data
does not include results for the tube Geometries of interest in this
study. Consequently the heat transfers and hydraulic resistances were
measured experimentally for a single row of small diameter, closelyspaced
tubes situated in a crossflowing fluid stream. The Reynolds
numbers (bc,.s ed on the mainstream fluid velocity and the tube diameter)
ranged between 300 and 6500. Where possible the accuracy of the experimental
procedure was checked by comparing the present results where
possible with those obtained by previous workers. Several arrangements
of both bare tubes and tubes fitted with extended surfaces were studied,
The results were analysed and discussed, and where appropriate compared
with those published in the open technical literature. In most comparative
cases excellent agreement was experienced and any departure, could
be explained.
For the bare tube arrangements the influences of flow blockage ratio,
mainstream turbulence intensity and surface roughness on the average
heat transfer performance were investigated. A comparison of the heat
transfers and pressure drops characteristics of the different tube
arrangements led to proposals for an optimal exchanger geometry. The
validity of the empirical corrections suggested by previous workers to
account for the influence of flow blockage on average heat transfers
was examinod. An alternative modified empirical expression was then
proposed for the particularly high flow blockage situations (D/1I>0.85).
It was found that at those high flow blockages the overall average heat
transfers were independent of the mainstream turbulence intensity. However
preliminary tests suggested that an increase in the tube surface
roughness increases the average tube heat transfers.
Since the proposed heat exchanger operates at the higheet possible mean
metal temperatures, it is likely that 'hot spots' occurring due to variations
in local heat transfers can lead to premature failure. Consequently
a detailed study of the local heat transfer distributions is presented for
various Geometrical conditions. The influences of blockage ratio and
mainstream Reynolds numbers are examined and the results are analysed, and
discussed, and where possible compared with other published data. The
accuracy of experimental procedure employed in these tests was checked by
comparing the results for a single cylinder ease with those reported by
other investigators.
The influence of fitting a single longitudinal fin to the rear of tubes
on both the heat transfer and pumping power was studied. The tube diameter-
in these tests was kept constant, at 6.0 mm, but the angle of
inclination of this longitudinal fin was varied incrementally so that an
optimal angle for maximum performance is recommended. In a similar
manner, transverse finned tubes with two different fin spacings were
also investigated. The heat transfers and pressure losses obtained for
the different finned tube arrangements were compared with each other and
with those obtained for the bare tube geometries so that an optimal tube
configuration was proposed .
The data presented in this thesis were generalized, where possible, so
that the results should be useful for future work. They should thus
contribute to an understanding of the basic phenomena associated with
modern compact heat exchangers. Recommendations for further work are
also presented.
|