Bristol Centaurus
Bristol Centaurus
Bristol Centaurus
FLIGHT
Survey of Britain's Most Powerful Radial Engine : An Example of
. r l^
O
gi
C
al Layout to Achieve Compactness with Power
T has been eommbn knowledge far some time past that
the Bristol Aeroplane Co., Ltd., have been engaged with
the production of a larger and improved model in the
age of radial, air-cooled, sleeve-valve engines with which
ey have for so long enhanced their reputation. This
mmon knowledgethe result of unofficial "leaks"
ibraced the facts that the new engine was an 18-cylinder
lit of over 2,000 h.p. and was called the Centaurug..
ther than this nothing much was generally known until
nctioned reference, to the engine was made with the
lease of the Short Shetland flying boat (Flight, May 17th,
145), when it was revealed that the Centaurus was of
'er 2,500 h.p.
Even now we are not permitted to give any indication
the power output other than
at it is well in excess of
500 h.p., so that, in assessing
it'cific qualities' relating
>wer, one can
ily use the
;ure of 2,500 as
'datum and
ite that the
uated result is
mewhat lower
an the factual
due.
Probably the
ost impressive
ature to strike
e casual ob-
rver of the
mtaurus is the
mpactness of
e unit in rela-
)n to its power
itput, and, as a
irollary, the achieve-
ent of a clean and efn-
ent fairing by refine-.
ent of design. It is,
course, obvious that
miponent in the engine is the
suit of months, in some cases
;ars, of development, and that
c only means by which such WOLIC-
anship can bear fruit is by way of
ie hard road of technical ability fol-
wed strongly by the incessant trial"
id error of testing. By such means a tremendous fund
E data is amassed, such data being the "history" only
uough which the future may be discerned. Looking
ack in order to look forward applies equally in the tech-
ical as in the ordinary world. .
Specific Powers
From the aspect of arrangement analysis the Centaurus
basically similar to its immediate forerunner, the Her-
jles, with the chief differences that it has four more
flinders and that entirely separate individual gear trains
re embodied front arid rear for driving the sleeves in the
-spective cylinder* banks. Swept volume is naturally
irger, but so is the unit swept volume as, although the
ylinder bores remain the same at 5.75m., the stroke of the
entaurus is 7.0m. as compared with the 6.5m. of the
lercules. This gives the new engine a capacity of 53.6
tres (3,270 cu. in.) in contrast to the Hercules' 38.7 litres
2.360 cu. in.), an increment of 38.5 per cent. In terms of
apacity and power output, at the restricted Centaurus
every
Compactness without complication.
Note the backswept exhaust stack
cooling muffs and the collected exhaust
tail pipes.
power figure of 3,500, the b.h.p. /litre of both Hercules
Centaurus is 46.5, although we know that the latter engine
is somewhat better than this, as may be rough!y indicated
by the figures for b. h. p. /sq. in. of piston area, these
being respectively:. Hercules 4.93. and Centaurus better
than 5.34.
f
: Cooling Arrangements
As an indication of refinement in the design of the cowl-
ing, if. we take as a datum the frontal area of the Hercules
at 2,122 sq. in. and give it the value of unity, then the
Centaurus frontal area of 2,402 sq. in. gives a comparative
ratio of 1.13:1, which is well below the relative h.p..ratio
of 1.385 :1, itself a conservative figure. Put another way,
this means that for an increase in
power output of 38.5 per cent, over
the Hercules, the Centaurus has only
a 13 per cent, greater frontal area,
its specific' power thus being
150.5 h.p./sq. ft.
It must be appreciated that
the goal of really close cowling
...- . . ,.., . -.. is hindered in at-
' . "' ' ' / tainment by the-
very critical fac-
tors of cooling,
and Bri st ol ' s
achievement is
rendered the more
deserving of praise
by virtue of their
having obtained,
with it, an en-
hanced cooling
efficiency. To a
- certain extent the
employment of a
cooling fan driven from the air-
screw spinner can be deemed
partially responsible for the in-
crease in cooling efficiency
but by no means wholly.
As the Centaurus is designed- as a
power plant, the cowling, fairing and
baffling arrangements naturally come
into the picture in addition to the
engine per se, so whilst on the sub
ject of cowling we might as well deal
with it before going on tc the engine'
proper. The subject is well worthy of study.
The main body of the engine in way of the cylinder
-'banks is closely cowled by four panels, two per side, hinged
together at top and bottom on the fore and aft centre
plane of the engine and secured together with three toggle-
type fasteners on each side. The latter are recessed to
give a smooth surface line. Forward of the cowling panels
is the-aerofoil-section annular nose fairing which, with the
very unusual spinner extension fairing (fitted around the
reduction gear casing), provides a smoothly contoured, but
surprisingly small, annular intake duct in the entry of which
rotates the- 18-blade cooling fan fitted to the rear of the
spinner.
It may thus be seen that cooling air entrained is ob-
structed not at all until it reaches the front cylinders.
Having done its due amount of fin circulation the air is
then allowed to escape to the outside atmosphere through
cooling gills in the skirt ring of the main cowling panels.
In consideration of the amount of development design
which has gone into producing the Centaurus, the instance
1*4
FLIGHT
BRISTOL CENTAURUS
; cylinder and head finning is interesting. By experiment
has been shown that the optimum pitch for fin location
in the region of o.iin.-^the pitch on the Centfturus is
[izin. and the cooling area of finning to each cylinder
id head is no less than 32.47 sq. ft.
Not the least novel feature of the cooling arrangements
the fact that, unexpectedly enough, the front,row of
flinders, due to the amazing control of the cooling air-
low, actually runs a trifle hotter than the rear row.
pressure drop in the cooling air through, or, rather, across
he cylinders is the rather high figure of 10 inches of water,
3d, very shrewdly, this is put to incidental, yet none the
useful, advantage in cooling the rear-swept front-"
ylinder exhaust pipes. These are shrouded' with tubular
jufis, open at each end, which extend nearly to the trans-.
jterse centre plane of the engine, and the pressure differ-
ential between front and rear open ends induces an air-
Bow through the muffs which effectively cools the exhaust
Wipes whilst, at the same time, isolating their cooling ajf
prom the ordinary flow to the rear cylinders.
Limits of Cooling
The cooling efficiency of the Centaurus is nearly as high
it is possible to get. There are for the future the addi-
tional refinements of a geared cooling fan running at higher
[speed so inducting a greater mass flow,'and also the closer
pitching of the cooling fins. Nevertheless, it would appear
that the ultimate is at- least within sight, and from this
aspect it is a reasonable assumption that some limit will
sooner or later enforce a barrier against the use of more
potent fuels and increase of engine rating by stepping up
the b.m.e.p. This reflection excludes consideration of the
(use of water/methanol injection and kindred devices which
conceivably might have an increasingly large part to play.
The reduction gear is the well-tried Bristol-Farman type
I bevel epicyclic which is so well known as to render a
[further description superfluous. One element is, however,
[worthy of note; the rear driving bevel and the front
[stationary (sun) bevel ring are both located on annular
Iseatings, the bearing faces of which are spherically curved.
I This measure ensures that thedriving and sun bevels can-
[each rock slightly and so permit the load to be distributed
I equally among all three planet pinnions. The three
["spokes" on which the planets are carried are integral
[with the airscrew shaft, the tailpiece of which is supported"
I in a bearing-in the forward end of the crankshaft. Airscrew'