You are on page 1of 8

I85

The Gas Turbine in Perspective


By Hayne Constant, M.A., M.I.Mech.E., F.R.S.*
A survey is made of the progress that has been achieved in Britain over the last decade in improving
the performance of gas-turbine components. The effect of these improvements on the performance
of jet-propulsion engines is then shown.
An attempt is also made to foresee the probable future trend of progress and to assess the nature of
the important problems which have still to be overcome before the gas turbine can be more generally
applied.
INTRODUCTION In consequence, it is becoming the practice to judge reliability
The rise of a new form of heat engine from the dream of engi- and capacity for endurance in terms of the basic factors involved
neers to the practical machine doing useful work in the service rather than by a direct test. Such an outlook is, of course, essential
of man is a rare event, and the possibility of studying the processes in the worlds of the power station and the marine engine, for
involved in such a development occurs very infrequently. none of us has sufficient patience or money to wait ten or twenty
The emergence of the gas turbine as a new prime-mover is an years to find out if a new piece of machinery is satisfactory.
example of such an event, and it is an example which has pro- The two most serious potential causes of unreliability in the
gressed at such a pace that it is possible to hold in perspective gas turbine are the blades and the combustion-chamber baffles.
the interplay of forces that have controlled the direction and The problem of blade vibration is one which has been inherited
nature of the progress made. from the steam turbine and, although there is much to be done
The gas turbine remained no more than an idea for years, before it can really be said that we fully understand the processes
awaiting the development of the techniques necessary to convert taking place, there is no reason to suppose that semi-empirical
the thermodynamic conception into the practical reality. The methods will not be as successful in the gas turbine as they have
most vital technique that was wanting was the capacity to design been in the steam turbine in providing a very high degree of
a compressor of sufficiently high efficiency. Another difficulty reliability. I hope that we shall not remain for ever dependent
that forbad a start being made was the necessity for using a gas on these semi-empirical methods, and that the elaborate studies
temperature at admission to the turbine several hundred degrees of the forcing impulses and their effects on blade rows now
greater than existing knowledge allowed. It was at first thought going on will ultimately yield a very precise knowledge of the
that this made necessary the development of new materials stress conditions in a vibrating blade. The combustion-chamber
capable of resisting creep and fatigue at high temperatures, and of the gas turbine, however, presents quite novel problems. The
it was not until such materials became availablethat serious work basic difficulty here is that the sheets of metal which form the
began. The steady improvement of such materials has continued inner baffles of the chamber may be subjected to high tempera-
ever since. It is now beginning to be appreciated, however, that ture, to serious temperature gradients, to unequal rates of heating
the non-existence of materials capable of operating satisfactorily and to pressure fluctuations. As a result of these conditions,
at high temperatures need never have prevented the development oxidation of the metal surface in the neighbourhood of hot spots,
of the gas turbine. For recent ideas on the cooling of metal buckling of the sheet due to temperature gradients, and fatigue
surfaces by fluids have shown that it is possible to operate a gas and ultimate failure of parts due to a combination of effects, may
turbine at sufficiently high gas-temperatures to obtain a reason- occur.
able performance, without the use of special heat-resisting There are a number of ways of dealing with this situation.
materials. Before the development could start, therefore, it was The first is that of making the parts sufficiently strong to
necessary to have either high-temperature materials or the stand up to the conditions. This may often be an effective, if
techniques of cooling. W e need to be constantly on our guard expensive, solution, but there are conditions under which an
against delaying the initiation of a new venture simply because increase in the thickness of a member may make it more liable
we have not to hand the technique of solving an outstanding to fail as a result of the stresses set up by unequal rates of heating,
problem by the most obvious means. A second, and better, basis of design depends on the avoidance
of redundant structures in the baffle system. I n such an arrange-
ment the baf3es are designed so that they are free to expand
P R O G R E S S IN R E L I A B I L I T Y under the imposed temperature gradients. Such a simply stiff
structure, if made of thin sheet metal, will not be heavily stressed
Once the development of the gas turbine had started, the most by internal thermal stresses.
urgent need was the achievement of sufficient reliability to allow By far the best way of attaining high reliability and long life
the operation of the machine to be properly studied. I n the early in a combustion-chamber lies in the avoidance of pressure
days a run of a few hours was an event, and it is not so many fluctuationsby designing for a low velocity and stable combustion
years ago that an endurance run of 100 hours was regarded as system, and the effective cooling o f all metal parts down to a
of great significance. Nowadays, aero-engines are expected to temperature at which the internal stresses are negligible.
run as much as 500 hours between overhauls. This rapid im- Techniques have now been developed which enable this to be
provement was expected, for one of the important advantages done adequately when using distillate fuels, but further develop-
always claimed for the gas turbine is its relative freedom from ment may be required before satisfactory cooling can be obtained
wear and deterioration and its inherently great reliability. with the highly luminous flames produced by residual oils.
We are now, however, coming to a position in which the Fig. 1 shows the layout of a combustion-chamber in which no
proving of reliability is becoming increasingly difficult. The part of the metal surfaces exceeds 530 deg. C. (990 deg. F.) when
marine and industrial worlds are not interested in tests of a few operating on gas oil, with an inlet-temperature of 400 deg. C.
hundred hours. Moreover, the power output of modern engines (750 dea. F.) and an outlet-temperature of 750 aen. C .
is increasing rapidly. An endurance test, of sufficient length to (1,380 dig. F.). -
be significant, is therefore becoming a very expensive matter. When the cooled combustion-chamber has been fully
The M S . of this Iecture was received at the Institution on 3rd developed, and our knowledge of vibration vut on a rational and
August 1950. quantiiative basis we shall well on the Gay to predicting, as a
* Director, National Gas Turbine Establishment. result ofofLeeds
Downloaded from pme.sagepub.com at University measurements
on June 5, 2016 taken over a few hours of running,roughly
11
186

AIR GAP INSULATION THRO~TLEFOR COOLING AIR


FOR OUTLET SECTION
Fig. 1. Design of Combustion-chamber for Long Life

for how long the machine will give satisfactory service. We are, in Fig. 2. Higher polytropic efficiencies have been obtained at
in fact, already in a position to design engines to run reliably for speeds other than the design speed. It is extremely difficult to
ten or twenty years on distillate oils, but owing to our lack of make accurate measurements of the efficiency of machines of
precise information on factors of safety, their performance would this kind. Unfortunately years of experience have shown that the
have to be unnecessarily conservative. more instruments are improved, the lower is the efficiency which
is recorded. However, the greatest ingenuity that we have been
able to apply towards the refinement of measuring techniques
PROGRESS IN T H E COMPRESSOR has failed to reduce the efficiency below the values shown in
While the struggle to improve the life and reliability of the Fig. 2. The efficiency has risen some 10 per cent in thirteen years
machine has been going on, a continuous effort has been
directed towards a different, and to some extent a conflicting,
+
and is now increasing at the rate of about per cent per year.
objective; that of improved performance.
In some respects an improvement in performance may give
improved life and reliability. In the compressor, for example, a
reduction in blade losses can reduce the pressure pulsations
which tend to force blade vibration and will reduce the tem-
perature rise in the air and make possible a generally lower
working temperature for the same cycle pressure-ratio. But
usually designers will take advantage of any improvements in
component efficiency that are offered them and rate the machine
that much higher; so that no gain in reliability results.
The history of the marine and industrial gas turbine is as yet
unmade and it will be many years before we can study the records
and note the progress achieved. But in the aircraft world the Fig. 2. Polytropic Efficiency of Axial Compressors
gas tprbine has now some ten years to its credit, and sufficient
engines and components have been tested to enable curves of The values given refer to the point of maximum efficiency at the design
progress to be drawn. speed.
Care must be exercised in arguing from the achievements of
one branch of engineering to another. It must be borne in mind, This improvement has been due, I think, more to better match-
for example, that the aircraft gas-turbine has components which ing of the different stages, than to raising the efficiency of a
are aerodynamically cramped and restricted in the interests of single stage. The foundation on which this progress has been
bulk and weight reduction, and that they represent compromises based has been the wind-tunnel tests of blades in cascade. An
to enable them to operate over a wide range of conditions. I t is account of the techniques used in this work is given in another
therefore usually safe to say that good efficiency should be easier lectllre in this series. During the next few years we may expect
to attain in an industrial or marine component than in one to obtain results from some of the fist industrial and marine
designed for the air. On the other hand, it is often true that it compressors at present under construction and test. I hope and
would be uneconomic to put so much detail refinement into believe that the less restricted geometry and lower Mach
the industrial or marine machine as into that for the air. In spite numbers of these compressors will enable an excellent per-
of these limitations a study of the progress achieved in the formance to be obtained and assist in maintaining ihe upward
aeronautical gas-turbine provides plenty of food for thought trend of the curve.
applicable to gas turbines in general. The end is in sight. There are certain unavoidable losses, such
Perhaps the most important advance that has been made so as those due to skin friction, which will always be with us. Taking
far is that in the efficiency of the compressor. A curve showing account of these losses we are led to the probability that the
the manner in which the polytropic efficiency of the axial com- highest adiabatic polytropic efficiency which can be anained
pressor at design speed has increased over Downloaded
the years is shown at without
from pme.sagepub.com impracticable
University of Leeds on June 5, 2016 elaboration is about 95 per cent, and this
THE G A S TURBINE I N PERSPECTIVE 187
I have marked on Fig. 2. At the present rate of progress this tendency in general seems to be for the temperature rise per
goal should be achieved in about ten years. But, in fact, the rate stage to slowly increase. A notable exception to the general
of progress is likely to decrease and the limit be approached trend is the compressor marked 109 in Fig. 4. This is an experi-
asymptotically. mental compressor which was designed with the specific object
It will be of interest to see how close to the ideal limit it is of obtaining the highest possible temperature rise per stage in
proved to be worth going. At the moment it appears that in a order to explore the possibilities of more compact compressors.
long-life engine a 1 per cent increase in compressor efficiency is I t is therefore not surprising that it does not conform to the
worth while if it can be obtained without increasing the cost of general trend of development.
the compressor by more than about 10 per cent. The final
answer will probably be that it will pay us to carry on our
researches right up to the limiting efficiency of 95 per cent, but
to apply them only if they do not involve blade forms of con-
siderably increased cost.
This improvement in efficiency is all the more remarkable
when one remembers that it has been accompanied by an increase
in pressure ratio. Owing to the deterioration in the boundary
layers on the casing and rotor and to the decrease in blade aspect-
ratio as one passes from the low-pressure to the higher-pressure
end of the compressor, it is usual for the blade efficiency to be
highest at entry and to diminish through each successive blade
row. The addition of further stages to attain higher pressure- DESIGN DATE
ratios therefore results in a lowering of the polytropic efficiency Fig. 4. Stage Temperature-rise of Axial Compressor
of the compressor. It is certain that for a given pressure ratio
the rate of progress has been greater than that shown in Fig. 2. The values given refer to the point of maximum efficiency at the design
The actual increase in pressure ratio that has taken place is speed.
shown in Fig. 3. This increase has latterly introduced difficulties
The reassuringly steady slope of the curve of compressor
progress combined with the comparatively small scatter of the
individual points is owing mainly, I think, to the fact that most
axial compressors have been designed to a single formula. Indeed
the laws of fluid dynamics, so far as can be seen at the moment,
will not allow any great departure from this formula so long as
we confine ourselves to subsonic flow. For there are fairly well-
defined limits to the adverse pressure-gradient under which a
boundary layer can remain stable.

P R O G R E S S IN T H E T U R B I N E
In the turbine a great deal more latitude in design can be, and
has in fact, been taken :everyone knows how to design a turbine,
but everyone's ideas are not always the same.
It is not surprising therefore that we cannot record the same
steady progress in turbine design as in compressor design. The
results shown in Fig. 5 are so scattered that I hesitated to
draw a line through them. The line I have drawn represents
L

Fig. 3. Pressure Ratio of Axial Compressors


The values given refer to the point of maximum efficiency at the design
speed.

of matching the compressor to the turbine, owing to the inherent


tendency of the high-pressure axial compressor to exhibit a surge
line with a kink or discontinuity in it. All that I need say further
on the matter of pressure ratio is that it is unlikely and certainly
undesirable that the pressure ratio in a single compressor should DATE OF FIRST RUN
be raised any further except for very special purposes. I have Fig. 5 . Polytropic Efficiency of Turbines
indicated my wish-thought in this matter by the dotted extra-
polation in Fig. 3. Further increase in engine compression-ratio The values given refer to conditions at the design point.
should be attained by double compounding, that is, by the use
of two mechanically independent compressors in series. Such a the mean, not of all the results shown, but of those due to a
system of compounding will give a more flexible engine with single manufacturer, in whose products a consistent trend is
greater freedom from the danger of surging, will reduce the observable.
danger of failure of low-pressure compressor blades due to I n Fig. 6 is plotted the corresponding curve of temperature
operating stalled at part loads, and in land and marine applica- drop per stage, on a non-dimensional basis. The curve is again
tions will make easier the use of intercoolers. drawn through the results obtained with turbines made by one
Another feature of compressor performance, which is of manufacturer, but it represents fairly well the general trend.
interest in that it determines the number of stages required to Very few turbine test results were available before about 1945.
attain a given pressure-ratio, is the temperature rise per stage. It is to this that we must attribute the comparatively static state
This is of importance in the aeronautical application, since it of the turbine until quite recently. A turning point, which is not
affects the size and weight of the compressor, and in other shown very clearly in Fig. 6, came in about 1948 with the
applications in its influence on the number of blades and hence realization that the turbine blades were overloaded; the air
the compressor cost. The change in the temperature rise per deflexions and lift coefficients were excessive. Since then, there
stage of axial compressors over the years is shown infrom
Downloaded Fig. 4. The at has
pme.sagepub.com been
University a ontendency
of Leeds June 5, 2016 to reduce this loading by an increase in
188 INTERNAL COMBUSTION T U R B I N E S
peripheral speed and in the number of stages, but a resulting Important reductions in the aerodynamic losses in the aero-
gain in efficiency is not yet clearly apparent. The extent to which engine combustion-chamber have been made, but these have not
this process can be carried in the aircraft gas-turbine is probably been reflected in any consistent improvement in pressure loss,
quite limited, owing to the weight penalty that has to be faced. since designers have taken advantage of these reductions to
For applications in which bulk is not a limiting factor the way enable them to use more compact chambers. As a result of this,
of advance appears to be by increasing the reaction and reducing heat release rates have risen.
I n the industrial and marine combustion-chamber the
direction of advance has, of course, been quite Merent: heat
release is of secondary importance and high efficiency and low
pressure-loss under most conditions are no real problems. The
points on which attention is now rightly focused are those of
obtaining good reliability and long life and of designing a
chamber capable of burning ash-forming fuels.

1938 1939 I940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951
P R O G R E S S I N MATERIALS
D4TE OF FIRST RUN
Another technology which has made a great contribution to
Fig. 6. Stage Temperature Drop Parameter for Turbines the progress of the gas turbine is that of heat-resisting materials.
The use of materials capable of resisting oxidation has been
The values given refer to conditions at the design point. widespread for many years, but these materials did not in general
g Acceleration due to gravity. have good properties at high temperatures. The need for a
J Mechanical equivalent of heat. material capable of resisting creep at moderate temperatures was
c Specific heat at constant pressure. made clear by the demands of the steam turbine in the years
A f i Temperature drop through a single stage. between the wars. The achievement of materials with good high-
u, Mean blade speed. temperature properties was apparently an incidental by-product
of other developments. It was certainly not the result of the
the lift coefficients. It is indeed paradoxical that turbine requirements of the gas turbine, for it took place before work
efficiencies should be consistently less than compressor on the gas turbine had gathered much momentum. This is
efficiencies; but it is likely that the increased effort now being shown in Fig. 7, in which the temperature at which a stress of
put into the study of turbines will soon improve the present 6 tons per sq. in. could be supported for a creep strain of
unsatisfactory position. 1 per cent in 10,000 hours is plotted against the date at which
the material became available. Although this particular criterion
is not by any means the best for showing the capability of a
PROGRESS I N COMBUSTION material, it shows the general trend sufficientlyfaithfully, and it
Progress in the combustion system is difficult to define, partly has the merit that many results in the past have been expressed
because so many changes have been occurring simultaneously in this form so that a curve can be plotted over an extended
and partly because of the difficulties of measurement. For many period.
years now the efficiency attainable at ground level has been only The continued upward trend of the curve in recent years is
1 or 2 per cent below 100 per cent and the new goal in the air- due in part to the introduction of cast alloys and of materials
craft world has been the attainment of higher efficiencies at high that are more difficult to work. It is probable that further progress
altitudes. will be controlled more by our ability to devise improved and

DATE

Fig. 7. Improvement of Creep-resisting Alloys


The points give the approximate temperature for 1 per cent creep strain in 10,000 hours at 6 tons per sq. in.
o Forged material.
Downloaded from pme.sagepub.com University of Leeds on June 5,T.Y.
o Castat material, 2016 Time yield.
THE GAS TURBINE I N PERSPECTIVE 189
novel methods of manipulating and working the material than matching of the components. The difference between the two
by its capacity to resist creep. I think myselfthat further progress lines for centrifugal engines is a measure of the gain due to the
will be small, and that we are approaching the maximum that changes in the cycle. As the maximum cycle temperature has
can be expected from the metallurgist. This is, perhaps, fortunate changed little over the years this gain reflects very nearly the
for the mechanical engineer because, however good the per- result of the gradual raising of the pressure ratio.
formance of a material in creep, fatigue, and other strength In the years to come we shall see a continuation of the tendency
characteristics, there are many awkward problems involved in for specific fuel consumption to fall owing to these same causes.
operating machinery at a bright red heat. Of these the problem It seems to me unlikely, however, that it will be found worth
of thermal expansion and distortion is probably the most while carrying the development beyond a point at which the
serious. I think it likely that a substantial reduction in coefficient consumption is about 0.73 lb. per hr. per lb. of thrust. This
of expansion would be a more desirable gift from the metal- looks to me like a practical limit to our quest for efficiency and
lurgists than a f i s h e r increase in the maximum temperature at one which, if present trends continue, will be attained within
which a machine can be operated j but the probability of obtain- the next twelve years.
ing a satisfactory material of low coefficient of expansion is small. Of course the aeronautical engineer is not really concerned
with the static fuel consumption but with that under conditions
of altitude and forward speed. Unfortunately this is harder to
PROGRESS I N JET ENGINE P E R F O R M A N C E measure. The big effort now is to improve the consumption in
The progress made in components and materials has been flight, and in particular, at high altitudes. When this has been
reflected in the performance of the complete engine. The done, any further reductions in the specific consumption will
improvement over the years of the specific fuel consumption of have to be attained by a change of cycle, such as the introduction
British jet-propulsion aero-engines is shown in Fig. 8. The values of the ducted fan. This is a development which is long overdue.
given are the static sea-level consumptions. The number of test Another yardstick of progress, and one of particular concern
points available is very large and the scatter is great j the lines I to the aeronautical engineer, is the engine specific weight. The
have drawn define the lower edges of the area of scatter rather improvement that has taken place in this respect is shown in
than the mean. I have done this deliberately since, for the purpose Fig. 9. Again I have drawn envelope curves for the best engines,
of my immediate argument, I am concerned with the best con- ignoring the more robust and conservative designs which lie
sumption that has been obtained, and I am prepared to ignore above the curves. The slow start made by the axial engines was
those engines in which consumption has been sacrificed for partly due to lack of development effort and partly to the fact
some other advantage. The lack of progress made in the that they represent a more difficult problem, but the axial specific
centrifugal-compressor engine since 1944 is noteworthy and weight has at last reached equality with that of the centrifugal.
reflects the general stabilization of the aerodynamic design that It is a curious fact that in most instances the specific weight of a
took place about that date. The necessity for a fresh line of given engine does not decrease much with time as a result of
attack is apparent, and this line was supplied by the axial-
1'0 --1----i --
compressor engine, The unusual shape of the curve for the axial 1 --1
engines is difficult to account for completely, but, in part, it is
certainly due to the fact that in the early days the bulk of our
1.5

1'4
@L
L
51 3
a
m
--I 1'2
a
.I
U

? I I
I
91 0
c
a
r
20 9
Z
0
U
0-8
2
g IT7
-
Y
U
w
% 0'6

0 5 DATE OF TEST
I940 1911 I942 1943 1944 '$45 1946 1947 1948 1947 1950 1951 1951
DATE OF TEST Fig. 9. Specific Weight of British Jet Engines
Fig. 8. Specific Fuel Consumption of British Jet Engines The values given refer to sea-level static conditions at take-off engine
The values given are the values under sea-level static conditions. speed.

resources were devoted to the centrifugal engine and it was not development, and progress is made by making a fresh start
until about 1947 that there was a large transfer of effort to the with a completely new design. The same feature is apparent in
axial type. the other criterion of aero-engine performance, thrust per unit
The drop in consumption shown in both curves is due only frontal area, and is illustrated in Fig. 10. The points defining
in part to the better matching of components and the improve- some of the curves have been shown to give some measure of
ment of their efficiencies, and the reduction in combustion- the scatter involved. The development lines of a number of
chamber losses. The consumption has fallen also as a result of specific engines, and the big steps in progress made with each
the gradual rise in the cycle pressure-ratio and in spite of the new design are also shown. This i s a characteristic of the early
tendency for the maximum cycle temperature to rise slightly. days of a new development. The great increase in thrust per unit
I have attempted to illustrate t h i s in Fig. 8 by an additional frontal area resulting from the change-over to the axial engine is
line drawn for the centrifugal engines. This line shows how the apparent. Within the next few years we may expect that the use
specific fuel consumption would have improved over the years of higher pressure-ratios and of anuular combustion-chambers
if the cycle temperature and pressure had remained unaltered. and the more compact layout of auxiliaries will raise the thrust
It represents the gain due to improvements in the efficiency and per unit
Downloaded from pme.sagepub.com at University frontal
of Leeds 2016 to over 1,000 lb. per sq. ft.
on June 5,area
190 INTERNAL C O M B U S T I O N TURBINES
In industrial and marine gas-turbines today and to a smaller D I F F I C U L T I E S D E L A Y I N G T H E A P P L I C A T I O N OF THE
extent in aeronautical gas-turbines, the performance actually G A S TURBINE
achieved is considerably below that which should be attainable The nature and importance of these' practical difficulties
with present knowledge. There are several reasons for this depend very much on the purpose for which the gas turbine is
discrepancy. intended.
In the first place, the matching of the various stages of the I n the aeronautical world there are no special engine problems
compressor is still far from perfect and the maximum efficiency holding back the application of the gas turbine. There are of
is often not developed under the design conditions. Again, it course the usual development difficulties that go hand in glove
will have been noted that there is a scatter of some 5 per cent with the search for improved performance. But we can rely on
in efficiency between the best and the worst compressors and a the genius of the aeronautical engineer to overcome these
rather greater scatter in the turbines. It nearly always seems to obstacles as they arise, even when, as sometimes happens, he
happen that if an engine is fortunate enough to get a good com- himself is responsible for their erection. The bogy of engine
pressor, its turbine design leaves much to be desired. We seldom failure due to ice formation is gradually being laid to rest, and
get an engine in which all the components are up to the best the fear that compressor blades would become quickly fouled
standards, and if ihey are, the matching of one to the other is by flies and dust is proving unfounded. We need better per-
poor, so that the overall performance is disappointing. It is formance at altitude, but that will come in time and cannot be
difficult to generalize on this, but I should say that many of our said to be a limiting factor. We need better airport control, but
that is outside the province of the gas-turbine engineer.

The Use of Residual Oil. In the marine world the outstanding


problem is that introduced by the need for all except small and
specialized ships to use residual fuel oil.
The work that has been done during recent years on the actual
combustion of residual oils under the conditions obtaining in
gas-turbine combustion-chambers has been successful, and it is
now possible to burn the heaviest oils with good efficiency. Our
knowledge, however, is by no means complete and we need
considerably more experience before we can with confidence
make departures from designs already proved.
The real difficulty with residual oils is not in the process of
combustion itself, but in the corrosive attack of inorganic salts
such as those of vanadium on the surfaces of the combustion-
chamber and turbine. Steady progress is being made in obtaining
an understanding of the processes involved, but an outright
solution of the problem is not yet in sight. The best that can
be looked for at the moment is a mitigation of the evil by the
removal of as much ash from the fuel as possible by cleaning;
1 1 . 1 1 1 I I 1 1 1 1 ' 1
by the addition to the oil of substances converting the vanadium
@

1940 1941 ' 1942 1943 1944 I 1945 I 1946 1947 ' 1948' 1949' 1950 ' 1951 ' 1952' salts to less active types and acting as diluents to the vanadium
DATE OF TEST in the ash; by coating of the blades with substances less
Fig. 10. Thrust per Unit Frontal Area of British Jet Engines susceptible to vanadium attack; and by reduction of the tem-
perature (by cooling) of the surfaces on which vanadium may
The values given refer to sea-level static conditions at take-off engine be deposited. An account of the stage to which our knowledge
speed. of this difficult problem has been taken is given in one of the
lectures in this series.
present-day engines could be improved by at least 10 per cent
in respect of their fuel consumption, if each component were The Heat Exchanger. Another difficulty that may be referred
replaced by one representing the best practice achieved so far, to in its application to the marine engine is that of size. The
and if these components accurately matched each other so that bulk of the present-day marine gas-turbine is considerable, but
they all gave their best performance at the same flow. so is the bulk of its steam and Diesel competitors. It is improbable
These little difficulties are, of course, inevitable and to be that engine bulk or weight is restricting the application of the
expected and give no cause for alarm about the future. On the gas turbine to marine use at the moment. Resistance on these
contrary they are reassuring to the extent that they imply that grounds will certainly stiffen if and when the steam turbine
we have much more performance available than would appear achieves the higher pressures and temperatures now being
at first sight. promised for it. When this threat materializes a drastic reduction
I have shown these curves to give some background to my in the size of the gas turbine will be necessary. Fortunately, ways
general contention that so far as aerodynamic and thermodynamic and means of achieving such an objective can be envisaged. The
performances are concerned, progress is being made, and made, most important of these possible reductions in bulk is to be
I think, at a satisfactory rate. Although there is not yet enough found in the largest single part of the engine, the heat exchanger.
data to show curves for marine and industrial gas-turbines, the It is theoretically possible to reduce the length of a con-
same potential progress is being made in these fields, in ventional stationary heat-exchanger almost indefinitely by
the sense that most of the improvements being made in the reducing the cross-sectional dimensions of its passage ways; but
efficiencies of aircraft components can be directly applied to no one has so far developed a satisfactory method.of constructing
give better performance on land and sea. such a heat exchanger with very small passages. If the shell and
These advances will continue for a few more years until &he tube type of construction is used there is the problem of fixing
law of diminishing returns makes further effort uneconomic. the tubes into the end plates, and since in a set of only 10,000 kW.
We can say therefore that, barring the introduction of novel the number of tubes would be measured in tens of thousands,
mechanisms which may open up at present unthought of lines of the difficulty is very great. In the plate type of cross-flow
advance, the aero-thermodynamic future is fairly clear. By this exchanger with secondary surfaces the problem is rather that
I mean that it is clear how far the laws of aerodynamics and of brazing or welding a considerable number of plates to get,
thermodynamics will allow us to go in practice. and maintain during running, good thermal contact. It is possible
What is now hindering our progress and preventing us from that these difficulties will be overcome, but in the meantime the
grasping the prizes that we have been working towards these rotary or capacity type of heat exchanger is a promising alterna-
last ten years, are the practical difficultiesthat always arise in an tive.
endeavour to apply principles. Work has been progressing in Britain on the high-pressure
Downloaded from pme.sagepub.com at University of Leeds on June 5, 2016
THE G A S T U R B I N E I N PERSPECTIVE 191
rotary heat-exchanger for several years, and already at the porous material. The practical problems of manufacture and
National Gas Turbine Establishment (N.G.T.E.) three different maintenance of clearances favour holes rather than slits; the
units have been designed and tested. porous wall, while giving more even cooling over the porous
Progress in reducing the leakage of compressed air-and this surface, has given us trouble at junctions where the porosity is
is the principal remaining problem-has been so good, that in lost and has shown a deterioration with time owing to choking
my opinion there is no doubt that this type of exchanger can be of the interstices by impurities in the air. If air cooling is
developed into a reliable and satisfactory element of the gas used in the combustion-chamber and associated ducting and
turbine. However, it is still a matter of opinion, and we cannot this air is later led into the gas stream, no heat loss is associated
yet offer proof. Once this proof has been provided, the way will with the cooling. There is therefore no advantage to be gained
be open to considerable reductions in heat-exchanger bulk, by the use of porous wall liners upstream of the turbine, but in
particularly in the type of engine whose layout favours a contra- the turbine itself porous walls for stationary members appear
flow arrangement. very attractive.
A description of the work that we have been doing on the The cooling of the blades can be effected in a number of
rotary heat-exchanger will be given in a lecture later on in this ways. Steam cooling by the closed thermosiphon method
series. involves dissipation of heat from the blade root by cooling-air,
and this is a very awkward process to carry out in the cramped
The Cooled Turbipze. While talking of the reduction in the space available. Cooling by the open thermosiphon method, in
size of marine gas-turbines I can with good continuity pass which the steam is led away from the turbine by a running
easily to the matter of the high-temperature or cooled turbine. gland, offers advantages in that the heat lost to the coolant can
For one of the greatest advantages we shall reap from running be used in an auxiliary steam turbine. We have, however, found
at higher gas temperatures is a reduction in size as a result of practical difficulties in operating such a system. If air cooling
lower specific air consumptions. High gas temperatures give is used the cooling-air will probably have to be discharged into
idvantages not only in reduction in size and weight, but also, the gas stream. The easiest place to discharge is at the blade
except in jet engines, in the raising of efficiencies ;these advan- tips; this may cause losses due to disturbance of the boundary
tages apply not only to marine engines but to all applications layer, but so far we have not been able to detect them. A better
of the gas turbine except the jet engine. place to discharge would be from the trailing edge of the blade,
It cannot be said that the absence of the cooled turbineis but considerable difficulty is attached to the design of the
positively holding back the gas turbine, but its advent will blade ducting to lead the air to this part of,the blade. Stator
certainly provide a stimulus, and one that is much needed for blade cooling may be achieved by the use of hollow blades with
certain applications. porous walls, and rotors and stators may be cooled by a protective
The high-temperature cooled turbine raises many questions boundary layer of air led into the gas stream near to the nose of
to which there are not at the moment precise answers. There the blade.
are several types of turbine in the running, each offering A difficulty that has to be overcome with any method of
advantages and exacting a price in return. The conventional internal blade-cooling is that of obtaining a reasonably uniform
axial-flow type of turbine promises the highest efficiency in the distribution of temperature along the blade length and along its
uncooled state; but whether it will still lead the way in efficiency, chord.
when the losses associated with cooling it are allowed for, is To my mind the importance of the cooled turbine cannot be
another matter. The construction of cooled blades for an axial- exaggerated. At present most designers are restricting them-
flow turbine appears at the moment to be difficult and expensive. selves, for long life plants, to the very moderate gas temperature
The axial-flow turbine is, in fact, the most complicated of the of 650 deg. C. (1,200 deg. I?.). By cooling to an extent far less
various possible types. The inward-flow turbine is a simpler than we have already used, this temperature can be raised to
piece of machinery to make, partly because its heat-drop per 850 deg. C . (1,560 deg. F.), without any rise in temperature of
stage is high and its number of blades per stage is small, As a the metal. The use of such a temperature, combined with double
result, we are led to a rotor with a comparatively small number compounding to give a pressure ratio of 911 would make possible
of blades which can be made quite robust. The probability is an overall efficiency of 27 per cent in a very compact plant
that more efficient methods of cooling can be adopted than are without the use of intercooling, reheat, or heat exchangers.
possible with the axial-flow type. On the other hand the inward- So far, we have operated three different forms of cooled turbine
flow turbine probably has a lower uncooled efficiency than its at the N.G.T.E. and are slowly finding out by practical experience
competitor. The radial entry ducting may be advantageous in the real difficulties. At this stage there is insufficient evidence
keeping hot gases away from the bearings and in fitting in with for certainty as to the best layout and best cooling medium for
the general geometry of the engine layout; but it will certainly a cooled turbine. Personally, I am leaning strongly towards the
introduce severe cooling problems in the ducting leading from conventional type of axial turbine with air cooling. You may be
the combustion-chamber. A third possibility i s a displacement in a better position to make up your own minds on this matter
turbine. In such a turbine the robust rotor design may make when in a later lecture in this series you hear the experience that
cooling comparatively easy to achieve and will keep the cooling we have gained.
losses to a low value. On the other hand, this type of machine is
relatively bulky and inefficient even when operating with The Use qf SoZid FzleZs. As another example of the special
clearances less than those normal to the flow type of machine. problems holding back the more general application of the gas
The special problems of the flow type of turbine are cooling turbine I will briefly mention the combustion of solid fuels;
the entry ducting and cooling the blades. The problem of cooling briefly, because as yet there is very little that I can say. It is
the entry ducting will be eased by carrying out the combustion accepted that the useof the gasturbine in base-load power stations
process as close to the turbine inlet as possible and avoiding depends on its capacity to burn cheap fuels, such as coal. We have
sharp changes of direction in the ducting. If this process were not as yet mastered the technical difficulties of burning solid fuel
carried to extremes i t would involve an annular chamber for under the special conditions required in a gas-turbine com-
the axial turbine and an inward-flow chamber of most unusual bustion-chamber. There is no doubt in my own mind that these
design for the radial turbine. Neither of these designs are difficulties-the pumping of the fuel and the disposal of the ash,
impossible but they both introduce problems which are not for example-will be overcome fairly easily. But it is not so clear
present in the more usual, pipe type of chamber. It is therefore to me whether the machinery and appliances embodying these
probable that the pipe chamber, with connecting ducting, will technical processes may not prove so complex as to nullify the
be favoured at first. It is important to reduce the loss of heat essential simplicity of the gas turbine. There is a danger in
to the coolant as much as possible, and this involves main- pushing a good thing too far, and we must be careful not to let
taining the wall surfaces at as high a temperature as is con- the soundness of our judgement be compromised by our mastery
venient. For this reason water cooling is unsuitable. Steam of technology.
cooling on the outside of the walls or air cooling on their inside
are possible alternatives. If air cooling is used, it may be led Scale Effects.Another possible application of the gas turbine,
into the main stream through slits, holes, orDownloaded
the interstices as a small
of a at University
from pme.sagepub.com of Leedspower
on June 5, plant
2016 for road traction, small ships, etc., is held
192 I N T E R N A L COMBUSTION TURBINES
up by the problem of scale effects. The well-known laws of many other good things, will have to be marked “for export
Osborne Reynolds still limit us, and ordain that with the types only”. My only hope is that in this case we shall not be required
of machines at present used losses must inevitably increase as to retain for our own use the austerity or coaI-burning model.
the scale is reduced. The result is that the gas turbine becomes
increasingly inefficient as its size is reduced. Knowledge of the
effects of Reynolds number on the performance of gas-turbine CONCLUDING REMARKS
components is still imperfect and subject to modification. Using It is not so very long ago that the belief was widespread that
the best information available to me, I have shown in Fig. 11 the there was very little difference between the general processes of
development for piston engines and for gas turbines. The basis
of these processes was that one made an engine, without over-
much intellectual strain, and ran it until something broke. One
then strengthened the part that had failed and ran the engine
again. After several thousand hours of this procedure the result
was a good engine, refined in the infallible furnace of trial and
error.
There was a great deaI to be said for this method of getting the
best out of a piston engine, because the intermittency of the
airflow and the reciprocation of the mechanism made precise
calculations on the airflow, the heat-flow, and the stresses quite
impossible.
I,dOO 2.030
BRAKE HORSE-POWER
There is now, I am glad to say, a growing body of opinion
that there are better ways of dealing with the gas turbine. In this.
Fig. 11. Effect of Engine Size on Specific Fuel Consumption, machine, or at any rate in its constant-pressure manifestation,
Expressed as a Percentage of the Fuel Consumption of a the airflow can be made comparatively steady, and the op-
Large Engine portunities for the development of dangerous vibrations can be
severely circumscribed, so that the whole problem becomes
probable change in fuel consumption of a simple form of axial amenable to the ordinary processes of induction.
gas-turbine with size, the size being expressed in terms of power It is becoming increasingly possible in the design stage to
output. The curve refers to geometrically similar units and consider each individual part of the machine in relation to the
indicates only the effects of alterations in the ratio of viscous to conditions under which ultimately it will have to work. The
inertia forces. I n practice, the losses in the small sizes will be blading and passages containing the airflow can be studied in
greater owingto the greater difficulties inmaking parts accurately, wind tunnels and their aerodynamic performances assessed. The
and will be less, because designers will evade the laws of similarity blading and other parts exposed to impulsive forces can be
by, for example, increasing the chords of the blades. The engine subjected to vibration simulating the conditions encountered in
with centrifugal components will have a higher consumption service and their responseschecked against predictions. The heat-
than the axial type at the larger powers and a smaller consumption flows and temperature gradients under steady and transient
at the lower powers, but the point at which the curves cross is conditions can be predicted and the resulting stresses, distor-
not known to me. tions, and changes of clearance determined.
There is no doubt that small engines can be made; that has I must make it clear that none of these processes can as yet
now been demonstrated in the most practical manner. Whether be performed with as much certainty as we would like, or indeed,
they will be used depends on whether their light weight and as is necessary, and empirical corrections and ad hoc measure-
special characteristics are so highly valued as to discount their ments still have to be widely used.
inherently high fuel-consumption. With the further increase in knowledge that I am sure will be
given to us we should reach a position in which the first test
of a major component, such as a compressor, will be regarded as
T H E LOCOMOTIVE the culmination of its development rather than the start; and
In a lecture which I gave before this Institution in November the fast assembly of all the components to form the complete
1948 I referred to the prospects of the locomotive gas-turbine. engine will be the signal not for the development engineer to
During the two years which have since elapsed a number of take off his coat and roll up his sleeves but for him to “call it
locomotives have been built. These engines have been of simple a day” and look round for fresh worlds to conquer.
design and have given efficiencies under operating conditions I must admit that I am to some extent anticipating and that
comparable with the steam locomotive. The caution which what I have said would still be regarded in some circles as
designers have displayed in the general layout and detail design heresy today, but I am convinced that it is the road along which
of these new locomotives is understandable, and their desire to we should travel. Trial and error is a very expensive technique.
introduce as few untried features as possible must be sympathized If it must be exercised it should be done on the smallest possible
with. Nevertheless, the result is a machine which in capital cost, scale. It is much cheaper to find out a stress concentration by
running cost, and thermal efiiciency can offer, I suspect, little vibrating a single blade than by wrecking a compressor or even
or no attraction to the railway engineer. It is to be hoped that the a complete engine.
future of the gas-turbine locomotive on the railways of the world Thus I come to the conclusion of my lecture and to its moral,
will not be judged by the behaviour of these early machines. which is simply “More haste, less speed”, a stricture which, of
These should be regarded as no more than pioneers, the gas- course it would be inappropriate to apply to British industry in
turbine equivalent of Stephenson’s “Rocket”. general.
Even when the high-performance gas-turbine locomotive has
been produced, it must be admitted that its prospects in Britain AcknowEedgement. Acknowledgement is made to the Chief
look a trifle bleak. The high traffic density experienced on many Scientist, Ministry of Supply, for permission to deliver this
of our lines greatly favours electrification, and works to the lecture. It is Crown copyright reserved, and reproduced with
disadvantage of the high-performance self-contained locomotive. the permission of the Controller of His Majesty’s Stationery
If this proves to be true then the gas-turbine locomotive, like Office.

Downloaded from pme.sagepub.com at University of Leeds on June 5, 2016

You might also like