Gas Turbine Cycle Design

Methodology: A Comparison of
Parameter Variation With
Numerical Optimization
In gas turbine performance simulations often the following question arises: what is
the best thermodynamic cycle design point? This is an optimization task which can
J. Kurzke be attacked in two ways. One can do a series of parameter variations and pick from
DASA-MTU Munchen GmbH, the resulting graphs the best solution or one can employ numerical optimization
Engine Performance Department TPSZ, algorithms that produce a single cycle that fulfills all constraints. The conventional
Dachauer Str. 665,
Munchen, 80995
parameter study builds strongly on the engineering judgement and gives useful infor-
Germany mation over a range of parameter selections. However, when values for more than
a few variables have to be determined while several constraints are existing, then
numerical optimization routines can help to find the mathematical optimum faster
and more accurately. Sometimes even an outstanding solution is found which was
overlooked while doing a preliminary parameter study. For any simulation task a
sophisticated graphical user interface is of great benefit. This is especially true for
automated numerical optimizations. It is quite helpful to see on the screen of a PC
how the variables are changing and which constraints are limiting the design. A
quick and clear graphical representation of trade studies is also of great advantage.
The paper describes how numerical optimization and parameter studies are imple-
mented in a Windows-based PC program. As an example, the cycle selection of a
derivative turbofan engine with a given core shows the merits of numerical optimiza-
tion. The parameter variation is best suited for presenting the sensitivity of the result
in the neighborhood of the optimum cycle design point.

1 Introduction 2 Parameter Studies

The traditional way to select the thermodynamic cycle of a The results of parameter studies are normally presented as
new gas turbine employs extensive parameter variations. For graphics. In a single graph one can show for given ranges of two
a complex engine with many design variables this is a time- parameters the results for several dependant quantities. In a cycle
consuming task. One looks for the optimum solution in a certain study for a single spool turbojet engine, for example, with com-
respect. pressor pressure ratio and burner exit temperature as design pa-
rameters, one can plot the specific fuel consumption over specific
Instead of screening a wide range for the design variables
thrust. In the resulting carpet one can additionally show lines for
with systematic parameter variations it is also possible to do an
other calculated parameters as for example the turbine pressure
automatic search for the optimum engine design with the help
ratio and the turbine exit temperature, see Fig. 1.
of numerical optimization routines. This can be done for exam-
ple with the Windows-based gas turbine performance program Let us assume, for example, that the design aim is a low cost
turbojet with a single stage turbine and an uncooled turbine exit
GasTurb developed by the author.
casing. The feasible region of design parameter combinations
Certainly it is not sufficient, to get a single cycle as the best
can be marked easily in the carpet since it is limited by the
solution from the computer program for two main reasons.
following constraints:
Firstly, the numerical optimization algorithm will find the
optimum of the mathematical model only as opposed to the turbine pressure ratio < 3.5 single stage turbine
"true" optimum. An exotic cycle as result of an optimization
run is mostly a hint to a deficiency of the model. In such a case turbine exit temperature < 120OK uncooled turbine exit casing
most probably a design constraint has been overlooked when From Fig. 1 one can read that with these constraints the cycle
defining the problem. with the highest specific thrust has a burner exit temperature of
Secondly, it is always of interest to know about the neighbor- 1600K and a compressor pressure ratio of around 16.5.
hood of the optimum solution. From a parameter study limited However, most gas turbine cycles are much more complex
to the region of interest it becomes obvious which design vari- than the turbojet example discussed above. A two-spool turbo-
ables and constraints have the biggest impact on the result. fan has the following five cycle design variables: low spool
One of the advantages of numerical optimization is, that the pressure ratio; high spool pressure ratio; burner exit tempera-
region where parameter studies should be performed is nar- ture; bypass ratio; and fan pressure ratio. There will also be
rowed down significantly. more constraints than with the turbojet example, as follows:
lp turbine inlet temp. < limit
Contributed by the International Gas Turbine Institute and presented at the uncooled lp turbine
International Gas Turbine and Aeroengine Congress and Exhibition, Stockholm,
Sweden, June 2 - 5 , 1998. Manuscript received by the ASME Headquarters April compressor exit temp. < limit
1, 1998. Paper No. 98-GT-343. Associate Technical Editor: R. Kielb. material of compr. disk

6 / Vol. 121, JANUARY 1999 Transactions of the ASME

Copyright © 1999 by ASME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Compr.
1700
32.5 7.5 Pressure
Ratio
<##» 1600,
10
30
qp- 1500^ 12.5

**1«M 15
27.5

25
Turbine
22.5 Pressure
Ratio
20
Fig. 2 Optimization strategy
17.5
12 14 16 18 20
Net Thrust [kN]

Fig. 1 Result of a parameter study The steepest ascent may, however, lead toward a border
(which is either the lower or upper limit of a design variable)
of the region. Then our mountaineer will walk along the border
until he reaches the place where each step leads downwards or
fan pressure ratio < limit out of the allowed region.
single stage fan Is that the end of the story? Not necessarily. There might be
several summits within the region. Our mountaineer may have
hp turbine press, ratio < limit found the highest peak by chance, but he cannot be sure of that.
single stage turbine He has to check other parts of the region. In mathematical
terms there might be "local" optimums besides the "global"
In a more detailed study there will be even more design optimum.
variables as for example the stage numbers for the high and the
Up to now we have not spoken of constraints. They are like
low pressure turbine. It is obvious that with a parameter study
fences. A part of the region is forbidden to our mountaineer.
it will be very difficult and time consuming to find the optimum
His task is made more difficult because on his way to the summit
values for the design variables.
he may have to walk downwards for a while to avoid a forbidden
region. The fences (the constraints) often exclude the summit
3 Numerical Optimization (where each step leads downwards) as an acceptable solution.
They create local optima that would not exist without fences.
By the way, how is the optimum defined in a mathematical Constraints make the task of optimization difficult.
sense? In a parameter study that question must not be answered
Let us turn to the mathematical algorithm now. The moun-
a priori. In a numerical optimization, however, a figure of merit
taineer who first makes test steps in several directions uses the
must be clearly defined before the calculation can commence.
"gradient strategy" as a search method. With the test steps he
The figure of merit might be the specific fuel consumption of
is looking for the partial derivatives dZ/dVt. For each optimiza-
a turbofan at cruise which is to be minimized. For a fighter
tion variable he must do one test step before he can start his
engine it might be that the specific thrust shall be maximized.
way in the "right" direction.
One can also think of a weighted combination of these parame-
ters. After the first step uphill the local gradients will be different.
The test steps could now be repeated to find the new direction.
When values for more than a few variables have to be deter-
Test steps take time, however, and it is therefore better to go
mined while several constraints are existing, then numerical
on in the same direction as long as the altitude increases. Reach-
optimization routines can help to find the mathematical opti-
ing a fence (violating a constraint) could be another reason for
mum (i.e., the minimum and maximum, respectively, of the
stopping the climb. Only then will new gradients be sought.
figure of merit) faster and more accurately. As shown above
The new direction will eventually take you along a fence.
with the turbojet example, in a parameter study with only two
variables it is easy to find an optimum solution. If there are The gradient search algorithm implemented in GasTurb was
three variables the situation is not so clear. With more than derived from [2]. The principle is the following (see Fig. 2).
three variables the picture may get obscure. In complex studies We begin at the point marked "Start 1", looking for the direc-
the true optimum may never be found with the conventional tion of the steepest gradient ("Direction 1"). Following this
parameter study. direction we walk to the highest point. Then we change the
direction by 90 deg (orthogonal). This can be done without
There are many numerical optimization algorithms known evaluating the local gradient. We again go for the highest point
from literature. They can be divided basically into the following here. To define the third direction we use the experience from
two major groups: methods that use gradient information and the first two directions. We connect the point "Start 1" with
others. In the program GasTurb there is one method from each the optimum point found along "Direction 2 " . We follow this
group implemented. A short explanation how these algorithms direction again as long as altitude increases.
work is given in the following chapters.
This procedure can be applied repeatedly until the search
3.1 Gradient Method. The following is a good example steps or the changes in the "figure of merit" become very
for the optimization task. A mountaineer shall climb the highest small. There is also a maximum limit for the number of optimi-
peak in a certain region. He has no map and the weather is zation steps. In the example of the figure the optimum is found
foggy. His only tool is an altimeter. What is he going to do? along search direction 6 (not marked in the figure, perpendicular
He will certainly check his surroundings first and then go in to direction 5).
the direction of the steepest ascent. In the end he will come to The dashed line in the figure shows how optimization would
the top of a mountain. This is a place where each step leads go on, if only local gradient information is used. With this
downwards. simple strategy, the search direction would change very often.

Journal of Engineering for Gas Turbines and Power JANUARY 1999, Vol. 1 2 1 / 7

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Up to now we have only dealt with optimization without
constraints. In the figure there is a shaded zone which suggests
a forbidden region. If we use the strategy just described the
search for the optimum will end at the point " A " along ' 'Direc-
tion 5 " . We cannot find the global optimum if we begin at
"Start 1". If we begin at "Start 2 " , however, we will be at
the top of the hill very quickly.
3.2 Random Search. The second optimization strategy
offered by GasTurb is based on [3]. In an adaptive random
search, random numbers are used for the optimization variables
that are concentrated around the best solution found before. The i HPkDugftobypns
6 NSVcooUng
algorithm is lookaoa from b)p«K c WT cooling
ovwtoirdblMds

Vt = Vf + — {2® - 1)*- Fig. 3 Configuration of the basic engine

kR
with
4.2 Design Variables, Constraints and Figure of Merit.
Vt •• new value for optimization variable Design Variables. Besides the pressure ratios of the new
V? - value of V, producing the best figure of merit booster and the fan there will be the bypass ratio and the burner
R, -- search region for variable V, exit temperature among the design variables of the growth en-
kR -- range reduction coefficient (positive integer) gine. A new low pressure turbine will be required while the gas
fcv - distribution coefficient (positive odd integer) generator remains unchanged. The configuration of the growth
® = random number between zero and one engine will be as shown in Fig. 4.
The core compressor of the new engine must not necessarily
To start an adaptive random search one should have a variable be operated at the same operating point as in the basic engine.
combination which fulfills all of the constraints. At the start of In fact that might even be impossible because doing that would
the search kR is 10 and k, is 1. In one search run, the program require an increase in the mechanical spool speed beyond the
tries (40 times the number of Optimization Variables) random limits of the original design. Thus, we get as two further design
engine cycles. When all cycles have been calculated, then kR variables for the derivative engine the core compressor mass
will be duplicated and k, will be increased by 2. The search flow and its pressure ratio.
region will get smaller. Another (40 times the number of Opti- It is standard practice not to read a compressor map with
mization Variables) cycles will be calculated and then kR will given mass flow and pressure ratio, but with given corrected
be duplicated again and k, will be further increased by 2. This speed and a value for an auxiliary coordinate (here called beta)
procedure will be repeated until all cycles for kR = 80 have (see, for example, [4]). In the list of the design variables we
been tried. Cycles that do not fulfill the constraints will be get instead of the compressor mass flow and its pressure ratio
ignored. the two equivalent variables corrected speed and map coordinate
beta.
4 Cycle Selection for a Derivative Turbofan Altogether there are six design variables for the derivative
engine.
A very common design task is to adapt an existing engine
for a new application. It is quite obvious that in this case there Constraints. There are several constraints for the new en-
are more constraints than during the design of a brand new gine design to be observed. The common core with the basic
engine. In this chapter at first the basic engine will be described engine requires that both high pressure turbines have practically
and then the design variables, the constraints and the figure of the same flow capacity. We want the Mach number at the core
merit for the numerical optimization of a derivative engine. exit also to be nearly the same and that has the consequence,
that the flow capacity of the low pressure turbine must also be
4.1 Description of the Basic Engine. Let us assume that very similar between both engines. As a consequence there will
we can start from an existing unmixed flow turbofan engine for be practically no difference between both engines with respect
a business jet. This type of engine has a rather low overall to the high pressure turbine pressure ratio.
pressure ratio and a moderate burner exit temperature compared
A further constraint is, that the low pressure turbine inlet
to the big turbofan engines used on commercial airliners. The
temperature T45 must be below say 1150K which allows to
main cycle parameters are shown in the table below.

Table 1 Main cycle parameters of the basic engine ©

Flight Condition 11 km/Mach 0.8
Max Climb
installed

Thrust 3.64 kN

SFC 19.7 g/kN*s

Bypass Ratio 4.5

Burner Exit Temperature 1350 K

Overall Pressure Ratio 17.82

• HPMugotobyput
b NGVo
Core Pressure Ratio 12 lotkagofrombypass^ ' ovwtoan] Htoos c HPT cooing

ISA Corrected Mass Flow 60 kg/s

Fig. 4 Configuration of the growth engine

8 / Vol. 121, JANUARY 1999 Transactions of the ASME

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 design an uncooled low pressure turbine from inexpensive mate- Table 2 Ranges for the design variables
rials.
The core compressor will eventually cause several constraints min value max value
for the design. There might be a temperature limit when the fanP/P 1.1 1.9
last stage is made from titanium, for example. Also, a mechani-
cal speed limit may exist. And last but not least the minimum booster P/P 1.4 2.3
surge margin requirements must be met. bypass ratio 4 6
Another constraint may come from the nacelle in which the burner exit temp 1300 K 1600 K
engine has to be installed. This will limit the fan diameter of
the growth engine. relative core compr. corr. speed 0.9 1.02
In our example the task is, to increase the Max Climb thrust core compr. map coord, beta 0.3 0.8
by 25 percent. Engine designs with less thrust than required
will not be acceptable, and, therefore, the thrust is a design
constraint for the growth engine. In summary, the design con-
straints for the derivative engine are: Before the numerical optimization algorithm can start, we
need to define a range for the design variables. On one side this
hp turbine flow capacity = reference value ±5 percent range should be as narrow as possible because then the search
for the optimum will take less effort. However, when the range
lp turbine flow capacity = reference value ±5 percent is too narrow, then the true optimum might be excluded from
lp turbine inlet temp TAi < 1150K the search unintentionally.

compressor exit temp T3 < 750K 5.2 Starting Point. Many numerical optimization algo-
rithms require that a set of design variables that fulfills all
core spool speed < reference + 5 percent constraints must be known before the calculation can com-
mence. The cycle of the basic engine is within the ranges of
fan tip diameter < 0.75m
all design variables, however, obviously it does not fulfill the
Max Climb thrust > 4.5 kN minimum thrust constraint.
How can we get a valid cycle to start with? One possible
Figure of Merit. The specific fuel consumption (SFC) for approach would be, to do a rough parameter study which has
Max Climb rating is the figure of merit which is to be mini- only the aim to find a feasible solution, but not the best solution
mized. This will automatically result in a low fuel consumption for the problem. However, this parameter study takes more
for cruise. effort than necessary. We can redefine the figure of merit for
the moment and do a slave optimization with the aim of max-
4.3 Mathematical Model of the Engine. A mathematical
imizing the Max Climb thrust. The minimum thrust constraint
model of the growth engine requires a mixture of design and
is dropped for that preliminary exercise which makes the cycle
off-design calculations. The components on the low pressure
of the basic engine valid as a start point.
spool will be newly designed while the core components will
be operated at some off-design condition compared to the design While the slave optimization is running, one can observe on
point of the basic engine. the computer screen the progress. As soon as a cycle is found
which has more Max Climb thrust than required (and fulfills
We select as the cycle design point for the growth engine
all constraints) we can stop. Now we redefine the figure of
the Max Climb rating at altitude. For this flight conditions the
merit as specific fuel consumption and introduce the minimum
optimum values for the design variables will be found.
thrust constraint. The final optimization can commence now.
The mathematical model of the engine must take into account,
that the design point efficiencies of the fan, the booster and the 5.3 Graphical User Interface. Figure 5 shows the opti-
low pressure turbine will change with the aerodynamic loading. mization window of GasTurb for the example of this paper with
For axial compressors an appropriate correlation has been pub- six horizontal gauges for the design variables on the left and
lished by Glassman [5] and for the low pressure turbine one seven gauges for the constraints in the upper right part. The
can use a simplified version of the preliminary turbine design gauges are continuously updated while the optimization is run-
routine from Warner [6]. ning. In the lower part of the screen the figure of merit is shown
The efficiency and the surge margin of the core compressor
will be read from the map dependent from the values for the
design variables core compressor corrected speed and map coor-
1-lnlxl
dinate beta. gow Strategy anltrt StvlMy &»mm Outgltt**
Note that the temperature limits for r 3 and r 45 in the list of
constraints are not applicable to the Max Climb rating, but for
1,1 Ouler Fan Pressure Ratio =T? | f Stan |

the flight case with the highest temperatures encountered in the 1A IP Compressor Pressure Ratio 2,3 » Slop | 40 LPTInl Corr Flow VW5R«td 53
flight envelope. That means, that the numerical model of the 4 Design Bypass Ralio 6 0 LPT InUl Ttmparaturt T45 1160
engine must be capable to simulate both the Max Climb flight
1300 Burner Exit Temperature 0 HPC Exit TtmptrMura T3 750
case at altitude (as a cycle design point) and the Take Off rating 1
for the "hot day" (ISA + 15K) at sea level, Mach 0.2 (as an 0^ rel NH/sqrt(T25/T$td) ^ 0J9 Design Pt HP Spool Speed

off-design condition). 0,3 Auxiliary Coordinale Beta

5 Optimizing the Growth Engine AljdjJ

Flgur»otMtrft
5.1 Ranges for the Design Variables. Another argument 18.9401
for setting the range of a design variable is that either the lower
or the upper limit represents a true limit for the engine design.
In our example, this is the case for the pressure ratio of the
single stage fan which is introduced with an upper limit of 1.9.
In the table below the ranges for all six design variables are
given. Fig. 5 Graphical user interface of Gas Turb

Journal of Engineering for Gas Turbines and Power JANUARY 1999, Vol. 121 / 9

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Table 3 Cycle parameter summary

Basic Engine Growth Engine

Max Climb Hot Day Max Climb Hot Day

Take Off Take Off

Thrust 3.61 13.10 4.50 16.94

[KN]

SFC 19.86 14.23 18.93 13.33

IflWs)]
Bypass 4.5 4.65 5.06 5.23
Ratio

Fan 1.775 1.62 1.73 1.60

P13/P2

Ideal Jet 0.761 0.886 0.726 0.839

Vel. Ratio

Booster 1.5 1.33 1.80 1.61 Net Thrust [kNJ

P24/P2
Fig. 7 Sensitivity for HPC map operating point
HPC 12 11.36 12.3 11.67
P3/P2S

T4N 1350 1479 1393 1530

low pressure turbine design and the operating points in the
W41Rstd 1,35 1,35 1,31 1,31 component maps at off-design.
W4SRstd 4,98 4,96 5,01 5,00 5.4 Local and Global Optima. As explained in the chap-
T3[K] 610 708 649 750 ter about the gradient search strategy with numerical optimiza-
tion there is always the danger that the algorithm finds only a
T45|K] 973 1076 1000 1108
local optimum but not the global optimum within the parameter
range. When there are several local optima within the feasible
region, then it depends on the starting point of the algorithm
both as numbers and as graphic with a dot for every valid which local optimum will be found. Therefore, one should re-
solution. peat the optimization run several times and pick from all local
One can immediately see from the gauges when a variable optima the best one.
or a constraint is driven toward a range boundary respectively One can find easily a new starting point for the optimization
limit. When the range boundary of a design variable happens by redefining the search direction. Instead of minimizing the
to be not a true limit for the engine design then one can stop specific fuel consumption one looks during a restart run for the
the calculation and redefine the range for the corresponding cycle with the maximum SFC. The random adaptive search will
design variable. lead for each restart run to a different starting point even when
In practice it happens quite often, that during the first attempts it commences several times from the same optimum.
the optimization problem is not formulated correctly. In such a In Fig. 5 one can see from the graphics for the figure of
case the numerical algorithm drives the mathematical model in merit, that twice a restart has happened. This was caused by
a direction which is obviously nonsense because a constraint the algorithm called "endless random search" which restarts
was forgotten, for example. Therefore, an easy to survey graphi- automatically after the algorithm has homed into an optimum.
cal user interface is very helpful for avoiding a waste of comput- The best solution found will be stored in memory and can be
ing time. restored as soon as the calculation is stopped.
The optimization can be stopped at any time, which allows Some important data for our growth engine example is sum-
to check the best solution found in more detail than possible marized and compared to the basic cycle in Table 3 below.
from the values for design variable and constraints alone. For
both the engine design point (Max Climb at altitude) and the 6 Discussion of the Results
off-design condition (SL Take Off ISA + 15K Mach 0.2) there The optimum growth engine is influenced by three of the
are all details accessible. This includes even graphs with the design constraints. It has a fan diameter of 0.75 m, i.e., it uses

27.5i

104 i .
3 8 4 4
-2 4
i. .T, 4
- 6
-^ 4
- 8
5 5.2 5.4
Net Thrust [kN]
Fig. 8 Sensitivity for bypass ratio and burner exit temperature
Fig. 6 Sensitivity for booster and outer fan pressure ratio

Transactions of the ASME

10 / Vol. 121, JANUARY 1999

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 the largest fan which was allowed in this exercise. The second Figure 8 deals with the operating point of the high pressure
constraint which had an impact on the design of the growth compressor. High values for the auxiliary coordinate beta go
engine is the compressor exit temperature which was limited to with a low surge margin. The square marks the point with
750 K for the hot day take off case. The third constraint was lowest SFC in the region of feasible designs.
the minimum high pressure turbine flow capacity W41Rstd.
All design variables did remain within the predefined range 7 Summary
during the optimization. The thrust increase for Max Climb With a conventional parameter study it is very difficult to
rating at altitude is 25 percent and at Take Off even 29 percent. find the optimum solution for a problem as soon as four or
Note that both engines run during Take Off speed with 7 percent more design variables and several constraints are involved. With
more mechanical high pressure spool speed than at Max Climb the help of numerical optimization algorithms one can easily
in this example. The specific fuel consumption at altitude is find the mathematical correct solution to the problem. Extensive
nearly 5 percent better for the growth engine. parameter studies around the solution will help to understand
In the table there is also a row for the ideal jet velocity ratio. why this combination of design variables is the best choice and
From theoretical considerations one can derive, that this ratio how sensitive the figure of merit is to small deviations from the
should be equal to the product of fan and low pressure turbine optimum.
efficiency when an unmixed flow turbofan is to be optimized As an example, the cycle selection of a derivative turbofan
for SFC. Note that the numerical optimization algorithm has engine with a given core shows the merits of numerical optimi-
automatically found a cycle for which the jet velocity ratio is zation. The parameter variation is best suited for presenting the
near to its theoretically best value. sensitivity of the result in the neighborhood of the optimum
It is quite useful to do a parameter variation in the neighbor- cycle design point. Sometimes this leads to a redefinition of the
hood of the optimum solution. This gives an insight to the figure of merit or the constraints imposed on the solution. In
sensitivity of the result. Figure 6 shows the influence of "outer rare cases even an outstanding solution is found which was
fan pressure ratio" and "booster pressure ratio" on thrust and overlooked while doing a preliminary parameter study.
specific fuel consumption. Note that the parameter on the verti-
cal axis was mainly selected because it spreads the carpet nicely. 8 References
1 Kurzke J., 1998, "Manual GasTurb 8.0 for Windows—A Program to Cal-
The corrected flow at the low pressure turbine exit is not a very culate Design and Off-Design Performance of Gas Turbines," available from the
important engine design parameter. author.
Actually, the limiting line for the fan diameter (0.75 m) is 2 Jacob, H. G., 1982, "RechnergestUtzte Optimierung statischer und dynam-
ischer Systeme," Fachberichte Messen—Steuern—Regeln, Springer-Verlag KG,
identical to the line for the minimum high pressure turbine flow Berlin.
capacity. To the left of the line for booster pressure ratio « 3 Kelahan, R. C , Gaddy, J. L., 1978, "Application of the Adaptive Random
1.79 the lower limit for the high pressure turbine flow capacity Search to Discrete and Mixed Integer Optimization," International Journal for
is violated and to the right of this line the fan diameter is too Numerical Methods in Engineering, Vol. 12, pp. 289-298.
4 Kurzke, J„ 1996, "How to Get Component Maps for Aircraft Gas Turbine
big. In this graph only one point is a valid solution: the square Performance Calculations," ASME Paper 96-GT-164.
which fulfills the thrust requirement. One can see, that the opti- 5 Glassman, A. J., 1992, "Users Manual for Updated Computer Code for
mum solution is pretty much boxed in by the design constraints. Axial-Flow Compressor Conceptual Design," NASA Contractor Report 189171.
6 Stewart, W. L„ 1961, "A Study of Axial-Flow Turbine Efficiency Charac-
In Fig. 7 two more design variables were systematically var- teristics in Terms of Velocity Diagram Parameters," ASME Paper 61-WA-37.
ied around the optimum solution. Again the square marks the 7 Cohen, H„ Rogers, G. F. C , Saravanamuttoo, H. I. H„ 1996, "Gas Turbine
only point which fulfills all constraints. Theory," 4"1 ed., Addison-Wesley Longman, London, United Kingdom.

Journal of Engineering for Gas Turbines and Power JANUARY 1999, Vol. 121 / 11

Downloaded From: http://gasturbinespower.asmedigitalcollection.asme.org/ on 02/13/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use