UNCLASSIFIED

An Investigation into Performance Modelling of a Small

Gas Turbine Engine

Zafer Leylek

Air Vehicles Division

Defence Science and Technology Organisation

DSTO-TR-2757

ABSTRACT

A small gas turbine performance modelling and testing project has been completed as part of
a Divisional Enabling Research Program (DERP). The main objective of the program was to
enhance DSTO's capability in understanding and modelling the thermodynamic and
performance characteristics of gas turbine engines. Secondary objectives of the program
included the investigation of thrust augmentation technologies and infrared suppression
modelling and analysis. This report presents the results of both numerical and experimental
investigation into engine performance simulation. It outlines the different tools and
techniques used in modeling engine component behaviour and discusses their advantages
and disadvantages. The results of two different tests designed to explore engine operating
regions close to compressor stall and surge is presented.

RELEASE LIMITATION

Approved for public release

UNCLASSIFIED
 UNCLASSIFIED

Published by

Air Vehicles Division

DSTO Defence Science and Technology Organisation
506 Lorimer St
Fishermans Bend, Victoria 3207 Australia

Telephone: (03) 9626 7000

Fax: (03) 9626 7999

© Commonwealth of Australia 2012

AR-015-427
October 2012

APPROVED FOR PUBLIC RELEASE

UNCLASSIFIED
 UNCLASSIFIED

An Investigation into Performance Modelling of a

Small Gas Turbine Engine

Executive Summary

A small gas turbine performance modelling and testing project has been completed as
part of a Divisional Enabling Research Program (DERP). The main objective of the
program was to enhance DSTO’s capability in understanding and modelling the
thermodynamic and performance characteristics of gas turbine engines. Secondary
objectives of the program included the investigation of thrust augmentation
technologies and infrared suppression modelling and analysis.

Engine performance modelling was conducted using commercial software and an in-
house developed code. The engine performance model of the small gas turbine engine
was validated through engine testing. A number of non-standard tests were also
conducted in order to “stress” the engine and compare results with the engine
performance model and to provide an insight into thrust augmentation options. A
number of auxilary tools and techniques were developed during this process.

The primary objectives of the project were successfully reached. An engine

performance modelling tool was developed and validated and different tools and
techniques were evaluated. Infrared suppression related studies were also conducted.

A number of other projects have been initiated based on the tools, technologies and
processes developed during this project. An alternate fuels research program will
directly utilise the testing infrastructure, tools and know-how. A PhD into infrared
prediction technologies is currently under way and a final year thesis completed. A
feasibility study into compressor and turbine performance testing will be completed. A
number of numerical tools will continue to be developed to aid and improve the
modelling of engine performance components.

The tools, techniques and conclusions derived from this project have direct impact on
Australian Defence Organisation outcomes. Engine performance modelling is critical
for ADF gas turbine engine performance analysis and requirements, structural, thermal
and creep assessment, performance degradation studies, infrared suppression
technologies to name a few. The tools and techniques developed in this project will be
directly used to model the performance of ADF gas turbine engines.

UNCLASSIFIED
UNCLASSIFIED

UNCLASSIFIED
 UNCLASSIFIED

Author

Zafer Leylek
Air Vehicles Division

Zafer Leylek has worked in the Infrared Signatures and

Aerothermodynamics group of the Defence Science and Technology
Organisation since 2010. He completed his undergraduate training
at the Royal Melbourne Institute of Technology (RMIT), in
Aerospace Engineering and a graduate certificate degree in
Computer Systems Engineering. Zafer has over ten years
experience in engine performance and aerodynamic design and has
worked on a large scale European engine development program.
____________________ ________________________________________________

UNCLASSIFIED
UNCLASSIFIED

UNCLASSIFIED
 UNCLASSIFIED
DSTO-TR-2757

Contents

NOMENCLATURE

ACRONYMS

1. INTRODUCTION............................................................................................................... 1

2. AMT OLYMPUS HP TURBOJET..................................................................................... 1

3. ENGINE NUMERICAL MODELLING .......................................................................... 3

3.1 System Performance Modelling............................................................................. 3
3.2 Intake........................................................................................................................... 4
3.3 Compressor ................................................................................................................ 4
3.3.1 Map Scaling .............................................................................................. 4
3.3.2 Compressor Performance Map.............................................................. 6
3.4 Combustor .................................................................................................................. 7
3.5 Turbine........................................................................................................................ 7
3.5.1 CFD Modelling ...................................................................................... 10
3.5.2 Mean-line and Through Flow Modelling........................................... 11
3.5.3 Turbine Performance Map ................................................................... 12
3.6 Ducts.......................................................................................................................... 14
3.7 Nozzle ....................................................................................................................... 14
3.8 Oil/Lubrication System.......................................................................................... 15

4. ENGINE TESTING........................................................................................................... 16
4.1 Engine Test Objectives and Configurations...................................................... 16
4.1.1 Engine Operating Line Test ................................................................. 16
4.1.2 Reduced Nozzle Area Test................................................................... 16
4.1.3 Nozzle Exit Flow Blockage Test .......................................................... 16
4.2 Engine Test Facility ................................................................................................ 17

5. RESULTS & DISCUSSION............................................................................................. 19

5.1 Engine Operating Line........................................................................................... 19
5.2 Reduced Nozzle Area ............................................................................................. 23
5.3 Nozzle Exit Flow Blockage.................................................................................... 25

6. CONCLUSION .................................................................................................................. 27

7. RECOMMENDATIONS.................................................................................................. 28

8. REFERENCES .................................................................................................................... 30

UNCLASSIFIED
 UNCLASSIFIED
DSTO-TR-2757

UNCLASSIFIED
 UNCLASSIFIED
DSTO-TR-2757

NOMENCLATURE
b = Combustor part load constant
f = Fuel to mass flow ratio or scale factor
h = Enthalpy
F = Force
P = Pressure
T = Temperature
W = Mass flow rate
V = Volume
π = total-total pressure ratio
η = total-total efficiency
Ω = Combustor/burner loading

Subscripts
t = total thermodynamic conditions
ref = reference
des = design
ds = design speed
D = discharge
FG = gross thrust
N = net
0 = atmospheric conditions
1 = compressor inlet
2 = compressor outlet/combustor inlet
3 = combustor outlet/turbine inlet
4 = turbine outlet/exhaust inlet
6 = exhaust outlet/nozzle inlet
8 = nozzle exit

Superscripts
‘ = relative value in the rotating coordinate system

ACRONYMS
rpm = Revolutions per Minute
CFD = Computational Fluid Dynamics
CORR = Corrected
DERP = Divisional Enabling Research Program
DP = Design Point
DP (MAP) = Component Map Design Point
EGT = Exhaust Gas Temperature
HPC = High Pressure Compressor
HPT = High Pressure Turbine
RNI = Reynolds Number Index
OL = Operating Line

UNCLASSIFIED
 UNCLASSIFIED
DSTO-TR-2757

UNCLASSIFIED
 UNCLASSIFIED
DSTO-TR-2757

1. Introduction
A small gas turbine performance modelling and testing project has been completed as part
of a Divisional Enabling Research Program (DERP). The main objective of the program
was to enhance DSTO’s capability in understanding and modelling the thermodynamic
and performance characteristics of gas turbine engines. A small gas turbine engine was
chosen to reduce the cost and time associated with testing larger engines. As part of this
objective DSTO’s small engine test capabilities were enhanced and an extensive testing
and validation program completed.

Secondary objectives of the program were to: investigate thrust augmentation options;
undertake modelling and analyses of infrared suppression technologies; provide a
foundation and test bed for future alternate fuels research; and initiate co-operative
research programs with universities.

Numerical modelling of engine performance and thermodynamics involved:

 engine disassembly,
 component geometry scanning,
[1]
 compressor and turbine mean-line and through flow analyses using TurbAero
[2]
and CompAero ,
 computational fluid dynamics modelling of the compressor and turbine using
[3]
Numeca , and
 development of a turbojet engine performance simulation code to enhance the
[4]
capabilities of Gasturb .

Testing and validation included:

 designing and constructing a small gas turbine engine performance test bed,
 designing and constructing a bell-mouth,
 instrumentation analysis and installation,
 data acquisition, and
 engine component modifications and installations.

2. AMT Olympus HP Turbojet

The AMT Olympus HP turbojet engine was used as the benchmark engine. Figure 1 shows
the AMT Olympus HP engine.

UNCLASSIFIED
1
 UNCLASSIFIED
DSTO-TR-2757

Figure 1 Olympus HP Gas Turbine Engine

The Olympus HP gas turbine engine is a single spool turbojet engine. It is comprised of a
standard inlet, single stage centrifugal compressor with vaned diffusers, annular
combustor, single stage axial turbine and convergent nozzle. The engine is controlled
using an Engine Control Unit (ECU) and comes with an electric starter motor.

The main engine performance characteristics provided by the engine manufacturer can be
seen in Table 1.

[5,6]
Table 1 Olympus HP engine performance parameters
Parameter Condition (ISA, SLS) Value
Thrust 108000 rpm 230 N
Pressure Ratio 108000 rpm 4
Mass Flow Rate 108000 rpm 450 g/sec
Maximum Speed 108000 rpm
Maximum Allowed Speed 112000 rpm
0
Exhaust Temperature 700 C
0
Maximum Exhaust Temperature 750 C
Fuel usage 108000 rpm 640 g/min
Fuel type Kerosene/paraffin/A-
1/white spirit

The electric starter motor was removed and replaced with an air starter system. The
reasons for removing the electric starter were to ensure uniform flow through the bell-
mouth for mass flow rate measurement, eliminate a source of loss and turbulence to the
engine compressor and provide a standard configuration for all tests.

A three stage process is used to start the engine. The first stage requires shop compressed
air to be directed onto the impeller through a slot in the intake to bring rotation up to 9000
rpm. After this, propane is pumped into the combustor to initiate combustion. Then, once
the engine speed reaches approximately 50,000 rpm, the propane gas valve is shut-off and
replaced with kerosene. Thereafter the engine can be operated normally.

UNCLASSIFIED
2
 UNCLASSIFIED
DSTO-TR-2757

3. Engine Numerical Modelling

The engine performance and thermodynamic cycle is simulated using a combination of
commercial and in-house engine performance codes, empirical based component analysis
tools and CFD. Figure 2 shows a breakdown of the engine components and the tools used
in predicting thermodynamic behaviour.

Engine Performance Simulation Station numbers shown below are used in subsequent
• Gasturb (commercial) analysis. Refer to nomenclature for definition.

• Engine Performance Simulator (Python script) 41

2 3 31 4 5 6 8
Compressor and Turbine Performance Maps
• CFD – Fluent & Numeca
• Empirical (Meanline and ThroughFlow) - TurbAero & Intake
CompAero
Nozzle
• Meanline solver - Python code
• Map scaling
Combustor R e c irc u la tin g
NGV
C o o l.
• Empirical loss models HPT
C o o lin g
Intake/Ducts & Nozzle
• Empirical loss models
Compr Turb

0.8
0.

0 .8
16 4 87

6
4
0.80
Reference 0.8
8
2nd operating line
14 3.6
0.8
9

N=1.1
N=1
N=1.05
N=0.9
N=0.8
12 84 3.2
Compressor 0.
85
Turbine
83

0.
0.
82
0.

10 2.8
HPC Pressure Ratio

Turbine Pressure Ratio

80

N=1.05
0.

N=1

8 2.4
N=0.95
N=0.9

6 75
2
0.
N= 0.85
N=0.8

4 1.6
P3q2

N=0.75

P4q5
N=0.7

0.70
2 1.2
N=

N=
0

0
.6

.5

0 .8
4 8 12 16 20 24 28 32 36 5.3 5.4 5.5 5.6 5.7 5.8 5.9
Mass Flow W2RStd [kg/s] W41*sqrt(T41)/(P4/Pstd) [Kg/s]

Figure 2 Engine and component performance models and tools used to predict their behaviour

3.1 System Performance Modelling

The performance of each component is combined in an engine performance model. A

[4]
commercial code, Gasturb , was used as the main performance simulator. In order to
enhance and adapt various models to the Olympus HP engine an in-house engine
performance code was developed. This code was subsequently validated with Gasturb.

UNCLASSIFIED
3
 UNCLASSIFIED
DSTO-TR-2757

The component performance models are combined to produce a system of non-linear

equations with the engine performance code. For a simple turbojet configuration the
system is solved using the following relationships:

Mechanical coupling: N HPC  N HPT

Energy balance between HPC & HPT: hT 3  hT 2   m (1  f )(hT 4  hT 5 )
Mass flow balance: WHPC (1  f )  WHPT

3.2 Intake

The intake is modelled as an adiabatic system. The intake loss is primarily correlated with
the inlet corrected flow rate or Mach number. In this instance, the dependency is expressed
using the intake inlet and outlet total pressure ratio.

The intake pressure loss ratio was assumed to be constant for the Olympus HP intake. This
assumption was made after considering a) the sensitivity of the intake loss model accuracy
on overall system performance, b) instrumentation constraints and uncertainties in a
comprehensive intake test, c) the lack of resources needed for testing and d) a literature
survey. Test results subsequently showed this assumption to be valid.

The equation set used in the engine performance calculations for the intake are:

Pt 2
 int ake   const  0.98
Pt 0
ht 0  ht 2

3.3 Compressor

The Olympus HP compressor consists of an impeller, radial vaned diffuser and axial guide
vanes. The impeller has seven blades and seven splitter vanes. The radial diffuser contains
fifteen vanes followed by thirty axial guide vanes.

The compressor is arguably the most challenging component for thermodynamic

performance prediction. A number of different approaches to generating performance
curves were evaluated with limited success. The techniques used to generate compressor
performance curves included CFD modelling, mean-line and through-flow modelling,
literature and map scaling.

3.3.1 Map Scaling

Map scaling is a commonly used technique where published performance maps of

compressors with the same configuration and capacity are scaled using known reference
points or curves. The reference point at which maps are usually scaled is the compressor

UNCLASSIFIED
4
 UNCLASSIFIED
DSTO-TR-2757

design point. Assuming that the Olympus HP compressor design point is known or can be
inferred from existing performance data, the equations used to scale the corrected mass
flow rate, efficiency and pressure ratio can be written as:

WCORR , DP
f mass 
WCORR , DP ( MAP ) * fW , RNI

 DP
f 
 DP ( MAP ) * f , RNI

P3 / P2 DP  1
f P3 / P2 
P3 / P2 DP ( MAP) 1

N DP
fN 
N DP (MAP )

Pt Rref Tt , ref  ref

RNI 
Pt , ref RTt 

0.022 ln( RNI )  1 ln( RNI )  0

fW , RNI  
 1 ln( RNI )  0

0.011ln( RNI )  1 ln( RNI )  0

f , RNI  
 1 ln( RNI )  0

Where f is the scaling factor and RNI is the Reynolds Number Index.

Figure 3 illustrates the map scaling process graphically.

Scaled map
Un-scaled reference map

Given Olympus HP
compressor DP:
Reference map
design point (DP) • corrected mass flow
• pressure ratio
• efficiency

Figure 3 Map scaling process

UNCLASSIFIED
5
 UNCLASSIFIED
DSTO-TR-2757

The standard map scaling process is better suited for gas turbine design applications
where the design point data of the new engine is known. Also the map scaling point only
has a single control lever, namely the scaling factor.

This technique has been modified for performance modelling of existing engines where
separate scaling factors are applied to each speed line to better approximate measured test
data. This results in a set of map scaling parameters and allows the reference map to also
be shifted and/or skewed for each speed line.

The reference map used to scale and construct the Olympus HP map belongs to the
[7]
Olympus engine as published in Rahman .

The Olympus HP compressor design point is not published by the manufacturer. The
pressure ratio and mass flow for the compressor is given at the maximum operating speed.
Hence the design point corrected mass flow rate, efficiency and pressure ratio for the
compressor has to be estimated. Although this may yield reasonable results, the engine
test setup and instrumentation were designed so that the compressor corrected mass flow
rate, pressure ratio and efficiency could be estimated at different rotational speeds. This in
turn allows the calculation of the compressor operating line which is used to further refine
or calibrate the map scaling process, enhancing overall engine performance prediction.

3.3.2 Compressor Performance Map

[7]
The Olympus compressor map published by Rahman was used as the baseline map. The
compressor map was scaled using the compressor performance parameters measured
during the operating line test. Figure 4 shows the unscaled and scaled Olympus
compressor map.

Figure 4 Unscaled and scaled compressor map

UNCLASSIFIED
6
 UNCLASSIFIED
DSTO-TR-2757

3.4 Combustor

The combustor is of an annular type with six fuel nozzles. The design of the combustor is
fairly crude and consists of an annular cylinder and six fuel nozzles.

The combustor thermodynamic performance is estimated using common empirical

models. Modern day combustion systems exhibit very good performance with very little
losses. The application of loss and performance models to a combustor with crude design
raises a number of issues and presents added complications. The only way to resolve these
complications is to conduct component testing which was deemed too costly in terms of
time and money. The test requires specialised expertise and testing equipment and the
gain in data was deemed insufficient when considering the overall scope of the project.

Another added complication to the Olympus HP combustor performance prediction is the

inclusion of oil in the fuel system. The fuel is pre-mixed with 4.5% Aeroshell 500 turbine
oil before use. The fuel line from the tank is split into two with one line leading to the
combustor fuel nozzles and the other onto the bearing system. The mechanical design of
the shaft/bearing system is such that the fuel/oil mix returns to the combustor once it has
passed over the bearings. Hence, the complication is doubled in that not only the effect of
the oil on combustion but also the affect of the lubrication fuel/oil returning to the
combustor needs to be accounted for.

The combustor efficiency can be correlated with the combustor loading (Gasturb) which is
defined as:
W
 31
1.8 T3 / 300
P e
3 V
An empirical loss model has been used to estimate the change in combustor efficiency at
off-design conditions. The combustor efficiency at off-design conditions can be calculated
as:
  
log(1   )  log(1   des )  b * log 
  des 
where b = 1.6.

The pressure loss in the combustor is dependent on inlet corrected mass flow rate which is
modelled as a duct (see section 3.6 for details).

3.5 Turbine
The Olympus HP turbine contains a single stage axial turbine with 22 Nozzle Guide Vanes
(NGV) and 29 rotor blades. The turbine is unshrouded. A picture of the Olympus HP
turbine NGV and rotor can be seen in Figure 5.

UNCLASSIFIED
7
 UNCLASSIFIED
DSTO-TR-2757

Turbine NGV Turbine rotor

Figure 5 Olympus HP turbine NGV and rotor

Initially two different techniques were used to predict the performance characteristics of
the turbine. The first method was based on CFD and the second method the mean-line and
through-flow code TurbAero. Subsequent modelling showed that the results needed to be
scaled in order to improve correlation between test and model results; hence map scaling
has also been applied to obtain the final turbine map.

The main input when conducting CFD modelling and to a lesser extent mean-line and
through-flow technique is a detailed knowledge of the geometry. Hence the engine was
dismantled and optical scanning was used to generate a geometric point cloud. Figure 6
shows the optical scan results of the turbine nozzle and rotor.

Turbine NGV Turbine rotor

Figure 6 Optical scan plot of the Olympus HP turbine nozzle and rotor
The optical scan yields a very large dataset. The accuracy is dependent on a number of
factors including part shape, size, surface finish and a number of optical scanning
parameters. The accuracy for the component scan was stated to be below 20-30 micron
which was deemed sufficient for modelling purposes.

UNCLASSIFIED
8
 UNCLASSIFIED
DSTO-TR-2757

Although data points generated with optical scanning are highly accurate, kinks are
formed in regions where optical visibility and access is poor and complex geometric
features are present. Hence, post-processing is required to eliminate these kinks before
proceeding to CFD and mean-line and through-flow analysis. Capturing and defining
sharp corners is also difficult to achieve with a cloud of geometric data points and so
requires special attention and treatment.

A number of blade section profiles were then generated from the post-processed geometric
data points using a 3D CAD package. This is required for clean geometry and mesh
construction. It is also required to define key blade profile parameters for mean-line and
through-flow analysis.

Figure 7 shows the blade and cross section profiles of the turbine nozzle and rotor.

Turbine NGV Turbine rotor

Rotor tip clearance

Blade profile cross section

Figure 7 Olympus HP turbine cross section and blade profile plot

Figure 8 shows some of the key blade profile performance parameters measured for mean-
line and through-flow analysis.

UNCLASSIFIED
9
 UNCLASSIFIED
DSTO-TR-2757

Figure 8 Main blade geometric parameters used in mean-line and through-flow analysis

3.5.1 CFD Modelling

[3]
A CFD model of the Olympus HP turbine was constructed and solved using Numeca
Fine/TURBO suit of tools. Figure 9 shows the Olympus HP turbine CFD model.

Rotor blades

Nozzle blades

Figure 9 Olympus HP turbine CFD model

The mesh contains approximately two million grid points. The air is modelled as a
compressible ideal fluid and specific heats a function of temperature. The solver
recommended standard Spalart-Allmaras turbulence model was used during simulations.
A standard turbine inlet total temperature of 1000K and 300 kPa total inlet pressure was
used as the standard operating condition. The turbine rotational speed was varied from

UNCLASSIFIED
10
 UNCLASSIFIED
DSTO-TR-2757

60K rpm to 110K rpm with 10K rpm increments to obtain 5 speed lines. For each turbine
rotational speed 12 different outlet static pressure boundary conditions were applied to
simulate performance ranging from slow subsonic to choked flow. Hence, a total of 60
CFD simulations were successfully completed to provide an accurate representation of the
turbine performance throughout its operating envelope.

It should be noted that the performance maps are generated in non-dimensionalised form.
Flow similarity in a rotating component is dependent on three independent parameters,
namely the inlet corrected mass flow rate, corrected rotational speed and Reynolds
number. These three parameters can also be expressed in different forms such as Mach
number, flow coefficient, loading coefficient etc. The corrected mass flow rate and
corrected rotational speed are the two primary parameters that determine the component
performance. It has been shown that Reynolds numbers above 100,000 do not alter rotating
component performance and Reynolds number effects are negligible. Since the CFD based
map generation approach ensures that the corrected flow rate and speeds are matched, the
error introduced by the arbitrary selection of inlet and outlet conditions is negligible.

3.5.2 Mean-line and Through Flow Modelling

Another technique used to predict turbine performance is the mean-line and through flow
modelling techniques. Both mean-line and through flow codes use blade cascade tests to
generate empirical performance curves as a function of key blade geometric parameters.
Mean-line codes apply the empirical performance curves at only one streamline surface
along the equivalent mean turbine radius. Through-flow codes extend mean-line tools by
modifying and applying loss and performance models in the span-wise direction. Hence
multiple stream surfaces are used to capture performance and loss from blade hub to tip.
The advantages of using mean-line and through-flow tools are: speed, stability, minimum
input data requirement, pre- and post-processing requirements, insight into turbine flow
physics, and accuracy.

TurbAero, which is a through-flow code, was used to model the Olympus HP turbine. The
code extends the well known Ainley-Mathieson and Dunham-Came (AMDC) mean-line
loss modelling system.

Mean-line techniques conceptualise the loss mechanisms based on turbine features. The
losses are attributed to blade profile, secondary air flow, blade clearance, trailing edge,
supersonic expansion and shock loses and is mathematically expressed as:

Y  YP  YS  YCL  YTE  YEX  YSH

Pt1'  Pt '2
Y
Pt '2  P2
where:

YP = Profile loss coefficient

UNCLASSIFIED
11
 UNCLASSIFIED
DSTO-TR-2757

YS = Secondary flow loss coefficient

YCL = Blade clearance loss coefficient
YTE = Trailing edge loss coefficient
YEX = Supersonic expansion loss coefficient
YSH = Shock loss coefficient

In most cases the profile and secondary flow losses are the primary parameters driving
turbine performance. The profile loss represents losses associated with the blade geometry
and flow incidence. The secondary losses are associated with the interaction between the
flow through the blade passage and the end wall boundary layer. Blade clearance loss
accounts for the additional losses due to tip clearance in unshrouded blade rows. Trailing
edge loss is treated as a classic sudden expansion loss where the effect of the blade trailing
edge thickness can be significant. The supersonic expansion loss and shock loss account
for shocks that occur in certain regions along the blade passage and when flow is
supersonic.

3.5.3 Turbine Performance Map

The results of the CFD generated turbine performance map can be seen in Figure 10.

Figure 10 CFD generated turbine map

A comparison of the CFD and mean-line/through generated turbine maps can be seen in
Figure 11.

UNCLASSIFIED
12
 UNCLASSIFIED
DSTO-TR-2757

Figure 11 Comparison of CFD and mean-line/through generated turbine maps at three different
speeds

From the relative mass flow rate comparison it can be seen that agreement between CFD
and mean-line/through flow results is good. The mean-line/through-flow choked mass
flow rate agrees to within 1% of CFD generated mass flow rate. At pressure ratios below
the choking point the corrected mass flow rate for mean-line/through-flow method
predicts greater corrected mass flow rate when compared to CFD. The agreement between
the total-to-total efficiency results show very good agreement, however; there is a positive
vertical shift in efficiency of approximately 3-5% for the mean-line/through flow results
when compared to the CFD generated efficiency at high rotational speeds. The shift in the
total-to-total efficiency is primarily associated to the fact that TurbAero does not apply a
sudden expansion correction factor to the step change in hub and shroud radii between
the turbine nozzle and rotor.

The turbine performance curve is arguably the most difficult component to experimentally
map. The combustor design is not ideal and adds a large number of complications related
to combustion completeness and profile. The addition of oil into the fuel and the fact that
the fuel/oil mixture exiting the bearing lubrication system is diverted back to the
combustor makes the measurement of turbine inlet conditions extremely difficult. Added
to this complication is the limitations imposed on instrumentation due to high
temperatures and poor accessibility raises a large number of issues regarding turbine map
quality as measured on an engine test stand. Hence, the agreement between CFD and the
mean-line/through flow solver provides a great deal of confidence with the results.

The CFD results were used as the baseline for the Olympus HP turbine map. Data
smoothing was applied to the CFD results to ensure continuity when looking at the multi-
dimensional turbine map from different directions.

Subsequent modelling showed that CFD generated turbine performance under-predicted

choking when calibrating the engine performance model. The under-prediction of choking
mass flow rate can be associated with a number of causes related to engine performance
modelling, engine component design, and instrumentation and testing. Another source of

UNCLASSIFIED
13
 UNCLASSIFIED
DSTO-TR-2757

error could be related to CFD related assumptions, boundary conditions and turbulence
models. CFD is a better predictor of trend-lines rather than absolute values and scaling
techniques are commonly used to better approximate actual performance (while
maintaining trend-line).

The final scaled turbine map along with the unscaled map for the Olympus HP engine can
be seen in Figure 12. Note that a number of speed lines have been omitted for clarity.

Figure 12 Olympus HP turbine performance map

3.6 Ducts
Ducts are used throughout the engine and represent the connection/transitions between
components. The turbine exhaust duct is the only main duct in the Olympus HP. The
combustor is also idealised as a duct for pressure loss purposes. Note that in multi-spool
by-pass configurations the number of ducts increases significantly.

Adiabatic expansion is assumed in the duct which enforces conservation of enthalpy. The
loss due to boundary layers and turbulence is reflected as a total pressure loss. This loss
can be correlated to inlet corrected mass flow rate. The relationship is expressed as:

2
 
P  W RTt 
1  t2  
Pt1 Pt
 
 Pt 2    W RTt  
1    
 Pt1  ds   Pt  
  ds 

3.7 Nozzle
The nozzle is broken down into two sections for performance and loss modelling. The first
section is the turbine exit duct.

UNCLASSIFIED
14
 UNCLASSIFIED
DSTO-TR-2757

The performance of the second part (converging section) of the nozzle is estimated using
discharge and thrust coefficients to account for the effect of boundary layer and flow
distortion on the nozzle exit plane. Isentropic expansion to ambient pressure is assumed.
A check for sonic conditions is required. If the nozzle exit plane is found to be choked then
the nozzle exit plane Mach number is set to 1.0 and static pressure calculated.

Discharge coefficient is a measure of change in effective mass flow rate to the ideal mass
flow rate (assuming ideal expansion through the nozzle). It is defined as:

Aeff
CD 
Ageom
The main parameters affecting discharge coefficient are nozzle pressure ratio and
geometric petal angle or nozzle area ratio. The nozzle discharge coefficient for the
Olympus was assumed to be constant.

The effect of boundary layer and flow distortion on the gross thrust is accounted for using
the thrust coefficient. The gross thrust is thrust component at the nozzle exit plane. The net
thrust is the negative thrust component arising from the intake. The definition of thrust
coefficient can be seen in the equation for net thrust, that is:

FN  W8V8 C FG ,8  A8 ( Ps ,8  Pambient )  W2V0

Where 0 denotes the atmospheric free stream conditions, 2 is the engine inlet and 8 is the
nozzle exit plane.

3.8 Oil/Lubrication System

The fuel flow rate is a primary parameter in engine performance modelling. The fuel/oil
mixture flow rate can only be measured before it is split into two where one leads to the
combustor and the other to the bearings. In order to estimate the ratio of combustor fuel
flow to the lubrication fuel a test was conducted.

The test was conducted under ambient conditions where the fuel line was disassembled
from the engine and fuel pumped through the piping system. The flow coming out of the
combustor and lubrication line was then collected into a container. The ratio of lubrication
to combustor fuel flow rate was found to be approximately 0.14.

The use of fuel flow in validation (and parameters closely linked to it) was avoided as
much as possible. In instances where this was not possible, it was assumed that the
combustor/lubrication fuel/oil flow ratio was constant as measured during the fuel flow
test.

UNCLASSIFIED
15
 UNCLASSIFIED
DSTO-TR-2757

4. Engine Testing

4.1 Engine Test Objectives and Configurations

The main objective of testing the Olympus HP engine is to validate engine performance
models. The global performance parameters such as thrust, engine inlet mass flow rate,
fuel flow rate and Exhaust Gas Temperature (EGT) are the four main parameters that are
of interest where engine performance modelling is concerned.

Restricting the test to the measurement of only these four parameters has a number of
disadvantages. The main disadvantage is that there is no scope for model adjustment and
root/cause analysis when discrepancies between the engine performance model and test
data are present. The validation will only yield a positive or negative outcome. The second
disadvantage is that no insight into the behaviour of the engine components will be
gained. Hence, measuring component thermodynamic behaviour is also considered to be a
primary objective of the engine tests.

It should also be noted that the secondary objectives of the project are to investigate thrust
augmentation options and flow field characteristics for fuel flow and infrared signature
studies. This also required the test to provide as much insight into the component
behaviour as practically possible.

Three different test configurations were utilised to achieve the test objectives.

4.1.1 Engine Operating Line Test

The engine operating line test formed the backbone of the test program and established
both global and component performance parameters. The engine, under normal operating
conditions is operated from idle to maximum engine rotational speed and measurements
taken at various rotational speeds. This test yields the global performance characteristics
(thrust, mass flow rate etc) and also estimation of the compressor pressure ratio and
efficiency as a function of corrected mass flow rate and rotational speed.

4.1.2 Reduced Nozzle Area Test

The operating characteristics of a typical turbojet under normal operational conditions are
well established. The compressor operating line (established in the engine operating line
test) does not significantly change even at altitude and different flight speeds. In order to
get a better understanding of the compressor behaviour and performance map a second
test is required to “stress” the engine so that the operating line shifts/distorts which can
then be measured and mapped.

4.1.3 Nozzle Exit Flow Blockage Test

Nozzle exit flow blockage test is another test in which the engine is “stressed” to shift its
operating characteristics from the nominal. A metallic plug in the shape of a dome is

UNCLASSIFIED
16
 UNCLASSIFIED
DSTO-TR-2757

placed behind the nozzle. The distance between the nozzle and plug is reduced and the
operating line test performance completed at each plug location.

4.2 Engine Test Facility

An engine test stand was constructed to test and measure engine global and component
[8]
level performance data . Figure 13 shows the engine test stand.

Figure 13 Engine test stand

The OEM supplied engine contained a single EGT thermocouple and a tachometer to
measure engine rotation speed. The engine was subsequently instrumented with 15 static
pressure tapings, 1 total pressure probe for mass flow rate measurement, 12
thermocouples to measure gas path total temperature, a load cell to measure force, fuel
flow meter to measure fuel flow rate, secondary tachometer and a number of auxiliary
instruments.

Engine instrumentation was selected based on detailed uncertainty analysis using both the
root-mean-square and Monte-Carlo techniques. The pressure, temperature and thrust
transducers were chosen to minimise the error range.

The mass flow measurement was completed using a specially design and built bell mouth
which was subsequently calibrated. Details relating to the test and instrumentation design
[8]
can be found in the engine test report .

UNCLASSIFIED
17
 UNCLASSIFIED
DSTO-TR-2757

The location and naming convention of the instruments can be seen in Figure 14.

Station 3
Station 5
Station 1A

Station 4
Station 2

Figure 14 Engine instrumentation diagram

The test program included a reduced nozzle area test in which a conic extrusion was
added to the end of the nozzle. The location of the exit thermocouple was subsequently
shifted. The blockage test involved placing a conic/bullet shaped object behind the nozzle
exit plane and its proximity controlled using a linear actuator mechanism.

UNCLASSIFIED
18
 UNCLASSIFIED
DSTO-TR-2757

5. Results & Discussion

5.1 Engine Operating Line

Figure 15 Comparison of the net thrust test with engine performance model and OEM
data.

Figure 15 Comparison of the net thrust test with engine performance model and OEM data
The net thrust agreement between the engine performance model and test data is very
good. The match between test data and engine performance model is better than for the
data given by the OEM. The prediction quality of the model degrades slightly at rotational
speeds below 75,000 rpm. This is expected as instabilities and secondary effects are usually
prominent at lower speeds. It should also be noted that the area of interest in a turbojet is
at high rotational speeds as this is where it will be operated during its complete flight
profile (except start-up and shut-down).

A calibrated bell mouth was used to measure the mass flow rate. Figure 16 shows the mass
flow rates versus engine corrected rotational speed. The performance model and test show
very good agreement. The trend-line in both cases is identical. It should also be noted that
the mass flow rate given by the manufacturer (0.45 kg/s at 108000 rpm) underestimates
measured mass flow rate by approximately 10%.

UNCLASSIFIED
19
 UNCLASSIFIED
DSTO-TR-2757

Figure 16 Fuel and air mass flow rate versus engine rotational speed

The engine fuel flow rate comparison can be seen in Figure 17. The fuel flow rate agrees
well with the OEM data. The engine performance model shows a negative shift in fuel
flow rate. This result is expected as the fuel flow rate given by the engine performance
model refers to the fuel flow being used by the combustor. The fuel flow rate given by the
OEM and measured during testing is measured at the fuel tank exit and includes the fuel
diverted to the lubrication system. The fuel has been modelled using the standard JP-4
lower heating value of 43.323 kJ/kg. The inclusion of oil would effectively reduce the
lower heating value which would also contribute to the shift in the engine fuel flow rate.

UNCLASSIFIED
20
 UNCLASSIFIED
DSTO-TR-2757

Figure 17 Fuel flow rate comparison

The exhaust gas temperature (EGT) can be seen in Figure 18. The agreement between test,
OEM and performance model is good where both EGT trend-line and absolute values
have been captured by the model.

UNCLASSIFIED
21
 UNCLASSIFIED
DSTO-TR-2757

Figure 18 Exhaust gas temperature vs. corrected speed

The compressor total pressure ratio and total-total efficiencies are plotted on the
compressor map in Figure 19 which also shows the engine operating line. As it can be seen
from the plots the compressor model generated using the map scaling technique agrees
very well with the engine behaviour.

Figure 19 Compressor map and operating line

UNCLASSIFIED
22
 UNCLASSIFIED
DSTO-TR-2757

5.2 Reduced Nozzle Area

The objective of the reduced nozzle area test was to “stress” the engine and force it to
operate under different operating regimes. Testing under different operating conditions
would allow for better mapping of the compressor and engine performance, provide
another layer of validation to the engine performance model and aid with thrust
augmentation studies.

The change in thrust, mass flow rate, fuel flow rate and EGT with corrected speed and
reduced nozzle area can be seen in Figure 20. The reduction in nozzle exit area increases
net thrust. This is expected as the reduction in area causes the nozzle exit velocity to
increase, hence increasing the momentum change between engine inlet and outlet.

Figure 20 Change in engine performance parameters with reduced in nozzle exit area

The results show that the increase in thrust is not proportional to the change in nozzle exit
area. That is, assuming identical atmospheric conditions and corrected speed, the
thermodynamic state at the reduced nozzle exit plane is different. This is due to the

UNCLASSIFIED
23
 UNCLASSIFIED
DSTO-TR-2757

reduction of mass flow rate at the nozzle exit plane which causes an in-balance in the
engine mass-flow rate. This in turn causes an adjustment in the engine operating
conditions by shifting the compressor and engine operating line.

The vertical shift in thrust for a nozzle area reduction of 28.8% is approximately 30-50%.
Note that the expected increase in thrust would be approximately 90-100% in the ideal
case where the engine operating conditions did not shift. Further reduction in nozzle exit
area does not significantly change thrust. This may be due to a combination of factors and
the compressor reaching the stall conditions.

The agreement between test and engine performance model for the net thrust is
reasonably good. In both cases the trend-line between test and model is identical. The
model does over estimate the thrust in both instances. This can be associated to the
compressor map and the map scaling technique used to estimate its performance. The map
scaling technique starts of with a known compressor map and scales according to given
data. The prediction of the compressor map boundaries and when stall and choke occur is
impossible using this technique. For the purpose of this investigation the absolute values
of these boundaries and the exact behaviour of the engine at regions close to stall/surge is
not required, hence not pursued further.

Engine mass flow rate data also shows similar behaviour as seen for the engine thrust. The
difference in this case is that the mass flow is reduced with reduced engine nozzle exit
area. The trend-line correlation between test and engine performance model is also good,
however the model under estimates the mass flow rate. This is also associated with the
constraints on estimating the compressor map using map scaling and the accurate
prediction of regions close to stall /surge and choke.

The fuel flow rate increase is substantial with the reduction in nozzle area. The increase in
fuel flow (addition of energy) is required to compensate for the inefficient operating
regime of the engine. The engine performance model trend-line agrees with the test data.
Although it may seem that the agreement between engine performance model and test
data is very good for the 28.8% nozzle area reduction test, this is not the case as the test
points show total fuel flow (combustor + lubrication flow) where as the model shows the
combustor fuel flow only. For a good absolute fit, a similar offset as seen in the standard
operating line test would be expected.

The effects of increasing fuel flow to compensate for the in-efficiencies of the engine
operating regime are clearly seen in the EGT plot. This EGT, hence turbine inlet runs at
much higher temperatures in both the reduced nozzle area cases. The increase in EGT is
the main limiter when conducting reduced area nozzle tests as the engine could not be
operated at reasonable speeds. The increase in EGT substantially limits the use of nozzle
area reduction for thrust augmentation.

The agreement between test and engine performance model for the prediction of EGT is
very good. Not only is the trend-line captured but also the absolute fit is well defined.

UNCLASSIFIED
24
 UNCLASSIFIED
DSTO-TR-2757

The shift in the compressor operating line with reduced nozzle exit area can be seen in
Figure 21.

Figure 21 Change in engine compressor operating line with reduced in nozzle exit area

The operating line shifts towards the stall/surge limit as expected. The engine
performance model predicts this shift reasonably well for the 28.8% area reduction case.
The prediction degrades with further nozzle area reduction while maintaining good trend-
line agreement. The test show that the shift is very small between the two test cases which
may also suggest that the compressor might by stalled.

The compressor efficiency plot shows a large degree of scatter in efficiency when nozzle
exit area is reduced. This scatter is prominent at the lower corrected mass flows (also
associated with lower corrected speeds). The increase in efficiency in the lower corrected
mass flow region raises a number of questions regarding the validity of the data and how
well the instrumentation and processing technique is able to capture such efficiencies at or
near stalled regions. The efficiencies at higher speeds/flows are reduced below the
operating line case as expected. The engine performance model under predicts the change
in efficiency at the higher speeds/flows.

5.3 Nozzle Exit Flow Blockage

The objective of the nozzle exit flow blockage test is the same as the reduced nozzle exit
area test. It is another technique used to “stress” the engine and compressor at non-
standard operating conditions to measure change in overall and compressor performance.

The test involved moving a cone/bullet shaped object towards the nozzle exit plane in
increments and running an engine operating line test at each station. The cone/bullet
results in a blockage to the nozzle exit air flow. This obstruction can be viewed as another
loss source which the engine needs to overcome for a given condition.

Adding a blockage at the end of the nozzle results in complex flow features which can not
simulated in an engine performance model without mapping the behaviour of

UNCLASSIFIED
25
 UNCLASSIFIED
DSTO-TR-2757

nozzle/blockage interaction and losses. The losses and performance penalties associated
with the cone/bullet would need to be accounted for in order to obtain comparable
results. Hence, it is not possible to compare the results of the nozzle exit blockage with the
engine performance model.

The results for thrust, mass flow rate, fuel flow rate and EGT can be seen in Figure 22. The
results mirror conclusions found in the reduced nozzle exit area test. Thrust increases with
blockage due to the additional losses and energy required to overcome the blockage. The
rate at which the thrust curve shifts is directly proportional to the proximity of the
obstruction to the nozzle exit plane.

Mass flow rate does not change significantly for a great portion of the blockage stations.
The mass flow rate drops when the cone/bullet is at 0 and -0.125 exit diameters from the
nozzle exit plane. The fuel flow rate increases with the reduction in blockage as expected.
The change in EGT is also significant in these tests as it was in the reduced nozzle exit area
test.

Figure 22 Change in measured engine performance parameters with nozzle exit blockage

UNCLASSIFIED
26
 UNCLASSIFIED
DSTO-TR-2757

The change in compressor operating line and efficiency can be seen in Figure 23.

Figure 23 Change in engine compressor operating line with nozzle exit blockage

The pressure ratio plot shows the operating line shift towards the stall/surge line as
expected. This shift also seems to be proportional to blockage distance from nozzle exit
plane.

The efficiency plot shows little change in total-total efficiency at higher corrected mass
flow rates with nozzle exit blockage distance. The scatter is also consistent with reduced
nozzle area tests at lower corrected mass flow rates. There does seem to be a parabolic
trend in the efficiency plot at the lower corrected mass flow rates which raise a number of
issues. It could be argued that the lower corrected flow rate efficiencies are not scatter due
to instrumentation and data processing but represents the actual performance of the
compressor. Optimum efficiencies for the compressor would be expected at higher
corrected mass flow rates and speeds which would in turn contradict this argument. This
inconsistency in the results can only be answered with component testing which is beyond
the scope of this project.

6. Conclusion
The primary and secondary objectives of the project were successfully completed. An in-
depth analysis of engine performance modelling tools and technologies was completed. A
commercial engine performance modelling tool was initially used to simulate the
Olympus engine performance. An in-house code was developed and successfully bench
marked against the commercial code, which was subsequently used to optimise
performance parameters to improve the model output. Also, a number of tools were used
in the prediction of engine component performance. The tools and techniques that were
used included mean-line and through flow analysis, CFD, component map scaling,
geometric data scanning and post-processing to name a few.

UNCLASSIFIED
27
 UNCLASSIFIED
DSTO-TR-2757

A comprehensive test program was successfully completed which involved the design and
construction of an engine test stand. A number of components such as the engine bell-
mouth, fittings and extended nozzle were designed, manufactured and installed. A large
number of tests were completed using an array of instruments including standard
pressure and temperature probes, traverse mechanisms and multi-holed probes.

The engine performance model agreed very well with test data. Both trend-line and
absolute values matched very well at the operating line. The operating line is where the
engine will work during its flight profile. The agreement of the performance parameters in
“stressed” conditions was also encouraging. The trend-line was predicted accurately for all
parameters with a small offset in absolute values for most of the cases. This offset was
associated with the behaviour of the compressor at regions close to stall/surge and the
difficulty in predicting it using the map scaling techniques.

The project also achieved its secondary objectives. A sound understanding of the engine
performance was achieved and thrust augmentation options evaluated. Thrust
augmentation was not pursued due to the cost and complexity in manufacturing and
integrating new components into an existing engine. Infrared related CFD studies were
conducted which has spurred a separate testing program to be reported subsequently. An
alternate fuels project has been started which relies heavily on the knowledge and
[9]
infrastructure created through this project. A final year thesis by Jones has been
completed and a PhD research project into infrared technologies has been started as a
direct result of this program.

7. Recommendations
The project has not only helped gain an insight into engine performance modelling and
testing but also raised a number of issues, obstacles and deficiencies that should be
addressed in order to effectively model gas turbine engine performance.

The issues needing further investigation can be broken into two categories, namely,
numerical and experimental. The following recommendations for future work can be
made in the numerical simulation area:

 Although using a commercial tool for engine performance modelling provided

valuable data at the early stage of the project, it also raised a large number of
issues. Commercial engine performance modelling tools are mainly structured
towards engine performance design. In a large number of instances this is not
compatible with performance modelling of existing engines. The use of “generic
maps”, scaling techniques and in-built assumptions resulted in a large set of
invalid performance models. Hence, it is recommended that DSTO pursue the
development of its own in-house engine performance tool which can be
customised for modelling existing engines based on limited data sets. An in-house

UNCLASSIFIED
28
 UNCLASSIFIED
DSTO-TR-2757

code has been developed for this project, however; it would be advisable to extend
this to other engine configurations.
 The use of CFD in predicting component performance did not yield fruitful results
for the compressor. A large number of CFD models were generated for the impeller
and diffuser system without success. This is mainly associated with the complexity
of the compressor geometry, tight spacing between impeller and radial diffuser
and issues related to CFD simulation of rotating parts. The results of the turbine
CFD analyses also raised a number of questions. Hence, it is advisable to start an
in-depth investigation into turbo-machinery related CFD. More importantly, the
project should aim to generate a set of validated process in the application of CFD
to turbo-machinery. This could also form a basis for the DSTO quality management
system where turbo-machinery based CFD simulations are concerned.

Experimental improvements that would greatly enhance engine performance modelling

capability are:

 Enhancement of component testing facilities. The two main drivers of engine

performance are the turbine and compressor. DSTO does not have a facility to
characterise the aerodynamic and performance behaviour of rotating turbo-
machinery components. Initially an investigation and feasibility study into a small
compressor and turbine aerodynamic and performance should be completed. If
feasible, a small compressor and turbine test facility would provide a very good
foundation for future research into engine performance modelling and
aerodynamic related issues and challenges faced by DSTO.

UNCLASSIFIED
29
 UNCLASSIFIED
DSTO-TR-2757

8. References
[1] TurbAero Software Website, http://www.turbo-aero.com/Pages/TurbAero.aspx
[2] CompAero Software Website, http://www.turbo-aero.com/Pages/TurbAero.aspx
[3] Numeca Software Website, http://www.numeca.com/
[4] GasTurb Software Website, http://www.gasturb.de/index.html
[5] “AMT Olympus HP Specifications”, Advanced Micro Turbines, Netherlands, Sourced
from http://www.amtjets.com/OlympusHP.php, 2009.
[6] AMT Olympus HP Operating Manual.
[7] Rahman, N, “Propulsion and Flight Controls Integration for the Blended Wing-Body
Aircraft”,
Cranfield University, UK, 2009.
[8] Anderson, W., “Development of the Microturbine Test Stand and Experimental
Investigation of a Microturbine Engine”,DSTO Technical Report, 2012.
[9] Jones, B., “Optimisation of a Small Gas Turbine Engine”, Monash University, 2011.

UNCLASSIFIED
30
 Page classification: UNCLASSIFIED

DEFENCE SCIENCE AND TECHNOLOGY ORGANISATION

DOCUMENT CONTROL DATA 1. PRIVACY MARKING/CAVEAT (OF DOCUMENT)

2. TITLE 3. SECURITY CLASSIFICATION (FOR UNCLASSIFIED REPORTS

THAT ARE LIMITED RELEASE USE (L) NEXT TO DOCUMENT
An Investigation into Performance Modelling of a Small Gas CLASSIFICATION)
Turbine Engine
Document (U)
Title (U)
Abstract (U)

4. AUTHOR(S) 5. CORPORATE AUTHOR

Zafer Leylek DSTO Defence Science and Technology Organisation

506 Lorimer St
Fishermans Bend Victoria 3207 Australia

6a. DSTO NUMBER 6b. AR NUMBER 6c. TYPE OF REPORT 7. DOCUMENT DATE
DSTO-TR-2757 AR-015-427 Technical Report October 2012

8. FILE NUMBER 9. TASK NUMBER 10. TASK SPONSOR 11. NO. OF PAGES 12. NO. OF REFERENCES
2012/1121408/1 DERP CAVD 30 9

13. DSTO Publications Repository 14. RELEASE AUTHORITY

http://dspace.dsto.defence.gov.au/dspace/ Chief, Air Vehicles Division

15. SECONDARY RELEASE STATEMENT OF THIS DOCUMENT

Approved for public release

OVERSEAS ENQUIRIES OUTSIDE STATED LIMITATIONS SHOULD BE REFERRED THROUGH DOCUMENT EXCHANGE, PO BOX 1500, EDINBURGH, SA 5111
16. DELIBERATE ANNOUNCEMENT

No Limitations

17. CITATION IN OTHER DOCUMENTS Yes

18. DSTO RESEARCH LIBRARY THESAURUS

Gas turbine, performance modeling, engine thermodynamic cycle, turbojet

19. ABSTRACT
A small gas turbine performance modelling and testing project has been completed as part of a Divisional Enabling Research Program
(DERP). The main objective of the program was to enhance DSTO's capability in understanding and modelling the thermodynamic and
performance characteristics of gas turbine engines. Secondary objectives of the program included the investigation of thrust
augmentation technologies and infrared suppression modelling and analysis. This report presents the results of both numerical and
experimental investigation into engine performance simulation. It outlines the different tools and techniques used in modeling engine
component behaviour and discusses their advantages and disadvantages. The results of two different tests designed to explore engine
operating regions close to compressor stall and surge is presented.

Page classification: UNCLASSIFIED