Journal of

Mechanical
Science and
Technology
Journal of Mechanical Science and Technology 22 (2008) 538~544
www.springerlink.com/content/1738-494x

Ignition and combustion characteristics of the gas turbine slinger

combustor
Seongman Choi1,*, Donghun Lee2 and Jeongbae Park3
1
Division of Mechanical and Aerospace System Engineering, Chonbuk National University, 664-14, Deokjin Dong 1 Ga,
Jeonju City, Chonbuk, 561-756, Korea
2
Power Systems R&D Center, Samsung Techwin Co., LTD., Sungju Dong 28, Changwon City, Gyeongnam, 641-717, Korea
3
Agency for Defense Development, Yuseong, Daejeon, Korea

(Manuscript Received June 4, 2007; Revised November 9, 2007; Accepted November 12, 2007)
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Abstract

This paper describes the ignition and combustion characteristics of a gas turbine slinger combustor with rotating fuel
injection system. An ignition test was performed under various airflow, temperature and pressure conditions with fuel
nozzle rotational speed. From the test, there are two major factors influencing the ignition limits: the rotational speed of
the fuel nozzle, and the mass flow parameter. Better ignition capability could be attained through increasing the rota-
tional speed and air mass flow. From the spray visualization and drop size measurement, it was verified that there is a
strong correlation between ignition performance and drop size distribution. Also, we performed a combustion test to
determine the effects of rotational speed by measuring gas temperature and emission. The combustion efficiency was
smoothly enhanced from 99% to 99.6% with increasing rotational speed. The measured pattern factor was 15% and
profile factor was 3%.

Keywords: Slinger combustor; Rotating fuel injection system; Ignition; Combustion

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

general, these injectors require high manufacturing

1. Introduction
and design quality for their components such as a
An annular combustor with rotating fuel injection swirl chamber and orifice. The rotating fuel injection
system, what is called a slinger combustor, has been system can be applied economically to the gas turbine
frequently employed in gas turbines. Such a system is combustor. [2]
successfully adopted in a number of engines of Tur- Yet, in spite of its simplicity, the detailed ignition
bomeca and others. The main advantages of this sys- and combustion characteristics of the slinger combus-
tem are its cheapness and simplicity. In terms of at- tor were for a long time not fully understood. [3-4]
omization, a rotating fuel injection system makes a Maskey and Marsh [3] briefly explained about the
fine fuel spray even at part load or idle conditions. [1] slinger combustor characteristics in terms of engine
The role of a gas turbine fuel atomizer is to distribute operating conditions by using a J69-T-25 turbo jet
the drops into the combustion zone. The spatial distri- engine; but they did not show combustor rig test re-
bution depends on the penetration of fuel spray into sults. Rogo and Trauth [4] explained the slinger com-
the primary zone and is closely coupled to the aero- bustor in terms of engine scale design by using a
dynamic flow pattern. A pressure swirl atomizer or YJ402-CA-400 turbojet engine. By using empirical
air-blast atomizer can meet these requirements. But in correlation of droplet size, which was obtained from a
*
simplex pressure atomizing nozzle, they explained
Corresponding author. Tel.: +82 63 270 3996, Fax.: +82 63 270 2472
E-mail address: csman@chonbuk.ac.kr combustion characteristics with droplet size. These
DOI 10.1007/s12206-007-1106-6 previous results could not provide details of ignition
 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544 539

and combustion characteristics. Fig. 3 shows the thermocouples at the hollow guide
In this study, ignition and combustion tests were vane. They are installed in four hollow guide vanes
performed at various conditions to verify combustion (five for each vane) and equally spaced. The ignition
behavior. At the same time, spray characteristics of performance is measured by twenty thermocouples
the rotating fuel injection system were investigated by installed in the leading edge of the hollow guide
using laser diagnostic techniques. From these experi- vanes.
ments, we could elucidate ignition and combustion Fig. 4 shows the rotating device with measuring
characteristics of the slinger combustor with a rotat- sensors. It is installed next to the hollow guide vane
ing fuel injection system. for measuring combustor exit gas temperature profile,
total pressure and emission. It consists of an array of
2. Experimental apparatus
pressure rake, two temperature rakes and a gas sam-
Combustion tests were performed in the combustor pling probe which is cooled by water. This device
test facility of KARI. Fig. 1 shows a schematic dia- rotates 360 degree per 3 minutes. To calculate the
gram of the test rig. The rotating fuel nozzle is di- combustion efficiency, concentrations of CO, CO2,
rectly linked with a high-speed electric motor located NOx, and unburned hydrocarbon are measured from
inside of the test rig. Kerosene fuel is supplied the sampling gas.
through a fuel channel and injected into the combus- Fig. 5 shows a schematic diagram of the spray test
tor by centrifugal force of the rotating nozzle. Ignition rig. The experimental apparatus consists of a high-
is conducted with two torches.
Fig. 2 shows the flow pattern of the combustor. The
combustor is composed of an outer liner, an inner
liner, a hollow guide vane and a rotating fuel nozzle.
The detailed flow structures of air, fuel and combus-
tion gas within the combustor are traced in Fig. 2.

Fig. 1. Schematic diagram of combustion test rig.

Fig. 3. Thermocouples at the hollow guide vane.

Fig. 2. Schematic of flow pattern in combustor. Fig. 4. Picture of the rotating device.
540 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544

250 45

Air Flow rate(kg/s) * 10

40
Liner temperature
200 Fuel flow rate
35
Air flow rate

Temperature [ C]
Temperature
30

o
at combustor exit
150
Ignition 25

-2
20
100

Fuel Flow rate(cc/s)

Torch flame start 15

10
50
Fuel start
5

0 0
0 5 10 15 20 25 30 35 40
Time [sec]

Fig. 6. Typical ignition sequence.

In Fig. 6, the gas temperature is increased by 60 ଇ

after the torch is on, and then it is over 100 ଇ after
successful ignition. From the preceding descriptions,
ignition limits are determined by the flame gas tem-
perature under certain fuel and air mass flow rate
conditions.
To verify the effect of shaft speed, it is smoothly
Fig. 5. Schematic diagram of spray test rig with PDPA sys- increased from 13,000 rpm to 30,000 rpm with con-
tem. stant equivalence ratio of 0.31, airflow of 1.5 kg/s, air
temperature of 450 K and air pressure of 3.5 bar. To
speed motor, a shaft, a rotating fuel nozzle, an acrylic see the effect of equivalence ratio, it is increased from
case and a PDPA system. The high-speed electric 0.1 to 0.3 with air flow rate of 0.6 kg/s and constant
motor and shaft are directly connected with the rotat- shaft speed of 13,000 rpm. To see the general per-
ing fuel nozzle. The spray nozzle is placed at the cen- formance of the slinger combustor, a combustion test
ter of the X and Y plane and rotates counterclockwise. was conducted at shaft speed 35,000 rpm, air tem-
The dimension of the acrylic case is a 1,000 mm(L) u perature of 450 K and air pressure of 3.5 bar. The
1,000 mm(H) u 131 mm(W) square box. The PDPA equivalence ratio is about 0.26. Also, combustion
measurement system is composed of a laser source, a pressure loss was measured at various combustor inlet
transmitter, a receiver, a signal processor and a 3D air speeds with air temperature 400 K, air pressure 3.5
traverse system. A 5W Ar-Ion(Innova 70, Coherent bar and equivalence ratio of 0.26.
Co.) laser and green and blue beam(wave length
514.5 nm and 488 nm) are used for measuring two-
4. Ignition and spray test
dimensional velocity component and droplet size. The
spray visualization was performed by a digital camera Ignition limits are shown in Fig. 7. The data inter-
with high shutter speed and pulse volume beam from polation is done by cubic B-spline connection.
120 mJ Nd:YAG laser. Equivalence ratio is 0.33 at 7,300 rpm, 0.17 at 10,400
rpm and 0.15 at 13,000 rpm respectively at constant
mass flow parameter of 5.0. In this result, the ignition
3. Experimental conditions
limit is considerably increased with the shaft speed
The ignition test was performed under ambient from 7,300 rpm to 10,400 rpm, but it is slightly in-
conditions with kerosene fuel. Fig. 5 shows a typical creased with the shaft speed from 10,400 rpm to
ignition sequence of this test. The procedure is de- 13,000 rpm. The change of ignition limit is consid-
scribed as follows. ered to be mainly due to the variance of droplet size
z STEP 1 - Supply air to the combustor distribution with the shaft speed.
z STEP 2 - Rotate the fuel nozzle to desired speed The water spray test result [5] is shown in Fig. 8.
z STEP 3 - Start the torch igniter The data are connected by smooth curve fitting.
z STEP 4 - Supply fuel to the combustor SMD is about 64 Pm on the 5,000 rpm, 48 Pm on the
z STEP 5 - Check the ignition by gas temperature 10,000 rpm and 47 Pm on the 20,000 rpm at the
 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544 541

0.40

7,300 RPM
0.35
10,400 RPM
13,000 RPM
0.30
Equivalence ratio

0.25

0.20

0.15

Fig. 9. Spray visualization with rotational speed.

0.10

single large ligament. From 15,000 to 20,000 rpm, the

4 5 6 7 8 9
Mass Flow Parameter
numbers of ligaments are increased and the ligament
leads to fine drop sizes. This condition corresponds to
Fig. 7. Ignition limits.
“super-critical liquid breakup.” In supercritical
breakup, inertial effects cause the film to break into
70
many smaller ligaments. The transition between sub-
60 critical to supercritical phenomena can be seen in
50
7,500 to 10,000 rpm. So, these visualization results
suggest a clear correlation with drop size and ignition
SMD(μm)

40
performance with rotational speed.
30 Moreover, In Fig. 7 at 7,300 rpm, while increasing
20 mass flow parameter from 5.0 to 8.7, the equivalence
ratio varies from 0.39 to 0.13. Also at 13,000 rpm, the
10
equivalence ratio varies from 0.16 to 0.08. It is con-
0 sidered that a higher mass flow parameter makes
0 5000 10000 15000 20000 25000
higher air velocity in the combustor, and it affects
RPM
good mixing of air and fuel. In these results, ignition
Fig. 8. Sauter mean diameter with rpm. limits are increased with mass flow parameter in the
same shaft speed condition. Consequently, there are
60 mm radial distance from rotating nozzle tip surface. two major factors influencing the ignition limits: shaft
The results clearly explain why ignition limits rapidly speed and mass flow parameter. Better ignition capa-
change from 7,300 to 10,400 rpm and smoothly bility can be attained through increasing the shaft
change from 10,400 rpm to 13,000 rpm. speed and the mass flow parameter.
Fig. 9 shows the liquid breakup process at rota-
tional speeds of 2,500, 5,000, 7,500, 10,000, 15,000
and 20,000 rpm. Dahm pointed out that the primary 5. Combustion test
liquid breakup process would occur in liquid column Gas temperature was measured to determine the re-
mode or near the transition between the liquid column lationship between shaft speed and combustor exit
mode and the bag mode. [6] gas temperature. Fig. 10 shows combustor exit gas
From 2,500 to 5,000 rpm, the liquid film that issues temperature with shaft speed. In this test, the equiva-
from the orifice hole is drawn into a single ligament lence ratio was kept constant at about 0.32. In the
and then undergoes breakup process. This condition range of 14,000 rpm to 16,000 rpm, gas temperature
corresponds to “subcritical liquid breakup.” In sub- is steeply increased from 150 qC to 450 qC, and over
critical breakup, surface tension is sufficiently strong 16,000 rpm, it is increased smoothly. However, up to
relative to the film inertia to draw the liquid into a 30,000 rpm, the gas temperature is flat and keeps a
542 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544

constant value. This experimental result shows that region. That would be mainly due to the rotating fuel
combustion is directly affected by the shaft speed. injection system which has a homogeneous droplet
According to droplet measurement results [5], it is distribution like an infinite number of fuel nozzle
considered that droplet size is the main factor in com- spray.
bustion characteristics. About the shaft speed of Fig. 13 shows the pattern and profile factor of
15,000 rpm, there are large temperature fluctuations, equivalence ratio 0.26. These temperature profiles
and that is due to shaft vibration. were calculated by 12 thermocouple temperatures
Fig. 11 shows combustion efficiency with the shaft from Fig. 10. The profile looks like a bow shape and
speed. Test conditions were as follows: air flow rate the maximum value position is about 40% turbine
of 1.5 kg/s, air pressure of 3.5 bar and air temperature height. The profile factor is about 3% and this value is
of 450 K. The combustion efficiency was calculated extremely good compared with other annular
by EI(Emission Index) method using sampling gas. combustors using a conventional fuel injector. It
The combustion efficiency is increased from 99.1% could be explained that fuel is sprayed by the rotating
to 99.62% while the shaft speed is increased from fuel injection system uniformly and is mixed with air,
15,000 rpm to 30,000 rpm. Over 30,000 rpm, the and finally it makes very homogeneous fuel-air
combustion efficiency is not changed and kept con- mixture at the entire flame tube region. The pattern
stant. From Fig. 7 and Fig. 8, combustion efficiency factor is about 15% and maximum value was found at
could be improved by increasing the shaft speed 75% turbine height. This pattern is a very good value
without changing air and fuel flow rate. compared with an annular combustor with air blast
Fig. 12 shows temperature profiles at the combus- spray nozzle [7], and the maximum position is very
tor exit with rotating device angle. The temperature desirable for turbine life.
profiles of 12 thermocouples on the rotating device Fig. 14 shows total pressure loss with critical Mach
have a similar pattern with the time. It means that number for combustion and without combustion con-
combustion took place uniformly at every flame tube ditions. The two pressure-loss curves increase with
critical Mach number. At the critical Mach number
700
1100
600

1000
500
Temperature[ C]

R/D Temperature[ C]

Trd_1_1
0

Trd_1_2
400 900 Trd_1_3
Trd_1_4
Trd_1_5
Trd_1_6
300 Trd_2_1
800 Trd_2_2
Trd_2_3
200 T_vane 1 Trd_2_5
T_vane 2 Trd_2_6
T_vane 3 700
100 T_vane 4
T_vane 5
600
0
0 60 120 180 240 300 360
14000 16000 18000 20000 22000 24000 26000 28000 30000 o
R/D Angle [ ]
Shaft Speed[RPM]
Fig. 12. Combustor exit gas temperature with rotating device
Fig. 10. Combustor exit gas temperature with shaft speed.
angle.
100.0
100
99.8 Mean Max Trd_6
Min
Combustion Efficiency[%]

99.6
Trd_5

99.4
Vane Height [%]

Trd_4
99.2
50
99.0 Trd_3

98.8 Trd_2
Sensor 1
98.6 Sensor 2
Trd_1
98.4 0
15000 20000 25000 30000 35000 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30
Shaft Speed [RPM] Temperature Raito

Fig. 11. Combustion efficiency with shaft speed. Fig. 13. Pattern and profile factor with vane height.
 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544 543

4.5
dynamic pressure loss is 84% and combustion pres-
4.0 sure loss is 16% of the total pressure loss.
3.5
Pressure Loss[%] Nomenclature-----------------------------------------------------------
3.0

P : Inlet pressure [bar]

2.5
T : Air temperature [K]
2.0 rpm : Revolutions per minute
1.5
With Combustion Tvane : Combustion gas temperature at hollow guide
Without Combustion
vane
1.0
0.10 0.11 0.12 0.13 0.14 0.15 0.16 W : Air flow rate [kg/s]
Critical Mach Number at Combustor Inlet
PDPA : Phase Doppler particle analyzer
Fig. 14. Total pressure loss with critical Mach number. SMD : Sauter mean diameter [μm]
Mass Flow Parameter : W T / P
0.16, pressure loss for the combustion condition is F : Fuel flow rate [kg/s]
about 4.3% and cold pressure loss is about 3.6%. The A : Air flow rate [kg/s]
relative difference of these two cases is about 16%. (F / A)actual
Equivalence ratio : G =
So this combustor aerodynamic pressure loss is about (F / A)stoichiometric
84% and combustion pressure loss is about 16%. Tlocal,max : Combustor outlet local maximum tempera-
ture [K]
6. Conclusions Tinlet : Combustor inlet temperature [K]
The objective of this study was to understand the Toutlet : Combustor outlet temperature [K]
ignition and combustion characteristics of an annular Tradial,max : Combustor radial average maximum tem-
combustor with a rotating fuel injection system. perature [K]
In the ignition test, ignition limit is largely in- Poutlet : Combustor outlet pressure
creased with the shaft speed from 7,300 rpm to Pinlet : Combustor inlet pressure
10,400 rpm, but it is only slightly increased with the Tmeasured : Measured combustor outlet temperature
T  Tinlet
shaft speed from 10,400 rpm to 13,000 rpm. Pattern Factor : 100 × local,max
From the spray visualization, the transition between Toutlet  Tinlet
sub-critical to super-critical breakup process could be Tradial,max  Tinlet
Profile Factor : 100 ×
seen in 7,500 to 10,000 rpm. Also SMD is about Toutlet  Tinlet
64 Pm on the 5,000 rpm, 48 Pm on the 10,000 rpm Poutlet  Pinlet
and 47 Pm on the 20,000 rpm. So we find that fuel Pressure Loss : 100 ×
Pinlet
drop size is inversely proportional to the shaft speed
T  Tinlet
below 10,000 rpm. However, there is almost no corre- Temperature ratio : measured
lation between fuel drop size and shaft speed above Toutlet  Tinlet
10,000 rpm, but the number of ligaments is increased
References
and it leads to a change in the regional distribution of
the fuel. [1] NREC, The design and development of gas turbine
In the combustion test, the measured value for combustors, Northern Research and Engineering
combustion efficiency is 99.6%, the pattern factor is Corporation, USA, (1980) 6. 32-6. 34.
15%, and the profile factor is 3%. In the results, com- [2] A. H. Lefebvre, Atomization and Spray, Hemi-
bustion efficiency can be improved by increasing sphere Publishing Corporation, USA, (1989) 134-142.
shaft speed without changing the air and fuel flow [3] H. C. Maskey, F. X. Marsh, The annular combus-
rate. In the accelerating test, flame is unstable near tion chamber with centrifugal fuel injection, SAE
15,000 rpm, but the flame is very stable over 20,000 444C, USA, (1962)
rpm. The combustor outlet temperature profile is ex- [4] C. Rogo, R. L. Trauth, Design of high heat release
tremely good and the maximum value position is slinger combustor with rapid acceleration require-
about 75% of the turbine height. The total pressure ment, SAE 740167, USA, (1974).
loss is 4.3% at critical Mach number 0.16. The aero- [5] S. Choi, K. Lee and D. Lee, J. Park, Spray charac-
544 S. Choi et al. / Journal of Mechanical Science and Technology 22 (2008) 538~544

teristics of the rotating fuel injection system by us- 32nd AIAA Fluid Dynamics Conference and Ex-
ing PDPA and PIV, 17th International Symposium hibit, USA, (2002) AIAA 2002-3183.
on Air Breathing Engine, Germany, (2005) ISABE [7] R. K. Mishra, P. S. Ramanujam, C. Badarinath and
2005-1217. M. N. Bhat, Influence of operating pressure on the
[6] W. J. A. Dahm, P. R. Patel and B. H. Lerg, Visuali- performance of an aero gas turbine combustor, 17th
zation and fundamental analysis of liquid atomiza- International Symposium on Air Breathing Engine,
tion by fuel slingers in small gas turbine engines, Germany, (2005) ISABE2005-1020.