Measurement of the periodic flow of an enclosed
lean premixed prevaporized stagnation flame
by
O. Schäfer, R. Koch, S. Wittig
Universität Karlsruhe
Institut für Thermische Strömungsmaschinen
Kaiserstr. 12, D-76128 Karlsruhe (Germany)
Abstract
Phase resolved velocity, pressure and OH-emission measurements and laser light sheet visualization have been
applied in the analysis of an unsteady reacting flow. The measurements reported herein are for an enclosed,
planar, lean premixed prevaporized kerosene flame, stabilized by means of a movable stagnation plate. The
experimental setup displayed in figure 1 comprises a co-annular jet arrangement with a shielding air flow around
a central fuel loaded jet, both impinging on the stagnation plate. The turbulent flames investigated show an
abrupt change in their mode of oscillation depending on the position of the plate with respect to the outlet of the
prevaporizing, premixing section. Two distinct modes can be identified. The lower branch mode exhibits
combustion oscillations in the frequency range of 110-125 Hz with higher harmonics. In the higher frequency
mode the oscillations show a single tone at a frequency of 175-195 Hz, depending on the flame position. The two
branches are separated by a small zone with no oscillations. This zone is only a few millimeters thick in terms of
the position of the stagnation plate, or in other words the position of the planar flame front.
photodiode
field of view
quartz glas tube
stagnation plate
y
premixing tube
secondary air
exhaust
primary air
z
fuel (kerosine)
x
movable
pressure probe
atomizing air
premixing duct
Fig. 1:
combustor
water cooled liner
Sketch of the experimental setup
The two modes appreciably differ in the phase relation between the pressure inside the combustor and the heat
release measured by the spontaneous emission of OH-radicals. Phase averaged visualization of the flow shows
large coherent structures in the shear layer of the co-annular flow in the lower frequency case. The periodic
vortex shedding results in appreciable spatial and temporal fluctuations in the reacting flow field which generate
large periodic fluctuations in the heat release. For the higher frequency branch the visualization showed rather
small spatial fluctuations. To further elucidate the difference between the two modes of oscillation phase
resolved LDV measurements of the axial velocity in the turbulent flow have been conducted. The measurements
show large non-linear fluctuations in the shielding air flow. These fluctuations induce velocity oscillations in the
primary air with a phase lag of about 180°. Due to the phase lag between the flows excitation of the separating
shear layer is evident, even though the time mean bulk velocities are equal. In the high frequency mode the
oscillations in both flows are in-phase. As the two flows have the same bulk velocities at the inlet and to the
combustor the effect on the shear layer is small. In this mode the combustion oscillations can rather be explained
in terms of the sound particle velocity modulating the flow at the combustor inlet.
1
1. Introduction
A major challenge to gas turbine engineering today is the substantial reduction of NOx emissions without
unacceptable values for CO (Stolarski et al, 1995, Tacina, 1990). One promising concept for liquid fueled gas
turbines is lean prevaporized premixed (lpp) combustion. Unfortunately, the high NOx reduction potential of
lpp-combustion is offset by the operational disadvantages, i.e. combustion instabilities, susceptibility to flash
back and autoignition. The experiment presented here is mainly designed for the investigation of the flash back
phenomenon of lpp-flames, that can be stabilized in different ways. During the first phase of the investigation on
the upstream flame propagation of a planar, turbulent, non-swirling flame combustion oscillations have been
encountered.
Combustion oscillations have been studied of many investigators covering a wide range of applications. In
general information on the coupling of mechanisms between pressure, velocity and heat release fluctuations is
required in order to understand and predict the onset of combustion oscillations. This will be done herein for the
baseline case of a planar turbulent flame. The stagnation flow configuration to be presented depicts an
established method for the study of turbulent premixed flames, but to our knowledge it is the first time that this
configuration has been enclosed, with the consequences on combustion stability discussed in the following. The
combustor exhibits two distinct instability modes which are analyzed by means of cross-correlated pressure and
OH-emission measurements to acquire the exact phase relation between the pressure and the heat release rate
within the ± ¼ π band of the Rayleigh criterion. To further elucidate the interaction of the flow field and the
reaction zone, phase-resolved laser light sheet visualization and LDV measurements are conducted. The
measurements clearly show the importance of secondary effects, like the vortex shedding of secondary air, on
the coupling of pressure and heat release eventually leading to combustion instabilities. In the analysis of
combustion instabilities with delay time models the propagation speed of the flow modulation has to be
measured by means of phase-resolved detection of the velocity field (Zangl et al., 1996).
2. Experiment and Measurement
Experimental Setup
The experimental setup has been designed to enable different modes of stabilization of lean premixed flames,
e.g. stagnation flow, dump combustor and swirl flow. Only the premixed stagnation flame will be discussed
herein. A sketch of the experimental setup (figure 1) shows the coaxial arrangement of the primary fuel-air
mixture and the secondary air, both impinging on a movable, air cooled stagnation plate and stabilizing a planar
turbulent flame in the vicinity of the plate. The inner and outer diameter of the primary and secondary flow at the
inlet of the combustor are 41 and 58 mm respectively giving an area ratio of one. The annular flow, a typical
feature of stagnation flame investigations, inhibits the entrainment of hot combustion products that recirculate in
the corners of the combustor. The liquid fuel is injected via an internally mixing twin-fluid atomizer. This
atomizer produces very small droplets (Sauter Mean Diameter < 15 µm) over its whole operating range, which,
in combination with a long premixing duct (400 mm), guarantees a homogenous mixture at the inlet of the
combustor. The experiments are performed at atmospheric pressure. Primary and secondary air mass flows can
be controlled and heated separately. The mass flows are determined by means of calibrated hot film sensors
before the flows are electrically heated up to maximum temperature of 973 K. For the investigations presented
here the inlet temperature of both flows is 573 K, the equivalence ratio of the primary fuel-air mixture is 0.85.
The temperatures are measured just before the inlet of the combustor accounting for the temperature deficit of
the primary flow due to evaporation of the kerosene droplets. Because of the additional injection of fuel and
atomizing air the primary mass flow is about 9% higher than annular mass flow. With those parameters the
planar flame is stabilized closely to the plate providing a well defined length for the calculation of the time delay
between a perturbation at the inlet and the fluctuations in the heat release rate close to the plate.
Measurement Techniques
The LDV-system consists of an argon-ion-laser, operated at the power of 1.5 W, and a four beam two color
Dantec Fiber Flow system in backscatter mode. Only the 514.5 nm wavelength is used in this study to measure
the axial velocity u in the radial x-y plane (see Fig. 1). The measurement volume is 0.219 x 4.611 mm in
diameter and length, respectively. A burst spectrum analyzer (BSA 57N20) is used to process the data. CaCO3
particles (d < 1 µm) are equally added to both flows for seeding. To analyze the propagating modulation of the
flow, the time resolved data at the different locations in the combustor are related to the well defined filtered
analog signal of a pressure probe (Lang, Vortmeyer, 1987, Reuter et al., 1989). The pressure probe consists of a
condenser microphone (Bruel & Kjaer Type 4134) in connection with a calibrated, water cooled, helium purged
tube assembled to the plenum of the combustor at x = 0 mm (see figure 1). The frequency response of the probe
in terms of amplitude and phase is depicted in figure 2. Both modes of oscillations exhibit sound pressure levels
of about 135 dB at the inlet plenum inside the combustor.
2
180
90
1,0
0
-90
0,0
phase shift [ º ]
amplitude ratio [ - ]
2,0
-180
0
1.000
2.000
3.000
4.000
frequency [Hz]
Fig. 2:
Frequency response of the helium purged, water cooled pressure probe
The zero-crossing on the positive rising slope of the filtered, sinusoidal microphone signal is used to produce a
trigger signal for the LDV system for repetitive sampling. Phase lags induced by the microphone (180°), the
probe and the filters are taken into account in the analysis of the interaction of velocity field, pressure and heat
release rate. In the analysis of the LDV data two types of analysis, namely in the time domain and in the
frequency domain, have been applied. The first method is depicted in figure 3. By repetitive sampling of the data
3
axial velocity [m/s]
2
1
0
-1
-2
-3
0
1
2
3
4
5
time [ms]
Fig. 3:
10000 burst events per period, achieved by repetitive sampling and phase averaged velocities
at x = y =0 in the high frequency mode (f = 182Hz, T = 5.49ms)
with respect to the trigger and the time of one period, all velocity data can be plotted within one period of
oscillation. Dividing the period in phase bins of 10° yields the phase averaged velocity values used in the
following analysis. The second method applied to the velocity data was based on the non equidistant Fourier
analysis proposed by Zangl (Zangl et al., 1996). Again repetitive sampling of the data was performed yielding
sample rates for the Fourier analysis of more than ten times the frequency of oscillation avoiding aliasing in the
frequency analysis. In contrast to the analysis of Zangl (Zangl et al., 1996) resampling by a simple “sample and
hold” function was applied to the non equidistant data. The results of this analysis showed large scatter and a
high sensitivity of the phase on the sampling parameters. The frequencies were predicted correctly for all
measurement points. At this point in time it is obscure if the low modulation of the periodic part of the velocity
compared to the turbulent fluctuations might be a factor influencing the results. Further work in the analysis in
the frequency domain is ongoing.
Visualization of the flow field is achieved by use of a copper-vapor laser (flaser=10kHz, tpulse=20ns) producing a
light sheet in the x-y plane of 1 mm thickness and 40 mm in width. Only the primary air flow is seed with
particles for visualization, yielding a sharp separation between the primary and secondary flow inside the
combustor. The scattered light is detected perpendicular to the plane of illumination in two ways. For the phaseresolved visualization a B/W camera (LaVision) is triggered at 0, 90, 180 and 270 degree with respect to the
zero-crossing of the pressure signal and an exposure time of 99 µs. For each phase 300 images are acquired and
processed. In the second case a 3CCD-camera is used with exposure times of 1 and 0.1 ms yielding the
streaklines in the flow field.
3
The OH-emission as a measure of the heat release is measured with a photodiode (Burr-Brown, OPT210) in
conjunction with a spectral band pass filter (305,4 nm, BW = 22,3 nm) and an amplifier (Stanford Research
Systems SR650). A full detection angle of 30° enables the measurement of the global OH-emission of the planar
flame.
3. Results
Pressure and Heat Release
Regardless of the thermoacoustic mechanisms coupling the energy release rate to the flow field the Rayleigh
criterion must be satisfied to drive the combustion oscillations (Fernandes et al., 1996, Culick 1996). Following
the work of Lang (Lang, 1986) the cross power spectrum of the conjugate complex Fourier transform of the heat
release rate (OH-emission) and the Fourier transform of the pressure signal is used to analyze the oscillations of
the stagnation flame. In figure 4 the results for frequency and phase difference are presented as a function of the
plate position.
180
200
4,0
120
180
60
30
Rayleigh criterion
160
0
-30
140
-60
-90
-150
120
frequency
measurement
probe corrected
100
60
70
x=60mm
0,2
0,0
5,0
x=53mm
0,0
0
80
plate position [mm]
Fig. 4:
0,4
2,5
-180
50
pressure signal [Volt]
0,0
90
-120
x=73mm
2,0
frequency [Hz]
phase difference ∆φP-OH [ º ]
150
200
400
frequency [Hz]
left: phase difference between pressure and heat release, frequency
right: spectra of the pressure signal for the three particular plate positions
The mode of oscillation clearly “jumps” from a lower frequency branch to a higher frequency branch at a plate
location of x = 60 mm, at which the oscillations almost vanish. The diagrams on the right in figure 4 exemplify
the amplitude spectra of the pressure for three characteristic plate positions. In the following only the two cases
Fig. 5:
Laser light sheet visualization of primary air flow for x = 73 mm
left: arbitrary phase on low frequency branch, exposure time 1/1000 s
right: ensemble average of 300 images; phase-locked 90° prior maximum pressure
4
of x = 53 and 73 mm, with frequencies of 182 and 117 Hz respectively, will be analyzed as examples for the two
distinct modes of oscillation of the enclosed stagnation flame. For both branches the Rayleigh criterion is
satisfied, but the phase relationship between pressure and heat release clearly changes within a few millimeters
of plate distance.
Visualization of the flow
The image of the flow on the left in figure 5 is just an arbitrary snapshot of the coherent structures in the low
frequency mode that dramatically alter the primary fuel loaded flow. The right image in figure 5 is the result of
ensemble averaging over 300 single phase locked exposures. The results of the visualization at different phase
lags with respect to the pressure in the combustor are depicted in figures 6 and 7.
40
40
φ
radial positi on [mm]
radial positi on [mm]
φ
30
20
10
0
0
10
20
30
40
30
20
10
0
50
0
10
axial position [mm]
40
radial position [mm]
radial position [mm]
40
50
φ
30
20
10
0
0
10
20
30
40
30
20
10
0
50
0
10
axial position [mm]
20
30
40
50
axial position [mm]
Phase resolved laser light sheet visualization of primary air flow for the high frequency mode
40
40
φ
radial position [mm]
radial position [mm]
φ
30
20
10
0
0
10
20
30
40
30
20
10
0
50
0
10
axial position [mm]
40
50
φ
30
radial position [mm]
radial position [mm]
30
40
φ
20
10
0
0
20
axial position [mm]
40
10
20
30
40
50
30
20
10
0
0
axial position [mm]
Fig. 7:
30
40
φ
Fig. 6:
20
axial position [mm]
10
20
30
40
50
axial position [mm]
Phase resolved laser light sheet visualization of primary air flow for the low frequency mode
The figures show large differences in the phase resolved behavior of the flows. In the high frequency mode the
primary flow exhibits a bulging motion. In the lower frequency mode vortex shedding in the shear layer and
large coherent structures are evident.
5
Time Averaged Velocities
0 ,8
1
0
10
20
30
40
25
20
1,3
15
1 ,3
1
10
0,8
1
1
1,3
radial distance y [mm]
To further elucidate this abrupt change, LDV measurements of the axial velocity component in the axisymmetric
flow are carried out (see figure 8). The diagrams for the mean velocity show the typical deceleration of the flow
and the merging of primary and secondary flow. There is no indication for the combustion instabilities
encountered at those plate positions. The distribution of the fluctuating component u rms one the other hand shows
a clear difference between the two cases at x = 35 mm. In the lower frequency case there is a spot of higher
1
5
0
50
0
10
20
30
40
50
60
70
axial distance x [mm]
axial distance x [mm]
3
4
2,5
1,5
3,5
5
0
50
0
10
20
30
40
50
60
70
axial distance x [mm]
axial distance x [mm]
mean velocity u : 1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6 6,5 7 7,5
Fig. 8:
2
5
4 ,5
3
5
2,5
3,5
6
6,5
10
6
5
4,5
4
5 ,5
7
40
15
7
30
6
20
5,5
20
6 ,5
10
25
5,5
0
6
6
radial distance y [mm]
u rms : 0,5 0,8 1 1,3 1,5 1,8 2 2,3 2,5 2,8 3 3,3 3,5 m/s
m/s
LDV measurements (not phase-resolved) for upper (left) and lower (right) branch
top: axial velocity fluctuations (root mean square)
bottom: mean axial velocity
turbulence, which can be explained by interaction of coherent vortices in the primary and secondary flow, as can
be seen in figure 5 – 7. In addition extremely high fluctuations in the secondary flow can be found at the inlet to
the combustor. Those findings suggest the separation of the actual velocity in the triplet form
u = u + u% + u′
with a time mean, a periodic and a turbulent part of the velocity. For the time averaged velocity in the standard
form in figure 8 the periodic part attributes a large portion to the “turbulent” fluctuations. In a flow with large
coherent structures the triplet form should be preferred. The periodic part of the velocity field will be further
analyzed in the next section.
Phase Resolved Velocities
Further investigations by means of phase-resolved LDV measurements display the differences in the flow field
between the two modes of oscillation (figure 9,10). The velocity distributions show the same structures found by
laser light sheet visualization. Again only a slight bulging motion in the velocity field is found for the high
frequency mode. In the low frequency mode large velocity fluctuations found in the secondary flow are
convected downstream. Those large scale structures dramatically change the velocity distribution in the primary
flow. Those induced fluctuations are also convected downstream leading to a temporal change in the heat
release.
Examination of the periodic part of the axial velocity, for example at the inlet (x = 3 mm) of the combustor (see
figure 11), clearly shows the interaction of the pressure and the velocity field inside the combustor. For the
higher frequency mode the oscillations in the axial velocity show a maximum deceleration briefly after the point
of maximum pressure (90° phase angle). Primary and secondary flow oscillate with almost the same phase with
the primary flow a few degrees ahead of the secondary flow. For the low frequency case on the other hand there
6
φP= 90° (maximum pressure)
0
10
3 ,5
4
50
0
5,5
5
4,5
4
3,5
20
3
10
7
radial distance y [mm]
3
3,5
4
4,5
5,5
20
6
40
30
6
radial distance y [mm]
6
30
φP= 270° (minimum pressure)
2,5
2
1,5
0
40
0
10
axial distance x [mm]
20
30
40
1
50
axial distance x [mm]
Phase resolved velocity measurement in the higher frequency mode (182 Hz)
φP= 90° (maximum pressure)
6
5,5 5
4,5
4
3,5
3
5,
5
7 ,5
5
4,5
7
6,5
8
7
6
7
0
5
0
20
40
60
0
10
axial distance x [mm]
φP= 180° (pressure node)
20
30
40
50
60
4
axial distance x [mm]
3
φP= 270° (minimum pressure)
2
30
1
5,5
30
40
4
4,5
3,5
7
6,5
6,
5
6
10
6
6
5, 5
5
20
7 ,5
radial distance y [mm]
6 ,5
6
4
10
3,5
u [m/s]
20
7
radial distance y [mm]
fP= 0° (pressure node)
30
6,5
Fig. 9:
20
axial distance x [mm]
30
20
3
5 ,5
0
φP= 180° (pressure node)
0
4,5
6
50
axial distance x [mm]
10
6,5
3
40
7
0
4
30
10
3,5
20
7,5
4,5
10
u [m/s]
20
6,5
0
30
5
radial distance y [mm]
3
3 ,5
4 ,5
4
5,5
10
5
6
20
6 ,5
radial distance y [mm]
fP= 0° (pressure node)
30
0
0
10
20
30
40
50
60
0
axial distance x [mm]
10
20
50
60
x
Fig. 10: Phase resolved velocity measurement in the lower frequency mode (117 Hz)
is a phase difference between the secondary and primary flow, with the primary flow lagging about 180° behind.
The measurement also shows that the amplitude of the velocity oscillations in the secondary flow is about four
times higher than in the primary flow and about 50% of the mean velocity of the secondary flow. The
measurement show that the non-linear behavior has to be taken into account in the correct prediction.
From the measurements it is quite obvious that the oscillation in the secondary flow induce the oscillation in the
fuel loaded primary flow. The LDV measurements can be further investigated to examine the speed of
propagation of those disturbances at the inlet to the combustor (see figures 12 and 13). This speed a necessary
parameter for the empirical delay time models. Such a model is discussed by Hermann et al.(1995). An approach
to determine the propagation speed is to express the phase lag between two axial locations in terms of a time
difference and to calculate the speed as the ratio of the spatial to the temporal difference. The figures show the
periodic part of the velocity as function of the phase angle and the axial position for two lines in the flow at
y =0mm (centerline) and. y = 25mm (middle of the annulus in the secondary flow).
7
0,6
10
0,3
0,6
0
0, 9
-0,6
-0 ,
6
-0,3
5
0,3
0, 25
3
-0 ,
0
15
0
0
0,25
5
20
-0,3
-2,
5
0
-0,3
-0,5
10
0
25
3,5
20
15
-0, 3
3,5
-2,5
25
radial distance y [mm]
1,5
0
-0,5
radial distance [mm]
u [m/s] -4,5 -2,5
0
100
200
300
100
u [m/s]
-1
-0,75
-0,5
-0,3
0
0,25
200
300
pressure phase angle [ º ]
pressure phase angle [ º ]
u [m/s] -0,9
0,5
-0,6
-0,3
0
0,3
0,6
0,9
1,2
1,5
Fig. 11: Fluctuating part of the axial velocity as function of pressure phase angle at x = 3 mm
left: low frequency branch (117 Hz, plate at x = 73 mm, divided for better visibility)
right: high frequency branch (182 Hz, plate at x = 53 mm)
5
0 ,1
0, 2
0,
3
0 ,4
0 ,1
0 ,1
0,
1
0 ,1
-0
,2
0
-0 ,
0, 1
,5
0
50
0
0, 3
-0
0
-0
100
-0 , 1
,3
,3
pressure phase angle [ º ]
0 ,2
150
1
-0
0
50
-0 , 1
0
- 0, 2
100
0
- 0,1
,2
0,
0
-0
- 0 ,2
-0 , 1
,4
- 0 ,6
-0
-0
,3
1
150
0
0,
3
0
200
- 0,
0 ,1
0,2
0
,1
-0
0
0
200
250
y = 25 mm
0 ,4
0,1
0 ,1
0
300
0 ,1
250
0
0 ,1
300
-0
,3
-0
,4
ressure phase angle [ º ]
350
y = 0 mm (centerline)
0
350
0
0
10
20
30
40
10
axial distance x [mm]
- 0,6
u [m/s]
-0,5
-0,4
20
30
40
axial distance x [mm]
-0,3
- 0,2
-0,1
0
0,1
0,2
0,3
0,4
0,5
0,6
Fig. 12: Fluctuating part of the axial velocity as function of pressure phase angle for high frequency
branch (182 Hz, plate at x = 53 mm) for two radial positions in the flow
left: y = 0mm (on the centerline); right: y = 25mm (in the middle of the annular flow)
50
0,
2
0,
0,
6
- 0 ,4
- 0 ,2
2
0
0
-3
pressure phase an gle [ º ]
-4
0
0 ,4
p ressure phase angle [ º ]
-2
1 00
0
10
20
30
40
50
60
10
20
axial distance x [mm]
u [m/s]
0
1
0
0 ,6
-0 , 6
1 50
0
,4
0
0
-2
-0 , 4
-0
-1
2
1
-1
0
0 ,2
0
3
2 00
-2
0, 4
0
-0 ,2
-1
50
0, 2
,2
-0
0 ,2
1
0, 2
0, 4
0, 4
6
0,
0
2 50
-3
0, 4
100
8
2
0 ,6
0,
4
150
0,
y = 25 mm
0
200
3 00
-1
250
3 50
4
y = 0 mm (centerline)
300
3
350
-1
-0, 8
-0,6
-0,4
-0,2
0
0,2
30
40
50
60
axial distance x [mm]
0,4
0, 6
0,8
-5
-4
-3
-2
-1
0
1
2
3
4
5
Fig. 13: Fluctuating part of the axial velocity as function of pressure phase angle for low frequency
branch (117 Hz, plate at x = 73 mm) for two radial positions in the flow
left: y = 0mm (on the centerline); right: y = 25mm (in the middle of the annular flow)
8
In the case of low frequency oscillations the speed of propagation can easily be calculated by following the
maxima or minima of velocity. In the high frequency mode such a propagation of disturbances can be found in
the secondary flow. On the axis of the primary flow this propagation speed becomes large compared to the
velocities in the flow as the phase difference approaches zero. This might be due to the fact that the disturbances
are pure acoustic velocity oscillations (sound particle velocity) propagating with the speed of sound. As pointed
out in chapter 2 the analysis in the frequency domain to obtain the phase of the velocity fluctuations at different
spatial positions is part of the ongoing work.
4. Summary
The results presented demonstrate that the combination of pressure and OH-emission measurements, phaseresolved laser light sheet visualization and phase-resolved LDV measurements provide excellent tools for the
investigation of the coupling of pressure, heat release rate and the velocity field in the case of combustion
oscillations. With the phase-resolved LDV measurements the periodic interaction between the coaxial flows,
typical of premixed stagnation flame investigations, can be examined and the propagating speed of disturbances,
crucial in the analysis of combustion oscillations can be deduced.
From the physical point of view the analysis of the combustion oscillations of an enclosed stagnation flame
configuration shows the importance of secondary effects, like the interaction of the shielding air flow with the
fuel loaded main flow, which are not encountered in an openly burning stagnation flame. Hence a thoughtful
design is necessary for the investigation of planar, turbulent, premixed flames at pressures above ambient, taking
into account the acoustic boundary conditions and the interaction of multiple flows.
References
Culick, F. E. C. (1996):
Combustion Instabilities in Propulsion Systems, Unsteady Combustion, NATO ASI Series, Series E: Applied
Sciences – Vol. 306, pp. 173 - 242
Fernandes, E. C., Heitor, M. V. (1996):
Unsteady Flames and the Rayleigh Criterion, Unsteady Combustion, NATO ASI Series, Series E: Applied
Sciences – Vol. 306, pp. 1 - 16
Hermann, J., Zangl, P. Gleis, S., Vortmeyer, D.. (1995):
Untersuchung der Anregungsmechanismen selbsterregter Verbrennungsschwingungen an einem
Verbrennungssystem für Flüssigkraftstoff, 17. Deutscher Flammentag, Hamburg, 12. und 13. September, VDIBericht Nr. 1193
Lang, W. (1986):
Dynamik und Stabilität selbsterregter Verbrennungsschwingungen beim Auftreten mehrerer Frequenze. Ein
erweitertes Stabilitätskriterium, Dissertation, TU Munich
Lang, W., Vortmeyer, D. (1987):
Messung der Schallschnelle in einer schwingenden Flamme mit einem Laser-Doppler-Anemometer, 13.
Deutscher Flammentag, VDI-Bericht Nr. 645
Reuter, D. M., Hedge, U. G., Zinn, B. T. (1989):
Flowfield Modifications of Cobustion Rates in unstable Ramjets,
27th Aerospace Sciences Meeting, Reno, AIAA 89-0105
Stolarski, R. S., Baughcum, S. L., Brune, W. H., Douglass, A. R., Fahey, D. W., Friedl, R. R., Liu, S. C., Plumb,
R. A., Poole, L. R., Wesoky, H. L. (1995):
The 1995 Scientific Assessment of the Atmospheric Effects of Stratosperic Aircraft, NASA RP 1381
Tacina, R. R. (1990):
Low NOx Potential of Gas Turbine Engines, Presented at the Twenty-Eighth Aerospace Sciences Meeting, Reno,
Nevada, January, AIAA-90-0550
Zangl, P., Hermann, J., Orthmann, A., Vortmeyer, D. (1996):
Measurement of the periodic flow instability during a self exited combustion oscillation by LDV,
Eighth Int. Symposium on Application of Laser Techniques to Fluid Mechanics, Vol. 2, paper 39.2
9