13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
CH Double-Pulsed PLIF Measurement in Turbulent Premixed Flame
Mamoru Tanahashi, Shohei Taka, Toshio Miyauchi
Dept. of Mechanical and Aerospace Engineering, Tokyo Institute of Technology, Japan, mtanahas@mes.titech.ac.jp
Abstract The flame displacement speeds in turbulent premixed flames have been measured directly by the
CH double-pulsed planar laser-induced fluorescence (PLIF). The CH double-pulsed PLIF system consists
of two independent conventional CH PLIF measurement systems, and laser beams from each laser system
are lead to same optical pass using the difference of polarization. The highly time-resolved measurements
are conducted in relatively high Reynolds number turbulent premixed flames on a swirl-stabilized combustor.
Since the time interval of the successive CH PLIF can be selected to any optimum value for the purpose
intended, both of the large scale dynamics and local displacement of the flame front can be discussed. By
selecting an appropriate time interval (100µs~200µs), deformations of the flame front are captured clearly.
Successive CH fluorescence images reveal the burning/generating process of the unburned mixtures or the
handgrip structures in the burnt gas, which have been predicted by three-dimensional direct numerical
simulations of turbulent premixed flames. To evaluate the local flame displacement speed directly from the
successive CH images, a flame front identification and displacement vector evaluation schemes are
developed. Direct measurements of the flame displacement speed are conducted by selecting a minute time
interval (≈30µs) for different Reynolds number (Reλ = 63.1~115.0). Local flame displacement speeds
coincide well for different Reynolds number cases, which suggests the correctness of the flamelet concept.
Furthermore, comparisons of the mean flame displacement speed and the mean fluid velocity show that the
convection in the turbulent flames will affect the flame displacement speed for high Reynolds number flames.
1. Introduction
The description of dynamics of flame front has been one of the most important subjects in the
turbulent combustion research since that is the basis of turbulent combustion models in the flamelet
concept (Clavin 1985, Williams 1985, Pope 1988, Peters 2000, Law 2000, Williams 2000). In the
concept of the flame stretch, increasing rate of the flame area is expressed by flame displacement
speed, flame curvature and strain rate at the flame front (Pope 1988, Candel et al. 1990). The
effects of the strain rate have been discussed based on the Markstein number (Peters 2000, Clavin
2000). In previous studies (Poinsot et al. 1992, Sinibaldi et al. 2003), the curvature and strain rate
effects have been investigated in the steady or unsteady laminar flames because the flame elements
in turbulent flames are assumed to be laminar flame under weak stretch with small curvature.
To confirm the theory of the flame stretch experimentally, measurements of the flame displacement
speed are required. In general, shadowgraph or laser tomography has been adopted to estimate
flame displacement (Sinibaldi et al. 2003, Kido et al. 2003). However, in these techniques, flame
propagation normal to the flame front is assumed. In turbulent flames, the flame elements do not
always move into the flame normal direction due to strong convection effects of turbulence. In the
numerical investigations by direct numerical simulations (DNS) (Haworth and Poinsot 1992, Baum
et al. 1994, Echekki and Chen 1996, Tanahashi et al. 2000, Oijen et al. 2005), consumption speed
and local heat release rate are commonly used to represent characteristics of flamelets instead of the
flame displacement speed, and effects of the curvature and stretching rate have been discussed. In
experiments, measurements of the local consumption speed or heat release rate are quite difficult
even by CH planer laser-induced fluorescence (PLIF) due to severe calibration problem.
To investigate turbulent flame structures experimentally, PLIF (Dyer and Crosley 1984, Hanson
-1-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
1986) of molecules and radicals produced in chemical reactions, such as OH (Smooke et al. 1992),
CO (Seitzman et al. 1987), CH (Allen et al. 1986, Carter et al. 1998, Mansour et al. 1998) and
CH2O (Böckle et al. 2000), are commonly used. Since CH radicals are promoted at the flame
front and have narrow width enough to represent the reaction zones, CH PLIF measurements have
been used to visualize instantaneous flame structures. Although conventional PLIF measurements
have been a very powerful tool to obtain instantaneous local flame structures, investigation of
dynamics of flame structures in turbulence has been difficult due to the limitation of time resolution.
The time resolution of conventional PLIF measurement is of the order of Hz, which is limited by
experimental instruments, in most cases by camera and laser power. Recently, several time
resolved PLIF measurement have been reported (Kychakoff et al. 1987, Seitzman et al. 1994,
Schefer et al. 1994, Kaminski et al. 1999, Nygren et al. 2001, Watson et al. 2002, Hult et al. 2005).
To investigate mechanism of flame stabilization of the turbulent jet diffusion flames, double-pulsed
CH PLIF and CH4 Raman scattering technique has been reported by Schefer et al. (1994) and
simultaneous CH double-pulsed PLIF and particle image velocimetry (PIV) has been conducted by
Watson et al. (2002). By using 4 head YAG laser cluster, Kaminski et al. (1999), Nygren et al.
(2001) and Hult et al. (2005) have investigated dynamics of the flame front from time resolved OH
PLIF.
In this study, CH double-pulsed PLIF measurement is adopted to evaluate flame displacement
speed in turbulent premixed flames. The developed time resolved CH PLIF system is applied to
relatively high Reynolds number turbulent premixed flames on a swirl-stabilized combustor, and
dynamics of local flame structures are discussed. To evaluate the flame displacement speed
directly from successive CH images, a flame front identification and displacement vector evaluation
schemes are developed.
2. Experimental Setup and Apparatus
2.1 CH Double-Pulsed PLIF Measurement System
The schematic diagram of the experimental setup for CH double-pulsed PLIF measurement is
shown in Fig.1. By combining two independent CH PLIF systems, the time-resolved PLIF system
is comprised. For CH-PLIF measurement, the Q1(7,5) transition of the B2Σ--X2Π(0,0) band at
390.30nm is excited and fluorescence from the A-X(1,1), (0,0) and B-X(0,1) bands between 420
and 440nm is detected. CH radical is excited by two laser systems. First dye laser (Lambda
Physik, Scanmate2) is pumped by a XeCl excimer laser (Lambda Physik, LPX 110I, 308nm,
200mJ/pulse) with BiBuQ dye in p-dioxane solvent and second dye laser (Sirah Precisionscan) is
pumped by a Nd:YAG laser (Spectra-Physics, Quanta Ray PRO-190, 355nm, 300mJ/pulse) with
Exalite389 dye in p-dioxane solvent. These laser systems generate laser pulses of about 1820mJ/pulse and about 20-22mJ/pulse, respectively. Laser beams from each laser systems are
polarized in vertical and horizontal direction and are lead to same optical pass by a polarizing beam
splitter. The combined beams are expanded into laser sheets by laser sheet forming optics.
Fluorescence from excited CH radicals are detected by two intensified CCD cameras (Andor
Technology, iStar DH734-25U-03, 1024 × 1024pixels) fitted with 105mm f2.8 lens (Nikon, Micronikkor) and optical filter (SCHOTT, KV418). These cameras are located on the opposite side of
the burner, and optical axes are set to be perpendicular to the laser sheets. Timing control of these
laser systems and cameras are conducted by a delay/pulse generator (SRS, DG535) and internal
delay generator of each camera. Therefore, time interval ∆t of successive PLIF could be selected
unrestrainedly. In this study, ∆t is set to the range of 30 to 200µs. The laser beam is shaped into
200µm thickness vertical sheet with 30mm height. The thickness of the laser sheet is measured by
-2-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Fig. 1. Schematic diagram of the CH double-pulsed PLIF measurement.
Table 1. Experimental conditions.
Q[l/m]
200
250
300
Rel
265.8
602.0
881.0
Reλ
63.1
95.0
115.0
u′rms[m/s]
1.24
1.59
1.92
l [mm]
3.42
6.02
7.27
λ[mm]
0.812
0.950
0.949
η[mm]
0.0575
0.0548
0.0498
l/δF
82.9
146.0
176.6
u′rms/SL
3.20
4.12
4.99
a photodiode scanning. Each camera affords a view of 30mm×30mm. Therefore, spatial
resolution of PLIF is 30µm×30µm×200µm. To optimize signal-to-noise ratio, image intensifier
gate time is set to 30ns; this value is given from preliminary experiments conducted on a steady
laminar flame in the slot burner with different gate time.
2.2 Experimental Apparatus
Figure 2 shows schematics of the swirl-stabilized combustor (Tanahashi et al. 2005) and a direct
photograph of methane-air turbulent premixed flame at stoichiometric condition with flow rate Q =
300L/min. This combustion rig consists of contraction section, swirl injection section and
combustion chamber. The inner diameter of 120mm in the contraction section is reduced to 40mm
diameter in the swirl injection section. The inner-cross section of combustion chamber is 120mm
×120mm, and the length of the chamber is 550mm. On each side of the combustion chamber, a
silica glass plate of 120mm×170mm and 5mm thickness is installed to allow optical access. The
swirl nozzle has swirl vanes of 14mm inner diameter and 40mm outer diameter with inclination of
45 degree from the nozzle axis. Although a secondary fuel nozzle is mounted at center of the swirl
vanes, this nozzle is not used in the present study. In this experiment, vertical and horizontal
directions of the laser sheet is defined as the x and y axis, respectively. The z axis is selected to be
normal to the laser sheet.
In this study, CH double-pulsed PLIF measurements are conducted for three different flow rates; Q
= 200, 250 and 300L/min, at stoichiometric condition. To investigate turbulence characteristics of
the unburned mixture, hot-wire measurements are conducted in inert flows (Tanahashi et al. 2005).
The CH double-pulsed PLIF measurement is conducted in a cross-section with the maximum u'rms,
where u'rms indicates turbulence intensity. A white box in the direct photograph denotes the
measurement region of PLIF. Turbulence characteristics obtained by the hot-wire measurement
are shown in Table 1. Here, l is integral length scale, Rel is Reynolds number based on integral
length scale and u'rms, Reλ is Reynolds number based on Taylor micro scale (λ) and u'rms, δF is a
-3-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Fig. 2. Schematic of a swirl combustor (left) and a direct photograph of turbulent premixed flame
for Q = 300L/min (right).
laminar flame thickness and SL is laminar burning velocity. With the increase of the flow rate, Reλ
changes from 63.1 to 115.0. These conditions are the same with those in our previous study for
simultaneous CH-OH PLIF and stereoscopic PIV (Tanahashi et al. 2005) and classified into the
corrugated flamelets regime in the turbulent combustion diagram by Peters (1999).
3. Dynamics of Flame Front in Turbulence
Typical successive CH fluorescence images for Reλ = 63.1 (Q = 200L/min) are shown in Fig. 3 for
three different realizations. Time interval of successive CH PLIF are 100µs for Fig. 3 (a) and (b),
and 200µs for (c), and each image shows 30mm×30mm regions. In these figures, black line
represents the integral length scale of turbulence. Flame front represented by high concentration
CH layer shows very complicated distribution and have several sizes of wrinkling; large-scale more
than integral length scale to small-scale less than Taylor micro scale (Tanahashi et al. 2005). In
addition to these instantaneous features of flame front, dynamics of flame front can be discussed
from the successive flame front images. As shown in Fig. 3, ring-like flame fronts are found in the
first images (see circle A). The size of these flame fronts is diminished or they disappear in the
second images. Isolating process of the unburned mixture from main flame front is also observed
as denoted by a circle B in Fig. 3. These deformations of flame front might represent the burning
and generating process of the unburned mixture in the burnt gas or the handgrip structure which has
been shown by three-dimensional direct numerical simulation (DNS) of turbulent premixed flames
(Nada et al. 2004).
4. Direct Measurement of Flame Displacement Speed
4.1 Evaluation of Flame Displacement Speed
In this study, from successive CH fluorescence images, flame displacement speed is evaluated by
the following procedures:
-4-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
(a)
(b)
(c)
Fig. 3. CH double-pulsed fluorescence images of CH4-air turbulent premixed flames ( φ = 1.0, Reλ
= 63.1, (a) and (b): ∆t = 100µs, (c): ∆t = 200µs ).
(1) Set interrogation region just like an analysis technique for the conventional PIV.
(2) Define the flame front from the mean intensity of CH fluorescence in each interrogation
regions.
(3) Evaluate flame displacement speed using the cross-correlation method.
Here, optimizations of the size of interrogation region and the magnitude of the flame displacement
are required to evaluate the displacement speed by the cross-correlation method. Time interval of
successive CH PLIF and interrogation region size are set to 30µs and 48 × 48pixels, respectively.
These conditions were determined from preliminary experiments with different time interval.
Note that thickness of clearly captured flame front is not more than 15pixels. The crosscorrelation algorithm based on fast Fourier transformation is used to calculate the local flame
displacement speed in two-dimensional cross section. This code is mainly constructed based on
the high resolution PIV algorithm (Tanahashi et al. 2002). To detect the flame front precisely,
interrogation regions are 50% overlapped. To evaluate the displacement within sub-pixel accuracy,
3-point Gaussian fitting is applied.
In the process of flame displacement speed evaluation, detection of the flame front is most
important. Figure 4 (a) shows probability density functions of mean fluorescence intensity in
interrogation regions. The mean fluorescence intensity is normalized by the maximum
fluorescence intensity of each image. Note that it is impossible to conform the photographic
-5-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
(a)
(b)
Fig. 4. Probability density functions of mean CH signal in the interrogation regions (a), and the
maximum of correlation function in the interrogation regions (b).
(a)
(b)
(c)
Fig. 5. Successive CH fluorescence images (a) and (b), and detected flame front (c) (Reλ = 63.1, ∆t
= 30µs).
sensitivity of two ICCD cameras and pulse energy of successive laser beams. Therefore, signalto-noise ratios of the successive CH fluorescence images are different. If the same threshold was
selected for the mean fluorescence intensity, flame front that does not exist in fact might be detected
incorrectly. As for the case with the independent threshold for each image, it is difficult to remove
the influence of accidental shot noises. In these situations, the threshold for flame front detection
should be selected based on the geometrical average of fluorescence intensities in the corresponding
interrogation regions. In Fig. 4 (a), probability density function of the geometrical average is also
plotted. This result suggests that 0.3 of the normalized geometrical average is suitable for the
threshold to detect the flame front correctly.
In the cross-correlation method, displacements of flame front in interrogation regions are evaluated
from the location with the maximum of correlation function. Figure 4 (b) shows probability
density functions of the maximum of correlation functions. Since low correlation gives spurious
displacement speeds, threshold of the maximum correlation coefficient is set to be 0.35.
Furthermore, due to 50% overlap of interrogation regions, several adjacent interrogation regions
might be detected as flame front. To eliminate the region in which flame front is NOT included
completely, fluorescence intensity in central area of interrogation regions is compared with mean
intensity. If the intensity in central area is stronger than the mean one, that region is treated as
flame front. This discrimination is effective to reduce the spurious vectors. Figure 5 shows the
successive CH fluorescence images and detected flame front from corresponding images.
Detected flame front coincide very well with the location where CH intensity is high. Figure 6
shows vector map of flame displacement speed for three different Reynolds number with successive
CH fluorescence images. In these figures, red and blue color represents 1st and 2nd CH
-6-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
(a)
(b)
(c)
Fig. 6. Successive CH fluorescence images and vector maps of flame displacement speed (red: 1st
image, blue: 2nd image, (a): Reλ = 63.1, (b): Reλ = 95.0, (c): Reλ = 115.0).
(a)
0.12
Reλ = 63.1
0.10
0
Reλ = 63.1
Reλ = 95.0
(c) 10
0
Reλ = 63.1
Reλ = 95.0
Reλ = 115.0
0.08
P
(b) 10
10
Reλ = 95.0
Reλ = 115.0
-1
10
Reλ = 115.0
-1
0.06
10-2
0.04
10-2
0.02
0.00
0
5
10
15
20
10-3
-20
-15
-10
-5
Sd [ m/s ]
0
5
10
Sd,u [ m/s ]
15
20
10-3
-20
-15
-10
-5
0
5
10
15
20
Sd,v [ m/s ]
Fig. 7. Probability density functions of flame displacement speed (a), and x: (b) and y: (c)
component.
fluorescence image, respectively. White vectors represent the local flame displacement speed
evaluated by the above-mentioned method. Directions of the flame displacement vectors coincide
very well with observations from the successive CH fluorescence images.
4.2 Flame Displacement Speed of Turbulent Premixed Flame
In Fig. 7 (a), probability density functions of local flame displacement speed Sd are presented. To
obtain these results, 120 sets of successive CH images and more than 5000 displacement vectors are
used for each condition. The probability for high displacement speed is a little bit high for high
Reynolds number case. Mean fluid velocity in the measurement region has been measured by
stereoscopic PIV (Tanahashi et al. 2005). The mean velocity and turbulence intensity increase for
high Reynolds number. Typical values are about 5.0m/s to 8.9m/s for mean streamwise velocity
and about 2.5m/s to 3.6m/s for turbulence intensity. Although the velocity field depends on
Reynolds number significantly, the flame displacement speed is slightly affected by Reynolds
number. These results reflect the correctness of the flamelet concept, because the turbulent
burning velocity increase due to the increase of flame surface area and the flame displacement
speed is not modified significantly.
Figure 7 (b) and (c) show probability density functions of x and y component of the flame
displacement speed. The probability density functions of each component are independent on
Reynolds number and coincide very well. The different Reynolds number effect in Sd and in Sd,u
and Sd,v is caused by the correlation between Sd,u and Sd,v, and closely related to hydrodynamical
structure as described below. It should be noted that probability density functions in Fig. 7 (b) and
-7-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
(a)
(b)
(c)
Fig. 8. Vector maps of mean flame displacement speed and distributions of the number density of
detected flame front, (a): Reλ = 63.1, (b): Reλ = 95.0, (c): Reλ = 115.0.
(a)
(b)
(c)
Fig. 9. Vector maps of mean flame displacement speed (white arrows) and mean fluid velocity
(black arrows), (a): Reλ = 63.1, (b): Reλ = 95.0, (c): Reλ = 115.0.
(c) show large peaks at 0m/s. This might be caused by several possible reasons. One is existence
of the re-circulation zone in the swirl-stabilized combustor. As the re-circulation zone play an
important role for flame stabilization, stationary flame with Sd ≈ 0 has high possibility even for the
turbulent state. Another is three-dimensionality of turbulent flames. Since the present
measurement is conducted in two-dimensional cross section in highly turbulent swirling flow, flame
fronts which propagate into out-of-plane can not be measured. Such flame elements are detected
as stationary flame with Sd ≈ 0.
Figure 8 shows mean flame displacement vector maps and distributions of number density of
detected flame front. As the present measurements were conducted in the swirl-stabilized
combustor, isolated flame islands which are shown in Fig. 3 move into the upper-left side in the
measurement region. With the increase of flow rate (or Reynolds number), the isolated unburned
mixture in the burned gas increases as shown in Fig. 6 and in our previous study (Tanahashi et al.
2005), which results in the increase of turbulent burning velocity in the flamelet regions.
Therefore, the magnitude of the mean flame displacement speed is larger for high Reynolds number
case. In our previous study (Tanahashi et al. 2005), simultaneous CH-OH PLIF and stereoscopic
PIV measurements have been conducted in the same plane. In Fig. 9, mean flame displacement
vectors are compared with the mean velocity vectors obtained from previous PIV measurements.
The mean velocity distributions are evaluated from more than 130 realizations, and the compared
region are denoted by a box with dotted lines in Fig. 8. With the increase of Reynolds number, the
mean flame displacement vectors tend to align with the mean velocity vectors. These results
suggest that effects of convection become strong for high Reynolds number case. Even for the
-8-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
highest Reynolds number case, misalignment still remains, which might be caused by the sampling
number for the present displacement speed measurement. However, this misalignment will be
explained from the simultaneous CH double-pulsed PLIF and stereoscopic PIV, which gives flame
displacement speed, flame curvature and the strain rare simultaneously, in the near future.
5. Summary
In this study, CH double-pulsed PLIF measurement has been developed to measure the local flame
displacement speed directly. The developed measurement system was applied for relatively high
Reynolds number methane-air turbulent premixed flames stabilized on a swirl combustor.
By selecting an appropriate time interval of successive CH PLIF, modifications of the flame front
are captured clearly. From the successive CH fluorescence images, the burning and generating
process of the unburned mixture in the burnt gas or the handgrip structure which have shown by
DNS can be observed.
To evaluate the local flame displacement speed directly from the successive CH images, a flame
front identification and displacement vector evaluation schemes are developed. Local flame
displacement speeds coincide well for different Reynolds number cases, which represent the
correctness of the flamelet concept. Furthermore, comparison of mean flame displacement speed
and mean fluid velocity shows that convection in the turbulent flames will affect the flame
displacement speed.
Direct measurement of local flame displacement speed described in this paper can be conducted
with stereoscopic PIV measurement simultaneously. The simultaneous CH double-pulsed PLIF
and stereoscopic PIV measurements will provide detail information about important factors such as
flame displacement speed, flame curvature and strain rate to describe the dynamics of flame surface
in turbulence.
Acknowledgments
This work is partially supported by Grant-in-Aid for Scientific Research (A) (No.15206023) of
Japan Society for the Promotion of Science.
References
Allen MG, Howe RD, Hanson RK (1986) Digital imaging of reaction zones in hydrocarbon-air
flames using planar laser-induced fluorescence of CH and C2. Opt. Lett. 11: 126-128
Baum M, Poinsot TJ, Haworth DC, Darabiha N (1994) Direct numerical simulation of H2/O2/N2
flames with complex chemistry in two-dimensional turbulent flows. J. Fluid Mech. 281: 1-32
Böckle S, Kazenwadel J, Kunzelmann T, Shin D -I, Schulz C, Wolfrum J (2000) Simultaneous
single-shot laser-based imaging of formaldehyde, OH and temperature in turbulent flames. Proc.
Combust. Inst. 28: 279-286
Candel SM, Poinsot TJ (1990) Flame stretch and the balance equation for the flame area. Combust.
Sci. and Tech. 70: 1-15
Carter CD, Donbar JM, Driscoll JF (1998) Simultaneous CH planar laser-induced fluorescence and
-9-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
particle imaging velocimetry in turbulent nonpremixed flames. Appl. Phys. B 66: 129-132
Clavin P (1985) Dynamic behavior of premixed flame fronts in laminar and turbulent flows. Prog.
Energy Combust. Sci. 11: 1-59
Clavin P (2000) Dynamics of combustion fronts in premixed gases: from flames to detonations.
Proc. Combust. Inst. 28: 569-585
Dyer MJ, Crosley DR (1984) Rapidly sequenced pair of two-dimensional images of OH laserinduced fluorescence in a flame. Proc. Int. Conf. Laser 84: 211-218
Echekki T, Chen JH (1996) Unsteady strain rate and curvature effects in turbulent premixed
methane-air flames. Combust. Flame 106: 184-202
Hanson RK (1986) Combustion diagnostics: planar imaging techniques. Proc. Combust. Inst. 21:
1677-1691
Haworth DC, Poinsot TJ (1992) Numerical simulations of Lewis number effects in turbulent
premixed flames. J. Fluid Mech. 244: 405-436
Hult J, Meier U, Meier W, Harvey A, Kaminski CF (2005) Experimental analysis of local flame
extinction in a turbulent jet diffusion flame by high repetition 2-D laser techniques and multiscalar measurements. Proc. Combust. Inst. 30: 701-709
Kaminski CF, Hult J, Alden M (1999) High repetition rate planar laser induced fluorescence of OH
in a turbulent non-premixed flame. Appl. Phys. B 68: 757-760
Kido H, Nakahara M, Nakashima K, Hashimoto J (2003) Influence of local flame displacement
velocity on turbulent burning velocity. Proc. Combust. Inst. 29: 1855-1861
Kychakoff G, Paul PH, van Cruyningen I, Hanson RK (1987) Movies and 3-D Images of
Flowfields Using Planar Laser-Induced Fluorescence. Appl. Opt. 26: 2498-2500
Law CK, Sung CJ (2000) Structure, aerodynamics, and geometry of premixed flamelets. Prog.
Energy Combust. Sci. 26: 459-505
Mansour MS, Peters N, Chen YC (1998) Investigation of scalar mixing in the thin reaction zones
regime using a simultaneous CH-LIF/Rayleigh laser technique. Proc. Combust. Inst. 27: 767-773
Nada Y, Tanahashi M, Miyauchi T (2003) Effect of turbulence characteristics on local flame
structure of H2-air premixed flames. J. Turbulence 5: pn.16
Nygren J, Richter M, Hult J, Kaminski CF, Alden M (2001) Temporally resolved single-cycle
measurements of fuel- and OH-distributions in a spark ignition engine using high speed laser
spectroscopy. Fifth Intl. Symp. on Diagnostics and Modeling of Combustion in Internal
Combustion Engines 572-580
Peters N (1999) The turbulent burning velocity for large-scale and small-scale turbulence. J. Fluid
Mech. 384: 107-132
Peters N (2000) Turbulent Combustion. Cambridge Press.
Poinsot TJ, Echekki T, Mungal MG (1992) A study of the laminar flame tip and implications for
premixed turbulent combustion. Combust. Sci. and Tech. 81: 45-73
Pope SB (1988) The evolution of surfaces in turbulence. Intl. J. Engng. Sci. 26: 445-469
Schefer RW, Namazian M, Filtopoulos EEJ, Kelly J (1994) Temporal evolution of
turbulence/chemistry interactions in lifted, turbulent-jet flames. Proc. Combust. Inst. 25: 12231231
Seitzman JM, Haumann J, Hanson RK (1987) Quantitative two-photon LIF imaging of carbon
monoxide in combustion gases. Appl. Opt. 26: 2892-2899
Seitzman JM, Miller MF, Island TC, Hanson RK (1994) Double-pulsed imaging using simultaneous
OH/acetone PLIF for studying the evolution of high-speed, reacting mixing layers. Proc.
Combust. Inst. 25: 1743-1750
Sinibaldi JO, Driscoll JF, Mueller CJ, Donbar JM, Carter CD (2003) Propagation speeds and stretch
rates measured along wrinkled flames to assess the theory of flame stretch. Combust. Flame 133:
323-334
- 10 -
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Smooke MD, Xu Y, Zurn RM, Lin P, Frank JH, Long MB (1992) Computational and experimental
study of OH and CH radicals in axisymmetric laminar diffusion flames. Proc. Combust. Inst. 24:
813-821
Tanahashi M, Fujimura M, Miyauchi T (2000) Coherent fine-scale eddies in turbulent premixed
flames. Proc. Combust. Inst. 28: 529-535
Tanahashi M, Murakami S, Choi GM, Fukuchi Y, Miyauchi T (2005) Simultaneous CH-OH PLIF
and stereoscopic PIV measurements of turbulent premixed flames. Proc. Combust. Inst. 30:
1665-1672
Tanahashi M, Ootsu M, Fukushima M, Miyauchi T (2002) Measurement of coherent fine scale
eddies in turbulent mixing layer by DPIV. Engineering Turbulence Modeling and Measurements
5: 525-534
van Oijen JA, Groot GRA, Bastiaans RJM, de Goey LPH (2005) A flamelet analysis of the burning
velocity of premixed turbulent expanding flames. Proc. Combust. Inst. 30: 657-664
Watson KA, Lyons KM, Carter CD, Donbar JM (2002) Simultaneous two-shot CH planar laserinduced fluorescence and particle image velocimetry measurements in lifted CH4/air diffusion
flames. Proc. Combust. Inst. 29: 1905-1912
Williams FA (1985) Combustion Theory. Benjamin Cummings
Williams FA (2000) Progress in knowledge of flamelet structure and extinction. Prog. Energy
Combust. Sci. 26: 657-682
- 11 -