Publisher’s version / Version de l'éditeur:
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
Combustion Institute Canadian Section, 2004 Spring Technical Meeting
[Proceedings], 2004
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright
NRC Publications Archive Record / Notice des Archives des publications du CNRC : https://nrc-publications.canada.ca/eng/view/object/?id=e65efa4f-e518-4339-9a98-b87d1de38fd5 https://publications-cnrc.canada.ca/fra/voir/objet/?id=e65efa4f-e518-4339-9a98-b87d1de38fd5
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Soot concentration profiles in a non-premixed methane laminar flame
at high pressures
Thomson, Kevin; Gulder, Omer L.; Weckman, E.; Fraser, R.; Smallwood,
Gregory; Snelling, David
SOOT CONCENTRATION PROFILES IN A NON-PREMIXED METHANE
LAMINAR FLAME AT HIGH PRESSURES
1K. THOMSON
*,2, Ö. GÜLDER
†, E. WECKMAN
*, R. FRASER
*,
G. SMALLWOOD
‡, D. SNELLING
‡*
Mechanical Engineering Department University of Waterloo Waterloo, Canada, N2L 3G1
†
Institute for Aerospace Studies University of Toronto Toronto, Canada, M3H 5T6
‡
ICPET, Combustion Research Group National Research Council of Canada
Ottawa, Canada, K1G 0R6
INTRODUCTION
Since most practical combustors operate at high pressures (i.e., 2 – 10 MPa), it is of interest to understand how pressure influences the combustion phenomena, in particular soot formation pathways. There have been a number of fundamental studies in this area, e.g. [1-5]; however, none have comprehensively addressed the issue of soot formation at high pressures.
Using a diffusion flame burner operating with ethylene, Flower and Bowman [5] report maximum line-of-sight integrated soot volume fractions
line
( )d
,
n
v v
f
=
∫
f r r
∝
P
(1)where n = 1.2 ± 0.1 for P = 0.1 to 1 MPa. Lee and Na [4] also show line-of-sight integrated soot volume fractions for pressures of 0.1 to 0.4 MPa in an ethylene laminar diffusion flame. Their measurements suggest n = 1.26. These results are difficult to interpret since they represent linearly weighted averages through an annular soot distribution, thus, neither the peak nor total soot. The only spatially resolved measurements of soot volume fraction as a function of pressure are reported by Lee and Na [4]. The published data is very limited, but suggests
max 2 v
f ∝P for P = 0.2 to 0.4 MPa at a height of 20 mm above the burner nozzle, where
max v
f
is the maximum soot volume fraction.It is evident that information on soot in non-premixed laminar flames at higher pressures is limited. In this paper, measurements of soot volume fraction obtained using spectral soot emission (SSE) and line-of-sight attenuation (LOSA) measurements for pressure of 0.5 to 4 MPa are presented and compared. Soot production trends are then calculated and discussed. These results greatly extend the currently available information on flame sooting propensity at high pressures.
EXPERIMENTAL METHODOLOGY
The experimental pressure vessel and burner used in this study are described in [6]. Flame stability was assessed at pressures of 0.5, 1, 2, and 4 MPa using a digital video camera for methane flow rates of 0.55 mg/s and air flow rate of 0.4 g/s. Flames exhibit good, long term stability for all pressures with an rms flicker of the flame tip less than 0.1 mm. A typical flame height was 8 mm.
1
Combustion Institute/Canadian Section Spring Technical Meeting, Queen’s University Kingston, ON, May 9-12, 2004.
2
The theory and overall experimental layout of the spectral soot emission diagnostic (SSE) are described previously [7]. In SSE, line-of-sight radiation emission from soot is measured along chords through the flame. A series of emission projections at a given height in the flame can be inverted to obtain radially resolved emission rates from which temperature and soot volume fraction can be determined when soot optical properties are known. For the current measurements a 300 mm focal length lens (f/45, 2:1 magnification) is used to image the object plane at burner centre onto a vertical entrance slit (height 0.500 mm, width 0.025 mm) of a spectrometer. Output from the spectrometer is imaged onto a 16-bit CCD detector (1100x330 pixels). Knife edge scans across a diffuse light source located at the object plane show a horizontal spatial resolution of 50 µm. The system is calibrated for radiation intensity using a calibrated filament lamp placed inside the chamber. The uncertainty in the spectral radiance temperature is ± 5 K. Soot emission is measured over a wavelength range of 690-945 nm. Spectra are averaged over the vertical height of the entrance slit as well as across 21 nm spectral widths, thus providing 12 spectral data points per line-of-sight acquisition. One-dimensional tomography is applied to each wavelength range using a three-point Abel inversion method [8]. Local temperatures are determined from the spectral shape of the inverted soot emission intensity. Soot volume fraction is then determined from the soot emission intensity using the measured temperatures. The soot refractive index function, E(m), is assumed to be constant and equal to 0.274 for the calculations. This assumption is consistent with results of Krishnan et al. [9]. Sensitivity of SSE results to E(m) is discussed in [7]. Modelling of the flame emission using the methods described in [7] shows that emission attenuation by soot introduces only a small error (i.e. < 2%) in the measurements for even the highest soot loadings observed in this flame. Therefore no attenuation correction is applied.
The line-of-sight attenuation (LOSA) diagnostic is a simplified version of the 2D LOSA diagnostic described in [21]. In LOSA, line-of-sight measurement of the intensity of small light beam transmitted through a flame is made. When divided by a measurement of the intensity of the beam transmitted along the same path without the flame present, the transmissivity of the flame along the chord can be determined. A series of transmissivity measurements at a given height in the flame can be inverted to obtain radially resolved extinction coefficients from which soot volume fraction can be determined. The optical layout for the LOSA measurements is included in Figure 1. Light from a mercury arc lamp is first focussed onto a 50 µm pinhole. Light transmitted through the pinhole is modulated using a chopper wheel and imaged at the centre plane of the burner with a 1.5:1 demagnification at a speed of f/19. Knife edge scans of the lamp beam at burner centre show the beam width to be less than 40 µm across the diameter of the burner nozzle. A collection lens downstream of the burner re-focuses the transmitted lamp light onto a photodiode detector coupled to a lock-in amplifier. The collection lens is large (i.e., 100 mm dia.) to accommodate beam steering of the light transmitted through the flame, which becomes quite pronounced at 4 MPa. A glass plate located between the imaging lens and chamber reflects a portion of the lamp light onto a second photodiode which is used to ratio out temporal variation in the lamp intensity. Both photodiodes are filtered with 830 nm narrow band filters. For each measurement height, two scans are required, one with the flame lit and the second with the flame extinguished. The method used to calculate soot volume fraction measurements from line-of-sight transmissivity measurements is described in [10].
Measurements of soot volume fraction were obtained using SSE and LOSA for pressures of 0.5, 1, 2, and 4 MPa. A constant mass flow rate of methane and air of 0.55 mg/s and 0.4 g/s, respectively, was maintained across all pressure. For each pressure, measurements were obtained at height increments of 0.5 mm from the base to the tip of the flame and at horizontal increments 50 µm; however, plots of spatially resolved soot volume fraction are only reported for height increments of 1.0 mm because of space limitations.
Figure 1 – Line-of-sight attenuation layout RESULTS AND DISCUSSION
Soot volume fraction measurements are included in Figures 2 – 5, for pressures of 0.5, 1, 2, and 4 MPa, respectively. It is observed that the overall agreement between the SSE and LOSA soot volume fraction measurements is good. At a pressure of 0.5 MPa, both diagnostics are at their sensitivity limits and the measurements are therefore noisy. This likely explains the observed measurement differences near burner centreline, especially in the upper half of the flame. For all pressures, in the lower half of the flame, LOSA measurements are typically higher than SSE measurements. The differences can be as large as 80%, however, are more typically below 30%. The curve shapes and peak locations from each diagnostic agree well. With increasing height in the flame, the differences between the diagnostics diminish. The agreement is good in the upper half of the flame, though at 4 MPa the LOSA measurements are higher.
There are several possible explanations for the observed differences in the measurements from the two diagnostics. First, the LOSA diagnostic measures the scatter coefficient of soot rather than the absorption coefficient (from which soot volume fraction is determined). Therefore, the measurements will consistently overestimate the soot volume fraction. Scatter contribution increases with soot particle and aggregate size. Therefore the tendency of the LOSA measurements to exceed the SSE measurements low down in the flame is very interesting and not intuitive. Measurement of light scatter or of soot morphology could help to explain the observed differences between the measurements. Second, small variation in the z-axis location of the burner between LOSA and SSE measurements could lead to different soot volume fraction curves, since the soot volume fraction changes quite rapidly with z, particularly at the tip of the flame. Finally, the SSE measured soot volume fraction is coupled to the measured soot temperature so an error in temperature will lead to an error in soot volume fraction. Conversely, good agreement of the soot volume fraction measurements indicates that the SSE temperatures must be quite accurate since
I
∝
f
vexp(1/ )
T
−1 where I is the local emission intensity,f
vis the soot volume fraction and T is the temperature.
Figures 2 - 5 illustrate the narrowing of the flame with increasing pressure. Theoretical analysis suggests that the height of a constant mass flow rate diffusion flame is invariant with pressure [11]. This prediction is approximately true over the pressure range studied with a flame height of 8.5 mm for P = 0.5 – 2 MPa, dropping to 7.5 mm for 4 MPa. Residence time is also thought to be independent of pressure which can only be possible if the flame area decreases inversely with pressure (neglecting the effects of air entrainment or temperature distribution shape changes). The cross-sectional area of the flame,
A
cs, as measured using the radial location of the outer edges of the sooting region at each measurement height, is observed to decrease with pressure asA
cs∝
P
−n, where n = 1.0 ± 0.1. Although the observed n is consistent with the above argument, n is approximately double the value suggested by Glassman [12].arc lamp collection lens chopper motor mirrors focusing lens narrowband filter beam splitter
photodiode chamber windows
burner object plane collection lens narrowband filter photodiode pinhole
aperture collectionlens arc lamp chamber
chopper motor mirrors focusing lens narrowband filter beam splitter
photodiode chamber windows
burner object plane collection lens narrowband filter photodiode pinhole aperture chamber
0 10 20 30 40 50 60 70 -1 .4 -1 -0 .6 -0 .2 0. 2 0 .6 1 1 .4 r [ mm] SVF [ppm] LOSA SS E 2.0 MPa 0 20 40 60 80 100 120 140 -1 .4 -1 -0 .6 -0 .2 0 .2 0 .6 1 1 .4 r [ mm] SVF[ppm] z = 1. 0m m z = 2. 0m m z = 3. 0m m z = 4. 0m m z = 5. 0m m z = 6. 0m m z = 7. 0m m z = 8. 0m m LOS A SSE 4.0 MPa Figur e 2 - LOSA (left) an d S S E (rig h t) me as u re m en ts, P = 0 .5 MP a (se e
Figure 5 for legen
d . Figur e 4 - LOSA (left) an d S S E (rig h t) me as u re m en ts, P = 2 .0 MP a (se e
Figure 5 for legen
d . Figur e 3 LOS A ( le ft) and SSE (right ) m easu rements, P = 1.0 MPa ( se e F igure 5 fo r leg en d . Figur e 5 LOS A ( le ft) and SSE (right ) m easu rements, P = 4.0 MPa . Note: n ot all heig h ts d isplaye d in all figu re s. LOS A SSE 1.0 MPa 0 2 4 6 8 10 12 14 16 18 20 -1 .4 -1 -0 .6 -0 .2 0. 2 0 .6 1 1 .4 r [mm] SVF[ ppm] LOSA SS E 0.5 MPa 0 0. 5 1 1. 5 2 2. 5 3 3. 5 4 -1 .4 -1 -0 .6 -0 .2 0.2 0 .6 1 1 .4 r [mm] SVF [ppm]
Based on the above observations, residence time is assumed to be independent of pressure and measurements at the same height above the burner exit are therefore deemed comparable.
Both SSE and LOSA measurements indicate that the maximum soot volume fraction increase as
max 2 v
f ∝P over the pressure range of 0.5 to 2 MPa which is consistent with the limited results of Lee and Na [4] for an ethylene flame. Comparing the 2 and 4 MPa results, the rate of increase in soot volume fraction drops to
max 1.2 v
f ∝P , suggesting a change in the soot formation mechanism.
Higher soot volume fractions are not unexpected with increasing pressure as all species are in higher concentration and the flame is narrowing. To quantify the sooting propensity of the flame at different pressures it is useful to calculate the percentage of total carbon converted to soot as a function of height. The mass flow rate of carbon, in the form of soot, can be determined through the relationship
s
( )
z( )
s2
v( , )d ,
m z
=
v z
ρ
∫
π
f r z r
(2)where vz is the axial velocity and
ρ
s= 1.8 g/mL is the soot density. The axial velocity is estimated usingthe relationship v zz( )= 2az, where a is an acceleration constant commonly assumed to be 25 m/s2 [5, 13]. The percentage of carbon in the fuel converted to soot is simply
η
s=
m m
s c,
wherem
c is the carbon mass flow rate at the nozzle exit. The results of this calculation are included in Figure 6. The curves support the previous observation that the difference between the LOSA and SSE measurements diminishes with height above the burner tip. Peak carbon conversion occurs at a height of about 5.5 mm above the burner nozzle for pressures of 0.5 and 1 MPa, 5 mm for a pressure of 2 MPa, and 4 mm for a pressure of 4 MPa. Up to the peak carbon conversion, the curves are approximately linear with height and the slope increases with pressure. Extrapolation of the curves in Figure 6 to zero conversion shows that soot inception moves closer to the burner with increasing pressure. This suggests that fuel pyrolysis and soot nucleation are enhanced by pressure. Comparing peak carbon conversion to soot, it is observed thatm
s∝
P
n, where n = 1 for the pressure range of 0.5 to 2 MPa and n = 0.1 for the pressure range of 2 to 4 MPa. Thus, it is shown that soot formation is enhanced by pressure even when density effects are removed. However, at pressures between 2 and 4 MPa, the sensitivity is quite low and it is speculated that the maximum carbon conversion to soot could begin to drop at pressures above 4 MPa. Further measurements about 4 MPa are needed to better understand the sooting trends.To compare the current results with those of Flower and Bowman [5] and Lee and Na [4], line integrals of the soot concentration profiles were calculated, e.g., Eq. 2. It is found that the maximum line integrated soot volume fraction,
line v
f
, varies as line n vf ∝P , where n = 1.3 ± 0.1 for P = 0.5 to 2 MPa and n = 0.9 ± 0.1 for P = 2 to 4 MPa. It is noted that for the lower pressure range, the correlation agrees with results of Flower and Bowman [14] and Lee and Na [4] and that the agreement with the carbon conversion rate is fair; however, for the 2 and 4 MPa range, the line-of-sight integrated measurements do not capture the dramatic change in the peak soot conversion. It is therefore concluded that line-of-sight integrated soot volume fraction can be a misleading measure of sooting tendency.
CONCLUSIONS
The work presented here represents a significant addition to the available database of information on soot formation tendencies as a function of pressure for non-premixed laminar flames. For the first time in laminar non-premixed flames, spatially-resolved soot volume fraction measurements over the pressure
range of 0.5 to 4 MPa have been made. Flame cross sectional area is observed to decrease with pressure as
A
cs∝
P
−n, where n = 1 ± 0.1. SSE and LOSA measurements of soot volume fraction agree within 20% and show that the peak soot concentration varies asmax n v
f ∝P , where n = 2 for P = 0.5 to 2 MPa and n = 1.2 for P = 2 to 4 MPa. Peak carbon conversion to soot mass increases with pressure as
s n
m
∝
P
, where n = 0.9 for P = 0.5 - 2 MPa and n = 0.1 for P = 2 to 4 MPa. It is apparent from these correlations that soot formation is enhanced by pressure but that it becomes less sensitive to pressure above 2 MPa. The soot measurements, when appropriately transformed, are consistent with the line-averaged measurements of Flower and Bowman [14] and Lee and Na [4]. Finally, it is concluded that line-averaged soot volume fraction measurements do not provide a clear picture of soot formation trends.Figure 6 – Percentage carbon conversion to soot ACKNOLEDGEMENTS
This work was performed at the National Research Council of Canada and supported in part by the AFTER POL of the Canadian Government’s PERD program. Funding for K. A. Thomson has been provided in part by a Carl A. Pollock Fellowship from the University of Waterloo, research grants from the Natural Sciences and Engineering Research Council (NSERC) of Canada, as well as a Graduate Student Supplement Scholarship from the National Research Council of Canada.
REFERENCES
1. T. Heidermann, H. Jander & H.G. Wagner, Physical Chemistry Chemical Physics 1 (15) (1999) 3497-3502. 2. R.V. Ravikrishna, D.D. Thomsen & N.M. Laurendeau, Comb. Sci. Tech. 157 (2000) 243-261.
3. I.M. Miller & H.G. Maahs, NASA Technical Paper TN D-8407 (1977). 4. W. Lee & Y.D. Na, JSME Int. Journal Series B 43 (4) (2000) 550-555. 5. W.L. Flower & C.T. Bowman, Proc. Combust. Inst. 21 (1986) 1115-1124.
6. K. Thomson, Ö. Gülder, E. Weckman, R. Fraser & G. Smallwood, "Preliminary characterization of a high pressure laminar diffusion flame combustion facility," Combustion Institute - Canadian Section, Spring
Technical Meetings, Vancouver, BC, 2003.
7. D.R. Snelling, K.A. Thomson, G.J. Smallwood, Ö.L. Gülder, E.J. Weckman & R.A. Fraser, AIAA Journal 40 (9) (2002) 1789-1795.
8. C.J. Dasch, Applied Optics 31 (8) (1992) 1146-1152.
9. S.S. Krishnan, K.-C. Lin & G.M. Faeth, Journal of Heat Transfer 123 (2001) 331-339.
10. D.R. Snelling, K.A. Thomson, G.J. Smallwood & Ö.L. Gülder, Applied Optics 38 (12) (1999) 2478-2485. 11. F.G. Roper, Combust. Flame 29 (3) (1977) 219-226.
12. I. Glassman, Proc. Combust. Inst. 27 (1998) 1589-1596.
13. F.G. Roper, C. Smith & A.C. Cunningham, Combust. Flame 29 (3) (1977) 227-234. 14. W.L. Flower, Combust. Flame 77 (1989) 279-293.
0 2 4 6 8 10 12 0 1 2 3 4 5 6 7 8 z [mm] ca rb o nc o nve rsi o n[ % ] 0.5 MPa 1.0 MPa 2.0 MPa 4.0 MPa LOSA SSE
