I
ALEXANDER WEIR, Jr., S. W. CHURCHILL, L. F. ORNELLA, and M. E. GLUCKSTEIN The University of Michigan, Ann Arbor, Mich.
Emission Spectra of Propane-Air Flames Irradiated with a 1000-Curie Gold Source intense irradiation of propane-air mixtures and flames increases the emission due to CH and C2 but hardly affects the emission due to OH
T m s report presents one phase of an experimental investigation of the effect of intense nuclear radiation upon flames. The rate of propagation of irradiated Bunsen flames has been reported (7). Studies of combustion in ramjet burners and low-pressure flames have indicated that emission spectra provide more insight into the mechanism of the combustion process than do flame speed and stability measurements (4, 6, 7). Accordingly, equipment was designed to obtain the spectral emission at a series of elevations through a flat flame sub,jetted to nuclear radiation. The intensities of the Cz, CH, and OH bands were measured quantitatively. The source of radiation for the emission studies was the same 30 grams of irradiated gold wire used in the flamespeed studies. The flame-speed measurements were carried out at source intensities as high as 10,000 curies. The reported data on spectral emission were obtained after the gold had decayed to 1000 curies. Data were obtained at a series of propane-air ratios and a series of subatmospheric pressures as the source decayed.
Experimental Equipment The equipment consisted of a premixed propane-air supply and metering system, a burner, a vacuum tank and exhaust system, shielding, a source of radiation, and an optical traversing system (Figure 1). All but the traversing system were used in the flamespeed studies, and details of the burner and source assembly were given in that
report (Figure 1,7) as well as a description of the shielding used. As indicated in Figure 3, a frontsurfaced, aluminum-coated mirror was mounted in a 12-inch pipe. welded on the side of the tank. The mirror reflected light emitted by the flame through a plastic window in the tank wall and focused the image of the flame on the slit of the spectrograph. Dry air was blown on the mirror and window to prevent fogging. A slide valve was installed in the 12-inch pipe to keep the mirror dry when the tank was filled with water. The spectrograph was a student model with a plastic replica grating, mounted directly on the tank to prevent relative motion between the spectrograph and flame image. The construction of the mount is shown in Figure 1. The back of the spectrograph was pivoted on an axle set in a pair of pillow blocks fastened on a bed plate. The front of the spectrograph rested on an adjusting screw fastened to the same bed plate. Thus the position of the horizontal slit could
This is the age of high energy fuels for high fuel efficiency and high speeds. The chemistrv ind physics of all aspects of flames and flame propagation will assume an ever-increasing importance in industrial literature.
be adjusted up and down to traverse the flame image vertically. A dial indicator gage was mounted to read elevations to a thousandth of an inch. The light emitted by the flame passed through the spectrograph slit, to the reflecting spectrograph grating, and then to the photograph film where the intensity a t the various wave lengths was recorded.
Source The source consisted of about 30 grams of 99.95% gold wire, 0.005 inch in diameter. The gold was irradiated about 4.5 days in a high-flux section of the Materials Testing Reactor (Phillips Petroleum Co., Idaho Falls, Idaho). The wire was wound in l/lG-inch coils before irradiation and was lightly compressed into a hollow cylinder (Figure 1,7) after irradiation. With this source location both the propane-air mixture and the flame were irradiated intensely. From neutron flux calculations ( 5 ) and subsequent measurements (3), the strength of the gold source on removal from the Materials Testing Reactor was estimated to be 9600 curies due to gold198 and 5670 curies due to gold-199 for a total of 15,270 curies. The strength a t the time of the various spectral emission experiments was computed on the basis of half lives of 2.69 days for gold-198 and 3.15 days for gold-199. The results are summarized in Table I. The production of ion pairs in the gas phase in the cylinder and in the flame front would appear to be the most significant phenomenon resulting from the gold source. The disintegration scheme of gold-198 and gold-199 and the estiVOL. 49, NO. 9
SEPTEMBER 1957
1423
LEAD SHIELD7 MIRROR 8
MO
\
ro
VACUUM
\ RESSURE GAGE
RADIOACTIVE GOLD BURNER HEAD
UNISTRUT SPECTROGRAPH
NOTE: LIGHT PATH INDICATED
By Figure 1 .
mated rate of production of ion pairs per cubic centimeter per curie in these two regions have been presented (Figure 2, and Table 11, 7).
Procedure Premixed propane and air were passed from storage tanks through a metering system into the burner. The propane was 99.570 pure. The composition of the propane-air mixtures was established from blending pressures and was checked by analysis. All runs were made with a linear velocity of about 1.0 foot per second at the burner head. At this velocity a flat flame floating above the burner screen was obtained over the entire range of operating pressures and propane-air ratios. At somewhat higher velocities and low pressures the flame formed a Bunsen cone (Figure 1 , 7 ) and at somewhat lower velocities and high pressures it burned on the screen. As a further condition emission from the
1424
-e---
Optical traversing system
flame had to be sufficiently intense to produce a satisfactory spectrographic negative in 30 minutes. The flame was essentially a disk about 1 inch in diameter and of the order of 0.05-inch thick. The exact thickness and distance above the burner screen varied with pressure and propane-air ratio. This thin vertical dimension of the flame was traversed by the spectrograph. The temperature of Lhe propane-air mixture leaving the burner head was not measured because of failure of a thermocouple. However, subsequent experiments with an electrically heated coil in place of the radioactive gold indicate that the propane-air stream could not have been heated more than 5" F. by the gold wire. Such an increase in temperature is insufficient to produce a significant change in the spectral emission of the flame. The pressure and flow rate were first established. The spectrograph film pack was loaded with Royal Pan film (Eastman Kodak Co., Rochester, N. Y . ) ?and
INDUSTRIAL AND ENCINEERINC CHEMISTRY
the slit was adjusted to the drsired elevation. The room was darkened, the mask removed from the film pack, and the shutter opened for 30 minutes. The time of exposure was limited by darkening of the film by scattered gamma radiation. The shutter was closed, a new strip of the film was shifted into position, the slit elevation was adjusted, and the shutter was opened again. ExpowrrS were made at eight different elevations through the flame zone. This procedure was carried out for tank pressures of 6, 8, 10, and 14 inchm of mercury absolute at a propane-to-air mass ratio of 0.08; it was then used at propaneto-air mass ratios of 0.070, 0.0637, and 0.060 at a tank pressure of 14 inches of mercury absolute. The experiment was repeated on succeeding days at lower source strengths. The spectrographic films were developed under carefully controlled conditions to permit a quantitative comparison of the density of exposure at the various wave lengths. They were de-
IRRADIATED PROPANE-AIR FLAMES 1.0
I
I
1
I
I
I
SOURCE STRENGTH = 1022 CURIES
I
I
. ..
I
L
I
I
SOURCE STRENGTH
I
=40CURIES
L
-
I
I
I
SOURCE STRENGTH ~ 7 . 2CURIES
.
TANK PRESSURE 0 -14 in. Hp Ab&. i
A - 1 0 ~
18
18
0 - 8 .
II
01
IO
II
v-
6
01
ELEVATION THROUGH FLAME (arbitrary zero) -INCH
Figure 2.
Emission at 4 3 15 A. from irradiated flames Propane-air mass ratio = 0.08
veloped at constant temperature, agitation, and time in 15 gallons of DK-50 developer (Eastman Kodak Co., Rochester, N. Y . ) ,diluted 1 to 1 with water. The same batch of developer was used for all films to guard against changing developer strength. T h e quantity used was so great that it can be assumed its strength did not change appreciably with use. After development, the films were scanned with a microphotometer and the density of exposure was recorded as a function of wave length. A typical nicrophotometer tracing is shown. Results
Emission at 4315 A. The strongest emission from the flame occurred at 4315 A. and was due to CH. The relative intensity a t this wave length was read from the microphotometer traces and plotted us. elevation through the flat flame at various pressures and source strengths (Figure 2). The propane-air mass ratio for this set of data was 0.08. The maximum spectral intensities with respect to elevation were read from Figure 2 and plotted us. source strength in Figure 3. The data for different propane-air ratios a t a tank pressure of 14 inches of mercury absolute are presented in Figure 4. T h e maximum spectral intensities with respect to eleva-
Table 1.
Strength of Source
Curies Condition Removal from pile Experiment 1 Experiment 2 Experiment 3 Experiment 4 Experiment 5
AulSs
Total
5,670 484 205 111 24 4.8
15,270 1,022 403 207 40 7.2
Aulgs
Hours 0
9,600
538
268 362 429 595 772
198 96 16 2.4
10
09
,
08
Tank Pressure 0 - 1 4 In Hg abs, A - i O i n Hg abs. 0 - 8 in Hg obs.
v-6in Hg abs. I
I
I
06
05
04
0 31
I
I
I
10
100
1000
I 10,000
SOURCE STRENGTH-CURIES
Figure 3.
Effect of radiation on maximum emission at 4 3 15 A. Propane-air mass ratio = 0.08 VOL. 49, NO. 9
SEPTEMBER 1957
1425
-
I0
I
SOURCE STRENGTH
-
09
E
08
:1022
CURIES
I
,
I
I
I
SOURCE STRENGTH = 403 CURIES
I
0
I
I
I
,
I
SOURCE STRENGTH
:40
CURIES
I
I
I
SOURCE STRENGTH = 7 2 CURIES
- 0.08 - 0.0637
A -0.07
2
-eL
I
SOURCE STRENGTH = 207 CURIES
0
07
2I 06 5
z
ez 05 _I
2
04
Y %
03
w
f
O2 01
0 0
002
004
006
008
0
002
004
0 002 0 002 004 006 00%ELEVATION THROUGH FLAME(orbitrory zero)- INCHES
006
Figure 4.
008
004
006
008
0
GO2
004
006
008
Emission a t 431 5 A. from irradiated flames
Tank pressure = 14 inches of mercury absolute
2 5
+
. I
2
I
1
Y
I 10
Figure 5.
parameter. The dara did not permit construction of a plot such as Figure 5, ., shoiving the effect of propane-air ratio.
v
1
I IO0 SOURCE STRENGTH -CURIES
IO00
l0,oQO
Discussion
Effect o f radiation on maximum emission a t 4315 A.
The intensity of emission is observed in Figures 2, 4, and 6 to be almost sym-
Tank pressure = 14 inches of mercury absolute
'
DISTANCE (PROPORTIONAL TO WAVE NUMBER)
'
Microphotometer tracings of spectrographic films supplied relative intensities for CH,
1426
lNDUSTRlAL AND ENGINEERING CHEMISTRY
CZ,and OH
I R R A D I A T E D P R O P A N E - A I R FLAMES 0.20
I
I
I
I
I
SOURCE STRENGTH = 1022 CURIES
Figure
6. Emission at 5 165 and 3063 A. from irradiated flames Propane-air mass ratio = 0.08
metrical about the elevation a t which the maximum emission occurs, indicating that the maximum rate of reaction occurs midway between initiation and completion. The maxima decrease and shift away from the burner as the pressure is decreased. Figure 4 indicates a decrease in intensity of C H as the propane-air ratio is decreased, which is to be expected since the concentration of total carbon decreases. The elevation of the maxima does not appear to be much of a function of propane-air ratio. The scatter in the abstracted points in Figures 3 and 7, and particularly in Figure 5, is due in part to the difficulty in determining the maxima in the otherwise well-defined curves in Figures 2, 4, and 6. Some inprovement might have been obtained if the raw data were first fitted by empirical equations of appropriate form, but this refinement was not believed necessary. From Figures 3 and 7, it is apparent that the maxima at 4315 A. due to C H and a t 5165 A. due to CZincrease with source strength, while the maxima at 3063 A. due to O H are relatively constant. The effect of the radiation appears to increase with pressure, as would be expected since the probability of /3 interaction is greater at greater densities.
To compare the emission due to CH, CZ, and O H , the data for emission a t 4315, 5165, and 3063 A. are replotted us. elevation a t a propane-air ratio of 0.08 and a tank pressure of 14 inches of mercury absolute in Figure 8. The maxima occur a t different elevations.
Despite the scatter in the data in Figure 5, the effect of irradiation is clearly greatest at high propane-air ratios. This too is to be expected, because the probability of effective p interaction is greater with higher concentrations of fuel molecules and radicals.
OH
- 3063 A
-
I
A
I
I
I
'"00
SOURCE
I
I
a 0
I
1
STRENGTH-CURIES
Figure 7. Effect of radiation on maximum emission at 51 65 and
3063 A.
Propane-air mass ratio = 0.08 VOL. 49,
NO. 9
SEPTEMBER 1957
1427
ELEVATION THROUGH FLAME (arbitrary zero) -INCHES
Comparison of emission at
Figure 8.
3063,431 5, and 51 65 A. in irradiated flames
Tank pressures = 14 inches of mercury absolute Propane-air mass ratio = 0.08
The elevations a t which the maxima occur are plotted us. source strength for a propane-air ratio of 0.08 and pressures of 8 and 14 inches of mercury absolute in Figure 9. Contrary to expectations, the maxima due to Cz and OH occur a t about the same elevation, and the maxima due to C H occur a t a higher elevation. This suggests the possibility that carbon monoxide is formed in flames by the reaction C,
+ OH+
CH
+ CO
In view of contrary evidence from previous work, more information is needed before a positive conclusion can be drawn. Acknowledgment
This research was carried out through the Engineering Research Institute of The University of Michigan and was supported by the U. S. Air Force, through Tonk Pressure
8
HQ O b % .
the Office of Scientific Research of the Air Research and Development Command. This article was extracted in part from a previous report (2). Invaluable assistance and advice were provided by the following personnel of The University of Michigan: J. H. Enns, Department of Physics; A. H. Emmons, C. C. Palmiter, and W. R. Dunbar, Radiation Control Service; and R. B. Morrison, T. B. Khammash, R. L. Gealer, R. J. Kelly, R. E. Cullen, M . P. Moyle, and E. T. Howard, Aircraft Propulsion Laboratory.
(3) Emmons, A . H., Phoenix Memorial Laboratory, Ann Arbor. Mich., private communication. (4) Gaydon, A . G., Wolfhard, H. G., "Flames, Their Structure, Radiation, and Temperature," Chapman and Hall, London, 1953. ( 5 ) Lewis, W. B., Phillips Petroleum Go., Idaho Falls, Idaho, private communication. (6) Weir, A , , Jr., IND.ENG.CHESf.45,1637 (1953). (7) Weir, A , , Jr., Morrison, R. B., Univ. Mich. Enq. Research Inst. Rept. 2054-3-F(September 1954). RECEIVED for review October 8, 1956 ACCEPTEDMarch 8, 1957
literature Cifed
Division of Gas and Fuel Chemistry, 131st Meeting, ACS, Miami, Fla., April 1957.
(1) Churchill, S. W., Weir, Alexander, Jr., Gealer, R. L., Kelley, R. J., IKD. ENG.CHEM.49, 1419-22 (1957). ( 2 ) Churchill, S. W., Weir, A , , Jr., Ornella. L. F.. Gealer. R. L.. Kellev. ,, R. J., 'Gluckstein, M. E.; Univ. Mich. Eng. Research Inst. Rept. 2288-6-T, '4FOSR-TN-56-17 (December 1955 ),
Corrections
14 H Q obs
C2-51651 OW-30631
Disti Ilati on Improvement In the article, "Distillation Improvement by Control of Phase Channeling in Packed Columns." by R. E. Manning and M. R. Cannon [IsD. Esc. CHEM. 49, 347 (March 1957j], the present address of R . E. Manning should read: Cannon Instrument Co. State College, Pa. University Park, Pa., applies only to the Pennsylvania State University. Flotation
SOURCE S T R E N G T H - C b R I E S
Effect of radiation on location of maximum emission at
Figure 9. and 51 65 A.
Propane-air mass ratio = 0.08
1428
INDUSTRIAL AND ENGINEERING CHEMISTRY
3063, 4315,
In the Unit Operations Review on Flotation [IND.ENG. CHEM. 49, 496 (1957) ] reference 3D should have cited U. S. Patent 2,747,733.
