H,-D,, and HD-D, the thermal conductivity of any mixture ing from a binary mixture containing 1 mole of hydrogen

e of hydrogen and x
of the three gases can be calculated by means of the numerical moles of deuterium is given by the relations
relations below in which the chemical symbols represent the
mole fraction of each gas in the mixture. Gas Moles Mole fraction
HD 2x/(1 + x) 2x/(1 + x y
\I 0338 H2 1/(1 + x ) 1/(1 + x ) (18)
D* x2/(1 + x ) x2/(1 + xj2
Table I 1 contains the composition of the hydrogen-
deuterium mixtures and their observed thermal conductivities
as plotted in Figure 2. Table I1 also contains the composition
of the ternary mixtures as calculated by the Relations 18.
Finally the experimental thermal conductivities of the equili-
brated mixtures are compared with those calculated from the
curve for E = 4.00 in Figure 3.

29.56 (+)IH, + 0478]

~n \
?e) 10414
+
LITERATURE CITED
(1) Archer, C. T., Proc. Roy. SOC.(London)A165,478 (1938).
(2) Bolland, J. L. Melville, H. W., Tram. Faraday SOC.33, 1316
(1937).
(3) Buddenberg, J . W., Wilke, C. R., Ind. Ens. Chem. 41, 1345
(1 949).
(4) Farkas, A., Farkas, L., Proc.Roy. SOC.(London) A144,46i (1934).
(5) Ibbs, T. L., Hirst, A. A,, Zbzd., A123, 134 (1929).
(6) Lindsav. A. L.. Bromlev. L. A.. Znd. ERP.Chem. 42.1508 (1950).
( 7 ) Van Cieave, A. B., Maas, O., Can. J . R k a r c h 12, 57 (1935)
APPLICATION TO EQUILIBRATED MIXTURES (8) Ibid., p. 372.
If 4.00 is taken as the value of the equilibrium constant in (9) W a d j e w a A., Physik ?. 7,235 (1904).
Equation 1 the composition of the equilibrated mixture result- RECEIVED
for review July 7, 1958. Accepted December 11. 1958.

Effect of Molecular Structure on Burning Velocity

G. J. GIBBS and H. F. CALCOTE
Experiment Inc., Richmond 2, Vo.

T h e burning velocities of various compounds have been in- arevaporized and fed to the burner in the gaseous form ( 5 ) .
vestigated by many observers in an attempt to understand the Air is taken from the laboratory supply and metered with a
mechanism of flame propagation. These data have been ob- sonic-orifice flowmeter. Gaseous fuels are taken directly from
tained by four general methods: the propagation of flame in a tanks and the pressure is controlled with a Moore Products
tube, in a spherical bomb, in a soap bubble, and on a Bunsen Nullmatic pressure regulator; the flow rate is determined with
burner (3, 4, 8, 12, 16, 17, 22, 23, 30, 36). Such data on many a sonic-orifice flowmeter calibrated for each gas. A mixing
fuel types are important in verifying theories of flame propaga- chamber containing two porous stainless steel disks and filled
tion and ignition and the data reported herein have already with glass Raschig rings both aids mixing and prevents flash-
been so employed ( 7 1 , 24, 32). By altering the fuel structure back through the gas-feed tubes. All tubing and valves with
( 6 ) ,the data are made available for examining such considera- which the gas comes in contact beyond the liquid-fuel-injection
tions and for predicting the burning velocity of most new sys- system are Type 316 stainless steel. The burners are seamless
tems. Combustibles for which good burning-velocity data are stainless steel tubing which can be easily exchanged at a flange
already available have also been studied, to demonstrate the on the mixing chamber. Both the air meter and gaseous fuel
degree to which such data, taken in different laboratories and meter are calibrated by a water-displacement method.
by different methods, can be compared. The flame is photographed with a Speed Graphic camera
To carry out this program successfully, it is necessary to having an f 4.7 lens using Ansco SSS-Ortho film, varying ex-
have a simple yet reliable method of calculating the burning posure times from 1/25 to 1/100 second. Shadowgraphs of the
velocityfrom the flame dimensions. Although there are a num- flame are made by photographing the shadow formed by the
ber of methods for determining burning velocities from Bunsen heated gases in the inner cone refracting the light from a 2-watt
burner flame cones, those which have a reasonable theoretical Western Union concentrated arc lamp serving as a point of
basis and give reliable results require somewhat detailed meas- light source. A Speed Graphic camera without a lens is used,
urements or calculations (14, 15, 19, 37, 33, 34). This report and theexposure time (1/15 to 1/25 second) is controlled by a
presents a simple procedure which requires a minimum of focal-plane shutter. The burner is placed between the lamp and
calculations to obtain reliable results with data demonstrating the film, 61 cm. from the lamp and normally 30 cm. from the
the errors involved, a comparison with results from other labor- film. A scale is photographed on each picture for determining
atories, and the burning velocities of 77 compounds showing the magnification factor, M .
the effect of molecular structure on burning velocity. The methodology involved in calculating burning velocities
from the photographs or tracings was derived by considering
APPARATUS AND PROCEDURE the volume rate of flow through a section of the flame surface
The apparatus is basically the same as that used by other over which the burning velocity is essentially constant. This
investigators. Provision is made for controlling and measuring condition is satisfied if the tip and base of the flame are neg-
flow rates olair, gaseous fuels, and liquid fuels. The liquid fuels lected. Then, the burning velocity is equal to the volume rate
of gas flow through a lateral area of the frustum of flame surface
'Present address, AeroChem Research Laboratories, Inc., P. 0. divided by that lateral area. The volume rate of gas flow
Box 12, Princeton, N. J.

226 JOURNAL OF CHEMICAL AND ENGINEERING DATA

-
through such a frustum i s obtained by integrating the linear of the equivalence ratio, 4, the stoichiometric air-fuel ratio ., ,
flow velocity between the radii which define the annular flow divided by the actual air-fuel ratio. This method of determin-
entering the fNStUm of cone considered. This assum& that a ing burning velocities will be referred to as the “apex-cone”
flow stream does not change its direction after leaving the method.
burner port until it enters the combustion region. The burn-
ing velocity, S, is, in terms of the total volume flow rate, V, and VALIDIN OF EXPERIMENTAL TECHNIQUES
the slant height ofthe frustum, h : Before measuring the burning velocity of a series of fuels, one
must examine both the possible factors which may cause errors
S” = (Y/h)k (1 1
and the ability to check results obtained in other laboratories.
liameters, d : For any method to be of value, the burning velocity should he
independent of burner diameter, total laminar-flow velocity,
and what one chooses as the flame surface, because all of the
flame must be propagating at the same velocity relative to the
a given burner diameter, D: unburned gas mixture. Thus burning velocities should be the
DZ - c, 2 1 ~ ~ 4
same when measured on shadowgraphs or luminous cones if
= (3) proper considerations are made.
wgraph is sharp (Figure l ) , simpli- Burner Diameter. The effect of burner diameter is demon-
I ~ C ~ L I V L I 1 3 ~ J W K u y cuuruering the cone from the point where strated in Figure 2 for four different burner diameters between
the sides begin to curve to the apex ( d , then becomes equal 0.452 and 1.737.cm., a ratio of diameters of 3.8 to 1. No change
to 0), and Equation 2, for k, becomes in burning velocity is observed over this range. At least two
different burner diameters have always been employed to assure
that for the particular fuel the burner diameter had no effect.
Total Flow Rate. The effect of total flow rate on the burning
The calculations have been further simplified for a given burner velocity is demonstrated in Table 1. The only appreciable devi-
by setting up tables of k values for all necessary values of d,. ation occurs at the very low flow rate of 31 cc. per second, where
Then, to calculate the burning velocities, one needs only to the flame is extremely small. Doubling the flow rate, 62 to 122
place V, k, and h in Equation 1. cc. per second, gave no deviation beyond the experimental
The diameter and the slant height are measured from photo- error. Since all of the flame cones outside of this range (on the
graphic enlargements and the actual sizes obtained from the size of burners employed) are either so tall or so short as to be
magnification factor, M. When a cone is taken from shadow- obviously poor subjects for study, there is little chance for error
graphs, the negatives are projected in an enlarger on a sheet due to flow velocity. The data were taken over a period of sev-
of paper where lines are drawn parallel to the sides. Measure- eral days, hence show the reproducibility to be expected.
ments are made from this tracing without the necessity of Film to Flame Distance. Other observen (7, 2, 27) agree that
making a print. The burning velocities are plotted as a function the outer edge of the shadow cone-i.e., the outer edge of the
light area on the shadow negative-is a true image of the reac-
Figure 1. Luminous cone tion zone and that extrapolation to zero distance between the
photographic plate and the flame shows that the outer cone and
and shadowgraph of
the sharp inner cone coincide, and that the outer cone is the
Bunsen burner flame proper place to make burning-velocity measurements because
nPentane-air flame at 25’ C.; this image is not affected by refraction of the light as is the cone
Burning velocily, 42 cm./rec.; defined by the inner black-white edge. The real objection to
Volumetric flow rate, 124 cc./sec.: using the outer cone, however, lies in the inability to make ac-
Burner diometer. 1.10 cm.; curate, reproducible measurements because of the lack of sharp-
.# = 1.00 ness. At a film-flame distance of 30 cm., using the apex-cone
LYUlNOUS I*IDDWG*LP” method on the inner and outer cones of randomly selected flame
shadowgraphs with varying fuels and fuel-air ratios, burning-
velocity values are obtained which compare favorably. Sixty-
Effects on Burning Velocity
nine different shadowgraphs of 14 fuels and fuel mixtures at
I I I various air-fuel ratios and initial temperatures were compared.

? U T E O t O I l l O E ?WW W l Y I I W L R . I E c+ TLa*E, em Figure 4. Tracings of ethane-

Figure 2. Shadowgraphs of propone-oir flamer Figure 3. Ethane-airot4 = 1.02; Total flow, air flames (+ = 1.02) a t film
(apex~onemethod) 88 cm./rec.; Burner diameter, 1.10 cm. to flame distances. L

VOL 4, No. 3, JULY 1959 227

 reported by Grove, Hoare, and Linnett (27). There are in-
Table I. Effect of Total Flow Rate on the Burning sufficient data in the latter article to explain this difference. The
Velocity of Propane-Air Flames results from the total-area and apex-cone methods are also re-
(Apex-cone method)
ported. When the outer black boundary was used, the burning
Burner diam: 0.766 cm. velocity did not vary with film-flame distance; then the burning
Flow Burning Velocity, velocity was 46.2 cm. per second. The burning velocity at
Cm. / Sec. Av. % 4 = 1.02 from the curve of burning velocity us. equivalence
Equivalence Rate, yoDev. Dev. from ratio obtained by the standard procedure was 45.2 cm. per
Ratio @ Cc./Sec. Measured Mean from Mean Mean
second.
0.922 31.0 (49.98)o (+ 17.57) The results of Figure 3 are easily understood by reference to
31.0 (44.87) (+ 5.55) Figure 4. The shadowgraphic tracings of the same ethane-air
62.3 41.36 - 2.71 flame are presented at three different film-flame distances. The
62.3 41.51 - 2.35 inner three lines are for the black-white boundary. The outer
83.0 43.21 + 1.65 line is for the outer black boundary. Inspection will show that
83.0 43.98 42.51 + 3.46 2.54 any method based upon the outer black boundary will give re-
0.997 31.0 (51.59) (+ 12.74)
sults independent of film-flame distance as pointed out by
62.3 45.60 - 0.35 Grove, Hoare, and Linnett. Results based upon the black-white
62.3 46.18 + 0.92 boundary (inner three lines) will depend upon the method of
83.0 46.68 + 2.01 analysis. Any method based upon the angle the flame surface
122.0 44.58 45.76 - 2.58 1.46 makes with a flow line should give a decrease in burning ve-
locity with increasing distance. Any frustum method should
1.18 62.3 43.09 - 3.83
62.3 44.53 - 0.47
122.0 44.57 - 0.38 Table 11. Calculation of Burning Velocities by the Frustum
122.0 45.28 + 1.21 and Total-Area Method
122.0 46.21 44.74 + 3.29 1.84
(Luminous cone, butane-air mixture)
Av. errorb = 2.08% Maqnification, 10.5. Flow rate, 85.8 cc./sec
Burner diameter, 0.766 cm. @ = 1.05
il Figures in parentheses not included in determining the averages.
Includes experiments not recorded in table. Slant Burning
Cone Diameter, Height, Area, A , Velocity,
The average deviation of the outer cone value from the inner Section" d,, Cm. Cm. Sq. Cm. S, Cm./Sec. S X A
cone value was only 3.957'. The average deviation when the 1 8.80 1.11 0.2435 17.38 4.233
outer cone value exceeded the inner cone value was 5.87%, 2 7.60 1 .oo 0.2120 28.01 5.937
and when it was less, 2.04%. 3 6.88 1.68 0.308 36.06 11.107
The burning velocity reported herein, decreased with in- 4 5.99 1.68 0.2638 38.44 10.138
creasing film to flame distance, while others ( 7 , 27) found an 5 5.30 3.80 0.5035 41.06 20.672
increase with distance. This led to a closer examination of the 6 4.00 3.80 0.3687 46.03 16.971
7 2.81 3.21 0.2104 52.07 10.956
results. Figure 3 shows the data obtained by analyzing a series 8 1.79 2.60 0.0997 59.32 5.913
of shadowgraphs by different methods, all employing the inner 9 0.90 0.50 0.0113 78.15 0.880
black-white edge. I n the top curve the burning velocity at each 10 0.68 0.40 0.0039 302.4 1.170
point was obtained by a weighted average according to surface 0.00
areas for seven to nine frustums excluding the bases. I n the sec- Total area = 2.225 87.977
ond curve the burning velocities from four to five frustums, ex-
87.98
cluding the tips and bases, were averaged. The next curve was Area weighted average of burning velocity = - = 39.54 cm./sec.
obtained by using the angle method of Sherratt and Linnett 2.225
(37), which gives results in the opposite direction from those "From base of cone to apex.

Table Ill. Comparison of Methods of Calculating Burning Velocities

(Shadowgraph, butane-air flame)
Magnification, 15.67 Flow rate, 85.8 Cc./Sec. Burner diameter, 0.776 Cm. 4 = 1.05
Vertical Slant Linear Frustum Burning Velocity, Cm./Sec.
Cone Diameter, Height, Height, Flow Rate, Area: S and L's Frustum Angle
Section5 d,, Cm. Cm. Cm. Cm./Sec. Sq. Cm. Sin ab Method' Method Method
1 11.50 1.41 1.60 79.84 0.2194 0.3838 37.85 38.04 30.64
2 9.98 1.30 1.35 141.89 0.1637 0.3507 49.96 51.86 49.76
3 9.00 1.11 1.20 179.64 0.1334 0.2703 55.45 53.38 48.56
4 8.28 1.58 1.60 21 1.28 0.1605 0.2507 55.55 55.59 52.97
5 7.43 1.80 1.80 241.91 0.1618 0.2007 53.68 54.00 48.55
6 6.61 1.71 1.80 265.21 0.1455 0.1729 47.45 44.44 45.86
7 6.00 2.31 2.30 285.64 0.1652 0.1590 48.84 48.91 45.42
8 5.21 3.45 3.49 309.63 0.208 1 0.1561 48.83 47.93 48.33
9 4.11 3.30 3.39 330.60 0.1571 0.1534 47.57 45.83 50.71
10 3.15 3.61 3.68 344.94 0.1233 0.1420 50.02 49.27 48.98
11 2.09 3.45 3.48 356.22 0.0732 0.1420 44.96 44.44 50.58
12 1.21 4.75 4.8 361.99 0.0370 0.1420 46.05 44.85 51.40
0.00
Total area, A , = 1.748 Sa", = 48.85 48.21 47.65
Total area method, S = V/AT = 49.07
'From base of cone to apex.
'Sin a measxed by drawing a tangent to curve for angle method
'Sherratt and Linnett, (37).

228 JOURNAL OF CHEMICAL AND ENGINEERING DATA

give a decrease in burning velocity with increasing distance be-
cause the area of the frustum increases with distance. This is Table IV. Burning Velocity of N-Butane-Air Flames
shown by drawing vertical lines for flow lines (Figure 4) which by Different Methods
define the frustum diameters; the slant height increases with 4 . Shadowgraph B. Luminous Cone
increasing distance. The frustum method on the outer black
6 = 1.05 Q = 1.16 6 = 1.29
edge and the black-white boundary will give the same result
when these two edges are parallel; the outer cone is simply con- Method A B A B A B
sidered to be displaced downstream. Because the total area of Frustum, av. of 3 46.9 47.9 42.4 44.1 37.5 35.0
the three inner tracings decreases with increasing distance, the Frustum, top only or
burning velocity should increase for any total-area method apex cone 47.0 ... 42.8 ... 35.4 ...
Frustum, weighted height 47.4 49.8 47.3 45.6 39.7 39.3
which utilizes the shadowgraph technique. Compare the areas
Method of Sherratt
2.225 sq. cm. with 1.748 sq. cm. (Tables I1 and 111) for evi- and Linnett“ 47.6 .,. ... .,. ... ...
dence that the total surface area is less for the shadowgraph Angle method 47.6 ... ... ... ... ...
than for the luminous cone. The total-area method assumes that Frustum, av. 48.2 ... 49.0 .., 41.6 . .,
the apparent increase in burning velocity at the apex of the Frustum, weighted area 48.2 39.5 49.1 38.7 41.3 33.1
luminous cone and the decrease in burning velocity near the Total area, S = V ’ A , 49.1 38.6 48.4 37.7 42.4 33.2
base cancel each other, which is not the case. The area at the
base with low burning velocity is considerably larger than the ‘’ Averaged from cone sections 6 to 12 in Table 111.
area at the apex where burning velocity is high (Table 11)
There is no appreciable change in burning velocity near the
apex of a shadowgraph (Table 111), although the burning suggest that the “maximum burning velocity for propane-air
velocity is reduced near the base. The frustum method also mixture is 44 1 cm. per second when the percentage of pro-
gives a low value for luminous cones when a large number of pane present is 4.3 f 0.15’’ ($ = 1.02). This value was ap-
frustums are averaged. parently determined from an average of the maximum values of
Methods of Data Analysis. Four different methods of analyz- the selected curves, that of Andersen and Fein being obtained
ing a single cone have been applied to the data in Table 111. at 25’ C. with 0.07 mole yoof water in the fuel-air mixture and
The agreement among the methods is rather good. However, Gray’s being obtained at 11 C. with the moisture content not
O

Table I V demonstrates that by considering only a large top reported. The data given in Table 3 of Gray’s article (20) for
frustum, actually a cone of a shadowgraph, values may be ob- the schlieren “slit” procedure at 11 C. and 1 .O atm. coincide
O

tained which are comparable with those obtained by any of the with the authors’ data which were obtained with 3.1 mole yoof
acceptable methods either on shadowgraphs or luminous cones. water in air at 25” C. and 1 .O atm. If Gray’s fuel-air mix-
Analysis of randomly selected shadowgraphs for seven com- ture was relatively dry, a correction for initial temperature
pletely different fuels at various initial temperatures and air- would raise their curve to coincide with the authors’ 0.31 mole
fuel ratios curroborated this statement. Therefore, it is possible 70of water in air curve.
to obtain reliable results by the apex-cone method. Ethylene-Air Flames and Effect of Water. Further comparison
Propane-Air Flames and Effect of Water. The results for pro- with Linnett and his associates is illustrated in Figure 6 for
pane-air flames measured by different methods in different ethylene-air. Three methods were employed by Linnett to ob-
laboratories are summarized in Figure 5. Those obtained in this tain the data shown. Pickering and Linnett (29)used a modified
work contained normal laboratory air (0.31 mole yo of water angle method on shadowgraphs which neglected the burning
in air) at 25°C. and air saturated with water (3.1 mole 7oof velocity near the apex and base of the cone. These values were
water in air). The data were obtained on various sizes of therm- taken from the article of Linnett and Hoare (26) (which were
ally jacketed burner tubes at widely varying flow rates using corrected by extrapolation to zero film-flame distance). The
the apex-cone method and shadowgraph technique. Results other values are those obtained by Conan and Linnett ( 9 ) using
were checked in a closed system and in open room air. The the angle method on schlieren photographs and those of
curves of Singer (33) were obtained on a cylindrical and a slot Linnett, Pickering and Wheatley (27) using schlieren photog-
burner tube using air dried over calcium chloride and calcula- raphy and the soap-bubble technique. Strehlow and Stuart
tions made using. Dery’s truncated-cone method described in (35),employing a soap-bubble technique, recorded data agree-
reference (25). The slot-burner data agreed with results ob- ing with those of Linnett and associates. The three methods are
tained in a spherical bomb (28). The curve of Andersen and compared with the two ethylene-air curves obtained by the
Fein is a composite of two curves (I), the lean side of which authors with 0.31 and 3.1 mole 70of water in air. Linnett and
was obtained using the total-area method on shadowgraphs, Hoare have stated their initial temperature to be 16” + 2”C.,
extrapolated to zero film-flame distance, and the rich side ob- but Conan and Linnett ( 9 ) and Pickering and Linnett (29) did
tained using stroboscopically illuminated particle tracks. They not specify the initial temperature. No reference was made in
used a Mache-Hebra nozzle at 25“ C. with less than 0.07 mole either case to the moisture content of the gaseous mixture.
yoof water in the fuel-air mixture. Linnett and Hoare (26) have studied flame propagation along
Gray, Linnett, and Mellish (20) have compared the burning horizontal tubes, reporting a maximum burning velocity of
velocities of propane-air mixtures obtained by eleven different 69.7 cm. per second at $ = 1 05 (6.87 volume yo of ethylene).
observers. Figure 3 of their article illustrates the variety of data The maximum burning velocity is in agreement with the values
which exists on burning velocity. The highest maximum burn- obtained using the burner-tube and soap-bubble methods; how-
ing velocity shown in that curve is 1.5 times as great as the ever, the concentration of ethylene at the maximum is lower.
lowest maximum value. Linnett and associates have discussed They were unable to offer any explanation for this difference
the various methods of determining burning velocities, the but have pointed out that the maximum in the curve of burning
reasons for data being high and low, etc., but they do not men- velocity us. composition for propylene-air flames was at the
tion the effects of initial mixture temperature and moisture in same percentage for both tube (78)and burner methods (20);
the gaseous mixture. Unfortunately, many observers fail to re- however, for propane-air flames all burner methods (20) gave
cord these two very important parameters. Before a usefui value the maximum burning velocity at the same percentage, but the
for the burning velocity of any fuel can be declared, the initial maximum in the curve obtained by the tube method (78) was
temperature, pressure, moisture content, and fuel composition at a higher percentage. The data from the Bureau of Mines
must be stated. Gray, Linnett and Mellish (20), “after con- were obtained on a slot burner and agreed with spherical-bomb
sidering the various methods,” decided that their curve and results. Dugger (70) reports a maximum burning velocity for
that of Andersen and Fein (Figure 5, this article) “are ob- ethylene of 64.0 cm. per second (4 = 1.16) at 2 5 ° C . using
tained by methods least likely to lead to serious error.” They the Bunsen burner technique.

VOL. 4, No. 3, JULY 1959 229

 I I I I I I I I 1

m-
IeoC -I
eo-

a
g $0-
3
Y

A '

OB 09 IO II 12 13
i Figure 6. Ethylene-air flamer, 2 5 " C. FDIIIlLElbX RATIO
EWIVALEWE RATIO

Burning Velocities Figure 7. Acetylene-air flames

Figure 5. Propane-air flames, 25' C.
METHOD AND CONDITICNS

CALDWELL, BRolDA,)ND WVER 0 OB %H@

IN 'IXTURE 7 ROUND BURNER NOZZLE,
),SCHLIEREN

Y
I TUBE, DERY'S
N E METHOD,
c'0 +LIEREN

TUBE,
" IOD.
z
c
a

' I ..., " DADCP nxzu.n

- "IN'-"T
AIR
ROUND BURNERS.
I APEX-CONE METHOD,
SHADOWGRAPH

I I I I I Figure 8. Methane-air flames

08 09 10 I / 12 I3 14
EOUIVALENCE RATIO

Acetylene-Air. A comparison of the Bunsen burner, the fall on the curve and acetylene above it. The relationship is
flame-tube, and the soap-bubble methods is offered in Figure 7 . linear, the maximum burning velocity occurring slightly on the
The agreement for burning velocities is not too unsatisfactory rich side for the tube data. Figure 10 is a correlation at max-
considering the difference in conditions. However, the differ- imum burning velocities as obtained by the two laboratories;
ent equivalence ratios at which the maximum burning velocity the tube method gives lower values.
occurs is somewhat surprising. The Bunsen burner methods Singer (33) has recently compared burning yelocities deter-
give maxima at the same composition, while the soap-bubble mined on cylindrical and slot burners and in a spheriFal bomb.
and tube methods agree with respect to the composition for He has demonstrated by analysis and comparison of Bunsen
maximum burning velocity. This difference might be due to the flames from both cylindrical and slot burners that lower burn-
methods, except that Friedman and Burke (73) employing the ing velocity results when the latter method is used, owing to re-
tube method found the maximum flame velocity at $ = 1.25 moval of the effect of curvature of combustion surface as pre-
in agreement with the authors' value dicted by theory (25). Furthermore, he has found agreement
Methane-Air. Another comparison is made in Figure 8 ; this between the slot-burner method and the spherical-bomb
time for methane-air flames. The most outstanding feature of method which is within experimental error. Preliminary studies
this comparison is that the maximum burning velocity occurs on slot burners using a propane-air flame at Experiment Inc.
at approximately the same equivalence ratio, whereas for most are in very close agreement with the results reported by Singer.
other fuels different observers obtain widely scattered maxima. Although questions still remain about correct methods for ob-
Caldwell, Broida, and Dover (7) have obtained a slight increase taining "absolute" values of burning velocities, the method em-
in equivalence ratio at which the burning velocity is a max- ployed in this report is useful in determining effects due to
imum by the addition of water vapor. The burning velocity de- changes in molecular structure.
creased. The two curves shown in Figure 8, which were ob- RESULTS AND DISCUSSION
tained by Singer (33), illustrate the effect of using slot burners
as compared with round burners. The truncated-cone method The results are summarized in Tables V and VI for a range of
of Dery (25)was employed in both cases at 25" C. and 1.0 atm. equivalence ratios at initial temperatures of 25" and 100" C.
The curve of this article was obtained as explained for other These data have been taken from an averaged curve through
fuels. experimental points with an accuracy as indicated in the previ-
Comparison with Flame Propagation in a lube. The present ous section. In each experiment the equivalence ratio incre-
burning velocity data obtained by the Bunsen burner method ments were less than 0.1 unit. The averaged curves are pre-
and the flame speed of flame propagating in a tube are com- sented in Figures 11 through 25 for the 25" C. data and Fig-
pared with those of Gerstein, Levine, and Wong at NACA (78) ures 26 through 32 for the 100°C. data. Fuels are grouped
in Figure 9, where volume per cent fuel at maximum flame ve- according to molecular structure for purposes of comparison.
locity from NACA (78) is plotted against that obtained in this
laboratory. If this line were extended to 10yc, methane would (Text contrnued on page 235)

JOURNAL OF CHEMICAL AND ENGINEERING DATA

 I I I I I I BO I I I I I I

EIYILENE
/
-z
E T H Y L E HE

7.0 PII)P"HE

0
G ETMANE
Pg 6.C 3
m
ZLX
PRWlNf a
W z
3
b.

$ 5.c
z
-.
B
v)

2 4.c
.
I
3
W

3
Y 3.c
3
9 2-YETMILBUTANE I I 1 I 1
30 40 50 60 70 eo
2.c MAXIMUM BURNING VELOCITY ON A BUNSEN BURNER, cm./ snc.

I I I I I 1 Figure 10. Maximum burning velocities obtained on a Bunsen

2.0 3.0 40 5.0 60 7.0
burner and in a tube
V O L U M E % F U E L AT ~ . , I E X P E R l M E N T INC., B U N S E N BURNER )

Figure 9. Compositions at which maximum burning velocity

occurs on a Bunsen burner and in a tube

Burning Rates at
25" C.

20 30 60 SO
EWIV&ENCE RATlO

Figure 1 1 . Normal alkane hydrocarbons Figure 12. Acetylenes and hydrogen

so I I I I I

'O-

f
$
m-

20 I I I 1 ,
06 oa to 12 14 16 in 20
os on IO 12 14 16 in
EWlVALENE RbTlO EWlVbLEHCE RbllO

Figure 13. Ethylenes Figure 14. Branched hydrocarbons

VOL. 4, No. 3, JULY 1959 23 1

 Burning Rates at 25' C.

T-l---- fl

20 I I I I I

I I I I I 20 I I I I I
08 IO I2 I. 1s 06 08 IO I2 I4 16
EWiVbLfNCE RPTiO EWIVALENCE RATIO

Figure 15. Branched hydrocarbons and Figure 17. Secondary substituted

tetromethylsilane hydrocarbons

I I I I I

I I I I I
I I I I I

I I I I I
6 08 LO 12 14 I6
EWIVLLENCE R A T 0

Figure 19. Ketones and esters 1 I I I I

, OB 10 12 I. I6
EOYIVLLENCE RATW

Figure 20. Ethers

EOUIVALENCE RATIO

Figure 18. Aldehydes

20
06 08 10 12 14 16 I6
EWIVUENCE RATCO

Figure 22. Tetramethylethoxysilane,

EOUIWLENCE RAT0 di-fert-butyl peroxide,
Figure 21. Ethers and sulfides and triethylamine Figure 23. Three-membered rings

JOURNAL OF CHEMICAL A N D ENGINEERING DATA

232
 Burning Rates at 100" C.

I I I 1 I I

i
a
50-
Y

40 -

I
C~<"*-U~.Ulz-"*~~-C,

3 I I I I I
06 01 IO 12 14 16
EQUIVALENCE milo
YI I Figure 25. Six-membered rings
I I I / 1 I YI E M V A L E W RATIO
6 OB 10 12 I4 16
E W N A L C W 1)ATIO
Figure 26. Hydrocarbons
Figure 24. Five-membered rings

I I I I I 1

45 I 1 I 1 I

40 I I I \ I 1
06 08 10 12
EWIVbLENCE RATIO
I* 16
20
06
1
01
I
IO
I
I2
I
I4
I
I&
1
Id
os 01 10 I2 14 I6
E M W L W RATIO Figure 28. n-Butyl ether and diethyl Cellosolve EUJIULLNCE R A T 4

Figure 27. Alcohols and furfural Figure 29. Ethyl nitrate, ethyl nitrite,
and nitroethane

I t 1

40
Ob
1
08
I
lo
I
12
I
~4
I
16
I
10
das I
on
I
10
I
LZ 14
I I
16
I
11
E W N I U N C E RAT0 EOUIVALENCL R A T D

Figure 31. Three-membered ring compounds Figure 32. Hydrazine and carbon disulfide

Figure 30. Pyrrole, dimethylformamide,

and aniline

VOL 4, No. 3, JULY 1959 233

Effect of Molecular Structure
on Burning Velocities, 25" C.

t
80 155 I I I 1 I I
C k
C.C
W C
50 c-o*_- i

C.
,C-NHI
C
C-C-C-CI
C.
,c-c,
C

Figure 34.
Substituted
hydrocarbons

60
- - c-c-c
0
7
c-si-c
'n
c
n-s-n

C - C - C I O\ n

1
30
c-s-c

Figure 35. Com-

pounds containing
I
c-c>*;
0
c-c-c
" -0

c-;-c-c 9
C-C-H.C

P
c-c-0-c-c

oxygen, sulfur,
nitrogen, or
silicon Figure 36. Alde-
hydes, ketones,
and esters

Figure 37. Three-membered ring

compounds Figure 38. Five- and six-membered rings

JOURNAL OF CHEMICAL AND ENGINEERING DATA

234
 Appearance of flames. Aniline. The inner cone was blue Compounds with insufficient vapor pressure to be studied at
with a reddish yellow outer cone. At @ = 1 a yellow tip ap- 25” C were examined at 100”C with enough repeats to make
peared at the apex of the inner cone and became larger for comparisons (Figures 39 and 40). The results are consistent
richer mixtures. withobservations made at 25°C. Pinene shows no change in
Ethylamine. A yellow tip appeared at approximately @ = burning velocity from the alkane containing the same number
1.25 of carbon atoms.
Pyrrole. At @ = 1.25 a yellow tip appeared on the apex of To predict the burning velocity of any specific substance,
the inner cone. At @ = 1.35 three distinct cones appeared; a comparison of its structure with these examples should be
from inner to outer these were blue, dark blue, and greenish fruitful. However, before any more detailed analysis of the
blue with a yellow tip on the inner cone. molecular structure effects can be made profitably, the thermo-
dynamic. energy and mass transport, and chemical-kinetic
Thiophene. The inner cone was blue with a deep purple properties of the systems involved must be considered. Such an
outer cone which increased in intensity for richer mixtures. At
@ = 1.25 multiple cones began to appear; from inner to outer, analysis should help in unraveling the intricate relationship
these were blue, white, purple, and blue white with a yellow between the above-mentioned principles and make available a
tip just above the inner cone. better understanding of the mechanism of flame propagation.
The effect of molecular structure on the maximum velocity ACKNOWLEDGMENT
is presented graphically in Figures 33 to 40. The maximum The authors acknowledge the contributions of the following,
burning velocity in centimeters per second is recorded on the who at one time or another participated in this program: W. G.
ordinate, and the abscissa indicates structural changes and gen- Atkins, C . B. Barnett, R. B. Huettel, M. R. Irby, and A. R.
erally the number of carbon atoms in the molecule. O n each Sharpe, Jr.
figure a horizontal line has been drawn to indicate the burning
velocity of the normal alkane hydrocarbon. LITERATURE CITED
Hydrocarbon burning velocities fall in order: alkynes > Andersen. J . W., Fein, R. S., J . Chem. Phys. 17, 1268-73 (1949);
alkenes > alkanes. Conjugation generally increases the burning 18,441-3 (1950).
velocity; chain lengthening and branching decrease the burn- Ashforth, G. K., Fuel 29,284-7 (1950).
ing velocity, but structural alterations become less effective as Badin, E. J., Stuart, J . G., Pease, R. N., J . Chem. Ph.ys. 17,
the chain length is increased (Figure 33). 314-6 (1949).
Negative substituent groups decrease the burning velocity. BartholomC, E., <.Elektrochem. 53, 191-6 (1949); NaNalurw’ssen
In order of increasing effectiveness they are alcohol, mercaptan, schaflen 36, 171-5, 206-13 (1949).
amine, and chloride, which is about equal to the epoxy group, Calcote, H. F., Anal. Chem. 22, 1058-60 (1950).
and is more effective than a double bond-z.e., allyl chloride Calcote, H. F., Gregory, C. A., Jr., Barnett, C. M., Gilmer,
burns considerably slower than propylene, but only slightly R . B., Ind. Eng. Chem. 44,2656-62 (1952).
Caldwell, F. R., Broida, H. P., Dover, J . J.,Ibid., 43, 2731-9
slower than n-butyl chloride because the effect of molecular (1951).
alterations becomes less important as the chain length is in- Clingman, W. H., Brokaw, R. S., Pease, R. N . , “Fourth Sym-
creased (Figure 34). posium (International) on Combustion,” pp. 310-13, Williams &
Silicon in the molecule increases the burning velocity (Figure Wilkins Co., Baltimore, 1953.
35). The sulfide group (Figures 35 and 38) and nitroger, (Fig- Conan, H. R., Linnett, J . W., Trans. Faraday SOC. 47, 981-8
ures 34, 35,37, and 39) generally decrease it, excepting in ethyl (1951).
nitrate (Figure 40). Dugger, G . L . , J . Am. Chem. Soc. 72,5271-4 (1950).
Peroxide, aldehyde, and ether groups (especially in cyclics) Fenn, J . B., Calcote, H. F., “Fourth Symposium (International)
have high burning velocities, but ketones and esters have some- on Combustion,’’ pp. 231-9, Williams & Wilkins Co., Baltimore,
what lower burning velocities than the parent hydrocarbon 1953.
(Figures 36, 38, and 40). Fiock, E. F., “Physical Measurements in Gas Dynamics and
Combustion,’’ Part 2, pp. 409-38, Princeton University Press,
Three-membered ring compounds except ethylenimine, Princeton, N. J., 1954.
which contains nitrogen, have higher burning velocities than
Friedman, R., Burke, E., I d . Erg. Chem. 43,2772-6 (1951).
the corresponding alkane, the oxygen-containing ring being Garner, F. H., Long, R., Ashforth, G. K., Ful28,272-6 (1949);
particularly effective (Figure 37). Five- and six-membered 30,17-19 (1951).
rings as well as aromatic structures have little effect on the Garner, F. H., Long, R., Thorley, B., “Fourth Symposium (In-
burning velocity relative to the linear hydrocarbon containing ternational) on Combustion,” pp. 386-92, Williams & Wilkins
the same number of carbon atoms. (Figure 38). CQ.,Baltimore, 1953.

1 )
CiC-C-OH

90

/1
C-C-0-NOe

-
G-C-C-C
\ I
- C-OH
GC-C-C-0
C-C-0-C-C-0-C-C
-?-C-C-C
-
0
C-C-C-OH C-C-C- C-N 0 c-c-c-c
c- c- c-c c-C-c-c-c-c-c

C-C-NO1
F -
C-C-C-C-C
c c
50t c. 1
Figures 39 and 40. Maximum burning velocities at 100 C.

VOL. 4, No. 3, JULY 1959 235

 Table V. Burning Velocities of Various Fuels at 25" C. Air-Fuel Temperature
(0.31 mole-7' H?O in air)
Burning velocity, S, as a function of equivalence ratio, 4, in cm sec.
Q = 0.7 0.8 0.9 I.o 1.1 1.2 1.3 1.4
Saturated hydrocarbons
Ethane 30.6 36.0 40.6 44.5 47.3 47.3 44.4 37.4 47.6 1.14
Ropane 42.3 45.6 46.2 42.4 34.3 46.4 1.06
n-Butane 38.0 42.6 44.8 44.2 41.2 34.4 25.0 44.9 1.03
Methane 30.0 38.3 43.4 44.7 39.8 31.2 44.8 1.08
n-Pentane 35.0 40.5 42.7 42.7 39.3 33.9 43. 1.05
n-Heptane 37.0 39.8 42.2 42.0 35.5 29.4 42.8 1.05
2, 2, 4-Trimethylpentane 37.5 40.2 41 .O 37.2 31 .O 23.5 41.0 0.98
2, 2, 3-Trimethylpentane 37.8 39.5 40.1 39.5 36.2 40.1 1.00
2, 2-Dimethylbutane 33.5 38.3 39.9 37.0 33.5 40.0 0.98
Isopentane 33.0 37.6 39.8 38.4 33.4 24.8 39.9 1.01
2, 2-Dimethylpropane 31 .O 34.8 36.0 35.2 33.5 31.2 36.0 1.10
Unsaturated hydrocarbons
Acetylene 107 130 144 151 I54 154 152 155 1.25
Ethylene 37.0 50.0 60.0 68.0 73.0 72.0 66.5 60.0 73.5 1.13
Propyne 62.0 66.6 70.2 72.2 71.2 61 .O 72.5 1.14
1, 3-Butadiene 42.6 49.6 55.0 57.0 56.9 55.4 57.2 1.23
n- 1-Heptyne 46.8 50.7 52.3 50.9 47.4 41.6 52.3 1.00
Propylene 48.4 51.2 49.9 46.4 40.8 51.2 1.00
n-2-Pentene 35.1 42.6 47.8 46.9 42.6 34.9 48.0 1.03
2, 2, 4-Trimethyl-3-pentene 34.6 41.3 42.2 37.4 33.0 42.5 0.98
Substituted alkyls
Methanol 34.5 42.0 48.0 50.2 47.5 44.4 42.2 50.4 1.08
Isopropyl alcohol 34.4 39.2 41.3 40.6 38.2 36.0 34.2 41.4 1.04
Triethylamine 32.5 36.7 38.5 38.7 36.2 28.6 38.8 1.06
*Butyl chloride 24.0 30.7 33.8 34.5 32.5 26.9 20.0 34.5 1.00
Allyl chloride 30.6 33.0 33.7 32.4 29.6 33.8 0.89
Isopropyl mercaptan 30.0 33.5 33.0 26.6 33.8 0.94
Ethylamine 28.7 31.4 32.4 31.8 29.4 25.3 32.4 1.00
Isopropylamine 27.0 29.5 30.6 29.8 27.7 30.6 1.01
n-Propyl chloride 24.7 28.3 27.5 24.1 28.5 0.93
Isopropyl chloride 24.8 27.0 27.4 25.3 27.6 0.97
n-Propyl bromide No ignition
Silanes
Tetramethylsilane 39.5 49.5 57.3 58.2 57.7 54.5 47.5 58.2 1.01
Trimethylethoxysilane 34.7 41 .O 47.4 50.3 46.5 41 .O 35.0 50.3 1.00
Aldehydes
Acrolein 47.0 58.0 66.6 65.9 56.5 67.2 0.95
Propionaldehyde 37.5 44.3 49.0 49.5 46.0 41.6 37.2 50.0 1.06
Acetaldehyde 26.6 35.0 41.4 41.4 36.0 30.0 42.2 1.05
Ketones
Acetone 40.4 44.2 42.6 38.2 44.4 0.93
Methyl ethyl ketone 36 0 42.0 43.3 41.5 37.7 33.2 43.4 0.99
Esters
Vinyl acetate 29.0 36.6 39.8 41.4 42.1 41.6 35.2 42.2 1.13
Ethyl acetate 30.7 35.2 37.0 35.6 30.0 37.0 1.00
Ethers
Dimethyl ether 44.8 47.6 48.4 47.5 45.4 42.6 48.6 0.99
Diethyl ether 30.6 37.0 43.4 48.0 47.6 40.4 32.0 48.2 1.05
Dimethoxymethane 32.5 38.2 43.2 46.6 48.0 46.6 43.3 48.0 1.10
Diisopropyl ether 30.7 35.5 38.3 38.6 36.0 31.2 38.9 1.06
Thio ethers
Dimethyl sulfide 29.9 31.9 33.0 30.1 24.8 33.0 1.00
Peroxides
Di-lerl-butyl peroxide 41.0 46.8 50.0 49.6 46.5 42.0 35.5 50.4 1.04
Aromatic compounds
Furan 48.0 55.0 60.0 62.5 62.4 60.0 62.9 1.05
Benzene 39.4 45.6 47.6 44.8 40.2 35.6 47.6 1.00
Thiophene 33.8 37.4 40.6 43.0 42.2 37.2 24.6 43.2 1.03
Cyclic compounds
Ethylene oxide 57.2 70.7 83.0 88.8 89.5 87.2 81.0 73.0 89.5 1.07
Butadiene monoxide 36.6 47.4 57.8 64.0 66.9 66.8 64.5 67.1 1.24
Propylene oxide 41.6 53.3 62.6 66.5 66.4 62.5 53.8 67.0 1.05
Dihydropyran 39.0 45.7 51.0 54.5 55.6 52.6 44.3 32.0 55.7 1.08
Cyclopropane 40.6 49.0 54.2 55.6 53.5 44.0 55.6 1.10
Tetrahydropyran 44.8 51 .O 53.6 51.5 42.3 53.7 0.93

JOURNAL OF CHEMICAL AND ENGINEERING DATA

 Table V. Continued
(0 31 mole-% H 2 0 in air)
Burning velocitv, S, as a function of equivalence ratio, 4, in cm /sec
4 =07 08 09 10 11 12 1.3
Cyclic compounds
Tetrahydrofuran 43.2 48.0 50.8 51.6 49.2 44.0 51.6 1.19
Cyclopentadiene 36.0 41.8 45.7 47.2 45.5 40.6 32.0 47.2 1 .oo
Ethylenimine 37.6 43.4 46.0 45.8 43.4 38.9 46.4 1.04
Cyclopentane 31.0 38.4 43.2 45.3 44.6 41.0 34.0 45.4 1.03
Cyclohexane 41.3 43.5 43.9 38.0 44.0 1.08
Inorganic compounds
Hydrogen 102 120 145 170 204 245 213 290 325 1.80
Carbon disulfide 50.6 58.0 59.4 58.8 57.0 55.0 52.8 51.6 59.4 0.91
Carbon monoxide 28.5 32.0 34.8 38.0 52.0 2.05
Hydrogen sulfide 34.8 39.2 40.9 39.1 32.3 40.9 0.90

Table VI. Burning Velocities of Various Fuels a t 100” C. Air-Fuel Temperature

(0.31 moIe-Yc H,O in Air)
Burnins velocitv. S.as a function of equivalencc ratio, 6 in cm. sec.
@ = 0.7 0.8 0.9 1 .o 1.1 1.2 1.3 1.4 S,;,, d at S,;,
Propargyl alcohol 76.8 100.0 110.0 110.5 108.8 105.0 85.0 11O.j 1.08
Propylene oxide 74.0 86.2 93.0 96.6 97.8 94.0 84.0 71.5 97.9 1.09
Hydrazine‘ 87.3 90.5 93.2 94.3 93.0 90.7 87.4 83.7 94.4 0.98
Furfural 62.0 73.0 83.3 87.0 87.0 84.0 77.0 65.5 87.3 1.05
Ethyl nitrate 70.2 77.3 84.0 86.4 83.0 72.3 86.4 1.oo
Butadiene monoxide 51.4 57.0 64.5 73.0 79.3 81.0 80.4 76.7 81.1 1.23
Carbon disulfide 64.0 72.5 76.8 78.4 75.5 71 .O 66.0 62.2 78.4 1.00
n-Butyl ether 67.0 72.6 70.3 65.0 72.7 0.91
Methanol 50.0 58.5 66.9 71.2 72.0 66.4 58.0 48.8 72.2 1.08
Diethyl cellosolve 49.5 56.0 63.0 69.0 69.7 65.2 70.4 1.05
Cyclohexene
monoxide 54.5 59.0 63.5 67.7 70.0 64.0 70.0 1.10
Epichlorohydrin 53.0 59.5 65.0 68.6 70.0 66.0 58.2 70.0 1.10
n-Pentane 50.0 55.0 61.0 62.0 57.0 49.3 42.4 62.9 1.05
n-Propyl alcohol 49.0 56.6 62.0 64.6 63.0 50.0 37.4 64.8 1.03
n-Heptane 41.5 50.0 58.5 63.8 59.5 53.8 46.2 38.8 63.8 1.00
Ethyl nitrite 54.0 58.8 62.6 63.5 59.0 49.5 42.0 36.7 63.5 1 .00
Pinene 48.5 58.3 62.5 62.1 56.6 50.0 63.0 0.95
Nitrwthane 51.5 57.8 61.4 57.2 46.0 28.0 61.4 0.92
Iso-octane 50.2 56.8 57.8 53.3 50.5 58.2 0.98
Pyrrole 52.0 55.6 56.6 56.1 52.8 48.0 43.1 56.7 1.oo
Aniline 41.5 45.4 46.6 42.9 37.7 32.0 46.8 0.98
Dimethyl formamide 40.0 43.6 45.8 45.5 40.7 36.7 46.1 1.04
“Results questionable because of an indication of decomposition in the stainless-steel feed system.

Garside, J. E., Forsyth, J. S., Townend, D. T. A,, J , Zmt. Fuel Linnett, J. W., Pickering, H . S., Wheatley, P. Zbid., 47,
J.?

18,175 (1945). 974-80 (1951).

Gaydon, A. G., Wolfhard, H. G., “Flames, Their Structure, Manton, J., von Elbe, G., Lewis, B., “Fourth Symposium (In-
Radiation and Temperature,” Chapman and Hall, London, ternational) on Combustion,” pp. 358-63, Williams & Wilkins
1953. Co., Baltimore, 1953.
Gerstein, M . , Levine, O., Wong, E. L., 3. A m . Chm. SOC. 73, Pickering, H . S., Linnett, J. W., T r a m . Farmfay Soc. 47, 989-92
418-22 (1951); Znd. Eng. Chem. 43,2770-2 (1951). (1951); 47, 1101-06 (1951).
G u y , M., Ann. chim.phys. 18,27 (1879). Reynolds, T. W., Gerstein, M., “Third Symposium on Com-
Gray, K. L., Linnett, J. W., Mellish, C. E., T r a m . Faraday SOC. bustion and Flame and Explosion Phenomer.a,” pp. 190-4,
..
48,1155-63 (1952). Williams & Wilkins Co.,Baltimore, 1949.
Grove, J. R., Hoare, M. F., Linnett, J. W., Zbid., 46, 745-55 Sherratt, S., Linnett, J. W., Tram. Faraday SOC.44, 596-608
(1950). (1948).
Jost, Wilhelrn, “Explosion and Combustion Processes in Gases,” Simon, D. M., Ind. Eng. Chem. 43,2718-21 (1951).
trans. by H. 0. Croft, McGraw-Hill, New York, 1946. Singer, J. M., “Fourth Symposium (International) on Com-
Kapp, N. M., Snow, B., Wohl, K . , “Effect of Water Vapor on bustion,” pp. 352-358, Williams 81 Wilkins GI., Baltimore, 1953.
the Normal Burning Velocity and on the Stability of Butane-Air Smith, F. A., Pickering, S. F., J . Research Natl. Bur. Standards
Flames Burning above Tubes in Free Air,” United Aircraft 1 7 , 7 4 3 (1936).
Corp. Meteor Rep. UAC-30 (Nov. 1948).
(35) Strehlow, R. A,, Stuart, J. G., “Fourth Symposium (Inter-
(24) Levine, O., Gerstein, M., “Fundamental Flame Velocities of national) on Combustion,” pp. 329-36, Williams & Wilkins Co.,
Pure Hydrocarbons, 111,” NACA Research Mem. R M JL51J05
Baltimore, 1953.
(Dec. 3, 1951).
Lewis, B., von Elbe, G., 3. Chem. Phys. 11, 75 (1943); “Com- (36) Wheatley, P. J., Linnett, J. W., T r a m . Faraday SOC.48, 3 3 8 4 5
bustion, Flames and Explosions of Gases,” Academic h s s , New (1952).
York, 1951. RECEIVED for review January 21, 1958 Accepted October 8, 1958.
Linnett, J. W., Hoare, M . F., Trans. Faraday SOC.47, 179433 This work was carried out for the Department of the Navy, Bureau of
(1951); 49,103849 (1953). Ordnance, under contract NOrd-9756, as part of Project Bumblebee.

VOL. 4, No. 3, JULY 1959 237