(SANITIZED)UNCLASSIFIED SOVIET RESEARCH PAPERS ON COMBUSTION(SANITIZED)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80-00247A003100420001-9
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 10
• Document Creation Date: December 27, 2016
• Document Release Date: February 28, 2014
• Sequence Number: 1
• Publication Date: October 27, 1964
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
R
50X1 -HUM
Next 4 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
-
?e-
DETERMINATION OF THE RATE CONSTANT FOR THERMAL
CRACKING OF METHANE BY ADIABATIC COMPRESSION
AND EXPANSION
V. N. Kondratiev
USSR Academy of Sciences
The method of adiabatic compression and expansion worked out by
Ryabinini/ more than ten years ago has not found due application in
measuring rate constants of elementary chemical reactions at high
temperatures and pressures. The possibility of such an application
was shown only recently by Markevich et al.-" for the thermal,cracking
of methane. It was ascertained that due to specific temperature
variations in the course of compression and expansion, calculation
may be made considerably less tedious.
The experimental data obtained in this work permit fairly accurate
determination of the rate constant for unimolecular decomposition of
methane molecules. The following conditions may be considered as most
favorable.
1. With a system containing a low percentage of the reacting
gas considerably diluted with some inert gas, for example argon, the
reaction-induced variations in temperature may be neglected and the
reaction considered as strictly adiabatic.
2. Since under adiabatic compression the reaction becomes
measurable only at temperatures and pressures higher than certain
limiting T* and p* values which differ only slightly from Tmax and
max' e
= P may be considered as constant throughout the ranges of
P Cv
*
- T and p P.
Tmax max -
3. Hydrogen and. ethane were shown to he the main reaction
products of the initial reaction (i.e., at low compression)21/.
Consequently, under these ,conditions the reaction will follow a simple
mechanism.
CH CH3 + H
H + CH4 = CH3 + H2
2CH32H6
Under the assumption of a stationary H atom concentration, the
rate of hydrogen formation mill be
d(H2)
= Al(CH4)e-E/RT
dt
Atch 1
STAT
nAciassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
.
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
where Al is the frequency factor in the rate constant for the reaction
CH4 = CH3 + H. The data obtained by Markevich et al.,Al2-/, as well
as the majority of other published data, seem to be in favor of a
unimolecular mechanism for this reaction (under the experimental
conditions used).
4. Due to the high activation energy E, the reaction proceeds
at a temperature close to Tmax. This assumption is similar to that
taken as a basis for the theory or thermal flame propagation by
Zeldovich and Frank-Kamenetzkiili, namely that combustion proceeds
at a temperature differing only slightly from themaximum combustion
temperature.
(H2) and making use of the first
Introducing the value ip
(CH4)
three conditions, we obtain, after certain transformations, on the basis
of the assumed reaction mechanism and from the equation for piston
motion,
-xo
A e-E/RTdT
2(7- 1) poV_
T ? (Tmax T)
where xo is the length of the reactor channel, in is the piston mass
and po, Vo, To are the initial pressure, volume and temperature.
Integration, making allowance for the fourth of the above
conditions, gives
Ale-E/RTmax
oc
= xo
VI
E( -1) p0V0
2 m RTo
x*
where'y, is the finaf value of the (H2)/(CH4) ratio,o> Al
where A2 is the frequency factor in the rate constant of a bimolecular
reaction. The validity of this inequality ,ill be apparent if we assume
that AI = l015sec-1, A2 = 1021cm3mole-Isec- (see above) and (Ar) =
(Ar)00(= 2.5 x 1019 x 20 : 6 x 1023 = 10-3 mole cm3. This may be
considered as evidence that the thermal dissociation of methane under
experimental conditions212/ obeys a strictly unimolecular law.
For the determination of rate constants of elementary chemical '
reactions, the method of adiabatic compression and expansion seems no
less accurate than the shock tube method.
References
1. Yu. N. Ryabinin: Zhur. Exp. Teor. Fix. 23, 461 (1964).
2. a. A. M. Markevich, V. V. Azatyan, N. A. Sokolova: Kinetika i Kataliz
3, 431 (1962); b. I. E. Volokhonovich, A. M. Markevich, 1. F.
Masterovoy, V. V. Azatyan, Dokl. Akad. Naukk SSSR 146, 387 (1962).
3. Ya. B. Ze1dovich, D. A. Frank-Kamenetzkii: Zhur. Fiz. Khim 12, 100 (1938)1.
4. G. B. Skinner, R. B. Ruherwein: J. Phys. Chem. 63, 1736 (1959).
3
Atch 1
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28 CIA-RDP80-00247A003100420001-9
?
ON THE NATURE OF THE CRITICAL DIAMETER
-
A. N. Dremin and V. S: Trofimov
USSR Academy of Sciences
It has been established by many investigators concerned with the
detonation of liquid explosives that the shock front of a detonation
wave is not smooth. As in gas detonations, oblique shocks due to
instability of the ignition plane travel over the detonation wave
front. The ignition plane instability is a function of the temperature
dependence of the reaction.
In traveling over the shock front of a detonation wave, the
oblique waves collide and initiate a reaction. The induction period
of the latter at the points of collision is considerably shorter than
that observed beyond the shock front.
Consequently, initiating centers will not form when the conditions
for intercollisions and collisions with reactor walls are unfavorable.
When detonation passes from the tube into the air, the oblique waves
at the edge find nothing to collide with. Thus no active centers will
form at the shock periphery at the point of detonation exit. As a result,
there will be no formation of new oblique shocks at the periphery of the
detonation near the exit, and this, in turn, will lead to the disappearance
of neighboring initiating centers. This may be seen from smear-camera
records obtained for the detonation of transparent explosives, as
observed from the butt end.
It was shown experimentally that detonation will proceed at the
same rate if reaction-quenching waves do not penetrate as far as the
axis of the charge. When they reach the latter, detonation stops. It
will be apparent from this that the reaction-quenching waves may penetrate
to a distance equal to one-half the critical diameter.
Let us consider what governs the penetration of reaction-quenching
waves. After destruction of the oblique shocks, there remains a smooth
shock wave at the detonation periphery; it is compressed by detonation
products, as by a piston. The main shock compresses the explosive and
initiates a reaction after a certain induction period T'which depends
upon the temT,erature and pressure behind the wave- . A rarefied wave
comes from the charge edge at a rate C, thus the reaction has to occur
at a place unattained by this wave, after a time The reaction will
then proceed into the charge as a detonation over the shock-compressed
substance. It overtakes the reaction-quenching wave at a distance der
and stops the destruction of oblique waves.
Atch 2
STAT
nnr?Inecifiarl in Part - Sanitized Copy Approved for Release 2014/02/28 CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28 : CIA-RDP80-00247A003100420001-9
An equation for the Critical diameter of detonation of a liquid
explosive may be obtained on the basis of the above considerations.
d = 2 vru ( 1 1
Cr D - U U - C
(1)
Here V is the velocity of oblique shocks traveling over the
detonation front; U is the velocity of the explosive compressed by the
shock. relative to the boundary of the quenching, wave; D is the rate of
detonation propagation over the compressed substance.
It follows from the above equation that, besides depending on
kinetic parameters of the explosives, dCr is a function of the mechanical
properties of the explosive and its detonation products. It is Also
quite conceivable, that a more active substance would have a higher dcr,
due to its mechanical properties.
Equation (1) was used to calculate dc for nitromethane (NM) at
1? to 20?C and for melted trotyl (TNT) at g20 to 830C. The value obtained
for NM was dcr = 14.3 mm, as compared to the experimental 18 mm; for
TNT the calculated value was 62, and the experimental was 68 mm.
2
Atch 2
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
'Lc
CHARACTERISTICS OF LAMINAR AND TURBULENT FLASHBACK
L. N. Khitrin, Ph, B. Moi n, B: B. Smirnow and V. U. Shewchuk
The Krzhizhanovsky Institute Of Energetics, Moscow
The laminar flame flashback mechanism seems to be clear if one
considers a combustible gas mixture issuing from the port of a cylindrical
tube'. The Lewis and von Elbe "gradient flame flashback theory" has
now received conclusive proof based on numerous experimental data for
the burning of many gaseous but unheated mixtures.
Fewer investigations, however, have been dedicated to flashback
under other conditions: at _high laminar burning velocities, particu-
larly in preheated mixtures, or when the gas velocity profile differs
from the classical type for laminar steady flow in tubes (parabolic
profile). The fundamentals of the flashback in turbulent flOw also need
further clarification. In this field, experimental data are limited
and no attempts have been made to establish criteria for their evaluation.
This paper describes and discusses the influence of some factors
which are essential from toe standpoint of the flame flashback, such
as the wide variation of the laminar burning velocity, the type of gas
velocity profile of the stream issuing from the burner port (tubes,
nozzles), the burner diameter variation, the state of flow (laminar
or turbulent), and the type of mixture.
Experimental Results
Laminar flashback in tubes. The influence of pre-heating methane-
oxygen mixtures on the flashback velocities was investigated at .temper-
atures ranging.from 200 to 400?C., for a wide range of mixture compo-
sitions. Such investigations were made also with regard to the laminar
(normal) burning velocity. It was noted that the velocity gradient, g;
increased considerably with increasing temperature but somewhat more
slowly than the laminar burning velocity, Hip, and not aswould,have
been expected from the results for unheated mixtures, where it was.
proportional to U21). This is explained by the nature of the heat
conductivity coefficient, a, which is practically independent of the
mixture composition but rises markedly-with the temperature.
Taking into account the temperature dependence of a.and using the
- 'experimental data for the flashback velocity w and for lib., one can
obtain good agreement with the well-known Putnam and Jensen 'relation,
Atch 3
STAT
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28 : CIA-RDP80-00247A003100420001-9
Pe = kS2 (1)
wd Ubd
where Pe =, and S
a a
The coefficient k was found to be 0.01 which is Close to the
corresponding value of Putnam and dense, k = 0.0125.
It should be noted here that the maximum value of lib reached in
the experiments under discussion was Ubew 1500 cm/sec., which is 30 times
higher than that obtained hitherto when investigating flashback
phenomena.
Laminar flashback in nozzles. Here, methane-air mixtures and
nozzles of different shapes and sizes were used. It was found that the
flashback depended upon the ratio of the inlet nozzle diameter to the
outlet nozzle (Dc1); if this ratio increase's, the flashback velocity
increases too, even when the outlet diameter is smaller. Correspondingly,
the greater difference between these diameters, the higher the flash-
back velocity, so that the latter may sometimes reach the blowout
velocity. This is due to the great uniformity of the velocity profile
at the nozzle port, which is characterized by a sharp drop in velocity
at the center of the stream. Special washers of different thickness
were used to vary the velocity distribution at the nozzle orifice. Thus,
different flashback velocities could be achieved. Two typical regimes
were observed; with a short nozzle the flame had a meniscus shape and
flashed back along the nozzle axis; when on the contrary the nozzle was
long, the flame had the usual Bunsen cone form and flashed back near
the tube wall.
Correspondingly, in these two regimes one may observe different
relations between the flashback velocity and the length of the nozzle
orifice, h; for a short nozzle w diminished with increasing h; for a long
nozzle the result was quite the reverse. When the velocity profile was
plane, the flashback velocity reached minimum and equaled Ub. Thus,
quite different results may be obtained depending upon the actual
radial velocity profile of a nozzle orifice.
Turbulent flashback. Turbulent flame flashbackin methane-oxygen,
methane-air and hydrogen-air mixtures was investigated using cylindrical
tubes of different diameters. It was found that contrary to laminar
flashback, the burner diameter effect upon the flashback velocity was
rather negligible. .
Evaluation of the laminar boundary layer thickness 6 , corresponding
to the turbulent flashback velocity indicates that its value must be
1.5 - 2.5. times the possible value of the so-called "penetration'
distance,'"X in the mixtures under consideration.
Atch 3
neclassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28 : CIA-RDP80-00247A003100420001-9
?
From this deduction it follows that in turbulent flow the flashback
process would have to develop in the vicinity of a tube wall, namely
in the boundary layer; generally speaking, visual observations confirm
this.
Hence, according to the gradient-flashback theory the following
equation may be obtained:
1.11
Re = kPr0.55S
(2)
V
where Re ,10_, Pr =-57 and -N) is the kinematic viscosity coefficient.
This relation proved to be in good agreement with the experimental
data mentioned above. Thus, the flashback mechanism in turbulent flow
from the port is the same as the laminar flashback mechanism.
AtCh 3
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9
ON THE EFFECT OF KINETICS OF ELEMENTARY REACTIONS
ON ONIZATION IN STATIONARY AND NONSTATIONARY SUPER-
SONIC EXPANSION AND COMPRESSION OF GASES
N. I. Yushchenkova and S. I. Kosterin
USSR Academy of Sciences
STAT
The present investigation was designed tO determine the ionization
ratio of the air flow, as well as the ionization ratio in. the presence '
of an alkali metal in the air for different laws of supersonic expansion,
at the initial temperatures in the adiabatic slowdown flow of T0 6000-
20000K and initial pressures of Po --- 100-0.1 atm. and for(Na. = 10-4
10-2. The results of the calculations indicated significant deviation
of the ionization ratio from equilibrium during supersonic expansion
(in the range 1.2 4 ?, 10 at 1 cm 4:-. 10 At, 100 cm) and Made it,
possible to determine the dependence of the limit ionization ratio upon
the initial expansion parameters, the law of aerodynamic expansion and
a specific size of flow 10. It is shown that there is a flow region
for some range of Z? with a gradual transition of the equilibrium
ionization ratio toward the freezing one, while the size of the region
depends significantly on the law of aerodynamic expansion. Fdr example,
the region of transition of 4. for an axisymmetrical source is two or .
three times smaller than for a ,conical nozzle with a small aperture
angle. The latter condition restilts in the fact that for flow from: a
supersonic source and during expansion of a supersonic stream-in the
vacuum, the limit ionization ratio is maximum and produces flOw regions
with negative dielectric constant. For the laws of expansion with
slower variation of t, = f(/) during flow from nozzles and flow around
a blunt body, the existence of a wide range Of Z, with negligible
deviation from the equilibrium value is typical. At the same :timethe
limit ionization ratio is significantly less than for flow from a,
supersonic source. For example, for initial values of the adiabatic
slowdown flow of To = 60000K and Po = 1 atm, the limit ionization ratio
of air for a supersonic source is 2-1. or 3 times greater than for ex-
pansion with a supersonic conical nozzle * = 100. Similar relations
apply in the presence of an alkali metal in the air but the limit
ionization ratio in this case corresponds to higher values of ?Ix .
Atch 4
Declassified in Part - Sanitized Copy Approved for Release 2014/02/28: CIA-RDP80-00247A003100420001-9