ENGINEERING, DESIGN, AND PROCESS DEVELOPMENT
velocity a t 3 megacycles per second has been measured a t various James (S), and Furth (I) have been used for solids in recent
mole fractions in 16 binaries of varying degrees of nonideality in years. They are described and compared with data of Bridgman
studies conducted in this laboratory. Earlier experimental and with computations made by the Fermi-Thomas method.
studies of this kind have been reviewed by Wada (6). These equations are used with the hydrodynamic theory to
The free volume theory is useful for interpreting and correlating obtain expressions for the pressure, temperature, and entropy
sonic velocity data for liquid solutions. The application of the increase in various metals as a result of shock wave passage
simple free volume theory to solutions has been well presented by through the solids. For the Pack-Evans-James equation of
Prigogine and Bellemans (4). For the consideration of sonic state, which appears to be adequate for interpolating between the
velocity data, the theory is extended as f o l l o ~ s . The internal Bridgman data and the Fermi-Thomas computations, the differ-
pressure, Pi,is given by the free volume equation of state ence between Hugoniot and adiabatic pressure to fourth order in
the compreasion is
(3)
where V and V Onow refer to the solution molar volume and in-
compressible molar volume, respectively. The ratio ( V0/V)1/3 The increase in entropy acrose the shock front to the same ap-
is evaluated by means of Equation 1. By the use of the partial proximation is
molar concept
where a, p are parameters of the Pack-Evans-James equation,
where N A and N B are the mole fractions and VA, v~,
and V ~ are
B VO = l / p o is specific volume a t room temperature, To, and zero
the partial molar volumes and incompressible molar volumes of pressure, C, is specific heat, g( Po)= ( ~ P / ~ T and
) T ~p ,is density
the components A and B, respectively. The two right-hand behind the shock front.
terms in Equation 4 may be individually determined by the well-
known slope-intercept method from a plot of Pi from Equation 3 Liieraiure Cited
versus mole fraction. The values of v&41’3 and V O B ~which
/~
are thus obtained vary slightly with mole fraction in the direction (1) Furth, R., Proc. Roy. Boc. (London), A183, 87 (1944).
predicted by Prigogine and Bellemans (4). The excess internal (2) Murnaghan, F. D., “Finite Deformation of an Elastic Solid,”
Wiley, New York, 1951.
pressure for the nonideal solutions does not correlate directly with (3) Pack, D. C., Evans, W. M., and James, H. J., Proc. Phys. Soc.,
the heat of mixing. The relation between the behavior of the 60, 1 (1948).
excess internal pressure and the heat of mixing may be useful
in future .studies of the molecular internal degrees of freedom
in solution, as implicitly suggested by Bondi (11.
A N INSTRUMENT TO STUDY RELAXATION
Literaiure Cited RATES BEHIND SHOCK WAVES
(1) Bondi, A., J . Phys. Chem., 58, 929 (1954).
(2) Collins, F. C., and Navidi, M. H., J . Chem. Phys., 22, 1254 E. L. RESLER, JR., AND M. SCHEIBE
(1954). lnsfitute for Fluid Dynamics and Applied Mathemafics, University of Maryland
(3) Collins, F. C., and Raffel, H., Ibid., 22, 1728 (1954). College Park, M d .
ngogine, I., and Bellemans, A., Trans. Faraday Soc., 49, 80
(4) (1953).
(5) Wada, Y . , J . Phys. SOC.J a p a n , 4, 280 (1949).
Supported in p a r t by Wright Air Development Center of U. 9. Air Force
A N INSTRUMENT is described which combines the schlieren
technique, a photomultiplier tube, and an oscilloscope to
measure the density distribution behind shock waves (in gases)
under Contract No. AF33 (616b373.
produced in a shock tube. The use of shock waves and of optical
techniques to study the chemical kinetics of reactions has been
reported (1, 2 ) . This instrument however, is simpler in many
ENTROPIC EQUATIONS OF STATE respects than those previously used and is capable of better space
resolution and higher sensitivity while still recording the same
AND THEIR APPLICATION TO information.
SHOCK WAVE PHENOMENA Basically the instrument functions in the following manner.
A photomultiplier tube intercepts all the light clearing the knife
GEORGE E. DUVALL AND BRUNO J. ZWOLlNSKl edge of a schlieren system, the light beam of which passes through
Poulfer laborafories, Stanford Research Institute, Stanford, Calif. a section of the shock tube. A property of the schlieren system is
that light will escape the knife edge in an amount proportional
T H E propagation of a shock wave through an inviscid fluid is de-
scribed by a set of jump conditions, representing conservation
of mass, momentum, and energy, which are the same for all
to the density derivative a t each point in the field of the light
beam. Since the photomultiplier tube gathers all the light pass-
ing the knife edge it integrates the density derivative over the
mediums, and by an equation of state and the specific heat char- light beam or puts out a signal proportional to the density differ-
acteristic of the particular medium. If P P / d V z is positive, ence a t the edges of the light beam. Thus the system has the
only compressive shocks can exist and the entropy increases dis- sensitivity of a schlieren and photomultiplier combination and the
continuously in the shock transition space resolution is determined not by the beam width but by how
The hydrodynamic theory of shock waves in solids is based on well the edges of the light beam are defined or to what degree
the assumption that the shear modulus is completely negligible the light in the beam is parallel. Therefore, if in the shock tube
a t very high pressures and that the transition from the undis- a shock wave with a relaxation zone behind i t characterized by a
turbed to the shocked state occurs discontinuously a t the shock density distribution passes through the light beam of the schlieren
front. The region behind the shock front is then treated as an system and if the length of the relaxation zone behind the shock
ideal fluid with a hydrostatic equation of state. Three equations wave is less than the width of the schlieren light beam, then the
of state that bear the names of Murnaghan ( 2 ) , Pack-Evans- recorded signal from the photomultiplier tube will be the density
1182 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 47,No. 6
PULSATION AND VIBRATION
distribution behind the shock wave (assuming the index of re- shown that the steady-state composition gradient is, a t most,
fraction of the gases do not change in the transition zone.) as the very small.
shock wave plus transition zone enter the light beam. By meas- Hence, for practically all types of ultrasonic waves, no appre-
uring these density distributions the way the gas or gases relax t o ciable steady-state separation can be achieved in the gaseous
equilibrium after the enthalpy of the gas is increased suddenly state. This statement applies specifically to isotopic mixtures
a calculable amount by a shock wave can be determined. and becomes less valid for mixtures of widely differing molecular
The theoretical aspects of the instrument and its predicted weight. It is also possible that under conditions where the Chap-
performance were verified experimentally by measuring vibra- man and Cowling assumptions of the continuity of the hydro-
tional heat capacity relaxation times behind shock waves in dynamic medium break down, some separation might be achieved.
oarbor! dioxide containing water vapor. The instrument in these Debye ( 2 ) showed that a potential wave, due to partial un-
tests demonstrated a sensitivity sufficient to record a change in mixing, should exist if an electrolytic solution is irradiated with a
atmospheric density of 0.5% over I-mm. distance and a space traveling ultrasonic wave. This effect was confirmed experi-
resolution of the density in the shock tube of 0.1 mm. correspond- mentally ( 5 )with a standing wave. I n this paper the magnitude
ing to times of the order of 0.1 microsecond. of the unmixing associated with the Debye effect is shown to tie
only about mole fraction at 100 megacycles and 0.1 watt
*
Literature Cited per sq. em. This is true also for nonionic solutions. The
unmixing is about the same for a standing wave as for a traveling
(1) Carrington, Tucker, and Davidson, Sorman, J . Phys. Chem., 57, wave, although in the former case the potential wave is a stand-
418 (1953). ing wave 90” out of phase from the velocity wave.
(2) Smiley; E. E”.,
Winkler, E. H., and Slawsky, Z. I., J . Chem. Phys.,
No treatment has been found in the literature of the separation
20, No. 5, 923 (1952).
to be expected on passing either asymmetrical or damped sine
This research was partially supported by the United States Air Force through waves through a liquid mixture. However, as Debye points out,
the Office of Scientific Research of the Air Research and Development the frictional coefficients are far larger than the dynamical co-
Command. efficients in liquid systems. Hence, it is not probable that ap-
preciable separations could be reached a t present ultrasonic
frequencies with either distorted or damped waves.
Despite the intense accelerative effects of ultrasonic radiation,
ULTRASONIC UNMlXlNG OF it is shown that negligible steady-state separation can be ex-
ISOTOPIC SOLUTIONS pected by passing either symmetrical or distorted sound waves,
standing waves or damped waves through gaseous, and probably
S. G. BANKOFF’ AND R. N. LYON also liquid mixtures of isotopic constituents.
Oak Ridge National laboratory, Oak Ridge, Tenn.
Literature Cited
T HE high local accelerations and the multiple stage nature of
ultrasonic radiation apparently make it attractive for sepa-
rations based on small differences in mass. Despite this apparent
Chapman, S., and Cowling, T. G., “Mathematical Theory of
Non-Uniform Gases,” Macmillan, New York, 1939.
Debye, P., J . Chem. Phys., 1, 13 (1933).
attractiveness, several authors have reported negative results Frei, H., and Schiffer, AT., Phys. Rev.,71, 555 (1947).
from theoretical and experimental investigations of this possibility. (4) Passau, P., Ann. SOC.Sci. Bruzelles, 62, Ser. I, 40 (1948).
However, these analyses deal with the simpler cases; and some (5) Yeager, E., Bugosh, J., Hovorka, F., and McCarthy, J., J. Chem.
Phys., 17, 411 (1949).
instances of relatively large unmixing have recently been re-
ported (3, 4). It seemed advisable, therefore, to institute a more
comprehensive analysis, with special reference to isotopic sepa-
rations.
A generalized theory for gaseous isotope separation is developed, COMBUSTION OSCILLATIONS IN
based on integrating the binary diffusion equation ( 1 ) over one DUCTED BURNERS
period a t cyclical steady state. This yields
JOHN C. TRUMAN
Aeronautical Engineering Dept., USAF lnsfitute of Technology,
Wright-Patterson Air Force Base, Ohio
x ROGER T. NEWTON
Small Aircraft Engine Dept., General Electric Co., Lynn, Mass.
where y is the mole fraction, z the distance along the column, t the
A PROBLEM of increasing importance in t,he development of
modern aircraft propulsion systems is that of combustion
oscillations or combustion instability. These terms refer to
time, X the wave length, c the velocity of sound, m the mass, p the
pressure, k~ the thermal diffusion ratio, and T the temperature. periodic, large amplitude variations in pressure which are main-
These terms represent, respectively, the driving forces for diffusion tained in some manner by the combustion process. Such varia-
tions usually occur in the audiofrequency range. Their effects
due to concentration, pressure, and temperature gradients. As-
include changes in the chemical and thermodynamic processes of
suming a perfect gas adiabat, the thermal diffusion term vanishes.
For isotopic mixtures, the coefficient of the pressure gradient is combustion, and hence in burner performance, and structural
damage to burner components resulting from high amplitude pres-
very nearly constant, and hence the pressure gradient term van-
sure oscillations and from locally increased heat transfer rates.
ishes, or nearly so. Hence, the steady-state concentration
Most cases of combustion instability which have been reported in
gradient is zero, irrespective of wave form, for an undamped
the literature fall into one of three classes: ( 1 ) oscillations asso-
wave. This is to be expected from thermodynamic considerations.
ciated with failure of the flame to stabilize on a flame holder; (2)
The analysis is somewhat more difficult for damped traveling
oscillations depending on the existence of a time lag between the
waves, but b y a method of approximate series solution, it is
injection of propellants into the burner, and their transformation
1 Permanent address, Rose Polytechnic Institute, Terre Haute, Ind. to high temperature gases; and (3) oscillations, initiated or
June 1955 INDUSTRIAL AND ENGINEERING CHEMISTRY 1183
___ -~
