]\gt ; 277 ] XV .
On the Thermodynamic Theory of Waves of Finite Longitudinal Disturban W. J. MACQUORN RANKINE , C.E. , .D .
, F.R. .
Lond. ae Edin .
, 13 , \mdash ; Read Deaember 16 , 1869 .
object of the present investigation is to determine the relations which must exist between the of the elasticity of any , whether gaseous , liquid , or solid , and those of the wave-like propagation of a finite longitudinal bance in that substance ; in other words , of a disturbance consisting in displacements of particles along the direction of propagation , the velocity of displacement of the particles bein so great that it is not to be neglected in comparison with the velocity of propagation .
In particular , the investigation aims at ascertaining whnt conditions as to the transfel of heat from particle to particle must be fulfilled in order that a finite longitudinal disturbance may be ated along a prismatic or cylindrical mass without loss of energy or of type : the word type being used to denote the relation between the extent of disturbance at a given instant of a set of particles , and their respective undisturbed positions .
The disturbed matter in these inquiries may be conceived to be contained in a straight tube of uniform cross-section and indefinite length .
Mass-Velocit ?
/\mdash ; A convenient quantity in the present investigation : is what may be termed the or somatic \mdash ; that is to say , the mass of matter through which a disturbance is propagated in a unit of time while advancing along a prism of the sectional area unity .
That mass-velocity will be denoted by Let denote the bulh.ness , or the space filled by unity of mass , of the substance in the undisturbed state , and the linear velocity of advance of the wave ; then we have evidently . . . . . .
( 1 ) .
Cinematical Condition of Permanency of Type.\mdash ; If it be possible for a wave of disturbance to be propagated in a uniform tube without change of type , that possibility is expressed by the uniformity of the mass-velocity for all parts of the wave .
Conceive a space in the supposed tube , of an invariable length , to be contained between a pair of transverse planes , and let those planes advance with the linear velocity in the direction of propagation .
Let the values of the bulkiness of the matter at the foremost and ltftermost planes respectively be denoted by and , and those of the velocity of longitudinal disturbance by and .
Then the linearvelocitieswith which the particles traverse the two planes respectively are as follows : for the foremost plane , for the aftermost plane .
The uniformity of type of the disturbance involves as a condition , that equal masses of matter traverse the two planes l.espectively in a given Then in each unit of time the difference of pressure , , impresses on the mass the acceleration , and consequently , by the second law of motion , we have the following value for difference of pressure : . . .
( 4 ) Then substituting for the acceleration its value in terms of the change of bulkiness as given by equation ( 3 ) , we obtain , for the dynamical condition of permanency of type , the following equation , .
( 5 ) which may also be put in the form of an expression giving the value of the square of the mass-velocity , viz. . . . .
( 6 ) The square of the linear velocity of advance is given by the following equation : . .
( 7 ) The integral form of the preceding equations may be expressed as follows .
Let , as before , be the bulkiness in the undisturbed state , and the longitudinal pressure ; then in a wave of disfurbance of permanent type , we must have the following condition fulfilled : .
( 8 ) .
Waves of Sudden Disturbance.\mdash ; The condition expressed by the equations of the preceding section holds for any type of disturbance , continuous or discontinuous , gradual or abrupt .
To represent , in particular , the case of a single abrupt disturbance , we must conceive the foremost and aftermost advancing planes already mentioned to coalesce into one .
Then is the longitudinal pressure , and the bnlkiness , in front of tJle advancing plane ; is the lonigtudinal pressure , and the bulkiness , behind the advancing plane ; and the advancing plane is a wave-front of sudden compression or of sudden rarefaction* *August , Sir WILLIAN TKOIrSON has pointed out to the author , that a wave of udden rarefaetion , though mathematically possible , is an unstable condition of motion ; any deviation \amp ; om absolute suddenness tending to make the disturbance become more and more dual .
Hence the only wave of sudden disturbance whose permanency of type is physically possible , is one of compression ; and this is to be taken into account in connexion with all that is stated in the paper respecting such waves .
TIIEORY OF WAVES OF FINITE ONGITUDINAL DISTURBANCE .
according as is greater or less than P. The squares of the -velocity and of the linear velocity of advance are respectively as follows : ; ( 9 ) .
( 10 ) The velocity of the disturbed particles is as follows : ; ( 11 ) and it is forward or backward according as the wave is one of compression or ofrarefactionThe energy expended in unity of time , in .
any such wave , is expressed by pu ; for the wave may be conceived to be produced in a tube closed at one end by a moyeable piston of inappreciable mass , to which there is applied a pressule different from the undisturbed pressure , and which consequently moves the velocity .
The way in which that energy is disposed of is as follows : actual energy of the distulbance , ; work done in altering bulkiness , ; and the equation of the conselvation of energy is . . . . .
( 11A ) .
Thermodynamic \mdash ; While the equations of the two preceding sections impose the constancy of the rate of variation of pressure with bulkiness the disturbance as an indispensable condition of permanency of type of the wave , they leave the limits of pressure and of bulkiness , being four quantities , connected by one equation only .
Two only of those quantities can be arbitrary ; therefore one more equation is required , and that is to be determined by the aid of the laws of thermodynamics .
It is to be observed , in the first place , that no substance yet known fulfils the condition expressed by the equation constant , between finite limits of disturbance , at a constant temperature , nor in a state of non-conduction of heat ( called the adiabatic state ) .
In order , then , that permanency of type may be possible in a wave of longitudinal disturbance , there must be both change of temperature and conduction of heat during the disturbance .
The cylindrical or prismatic tube in which the bance is supposed to take place being ideal , is to be considered as non-conducting .
Also , the foremost and aftermost transverse advancing planes , or front and back of the wave , which contain between them the particles whose pressure and bulkiness are in the act of varying , are to be considered most and behind the aftermost plane , to resist conduction .
The transfer of heat , therefore , takes place wholly amongst the particles undergoing variation of pressure and bulkiness ; and therefore for any given particle , during its from the front to the back of the wave , the integral amount of heat receivel must be nothing ; and this is the thermodynamic condition which gives the required equation .
That equation is expressed as follows : ; . . . . .
( 12 ) in which denotes absolute tPmpelature , and the " " thelmodynamic function The value cf that function , as explained in various papers and treatises on thermodynamics , is given by the formula : hyp , . . . . .
in which the dynamical alue of a unit of heat ; , the real specific heat of the substance ; , a function of the temperature alone , which is for all temperatures at which the substance is capable of indefinitely to the perfectly gaseous state , and is introduced into the formula solely to provide for the possible existence of substances which at some tempelatures are incapable of approximating to the perfectly gaseous state ; and , the work which the elastic forces in unity of mass are capable of doing at the constant tempelature .
The substitution for the in equation ( 12 ) of its value in terms of and for any particular substance , gives a relation between the limits of pressure and , and the limits of bulkiness and , which being combined with equation ( 5 ) , or with any one of the equivalent equations ( 6 ) , ( 8 ) , or ( 9 ) , completes the expression of the laws of the gation of waves of finite longitudinal disturbance and permanent type in that particular substance . .
Assumption as to ansfer of IIeat.\mdash ; In the principles of the preceding section to the propagation of waves of longitudinal disturbance , it is obviously assumed that the transfer of heat takes place between the various particles which are undergoing disturbance at a given time , in such a manner as to ensure the fulfilment of the dynamical condition of permanency of type .
It appears highly probable , that how great soever the resistance of the substance to the conduction of heat may be , that assumption as to the transfer is ealized when the disturbance is sudden , as described in ; for then particles in all the successive stages of the change of pressure and bulkiness within the limits of the disturbance are at inappreciable distances each other ; so that the resistance to the transfer of heat between them is inappreciable .
But when the bance is not sudden , it is probable that the assumption as to the transfer of heat is fulfilled in an approximate manner only ; and if such is the case , it follows that the only longitudinal disturbance which can be propagated with absolute per manence of type is a sudden disturbance . .
Combination of the and Equations.\mdash ; In every fluid , and THEORY OF WAVES OF FINITE LONGITUDINAL DISTURBANCE .
probably in many solids , the quantity of heat received during an indefinitely small change of pressure and of bulkiness is capable of being expressed in either of the following forms : ; in which and denote the specific heat at constant bulkiness and at constant pressure respectively ; and the differential coefficients and of the absolute temperature are taken , the former on the supposition that the bulkiness is constant , and the latter on the supposition that the pressure is constant .
Let it now be supposed that the bulkin varies with the pressure according to some definite law ; and let the actual rate of variation of the bulkiness with the pressure be denoted by .
Then equation ( 12 ) may be expressed in the following form : Now , according to the dynamic condition of permanence of type , we have by equation ( 6 ) , ; which , being substituted in the preceding integrals , gives the following equations from which to deduce the square of the mass.velocity : . .
( 13 ) It is sometimes convenient to substitute for the following value , which is a known consequence of the laws of thermodynamics : .
the differential coefficient being taken on the supposition that is constant .
The equations ( 13 ) and are applicable to allfluids , and probably to many solids also , especially those which are isotropic .
The determination of the squared mass-velocity , , enables the bulkiness for any given , and the corresponding velocity of disturbance , to be found by means of the following formulae , which are substantially identical with equations ( 8 ) and ( 3 ) respectively : ; ( 14 ) . .
( 15 ) 282 PROBBSOR W. J. M. RANXINE ON THB THERMODINAMIC Equation ( 16 ) also serves to calculate the pressure corresponding to a given velocity of disturbance .
It may here be repeated that the linear velocity of advance is ( equation 1 ) . .
Application to a ct Gas.\mdash ; In a perfect , the specific heat at constant volume , , and the specific heat at constant pressure , , are both constant ; and consequently bear to each other a constant ratio , , whose value for air , oxygen , nitrogen , and hydrogen is nearly , and for steam-gas nearly .
Let this ratio be denoted by 7 .
Also , the differential coefficients which appear in equations ( 13 ) and have the follo ing values:\mdash ; . . . .
( 16 ) When these substitutions are made in equation ( 13 ) , and constant common factors cancelled , it is reduced to the following : . . . . . . . . .
( 17 ) But according to the dynamical condition of permanence of type , as expressed in equation ( 8 ) , we have ; whence it follows that the value of the integral in equation ( 17 ) is ; which , being divided by , gives for the square of the mass-velocity of advance the following value : . . . . . .
( 18 ) The square of the linear velocity of advance is ( 19 ) The velocity of disturbance corresponding to a given pressure , or , conversely , the pressure corresponding to a given velocity of disturbance , may be found by means of equation ( 15 ) .
Such are the general equations of the propagation of waves of longitudinal disturbance of permanent type along a cylindrical mass of a perfect gas whose undisturbed pressure and bulkiness are respectively and S. In the next two sections particular cases will treated of . .
Wave of Oscillation in a Perfect Gas.\mdash ; Let the mean between the two extreme THEORY OF WAVES OF FINITE pressures be equal to the undisturbed pressure ; that is , let , ; ( 20 ) then equations ( 18 ) and ( 19 ) become simply , ( 21 ) and ; ( 22 ) the last of which is LAPLACE'S well-known law of the propagation of sound .
The three equations of this section are applicable to an indefinitely long series of waves in which equal disturbances of pressure take place alternately in opposite directions . .
Wave of Permanent Compression or Dilatation in a Tube of Perfect \mdash ; To adapt equation ( 18 ) to the case of a wave of permanent compression or dilatation in a tube of perfect gas , the pressure at the front of the wave is to be made equal to the undisturbed pressure , and the pressure at the back of the wave to the final or permanently altered pressure .
Let the final pressure be denoted simply by ; then , and ; giving for the square of the mass-velocity , ( 23 ) for the square of the linear velocity of advance , ( 24 ) and for the final velocity of disturbance ( 25 ) Equations ( 23 ) and ( 24 ) show that a wave of condensation is propagated faster , and a wave of rarefaction slower , than a senes of waves of oscillation .
They further show that there is no upper limit to the velocity of propagation of a wave of condensation ; and also that to the velocity of propagation of a wave of rarefaction there is a lower limit , found by making in equations ( 23 ) and ( 24 ) .
The values of that lower limit , for the squares of the mass-velocity and linear velocity respectively , are as follows:\mdash ; ; ( 26 ) ; ( 27 ) and the corresponding value of the velocity of disturbance , being its negative limit , is ( 28 ) ; and its alternative roots are given by the following formula : . . .
( 29 ) The sign to be used , according as the piston moves inwards so as to produce condensation , or outwards so as to produce rarefaction .
Suppose , now , that in a tube of unit area , filled with a perfect gas whose undisturbed pressure and volume are and there is a piston dividing the space within that tube into two parts , and moving at the uniform velocity : condensation will be propagated from one side of the piston , and rarefaction from the other ; the pressures on the two sides of the piston will be expressed by the two values of in equation ( 29 ) ; and the force required in order to keep the piston in motion will be the difference of these values ; that is to say , . . . .
( 30 ) Two limiting cases of the last equation may be noted : first , if the velocity of the piston is very small compared with the velocity of sound , that is if is very small , we have nearly ; ( 30A ) secondly , if the velocity of the piston is very great compared with the velocity of sound , that is if is very small , we have nearly ( 30B ) .
Absolute Temperature.\mdash ; The absolute temperature of a given particle of a given substance , being a function of the pressure and bulkiness , can be calculated for a point in a wave of disturbance for which and are given .
In particular , the absolute temperature in a perfect gas is given by the following well-known thermodynamic formula : ; ( 31 ) and if , in that formula , there be substituted the value of in terms of , given by equa THEORY OF WAVES OF FINITE LONGITUDINAL DISTURBANCE .
tions ( 8 ) and ( 18 ) combined , we find , for the absolute temperature of a particle at which the pressure is , in a wave of permanent type , the following value : ; ( 32 ) PS in which the first factor \mdash ; is obviously the undisturbel value of the absolute temperature .
For brevity 's sake , let this be denoted by T. The particular cases may be noted .
In a wave of oscillation , as defined in , we have ; and consequently In a wave of permanent condensation or rarefaction , as described in , let ; then the final temperature is .
Types of Disturbance capable of Permanence.\mdash ; In order that a particular type of disturbance may be capable of permanence during its propagation , a relation must exist between the temperatures of the particles and their relative positions , such that the conduction of heat between the particles may effect the transfers of heat required by the thermodynamic conditions of permanence of type stated in During the time occupied by a given phase of the bance in traversing a unit of mass of the cylindrical body of area unity in which the wave is travelling , the quantity of heat received by that mass , as determined by the thermodynamic conditions , is expressed in dynamical units by The time during which that transfer of heat takes place is the reciprocal of the massvelocity of the wave .
Let be the rate at which temperature varies with longitudinal distance , and the conductivity of the substance , in dynamical units ; then the same quantity of heat , as determined by the laws of conduction , is expressed by .
The equality of these two expressions gives the following general differential equation for the determination of the types of disturbance that are capable of permanence : ( 33 ) The following are the results of two successive integrations of that differential equation:\mdash ; ( 33A ) ; ( 33B ) and It is unnecessary to occupy space by giving the whole details of the calculation ; and it may be sufficient to state that the following are the results .
Let ; then ( 34 ) .
hyp hyp . . .
( 34A ) In equation it is obvious that is reckoned from the point where ; that is , where the pressure ; a mean between the greatest and least pressures .
The direction in which is positive may be either the same with or contrary to that of the advance of the wave ; the former case represents the type of a wave of rarefactiom , the latter that of a wave of compression .
For the two limiting pressures when becomes infinite , and becomes positively or negatively infinite ; so that the wave is infinitely long .
The only exception to this is the limiting case , when the conductivity is indefinitely small ; and then we have the following results : when , or is infinite , and is indefinite ; and for all values of between and and are each indefinitely small .
These conditions evidently represent the case of wave of abrupt rarefaction or compression , already referred to in and 7 .
THEORY OF WAVES OF FINITE LONGITUDINAL DISTURBANCE .
Supplnnent to a Paper " " On the Thermodynamic Theory of Waves of Finite Longitudinal Disturbance by W. J. MACQUORN RANKINE , C.E. , .D .
, P.R. .
Lond. ae Edin .
Received October l , \mdash ; Read December 16 , 1869 .
Note as to previous investigations.\mdash ; Four previous investigations on the subject of the transmission of waves of finite longitudinal disturbance may be referred to , in order to show in what respects the present investigation was anticipated by them , and in what respects its results are new .
The first is that of POISSON , in the Journal de l'Ecole Polytechnique , vol. .
Cahier14 , p. 319 .
The author arrives at the following general equations for a gas fulfilling MARIOTTE'S ; in which is the velocity-function ; the velocity of disturbance , at the time , of a particle whose distance from the origin is is the limit to which the velocity of propagation of the wave approximates when becomes indefinitely small , viz. being the undisturbed pressure and the undisturbed density ; and denotes an arbitrary function .
This equation obviously indicates the quicker propagation of the parts of the wave where the disturbance is forward ( that is , the compressed parts ) and the slower propagation of the parts where the disturbance is backward ( that is , the dilated parts ) .
The second is that of Mr. SrOXES , in the Philosophical Magazine for November 1848 , 3rd series , vol. xxxiii .
p. 349 , in which that author shows how the type of a series of waves of finite longitudinal disturbance in a perfect gas alters as it advances , and tends ultimately to become a series of sudden compressions followed by gradual dilatations .
The third is that of Mr. AIRY , Astronomer Royal , in the Philosophical Magazine for June 1849 , 3rd series , vol. xxxiv .
p. 401 , in which is pointed out the analogy between the above-mentioned change of type in waves of sound , and that which takes place in sea-waves when they roll into shallow water .
The fourth , and most complete , is that of the Rev. SAMUEL EABNSHAW , received by the Royal Society in November 1858 , read in January 1859 , and published in the Philosophical Transactions for 1860 , page 133 .
That author obtains exact equations for the propagation ofwaves of finite longitudinal disturbance in a medium in which the pressure is any function of the density ; he shows what changes of type , of the kind already memtioned , must go on in such vaves ; and he points out , finally , that in order that the type may be permanent ( in the notation of the present paper ) must be a constant ON THE THERMODYNAMIC THEORY OF WAVES .
quantity ; being the proposition which is demonstrated in an elementary way near the beginning of the present paper .
Mr. EARNSHAW regards that condition as one which cannot be realized .
The new results , then , obtained in the present paper may be considered to be the following : conditions as to transformation and transfer of heat which must be fulf in order that permanence of tyPe may be realized , exactly or approximately ; the types of wave which enable such conditions to be fulfilled , with a given law of the conduction of heat ; and the velocity of advance of such waves .
The method of investigation in the present paPer , by the aid of mass-velocity to express the speed of advance of a wave , is new , so far as I know ; and it seems to me to have great advantages in point of simplicity , enabling results to be demonstrated in a very elementary manner , which otherwise would have required comparatively long and elaborate processes of investigation .

