Sound waves of finite amplitude

University of Iowa
Iowa Research Online
Theses and Dissertations
1916
Sound waves of finite amplitude
Victor August Hoersch
State University of Iowa
This work has been identified with a Creative Commons Public Domain Mark 1.0. Material in the
public domain. No restrictions on use.
This thesis is available at Iowa Research Online: http://ir.uiowa.edu/etd/3978
Recommended Citation
Hoersch, Victor August. "Sound waves of finite amplitude." MS (Master of Science) thesis, State University of Iowa, 1916.
http://ir.uiowa.edu/etd/3978.
Follow this and additional works at: http://ir.uiowa.edu/etd
SOUND WAVES OF FIN ITE AMPLITUDE
by
%
V ic t o r A. Hoersch
A th e s is
s u b m itte d to th e f a c u lt y o f
th e G raduate C o lle g e o f
The
S ta te
U n iv e r s it y
of
Iowa
in p a r t i a l f u l f i l l m e n t o f th e re q u ire m e n ts
f o r th e degree o f
M a s te r
o f
S c ie n c e
August 25 ,19 16 .
SOUND WAVES OF FINITE AMPLITUDE
OUTLINE.
A.
H ydrodynam ica l E q u a tio n s .
I.
II.
B.
System o f E u le r.
System o f Lagrange.
D is c o n t in u it ie s .
I . G eneral
P ro p e rtie s o f D is c o n t in u it ie s and T erm in ology
to be A dopted.
II.
G e n e ra lity o f a D is c o n tin u it y o f th e P ressu re
as a
F u n c tio n W h o lly o f the D e n s ity .
III.
K in e m a tic a l R e la tio n s P e r ta in in g to a D is c o n t in u it y .
IV .
S ta tio n a r y D is c o n t in u it ie s .
V.
D is c o n tin u o u s Waves.
1. D e r iv a tiv e s o f th e D e n s ity .
2. Second and H ig h e r O rder D is c o n t in u it ie s .
3. F i r s t O rder D is c o n t in u it ie s .
a. C o n tin u ity R e la tio n .
b.
Dynam ical E q u a tio n and V e lo c ity F orm ula.
c.
E q u a tio n o f E nergy.
d.
H u g o n io t's Law.
e.
f.
E n tro p y and Thermodynamic P o t e n t ia l.
A p p lic a t io n o f th e E n tro p y C o n d itio n to a
F i r s t O rder D is c o n t in u it y .
g.
h.
4.
M o tio n F o llo w in g an I n i t i a l Plane Wave D is c o n tin u it y .
(1 ).
R iem ann's T re a tm e n t.
(2 ).
Hadamard's T re a tm e n t.
O b je c tio n to H u g o n io t's Law.
Summary.
OUTLINE. - 2 G.
I n f in it e s im a l A m p litu d e Sound Waves.
D.
C ontinuous F i n i t e A m p litu d e Sound Waves.
I.
Plane Waves.
1. G eneral Theory o f C h a r a c te r is tie s f o r Plane Waves.
2. S in g le P ro g re s s iv e Waves.
a. E arnshaw 's C o n d itio n .
b . G eneral S o lu tio n and P ro p e rtie s o f a S u rfa ce I n t e g r a l,
c.
S o lu tio n W ith G iven M o tio n o f P is to n .
d.
S in g u la r it y in th e S o lu tio n .
3. Waves, n o t S in g le P ro g re s s iv e .
a. G eneral S o lu tio n .
b.
A p p lic a tio n to S p e c ia l Cases.
( 1 ) . Riemann’ s S p e c ia l Case.
( 2 ) . L in e a r
Type Waves in r and s.
( 3 ) . L im ite d I n i t i a l D is tu rb a n c e .
II.
S p h e ric a l Waves.
1. I n f in it e s im a l D is tu rb a n c e on a Steady F low .
2. F in it e A m p litu d e Waves.
E. Waves o f Permanent Regime W ith D is s ip a tiv e F o rce s.
I . E n tro p y C o n d itio n .
II.
Heat C o n d u ctio n .
III.
V is c o s it y .
IV .
V is c o s ity and Heat C o n d u ctio n .
F. Summary.
SOUND WAVES OF FINITE AMPLITUDE
The problem o f sound in a gas, in th e most g e n e ra l case,
is a problem o f the most g e n e ra l m o tio n t h a t a gas may have.
We
are le d to a c o n s id e ra tio n o f th e h yd ro d yn a m ica l e q u a tio n s .
These
are the d i f f e r e n t e q u a tio n s w hich a sound wave must s a t is f y .
A.
H ydrodynam ical E q u a tio n s ,1
The h y d ro d yn a m ica l e q u a tio n s a re most g e n e ra lly ex­
pressed in two systems o f r e p r e s e n ta tio n , t h a t o f E u le r
and
th a t o f Lagrange.
I.
System o f E u le r.
In t h is system o f re p re s e n tin g th e m o tio n o f a gas , i f
we adopt r e c ta n g u la r c o o rd in a te s , we lo o k upon th e gas from a system
o f f ix e d re c ta n g u la r axes and examine what happens a t a p o in t whose
c o o rd in a te s are x , y , z.
In d e s c r ib in g what i s happening a t a p o in t
we are in te r e s te d in knowing th e v e lo c it y o f flo w o f th e gas, in
-the eUrisity »f>cl+K€ ^YtiSUrt
m agnitude and d ir e c t io n , a t t h is p o in t .
These q u a n t it ie s w i l l , in
g e n e ra l, v a ry w ith the tim e .
A knowledge o f these q u a n t it ie s a ls o
g iv e s us a com plete knowledge o f th e s ta te o f the gas, f o r th e tem­
p e ra tu re is d e term in ed by th e c h a r a c t e r is t ic e q u a tio n o f the gas
as a fu n c tio n o f th e d e n s ity and p re s s u re ,
homogeneous.
th e gas b e in g assumed
Our d e s c r ip tio n o f th e m o tio n o f the gas, expressed
in m ath e m a tica l language would b e :-
1. I have s e le c te d th e m a te r ia l in t h is s e c tio n from Lamb's H ydro­
dynam ics, 1906 ;C h a p .I, Chap. I I , A r t s . 21 -25 , Chap. X, A r ts . 272-276,
281-282. The dynam ical e q u a tio n f o r a s y m m e tric a l s p h e r ic a l wave is
n o t e x p l i c i t l y g iv e n b u t may be e a s ily o b ta in e d by tra n s fo rm a tio n
o f the g e n e ra l e q u a tio n to s p h e r ic a l c o o rd in a te s .
-2 -
u = f^ ( x ,y ,z , t)
v = f 2(x ,y ,z ,t)
w = f g i 'x , y , z , t )
J° « P ( x , y , z , t )
P=
T T (x ,y ,z ,tj
where u ,v ,w , are the components o f th e v e lo c it y o f flo w a t the
p o in t x , y , z a t th e tim e t in th e d ir e c t io n s o f the x ,y ,a n d z axes
r e s p e c tiv e ly , where
is th e d e n s ity , and where p is th e p re s s u re
a t th e p la c e and tim e .
be p e r f e c t ly a r b i t r a r y .
We can e a s ily see t h a t these fu n d tio n s cannot
f o r exam ple, l e t us im agine any c lo s e d
s u rfa c e f ix e d w ith re s p e c t to the axes o f x ,y ,a n d z.
The flo w o f
gas outward ove r t h is c lo s e d s u rfa c e , d e te rm in e d by th e f i r s t th re e
fu n c tio n s , e v id e n tly has an e f f e c t on the d e n s ity w it h in t h is
s u rfa c e .
T h is in te rde pe nd en ce o f the d e n s ity upon th e v e lo c it y
components is expressed by the s o - c a lle d e q u a tio n o f c o n t in u it y :
(1)
A g a in , we e a s ily see t n a t the p re s s u re cannot be p e r f e c t ly a r b it r a r y
because the gas i s alw ays a c c e le ra te d so as to move from re g io n s
o f h ig h p re s s u re to re g io n s o f lo w e r p re s s u re .
t h is in v o lv e s a re c a lle d
The r e la t io n s w hich
the d yn a m ica l e q u a tio n s .
These a re :
(2)
,
An im p o rta n t case o f the m o tio n o f a gas is r e c t i l i n e a r
t
o r u n ^ ir e c t io n a l m o tio n .
I n t h is case n o t o n ly is the v e lo c it y
o f flo w in one d ir e c t io n b u t i t
is th e same in m agnitude a t e ve ry
p o in t in a p la n e a t r i g h t a n g le s to t h is d ir e c t io n and th e d e n s ity
and p re s s u re are a ls o th e same a t e ve ry p o in t in t h is p la n e .
-3 -
L e t us ta k e the d ir e c t io n o f flo w as t h a t o f the x a x is ,
i'hen e ve ry
q u a n tity in v o lv e d is a f u n c tio n s o le ly o f x and t and v and w
disap-pear from the above e q u a tio n s .
The e q u a tio n o f c o n t in u t iy
thus becomes:
(3)
and we have l e f t th e one d yn am ical e q u a tio n :
(4)
If
, in a d d it io n , a t e ve ry p o in t ,
the c o n d itio n o f th e gas does n o t
v a ry w ith tim e , we have ste a d y r e c t i l i n e a r m o tio n .
u , p , and p a re fu n c tio n s s o le ly o f x .
T h is means th a t
The p a r t i a l d i f f e r e n t i a l
e q u a tio n s the n become o rd in a r y and we o b ta in by in t e g r a t io n :
(5)
(6)
The l a s t e q u a tio n expresses th e c o n s e rv a tio n o f e n e rg y.
The f i r s t
term is the k in e t ic energy o f u n it mass o f the gas and the l a t t e r
its
p o t e n t ia l e n e rg y.
T h is is c a lle d . B e r n o u lli's e q u a tio n .
Besides r e c t i l i n e a r m o tio n , a n o th e r im p o rta n t case o f the
m otio n o f a gas is s y m m e tric a l s p h e r ic a l m o tio n in w h ich th e ve­
lo c it y
is everywhere in th e d ir e c t io n o f the r a d ic a l lin e s from a
f ix e d p o in t w hich i s th e c e n te r o f a system o f spheres upon each
o f w hich th e v e l o c i t y , d e n s ity , and p re s s u re are c o n s ta n t,
i f we
adopt s p h e r ic a l c o o rd in a te s and take the c e n te r p o in t as the p o le ,
a l l the q u a n t it ie s in v o lv e d become fu n c tio n s s o le ly o f r and t .
Then, i f u denote th e v a lu e o f th e r a d ic a l v e l o c i t y ,
th e e q u a tio n
o f c o n t in u it y is :
(7)
and the dynam ical e q u a tio n is :
(8 )
-4 -
I f the m otio n is ste a d y u,^0 , and p a re fu n c tio n s s o le ly o f r ,
th e
above p a r t i a l d i f f e r e n t i a l e q u a tio n s become o r d in a r y and we o b ta in
by in t e g r a t io n :
(9)
(10)
I n case the gas is a cte d upon by a system o f fo rc e
th ro u g h o u t i t s
volum e, we must s u b s t it u t e i n th e dyn am ical
e q u a tio n s f o r -
!
d t ° , iv,
/
5 / where 7) Jo is th e deJ z ~ 'W
r iv A t iv e o f th e p re s s u re in a d ir e c t io n n and Jn th e component o f
the r e s u lt a n t o f a l l th e fo rc e s p e r u n i t mass a c t in g a t a p o in t
x , y , z in th e same d ir e c t io n ,
i ’o r the e q u a tio n s in r e c ta n g u la r
c o o rd in a te s we must s u b s t it u t e in e q u a tio n s ( 2 ) :
(ID
where X,Y ,Z a r e 'th e c o o rd in a te s o f th e r e s u lt a n t o f a l l th e fo rc e s
p e r u n it mass a c tin g a t the p o in t x , y , z a t a tim e t and where each
is supposed to be g iv e n as a f u n c tio n o f x , y , z , t .
To summarize, i have in t h i s s e c tio n g iv e n the d i f f e r e n t i a l
e q u a tio n s o f m o tio n , w h ich must be s a t is f ie d by th e v e l o c i t y , den­
s i t y , and p re s s u re .
These th re e q u a n t it ie s are th e d e te rm in in g
fa c to r s o f the m o tio n o f a gas in the system o f if u le r .
1 have g iv e n
the reduced e q u a tio n s f o r the s p e c ia l cases o f r e c t i l i n e a r and sym­
m e tr ic a l s p h e ric a l m o tio n and have g iv e n the e q u a tio n s o f s te a d y
m o tio n in these two cases,
l have a ls o in d ic a te d th e changes in
the g e n e ra l e q u a tio n s when fo rc e s are a llo w e d to a c t upon th e gas
th ro u g h o u t i t s
volum e.
-5 I I System o f Lagrange
L e t us c o n s id e r an elem ent o f the gas w h ic h , in the i n i t i a l
c o n d itio n was a t the p o in t a , b , c w it h re s p e c t to a f ix e d system o f
r e c ta n g u la r axe s.
A t a tim e t t h i s elem ent o f gas w i l l have moved
to a new p o s it io n x , y , z w it h re s p e c t to th e same system o f axes.
Then i f we d e te rm in e x , y , and z each as a f u n c tio n o f a , b , c , t , we
know the p o s it io n o f e ve ry elem ent o f th e gas a t e v e ry in s t a n t .
If
we a ls o know th e d e n s ity and p re s s u re a t e ve ry p o in t o u r d e s c r ip tio n
o f th e m o tio n is com plete f o r the te m p e ra tu re i s d e te rm in e d by th e
c h a r a c t e r is t ic e q u a tio n o f th e gas as a f u n c tio n o f the d e n s ity and
p re s s u re . We may, th e n , express our d e s c r ip tio n o f the m o tio n
m a th e m a tic a lly in the form :
x = f 1 ( a ,b ,c ,t )
y = f 2(a ,b ,c ,t)
z = f 3 (a ,b ,c , t)
= F ( a ,b ,c ,t)
ρ
p = π (a ,b ,c ,t)
We can e a s ily see t h a t these fu n c tio n s cannot be p e r f e c t ly a r b i t r a r y .
It
is e v id e n t t h a t th e d e n s ity in th e in s ta n ta n e o u s s t a t e , compared
w ith t h a t in the o r i g i n a l s t a t e , v a r ie s w ith the clo se n e ss o f the
p a ckin g w hich the elem ents o f gas undergo and the clo se n e ss o f the
pa ckin g depends on th e fu n c tio n s f]_, f ^ and f ^ .
T h is in te rde pe nd en ce
is expressed by th e e q u a tio n o f c o n t in u it y :
( 12)
-6 Where JD is the d e n s ity in th e o r i g i n a l s ta te a t th e p o in t a , b , c .
A g a in , th e gas is always a c c e le ra te d in th e d ir e c t io n o f tne
p re s s u re g r a d ie n t and hence th e re i s an in te rd e p e n d e n ce between the
p re s s u re d i s t r i b u t i o n and th e m o tio n o f th e gas as expressed by
th e fu n c tio n s f i , f 2 , and f
by the dyn am ical e q u a tio n .
•
T h is in te rd e p e n d e n ce is expressed
I s h a ll here g iv e th d s e q u a tio n i n a
form in v o lv in g b o th th e E u le r ia n and la g ra n g ia n form s o f re p re s a n ta t io n . L e t us use the sym bol^ to denote d i f f e r e n t i a t i o n i n th e f o r ­
mer system and 'b f o r d i f f e r e n t i a t i o n i n th e l a t t e r system .
Then th e
dynam ical e q u a tio n s a re :
da)
D e r iv a tiv e s in th e two systems are r e la t e d by the f o llo w in g fo rm u la e :
(14)
l e t us c o n s id e r th e s p e c ia l case o f r e c t i l i n e a r m o tio n
and take as th e d i r e c t io n o f m o tio n th a t o f th e x a x is .
S ince th e
c o n d itio n o f the gas is th e same a t a l l p o in ts h a v in g th e same abs<iis$a
x a t the same tim e , x is a fu n c tio n s o le ly o f a and t , and y and z
rem ain eq ua l to t h e i r o r i g i n a l v a lu e s b and c r e s p e c t iv e ly . Then th e
e q u a tio n o f c o n t in u it y becomes:
(15)
We have le x t one d yn a m ica l e q u a tio n :
Since a r e l a t io n between d e r iv a tiv e s in the two systems is :
-7 -
We may s u b s t it u t e
in the above e q u a tio n and we o b ta in :
(16)
I n case the gas is a cte d upon by a system o f fo rc e s
th ro u g h o u t i t s
volum e, we must s u b s t it u t e in th e d yn am ical e q u a tio n s :
(17)
where X ,Y ,Z , are th e components o f th e r e s u lt a n t o f a l l th e fo rc e s
p e r u n i t mass a c tin g on an elem ent o f gas whose c o o rd in a te s are
a ,b ,c a t a tim e t and where each is supposed to be g iv e n as a fu n c tio n
o f a ,b ,c ,t.
To summarize, 1 have i n t h is s e c tio n g iv e n th e d i f f e r e n t i a l
e q u a tio n s o f m o tio n , w hich must be s a t is f ie d by the p o s it io n s o f the
p a r t ic le s , the d e n s ity , and th e p re s s u re .
These th re e q u a n t it ie s
are th e d e te rm in in g f a c t o r s o f th e m o tio n o f a gas in th e system o f
Lagrange.
m o tio n .
1 have g iv e n th e reduced e q u a tio n s f o r th e case o f r e e t lin e a r
I have a ls o in d ic a te d th e changes in the g e n e ra l e q u a tio n s
when fo rc e s are a llo w e d to a c t upon the gas th ro u g h o u t i t s
volum e.
-8 B.
D is c o n t in u it ie s
One o r more of the q u a n t it ie s used i n d e s c r ib in g th e m o tio n of
a gas may a t a c e r t a in tim e he d is c o n tin u o u s a t c e r t a in p o in ts o r the
d e r iv a tiv e s o f these q u a n t it ie s may he d is c o n tin u o u s a t a c e r t a in
tim e and a t c e r t a in p o in t s . The d is c o n tin u itie s w h ich I w ish to d is ­
cuss are those t h a t o ccu r a t is o la te d s u rfa c e s w ith o u t s i n g u l a r i t i e s .
I t may appear a t f i r s t s ig h t as i f such d is c o n t in u it ie s a re e x c e p tio n a l
bu t I s h a ll l a t e r show t h a t the occurence o f such d is c o n t in u it ie s i s
g e n e ra l and is o f im p o rta n ce in th e s tu d y o f th e m o tio n o f a gas. A
sep ara te a n a ly s is must he made as the e q u a tio n s o f m o tio n do n o t a p p ly
to th e d is c o n t in u it y .
I.
G eneral P ro p e rtie s o f D is c o n t in u it ie s and Te rm in o lig y
to he Adopted °
L e t ϕ be some q u a n t it y in v o lv in g th e c o n d itio n o f th e gas
w hich i s d is c o n tin u o u s a t a s u rfa c e , whose e q u a tio n in i n i t i a l c o o rd in a te s
is :
f(a ,b ,c ,t)
= 0
(18)
and in term s o f in s ta n ta n e o u s c o o rd in a te s is :
f 1( x , y , z , t )
= 0
(19)
L e t us te rm th e two re g io n s on e it h e r s id e o f t h is s u rfa c e th e re g io n s
1 and 2 .
Now ϕ
is c o n tin u o u s in b o th th e se re g io n s and th e r e fo r e as
we approach a p o in t on the s u rfa c e o f d is c o n t in u it y by any p a th w hich
l i e s w h o lly in re g io n 1, ϕ
c a ll ϕ 1 .
approaches a d e f in i t e l i m i t w hich we s h a ll
L ik e w is e as we approach a p o in t on the s u rfa c e o f d is c o n t in u it y
° Lecons s u r la H ro p o g a tio n des Ondes a t le s E o u a tio n s de I 'H y d r o dynamiaue, 1903; pp . 83 -85 , 87.
-9 by any p a th w h ich l i e s w h o lly in re g io n 2, ϕ
w h ich we s h a ll c a l l
ϕ 2 . Thusϕ
v a rie s c o n tin u o u s ly u n t i l we rea ch
th e s u rfa c e o f d is c o n t in u it y where i t
tude
approaches a d e f in i t e l i m i t
ta ke s a sudden jump o f m agni­
- ϕ 2 , w hich we s h a ll d e s ig n a te by th e symbol [ ϕ ] a f t e r w n ich
1
ϕ
i t a g a in v a r ie s c o n tin u o u s ly .
Now two p o in t s , P and Q, on the s u rfa c e o f d is c o n t in u it y
may be connected by pa th s i n re g io n s 1 and 2 w h ich d i f f e r by a d is ta n c e
as s m a ll as we p le a se from a p a th ly i n g in th e s u rfa c e o f d is c o n tin ­
u ity .
L e t th e two va lu e s o f ϕ a t P be ϕ 1 P , ϕ 2 P and the two va lu e s
a t Q be ϕ1Q, ϕ2Q. A lo ng th e p a th in re g io n 1 ϕ changes c o n tin u o u s ly
fro m ϕ1P to ϕ1Q and a lo n g th e p a th i n r e g io n 2 <js changes c o n tin u o u s ly
from ϕ2P t o ϕ 2 Q .
We may choose o u r p a th s such t h a t as the p o in t Q
i s made to approach the p o in t P a lo n g a p a th in t he s u rfa c e o f
d is c o n t in u it y , the le n g th s o f each o f these p a th s approaches z e ro .
As the le n g th s o f the se p a th s approach z e ro , s in c e f )
upon each o f them, <jf ^ must approach j)t ^ and
Thus
<f)t
is c o n tin u o u s
p must approach j %^
•
, must be a c o n tin u o u s f u n c tio n on th e s u rfa c e o f d is c o n ­
t i n u i t y and a ls o
must be c o n tin u o u s on t h is s u rfa c e .
The L a g ra n g ia n system o f r e p re s e n ta tio n is most u s e fu l
in the s tu d y o f d i s c o n t i n u it ie s .
The q u a n t it ie s x , y , z o r any o f t h e i r
d e r iv a tiv e s in th e la g ra n g ia n system o f c o o rd in a te s may be d is ­
co n tin u o u s when th e se v a lu e s are a ssig n e d to t h e i r p o s itio n s i n a
sphere whose c o o rd in a te s are a , b , c .
As a m a tte r o f te rm in o lo g y ,
we s h a ll term th e o rd e r o f th e d is c o n t in u it y th e o rd e r o f th e lo w ­
e s t d e r iv a tiv e w hich s u f f e r s a d is c o n t in u it y .
is o f the n ‘ th o rd e r and i f
t h is
Thus
3 ^ X_______
is th e lo w e s t o rd e r d e r iv a t iv e w hich
is d is c o n tin u o u s a t the d is c o n t in u it y , th e d is c o n t in u it y is o f th e n 't h
o rd e r.
(1 0 )
To sum m arize, we have pro ven an im p o rta n t p r o p e r ty o f a
d is c o n t in u it y ,
the c o n t in u it y o f the two v a lu e s th e re assumed ove r
the s u rfa c e o f d is c o n t in u it y .
T h is a llo w s us to form a t o t a l de­
r iv a t i v e o f e it h e r o f these v a lu e s on th e s u rfa c e .
We have a ls o i n ­
d ic a te d the arrangem ent o f d is c o n t in u it ie s a c c o rd in g to t h e i r o rd e r.
I I . G e n e r a lity o f th e E x is te n c e o f a D is c o n t in u it y i f
th e
P ressure is a F u n c tio n w h o lly o f th e D e n s ity 0
I f in a gas, th ro u g h o u t i t s
supposed to d i s t r i b u t e
m o tio n , h e a t developed is
i t s e l f im m e d ia te ly so t h a t th e gas is always
a t a c o n s ta n t te m p e ra tu re , th e p re s s u re o f t h e gas a t e ve ry p o in t
is p r o p o r tio n a l to the d e n s ity , in accordance w ith Bo y le 's law :
(20)
If,
on th e o th e r hand, th e h e a t c o n d u c t iv it y o f th e gas may he con­
s id e re d z e ro , the change o f s t a te o f each elem ent o f gas is a d ia b a tic
and th e p re s s u re v a rie s as th e d e n s ity to t h e Y power where Y = Sp/S
v
is the r a t i o
o f th e s p e c if ic h e a t a t c o n s ta n t p re s s u re to t h a t a t
c o n s ta n t volume, i n accordance w ith P o is s o n 's a d ia b a tic la w , nam ely:
(21)
We may w r it e to ta ke care o f e it h e r case:
C22)
Now in any m ech an ica l system , th e d i f f e r e n t i a l e q u a tio n s of m o tio n
a llo w us to c a lc u la te th e a c c e le r a tio n s o f i t s
component p a r t ic le s
a t any in s t a n t when we have g iv e n the p o s it io n s o f a l l th e se p a r­
t i c l e s Land t h e i r v e lo c it ie s a t t h a t in s t a n t s u b je c t to th e c o n d itio n
° Hadgmard, lo c , c i t .
pp . 139, 142.
(11)
th a t the p o s itio n s o f th e d i f f e r e n t p a r t ic le s and t h e i r v e lo c it ie s
must s a t is f y the c o n s tr a in ts imposed upon th e system .
m echanical system w ith an i n f i n i t e
number o f degrees o f freedom . In
the case o f th e m o tio n o f a gas, e q u a tio n s (1 2 ),
g iv e ilipon e lim in a t io n
o f p and
A gas is a
(13) and (E2)
the components o f a c c e le r a tio n
, (Y*-^ i f we are sciven x ,y ,a n d z as fu n c tio n s o f a ,b ,c
'S t*
a t th e tim e a t w h ich we are making ou r o b s e rv a tio n and a ls o
»
3 i:
d V _, ?) ^ , as fu n c tio n s o f a ,b ,e a t th e same tim e .
The c o n s t r a in t imposed upon th e gas is t h a t a t th e bounding
w a ll, th e elem ents o f gas in c o n ta c t w ith th e w a ll must move so as
to rem ain always i n c o n ta c t w ith t h is w a ll. We s h a ll he re exempt from
d is c tis s io n a t t h is tim e eases where th e v e lo c it y o f th e bounding w a ll
is oneof decom pression and o f such r a p i d i t y th a t th e gas is n o t a b le
to f o llo w the bounding w a ll b u t le a v e s a vacuum between th e s u rfa c e
o f th e bounding w a ll and th e s u rfa c e o f th e gas.
L e t th e e q u a tio n
o f th e bounding s u rfa c e be:
f
(x ,y ,z ,t)
* 0
(23)
S ince a p a r t i c l e , whose i n i t i a l c o o rd in a te s are a ,b ,c and w h ich o r i g i n a l ly
l i e s on t h is s u rfa c e , must c o n tin u e to l i e
on t h is s u rfa c e , by d i f ­
f e r e n t ia t in g th e e q u a tio n o f t h is s u rfa c e w it h re s p e c t to tim e , we o b ta in
as a ne cessa ry c o n d itio n upon th e g iv e n v a lu e s o f th e v e lo c it y a t
p o in ts on the boundary
the e q u a tio n :
(24)
We suppose th a t the g iv e n va lu e s o f th e v e lo c it y a t the boundary s a t is f y
t h is e q u a tio n .
D i f f e r e n t i a t i n g once more w it h re s p e c t to tim e we o b ta in :
(25)
-1 2 Now th e components o f a c c e le r a tio n are a lre a d y d e term in ed by the
e q u a tio n s o f m o tio n (13) th ro u g h o u t th e volume and hence upon th e bound­
in g s u rfa c e .
The compnnents o f a c c e le r a tio n th u s de term in ed w i l l n o t,
in g e n e ra l, s a t i s f y t h is e q u a tio n i f th e m o tio n o f th e bounding w a ll,
and hence th e f u n c tio n f , may be c o n s id e re d as a r b i t r a r i l y chosen.
The c o n t r a d ic t io n here in v o lv e d may be re s o lv e d by th e f o llo w in g con­
c e p tio n .
We may approach the boundary a lo n g a p a th ly i n g w h o lly in the
enclosed gas.
As we approach th e boundary a lo n g t h i s p a th by a d is ta n c e
as s m a ll as we p le a se , th e a c c e le r a tio n is g iv e n th ro u g h o u t by th e
e q u a tio n s o f m o tio n (13) b u t th e p a r t i c l e a t th e te r m in a l p o in t on th e
boundary has an a c c e le r a tio n w h ich s a t is f ie s e q u a tio n (25)
S ince
a re p r o p o r tio n a l to th e d ir e c t io n c o sin e s o f th e norm al
S t b
2
to the* s u rfa c e (19)
we see t h a t e q u a tio n (25) a f f e c t s o n ly th e norm al
component o f the a c c e le r a tio n .
if
the norm al component o f a c c e le r a tio n
o b ta in e d as a l i m i t by approach a lo n g a p a th fro m w it h in the gas does
n o t agree w ith the va lu e o b ta in e d fnrom e q u a tio n (2 5 ), th e re e x is ts a
d is c o n t in u it y a t th e bounding s u rfa c e ,
, b e in g the no rm al component
o f a c c e le r a tio n , o b ta in e d from e q u a tio n (25) and (p^ b e in g th e v a lu e o f
t h is component o b ta in e d as a l i m i t by approach fro m w it h in th e gas.
Thus in t h is case a second o rd e r d is c o n t in u it y is generated and t h is
w i l l be propogated in t o th e gas a c c o rd in g to c e r t a in laws w h ich we
s h a ll f u r t h e r in v e s t ig a t e .
L e t us suppose t h a t the va lu e s o f th e a c c e le r a tio n o b ta in e d
from e q u a tio n s (13) agree w ith e q u a tio n (25)
have f u r t h e r c o n d itio n s to be f u l f i l l e d .
on the boundary. We then
When th e p o s it io n s and ve -
1
D i f f e r e n t i a l and in t e g r a l C a lc u lu s - G r a n v ille , 1911: p . 266
-1 3 -
lo c it ie s
o f the p a r t ic le s are g iv e n n o t o n ly th e second d e r iv a tiv e s o f
x , y , and z may he o b ta in e d b u t e ve ry d e r iv a t iv e .
The second d e r iv a t iv e s ,
∂2x/∂a2, ∂2x/∂a∂b, ∂2x/∂a∂c, ∂2x/∂b2, -----, ∂2z/∂c2 a re known because x , y , and z
are g iv e n as fu n c tio n s o f a , b , c .
The second d e r iv a t iv e s , ∂2x/∂a∂t, ∂2x/∂d∂t,
----------- , ∂2z/∂b∂t, ∂2z/∂c∂t a re known because ∂ x / ∂ t , ∂ y / ∂ t , and ∂ z / ∂ t are
g iv e n as fu n c tio n s o f a , b , c .
As shown in th e p re c e e d in g pa ra g ra p h th e
second d e r iv a t iv e s , ∂2x/∂t2, ∂2y/∂t2 and ∂ 2 z /∂ t2 a re known fro m the e q u a tio n s
o f m o tio n (1 3 ).
o f a ,b ,c .
Thus a l l th e second d e r iv a tiv e s are known as fu n c tio n s
T his a llo w s us to c a lc u la te
by d i f f e r e n t i a t i o n a l l the
t h i r d d e r iv a tiv e s e xce p t ∂ 3 x /∂ t3 , ∂ 3 y /∂ t3 and ∂ 3 z /∂ t3 .
By d i f f e r e n t i a t i n g
each o f e q u a tio n s (13) w ith re s p e c t to t we o b ta in these d e r iv a tiv e s
and hence a l l the t a i r d d e r iv a tiv e s are known.
E q u a tio n (25) may be
d if f e r e n t ia t e d w ith re s p e c t to t f o r a c o n d itio n on the boundary and
i f t h is r e l a t io n is n o t s a t is f ie d by the t h i r d o rd e r tim e d e r iv a tiv e s
determ ined fro m th e g iv e n s ta te o f th e gas, th e re is a t h ir d
d is c o n t in u it y a t th e bounding s u rfa c e .
o rd e r
A l l th e f o u r t h o rd e r d e r iv a tiv e s
except ∂4x/∂t4, ∂4y/∂t4, and ∂4z/∂t4 a re found by d i f f e r e n t i a t i n g th e t h i r d
o rd e r
d e r iv a tiv e s w it h re s p e c t to a ,b ,c and hence are known, w h ile th e se de­
r iv a t iv e s are known by d i f i e r e n t i f c t i n g tw ic e e q u a tio n s (1 3 ).
T his
process can o b v io u s ly be re p e a te d and hence a l l th e d e r iv a tiv e s o f x ,
y , and z are known.
L ik e w is e by d i f f e r e n t i a t i n g s u c c e s s iv e ly e q u a tio n
(25) we o b ta in added c o n d itio n s upon the se d e r iv a tiv e s so t h a t even
sho uld th e t h i r d d e r iv a tiv e s a t the boundary as d e te rm in e d from th e g iv e n
s ta te agree w ith the boundary c o n s t r a in t , th e f o u r t h o rd e r d e r iv a tiv e s
m ig h t n o t agree and we sh o u ld have a f o u r t h o rd e r d is c o n t in u it y .
Thus,
in g e n e ra l, an a r b it r a r y m o tio n o f th e boundary im p lie s a d is c o n t in u it y
in the gas.
In th e f u r t h e r p ro g re s s o f th e m o tio n beyond th e g iv e n s ta te
should th e boundary w a lls d e s c rib e a m o tio n d is c o n tin u o u s a t a c e r t a in
-1 4 time t in th e n 't h o rd e r d e r iv a t iv e s , an n ’ th o rd e r d is c o n t in u it y w i l l be
formed in th e gas.
To summarize, i f
th e p o s itio n s and v e lo c it ie s o f th e p a r t ic le s
o f gas are g iv e n so as to be c o m p a tib le w ith the p o s it io n and v e lo c it y
o f the bounding w a lls , th e a c c e le r a tio n and tim e d e r iv a tiv e s o f h ig h e r
o rd e r w i l l n o t, in g e n e ra l, be c o m p a tib le w it h the m o tio n o f the
bounding w a lls and d is c o n t in u it ie s o f the second o r h ig h e r o rd e r
w i l l ensue.
III
K in e m a tic a l R e la tio n s P e r t a in in g to a D is c o n t in u it y
I w is h f i r s t to ta ke up th e v e lo c it y o f p ro p a g a tio n
o f a d is c o n t in u it y in two systems o f measurement.
The one system o f
measurement has to do w ith th e e q u a tio n o f th e s u rfa c e o f
d is c o n t in u it y in th e fo rm :
f
(a ,b ,c ,t ) = 0
(26)
and the o th e r in th e fo rm :
ϕ (x ,y ,z ,t)
=
0
L e t us f i r s t c o n s id e r th e fo rm e r system .
(27 )
In t h i s system
we im agine th e v a lu e s x , y , z a ssig n e d to e v e ry p o in t in a space whose
c o o rd in a te s are a , b , c .
The d is c o n t in u it y i s s it u a t e d on a moving
s u rfa c e in t h i s space whose e q u a tio n we may re p re s en t by (2 6 ).
We
may im agine a p o in t a ,b ,c w h ich moves w it h a v e l o c i t y alw ays norm al
to th e s u rfa c e ( 3 f ) and so as to rem ain alw ays in t h i s s u rfa c e .
The
■V elocity o f t h i s p o in t we s h a ll c a l l th e v e lo c it y o f pro pa g a tio n o f
th e d is c o n t in u it y w it h re s p e c t to th e i n i t i a l s ta te a t th e p o in t
c o n s id e re d .
We s h a ll d e s g in a te t h i s v e lo c it y by th e l e t t e r θ
1. Hadama^d, lo c . c i t . , pp. 101-105.
I have e la b o ra te d on th e p r o o f o f
the fo rm u la f o r th e v e lo c it y o f p ro p o g a tio n w h ich Hadamard o n lv
b r i e f l y o u t lin e s .
-1 5 -
The li n e
jo in in g a p o in t a ', b ' , c ' o u ts id e the s u rfa c e o f d is c o n t in ­
ui t y to th e p o in t a, b , c on t h i s s u rfa c e , c o n s id e re d as a v e c to r ,
has components, a - a ', b - b ', c- c ' ,
The p r o je c t io n o f t h i s l i n e segment
upon th e norm al a t a , b , c i s th e sum o f th e p r o je c tio n s of these components
on
th e n o rm a l. To fo rm th e p r o je c t io n
o f any component we must m u lt ip ly
i t by th e d ir e c t io n co sin e o f th e norm al w it h re s p e c t to th e c o o rd in a te
a xis c o rre s p o n d in g to t h i s component. Now th e d ir e c t io n c o s in e s o f th e n o r­
mal a re :
si n ce th e y are p r o p o r tio n a l t o ∂f/∂a, ∂f/∂b, ∂f/∂c1 and s in c e th e sum o f t h e i r
squares must e q u a l one.
Then the p r o je c t io n o f the li n e
segment upon
the norm al i s :
(28)
If th e p o in t a ' , b ’ , c ' is f ix e d and th e p o in t a ,b ,c is moving w ith a ve­
lo c it y whose components are d a /dt, db/dt, dc/dt, th e n the r a t e o f change o f th e
normal component o f th e l i n e
segment i s o b ta in e d by d i f f e r e n t i a t i n g th e
above e x p re s s io n f o r n , nam ely:
1.
D i f f e r e n t i a l and I n t e g r a l C a lc u lu s - G r a n v ille , 1911 p . 266.
-1 6 Now when th e p o in t a ' , b ' c ' c o in c id e s w it h th e p o in t a , b , c , i t
is ob­
v io u s th a t d n /d t i s th e v e l o c i t y o f th e p o in t a , b , c such as we have
d e s c rib e d i t
and hence i s e q u a l t o th e v e l o c i t y o f p ro p o g a tio n θ a t
t h is p o in t .
D i f f e r e n t ia t in g e q u a tio n (25) o f th e s u rfa c e , we o b ta in
f o r th e p o in t a , b , c , s in c e i t
lie s
on t h i s
s u rfa c e , th e ne ce ssa ry con­
d itio n :
(30)
S u b s t it u t in g th e sum o f th e f i r s t th r e e term s o b ta in e d fro m t h i s equ­
a tio n in th e f i r s t te rm o f th e r i g h t hand member o f (29) and p la c in g
a = a ’ , b = b ' , c = c ', and dn/dt = θ we o b ta in :
(31)
We s h a ll adopt th e fo rro w in g c o n v e n tio n as t o s ig n . The f u n c t io n
f i n the l e f t member o f (26) s h a ll be p o s it iv e i n re g io n 2. The
v e lo c it y o f p ro p a g a tio n s h a ll be c o n s id e re d p o s it iv e o r n e g a tiv e
a c c o rd in g as th e s u rfa c e o f d is c o n t in u it y moves fro m re g io n 1 in t o
re g io n 2 o r i n th e in v e rs e d ir e c t io n .
By these same methods we may f i n d th e v e lo c it y o f p ro p a g a tio n
T o f the s u rfa c e o f d is c o n t in u it y w it h re s p e c t to th e a c tu a l s ta te
t
m e re ly u s in g e q u a tio n (27) o f th e s u rfa c e and c o o rd in a te s x , y , z
in s te a d o f a , b , c . The v e l o c i t y o f p ro p o g a tio n T i s th e in s ta n ta n e o u s
v e lo c it y o f a p o in t in th e s u rfa e e
norm al to th e s u rfa c e . F o r t h i s
v e lo c it y we have a v a lu e analogous t o Q i n e q u a tio n ( 3 1 ) , nam ely
(32)
The v e lo c it ie s T
and #
a re d i s t i n c t
even i f th e i n i t i a l s ta te
-1 7 c o in c id e s w it h the a c tu a l s ta te a t th e in s t a n t c o n s id e re d ,
case x , y , z are r e s p e c t iv e ly e q u a l to a ,b ,c a t t h is
ϕ
and f are eq u a l a t t h is
in s t a n t .
how in t h i s
in s t a n t and henc e
T h e re fo re ∂ϕ /∂x, ∂ϕ /∂y, ∂ϕ /∂z are
r e s p e c tiv e ly eq u a l to ∂ f/∂ a , ∂ f/∂ b , ∂ f/∂ c . On th e o th e r h a n d we have:
(33)
where u ,v ,w are components o f th e v e l o c i t y .
D iv id e t h i s e q u a tio n
th ro u g h o u t by √((∂ϕ/∂x)2+(∂ϕ/∂y)2+(∂ϕ/∂z)2) o r i t s e
c.
b
+
(∂f/)2
l√
a
u
q
Then by v ir t u e o f the s t atem ent p ro c e e d in g e q u a tio n (2 8 ), th e f i r s t
th re e term s o f the r i g h t hand member o f e q u a tio n (35) become the norm al
component o f th e v e lo c it y a t th e d is c o n t in u it y w hich we s h a ll d e s ig n a te
by 1^,
.
A lso u s in g e q u a tio n s (31) and (3 2 ), e q u a tio n (33) becomes:
or
(34)
The p r i n c i p a l o f c o m p o s itio n o f v e lo c it ie s would a ls o serve to g iv e us
t h is r e s u lt f o r when the i n i t i a l s ta te is ta ke n c o in c id e n t w ith the
a c tu a l s t a t e , th e v e lo c it y 0 is th e v e lo c it y o f p ro p a g a tio n o f th e s u r­
face o f d is c o n t in u it y w ith re s p e c t to th e medium in i t s
a c tu a l s ta te and
the a c tu a l v e lo c it y o f p ro p a g a tio n T may be c o n s id e re d as th e r e s u l t ­
a n t o f th e v e lo c it y o f p ro p a g a tio n 0 w it h re s p e c t to th e medium and the
v e lo c it y , UVi o f the medium i t s e l f .
Under th is head I w is h a ls o to d is c u s s th e k in e m a tic a l r e l a t io n
e x is t in g between th e s e v e ra l n ' t h o rd e r d e r iv a tiv e s in an n 't h o rd e r
d is c o n t in u it y ,
l e t <J) be a q u a n t it y w hich has an n ’ t h o rd e r d is c o n ­
t i n u i t y a t a s u rfa c e re p re s e n te d by e q u a tio n ( 2 $ ) .
l s h a ll d e s ig n a te
by s u b s c rip ts 1 and 2 th e v a lu e s o f q u a n t it ie s p e r t a in in g to th e r e ­
g io n s 1 and 2 on e it h e r s id e o f th e s u rfa c e o f d is c o n t in u it y .
Then
agree in v a lu e and so do c o rre s p o n d in g d e r iv a tiv e s o f o rd e r
le s s tha n n o f these q u a n t it ie s .
Thus we have.
-1 8 -
on the s u rfa c e o f d is c o n t in u it y f o r p , q , r , h , a n y p o s it iv e in te g e rs
whose sum is n - 1 . D i f f e r e n t ia t in g t o t a l l y w ith re s p e c t to tim e ,
we o b t a in : -
Now a c c o rd in g to o u r assum ption o f th e e x is te n c e o f an n' th o rd e r
d is c o n t in u it y , s im i la r l y p la c e d term s in th e two members o f t h is equa­
t io n are n o t, in g e n e ra l , e q u a l.
B rin g in g a l l the term s to one s id e
o f the e q u a tio n we may w r ite :-
where the b ra c k e ts denote th e v a r ia t io n s o f the q u a n t it ie s th e y enclose
a t the s u rfa c e o f d is c o n t in u it y .
Now e q u a tio n (3 0 ) must lik e w is e h o ld
s in c e the t o t a l d i f f e r e n t i a t i o n can o n ly a p p ly on th e s u rfa c e (26) .
D iv id in g the e q u a tio n (30) th ro u g h o u t by √((∂f/∂a)2 +(∂t/∂b)2 +(∂f/∂c)2) and making use
o f the sta te m e n t p re c e d in g (28) and o f e q u a tio n (3 1 ), we o b t a in : (36)
w h e re α
γare th e d ir e c t io n co sin e s o f th e norm al a t the p o in t c o n s id e re d .
,β
S ince da/dt, db/dt, dc/dt, in e q u a tio n s (35) and (36)
s t i l l rem ain a r b i t r a r y ,
the c o e f f ic ie n t s o f the se q u a n t it ie s and th e n o n - a r b itr a r y term must be
p r o p o r tio n a l in these two e q u a tio n s and th e r e fo r e we h a v e :-
S ince the r e s u lt i s v a lid f o r p , q , r , h any p o s it iv e in te g e r s whose
sum is n - 1, we may w r ite : (37)
P la c in g ϕ s u c c e s s iv e ly e
q
u
a
l to x , y , z and l e t t i n g th e v a lu e s o f the
c o n s ta n t in (37) b e λ , μ , ν , r e s p e c t iv e ly , we o b t a in : -
(38)
As a s p e c ia l case we have : -
-1 9 -
(39)
The l e f t members o f the above e q u a tio n s a re components o f a v e c to r
∂ η s /∂ tη and hence λ,μ
ν
a re components o f a v e c to r c o in c id e n t w ith t h is
v e c to r and c a lle d by Hadamard th e c h a r a c t e r is t ic segment.
Now we
see from (38 ) th a t as f a r as the v a r ia t io n i n th e n 't h d e r iv a tiv e s are
concerned, the d is c o n t in u it y depends o n ly upon th e c h a r a c t e r is t ic
segment and th e v e lo c it y 0
To summarize, I have developed fo rm u la e f o r the v e lo c it y
o f p ro p a g a tio n o f a d is c o n t in u it y w ith re s p e c t to the i n i t i a l and
a c tu a l s ta te o f the gas and have shown th e r e l a t io n between these two
v e lo c it ie s when th e i n i t i a l s ta te c o in c id e s w ith th e a c tu a l s ta te
a t th e in s ta n t c o n s id e re d .
I have a ls o developed n e ce ssa ry r e la t io n s
between th e n ' t h o rd e r d e r iv a tiv e s in an n ' t h o rd e r d is c o n t in u it y
and have shown th a t a d is c o n t in u it y is c h a ra c te riz e d by a c h a ra c te r­
is tic
segment a t e v e ry p o in t on i t s
s u rfa c e and a v e lo c it y o f p ro ­
p o g a tio n 0 a t e ve ry p o in t .
IV
I V . S ta tio n a r y D is c o n t in u it ie s 1
A s ta tio n a r y d is c o n t in u it y is one w h ich a f f e c t s alw ays th e
same p a r t ic le s , t h a t i s , a d is c o n t in u it y f o r w h ich e q u a tio n (26)
does n o t in v o lv e t o r f o r w hich
th e r e fo r e , th a t 0 is
z e ro .
b ± » 0 . We see from e q u a tio n (3 1 ),
Z>t
P la c in g 0 * o in e q u a tio n s (38) we see th a t
[∂ηx/(∂ad∂bk∂cl∂tm)], [∂ηy/(∂ad∂bk∂cl∂tm)], [∂ηz/(∂ad∂bk∂cl∂tm)] a re zero u n le s s m = o .
LdoiJ
^
I
Thus we see th a t ^ f o r a s t a t io n a r y d is c o n t in u it y , th e f i r s t d e r iv a tiv e s
th a t are d is c o n tin u o u s are those n o t in v o lv in g t .
the n ' t h o rd e r d e r iv a tiv e s
1. Hadamard
lo c . c i t . ,
T^x >
,
~ T r» '
MV
1903, p p .99-101
2 )\:
We th u s see th a t
can o n ly be d is c o n -
-2 0 tin u o u s i f x , y , z are d is c o n tin u o u s on th e s u rfa c e (2 6 ),
Then we may
ta ke two p o in ts a1, b1, c1 and a2 ,b 2 c 2 as c lo s e to g e th e r as we p le a s e
and se p a ra te d by the s u rfa c e o f d is c o n t in u it y , w h ile t h e i r a c tu a l
p o s itio n s x i , y _ , z_ and x , y g Zg a re a f i n i t e
d is ta n c e a p a r t,
le t
us c o n s id e r th e e q u a tio n o f the s u rfa c e in a c tu a l c o o r d in a te s , ( 2 7 ) .
Then s in c e the d is c o n t in u it y a fr e e ts alw ays th e same p a r t i c l e s , x^
y^ z^ and x g y g Zg must s a t is f y t h is e q u a tio n th ro u g h o u t th e m o tio n .
Hence a s t a t io n a r y d is c o n t in u it y re p re s e n ts a m o tio n o f th e gas on
e it h e r s id e o f th e d is c o n t in u it y as i f th e s u rfa c e o f d is c o n t in u it y
was a bounding s u rfa c e f o r the two masses o f gas on e it h e r s id e .
We
thu s see th a t th e v e lo c it y must alw ays be t a n g e n tia l to t h i s s u rfa c e .
Y.
D is c o n tin u o u s Waves
A wave ty p e o f d is c o n t in u it y i s one w h ich is p ro p o g a te d from
one p a r t ic le
to a n o th e r, t h a t i s , one f o r w h ich 6 is n o t z e ro , and hence
t h is typ e o f d is c o n t in u it y and th e s t a t io n a r y typ e are m u tu a lly ex­
c lu s iv e .
A d is c o n t in u it y in x , y j
and z i s th e r e fo r e e lim in a te d from
d is c u s s io n and we see fro m e q u a tio n s (38) th a t a d is c o n t in u it y o f the
wave type always a f f e c t s the tim e d e r iv a tiv e s o f x , y , z .
1. D e r iv a tiv e s o f th e n e n s ity 1 .
The d e n s ity is d e fin e d by the e q u a tio n o f c o n t in u it y (12)
I t i s , a c c o r d in g ly , a f u n c tio n o f th e f i r s t d e r iv a tiv e s o f x , y , and z
w ith re s p e c t to a , b , c .
A d is c o n t in u it y in the se f i r s t o rd e r d e r iv a tiv e s
pro du ces, in g e n e ra l, a d is c o n t in u it y in th e d e n s ity .
A t an n 't h o rd e r
d is c o n t in u it y , t h e ( n - l ) ’ 6ii o rd e r d e r iv a tiv e s o f th e d e n s ity a re the
d e r iv a tiv e s f i r s t a f f e c t e d .
E q u a tio n (12) is :
1. Hadamard lo c . c i t . pp. 7 0 .7 1 .1 1 3 .
-2 1 l e t us a tte m p t to fo rm th e v a r ia t io n o f an ( n - l ) s t o rd e r d e r iv a t iv e
of
ρ0/ρ a t a s u rfa c e o f d is c o n t in u it y , t h a t is :
(40)
where th e f i r s t term i n th e r i g h t hand member o f t h is e q u a tio n is an
( n - l ) s t o rd e r d e r iv a t iv e by approach fro m th e re g io n 1 and the second
term th e same d e r iv a t iv e by approach fro m re g io n 2 .
In fo rm in g the se
two term s by d i f f e r e n t i a t i n g e q u a tio n (1 2 ), term s c o n ta in in g o n ly
d e r iv a tiv e s o f x , y o r z o f o rd e r le s s th a n n may be o m itte d s in c e the
value s o f th e se d e r iv a tiv e s by approach from re g io n s 1 and 2 agree and
hence th e y d is a p p e a r fro m e q u a tio n (4 0 ).
The re m a in in g term s are made
up o f the ( n - l ) s t d e r iv a t iv e o f each elem ent m u lt ip lie d by the m in o r
o f th a t elem ent w ith the a lg e b r a ic s ig n c o rre s p o n d in g to th e p o s it io n
o f the elem ent in th e d e te rm in a n t.
A s im p l i f i c a t i o n i s in tro d u c e d
i f we choose as i n i t i a l c o o rd in a te s the a c tu a l c o o rd in a te s in re g io n
2 a t the in s t a n t c o n s id e re d .
Then in re g io n 2, x= a , y= b , z=c and these
e q u a tio n s may be d i f f e r e n t i a t e d w it h re s p e c t to a , b ,c f o r a l l h ig h e r
d e r iv a tiv e s n o t in v o lv in g t .
I n re g io n 2 we th e n have ∂x/a
,
zc1
yb
=
∂x/∂b = ∂x/∂c = ∂y/∂a = ∂y/∂c = ∂z/∂a = ∂z/∂c = o and a l l h ig h e r d e r iv a tiv e s w it h r e ­
sp e ct to a ,b ,c z e ro .
The d e r iv a tiv e s o f o rd e r le s s than n on th e d is c o n t in u it y by
approach from re g io n 1 agree w it h th e c o rre s p o n d in g d e r iv a tiv e s by
approach fro m re g io n 2 . The e v a lu a tio n ju s t made f o r th e d e r iv a tiv e s in
re g io n 2 thus h o ld s f o r d e r iv a tiv e s o f o rd e r le s s th a n n on th e d is ­
c o n t in u it y by approach fro m re g io n 1. l o r a d is c o n t in u it y o f the second
o r h ig h e r o rd e r we th u s have
∂=
y
c
a
z
b
/
o
∂ x /∂ a = ∂ y /∂ b = ∂ z /∂ c = 1 and ∂x/∂b = ∂x/∂c =
and the d e te rm in a n t D becomes one, a l l
elem ents e xce pt those on th e p r in c ip a l d ia g o n a l becoming one t
o f an elem ent n o t on th e p r i n c i p a l d ia g o n a l is zero w h ile th e
|| ~~
The m inor
m in o r o f
A,p"i
-2 2 a p r in c ip a l elem ent is one.
Thus th e o n ly term s in the r i g h t hand
member o f (40) re m a in in g are t h e ( n - l ) s t d e r iv a tiv e s o f th e th re e p r i n ­
c ip a l elem ents m u lt ip lie d by t h e i r re s p e c tiv e m in o rs w h ich are 1.
Thus
e q u a tio n (40) becomes:
S u b s t it u t in g th e va lu e s of the v a r ia t io n s in the r i g h t hand member from
e q u a tio n s (3 8 ), we have:
(41)
Now any fu n c tio n o f ρ may be expressed in the form F ( ρo/ρ) and an
( n - l ) s t d e r iv a t iv e o f t h is f u n c tio n is F'1-(ρo/ρ) m u lt ip lie d by th e co­
re sp o n d in g d e r iv a t iv e
I f we assume an
o f (ρo/ρ), t h a t is
:
i n i t i a l s ta te c o in c id e n t w ith th e a c tu a l s ta te in
re g io n 2 we have f o r a second o r h ig h e r o rd e r d is c o n t in u it y
ρ = ρ0 and we may the n ta k e v a r ia t io n s a t the d is c o n t in u it y o f the
above q u a n t it y . We th u s o b ta in , m aking use o f e q u a tio n (4 1 ):
2.
Second and H ig h e r O rder Dis c o n t in u i t ies1
L e t us le a v e d is c o n t in u it ie s o f th e f i r s t o rd e r f o r l a t e r
d is c u s s io n .
The e q u a tio n s (13) h o ld v a l id f o r th e re g io n s 1 and 2.
S u b s t it u t in g the va lu e o f p from (1 8 ), we o b ta in :
(43)
Now f o r a second o r h ig h e r o rd e r d is c o n t in u it y , i f
th e i n i t i a l
c o o rd in a te s are ta ke n as c o in c id e n t a l w it h the a c tu a l in s ta n ta n e o u s
c o o rd in a te s in re g io n 2, we see fro m the r e la t io n s
(14) t h a t the
t io n s , δ /δ x , δ /δ y , δ /δ z are e q u iv a le n t r e s p e c t iv e ly to
∂/∂a, ∂/∂b, ∂/∂c
1. Hadamar d , Sur l a P ro p o g a tio n des Ondes, 1903 p p .225-229.
opera­
on
-2 3 th e s u rfa c e o f d is c o n t in u it y .
Then f o r a second o rd e r d e s c o n tin u ity
e q u a tio n s (43) g iv e , s in c e ρ = ρ 0,
(44)
S u b s t it u t in g the v a lu e s o f the v a r ia t io n s in the l e f t members from
e q u a tio n s (38) and th e v a lu e s o f th o se in the r i g h t hand members from
( 4 1 ) , we o b ta in :
(45)
Since θ is n o t eq u a l to z e ro , we may d iv id e one e q u a tio n by a n o th e r and we
o b ta in :
λ :μ
: ν = α : β : γ
(46)
T h is shows t h a t the c h a r a c t e r is t ic segment is c o in c id e n t w ith the norm al
to th e s u rfa c e and hence we see t h a t the v a r ia t io n o f a c c e le r a tio n a t
the d is c o n t in u it y is a v e c to r p e rp e n d ic u la r to th e s u rfa c e .
q u a n tity λ α +μ
β
+ νγ
by i t s
The
form d e s ig n a te s th e p r o je c t io n o f the
c h a r a c t e r is t ic segment upon the no rm al b u t, s in c e th e c h a r a c t e r is t ic
segmen t is c o in c id e n t w ith th e n o rm a l, i t
fo rm e r.
is the a b s o lu te le n g th o f th e
I f th e r i g h t hand members o f th e above e q u a tio n s we have t h is
q u a n tity m u lt ip lie d by
, β
α
, γ
, w hich g iv e s th e p r o je c tio n s o f the
c h a r a c t e r is t ic segment upon th e c o o rd in a te axes, t h a t is
r e s p e c t iv e ly .
λ, μ,
ν
The th re e e q u a tio n s tha n reduce t o :
(47)
The v e lo c it y o f p ro p o g a tio n r e l a t i v e
to th e medium in i t s
a c tu a l s ta te
i s a c c o rd in g ly t h a t o f o r d in a r y sound o f i n f i n i t e s i m a l a m p litu d e 1 . I f
th e p re s s u re p, is n o t o n ly a f u n c tio n o f ρ b u t a ls o o f a , b , c in the
e q u a tio n s o f m o tio n used above we must have in th e r i g h t members
Lamb, Hydrodynam ics, 1911, p 455.
-2 4 ((1/ρ)(δp/δx)), ((1/ρ)(δp/δy)), ((1/ρ)(δp/δz)) r e s p e c t iv e ly .
S ince no d i s t i n c t i o n is in v o lv e d
b e tw e e n the c o o rd in a te axes we need o n ly c o n s id e r one o f the se ex­
p re s s io n s and r e s u lt s f o r th e o th e rs may be w r i t t e n down by c y c lic
change o f c o o rd in a te s .
U sing as i n i t i a l c o o rd in a te s th e a c tu a l in s t a n t ­
aneous c o o rd in a te s in re g io n 2, we have:
and
s in c e
∂p/∂ρ and ∂ p /∂ a are c o n tin u o u s a t th e d is c o n t in u it y .
On s u b s t it u t io n
o f t h is r e s u lt , the o n ly change a ffe c te d is a d isp la ce m e n t ofφ
'(ρ)
by
∂p/∂ρ.
Thus, even under the se c o n d itio n s
a c c e le r a tio n
, th e v a r ia tio n ;,
in
is norm al to th e s u rfa c e and th e v e l o c i t y o f p ro p o g a tio n
√(∂p/∂ρ).
We may g e n e ra liz e to a d is c o n t in u it y o f th e n ’ th o rd e r1 .
D i f f e r e n t ia t in g the e q u a tio n o f m o tio n w it h re s p e c t to t in th e i n i t i a l
system o f c o o rd in a te s n -2 tim e s and ta k in g v a r ia t io n s we o b ta in , i f the
i n i t i a l c o o rd in a te s a re th e in s ta n ta n e o u s c o o rd in a te s in re g io n 2,
(4 8 )
l e t us c o n s id e r o n e *o f th e q u a n t it ie s whose v a r ia t io n occu rs in th e r i g h t
member o f the above e q u a tio n s . R e s u lts f o r th e o th e rs may th e n be
deduced by c y c lic change o f c o o rd in a te s .
Thus
tThere A is an e x p re s s io n c o n ta in in g ^ d e r iv a tiv e s o f" ^
o f o rd e r le s s
tha n n -1 and d e r iv a tiv e s o f x , y , z o f o rd e r le s s th a n n .
From the la s t
e x p re s s io n we may tra n s fo rm to th e f o llo w in g e x p re s s io n :
Where B is
o f o ra e r le s s th a n n -1 i n
This was suggested by hadamard, lo c , e i t . ,
P0 and o f o rd e r le s s
b u t done here f a r th e f i r s t tim e *
-2 5 than n in x , y , z .
A t th e d is c o n t in u it y δ
/x
is e q u iv a le n t in o p e ra tio n
to ∂/∂a and in ta k in g v a r ia t io n s B is c o n tin u o u s and hence d is a p p e a rs .
Thus we o b ta in :
S im ila r ly , by c y c lic change o f c o o rd in a te s we have:
S u b s t it u t in g these e q u a tio n s in the e q u a tio n s (48) we o b ta in :
S u b s t it u t in g
le ft
the v a lu e s p r e v io u s ly fou nd f o r th e v a r ia t io n s in the
(38) and the r i g h t (41) members o f the e q u a tio n s , we o b ta in :
(49)
For θ n o t eq ua l to z e ro , we may d iv id e each o f the se e q u a tio n s by
(-θ )η
2
-
and th e e q u a tio n s become id e n t i c a l w it h the e q u a tio n s fou nd f o r
a second o rd e r d is c o n t in u it y .
The c o n c lu s io n s th e re drawn, th a t the
c h a r a c t e r is t ic segment is no rm al to th e s u rfa c e and t h a t th e v e lo c it y
o f p ro p a g a tio n is
√(φ'(ρ0)) th e r e fo r e s t i l l h o ld f o r a n ' t h o rd e r d is c o n ­
tin u ity .
3 . F i r s t O rder D is c o n tin u it y
a
. C o n tin u ity R e la tio n 1
l e t us now c o n s id e r a d is c o n t in u it y o f th e f i r s t o rd e r. L e t
Riemann, Uber d ie 1 o rtp flc ja z u n g ebener lu f t w e ll e n von e n d lic h e r SchwiBgangsw e ite - G o e ttin g e n Abhandlung t V I I I , 1860
-2 6 -
the v e lo c it y components, th e d e n s ity and th e p re s s u re
v a lu e s u 1
, v1, w1
, ρ1, p1 and u 2
, v2 , w
2
take on the
, ρ
,p2 , on the s u rfa c e o f d is c o n ­
2
t i n u i t y as we approach t h is s u rfa c e fro m th e re g io n s 1 and 2 r e s p e c tiv e ­
l y and l e t us choose th e x - a x i s c o in c id e n t w ith th e norm al to the
s u rfa c e and reckoned p o s it iv e i n the d ir e c t io n from the re g io n 1 to
the re g io n 2 .
L e t θ be th e v e lo c it y o f p ro p a g a tio n w ith re s p e c t to
any i n i t i a l s ta te in w hich the d e n s ity has the v a lu e ρ0 .
Then we have
from (41) and (3 8 ), s in c e α = 1, n = 1,
(50)
(51)
B lim in a t in g A
(52)
Ifθ
1
and θ2 are th e v e lo c it ie s o f p ro p a g a tio n w ith re s p e c t to the
re g io n s 1 and 2 reckoned as i n i t i a l s ta te s , we may s u b s t it u t e θ 1 a n d θ 2
f o r θ and ρ1 and ρ2 r e s p e c t iv e ly f o r ρ .
We thus see fro m th e above
e q u a tio n th a t the f o llo w in g r e l a t i o n must e x is t between these q u a l it ie s :
(53)
b . Dynam ical E q u a tio n and V e lo c it y Form ula
•
C o nside r a c y lin d e r o f i n f i n i t e s i m a l cro ss s e c tio n are a S
•T
norm al to the s u rfa c e o f d i s c o n t in u it y .
In a tim e d t a mass o f gasρ
t
d
S
θ
0
flo w s in t h is c y lin d e r acro ss th e s u rfa c e o f id is c o n t in u it y . T h is mass
o f gas may be co n sid e re d to be en clo se d by p is to n s in t h is tu b e .
D u rin g the passage o f t h is gas a cro ss th e s u rfa c e o f d is c o n t in u it y ,
one p is to n is on one s id e o f th e s u rfa c e and th e o th e r p is to n on th e
o th e r s id e .
Iffi,d t be ta k e n n e g lig ib le
in com parison w ith S, we may
n e g le c t the e f f e c t o f the p re s s u re on th e gas a cro ss the l a t e r a l w a ll
Riemann. lo c . c i t . .
-2 7 -
o f the c y lin d e r .
The fo rc e th e n a c t in g on t h i s gas is
w hich a c ts norm al to th e s u rfa c e .
A c c o rd in g to dyn am ical p r in c ip le s
the v a r ia t io n in momentum a t th e d is c o n t in u it y must he c o in c id e n t
w ith th e d ir e c t io n o f t h is fo rc e and hence th e v a r ia t io n o f momentum
and th e r e fo r e th e v a r ia t io n o f v e lo c it y ,
fa c e .
is a v e c to r norm al to the s u r­
We o b ta in the dynam ical e q u a tio n by e q u a tin g th e im pulse o f the
fo rc e to th e change in momentum,as f o llo w s :
o r d ro p p in g th e f a c t o r S d t ,
(54)
Both th e c o n t in u it y e q u a tio n and th e dyn am ical e q u a tio n were f i r s t
g iv e n by Riemann.
He a p p lie d them s p e c i f i c a l l y to p la n e waves.
By
e lim in a t in g u 1 - u 2between th e se two e q u a tio n s we o b ta in th e fo rm u la
f o r th e v e lo c it y o f p ro p o g a tio n :
(55)
As ρ1 approaches ρ
2
and p 1 approaches
p2
t h is e x p re s s io n approaches
the v a lu e : θ = ρ 1 /ρ 0
o r from e q u a tio n √(dp/dρ)
(53),θ
= d p /d ρ = θ 2
1
(56)
and hence the v e l o c i t y o f p ro p o g a tio n w ith re s p e c t to the medium on
e it h e r s id e o f th e d is c o n t in u it y approaches th e v e lo c it y o f p ro p o g a tio n
o f o r d in a r y sound o f i n f i n i t e s i m a l a m p litu d e
.
If
, however,
we assume P o is s o n ’ s a d ia b a tic law ( 21) we have:
and f o r g iv e n va lu e s^a s la rg e as we p le a s e by ta k in g p 2 s u f f i c i e n t l y
[of p1 and ρ1, θ becomes]
la r g e .
c . E q u a tio n o f energy
lo r d Rayleigh-*- o b je c te d to t h is s o lu t io n f o r p la n e waves
because the e q u a tio n o f energy is n o t s a t i s f i e d . .
L. R a y le ig h , Theory o f Sound - 1896, p p ,32 -3 2 , 40, 41.
By im p re s s in g upon
-2 8 -
the gas as a whole a v e lo c it y eq ua l and o p p o s ite to th e v e lo c it y o f
p ro p a g a tio n o f the d is c o n t in u it y , th e l a t t e r is b ro u g h t to r e s t w ith
re s p e c t to our f ix e d asses.
The phenomenon is th e n one o f ste a d y
m otio n and B e r n o u lli’s e q u a tio n o f energy (6) is a p p lic a b le .
T his g iv e s :
When th e d is c o n t in u it y is b ro u g h t to r e s t , u 1 = - θ1, u 2 = - θ2, and
from e q u a tio n (53) we o b ta in :
(57)
B e r n o u llis e q u a tio n o f energy mayAbe w r i t t e n as:
(5 8 }
Here u(, f>, and p f may be c o n s id e re d as c o n s ta n t and u ^ , & , p z as
v a r ia b le .
Then d i f f e r e n t i a t i n g we have:
and in t e g r a t in g we o b ta in :
(59)
T his is the o n ly law o f p re s s u re f o r w hich the energy c o n d itio n and th e
e q u a tio n o f c o n t in u it y a re s a t i s f i e d .
T his is n o t f u l f i l l e d
f o r any
a c tu a l gas and on t h is b a s is R a y le ig h con clu de s t h a t a f i r s t o rd e r
d is c o n t in u it y i s im p o s s ib le w ith o u t d is s ip a t iv e fo r c e s .
A fo rc e
a c tin g on th e gas is a new q u a n t it y e n te r in g the d yn a m ica l e o u a tio n and
may be de term in ed so as to produce a ste a d y s ta te w ith any la w o f
p re s s u re .
Weber'*' c o n s id e rs th e energy change in a d is c o n t in u it y o f
the f i r s t o rd e r ab i n i t i o
and f in d s f o r th e a d ia b a tic law o f P o isson (21)
r
a change o f energy i n p a s s in g the d is c o n t in u it y .
He says t h a t we must
in t h is case c o n s id e r the lav/ o f C a rn o t, in accordance w ith w hich th e re
may be a lo s s o f energy b u t ne ver a g a in .
Prom t h i s c r i t e r i o n he
1. Weber, D ie P a r t ie lle n D if f 'e r e n t ia l- g le ic h u iz e n , Y o l . I I 1901, p p .489-498.
-2 9 -
determ ines t h a t a d is c o n t in u it y o f c o n d e n sa tio n is p o s s ib le b u t one o f
r a r e f a c t io n im p o s s ib le .
Weber says t h a t th e lo s s o f energy must be made
up by the form es th a t move th e p is to n s .
does n o t go in h is e x p la n a tio n .
Beyond
t h is s ta te m e n t he
I t h in k a s o lu t io n to t h i s q u e s tio n
must be o b ta in a b le by a c o n s id e r a tio n s o le ly o f what happens in th e
im m ediate ne ighborhood o f th e d is c o n t in u it y and t h a t , as R a y le ig h
says, f o r th e a d ia b a tic law o f Pois s o n th e d is c o n t in u it y does n o t obey
th e law o f energy and t h a t t h is s o lu t io n is th e r e fo r e in v a l i d ,
d. H u g o n io t's la w .
H u g o n io t1 o b je c te d to th e use o f the a d ia b a tic law o f
P oisson (21) as
the v a l i d i t y o f t h is law depends on
w hich is c o n tin u o u s .
He s a id th a t t h is law is no more v a l id f o r a b ru p t
d is c o n tin u o u s changes o f d e n s ity .
the l a t t e r case.
a change o f d e n s ity
He d e riv e s a law w h ich i s v a lid f o r
He uses th e changeofi n t e rn a l energy o f a gas a
s
the energy
change n o t accounted f o r by th e k in e t ic energy andth
y e x te r n a l
b
n
rkd
o
w
e
fo rc e s b u t w h ich must e x is t because o f the law o f c o n s e rv a tio n o f e n e rg y.
The in t e r n a l energy o f a mass o f gas i s a f u n c tio n s o le ly o f i t s
s ic a l s ta te and s in c e a l l energy is a d d itiv e i t
gas.
L e t us c o n s id e r a g a in the mass o f gas,
phy­
v a rie s as th e mass o f
ρ0θsd t, w h ich flo w s acro ss
the s u rfa c e o f d is c o n t in u it y th ro u g h a c y lin d e r o f i n f i n i t e s i m a l c ro s s s e c tio n S in th e tim e d t .
Then th e in t e r n a l energy o f t h i s gas s u f­
fe r s a change o f m agnitude [η(ρ1,p1) - η(ρ2,p2)]ρ0θsdt where η (ρ ,p ) is the
s p e c if ic in t e r n a l e n e rg y, t h a t i s ,
th e in t e r n a l energy £ e r u n i t mass.
Since the work done by th e -pressures in i t s
passage from one s id e o f th e
s u rfa c e o f d is c o n t in u it y to the o th e r i s ( p ( u ( - p ^ u J S d t and the change
in k in e t ic energy is J _ p eQ S d t ( u * - u * ) , th e e q u a tio n o f energy beX
comes, d ro p p in g th e f a c t o r S d t ,
(60)
H ugoniot,, J o u rn a l de l ’ eeole P o ly te c h n iq u e , 1887, 1889
- 30 -
T his e q u a tio n should rem ain v a lid
if
any v e lo c it y is added to the gas
as a whole and the above e q u a tio n s h o u ld , t h e r e fo r e , c o n ta in o n ly th e
q u a n tity u 1 - u 2. To make t h is e v id e n t, m u lt ip ly the dyn am ical e q u a tio n
(54) by (u 1 + u 2)/2) and s u b tr a c t fro m the p re c e e d in g e q u a tio n and we o b ta in :
(61)
E lim in a tin g u 1 - u 2 by means o f th e e q u a tio n o f c o n t in u it y (5 2 ), we o b ta in
(62)
For a p e r f e c t gas the in t e r n a l energy has th e fo rm 1 η (ρ1p) =
(1/(y-1))(p/ρ)
and in t h is case we o b ta in :
(63)
or
(64)
T his is th e a d ia b a tic law o f H u g o n io t the v a l i d i t y o f w h ich is r e ­
cognized w ith o u t q u e s tio n by a l l the French s tu d e n ts o f t h is q u e s tio n .
H ugoniot c a lle d t h is law th e d yn am ical a d ia b a tic law in c o n tr a s t w ith
P o is s o n 's r e l a t io n w hich he c a lle d th e s t a t i c a d ia b a tic la w .
approaches
1
ρ
2
ρ
form :
As
and p 1 approaches p 2 , H u g o n io t's law approaches the
(dp/dρ) = (yp/ρ)
and hence agrees w ith the law o f P o is s o n .
By the v e ry method o f i t s
d e r iv a tio n the la w o f H u g o n io t s a t i s f i e s th e c o n d itio n o f energy and
hence is n o t open to the o b je c t i o n w h ich we fou nd in th e a p p lic a t io n
o f P o is s o n 's la w to a f i r s t o rd e r d is c o n t in u it y
e. En tro p y and Thermodynamic P o t e n t ia l
We s h a ll need th e q u a n t it ie s , e n tro p y and thermodynamic
p o t e n t ia l f o r l a t e r d is c u s s io n and i t
is a d v is a b le here to d e fin e them.
1. Sur le P ro p o g a tio n des Ondes * Hadamard, 1903, p 191
-3 1 -
I f a system o r p a r t o f a system a t an a b s o lu te te m p e ra tu re T re c e iv e s
a q u a n tity o f h e a t dQ, the in c re a s e o f e n tro p y 12 i s :
(65)
I f th e te m p e ra tu re o f th e system i s n o t c o n s ta n t, we must b re a k i t up
in t o elem ents and sum up th e e n tro p y o f each e le m e n t.
How the change
in th e in t e r n a l energy in ch a n g in g fro m one s ta te to a n o th e r is made
up o f th e h e a t g a in and o f the w ork done on th e gas, when the work
n e c e s s a ry to produce th e change in k in e t ic
ed from c o n s id e r a tio n .
We s h a ll l e t
^
energy as a whole is a b s t r a c t ­
be the s p e c if ic in t e r n a l e n e rg y,
and we s h a ll re cko n th e w o rk, W, as p o s it iv e when i t
the gas and n e g a tiv e i f
it
i s done on the gas.
is work done by
Then we must have, f o r
a u n i t mass o f gas
d Yj = dQ - dW
S u b s t it u t in g th e va lu e o f dC fro m ( 6 5 ) we o b ta in :
(66)
From the d e f i n i t i o n o f w ork we have:
dW=pdv, where v is th e s p e c if ic volume o f th e gas
S u b s t it u t in g t h is
va lu e o f dW in
(66) we o b ta in :
(67)
The therm odynam ical p o t e n t ia l is d e fin e d
2
as :
(68)
D i f f e r e n t ia t in g we have:
S u b s t it u t in g th e va lu e o f d »-j from (67) we o b ta in :
1. G. H. B ryan, Thermodynamics, 1107, p . 58
2. B ryan, Thermodynamics, 1907, p p .9 1 ,9 2 .
-3 2 -
Hence we have:
(69)
(70)
F or ir r e v e r s ib le processes th e e n tro p y o f a system must in c re a s e .
In f a c t the i n crease o f e n tro p y may be ta ke n as a measure o f the
i r r e v e r s i b i l i t y o f a p ro c e s s .
A r e v e r s ib le pro cess is o n ly an id e a l
and we may c o n s id e r th e in c re a s e o f e n te ro p y a ne cessa ry c o n d itio n in
a l l a c tu a l p h y s ic a l changes.
f . A p p lic a tio n o f the E n tro p y C o n d itio n to a F i r s t O rder D is c o n t in u it y .
Duhem1
uses as a fo u n d a tio n o f h is s tu d y o f f i r s t o rd e r
d is c o n t in u it ie s e q u a tio n s (6 8 ),
(52) and ( 54 ) .
(69)
(70)
and the e q u a tio n s o f Riemann
The c o n d itio n o f in c re a s e o f e n tro p y , he wr i t e s by means
o f (6 9 ):
(71)
He a ls o uses th e c o n d itio n t h a t the p re s s u re must in c re a s e w ith in c re a s e
o f d e n s ity o r t h a t d
/ρ
p
is p o s it iv e a lw a ys.
By means o f e q u a tio n (70)
t h is e q u a tio n may be w r it t e n :
(72)
He a ls o uses H u g o n io t's law in th e g e n e ra l fo rm
(6 2 ).
are
The d is c o n t in u it y is
c o n s id e re d fu n c tio n s o f ρ
and T.
A l l fu n c tio n s
supposed to have v a lu e s o f th e d e n s ity and te m p e ra tu re ρ 1 ,T 1 by
approach from th e re g io n 1 and ρ 2 ,T2 b y approach from the re g io n 2.
A t the d is c o n t in u it y ρ1 and T (
are the n fu n c tio n s o f Px .
a re c o n s id e re d as f ix e d and a l l fu n c tio n s
We may th u s w r it e :
S u b s titu te in e q u a tio n (62) th e v a lu e o f b from (68) and we o b ta in :
(73)
D i f f e r e n t ia t in g w ith re s p e c t to
we o b ta in :
and u s in g e q u a tio n s (69) and (70)
(74)
. Duhem, Z e i t s c h r i f t f u r P h y s ik a lis c h e Chemie 69, 1909, pp. 169-186.
-3 3 -
By d i f f e r e n t i a t i n g the v a lu e o f σ (ρ2,T2) = ∑ ( ρ2) from e q u a tio n (69)
and the va lu e o f ρ2 = Π ( ρ 2 ) from e q u a tio n (70) w ith re s p e c t to ρ2 we
o b ta in :
(75)
(76)
I t is o f advantage to in tro d u c e here th e q u a n t it y V ( ρ2,T2) = ^ ^-TRpJj
w hich is th e o r d in a r y v e l o c i t y o f sound o f i n f i n i t e s i m a l a m p litu d e in
the re g io n 2 w it h re s p e c t to t h i s medium.
q u a n tity
(dp2/dT2)(p2
U sin g t h is q u a n t it y and the
c o n s ta n t) we may express th e r e s u lt o f the e lim in a t io n
o f dθ(ρ2)/dρ2 between the above two e q u a tio n s in th e f o llo w in g fo rm :
(77)
In e q u a tio n (74) s u b s t it u t e :
o b ta in e d from th e v e lo c it y e q u a tio n o f Reimann (5 5 ).
s u lt in g e q u a tio n and e q u a tio n (7 7 ), e lim in a te
Between th e r e ­
dΠ(ρ2)/dρ2and we o b ta in :
(78)
Duhem1
w r it e s t h is e q u a tio n w ith th e r i g h t member o f o p p o s ite s ig n
b u t h is er r o r is e v id e n t on com plete in v e s t ig a t io n .
him to advance some wrong
T h is e r r o r
le d
c o n c lu s io n s b u t th e y are e a s ily r ig h t e d .
T his e r r o r is e v id e n t a ls o fro m th e c o n t r a d ic t io n in the c o n c lu s io n s
w hich he draws in the g e n e ra l case and in th e case o f a p e r f e c t gas.
From t h is e q u a tio n we see t h a t f o r ρ2 =ρ1
Duhem2
does n o t m e n tio n t h a t t h i s c o n c lu s io n m ig h t have been
from e q u a tio n (71) s in c e Π ( ρ 2 ) = P1 , whenρ
2
= ρ
. D iffe r e n tia tin g
1
e q u a tio n (78) we o b ta in an e q u a tio n o f the f o llo w in g form*.
L. Loc. c i t . p.
1. Loc. c i t . p .
o b ta in e d
-3 4 -
(80)
where A and B a re f i n i t e
fo r ρ 1 = ρ 2 .
Then f o r ρ1 = ρ2 we must have from
the above e q u a tio n ,
(81)
Thus f o r a d is c o n t in u it y w it h an i n f i n i t e s i m a l v a r ia t io n in d e n s ity ,
the v a r ia t io n i n e n tro p y is an i n f i n i t e s i m a l o f a t le a s t th e t h i r d o rd e r.
Since
ρ1θ1 = ρ2θ2(see e q u a tio n (53)
)
θ1
and θ2
a re o f the same s ig n .
A c c o rd in g to th e c o n v e n tio n o f s ig n a lre a d y used, i f θ 1 a n d θ 2
a re
p o s it iv e , the gas passes from th e re g io n 2 to the re g io n 1 and
∑ (ρ2)-σ(ρ1,Τ1)
and θ2
must he n e g a tiv e s in c e th e e n tro p y must in c re a s e ;
ifθ
1
are n e g a tiv e , th e gas passes from the re g io n 1 to th e re g io n 2
and ∑(ρ2)-σ(ρ1,Τ1) must be p o s it iv e .
Thus th e d iffe r e n c e in e n tro p y ,
∑(ρ2)-σ(ρ1,Τ1) is o p p o s ite in s ig n to θ1 and θ 2 . S ince th e v a r ia t io n
o f e n tro p y is o f th e t h i r d o rd e r ∑(ρ2)-σ(ρ1,Τ1) has th e s ig n o f (ρ2-ρ1)[(d3∑
(ρ2)/dρ23]ρ2=ρ1
as lo n g as ρ1 -
does n o t exceed in a b s o lu te v a lu e a c e r t a in l i m i t .
2
ρ
By d i f f e r e n t i a t i n g e q u a tio n (80) and p la c in g ρ2 = ρ1 we e a s ily o b ta in :
(82)
For b r e v it y l e t us in tro d u c e the f u n c tio n
(83)
Then ∑(ρ2)-σ(ρ1,Τ1)
has th e s ig n o f (ρ 2 - ρ 1 ) H ( ρ1,Τ1)
s ig n s o p p o s ite to t h i s q u a n t it y .
a p o s it iv e d is c o n t in u it y
and θ1 and θ2 must have
In gases f o r w h ich H ( f,,T, ) is p o s it iv e
(6>tf) can o n ly he p ro po ga te d i f
p, >
or i f
the wave is one o f co n d e n s a tio n and f o r a gas f o r w hich H (^?,T, ) is
n e g a tiv e , a p o s it iv e d is c o n t in u it y can o n ly be p ro p o g a te d i f
i f the wave is one o f r a r e f a c t io n .
P, < p x o r
In th e same way, o f c o u rs e , i f
the
- 35 -
d is c o n t in u it y is p ro p o g a te d in th e n e g a tiv e d ir e c t io n , co n d e n s a tio n o r
r a r e f a c t io n are p o s s ib le a c c o rd in g as H ( ρ1, T 1 ) is p o s it iv e o r n e g a tiv e
r e s p e c tiv e ly .
For a c tu a l gases H (ρ1,T1) is p o s it iv e and hence o n ly
con de nsa tion s a re p o s s ib le .
E q u a tio n (78)
in v o lv e s the q u a n t it y V ( ρ2T2) - θ 2 2 . By
d e te rm in in g the s ig n o f t h i s q u a n t it y , we d e te rm in e th e r e l a t i v e m agni­
tude o f the v e lo c it y o f p ro p a g a tio n , w it h re s p e c t to th e gas i n re g io n
2 compared w ith the v e lo c it y o f p ro p a g a tio n o r o r d in a r y sound in t h i s
re g io n . In t h is e q u a tio n th e c o e f f i c ie n t o f
dΣ(ρ2)/dρ2 i s p o s it iv e
if
i n a b s o lu te v a lu e is l e ss th a n a c e r t a in l i m i t s in c e 2ρ1ρ2θ(ρ2) is
p o s it iv e .
A ls o
dΣ(ρ2)/dρ2
has the s ig n o f H (ρ1,T1) and th e r e fo r e is
p o s it iv e o r n e g a tiv e ac c o rd in g as th e wave is
r a r e f a c t io n .
one o f c o n d e n s a tio n o r
We may express the se r e s u lt s by s a y in g t h a t
dΣ(ρ2)/dρ2 and
ρ2-ρ1 have th e same s ig n i f th e re g io n 2 is b e h in d th e wave and o p p o s ite
s ig n s i f
t h is re g io n is in f r o n t o f t h i s wave.
From the above m entioned
e q u a tio n we th u s see th a t w it h re s p e c t to th e re g io n behind the wave,
th e v e lo c it y o f p ro p a g a tio n is le s s th a n t h a t o f o r d in a r y sound and w it h
re s p e c t to th e re g io n i n f r o n t o f th e wave i t
o r d in a r y sound.
is g re a te r th a n th a t o f
These r e s u lt s h o ld , o f co u rs e , p r o v i d e d ρ
does n o t
-1
2
exceed a c e r t a in l i m i t in a b s o lu te v a lu e .
Duhem1 a ls o examines th e s ig n o f the q u a n t it y θ 2 2 - [V ( ρ 1 ,T 1 ) ]2 . In so
d o in g he compares the v e lo c it y o f p ro p a g a tio n w ith re s p e c t to one re g io n
w ith th e v e lo c it y o f o rd in a r y sound in th e o th e r medium.
q u a n tity is
z e ro .
For ρ 2 = ρ 1 t h is
By d i f f e r e n t i a t i n g th e e x p re s s io n f o r the v e lo c it y
as g iv e n by Riemann (55) we o b ta in :
(84)
where β
-1
)=
2
(ρ
For
ρ 2 = ρ 1 , β (ρ 2 )= o
.. lo c . c i t . p .
dΠ(ρ2)/dρ2 - (2ρ2-ρ1)[π(ρ2)-p1]
th e r e fo r e we must examine
<*3j o j . By d i f f e r e n t -
-3 6 -
i a t i n g we have:
(85)
For ρ 2 = ρ 1
t h is
is
zero and we must examine th e n e x t d e r iv a t iv e .
F o r t h is we o b ta in th e e x p re s s io n :
(86)
and
say
Thus we see t h a t f o r
β ( ρ2) has i t s
(87)
le s s th a n a c e r t a in l i m i t in a b s o lu te v a lu e -
s ig n o p p o s ite to t h a t o f K (ρ 1 T1)and θ22- [V(ρ1T1)]2 has the
s ig n o f - (ρ2 - ρ1)K (ρ l Tl)
I f K ( on T(J ls p o s it iv e , th e v e lo c it y o f p ro p a g a tio n o f th e d is c o n t in u it y
w ith re s p e c t t o th e le s s dense re g io n i s g re a te r th a n the v e lo c it y o f
sound in the more dense re g io n ;
if
is n e g a tiv e , th e v e lo c it y o f
p ro p a g a tio n o f th e d is c o n t in u it y w it h th s p e c t to th e more dense re g io n
is le s s th a n the v e lo c it y o f sound in th e le s s dense re g io n .
Duhem'1" the n a p p lie s these th e o r ie s to a p e r f e c t gas.
I f C and c are
the s p e c if ic h e a ts u n de r c o n s ta n t p re s s u re and volume r e s p e c t iv e ly we
have iiu g o n io t's law (64) in th e f o llo w in g fo rm :
(88)
E it h e r th e n u m era to r o r d e no m in ator o f the f r a c t io n in th e r i g h t
member a te p o s it iv e and s in c e the p re s s u re s must be p o s it iv e we must
have b o th n u m era to r and de no m in ator p o s it iv e .
(89)
We may w ith o u t lo s s o f g e n e r a lit y a t t r i b u t e the s u b s c r ip t 2 to th e
g re a te r o f th e two d e n s it ie s .
o f its e lf.
Then th e f i r s t s n e g u a lity is f u l f i l l e d
The second may be w r i t t e n in th e
fo rm :
(90)
Loc. c i t . p
-3 7 -
For most d ia to m ic gases C
v a lu e n e a r 6.
p2
/ c
.4 a p p ro x im a te ly and ρ 2 /ρ 1 has a l i m i t i n g
1
=
As the d e n s ity in c re a s e s w ith a f i x e d ρ1 the p re s s u re
becomes i n f i n i t e
as ρ 2 approaches
(C + c)/(C-c)ρ1.
E lim in a t in g p 2 between
e q u a tio n s (55) and (88)we o b ta in , when we p la c e $ = θ1, and ρ 0 = ρ1
and
Then we have :
(91)
T h is shows t h a t th e v e lo c it y o f p ro p a g a tio n o f a d is c o n t in u it y w it h r e ­
spe ct to the re g io n be hind the wave is le s s tha n th e v e lo c it y o f sound
in th e same re g io n and the v e lo c it y w ith re s p e c t to th e r e g io n in f r o nt
o f the wave is g re a te r th a n th a t o f sound in t h is r e g io n .
the fu n c tio n
o f e q u a tio n
We see t h a t
( 87), K(ρ1,T1),has th e f o llo w in g fo rm :
(92)
T his i s p o s it iv e o r n e g a tiv e a c c o rd in g as C/c is le s s o r g re a te r th a n 3
I n a l l known cases the fo rm e r is t r u e .
shown to he p o s it iv e f o r ρ
/2
1
In t h is case θ22-[V(ρ1,T1)]2
le s s th a n 2c/(C-c)
and n e g a tiv e i f
it
is e a s ily
is
g re a te r th a n t h i s q u a n t it y .
M. E. Jouguet^ o b ta in e d some o f the above r e s u lt s p re v io u s
to M. huhem b u t by analagous m ethods.
e n tro p y and d e riv e d the id e n t ic a l
He a ls o used th e d r i t e r i o n o f
o n d itio n s to those g iv e n above f o r
the p o s s i b i l i t y o f a c o n d e n sa tio n and r a r e f a c t io n . He shows th a t i f the
s
medium i f o f s m a ll v is c o s it y , a s m a ll s tra tu m o f g re a t change in v e lo c it y
and d e n s ity w i l l r e s u lt such t h a t the v is c o s it y may be n e g le c te d e v e ry »d
where save in t h i s s tra tu m .
I f th e th ic k n e s s o f t h is s tra tu m is so
Jouguet, Comptes Hendue 138 p 1685, 1904; Comptes Hendus 139, p .7 8 6 , 1904
-3 8 -
i n f i n i t e l y s m a ll, th e laws o f p ro p a g a tio n are tho se g iv e n by the fo rm ­
u la e o f Riemann (e qu a tio n s 52 and 54) and o f Hu g o n io t (e q u a tio n 6 4 ) .
g . M o tio n F o llo w in g an I n i t i a l P lane Wave D is c o n t in u it y .
An i n i t i a l d is c o n t in u it y o f th e f i r s t o rd e r w i l l n o t in
g e n e ra l, be p ro pa ga te d a s such f o r th e e q u a tio n s o f Riemann,
(5 4 ), to g e th e r w ith (18)
, g iv in g th e law o f p re s s u re , w i l l n o t, in gen­
e r a l be s a t is f ie d f o r a r b i t r a r y va lu e s o f u 1
, ρ1,
d is c o n t in u it y .
(52) and
u 2 ,ρ 2 in an i n i t i a l
The i n i t i a l d is c o n t in u it y may re s o lv e i t s e l f in t o
two d is c o n t i n u it ie s o f th e f i r s t o rd e r o r in t o d is c o n t in u it ie s o f the
second o rd e r.
(1) R iem ann's Treatm ent
Riemann1
tr e a te d o f an i n i t i a l p la n e wave d is c o n t in u it y
in an unbounded gas in w h ich the re g io n s 1 and 2 were in a c o n s ta n t
u n ifo rm c o n d itio n o f v e lo c it y and d e n s ity suchthat th ro u g h o u t re g io n s 1 and 2
andUi
f>, and fares pe-csirively
th e v e lo c it y was u, and the d e n s ity . . Weber g iv e s Riem ann's tre a tm e n t
i n a v e ry c o n c is e , c le a r manner.
Riemann showed,
1 s h a ll here f o llo w h is tre a tm e n t.
t h a t , co n fo rm a b ly w ith the e q u a tio n s o f m o tio n (3)
and (4) a p p lic a b le to re g io n s where a l l th e q u a n t it ie s in v o lv e d are
c o n tin u o u s and w ith e q u a tio n s (52) and (54) a p p lic a b le to a f i r s t
o rd e r d i s c o n t i n u it y , assum ing th e p re s s u re g iv e n by ( 1 8 ;, an i n i t i a l
p la n e wave d is c o n t in u it y , c o u ld be s a t i s f i e d by one o f fo u r cases de­
pending upon th e i n i t i a l v a lu e s o f the v e lo c it y and d e n s ity in re g io n s
1 and 2 .
These eases in v o lv e r e s o lu t io n in t o d is c o n t in u it ie s a f
com pression and a s p e c ia l ty p e o f r a r e f a c t io n wave.
a lre a d y been d is c u s s e d .
The fo rm e r have
The l a t t e r type o f wave in v o lv e s th e a p p l i­
c a tio n to a c e r t a in domain o f the x - t p la n e o f a s o lu t io n o f th e typ e
u
+ c = rF Jfyfijo) + *
T h is s o lu t io n may e a s ily be v e r i f i e d
. doc. c i t .
. '7eber , D ie f a r t i e l l e n D if f e r e n t ia l- g le ig h u n g e n IIp p 4 8 0 -8 8
-3 9 -
by s u b s t it u t io n in e q u a tio n s (3) and ( 4 ) .
1 s h a ll d is c u s s t h i s s o lu t io n
more f u l l y l a t e r h u t 1 s h a ll need the r e s u lt f o r use in the p re s e n t
case.
K iem ann's fo u r cases a re , th e n :
(1) a r e s o lu t io n in t o two
d is c o n t in u it ie s o f com pression pro p a g a te d in o p p o s ite d ir e c t io n s w it h
re s p e c t to th e gas,
(£) a r e s o lu t io n in t o two r a r e f a c t io n waves p ro ­
pagated in o p p o s ite d ir e c t io n s w ith re s p e c t to th e gas,
(3) a re s o ­
l u t i o n in to a p o s i t i v e l y p ro p a g a te d d is c o n t in u it y and a n e g a tiv e ly
porpagated r a r e f a c t io n wave, and (4) a r e s o lu t io n in t o a n e g a tiv e ly
propagated d is c o n t in u it y and a p o s i t i v e l y pro pa ga te d r a r e f a c t io n wave.
e may ro u g h ly see the c o n d itio n s f o r some o f these case s, l o r the f i r s t
case, the in te rm e d ia te re g io n c re a te d i s a re g io n o f g re a te r d e n s ity
th a n e it h e r o f the domains 1 and 2 and th e v e lo c it ie s must o b v io u s ly
be such th a t the two masses o f gas in re g io n s 1 and 2 t& jodcth'U jaiatl,
a n o th e r.
one
L e t us s tu d y the se cases in d e a t i l .
C o nside r K iem ann's f i r s t case.
The i n i t i a l d is c o n t in u it y
re s o lv e s i t s e l f in t o two d is c o n t in u it ie s o f com p ression .
In f r o n t o f
the f i r s t d is c o n t in u it y u and p r e t a in t h e i r i n i t i a l v a lu e s u t and
and be h in d th e second d is c o n t in u it y th e y r e t a in t h e i r i n i t i a l v a lu e s
u,>
w h ile in the re g io n between th e two d is c o n t in u it ie s th e y have
the c o n s ta n t v a lu e s u ^ p
. Then th e v e l o c i t i e s o f p ro p a g a tio n o f th e
two d is c o n t in u it ie s a re c o n s ta n ts .
We may re p re s e n t th e p o s it io n o f
each o f the d is c o n t in u it ie s a t a tim e t by a p o in t in a p ane in w hich
x and t are ta ke n as r e c ta n g u la r c o o rd in a te s .
In t h i s p la n e th e m o tio n
o f each d is c o n t in u it y is re p re s e n te d by a l i n e
s in c e the v e lo c it y o f
p ro p a g a tio n
i i s
c o n s ta n t.
x - a x is is such th a t i t s
The angle made by t h i s l i n e w it h th e
c o ta n g e n t is
dX ,th e v e l o c i t y o f p ro p a g a tio n ,
"2SF
In the accompanying h ig . 1 th e d is c o n t in u it ie s are pro p a g a te d a lo n g the
lin e s 1 and 2 w hich make a n g le s
o f e q u a tio n (34) we th e n o b ta in :
and <Xj_ w ith the x - a x is .
By means
-4 0 -
(93)
(94)
F ig . 1.
From the above fo rm u la e ,
(95)
(96)
A dding we have:
(97)
The fu n c tio n , ((ρ-ρ1)φ(ρ)-φ(ρ1))/ρρ1 = ( (1/ρ1)- (
1
/
ρ
)
) [φ(ρ)-φ(ρ1)]for
is an in c r e a s in g f u n c tio n o f a s in c e each f a c t o r is p o s it iv e and in c re a s e s
w it h in c re a s e o f ρwhenρ>ρ1.For Riem ann's f i r s t cas e, ρ' must b e g r e a te r tha n
b o th ρ1 andρ
2
and hence th e r i g h t member o f the above e q u a tio n i s an i n ­
c re a s in g f u n c t io n o f ρ ' .
When
'=ρ1 and when ρ ' =ρ 2 , the r i g h t member
ρ
o f the above e q u a tio n becomes ((ρ1-ρ2)[φ(ρ1)-φ(ρ2)])/ρ1ρ2
and th e r e fo r e we
must have the f o llo w in g in e q u a lit y :
(98)
I f t h is in e q u a lit y i s
fu lfille d ,
th e re is a s in g le v a lu e o f ρ' w hich
s a t is f ie s e q u a tio n (97) and when ρ ' is fo u n d , u'
o f e q u a tio n s (95) o r (9 6 ).
The v e l o c i t i e s
is o b ta in e d from e it h e r
o f p ro p a g a tio n are found
from (93) and (94) and th e n e ce ssa ry r e l a t io n ,
cot
is
α2> c o t α1
, o f i t s e l f , v e r ifie d .
I f the c r i t i c a l in e q u a lit y
(98)
(98) above is an
e q u a lity , we see from th e l a s t e q u a tio n t h a t ρ ' equals e it h e r ρ
1
o r ρ2 and
-4 1 we have a s in g le d is c o n t in u it y p ro p a g a ted backward o r fo rw a rd re s p e c t­
iv e ly .
C o nside r n e x t Riem ann's second case.
Fig . 2
we ta ke fo u r s t r a ig h t li n e s ,
In the accompanying
1 ,2 ,3 ,4 w ith th e a n gles α 1 ,α 2 ,α 3 ,α 4
w ith th e x a z is r e s p e c t iv e ly and assume in th e s e c to r ( - ∞ 0 1 ) u and
p c o n s ta n t and e q u a l to t h e i r i n i t i a l v a lu e s u 1,ρ1 and in the s e c to r
(4 , 0 +∞ ) lik e w is e c o n s ta n t and equ a l to u 2 ,ρ 2 .
l e t u and ρ
be c o n s ta n t and eq u a l to u ',ρ'.
In the s e c to r (2 ,0 ,3 )
In th e s e c to rs ( 1 , 0 , 2 )
and ( 3 ,0 ,4 ) we assume s o lu tio n s o f th e t y pe u =
±∫((√φ'(ρ))/ρ)dρ + c = ±√(φ'(ρ)) + (x/t)
ta ke n so th a t a lo n g the lin e s 1 , 2 ,3 ,4 these s o lu tio n s s h a ll b e con­
tin u o u s w ith the va lu e s in th e b o rd e rin g dom ains.
t y pe has the c h a r a c t e r is t ic t h a t f o r
s t a n t,
in o th e r w ords, u and ρ
x /t
A s o lu t io n o f t h i s
c o n s ta n t, u and ρ are c on­
have c o n s ta n t v a lu e s a lo n g e v e ry li n e
em anating from 0 .
Fi g . 2.
L e t us f o r b r e v it y w r it e f ( ρ ) = ∫((φ'(ρ ))/ρ)dρ.
We may ta b u la te the above
c o n d itio n s as f o llo w s :
(99)
The c o n d itio n o f c o n t in u it y between these re g io n s g iv e s th e f o llo w ­
in g r e l at io n s :
-4 2 -
(10 0)
From these e ig h t e q u a tio n s we must d e te rm in e the e ig h t q u a n t it ie s ,
c , c ' , u ' , ρ ', α 1 , α 2 , α 3 , α 4 .
E lim in a t in g c and c ' we o b ta in :
(101)
(102)
S u b tra c tin g we o b ta in :
(103)
S ince the waves are o f r a r e f a c t io n , ρ' must be s m a lle r th a n b o th
ρ1 and ρ 2 .
The d e r iv a t iv e fu n c tio n :
f '( ρ ) =
is p o s it iv e and f (ρ
(√φ'(ρ))/ρ
) th e r e fo r e decreases as ρ
d e cre a se s.
u 1 - u 2 , g iv e n by th e above e q u a tio n , is le s s than i t s
ρ' =
For ρ
,
2
>
1
v a lu e f o r
and we have:
2
ρ
(104)
For ρ2 > ρ1, u 1- u 2 is le s s th a n i t s
va lu e f o r ρ ' = ρ1 and we have:
(105)
I n o rd e r t h a t a vacuum he n o t c re a te d in th e gas, f o r c e r t a in laws o f
pre s s u re , we must f u r t h e r r e s t r i c t u 1 - u 2 to b e in a b s o lu te v a lu e le s s
than a l i m i t d e term in ed by th e law o f p re s s u re .
P o is s o n 's la w , φ(ρ)=aρ y
F o r exam ple, f o r
and f (ρ ) = (2a(√y))/(y-1))ρ(y-1)/2 w h ic h , s in c e y is
g re a te r th a n 1 and s in c e ρ
From e q u a tio n (103) t h is
is e s s e n t ia lly p o s it iv e , must be p o s it iv e .
in v o lv e s th e l i m i t a t i o n
u1 - u 2< - f (ρ 1 ) - f ( ρ 2 ).
For B o y le 's law no l i m i t a t i o n is ne ce ssa ry s in c e f ( ρ ) = a lo g ρ
be any v a lu e from - ∞
to +∞
.
can
I f th e l i m i t a t i o n ju s t m entioned and
the in e q u a lit y (104) o r (105) w h ich e ve r i s a p p lic a b le , is f u l f i l l e d ,
the n e q u a tio n (103) g iv e s a u n iq u e va lu e f o r ρ '.
Knowing ρ ' , from
e it h e r e q u a tio n (101) or (102) we may o b ta in u ' w hich l i e s between
-4 3-
u 1andu
2
and from e q u a tio n s (100) we may th e n o b ta in th e c o ta n g e n ts
o f th e a n g le s α 1 , α 2 , α 3,
and α 4 .
These a re in c o r r e c t o rd e r o f
m agnitude w ith o u t f u r t h e r r e s t r i c t i o n s .
C o nside r n e x t Ri emann's t h i r d case.
In t h i s case a d is c o n ­
t i n u i t y o f c o n d e n sa tio n is p ro p a g a te d fo rw a rd s and a wave o f r a r e f a c t io n
backward.
T his in v o lv e s th e assum ption t h a t ρ1 is g re a te r th a n ρ 2 .
In the x - t p la n e (see F ig . 3) we have th e wave o f r a r e f a c t io n in th e
s e c to r ( l , o , 2 ) and th e p o s it iv e ly p ro pa ga te d d is c o n t in u it y a lo n g th e
lin e
(o , 3)
.
F ig . 3.
W ith the te rm in o lo g y in d ic a te d in the accompanying d iag ram , we may
ta b u la te th e f o llo w in g c o n d itio n s :
( - ∞ , o , l ) : u = u 1 ρ = ρ 1 , c o n s ta n t
( l,o ,2 )
:
( 2 ,0 ,3 )
: u
u = - f ( ρ ) + c = (√φ'(ρ)) + (x/t)
(106)
( 3 ,o ,
): u
= u'
ρ = ρ ' , c o n s ta n t
= u2
ρ = ρ 2 , c o n s ta n t
We have f o r th e v e lo c it y o f p ro p a g a tio n o f th e d is c o n t in u it y :
(107)
The c o n d itio n o f c o n t in u it y a lo n g th e lin e s
( o ,l)
and (o ,2 ) demands:
(108)
(109)
From these e q u a tio n s we o b ta in :
-4 4 -
(110)
(111)
S u b tr a c tin g we o b ta in :
(112)
The r i g h t member o f th e above e q u a tio n in c re a s e s w ith in c re a s e
of ρ' i f ρ
'
andρ
1
ρ
2
must l i e
is g re a te r th a n ρ2 as is n e c e s s a ry .
in v a lu e ,
S ince ρ ' l i e s between
hence u 1 - u 2, g iv e n by th e above e q u a tio n ,
between i t s
v a lu e s f o r ρ'=ρ1 and ρ'=ρ2. T h e re fo re we have:
(113)
I f t h is in e q u a lit y is
ρ ' and when t h i s
fu lfille d
e q u a tio n (112) g iv e s a un iq ue v a lu e f o r
is known, e it h e r e q u a tio n s (110 o r (111) d e term in e
u ' and α 1 , α2, α3
a re th e n found fro m e q u a tio n s ( 1 0 7 ) ,
(108) and (109)
o f t h is pa ra g ra p h and a re in c o r r e c t o rd e r o f m agnitude w ith o u t f u r ­
th e r r e s t r i c t i o n s .
Riemann’ s f o u r t h case is n o t e s s e n t ia lly d i f f e r e n t from the
t h i r d and may b e o b ta in e d fro m i t by changing th e d ir e c t io n o f th e
x - a x is .
We may now sum up th e p re c e e d in g r e s u lt s .
For b r e v it y , l e t us
w r it e :
f ( ρ1) - f ( ρ2) = ± ∆ where th e s ig n o f ∆
th a t
is ta k e n such
∆ ( ρ 1 -ρ 2 )> o
Then we have:
Case I :
[u ] > R
Case I I :
- ∆ > [u ]
Case I I I :
R > [u ] > - ∆
, ρ1>ρ2
Case I v :
R > [u ] > - ∆
, ρ1> ρ2
W ith th e e x c e p tio n o f th e l i m i t a t i o n on u1 - u2 in Case I I
-4 5 -
to p re v e n t fo rm a tio n o f a vacuum, one o f these fo u r g iv e s an a c c e p ta b le
s o lu t io n f o r e ve ry s e t o f v a lu e s o f th e g iv e n q u a n t it ie s , u 1, u 2
, ρ1, ρ2.
The q u e s tio n s t i l l rem ains open as to w h e th e r t h is s o lu t io n is u n iq u e .
Riemann1 d id n o t d is c u s s t h is q u e s tio n
a
d
W
e
b
r2
a
lso
m
a
ke
sn
o
m
e
tio
n
f.H
a
d
m
r3
h
o
w
e
vr,te
a
sh
q
u
e
stio
n
q u ite th o ro u g h ly and shows th a t
s e v e ra l s o lu tio n s are sometimes p o s s ib le w h ic h s a t i s f y th e e q u a tio n s
o f m o tio n .
R iem ann's s o lu t io n a p p lie s s p e c i f i c a l l y to a d is c o n t in u it y
s e p a ra tin g two unbounded re g io n s in w hich u and ρ
a re c o n s ta n t, bu t
the s o lu t io n a ls o h o ld s f o r th e im m ediate n e ig hb orh oo d o f any d is c o n ­
t i n u i t y f o r the f i r s t in s t a n t o f tim e .
The g e n e ra l problem o f d is c o n ­
t i n u i t i e s w ith g iv e n i n i t i a l c o n d itio n s has n o t been s o lv e d .
2. Hadamard' s tre a tm e n t.
L e t us c o n s id e r the case o f d is c o n t in u it ie s as tre a te d by
Hadamard1
.
Ea ch o f th e waves o f r a r e f a c t io n tr e a te d by
Riemann
are c la s s if ie d two d is c o n t in u it ie s o f th e second o rd e r by Hadamard.
Hadamard t r e a t s o f th e r e s o lu t io n in to d is c o n t in u it ie s o f th e f i r s t
o rd e r.
Between th e two re g io n s in w h ich u and p and p have t h e i r
o r i g i n a l v a lu e s u1, ρ1, p 1and u 2 , ρ 2 , p 2 , we have an in te rm e d ia te
re g io n in w hich u, ρ and p have v a lu e s u ' , ρ ' , p ' . Hadamard c o n s id e rs
P o is s o n 's law to h o ld between the se th re e s ta t e s , namely pwy = c o n s ta n t
where w = 1 /ρ
is in tro d u c e d f o r co n ve n ie n ce .
He v is u a liz e s t h i s law
as a curve i n a p la n e in w h ich w and p are r e c ta n g u la r c o o rd in a te s . The
th re e p o in ts
(w 1 ,p1 ) ,
( w 2, p 2 ) , and (w'
,p' ) l i e
upon t h is c u rv e .
T his assum ption makes Hadamard's tre a tm e n t le s s g e n e ra l th a n Riem ann's
fo r i t
im p lie s t h a t o n ly th re e o f th e q u a n t it ie s , w1 , p 1 , w2 , p2 a re
a r b it r a r y .
In e f f e c t h e assumes t h a t t h is m o tio n is g e ne rated fro m a
gas i n i t i a l l y homogeneous such, i o r example, as by th e sudden move­
ment o f a p is to n ,
in w h ich case th e two s ta te s w ould H e on th e same
. Loc. c i t .
. Loc. c i t .
. Hadamard, Sur la P ro p a g a tio n des Undes, 1903 , p p .194-200
-4 6 -
a d ia ba t i c , n o t ta k in g acco un t o f H u g o n io t's o b je c tio n , We have fro m the
e q u a tio n o f Riemann (52) and (54) by e lim in a t io n o f 0
(114)
(115)
S u b tra c tin g we have:
(116)
where e ve ry c o m b in a tio n o f upper and lo w e r s ig n s b e fo re th e two
r a d ic a ls i s p e rm is s ib le b u t f o r e it h e r r a d ic a l th e s ig n is e it h e r
upper o r lo w e r f o r a l l th re e e q u a tio n s and t h i s r u le w i l l be extended
to the subsequent e q u a tio n s .
By t r a n s p o s it io n and s q u a rin g o f b o th
members we o b ta in f o r the above:
(117)
(118)
S q ua ring b o th o f th e se e q u a tio n s , we amalgamate th e two branches ex­
pressed by the ambiguous s ig n in t o a s in g le c u rv e , as fo llo w s :
(119)
(120)
These e q u a tio n s are e q u iv a le n t.
The p re c e e d in g e q u a tio n s , however, are
d i f f e r e n t as th e y d iv id e th e cu rve in to branches a t d i f f e r e n t p o in ts .
P 1 and P2 are q u a d ra tic e x p re s s io n s b u t P - P _ is li n e a r .
Thus t he
curve is o f th e second degree and hence, a coni c s e c tio n .
For P1 - P2+a2=o ,
from e q u a tio n (119) we' must have P1 = (p' - p 1
) (w1- w') = o and hence e it h e r
p' = p 1
, o r w' = w1. Since th e e x p re s s io n P 1- P2 + a2 occurs tw ic e as a
f a c t o r in equ a tio n (119) th e coni c is ta n g e n t to th e lin e s p ' = p 1and
w, = w ( , p a r a l l e l to th e c o o rd in a te axes a t t h e i r p o in ts o f in t e r s e c t io n
w ith the s t r a ig h t l i n e P, - P^+ a = o.
a c c o rd in g to e q u a tio n (120)
lik e w is e th e l i n e P, - Pa - a = o
in te r s e c t s th e c o n ic a t i t s
p o in ts o f
-4 7 -
tangency w ith th e lin e s p' = p2 and w' = w2
. S ince th e e q u a tio n o f th e curve
in v o lv e s the s in g le p a ra m e te r a , the cu rve is one o f a fa m ily o f c o n ic s
ta n g e n t to th e s id e s o f a re c ta n g le composed o f th e lin e s p ' = p1, w' = w1,
p' = p 2
, w' = w2 drawn w ith th e g iv e n p o in ts
o p p o s ite c o rn e rs .
(w1 ,p1, ) and (w2 , p2 ) a t
We e a s ily see th a t f o r a
com prised between o and
(p1 - p 2
) (w2 - w1 ) th e c o n ic is an e llip s e and i f a
value i t
surpasses t h is
is an h y p e rb o la . In th e accom panying Fi g . 4 , A 1and A2 are the
g iv e n p o in ts
(w1p 1 ) and (w2 p2) r e s p e c t iv e ly and A1B1A2B2 the re c ta n g le
to th e s id e s o f w hich the c o n ic is ta n g e n t.
as an e llip s e
in s c r ib e d in t h i s r e c ta n g le .
The c o n ic is re p re s e n te d
The cho rd C1D1has f o r i t s
e q u a tio n P 1 - P2 + a2 = o and th e chord C D has f o r i t s
e q u a tio n P1 - P2- a2 =o
such t h a t P1 - P2 = o re p re s e n ts the d ia g o n a l B1B2
T his c o n ic c u ts the a d ia b a tic between A1and A 2 in two p o in ts A' and A' ' .
Thus we have two va lu e s f o r th e i n te rm e d ia te s t a t e .
There is however a
f u r t h e r c r i t e r i o n to d is t in g u is h between the se two v a lu e s . C o nside r
the e q u a tio n s o f th e cu rve d iv id e d in t o b ra n ch e s.
The p o in t A ' l i e s
on
the a rc C1, D1, and th e p o in t A ' ' on th e a rc C2D2. F o r th e fo rm e r a r c ,
P1 - P2 + a 2
< o and f o r th e l a t t e r , P1 - P2 - a 2>o
We see by com paring the
s ig n b e fo re the r a d ic a ls in e q u a tio n s (117) and (118) w ith th e s ig n
b e fo re the same r a d ic a ls in e q u a tio n s (115) and (116) t h a t f o r th e a rc
C1D1, a (u '- u1
o
)>
and a (u'
- u 2 )> o and f o r th e a rc C2 D2 , a ( u ' -u 1) <o
and a (u' - u 2
)<o w h ile f o r th e r e s t o f the c o n ic , a ( u ' - u1)<o
a (u ' - u 2
) > o. L e t θ
1
andθ
2
and
be th e v e lo c it ie s o f p ro p a g a tio n o f the two
-4 8 -
d is c o n t in u it ie s w ith re s p e c t to an i n i t i a l s ta te in w h ich the d e n s ity
is
. Then th e e q u a tio n s o f c o n t in u it y
(52) become:
(121)
(122)
D iv id in g one e q u a tio n by th e o th e r we o b ta in :
(123)
F or b o th the a rc s C1D1and C2D2 u' - u 1 , and u' - u2 have the same s ig n
and hence we have:
(124)
A ls o , i f
th e c o n ic is an e l l i p s e , w'
l i e s between th e v a lu e s w1and w2
and hence w1 - w' and w2 - w ' have o p p o s ite s ig n s o r
T h e re fo re from th e p re c e e d in g e q u a tio n we have:
(125)
Thus on b o th o f the se a rc s , th e d i s c o n t in u it ie s t r a v e l in o p p o s ite
d ir e c t io n s , F or th e re m a in in g two a rc s o f ou r e l l i p s e , u ' - u 1and u ' - u 2
have o p p o s ite s ig n s and hence we have:
(126)
Thus on these a r c s, th e two d is c o n t in u it ie s t r a v e l in the same d ir e c t io n .
For ou r p o in ts A' and A'' on the a rc s C1D1 and C2D2 r e s p e c t iv e ly i t
c e s sa ry to in q u ir e i f
the c o n d itio n
θ1 < o is s a t is f ie d f o r b o th the se
p o in ts and i f f o r o n ly one, w h ich one.
th a t i f θ1 is to be n e g a tiv e w1<w' o r
is ne­
From e q u a tio n (1 2 1 ), we see
w1>w' a c c o rd in g as u' - u 1= +√
)
1
(P
or
u ' - u1
=- √ ( P 1 ) . F o r w 1 < w 2 , the fo rm e r a lt e r n a t iv e must be chosen and
t h is e n fo rc e s th a t th e in t e r s e c t io n w h ich is a c c e p ta b le must be on the
a rc C1D1 i f
the i n i t i a l d is c o n t in u it y is one o f co n d e n s a tio n (See
-4 9 -
e q u a tio n 98)
(a >o)
and on the a rc C2D2 i f
is one o f r a r e f a c t io n ( a< o)
the i n i t i a l d is c o n t in u it y
(See e q u a tio n s (1 0 4 )and(105).
Fo r w1> w2
th e l a t t e r a lt e r n a t iv e must be chosen and t h i s e n fo rc e s th e c h o ic e o f
th e a rc C2D2 i f the i n i t i a l d is c o n t in u it y is one o f co n d e n s a tio n (a> o)
and the a rc C1D1i f th e i n i t i a l d is c o n t in u it y is
We may p u t t h is in a l i t t l e
u ity
d i f f e r e n t fo rm .
one o f r a r e f a c t io n ( a < o ) .
I f the i n i t i a l d is c o n t in ­
is one o f c o n d e n s a tio n the p o in t A must be chosen n e a re r t h a t one
o f the two p o in ts A 1and A 2 w h ich corre spo nd s to the g re a te r d e n s ity ; i f
the i n i t i a l d is c o n t in u it y is one o f r a r e f a c t io n the p o in t A must be
n e a re r th e p o in t c o rre s p o n d in g to th e le s s e r d e n s ity .
Thus i f
the
c o n ic is an e llip s e and the o n ly in te r s e c t io n s o ccu r on the a rc s C1 D1
and C2D2 the s o lu t io n is u n iq u e .
i f th e c o n ic is an h y p e rb o la , w ' is
e x t e r io r to the range from w1to w2.
i n i t i a l d is c o n t in u it y is
w1 and w2 and i f
to b o th w 1 and
the
We know, how ever, t h a t i f the
one of c o n d e n s a tio n , w' is
i s c o n t i n u i t y is
d
i n f e r i o r to b o th
one o f r a r e f a c t io n , w' is s u p e r io r
w2 . S ince one in t e r s e c t io n w i l l always o ccu r on the
s id e i n f e r i o r to t h is range and one on the s id e s u p e r io r to t h i s ra n g e ,
t h is c r i t e r i o n serve s to d is t in g u is h th e c o r r e c t s o lu t io n .
more than two in t e r s e c t io n s of th e a d ia b a tic w it h th e c o n ic .
e may have
W
In the
case o f th e e llip s e p ic tu r e d in th e f ig u r e , we may have an in t e r s e c t io n
a lo n g the a rc C1D2 . That such p o in ts o f in t e r s e c t io n occu r may be seen
i f the e llip s e a p pro xim a te s u f f i c i e n t l y c lo s e ly i t s
l i m i t i n g p o s it io n
as a double li n e a lo n g the d ia g o n a l A1A 2
. These two in t e r s e c t io n s , as
we have seen by e q u a tio n (126) r e p re sen t m o tio n s f o r w h ich θ1 and θ2 have
the same s ig n .
I n the se cases the s o lu t io n is n o t u n iq u e b u t th e re are
s e v e ra l s o lu t io n s , a l l o f w hich a re p o s s ib le .
employ the c r i t e r i o n o f e n tro p y ,
Hadamard does n o t
f ######################
as
the a p p lic a t io n o f t h is p r i n c i p le had n o t been advanced a t the tim e he
2.
3•
U
5.
-5 0 -
w ro te h is book.
O th e rw is e , o f c o u rs e , he w ould r e je c t d is c o n t in u it ie s
o f r a r e f a c t io n as Zemplen1 , Jou q u e t2 and Duhem3 h a te done
(See page 2 5 ).
h . O b je c tio n to H u g o n io t's Law
A lth o u g h th e p r i n c i p le o f e n tro p y has been e x te n s iv e ly employ­
ed in c o n n e c tio n w ith f i r s t o rd e r d is c o n t in u it ie s ,
it
is d i f f i c u l t to
see how in the absence o f any d is s ip a t iv e fo r c e s , the gas in p a s s in g
th ro u g h a d is c o n t in u it y can s u f f e r a g a in in e n tro p y ,
d e c la re s t h a t i t c a n n o t.
lo r d R a y le ig h ^
He shawa th a t we may o b ta in the e q u a tio n o f
H ugoniot by c o n s id e rin g a wave o f perm anent wave form in w h ich the
t r a n s it io n between two s ta te s need n o t be a b ru p t b u t t h a t d is s ip a t iv e
fo rc e s must be a llo w e d to m a in ta in th e permanency o f the wave.
A
r e l a t io n id e n t ic a l to H u g o n io t’ s was found f i f t e e n y e a rs e a r l i e r by
Rankine
f o r a wave o f permanent w$ve form m a in ta in e d by h e a t con­
d u c tio n .
R a y le ig h says t h a t the e o u a tio n o f ehergy is s a t i s f i e d b u t
b u t n o t the second law o f therm odynam ics.
s id e o f the d is c o n t in u it y l i e ,
He s a y s ,in e f f e c t ,
The two s ta te s on e it h e r
in g e n e ra l, upon two d i f f e r e n t a d ia b a tic s .
th a t one cannot g e t from one a d ia b a tic to a n o th e r
w ith o u t a t r a n s f e r o f h e a t,
in th e d e r iv a t io n o f H u g o n io t’ s la w th e
in t e r n a l energy is e v a lu a te d fro m th e e x p re s s io n :
The d iffe r e n c e o f in t e r n a l e n e rg ie s in th e t r a n s f e r acro ss th e d is ­
c o n t in u it y is ta ke n as:
R a y le ig h p o in ts o u t t h a t t h is
in v o lv e d in th e passage a t
in v o lv e s the assum ption t h a t n o th in g is
w = c& from one a d ia b a tic to a n o th e r and th a t
Loc. c l 't .
HOC. e x t.
R a y le ig h , P roceedings o f th e R oyal S o c ie ty , 1910, yp 267-269.
R ankine, ” 0n the Thermodynamic Theory o f Waves o f F in i t e lo n g it u d in a l
D is tu rb a n c e " - P h il. T rans. 1870, v o l. 160. H a rt I I , p , 277.
- 51-
th e re is a c t u a lly re q u ire d the com m unication o f an i n f i n i t e s i m a l q u a n tity
o f h e a t. To g e t from one a d ia b a tic to th e o th e r a t r a n s f e r o f h e a t is
r e q u ir e d .
Heat can o n ly be gained a t the expense o f w ork and hence d is ­
s ip a t iv e fo rc e s must e n te r and an a b ru p t d is c o n t in u it y becomes th e n an
im p o s s ib ilit y ,
That d is s ip a t iv e fo rc e s c o u ld n o t be a llo w e d in a f i r s t
o rd e r d is c o n t in u it y is a ls o re c o g n iz e d by Duheni*' and f o r t h is
reason he
is a t a lo s s f o r an e x p la n a tio n o f the e n tro p y change f o r t h is case b u t
in s is t s t h a t , a c c o rd in g to th e c o n c lu s io n s o f Jouguet'-', h is r e s u lt s are
a p p lic a b le to a wave w ith v e ry sudden t r a n s i t i o n la y e r m a in ta in e d by
d is s ip a t iv e fo rc e s .
4.
i t h in k t h i s
s t i l l rem ains a d is p u te d p o in t .
Summary
I t has been shown t h a t d is c o n tin u o u s waves alw ays produce
v a r ia t io n s in the tim e d e r iv a tiv e s o f x , y , and z.
The d e r iv a tiv e s o f
one
the d e n s ity , o f o rd e r^ le s s th a n th e o rd e r o f d is c o n t in u it y are the f i r s t
to become d is c o n tin u o u s .T h e f i r s t o rd e r d is c o n t in u it y i s th e o n ly case
where th e re i s a d is c o n t in u it y in the d e n s ity i t s e l f .
I n e v e ry d i s ­
c o n tin u o u s wave th e c h a r a c t e r is t ic segment i s norm al to th e s u rfa c e
and in waves o f th e second o r h ig h e r o rd e r th e v e lo c it y o f p ro p a g a tio n
is t h a t o f o r d in a r y sound o f i n f i n i t e s i m a l a m p litu d e w ith re s p e c t to
the medium in i t s
c o n t in u it y .
a c tu a l s ta te on e it h e r s id e o f th e s u rfa c e o f d is ­
The c o n t in u it y and d yn am ical e q u a tio n s f o r a f i r s t o rd e r
d is c o n t in u it y were de ve lo p e d .
P’o r the a d ia b a tic law o f P o isso n the
v e lo c it y o f p ro p a g a tio n may be made as g re a t as we p le a s e by in c re a s in g
the p re s s u re on one s id e o f th e d i s c o n t i n u it y .
The energy e q u a tio n is
n o t s a t is f ie d by P oissons law o f p re s s u re and H u g o n io t's lav/ was de1. Loc. c i t .
2. Hoc. c i t .
-5 2 -
velop ed so as to s a t is f y t h is e q u a tio n .
The e n tro p y o f th e gas,
however, under t h is law s u f f e r s a change on p a s s in g th e d is c o n t in u it y .
The c o n d itio n o f in c re a s e o f e n tro p y a p p lie d to a c tu a l gases im p lie s
th a t o n ly a d is c o n t in u it y o f c o n d e n s a tio n is p o s s ib le .
The v e lo c it y o f
p ro p a g a tio n under n u g o n io t’ s law is alw ays g re a te r th a n t h a t o f o rd in a r y
sound o f in f i n i t e s i m a l a m p litu d e when b o th are measured w ith re s p e c t to
the medium in f r o n t o f th e wave and le s s th a n th e v e l o c i t y o f o rd in a ry
sound o f in f i n i t e s i m a l a m p litu d e when b o th are measured w ith re s p e c t to
the re g io n b e h in d th e wave.
The r e s o lu t io n o f an i n i t i a l d is c o n t in u it y
was d is c u s s e d a c c o rd in g to the method o f Riemann and o f Hadamard.
U sing
P o is s o n 's law i t was shown th a t a r e s o lu t io n in to two f i r s t o rd e r d is ­
c o n t in u it ie s s a t is f y in g the e q u a tio n o f c o n t in u it y and the dynam ical equa­
t io n c o u ld sometimes be made in more th a n one way.
The o b je c tio n o f
lo r d R a y le ig h to th e a p p lic a t io n o f n u g o n io t's law to a d is c o n t in u it y
was d is c u s s e d .
T his o b je c tio n , is i n su b sta n ce , t h a t the e n tro p y can
n o t in c re a s e w ith o u t th e e n tra n c e o f d is s ip a t iv e fo rc e s in w hich case
the t r a n s it io n between two s ta te s cannot be a b ru p t.
-5 3 C. I n f in it e s im a l A m p litu d e Sound Waves?
B e fo re ta k in g up th e s u b je c t o f sound waves o f f i n i t e
a m p litu d e i t
is p ro p e r to in d ic a t e th e g e n e ra l r e s u lt s a tta in e d
f o r in f i n i t e s im a l a m p litu d e waves as these waves a re r e a l l y the
f i r s t a p p ro x im a tio n to th e case o f f i n i t e
For convenience i t
a m p litu d e waves.
is custom ary to change from th e v a r ia b le p
to the v a r ia b le S d e fin e d by
(127)
The q u a n tity S is c a lle d the c o n d e n s a tio n .
The case where
u , v , w , s , and t h e i r f i r s t d e r iv a tiv e s w ith re s p e c t to x ,y ,z ,a n d
t are reg ard ed as so s m a ll t h a t th e second degree term s in the se
q u a n t it ie s may be dropped fro m th e e q u a tio n s o f m o tio n (1)
and ( 2 ) , may be c a lle d th e case o f sound waves o f in f i n i t e s i m a l
a m p litu d e .
The e q u a tio n s o f m o tio n the n become li n e a r .
I f the v e lo c it y has a p o t e n t ia l f u n c t io n , t h a t is
, if
th e re
e x is ts a fu n c tio n (p such t h a t : (128)
th e m o tio n is found to be i r r o t a t i o n a l .
When the se v a lu e s
o f u ,v ,w , a re s u b s titu te d in th e dyn am ical e q u a tio n s ( 2 ) , we
o b ta in by in t e g r a t io n and d is c a r d in g o f second degree term s
f o r the case o f in f i n i t e s i m a l a m p litu d e waves th e s in g le e q u a tio n :
(129)
where
C =
is to be c o n s id e re d c o n s ta n t s in c e th e d e n s ity
changes are s m a ll.
o f c o n t in u it y
W ith th e same s u b s t it u t io n
the e q u a tio n
(1 ) becom es:(130)
1.
The m a te r ia l f o r t h is s e c tio n is s e le c te d from
Lamb's H ydrodynam ics,1906, pp. 4 5 4 ,4 5 5 ,4 6 6 -4 7 2 .
-5 4 The e lim in a t io n o f S between e q u a tio n s (129) and (130) g iv e s :(131)
I f the gas is c o n s id e re d as unbounded and th e i n i t i a l
v a lu e s o f the v e lo c it y and d e n s ity a re g iv e n , th e i n i t i a l
v a lu e o f ϕ is every where g iv e n by in t e g r a t io n o f any one o f
e q u a tio n s (1 2 8 ), th e c o n s ta n t o f in t e g r a t io n re m a in in g a r b i t r a r y ,
and the i n i t i a l v a lu e o f ∂ ϕ / ∂ t is everyw here g iv e n by v ir t u e
o f e q u a tio n (1 2 9 ).
We may w r it e f o r the se i n i t i a l c o n d it io n s : (132)
(133)
The v a lu e o f ϕ a t any p o in t P a t a tim e t,
the o r ig in a l v a lu e s o f ϕ
about the p o in t P.
ifϕ
and ∂ϕ
/t
is a f u n c tio n o f
on a sphere o f ra d iu s c t
The e f f e c t a t P , m oreover, is
unchanged
and ∂ϕ/∂t are c o n s ta n t and eq u a l to t h e i r re s p e c tiv e mean
v a lu e s on th e sphere.
L e t ϕm and (
/t)m d e s ig n a te th e mean
∂ϕ
v a lu e s o f ϕ and ∂ ϕ /∂ t r e s p e c t iv e ly , on a sphere o f ra d iu s r
about th e p o in t P in the o r i g i n a l c o n d itio n .
I f co d e s ig n a te
a s o lid a n g le , we h a v e :(134)
(135)
where P is th e p o in t x , y , z , where r is the ra d iu s v e c to r from
t h is p o in t and
1, m, n i t s
d ir e c t io n c o s in e s .
The e f f e c t a t P
is the same as i f we were g iv e n a s y m m e tric a l s p h e r ic a l m o tio n
w ith i n i t i a l v a lu e s o f th e v e lo c it y p o t e n t ia l and i t s
d e r iv a t iv e , ^ =
v) and | ^ * l = / I (.>■) .
tim e
The v a lu e o f <f) as a fu n c tio n
o f t a t th e p o in t P in a s y m m e tric a l s p h e r ic a l m o tio n about t h is
p o in t w ith th e above i n i t i a l c o n d itio n s i s : (136)
-5 5 -
S u b s t it u t in g th e v a lu e s o f Γ and Λ
from e q u a tio n s (134)
and (135)
in e q u a tio n (136) and s in c e :r = c t , 1 = s i n θ cosϕ , m = s i n θ s i nϕ ,n = cosθ ,dω= s in θd θ d ϕ
we o b t a in : (137)
T h is is a com plete s o lu t io n f o r an i r r o t a t i o n a l sound wave o f
in f i n i t e s im a l a m p litu d e in an unbounded gas.
The s o lu t io n f o r
a plan e wave is
(138)
The s o lu t io n f o r a s y m m e tric a l s p h e r ic a l wave is :(139)
In b o th cases th e fu n c tio n s f and F a re to be d e te rm in e d by
the i n i t i a l c o n d itio n s .
-5 6 -
D. C ontinuous F in i t e A m p litu d e Sound Waves.
I s h a ll now d is c u s s c o n tin u o u s f i n i t e
a m p litu d e sound waves,
in t h is case no s o lu t io n has been o b ta in e d f o r th e e q u a tio n s in t h e i r
g e n e ra l form (1) and (2) o r (12) and ( 1 3 ).
I s h a ll , th e r e fo r e , c o n s id e r
f i r s t the case o f p la n e waves o r r e c t i l i n e a r d is tu rb a n c e s o f f i n i t e
a m p litu d e .
I.
Plane Waves
For p la n e waves we must o b ta in a s o lu t io n o f e q u a tio n s (3) and
(4) o r (15)and (1 6 ).
I n e it h e r system o f r e p r e s e n ta tio n th e re are b u t
two independent v a r ia b le s x and t o r a and t .
W hichever o f the two
dependent v a r ia b le s we w ish to c o n s id e r may be c o n s id e re d as a space
c o o rd in a te w ith the in de pe nd en t v a r ia b le s
th e o th e r two c o o rd in a te s .
A s o lu t io n o f the d i f f e r e n t i a l e q u a tio n s co rre sp o n d s th e n g e o m e tr ic a lly
to a s u rfa c e in w h ich these a re the c o o rd in a te s .
I s h a ll now d is c u s s
the p r o p e r tie s o f th e s o lu tio n s o f the e q u a tio n s o f m o tio n and o b ta in
s o lu tio n s f o r d i f f e r e n t c o n d itio n s .
1. G eneral Theory o f C h a r a c te r is t ic s f o r P lane WavesiL e t us c o n s id e r e q u a tio n s (15) and (16) f o r pla n e waves.
In s te a d o f e x p re s s in g p as a fu n c tio n o f ^
as in e q u a tio n ( 2 2 ) , l e t
us adopt a new v a r ia b le .
and express p as a f u n c tio n o f t h is v a r ia b le , namely
(140)
E lim in a tin g j °
between e q u a tio n s (15) and (16) and s u b s t it u t in g the
above v a lu e o f p, we o b ta in :
P la c in g - ---- w fa *
=
(wj we o b ta in :
(141)
L. Hadamard, lo c . c i t .
pp. 154-159, 173-175
-5 7 T h is is th e e q u a tio n to be s o lv e d to o b ta in x as a fu n c tio n o f a and t .
Since the v e lo c it y is u = ∂ x / ∂ t
and w = ∂ x / ∂ a , we have:
(142)
(143)
E lim in a tin g ∂2x/∂a
2 and ∂2x/∂t2
between e q u a tio n s (141)
(142) and (143)
we o b ta in :
(144)
The e q u a tio n :
(145)
When the s u rfa c e in t e g r a l is known, d e te rm in e s two f a m ilie s o f curve s
on t h is s u rfa c e , one curve o f each fa m ily p a s s in g th ro u g h each p o in t
on th e s u rfa c e .
The curve s th u s d e fin e d are c a lle d th e c h a r a c t e r is t ic s
o f t h is s u rfa c e .
f in ite .
I f o u r s o lu t io n is to be c o n tin u o u s , ∂2x/∂a∂t
must be
A lo n g one o f th e c h a r a c t e r is t ic s d e fin e d by (1 4 5 ), th e deno­
m in a to r o f th e r i g h t member o f e q u a tio n (144) is
be f i n i t e ,
zero and i f ∂2x/∂a∂t is to
the n u m e r a to r o f th e r i g h t member o f e q u a tio n (144) must
a ls o be z e ro .
T h e re fo re we must have
(146)
S u b s t it u t in g th e v a lu e o f da from e q u a tio n (19) in t h is e q u a tio n we
o b ta in
(147)
I n t e g r a t in g we o b ta in :
(148)
T his is a d i f f e r e n t i a l e q u a tio n w hich must be s a t i s f i e d a lo n g th e
c h a r a c t e r is t ic s o f o u r s u rfa c e in t e g r a l.
l e t us c o n s id e r the c o n d itio n th a t a second s u rfa c e in t e g r a l
-5 8 -
may ex i s t ta n g e n t to the f i r s t a lo n g a c u rve Γ
. Now the e q u a tio n o f
th e ta n g e n t p la n e to th e s u rfa c e in t e g r a l a t the p o in t x , a
, t
, i s:
(149)
Now a lo n g the cu rve Γ ,
the ta n g e n t p la n e s f o r th e two s u rfa c e s c o in c id e
and from the above e q u a tio n we see t h a t w and u must be th e same f o r
th e two s u ffa c e s a lo n g t h is c u rv e ,
i s o f the f i r s t o rd e r,
a lo n g the curve
in d e te rm in a te .
7 )V
i f th e c o n ta c t o f th e two s u rfa c e s
w i l l be d i f f e r e n t f o r the two s u rfa c e s
and hence i t s
e x p re s s io n fro m e q u a tio n (18) must be
T h is is o n ly in d e te rm in a te a lo n g a c h a r a c t e r is t ic and hene<
the curve JT* must be a c h a r a c t e r is t ic .
between two s u rfa c e in t e g r a ls i t
must be a c h a r a c t e r is t ic .
F or any f i n i t e
o rd e r o f c o n ta c t
i s p ro ve n t h a t th e cu rve o f tangency
The c h a r a c t e r is t ic
is th u s th e lo c u s ctf a
d is c o n t in u it y o f th e second o r h ig h e r o rd e r and e q u a tio n (145) agrees
w ith e q u a tio n (47) and shows t h a t the d is c o n t in u it y is p»opagated w ith
the o r d in a r y v e lo c it y o f sound o f in f i n i t e s i m a l a m p litu d e w ith re s p e c t
to the gas.
j
2. S in g le P ro g re s s iv e Yaves
A s in g le p ro g re s s iv e wave may be d e fin e d as one w h ich is p r o ­
pagated in t o a re g io n a t r e s t .
A d o p tin g o u r g e o m e tric a l re p r e s e n ta tio n ,
the re g io n a t r e s t is re p re s e n te d by the p la n e x = a .
A s in g le p r o ­
g re s s iv e wave, th e n , is one re p re s e n te d by a s u rfa c e in t e g r a l ta n g e n t
to th e p la n e x= a .
a . E arnshaw 's C o n d itio n 1
L e t us c o n s id e r a s u rfa c e in t e g r a l ta n g e n t a lo n g a c h a ra c te r ­
is t ic
o f th e f i r s t fa m ily o f e q u a tio n ( 1 4 5 ) ,nam ely:
(150)
Since w = 1 a lo n g t h is c h a r a c t e r is t ic w h ich p e r ta in s to the re g io n a t
. Hadamard, lo c . c i t . pp 174. 195
R a y le ig h , Theory o f Sound Y o l . I I pp . 34, 35.
-5 9 -
r e s t in w hich
is tic
x= a .
From any p o in t o f th e s u rfa c e , draw a c h a ra c te r ­
o f th e fa m ily o p p o s ite to t h a t o f th e c h a r a c t e r is t ic o f ta n g e n cy.
T h is c h a r a c t e r is t ic w i l l c u t the c h a r a c t e r is t ic o f tangency in a
p o in t P.
A long th e fo rm e r c h a r a c t e r is t ic we must have (e q u a tio n 14 8):
u = -X (w) + G
and a t the p o in t P, s in c e u = o, w = 1, we have:
o = -X (1) + C
T his de te rm in e s th e c o n s ta n t C and th e c o n d itio n a lo n g th e fo rm e r
c h a r a c t e r is t ic becomes:
u = X ( l)
- X(w)
(151)
Since t h is c h a r a c t e r is t ic was a r b i t r a r i l y chosen and s in c e e ve ry p o in t
on the s u rfa c e must have a c h a r a c t e r is t ic o f t h is fa m ily p a s s in g th ro u g h
it,
th e above is a c o n d itio n to be f u l f i l l e d
the wave.
The c h a r a c t e r is t ic o f tan ge ncy
o ve r th e whole s u rfa c e o f
g iv e s th e m o tio n o f the
f r o n t o f th e wave and fro m (150) we see th e v e l o c i t y o f th e f r o n t o f
the wave is p o s it iv e o r t h a t th e wave is a p o s it iv e p ro g re s s iv e wave.
E q u a tio n (151) is c a lle d E arnshaw 's c o n d itio n f o r a p o s it iv e p ro g re s s iv e
wave.
I f the c h a r a c t e r is t ic o f tangency be lo ng s to the o th e r fa m ily
o f e q u a tio n (145) t h a t i s ,
corre spo nd s to
we may s i m i l a r l y show t h a t t h is
s u rfa c e in t e g r a l s a t i s f i e s
u = X (w) - X (1)
the r e l a t io n :
(152)
The p ro p a g a tio n in the n e g a tiv e d ir e c t io n is n o t e s s e n t ia lly d i f f e r e n t
from t h a t i n the p o s it iv e d ir e c t io n s in c e th e one case becomes the
o th e r by r e v e r s in g the d ir e c t io n s o f th e a - and x - axes.
T h is r e v e r s a l
o f axes im p lie s a s u b s t it u t io n o f a = -a , x = -x , and u = -u and we see from
the above fo rm u la e t h a t th e one case tra n s fo rm s in to th e o th e r .
Thus the
two cases d i f f e r o n ly as an image in a p la n e m ir r o r d i f f e r s fro m i t s
-6 0 -
o b je c t.
Ho?/ from e q u a tio n (1 4 8 ), X ’ (w) - t / V (w) and from th e rem ark
p re c e e d in g e q u a tio n (1 4 1 ), [jj (w) = -
(w) . I f th e p re s s u re in / 0
creases w ith in c re a s e o f d e n s ity we see from e q u a tio n s (15) and (140)
th a t j) (w) is n e g a tiv e and hence
r e a l and p o s it iv e .
^
(w) is p o s it iv e and hence X ’ (w) is
X(w) is th e n an in c re a s in g f u n c tio n o f w.
Prom
e q u a tio n s (151) and (152) f o r p o s it iv e and n e g a tiv e waves r e s p e c tiv e ly
we see th a t f o r th e fo rm e r u decreases w it h w and f o r th e l a t t e r u i n ­
creases w ith w o r from e q u a tio n (15) we see t h a t f o r th e fo rm e r u i n ­
creases w ith
f
and f o r th e l a t t e r u daereases w ith
n e c e s s a r ily fo llo w s i f
f 1 . T h is r e s u lt
the r e s is ta n c e o ffe re d by the gas to th e m o tio n
o f th e p is to n , used to g e n e ra te a s in g le p ro g re s s iv e wave i n a c y lin d ­
r i c a l tu b e , in c re a s e s w ith i t s
speed i f t h i s is measured p o s it iv e i n
a d ir e c t io n com pressing the gas.
p is to n is pS where S
The re s is ta n c e to th e m o tio n o f the
is the c ro s s s e c tio n are a o f th e p is to n .
I f th e
re s is ta n c e in c re a s e s w ith u , p must in c re a s e w ith u and s in c e p is an
in c r e a s in g f u n c tio n o f
f
, p must in c re a s e w ith u .
We may o b ta in E arnshaw 's c o n d itio n Bp a method due to Lo rd
R a yle ig h ? - 1 e may im agine an;y p o in t w h ich we are exa m in in g to be b ro u g h t
to r e s t f o r the in s t a n t by im p re s s in g upon th e gas as a whole a u n i­
form v e lo c it y .
in th e neig hb orh oo d o f t h is p o in t th e c o n d itio n is th a t
o f a wave o f i n f i n i t e s i m a l a m p litu d e and the e q u a tio n s g iv e n in
s e c tio n C o b ta in ,
f o r a p o s it iv e p ro g re s s iv e i n f i n i t e s i m a l wave E=o in
e q u a tio n (138) and s u b s t it u t in g the v a lu e o f
(j) from t h i s e q u a tio n
f o r u , (128) and in the e q u a tio n i o r S, (1 2 9 ), we o b ta in on e lim in a t in g
f'
between th e r e s u lt in g e q u a tio n s :
1 .R a y le ig h , Theory o f Sound, 1896, V o l. I I , p p . 24 -35 .
-6 1 -
o r s in c e
c= √(ϕ'(ρ)), u/s = √(ϕ'(ρ))
Now in the wave o f f i n i t e
(1 5 3 )
a m p litu d e we must c o n s id e r i n f i n i t e s i m a l
v e lo c it ie s and co n d e n sa tio n s r e l a t iv e
to th e p o in t b ro u g h t to r e s t .
We m ust, t h e r e fo r e , use du f o r u and, by e q u a tio n (127), dρ/ρ f o r
and we must have f o r th e v i c i n i t y o f e ve ry
S
p o in t in th e wave:
(154)
I n t e g r a t in g t h i s e x p re s s io n from a p o in t on th e c h a r a c t e r is t ic o f
tangency where u = o and ρ
= ρ0 to any p o in t in the wave h a vin g a
v e lo c it y u and a d e n s ity ρ we o b ta in :
(155)
w hich is e q u iv a le n t to e q u a tio n (151) and i s Earnshaw’ s c o n d itio n f o r
a p o s it iv e p ro g re s s iv e wave.
b . G eneral S o l u t i o n and P r o p e r tie s o f a S u rfa ce I n t e g r a l.
S ince a p o s it iv e and
n e g a tiv e wave are n o t e s s e n t ia lly
d i f f e r e n t , l e t us c o n s id e r a p o s it iv e wave f o r w hich e q u a tio n (151)
h o ld s .
tic
A lo ng a c h a r a c t e r is t ic o f th e same fa m ily as th e c h a r a c te r is ­
o f ta n g e n cy,
s ig n .
(e q u a tio n 150) we have e q u a tio n (145) w ith th e upper
To t h is corre spo nd s e q u a tio n (148) w ith the upper s ig n , t h a t i s :
u = X ( w) + C
For th e s u rfa c e o f the wave in g e n e ra l we have e q u a tio n (1 5 1 ) nam ely:
u = X ( l)
- X(w)
A dding and s u b tr a c tin g , we see t h a t u and w a re b o th c o n s ta n t f or
p o in ts on the same c h a r a c t e r is t ic ; f o r w h ic h we have e q u a tio n (145) w ith
the upp e r s ig n .
T h is shows t h a t c o n s ta n t va lu e s o f u and w are tr a n s ­
m itte d th ro u g h the gas w ith
to th e i n i t i a l s t a t e .
v e l o c i t i e s e q u a l to √(ψ(w)) w ith re s p e c t
T his is th e v e lo c it y o f i n f i n i t e s i m a l sound in
a gas o f d e n s ity c o rre s p o n d in g to th e v a lu e o f w.
F or a p o s it iv e p ro g re s s iv e wave we have:
-6 2 -
dx=w∙d a + u ∙d t s in c e w = ∂ x /∂a
and u =
∂x/∂t
S u b s t it u t in g th e v a lu e o f u from e q u a tio n (151) we o b ta in :
dx = w∙ d a + [ X ( l)
- X (w )]
The s u rfa c e in t e g r a l is th e n found
x = w a + [X ( l )
o = a - X'
- X(w)]
(156)
by e lim in a t in g w between th e eq ua tion s
t +ϕ
)
(w
(157)
(w) t + ϕ ' (w)
When the a r b it r a r y f u n c tio n ϕ
c a r r ie d o u t.
dt
(158 )
is s p e c if ie d , th e e lim in a t io n may be
T h is is E a rn ’ shaw’ s1
method o f s o lu t io n .
E q u a tio n (157)
is th e e q u a tio n o f a ta n g e n t p la n e , as we see by com paring w ith e q u a tio n
(149) to th e s u rfa c e and, s in c e i t
the s u rfa c e is d e v e lo p a b le .
depends o n ly upon one p a ram e te r w,
I f u o r w is ta ke n as dependent v a r ia b le
in s te a d o f x , the dependent v a r ia b le is c o n s ta n t a lo n g th e ch a ra ct e r i s t i c s
o f th e fa m ily g iv e n by e q u a tio n (145) w ith the up pe r s ig n .
a c t e r is t ic s are the n h o r iz o n t a l g e n e ra tric e s o f a r u le d
The c h a r­
s u rfa c e i f
the a t - pla n e be ta k e n as h o r iz o n t a l.
E lim in a t in g a between e q u a tio n s (156) we o b ta in :
x - [X ( l )
- X (w) + w X ' ( w ) ] t =
S u b s t it u t in g X ( l)
ϕ (w) - W ϕ ' (w)
(1 57A)
- X(w) = u from (151) and wX ’(w) = √(ϕ'(ρ))
from e q u a tio n s '(1 4 8 ) an d(1 5 ’ )
we o b ta in :
(158A)
Where £
o f w.
is an a r b it r a r y f u n c tio n o f
a s in c e by (151) u in a fu n c tio n
R e ve rsin g th e f u n c tio n a l r e l a t i o n , we o b ta in :
(159)
T his i s known as P o is s o n 's in t e g r a l a lth o u g h P o isso n 3 o b ta in e d i t
f o r B o y le ’ s law (21) f o r w h ich ϕ '(ρ )= C2 is c o n s ta n t.
o n ly
T h is e q u a tio n
shows th a t va lu e s o f u are p ro p a g a te d w ith v e l o c i t i e s u + √(ϕ'(ρ)) w ith
re s p e c t to a fix e d s e t o f axe s.
From L o rd R a y le ig h ’ s2 method o f a n a ly s is
1. Earnshaw, P roceedings o f th e R oyal S o c ie ty , Jan 6, 1859.
2.R a y leig h , Theory o f Sound V o l I I 1896 pp 33 ,3 4
3. P o is s o n , Memoire s u r l a T h e o rie du Son, J o u rn a l de l'E c o le P o ly te c h n iq u e 1
9
3
p
l.7
o
,V
0
8
-6 3 -
we see t h a t t h is must be so s in c e any p o in t may be b ro u g h t to r e s t by im­
p re s s in g a c e r t a in v e lo c it y upon th e gas as a whole and in th e n e ig h ­
borhood o f t h is p o in t th e m o tio n must be t h a t o f a wave o f i n f i n i t e s im a l
a m p litu d e and henee v a lu e s o f th e v e lo c it y a re pro p a g a te d w ith a v e lo c
i t y
, t h a t o f o r d in a r y sound o f i n f i n i t e s i m a l a m p litu d e .
v e lo c it y o f p ro p a g a tio n in the f i n i t e
Hence the
a m p litu d e wave must be th e
v e lo c it y o f th e gas u p lu s the v e lo c it y o f o rd in a r y sound in the gas
w hich has th e v e lo c it y u .
Weber
ta ke s e q u a tio n ( 158A) and changes to a new v a r ia b le
(160)
from e q u a tio n (154)
. Then e q u a tio n (158A) becomes:
(161)
T his shows th a t V|
is c o n s ta n t a lo n g a s t r a ig h t l i n e i n the x t - p la n e ,
the c o ta n g e n t o f whose a n g le w ith the x - a x is i s - 7^ •
c . S o lu tio n w ith G iven M o tio n o f P is to n .
£
Hadamard t r e a t s the pro blem o f the m o tio n o f a gas i n i t i a l l y
a t r e s t in a c y l i n d r i c a l tube when th e m o tio n is gene rated by th e m o tio n
o f a p is to n .
I f th e re a re two m oving p is to n s e n c lo s in g th e gas h is
s o lu t io n h o ld s o n ly f o r such re g io n s and such tim e s as th e wave rem ains
a s in g le p ro g re s s iv e one, th e e n c o u n te r o f the waves i s s u i n g from the
two p is to n s p ro d u c in g a new p ro b le m .
..h e re ve r a r e f le c t e d wave e n te rs
h is s o lu t io n lik e w is e does n o t a n p ly .
L e t th e i n i t i a l a b s c is s a o f the
p a r t ic le s a t th e s u rfa c e o f th e p is to n c o n s id e re d be a=o.
L e t the
s u b s c r ip ts zero denote q u a n t it ie s p e r t a in in g to th e m o tio n o f the
p is to n o r the m o tio n o f th e gas a t a=o.
be g iv e n by:
L. Weber, lo c . c i t . p p .475-477
5. Hadamard, lo c . c i t . pp. 177-178
L e t th e m o tio n o f th e p is to n
-6 4 x o = F (to )
(162)
Then the v e lo c it y is g iv e n by
u o = F'
(t o )
(1 6 3 )
where F ' is th e d e r iv a t iv e fu n c t io n o f F .
Prom the f i r s t p a ra g ra p h o f
s e c tio n b above we have seen t h a t v a lu e s o f u o and wo a re p ro p a g a te d w ith
v e lo c it ie s √
)
o
w
(ψ
w it h re s p e c t to th e gas in i t s
i n i t i a l s ta te .
A c c o rd in g ly , a t a tim e t > t o th e v e lo c it y uo and c o rre s p o n d in g v a lu e o f
w o has been communicated to a p a r t i c l e whose i n i t i a l a b s c is c a is :
(164)
The a c tu a l v e lo c it y o f p ro g a g a tio n , a c c o rd in g to e q u a tio n (159) and
the succe ed in g d is c u s s io n is u + √(ϕ'(ρo)) o r from th e sta te m e n t im m e d ia te ly
f o llo w in g e q u a tio n (157 ) i t
is u o + wo X ' (wo) o r from e q u a tio n (148)
u o + wo√(ψ(wo)) . Thus th e a c tu a l a b s c is c a c o rre s p o n d in g to a v a lu e o f u
and w
is :
(165)
Now w ois expressed in term s o f u oby e q u a tio n (151) and u o is expressed
i n term s o f t o by e q u a tio n (163)
. Thus e q u a tio n s (164) and ( 165 )
b e sid es x , a, and t in v o lv e o n ly t oand the e lim in a t io n o f
t o between
these e q u a tio n s g iv e s the sou gh t f o r i n t e g r a l .
d . S in g u la r it y i n the S o lu tio n
We may perhaps most e a s ily p ic t u r e the s i n g u la r i t y in v o lv e d
i f we adopt Webers1 system o f r e p r e s e n ta tio n , e q u a tio n (1 6 1 ).
I n i t i a l l y we are g iv e n
as a f u n c tio n o f x , t h a t i s ,
lo n g the x a x is in th e x t p la n e .
th e v a lu e s o f η a­
Prom each p o in t on th e x - a x is c o n s tru c t
a s t r a ig h t l i n e making an angle w ith t h i s ax is whose co ta n g e n t is
η
Then as lo n g as two s t r a ig h t lin e s do n o t in t e r s e c t in a p o in t on the
p o s it iv e s id e o f the x - a x is , the s o lu t io n is u n iq u e , t h a t is
1 . Weber, lo c . c i t . p . 477.
^
is
.
-6 5 -
u n iq u e ly de te rm in e d f o r a l l p o in ts f o r p o s it iv e va lu e s o f t .
In g e n e ra l,
however, t h is f a m ily o f s t r a ig h t lin e s w i l l in t e r s e c t on the p o s it iv e
s id e o f th e x - a x is .
These lin e s w i l l o n ly f a i l to in t e r s e c t in th e
re g io n o u ts id e o f th e envelope o f th e f a m ily o f s t r a ig h t lin e s and i t
is o n ly th e re th a t η
is u n iq u e ly d e te rm in e d .
by the e lim in a t io n o f η
x = G(η
)
T h is envelope is g iv e n
between the e q u a tio n s ,
- η G '(η )
t = - G’ ( η )
(166)
A n o th e r in t e r e s t in g method o f v is u a liz in g the p ro ce ss is th a t
o f Lord R a y le ig h .2
A t any in s t a n t p l o t a curve w ith th e v e lo c it y
as o rd in a te and the a c tu a l a b s c is c a
curve is deformed w ith tim e .
as a b s c is c a . L e t us see how t h is
A v a lu e u is pr o pagated w ith a v e lo c it y
u + √ (ϕ '(ρ ))(See e q u a tio n (159) and f o llo w in g d is c u s s io n ) .
wave w i l l o n ly r e t a in i t s
T h is v e lo c it y
o r i g i n a l form i f u + √(ϕ'(ρ)) i s con s t a n t.
S u b s t it u t in g the v a lu e o f u from e q u a tio n (1 5 5 ), th e c o n d itio n f o r a
permanent wave form becomes:
D i f f e r e n t ia t in g and m u lt ip ly in g by
we have:
I n t e g r a t in g we have: √ ( ϕ '(ρ )) = B /ρ
S quaring we have: ϕ ' ( ρ ) = B 2 / ρ 2
I n t e g r a t in g a g a in we have
p = ϕ (p )= A - (B2
)
/ρ
(167)
I t is o n ly under t h i s law o f p re s s u re t h a t a v e lo c it y wave w i l l r e t a in
it s
o r i g i n a l fo rm . T his law is n o t f u l f i l l e d
the v e lo c it y wave is deform ed.
f o r any a c tu a l gas and hence
Fo r B o y le ’ s law (20) and P o is s o n 's
l . R a y le ig h , P roceedings o f th e R oyal S o c ie ty , J u ly 8, 1910, p p . 2 4 8 ,2 4 9 ,2 5 2 .
-6 6 a d ia b a tic law (2 1 ), u
tim e t ,
is an in c re a s in g fu n c tio n o f u .
A fte r a
e ve ry p o in t on th e v e lo c it y curve w i l l have moved fo rw a rd
p a r a l le l to th e x - a x is a d is ta n c e
u +
t . A lo n g a p o r t io n o f th e
v e lo c it y curve w hich s lo p e s downward in th e p o s it iv e d ir e c t io n , the
v e lo c it y o f p ro p a g a tio n , u +
, is g re a te r th e h ig h e ru p a p o in t is
on th e v e lo c it y curve such t h a t the s lo p e o f the wave becomes e ve r more
steep u n t i l a t a c e r t a in tim e i t becomes v e r t i c a l a t a c e r t a in p o in t
a f t e r w hich a v e r t i c a l l i n e
c u ts th e cu rve in more tha n a s in g le p o in t
and th e v a lu e o f u is no lo n g e r u n iq u e a t a p o in t .
Hence the s o lu t io n
must be d is c a rd e d beyond the p o in t where th e curve a c q u ire s a v e r t i c a l
ta n g e n t.
The o n ly ty p e
o f v e lo c it y curve w hich may escape u ltim a te
d is c o n t in u it y is one w h ich has no downward s lo p e b u t has an upward s lo p e
from - ao to
+ « in th e p o s it iv e d ir e c t io n o f th e x - a x is .
curve is lim it e d
, however, as to th e d u r a tio n o f i t s p re v io u s e x is te n c e
and when p r o je c te d
c e r t a in tim e .
Such a
back in to n e g a tiv e tim e is d is c o n tin u o u s a t a
-6 7 3 . Waves, n o t S in g le P ro g re s s iv e .
a . G eneral S o lu tio n .
A more g e n e ra l s o lu t io n th a n t h a t f o r S in g le p ro g re s s iv e
waves was found by Riemann1 .
I t o c c u rre d to Riemann to adopt as de­
pendent v a r ia b le s in s te a d o f u andρ
in e q u a tio n s (3 ) and (4) th e
q u a n t it ie s r and s d e fin e d by:
2 r = f(ρ ) + u
(168)
2 s = f (ρ
(169)
) - u
where
f ( ρ ) = ∫√(φ'(ρ)) d lo g ρ
T his fu n c tio n
(170)
f ( ρ ) is e q u iv a le n t to th e f u n c tio n X (w) and we
see fro m e q u a tio n (148) t h a t r and s are c o n s ta n t a lo n g th e c h a ra c te r ­
i s t i c s o f th e su r f a ce i n t e g r a l. S o lv in g th e above equ a tio n s f o r f ( ρ )
and u we have:
f (ρ ) = r + s
(171)
u = r - s
(172)
D i f f e r e n t ia t in g these e x p re s s io n s , we have by means o f e q u a tio n (1 7 0 ):
(173)
The e q u a tio n s o f m o tio n (3) and (4) may be expressed in th e fo rm :
(174)
(175)
S u b s t it u t in g the e x p re s s io n s e v a lu a te d in e q u a tio n s (173) in e q u a tio n s
(174) and (175) we o b ta in :
Riemann, lo c . c i t .
-6 8 -
Adding and s u b tr a c tin g we o b ta in the e q u a tio n s :
(176)
(177)
Since u and ρ
a re g iv e n as fu n c tio n s o f r and S by e q u a tio n s ( 171)
and (172) the above e q u a tio n s a re the e q u a tio n s sought f o r in w hich r
and S re p la c e u and ρ as dependent v a r ia b le s .
L e t us form the t o t a l d i f f e r e n t i a l s o f r and S.
We have by
means o f e q u a tio n s (176) and (177)
(178)
(179)
To f in d th e v e lo c it ie s w ith w h ich r and S a re propagated we must f in d
the r e l a t io n b e tween dx and d t f o r d r = o and ds=o r e s p e c t iv e ly .
From e q u a tio n (178) and (179) we see t h a t
f o r d r= o ,
and f o r ds=o
dx/d
√
+
u
t=
(ϕ '(ρ ))
dx/d
t =
(180)
u -√(ϕ
'(ρ)
(181)
Thus v a lu e s o f r ar e pro p a g a te d in th e p o s it iv e d ir e c t io n w ith a ve­
l o c i t y u +√
)
'ρ
(ϕ
and v a lu e s o f 8 are propagated in th e n e g a tiv e d ir e c t io n
w ith a v e lo c it y -u+ √
).
'ρ
(ϕ
Suppose i n i t i a l l y we are g iv e n th e v e lo c it y
and d e n s ity as fu n c tio n s o f x .
Then by e q u a tio n s (168) and (169)
we are a ls o g iv e n r and S as fu n c tio n s o f X.
L e t us p l o t two c u rv e s ,
ta k in g x as absci s sa and r as o rd in a te i n the one ease and S i n th e
o th e r.
We are g iv e n the curve s f o r r and S a t t= o .
A f t e r a tim e d t , a
p o in t on the r - c u r v e has moved a d is ta n c e (u+ (√ϕ'(ρ))) d t p a r a l l e l to the
x - a x is and a p o in t on th e s -c d rv e a d is ta n c e (u -(√ϕ'(ρ)))d t p a r a l l e l to
the x - a x is .
By ta k in g s m a ll enough in cre m e n ts o f tim e and u s in g the
above c o n s tr u c tio n s u c c e s s iv e ly we may a p p ro xim a te as c lo s e ly as we
please the p ro g re s s o f the w a v e ..
-6 9 -
Under c e r t a in c o n d itio n s we may o b ta in a s o lu t io n w ith o u t
t h is te d io u s g ra p h ic m ethod.
The p r i n c i p le
o f t h is s o lu t io n o b ta in e d
by Riemann depends upona change to l i n e a r e q u a tio n s by in te rc h a n g e o f
the r o le s o f independent and dependent
v a r ia b le s .
T h is may alw ays be
accom plished i f the e q u a tio n s a r e o f the f o llo w in g ty p e , w hich is the
type to w hich e q u a tio n s (176) and (177) b e lo n g :
(182)
(183)
where the c o e f f ic ie n t s U(
---------- are fu n c tio n s o n ly o f u , and u 2
,
n o t o f x f and x ^ . We have,
S o lv in g f o r th e d i f f e r e n t i a l s dx
and d x fc we o b ta in :
E q u a tin g the c o e f f ic ie n t s o f du and a u z on b o th s id e s o f th e se e q u a tio n s
we o b ta in :
(184)
S u b s t it u t in g these v a lu e s in e q u a tio n s (182) and (1 8 3 ), th e f a c t o r
may be dropned and we o b ta in :
(185)
(186)
Independent and dependent v a r ia b le s have here been in te rc h a n g e d and
s in c e th e c o e f f ic ie n t s are fu n c tio n s o f th e indepen d e nt v a r ia b le s
o n ly th e e q u a tio n s a re li n e a r and have th e p ro p e rty t h a t the sum o f
any number o f s o lu tio n s is a ls o a s o lu t io n .
A p p ly in g t h is
tra n s fo rm a tio n
-7 0 -
to eq u a tio n s (176) and (177) we o b ta in :
(187)
(188)
These e q u a tio n s may be
p u t in the e q u iv a le n t form :
(189)
(190)
From e q u a tio n s ( 171) and (172) we have:
(191)
(192)
Subs t i t u t i n g these q u a n t it ie s i n th e r i g h t members o f e q u a tio n s ( 189)
and (190) we have:
(193)
(194)
S ince the r i g h t members o f these e q u a tio n s a re equ a l in m agnitude b u t
o p p o s ite in s ig n , the l e f t members are eq u a l i n m agnitude and o p p o s ite
in s ig n . Hence we have:
(195)
T his is th e ne ce ssa ry and s u f f i c i e n t c o n d itio n f o r the e x i s t ance o f a
f u n c tio n w d e fin e d by:
(196)
(197)
S u b tra c tin g we o b ta in :
(198)
S u b s t it u t in g the l e f t members o f e q u a tio n s (196) and (198) in e q u a tio n
(193) o r the l e f t members o f e q u a tio n s (197) and (198)
in e q u a tio n (194)
we o b ta in :
(199)
-7 1 -
We may w r it e f o r b r e v it y ,
(200)
where m by com parison w it h the above is a f u n c t io n o f ρ
and th e r e fo r e ,
by e q u a tio n (1 7 1 ), a f u n c tio n o f r + s. The f u n c tio n m depends on the
law o f p re s s u re assumed.
F o r B o y le ’ s la w ,
(20) we have:
(201)
and f o r Po is s o n 's a d ia b a tic la w ,
(21) we have:
(202)
Thus we have reduced o u r problem to th e s o lu t io n o f a second
o rd e r l i n e a r d i f f e r e n t i a l e q u a tio n (200) w it h th e c o n d itio n o f s a t is ­
f y in g a g iv e n i n i t i a l c o n d itio n . When w is found as a f u n c tio n o f r and
S as a s o lu t io n o f t h is
e q u a tio n , equ a tio n s (196) and (1 9 7 ), when r
and S have been e lim in a te d by means o f e q u a tio n s (168) and (1 6 9 ), g iv e
two e q u a tio n s between u , ρ , x
o b ta in u and ρ
and t and by s o lv in g s im u lta n e o u s ly we
as fu n c tio n s o f x
t
. As to th e i n i t i a l c o n d itio n to be
s a t is f ie d by e q u a tio n (2 0 0 ), we see from e q u a tio n s (196) and (197)
th a t i n i t i a l l y we must have:
(203)
(204)
Now u and ρ and hence, a c c o rd in g to e q u a tio n s (168) a n d (169) r and S are
in itia lly
g iv e n as fu n c tio n s o f X .
Hence, a c c o rd in g to
( 2 0 4 ), w
is determ ined by in t e g r a t io n , exce p t f o r an a d d it iv e a r b i t r a r y c o n s ta n t
w hich has no e f f e c t on the s o lu t io n , as a f u n c tio n o f x f o r t= o .
The e q u a tio n s :
r = c o n s ta n t
S = c o n s ta n t
d e fin e two f a m ilie s o f curve s in the x t - p l ane. Fo r a cu rve o f th e
f i r s t fa m ily we have f r om e q u a tio n s (178) and (179)
-7 2 -
(205)
and f o r a cu rve o f the second f a m ily :
(206)
L e t us c o n fin e o u r a t t e n t io n to a le n g th o f α β
∂r/∂x
and ∂s/∂x
do n o t change s ig n .
o f the x - a x is f o r w hich
Then a lo n g th e curve r= c o n s ta n t
S in c re a s e s o r decreases th ro u g h o
t and a lo n g th e curve S= c o n s ta n t, r
u
in c re a s e s o r decreases th ro u g h outa c c o rd in g to e q u a tio n s (£05) and (£06)
I f from e v e ry p o in t o f the lin e o ^ j3 w e draw a cu rve o f each f a m ily ,
curves form a n e tw o rk and f i l l a p o r t io n o f th e p la n e
these
w hich is
hounded by the x - a x is , and the cu rve s
/
/
r = r
S = S
/
,
/
where r
is th e va lu e o f r a t <A and S th e v a lu e o f S a t ^3 .
The curves r= c o n s ta n t and S = co nstan t may the n serve as c o o rd in a te in
t h is domain.
In a p la n e in w h ich r and S a re ta k e n as re c ta n g u la r
c o o rd in a te s and in w h ich th e cu rve s r= c o n s ta n t and S =constant are s t r a ig h t
lin e s p a r a l l e l to the c o o rd in a te axes, the l i n e
w hich has no maximum p o in t between th e p o in ts 0( and
becomes a curve C
(3
in s o lv in g e q u a tio n (200) we use Stokes theorem a p p lie d to
a p la n e in th e form :
(207)
where U and V are any c o n tin u o u s fu n c tio n s o f r and s and where the
Rouble in t e g r a l is ta k e n ove r a domain in th e rs p la n e around th e bound­
a ry o f w h ich the li n e
be l a t e r de term in ed
in t e g r a l is ta k e n . L e t v be a f u n c tio n o f r , s to
and l e t
(£08)
Prom the above by d i f f e r e n t i a t i o n we have:
(209)
-7 3 -
L e t us determ ine th e v a lu e o f v so as to s a t is f y th e e q u a tio n :
(210)
From t h is e q u a tio n and from e q u a tio n (200) , e q u a tio n (209) becomes:
(211)
S to k e 's theorem , e q u a tio n (207)
, th e n g iv e s :
(212)
in w h ic h thein t e g r a l is ta ke n around th e boundary o f a domain in the rs - p la n e
A p p ly in g t h is to th e domain α β ξ , bounded by the cu rve C and t h e s t r a ig h t
lin e s r = r ' , S = S'
we have:
(213)
By p a r t i a l in t e g r a t io n we have:
(214)
S u b s t it u t in g the va lu e o f t h is
in t e g r a l in
(213) we o b ta in :
(215)
L e t us s u b je c t v to th e f o llo w in g c o n d itio n s :
(216)
Then e q u a tio n (215) becomes:
(217)
Thus the problem is changed to one o f d e te rm in in g th e f u n c tio n v f o r
when v is known, Wξ, is de te rm in e d by the above fo rm u la from th e va lu e s
o f w,∂w
/r
and ∂ w / ∂ s on th e curve C.
L e t us now c o n s id e r th e d e te rm in a tio n o f th e f u n c t io n v .
T his f u n c tio n must s a t i s f y e q u a tio n (210) and boundary c o n d itio n s (2 1 6 ).
The boundary c o n d itio n s may be s im p lif ie d by a change o f dependent
v a r ia b le .
L e t us s u b s t it u t e
-7 4 -
v = eν V
(218)
where ν is a f u n c tio n o f r and S s t i l l to be d e te rm in e d .
(216)
s u b s t it u t io n the boundary c o n d itio n s /b e c o m e
S ince m is a f u n c t io n o f σ
= r + S, i f we ta k e ν
. W ith t h is
= -∫
dth e above
'm
σ
c o n d itio n s become:
V =1
fo r
σ = σ'
where σ' = r ' + S'
and these are o b v io u s ly e q u iv a le n t to th e c o n d itio n :
V=1 f o r r = r ' and S = S' , t h a t is f o r th e s id e s α ξ and β ξ
By t h is change o f v a r ia b le e q u a tio n (210) becomes:
(220)
I f we suppose t h a t V is a f u n c t io n o f a v a r ia b le ,
z
where μ
= μ
(S - S' )
is a fu n c tio n
(219) may be f u l f i l l e d .
( r - r ')
(221)
o f σ to be d e te rm in e d , the boundary c o n d itio n
I f e q u a tio n (220) a llo w s o f t h is change o f
v a r ia b le s , t h i s e q u a tio n becomes an o rd in a ry d i f f e r e n t i a l e q u a tio n .
T his tr a n s fo r m a tio n a lth o u g h n o t in g e n e ra l p o s s ib le , is p o s s ib le f o r
the ease o f B o y le 's law (20) and P o is s o n 's law (21)
For B o y le ’ s la w , e q u a tio n (201) a p p lie s , and
e q u a tio n
becomes
(222)
(220)
-7 5 -
I f we l e t z
= ((r-r')(s -s '))/4 a 2 t h is tra n s fo rm s in to
(223)
T h is is reduced to B e s s e l's e q u a tio n by th e t r a n s f o r m a t i o n 4
z= - x 2 ,
g iv in g
(224)
The s o lu t io n is th u s :
(225)
The s o lu t io n o f ou r problem is now com plete f o r the case o f B o y le 's
la w .
Fo r P o is s o n 's la w , by e q u a tio n ( 20 2 ), we may w r it e
(226)
where λ
is a c o n s ta n t.
E q u a tio n (220) th e n becomes:
(227)
I f we l e t z = -(((r-r')(s-s'))/((r+s)(r'+s'))) t h i s tra n s fo rm s to :
(228)
The s o lu t io n may be expressed in term s o f a h y p e rg e o m e tric s e rie s as
fo llo w s :
(229)
The s o lu t io n o f o u r problem is now com plete f o r th e case o f P o is s o n 's la w .
P re v io u s to the work o f Riemann, Ampere1 succeeded i n re d u c in g
the d i f f e r e n t i a l e q u a tio n s o f m o tio n to a li n e a r form when the gas
obeyed B o y le 's la w .
He uses q u a n t it ie s
s ig n from th e q u a n t it ie s r ,
s
α,β a n d η w h ich d i f f e r o n ly in
and w r e s p e c t iv e ly o f Riemann. The
e q u a tio n he o b ta in s is
(230)
Ampere, J o u rn a l de l'E c o le P o ly te c h n iq u e 1820 C a h ie r X V I I I , t Xi
, p 177
-7 6 w hich agrees e x a c tly w ith Riem ann's e q u a tio n (200) when m is g iv e n by
e q u a tio n (2 0 1 ).
The in t e g r a t io n o f th is
e q u a tio n was, however, Riem ann's
c o n t r ib u t io n .
Hadamard1 tra n s fo rm s e q u a tio n ( 141) by cha ng in g in de pe nd en t
v a r ia b le s from a , t to u,w and cha ng in g the dependent v a ria b le
a v a r ia b le
from x to
d e fin e d by:
z
= w a + u t - x
(231)
The tra n s fo rm e d e q u a tio n is :
(232)
He then tra n s fo rm s independent v a r ia b le s a g a in from u,w t o ξ,η
e q u i va le n t to the q u a n t it ie s -2S , 2 r o f Riemann.
is :
4(∂2
o
)-f({}=
z/ξη
w h ic h are
The r e s u lt in g e q u a tio n
(233)
where the f u n c tio n f is d e fin e d by :
(234)
T his equ a tio n is o f th e same f orm as e q u a tio n (200) o f Riemann and is
s o lv e d in an analagous manner.
b . A p p lic a tio n to
S p e c ia l Cases
1 . R iem ann's S p e c ia l Case2
A s p e c ia l case tr e a te d by Riemann is one in w h ich i n i t i a l l y
r and S have d i f f e re n t c o n s ta n t v a lu e s on e it h e r s id e o f a segment
αβ
o f th e x - a x is and i n t h is segment change w ith o u t a maximum p o in t
from th e one c o n s ta n t va lu e to th e o th e r . W
e assume:
(235)
Hadamard, lo c . c i t . pp . 162, 163
Riemann, lo c . c i t .
-7 7 -
T h is case is g iv e n by Weber1 and h is tre a tm e n t is here fo llo w e d . In
the re g io n α β ξ o f the x t p la n e (See F ig . 5
) r and S are found as
fu n c tio n s o f x , t by the method o f the p re v io u s s e c tio n .
O u tsid e o f
t h is re g io n we assume a t le a s t one o f th e q u a n t it ie s r and S to be
c o n s ta n t.
By com parison o f e q u a tio n s ( 168 ) and (169 ) w ith e q u a tio n
( 155) we see th a t in t h is case the wave is s in g le p ro g re s s iv e .
c u rv e α ξ ,
r =r1 and S decreases c o n tin u o u s ly w ith o u t maximum o r m in­
imum p o in ts from α
to ξ .
On the curve β
ξ , S = S2
c o n tin u o u s ly w ith o u t maximum o r minimum p o in ts from β to
eve ry p o in t u on the curve
angle
θ
On the
and r decreases
ξ
. From
αξ c o n s tr u c t a s t r a ig h t l i n e u u ' m aking an
f o r w hich
(236)
Then, a c c o rd in g to th e s o lu t io n f o r a s in g le p ro g re s s iv e wave u and ρ
are c o n s ta n t a lo n g e ve ry such l i n e .
β ξ
A lso fro m e v e ry p o in t v on th e curve
c o n s tr u c t a l i n e v v ' m aking an a n g le θ
c o tθ
f o r w h ich
= u + √(ϕ'(ρ))
Then a lo n g e ve ry such li n e u and ρ
(237)
are c o n s ta n t.
C o n s tru c t fo u r
s t r a ig h t l i n e s , α 1 , ξ 2 , ξ 3 , β 4 w ith a n g le s θ 1 , θ 2 , θ 3, θ4
such th a t
(238)
Where u' and ρ'
are th e va lu e s o f u and ρ a t th e p o in t ξ.
Then we assume:
(239)
. Weber, lo c . c i t .
pp 515-518
-7 8 -
The r e s t o f th e x - t - p la n e is f i l l e d
by the l i n e s μ μ' and v v'
lin e s d iv e rg e and hence th e s o lu t io n is everyw here u n iq u e .
. These
Under o th e r
assum ptions th a n (235) the se lin e s w ould n o t d iv e rg e and a d is c o n tin ­
u i t y would s e t in a t a c e r t a in tim e .
When α β is made to approach
zero the above s o lu t io n becomes Riem ann's second case o f an i n i t i a l
d is c o n t in u it y .
2. L in e a r Type Waves i n r andS
1
Lo rd R a y le ig h f in d s a s o lu t io n f o r th e case where r and S
are lin e a r such t h a t we have:
r = Ax+B ,
S = Cx +D
(240)
where A ,B ,C , and B a re fu n c tio n s o f t .
B o y le 's la w ,
(20) is assumed,
Fo r t h is case e q u a tio n s ( 176) and ( 177) become:
(241)
(242)
S u b s t it u t in g th e v a lu e s o f r and s from e q u a tio n s ( 240) we o b ta in :
(243)
(244)
Since th e se e q u a tio n s are i d e n t i t i e s
in x , the c o e f f ic ie n t s o f x in the
two e q u a tio n s must be zero and term s w ith o u t x must be z e ro .
We thu s
have fo u r e q u a tio n s to d e te rm in e the fo u r q u a n t it ie s A, B,C, and D.
Four a r b it r a r y c o n s ta n ts e n te r in t o th e e x p re s s io n s f o r these q u a n t it ie s
b u t two o f them m e re ly d e s ig n a te an a r b i t r a r y o r i g i n o f tim e and a r b it r a r y
o r ig in o f a b s d is e a .
We are n o t in te r e s t e d in the tr a n s fo r m a tio n o f axes
and hence we may express th e se q u a n t it ie s i n term s o f two a r b i t r a r y
c o n s ta n ts .
On s o lv in g f o r these q u a n t it ie s and s u b s t it u t in g t h e i r
v a lu e s in e q u a tio n s (240) we‘"o b t a in the f o llo w in g e x p re s s io n # f o r r and S:
L. R a y le ig h , P roceedings o f th e R oyal S o c ie ty , d u ly 8, 1910, p 256
-7 9 -
2 r = ( H + l)(x/t)) + ( H2 - 1) a lo g t + M
(245)
2S = ( H - l) (x/t)) + (H2 - l )
(246 )
a lo g t + N
where M+N=-2aH
Then a c c o rd in g to e q u a tio n s (f7/-2) we o b ta in f o r th e v e lo c it y and d e n s ity :
u = r-S = (x/t) - a H
a lo g
o rρ
(247)
ρ = r + S = H(x/t) + ( H2-1 ) a lo g t - 1 /2 (M+N)
= Ct(H^2)- 1 e(H
t)where C=eh
x/a
(248)
Thus f o r t h i s case, th e v e lo c it y rem ains alw ays li n e a r ,
re m a in in g c o n s ta n t in m agnitude a t th e o r i g i n .
wave becomes e ve r more g ra d u a l.
ponentia l d i s t r i b u t i o n .
I f H=± l ,
The d e n s ity a t any in s t a n t has an e x we see from e q u a tio n s (245) and (246)
t h a t e it h e r r o r S is c o n s ta n t and in the se
s in g le p ro g re s s iv e waves.
The s lo p e o f th e v e lo c it y
cases we f a l l back upon
I f H = o, u=x / t . and 1 /ρ is p ro p or t i o n a l to t .
The d e n s ity is c o n s ta n t f o r a g iv e n v a lu e o f t th ro u g h o u t th e gas b u t
decreases w ith tim e in such a way t h a t the s p e c if ic volume o f th e gas
is p r o p o r tio n a l to t .
The a c c e le r a tio n e xp e rie n ce d by a p a r t i c l e o f
gas is (∂w/∂t) + (u(∂u/∂x)) and by s u b s t it u t io n o f u from e q u a tio n (247 ) we
see th a t t h is is z e ro .
3 . L im ite d I n i t i a l D is tu rb a n c e 1
Riem ann's s o lu t io n do es n o t a p p ly to a re g io n
r o r S has a maximum.
in w h ich
N e v e rth e le s s , we can le a r n som ething o f the
p ro p a g a tio n o f a sound wave from an i n i t i a l d is tu rb a n c e c o n fin e d to a
re g io n
o u ts id e o f w h ich the gas is everywhere a t r e s t and has a con­
s ta n t d e n s ity j- 0
by a s tu d y o f e q u a tio n s (178) and (179) o r s ta te m e n ts
(180) and (1 8 1 ).
From th e se e q u a tio n s
we see t h a t r is p ro p a g a te d in
the p o s it iv e d ir e c t io n w ith a v e lo c it y
u+
the n e g a tiv e d ir e c t io n w it h a v e lo c it y
-u+
and S is p ro p a g a te d in
. A t a c e r t a in tim e ,
vsihic!
is In d e te rm in a te u n t i l th e e q u a tio n s are s o lv e d , th e h in d e r l i m i t o f the
r v a r ie s meets th e h in d e r l i m i t o f th e re g io n in w hich
re g io n in w hichAS v a rie s a f t e r w h ich thetw o re g io n s se p a ra te and i n ­
c lu d e between them a p o r t io n o f the f l u i d in i t s
1. R a y le ig h , T h e o ry o f
Sound, 1896, Y ol I I p 40 .
e q u ilib r iu m c o n d itio n .
-8 0 -
A f t e r the s e p a ra tio n o f th e two re g io n s , S is c o n s ta n t in th e gas to
the r i g h t o f the in te rm e d ia te re g io n and r is c o n s ta n t in th e gas to
th e l e f t o f t h is re g io n and hence th e re a re two s in g ly
p ro g re s s iv e
waves, one p ro pa ga te d i n th e p o s it iv e d i r e c t io n and one in th e
n e g a tiv e d ir e c t io n .
-8 1 II.
S p h e ric a l Waves o f F in i t e A m p litu d e
S p h e ric a l waves o f f i n i t e
been s tu d ie d .
tio n s
a m p litu d e do n o t seem to have
The case o f steady r a d ia l flo w is g iv e n by equa­
(9) and (10) w hich d e te rm in e u and ρ as fu n c tio n s o f r .
the im p ortan ce o f s p h e r ic a l waves in e xp e rim e n t
to in v e s tig a te somewhat th e case o f f i n i t e
waves.
A t any p o in t o f a f i n i t e
c o n d itio n s
From
we have been le d
a m p litu d e s p h e r ic a l
a m p litu d e wave
we may a p p ro xim a te
in th e im m ediate n e ig hb orh oo d o f t h is p o in t a t th e
in s ta n t by a stea dy r a d ia l flo w w ith superim posed in f i n i t e s i m a l
d is tu rb a n c e .
a.
T h is f a c t le d us to in v e s t ig a t e th e l a t t e r case.
I n f in it e s im a l D is tu rb a n c e Upon a Steady F lo w .
We must use e q u a tio n s (7 ) and (8 )
S ince by e q u a tio n (9)
r 2ρu is
where p = ϕ(ρ )[eq u a tio n (2 2 )]
c o n s ta n t f o r a steady flo w , in
our case we h a v e :r 2ρu = c + s
where c is c o n s ta n t and s is
in fin ite s im a l.
(249)
E q u a tio n (7) then
becomes: (250)
Now l e t s = (∂σ/∂t).
S u b s t it u t in g t h i s in th e above and in t e g r a t in g
we o b t a in : (251)
or
(252)
where R i s an a r b it r a r y f u n c tio n o f r .
S u b s t it u t in g the v a lu e s o f u and ρ fro m e q u a tio n s (249) a n d (252)
in equ a tio n ( 8 ) we o b t a in : -
D ropping h ig h e r powers o f i n f i n i t e s im a l s th a n th e f i r s t , w e o b t a in : -
-8 2 -
(253)
E x p re s s in g ϕ'((R -(∂σ/∂r))/r)
in th e powers o f ∂σ
/r
tha n the f i r s t in
by th e f i r s t two term s in th e exp an sio n
and a g a in d ro p p in g term s o f h ig h e r degree
(253) we h a v e :-
(254)
The term s in d e r iv a tiv e s o f
a re in f i n i t e s im a l s and hence the
term s n o t c o n ta in in g th e se d e r iv a tiv e s must be p la c e d eq ua l to z e ro .
Thus we ha ve: (255)
Thus we have an o rd in a ry d i f f e r e n t i a l e q u a tio n o f th e f i r s t o rd e r
to d e te rm in e R.
Then e q u a tio n (254) becomes: (256)
Let σ =
e
where P is a f u n c tio n o f r to be d e fin e d l a t e r .
Then e q u a tio n (256) becomes: (257)
T h is may be p u t in t o th e fo rm :
(258)
and e q u a tio n (258) becomes o f th e f i r s t o rd e r and ta k e s th e fo r m :(259)
T h is may be p u t in th e fo r m :(260)
-8 3 Let θ = ξ((η2ϕ'(η)-1)/(η2ϕ'(η))).
Then th e above e q u a tio n becom es:(261)
Now η = R / r 2 is known when R is d e te rm in e d fro m e q u a tio n (255)
and hence (261)
may be s o lv e d f o r θ as a f u n c tio n o f r .
W orking
back th ro u g h th e tra n s fo rm a tio n s made ,P th e n is known as a fu n c tio n
of r.
V alues o f 0~ a re th e n p ro p o g a te d w ith a v e lo c it y - P'
S ince e q u a tio n
a s o lu t io n .
.
(256) is l i n e a r , th e sum o f any number o f s o lu tio n s is
By choosing "a " com plex, CTbecomes a damped p e r io d ic
fu n c tio n .
b . F in i t e A m p litu d e Wave.
L e t r 2ρ= ∂F/∂r
and s u b s t it u t e in e q u a tio n (7)
.
Then in t e g r a t in g w ith re s p e c t to r we o b t a in : /t)+
(∂F
The fu n c tio n f ( t )
F u = f
(t)
(262)
may be re g a rd e d as in c o rp o ra te d in th e f u n c tio n F.
The r i g h t member o f e q u a tio n (259) may th e r e fo r e be re g a rd e d as z e ro .
U sing these v a lu e s o f u and P in e q u a tio n (8) we o b t a in : - I
(263)
I have n o t been a b le to s o lv e t h is e q u a tio n .
The s o lu t io n o f
the p re c e d in g case sh o u ld be a b le to be employed to b u ild up a
s o lu t io n f o r t h is case s im ila r to Lo rd R a y le ig h 's developm ent o f
P o is s o n 's in t e g r a l f o r a p la n e p o s it iv e p ro g re s s iv e wave as
d iscusse d im m e d ia te ly a f t e r e q u a tio n (1 5 9 ).
-8 4 E.
Plane Waves o f Permanent Regime W ith D is s ip a tiv e
F orces.
Waves o f perm anent regim e a re those f o r w hich th e
v e lo c it y wave does n o t change i t s
by e q u a tio n ( 1 6 7 ) ,t h a t t h is
fo rm .
We have a lre a d y seen,
is o n ly tr u e f o r a c e r t a in law o f p re s s u re .
T h is law can o n ly be m a in ta in e d by c o n d u c tio n o f h e a t o r by v is c o s it y
o r o th e r body fo rc e s a c tin g upon th e gas.
A p la n e wave o f
permanent regim e can be b ro u g h t to r e s t by a p p ly in g a c o n s ta n t
v e lo c it y to th e gas as a w h ole.
a case o f steady m o tio* n .
It,
t h e r e fo r e , may be tr e a te d as
The e q u a tio n s a p p lic a b le a re (3 ) and
(4) in w hich a c c o rd in g to s u b s t it u t io n ( 1 1 ), we s u b s t it u t e f o r
— —'Qih ,X — —
*
Assuming ste a d y m o tio n th e d e r iv a tiv e s
P ^X
P <*Y
w ith re s p e c t to t d is a p p e a r fro m the se e q u a tio n s and th e e q u a tio n s
become o r d in a r y .
E q u a tio n s (3) and (4 ) th u s become
pu = m
(264)
(265)
We s h a ll c o n fin e o u rs e lv e s to a ste a d y wave in w hich a t s u f f i c i e n t
d is ta n c e s in th e p o s it iv e and d ir e c t io n s , th e gas becomes o f a
u n ifo rm s t a t e .
I f th e fo r c e X is due to v is c o s it y ,
a t b o th te r m in a l s ta te s .
the te r m in a l
A ls o th e n th e re i s
s ta te and hence J d Q ,
i t va n ish e s
no h e a t c o n d u c tio n a t
is zero where Q, is th e h e a t
re c e iv e d by an elem ent o f gas m oving to a p o in t x , and where
the in t e g r a l is ta ke n from one te r m in a l s ta te to the o th e r .
I.
E n tro py C o n d itio n .11
From e q u a tio n s (264) and (265) we h a v e :(266)
1. R a y le ig h ,P ro c e e d in g s o f th e R oyal S o c ie t y , J u ly 8 , 1910,p p .258-260.
-8 5 The fo rc e o p e ra tiv e upon an elem ent o f mass
is X p <)^and the
t o t a l momentum g iv e n to the elem ent i s : (267)
A c c o rd in g to e q u a tio n s (264( and (266) i f
the in t e g r a t io n is take n
from th e i n i t i a l te r m in a l s ta te in w hich th e v e l o c i t y , d e n s ity
and p re s s u re are u ( , p
, and p f R e s p e c t iv e ly , to th e f i n a l te rm in a l
s ta te in w hich the se q u a n t it ie s have th e v a lu e s u , P ,and p ,
r e s p e c tiv e ly .
The p r i n c i p le o f c o n s e rv a tio n o f momentum
a p p lie s to th e v is c o s it y fo rc e s between th e p a r t ic le s o f the
gas and hence lx p < J t/in e q u a tio n (267) is z e ro .
T h e re fo re we have:
(268)
which is e q u iv a le n t to e q u a tio n (167) a p p lie d to te r m in a l s ta te s .
The energy expended by th e v is c o s it y fo rc e s is
From
e q u a tio n s (264) and(266) we h a v e :(269)
By e q u a tio n (264) th e r i g h t member is a fu n c tr o n o f p a n d p .
F or B o y le 's la w , (2 0 ), and P o is s o n 's a d ia b a tic la w ,
(21) , t h i s
q u a n tity is p o s it iv e o r n e g a tiv e a c c o rd in g as p ^ is le s s o r
g re a te r tha n p^ .
The c o n d itio n o f in c re a s e o f e n tro p y o r th e con­
d i t i o n o f d is s ip a t io n o f energy im p lie s
n e g a tiv e .
T h e re fo re , s in c e
t h a t t h is q u a n tity be
th e f i n a l p re s s u re exceeding th e i n ­
i t i a l p re s s u re denotes a wave o f c o n d e n s a tio n ,o n ly waves o f con­
d e n s a tio n are p o s s ib le .
II.
Heat C o nd uctio n.
Rankine s tu d ie s th e h e a t tr a n s fe r s necessary to m a in ta in
Earnshaw 's law o f p re s s u re ,
1. R ankine, P h il.
(167) i
T r a n s . , 1870, v o l.
He f in d s f o r a p e r f e c t gas
160, P a rt I I ,
p .2 7 7 .
-8 6 th a t to e f f e c t a change o f p re s s u re dϸ , a c c o rd in g to t h is la w ,
the q u a n tity o f h e a t added must b e :(270)
I n t e g r a t in g from the i n i t i a l to th e f i n a l s t a t e , we have s in c e
(271)
where p0 and v0 a re th e p re s s u re and s p e c if ic volume a t any p o in t
in the wave.
I f p0 , v 0 , a re i d e n t i f i e d w ith p 1 , v 1 , we o b t a in : (272)
S ince v 1 = 1/ρ1, by e q u a tio n (264) m v1 is eq ua l to u 1 .
M ul­
t i p l y i n g e q u a tio n (272) by v 1 we th e r e fo r e o b t a in : u 12= m
2 v12 =v1{((1/2)(y-1 )p 1 ) + ( ( 1 / 2 ) ( y + 1 ) p 2 ) } (273)
which is the square o f th e v e lo c it y o f p ro p o g a tio n w ith re s p e c t
to re g io n 1.
T h is is g re a te r th a n v e lo c it y o f sound o f i n ­
f in i t e s im a l a m p litu d e , f o r w hich u 12= y p 1,v 1 , f o r a c o n d e n sa tio n
and le s s f o r a r a r e f a c t io n , as we see from th e above fo rm u la .
The a b s o lu te te m p e ra tu re
is d e te rm in e d f o r a p e r f e c t
gas as f o llo w s : (274)
by the r e la t io n in e q u a tio n (2 7 1 ).
everywhere K(d
/x),
θ
The flo w o f h e a t is
where th e c o e f f i c ie n t o f h e a t c o n d u c tio n K may
be a fu n c tio n o f th e c o n d itio n o f th e gas.
from
I f we re c k o n Q
the i n i t i a l c o n d itio n o f c o n s ta n t p re s s u re p 1 ,
(275)
F or the d i s t r i b u t i o n o f p re s s u re Rankine f in d s from th e above
e q u a tio n :(276)
Under th e s u p p o s itio n t h a t k is c o n s ta n t, Rankine £±xdx
in te g r a te s e q u a tio n (276) and o b t a in s : -
-8 7 (277)
where
q = p - (1 /2) (p1 + p2) and q 1 = (l/ 2 ) (p 2- p1 ) ,
from th e p la c e where q e q u a ls z e ro .
the wave is i n f i n i t e l y
x b e in g measured
A lth o u g h m a th e m a tic a lly
lo n g , p r a c t i c a l l y th e t r a n s it io n is e ffe c te d in
a d is ta n c e comparable w ith
k /(m C (y + 1 )).
change s ig n , th e n u m era to r in
S ince
dx/d
p must n o t
(276) must be alw ays p o s it iv e
,and hence
p2/p1 must n o t exceed (y+1)/(3-y).
III.
V is c o s it y . 1
For v is c o s it y we must p la c e in e q u a tio n (265)
X = (4/3ρ)(d/dx(μ(du/dx))) a c c o rd in g to Lamb.2
In t h is ease a ls o R a n k in e 's c o n d itio n h o ld s ,
see e q u a tio n (268)
E q u a tio n (265) in te g r a te s i n t o : (278)
For B o y le 's law we f in d by i n t e g r a t io n : -
Here d x /d v never changes s ig n and a permanent wave o f co n d e n sa tio n
is
always p o s s ib le no m a tte r what the v a lu e o f the r a t i o p1/p2
may be.
The v e lo c it y o f p ro p a g a tio n in t o the r a r e r medium is :(279)
IV .
V is c o s ity and Heat C o n d u ctio n .3
Here the h e a t change in an elem ent c o n s is ts o f two p a r ts
dQ1 and dQ2 , th e fo rm e r the h e a t re c e iv e d by c o n d u c tio n and th e l a t t e r
t h a t developed in t e r n a lly under v is c o s it y .
F or
dQ1/dx and dQ
2/dx
we f in d th e e x p re s s io n s : (280)
1. R a y le ig h ,P ro c e e d in g s o f th e R oyal S o c ie ty ,J u ly 8 ,1 9 1 0 ,p p .269-271.
2. Lamb, Hydrodynamics ,1906, p p .53 5 ,5 3 8 ,5 3 9 .
3. R a y le ig h ,
lo c . c i t .
pp.271 -2 8 2 .
-8 8 U sing Rankines e q u a tio n (2 7 8 ), e q u a tio n (280 becom es:-
From t h is e q u a tio n i f we recko n Q1from th e te r m in a l s ta te v ,
(y - 1) Q = 1/2( y + l ) m(v1-V)(v-V2) +(4/3)m(μv(dv/dx)).
E q u a tio n (278) s t i l l a p p lie s and e x p re s s in g θ as p v / R , we may
o b ta in ,b y s u b s t it u t in g
θ
th e v a lu e o f p from e q u a tio n (278) ,
as a f u n c tio n o f v .
Then s u b s t it u t in g in the e q u a tio n o f c o n d u c tio n , we o b t a in : (282)
R a y le ig h assumes M a x w e ll's th e o ry as c o n n e c tin g th e s p e c if ic h e a t C,
a t c o n s ta n t
v o lu m e ,w ith th e c o e f f i c ie n t o f h e a t c o n d u c tio n K.
F or the th e o ry w hich assumes a m o le c u la r r e p u ls io n in v e r s e ly as the
o f the d is ta n c e ,
C a llin g t h is r a t i o h and w r i t i n g
f o r μ/m e q u a tio n (282) becom es:(283)
W r itin g
ξ
f o r v2 and U
f o r μ'(dξ/dx'), t h is e q u a tio n is reduced to
one o f f i r s t o rd e r, nam ely,
where f ( ξ ) = ((3(y+1))/8y)((√ξ1+√ξ2)/(√ξ3)) - (3/2) - h
and
F ( ξ ) = (3/4)h ( y + 1)(√ ξ 1
- √ ξ )(√ ξ
When U is determ ined as a f u n c tio n o f
- √ ξ 2 ).
from t h is e q u a tio n , x
found by sim p le in t e g r a t io n o f the e q u a tio n :-
R a y le ig h s o lv e s e q u a tio n (285) by a s e rie s o f a p p ro x im a tio n s .
is
-8 9 F.
Summary and O u tlo o k .
we have tr e a te d c o n tin u o u s and d is c o n tin u o u s waves.
We were a b le to t r e a t d is c o n tin u o u s waves in th re e d im e n sio n s.
The v e lo c it y o f p ro p a g a tio n o f a second o r h ig h e r o rd e r d is c o n t in u it y
w ith re s p e c t to th e gas is t h a t o f sound o f in f i n i t e s i m a l a m p li­
tu d e .
A lth o u g h th e re is much l i t e r a t u r e
on th e s u b je c t o f f i r s t
o rd e r d is c o n t in u it ie s i t appears t h a t such waves are
p o s s ib le t h e o r e t i c a ll y .
In r e a l i t y ,
even i f
p ro b a b ly n o t
th e v is c o s it y and h e a t
co n d u c tio n were everywhere e ls e n e g l ig ib l y , th e y would n o t be neg­
l i g i b l e a t th e d is c o n t in u it y and these f a c t o r s would p re v e n t th e
fo rm a tio n o r th e p e rs is te n c e o f th e d is c o n t in u it y .
S ince second
o r h ig h e r o rd e r d is c o n t in u it ie s a re pro p a g a te d w ith a v e lo c it y o f
th a t o f o p d in a ry sound w ith re s p e c t to th e gas, P o is s o n 's in t e g r a l
f o r a c o n tin u o u s wave h o ld s i f
th e v e lo c it y wave is d is c o n tin u o u s
in s lo p e o r c u rv a tu re , o r h ig h e r d e r iv a tiv e s o f th e v e lo c it y w ith
re s p e c t to x a re d is c o n tin u o u s .
I f th e re is a g iv e n i n i t i a l r e c ­
t i l i n e a r d is tu rb a n c e in a lim it e d r e g io n , th e curves f o r Riem ann's
v a r ia b le s r and s a re pro p a g a te d in th e p o s it iv e and n e g a tiv e
d ir e c t io n s , r e s p e c t iv e ly .
A f t e r these waves s e p a ra te we have
two s in g le p ro g re s s iv e waves, one in th e p o s it iv e and one in th e
n e g a tiv e d ir e c t io n .
The tim e f o r com plete s e p a ra tio n to ta ke p la c e
can be a p p ro x im a te ly d e te rm in e d .
The l i m i t a t i o n
in v o lv e d in
P o is s o n 's in t e g r a l, b y w hich a f t e r a c e r t a in tim e the v e lo c it y and
d e n s ity a re no lo n g e r u n iq u e ly d e te rm in e d , is r e a l l y v e ry s e rio u s .
The th e o ry o f sound waves o f f i n i t e
a p p lic a tio n in e x p lo s iv e waves.
a m p litu d e sho uld f in d i t s
Now v e l o c i t i e s o f e x p lo s iv e
waves have been measured w hich a re more tha n tw ic e th e v e lo c it y
-9 0 o f o rd in a ry sound.
L e t us suppose t h a t th e v e lo c it y o f a wave
is measured by the v e lo c it y o f p ro p a g a tio n o f th e maximum p o in t
o f the v e lo c it y c u rv e .
The maximum p o in t o f th e v e lo c it y curve
under t h is c o n d itio n would o v e rta k e th e f r o n t o f the wave w hich
o rd in a ry
moves w ith th e Av e lo c it y o f sound in a d is ta n c e le s s th a n th e wave
1
in
le n g th o f the wave. Hadamard s ta te s t h a t Ath e e xp e rim e n ts o f
p
V ie lle ^ on e x p lo s iv e waves, d is c o n t in u it y would ta ke p la c e in
a few c e n tim e te rs .
The in t r o d u c t io n o f
v is c o s it y and h e a t con­
d u c tio n would seem to be ne cessa ry to p re v e n t th e wave
becoming d is c o n tin u o u s .
c o n d u c tio n th a t have
from
The o n ly waves u n de r v is c o s it y and h e a t
been s u c c e s s fu lly tr e a te d are waves o f
permanent regim e in w hich th e gas passes from one u n ifo rm c o n d itio n
in f r o n t o f th e wave by c o n tin u o u s change w ith o u t maximum o r
minimum p o in ts to a n o th e r u n ifo rm c o n d itio n back o f th e wave.
We have seen t h a t a s in g le p ro g re s s iv e wave g e ne rated from a
lim it e d i n i t i a l d is tu rb a n c e must have a maximum p o in t .
To such
a wave the s o lu t io n f o r the presence o f v is c o s it y and h e a t con­
d u c tio n does n o t a p p ly .
We c o u ld i n f e r p ro b a b ly t h a t th e fo rw a rd
s lo p e o f the wave would a t t a i n a s ta te a p p ro x im a tin g a permanent
regim e and t h a t f o r th e backward s lo p e o f th e wave, P o is s o n 's i n t e ­
g r a l would rem ain a p p ro x im a te ly c o r r e c t and t h a t , in accordance
th e r e w ith , t h i s slo p e would become more g ra d u a l.
As to f u t u r e in v e s t ig a t io n , we need a g re a te r t h e o r e t ic a l
knowledge o f waves o th e r tha n p la n e waves.
fin ite
A s o lu t io n f o r a
a m p litu d e s p h e r ic a l wave would be v a lu a b le f o r such waves
can be more n e a rly ap p ro xim a te d in e xp e rim e n t tha n p la n e waves.
Waves, supposed to be p la n e , p ro p a g a te d in tu b e s , a re a ffe c te d by
1. Hadamard, lo c . c i t . p .1 8 5 .
2. V ie ll e , Comptes Rendus , 1898-99,
-9 1 the f r i c t i o n
o f the gas upon th e tu b e .
The
in flu e n c e o f v is c o s it y
and h e a t c o n d u c tio n on such waves a ls o needs to be s tu d ie d .
a ls o need a s tu d y o f th e phenomenon o f r e f l e c t i o n .
fir s t,
as the most sim p le c a s e ,fo r p la n e waves.
We
T h is can be s tu d ie d
B e fo re we can stu d y
r e f r a c t io n we must know more a b ou t waves o th e r tha n p la n e waves.
As to e x p e rim e n ta l v e r i f i c a t i o n ,
th e v e lo c it y o f p ro p a g a tio n
o f a wave in to a re g io n a t r e s t can be measured, and in so f a r as the
wave a p pro xim a te s a p la n e wave, th e d i f f e r e n t fo rm u la e f o r th e v e lo c it y
o f p ro p a g a tio n can be te s te d .
The re s is ta n c e o f the gas to th e m o tio n
o f a s o lid th ro u g h i t , i f we assume th e wave ge n e ra te d in f r o n t o f
the s o lid to be p la n e , can be used to d e te rm in e th e r e l a t io n between
the p re s s u re and th e v e lo c it y in a p la n e wave. R a y le ig h 1 assumes f o r
p r o je c t i le s
p r o je c tile
th a t the fo r e p a r t o f th e wave g e n e ra te d in f r o n t o f the
is one o f perm anent regim e under v is c o s it y and h e a t con­
d u c tio n and t h a t from th e v e lo c it y a s s o c ia te d w ith th e f i n a l s ta te o f
t h i s wave th e re is a second tra n s fo rm a tio n under w hich as the head
o f the p r o j e c t i l e
is approached th e gas changes a c c o rd in g to th e
a d ia b a tic law and th e v e lo c it y becomes th a t o f th e p r o j e c t i l e .
These
two tra n s fo rm a tio n s are supposed to ta k e p la c e in such a s m a ll d is ta n c e
th a t in sound ph oto grap hs o f r i f l e
b u lle t s ,e t c ,
th e y appear as one.
The d e te rm in a tio n o f th e shape o f th e d e n s ity wave does n o t le n d i t s e l f
r e a d ily to e x p e rim e n t.
The use o f a membrane w ith a d e v ic e to re c o rd
i t s movements and upon w hich th e wave im pinges is o f l i t t l e
the fr e e p e rio d o f the membrane and i t s
a v a il s in c e
i n e r t i a a f f e c t the r e s u lt s in a
manner which cannot be c a c c u ra te ly a llo w e d f o r .
I f §. wave c o u ld be m ain­
ta in e d steady in a tube b y - im p re s s in g a v e lo c it y upon th e gas as a w hole,
1. R a y le ig h ,P ro c . R oyal S o c .,J u ly 8, 1910, pp . 282-284.
-9 2 -
the shape o f the wave m ig h t be fo u n d .
on th e r e f le c t io n o f f i n i t e
around o b s ta c le s ,
P o s s ib le e xp e rim e n ts
a m p litu d e waves and t h e i r c u rv a tu re
w hich can be s tu d ie d by means o f sound p h o to ­
g ra p h s, a w a it f u r t h e r advances in th e o ry .