k,1.,') T, , - I:., - .

I I -- -
. 'r , j , '. 11I. - ,,,
I :,
_, , ', " L
. . I6 , l , . , I1.f. -, _ _I _, -11 ', , .. ".I.. , r I - ' 11 1411
, I ,,

- -
.
::A,7-- f ! , ',!' ., , - - j 11
I - :57",Ty
. -, -, , - - ". '. , 5 - h, .
- - I - - .! - , , ' ,, - ,[ J , ?
IV- ,CL .; - - '. , 77- . Jl ,-jrj--r-.TU lp.
, ." I J -Tl," WT
- ."l
,,
: , " I- ,-- " i i, I--
1, l7 w, [j I, - ..-- ,,I
- 11 w ? 1-I,. , , ,- ,' , , , .z " -,- , ,- ; - -
I I ,171 . , , 111, .1 , A
. I Lil , -1 1I . , I.. - .,
'- ' fi -
!": 11
-, . , 4 - - I ", , - r , , ' ',
7
- -I' ;. , , -I J( . ), . ) .- - - ,I---
- , l I - - , I,, - -,, -
, , " X , . I --. , . r. Y , , ,- - , "- " : , I. " .- .M
I'll,--
; ... I - - - "
--- - - -- 7 -
- I. 11i--. , , - , . , , ,
11 1% M , IL 'L
-- , -,.!-
- . - I- 11
r :,,
, . _. . - ".! , - .,- , I " . I I - .1 ...
,- r. i . " l 11
: 11".. - - - . I 4 ,,, -
,,." , !
-- w- -,,,A - .-. - - -- - - - ; -- - . - -- - - - I- - - N -,... ! ,, ,, J. I
" , , I. . "
. -
Tt . - . ! . . , .,I -. -;'l , ..-,e . I -
, 1, Krtr
- - -' -'- - " L*
,
. -r- -,--.-. . - :C - - - .. . - " , I ,
- - ,
...
- - ...- .I - I--e, ''IF n I ,: , , -Ir - I , I . - -, - . 1:
.k , - --- - 1 , .- , -, A ! , - r
I JI- "Y ' .u . , . , il'r b.
'' I , ,
11,,
MIT LIBRARIES . , , !, a .'. ,i ._ ,-- , 4 . - '-,-%'r,-, I r,_ J. j -' " - . . " - , ,.
I
- _ ''r, " -,? , .",.11
- --! -, I -
.1
,
-
" j P. - J p - p " I- --- -,. -- .,- , ,. ,
,' ,
ti lL " ,
,,L - II P- ,r.,,-. I' _- . ±.' ' -) 721 f -. .1,,- ,, . , 11, ' ., I . . - L- j
T.", . , Ijft '. % -1
i'l - .,
.T!, !."K 's --, i
I 1. - - , , .. , , 1-1 , -
'fir * - T .1 ,, , .,, ,.- i - .v j
i": '-,, ". j-'-J---gl ".- " '-"-" L 'J, , -
7:1 . .. " ," , . ., I-
. . Ij - , 'PV.,,.!t .".Lv% , ll* 'r;.1 _
-- , i
I. -
.
-.& z. ! . 4 . , :
. 'r , ., . ', , R " I . ,- . . l , -lo -
- V. . LI ! - , , , . -.. . . F'r
.. - , ', j';,.1 r , ' ' ', . ,
-
. -'
- " "I . , q ,!- , 4 " "-",;Q - ,
. I 11111
- , I- ":.-, , P , IIV,I. U , ,;-. ,, '*rl ., ..,[I ., JL- Ul ,,,
, - "' "
, ": j II , LI r. . lp
- ;- 11
1111
1111
I 11111
11111111 , rj'. - -,,- , Li - h 4 -- . , i . P. - - . , - , .1. 4, I
rT I-- -, -, , J-1 % -, ,
" - ,: , , ,
. I-
-
j
-
, ,; : ,, . 'y - & " , " , '.*, Z
14 -, q j . L I ., _ ., 111. . -
'!. . 1,: _ _ , r I., - , ,, - - . I . 1:' r.

- I ,..-,,;;.
-
I
- - -TJ.-I
p1-. -I .-
- , ,1171 ; .. 41CITII --M7Ij ,. r . 1.'rrI-r, . !'_._ ' i' ' '' .rl-- , C.,4 -$.;", -,,. .,,:,-I-,11-
,I .,--.. , .L_.'---- I---:1r_-,-,r
111i
,.,,-'- , .'.r
,
, '-','
r''-; I,- L.r' _,-,r,'
. :'T .
. I
,..'r I I
'r
,,, "
T
,, ,
,
'.
" ., 7.
. Ir , . Y
- ,,.. -
. ,, _44 -
z
. . - I
I 11 - z
o 4- (1: ' ,,, "fI - -.-.
f,, .
I
- I ,
il
ff ;
,
:- -- 'j iI" l
-- (14 " .: - -
It . , -"
- "-, , , , , , . _ 7 :: I .
-'

.,I,I,. :,-_'.I,.--,- ,,1-1 r.!-' '- ,:-,- -. - ,:, , __- % . r : L, , 1.

-',- , , , , , I, , .
-'_"
:A , .; 'I -, . I , , -r , - -

. ,,Zi" I --. ; ,"

, , :

, . , ,, - ,,, j, ,. .. ,I
"'
, . '

.'-,,I t'I, :...-I.; ,rr,,--

,
Ili I - -. r 11- . ':r.,,,,,r-,,,.,
,-z
7 ,"t"1- .m ..r I'," - . --'. r ,_"'.--,-
-7
r __ ,. . r' , r,' ...
- A
, ,,-: I- 'i:""':
l.,. Ir I r--,',,, - ,,,. I - I :L..::: I - I r ' ,
'.. -i . .I I, r 1_,- -' ' r_.1, ,"-,'I ,, :L
, ., r ,- 1,
I r - ' - I
I ., -,- .,I- F , , ..,.:--- .I r- ,- '!
- -..
- .. , I- ,.

, , , _
: '- " ,I r -
r _ -
,r.
I - r . - ._, -, J1.,, -,__ I - .:., -,. .,.-. - I, .1-II [,, ,.'. , :
- f - ", -, .r,
' - 1,
. , , . ,,, j II
. Ill'. II,-', . L ,
r - - , . . .,I - Ir-- I . , L 1, . ." r.- - mt: ,.---, p'rr_r'_,I,.,,,r, .,-,,. I ",-,, L,- L ,- ,,- , - , -,Ir-j - "r
- I.'. ,_-, !,,.-,I.I 7,1--"1-,- 'rr", I "_,'!...,'_,:,;r, -
.,,f :rr-" - .- ,- o..II.... I- .I,I,,,.,*.
.
'1 -P .
.' r I_
,
I--"7
-I
- - .1:
I 1'
7
. _ -
r 1 r? , r,r ,
_ r
.
,
, , , -',, : , . -' ,
r ..
-l"-
-
, -,I, , , , L
, , -:" Ll
, - ,
.
;,;-r
,,', i, r',
: ", 'Z T : , ,
, ,I .,- ", - _:
-, _. ' I: ,
--, ,% ,
' 1, , ," ,,,r r
. 1, , ;_-
, , - ' Ir, ,_,
.
' ,1' - r' ,
t, ' '
,
.
', ,-
,
_
,._ , = , ,
-, - ,-, , ,:--- - -,, -
. I-, .;!
- r
. r
, , r. - : - , , ,
11
,
,
I . I
?
I I , -, F
;,, , 2 , I , .. ' . - I - --. ,- -- ' - . . , __ _ . r- .. -
I;-
, " ,:; - ,, '-' -- . , , ' ', ', r,, I . ' ;' ] r- e':r - ;.' - , .-- , . , ,- -, ; -,. , , - 1 .
l ""..,- . I , ,-,I, -
, ,,, " I -.
, - . - I . ' ; ,. - L I- , r I I I I" , . . - i - , ,,
- , . . 7 %, I
, , , - 'r , r,. . , ' -, ..!
! ,I- r I I, , . _ -- . L 11 . ; :- .7: L, ' . . ,
I ' iI , , ;,, ::, ' L , r: . . : - , , , , - - , :: , " : r,
, , I, , I , II
I--- , ., : _ , f .- - ., : : 7 r , . . . , , , , , _r . - , " - -'. - .- - ,- -;- -r r ..r , , - ,.. r I,- I . , I I I, , -- , , , ,I
-, r',
: 1. - I- ': : - ., I " , t;-, - - , , '__ 'r ; I I . I 1. , , _., , - " , -- . r ' o - ':,- , . I a ,I -: -I .
' -, ; : 7. , ' !.I'I
, - , , r!n . r - , - :-, - lr - I.. - -.1rI ".,; : I
- Ill 1'
'. ,
I : ',il"'
-' .. : . ,,, -,' L
r, '--, , r , I -, . i I ! I -11- , , ,. , - % 4 ,- , ,v , . , ,. ., . , - ' .' , _ _ , _
, _ , . , " , - r . ' "" '-
I 111'1] - 11' I
, ,. ' L,, - , 2', ',,' , ' - :-r,' , - ' ' , . ' L_ r . I , . ". , .I, _r : --,: ,, ,
_ - . , ., , - , .-.--, I-. ll l ,.11 I " ,- - '
- , ,' - r,I. , ., , I,-1,. - ., -- I , , . .. --
'_ ,-,r - , -, " ? 1, ., :, , , I -, . I - - I . 1:, I ,
.
1 4
. " ' -
_ -
, , 7 ., ,l ,-
- ' , ' " ,
' r .,
I
-';-- -'-:7L . ' -
-'- '-
1 ;-' r -'-_' . - , . , , . ,__
, L . 'r ' ., 1-11. :_
.
_ _ ',. r _
.
. _,
' ,
,,
' I., , ,_,
,- ,_ , , - _.
', , ._ v , _
11C., , , ..,
--!,
. .-
?-
I
" , - y - - ; I"
, , ,;:, , ;
. ,,-
--
, , 'r',
: "- . -' ':'r I I : r ..
-
,-
11..
-I - I'11,
- P?
pf , .

'7 r , r.r . , _ I ;; r,-- r'

, r ,
I
, I ., , -
1. ", , -, . I,- , .- ,-,,, II. I,
- ' _,
. . _
' --
, ; I
r_-r' ,,, -% ",
, -." ,
. rr , _r
,
r, - ':' , -r , , . ,
, 'r ' ,',
r " -.1-I ' , -L ?' , ; ,,r*
4 , - " :' 4 , ,
._ ,
' - ' (-_11 . -' _,,- .r
_ "',r ;r, , ,_
;l.- :' A n - , .. -, :,,,
"r
,
. , ", r,.,
-
,,,,r,
_z,. --I
,,I. ;," :.
, _,
_-
' :. '
., r,
..t.;,-
: , "- , I-' r,"
11
-
"' ,,,, .r-P,
, _
11_r " '.-,. -, ,
r"' " -. lp', q r ,-.r -;r ,-,,,r,'."
- . ,
-, _', -',. r": _:-_1' .,-, '- ,' '..' ' , 'r'

:- _ I !I-- . -,
I , :I:
, ''
r-,
, ,r
,-
II . -'
,-r I V -I.

- '- r-l, - ,,, r, -( .,-, A-' ' , - - , - , , ." , , "r ,

_ , ,i._r - r ' ' '- ." .r _ , -_i ,'-.. -. I,-,, ,', r ,
- , -1
, I ,
L
I I - I I . I- , . - ,. i , , 1: II. , I 1. , , -;; , .:, -, ;I, , 'I . - ',,'2:L tL. ,, , , . -,", Iel -'., ," , %,.,r . .) , - -' ., ' -' %. 'r :.' . - __, _ - ., . 'r ' , 1. I . , l , ..
- 1: : I -- .. " - . ' ' - ' : , - - . . :, 'I '*
. , i, ,. . - I,. , I ,- , .
4--. ; , -- ," : '' _ - . _,, I , . , -, , - 2 ty, , ,,,
; .,, -, !-
4, ,-; '
t ' rr;' -
.
,. I:., r , ' _
II .-i - __r ___ 5 -- _ __,- _7 _r ;, , . _ - rI . - _
- r - r,r, , ,;,.,-7,
I r - I -
-- 'L,1,1, - I , ; ., - .. ' ,- _r _ '. .r. ;,_ - - "
'. , - r, j . -r - I I'- . . , J - . , . , , I , ,'- " I- ,- ' - ,: r' - - r J,' ).. j .- , .-, .*P. , -,* 1-. , , , : .1, _ _' , ,
. , -, -- , - r._ , , .,, ',_
- -
- ,, ,,'
_r , , ,L4
, _ "
r.r -' '-'r
_,
, , --r
I
" ,
,' -,
2
_ !, , , L ,
- l . , u%
-,
_'_1 - ' I - -, ,, . j.I r . ' . -rl I ', .1 I - . I - " ,, r- r_ ._L ,-
.
, -
i 4 - ,f I - --. -" . . 6.1 )
',
I
_4 j
-
L ,
- I. I-I. , .-
, " , ", -
, , ,- , ,
, , .. - . - -. ' ., .- T.? i , L4 ,;,,-, -, - -- Ig , LI- , - , I I
T I.- If, I, I j. I , I . I. , - , , , -, Y . ! i , 0 , "
I
;L , j I - , , , -. 1. . . ,
r
-,.r
--, -- . . , , I - . 1, , _ . . . , -,I " -
I 11 . ' I.., , -,A. 1 . ,r. 7
- , , ..
-, I z _I 1,, , r I , '1 , - , 4 . I rLj ei , , j, " L, I i , -- , , ':'
I,A ", kC " ;, c j " : , '_ r. .
r ;."--. ". .
- I - . -- '... I ' . - - i 'if, , , _ . 0*' j ' - - . -' ' ' ' r, ' -4I. ' - " ' - -- , - IjI ,
.. - 11 Tj L- -, ] ;,.,' , 'I , J'.? I I 1';, , je r .
-, - i, -- .- '-, . - , ", ' ,' . --, I li l - - , , L ,O i ,'. - - .".-.76" ,h - - -
I , i 1.,, A' ".' ,' j, I ' --I
f!" -
, 'I . r -- " IL - ) II; , % , , 11 -, . I, , ) , , , , I ',
J1 ,- . .-
-,
_-r L
1'r . -
-
,
r,"
i... , , ,-,
- - _,,,,- , ,:
:- t _' ,-,--",---. - -. , - , -- ___ ,
- I -- , -- - , , , - -,-- , 1, r-
1,,-,
.
-,"r .- - '
-
.- r , u
3t
t ; e
, -_,
.-
.
.
..-
.
1. , ,i
t -:,-!.14 ?I,
, Y I, tj , 157
,
II
L.-
"CAV
,
. V , I L
Y ,. L I
.
- --
, .,..i,;.,, _.';q
-,.
= ,
;'I -,
,
-"
- 'V
. ,I
. --
, I, -, , , "- . I , - -
" !- ,' ,
r
- .,
,
, '- - 4 L; I % I - -: ' ,r
, J. L,
I , ,'
L I 11"
.1 j ,- i, , , -, , 'i
. .
.
I I
:,

:
I
, ,
,.: ,, , , ..
i1, - -
_ . _ , : , __ -, ' r - , i , , , I, -,
.- ' _ _ A . _ I_ I I . . I I 4 , .
i ,, ,, ' . .
: . . . . ,, il lf J , ,' ., - -,
,,. , ,,j -4. , , IJ, I" --I , ,
.,
, I " I -, .
-I---
), - ., -
-- - - - ; " .I. , , ,
L' I ' , I - I . ill- o". , I, , , -'II I
,
,.-V.' r.
-- t - I
,.. I
11 - a ;.W-
. , --
lv"
II ,
. hI. -I, . - .s., , v IsK -i, [ , -1 .,
, , . , ,,_,,
l :. ...A "
.
il i , ., 11 , ;. . I ' I ' ( ! r,.C-
.1
:, . r. 'Il I11
-
,- - . . I, A . , " . ,I, fT - , I _ L'
-,
-
-. , , , - -.',.
.' r.- ..
, I, - , ,r , .
, - :. . .,-, , - F', ,, ,
r ', ' , ;'' _
4-' 0
.1
- .I
I 1.
F , I ] .. I
I - , ?-ll 1, 1eA I"
.
,
, " "' -
-- ,-,- - - -11-1II ,, - j- , . ' : , - ,; - ? 'A, Lie ,- ,
. -
, '-,I'
-.r - . . _, *! , L.,., . , Yf l., -
r,'L
3
- !- -
,, ':.
L I .
-
-" 1, ,
" .
.. I ,. - - I . .1 P I'II , -'j , ,%
.rj -,.,
_ ,'-- , ,""-
- I ' , I
-rj..
T Itm Ll ;
I.. f - -
- ,
. .",T--- , . .
I
- -- .
, , ,, . -, , 4-9 . . ".1
1, . --- I!,,,. .11
'N ' ,o
" IT,L, ' , - -
. r .1 ' . , .
' '. i:
..1 I.,-, I %r. -,, 4----
V =Iq ,.i
" j-' I-'
- -! I rkl II
1'-, ' 'L'IT; ' L I i -
- -lk-.- :---.
'L, - - , - . '- -
,
I -
III -- r, . . -I Z
I
- I . - , . , i ;I, z-,
,
"
I11
.-
:,_ _I ,' , -- ,
-I -
I -f
,I
1
-
, - - '11, "
- - -, .
' -' r 1 -,,-t *.,. 1,
, '.F.. ' - , m -t _-
.'L - j I
.
, -.

k, ,-,Z
I
" r
, ; -
:,,
I ,
,
.. .- *
t, qN - !,
.. . -
,, , .. , ,
- . "

, - .1;
-
,,:t' 1k
i , 3, ; I'l "I L-.p - c
- k
, -4
-L , 4?
- i, _. L,
* ' j , I,
t .
. Il : ol .- 4.1 1t
-
kk
. , .
,
" : ?;,
-r
I,- . ..
r i'. I
" - .
" I
.
' '
r.
.
, I
-rr
- "I
- .. ,
,,,
, , , IJ. , , . - , I - 11A
. .); - - . , .. . -,, : . .k ' L . . I -' -
, L -.- ,- I . -',, A 7$ , -. , , -- , , , " , ,I
" ---
.,. - . , ; ,, ','-' 171 ), ,'.C, L I -;lf , , i
,, , . .; I; , - - , . I V i I II , , II '- 115:*-,
' ., ; K." ' . ;4 , -, ;,, ,' r . , - " , , - ,
-I
- - , 1, - '-',, .. L, ,, ,, - --- rl, , - , , ,
'I-' . - q
,) - ,- ' I - r. -- - r , i*P , * , ,I',- i . .- 1,., "W ."i, , ..; -I - , ,.-f " - - '.'
% .,"fY
., ,- , , IJ 1.1-11 -
." . _
. . . I . .
.. - .. , , ,
4 .: r
j . , ? 'I-, .-. . i , I
1-I, -,I ,
t., i 17 - I. ' - ,,') , , , r - ,
. , izIr'.7" -f r " .- . , '. " I )' . . . I - 1
. ,
" r.llr
;I C , Irr e , .. M
, ,, , , , --,- , . ..-. 'IF .41" .1 4 ; . 11
-
,
- '- - ',',Tl M -r.t j L- -. .. '. .t .N I r , -
- . , :. , , ,, - j
r ., . %. ,
., , . - I, , . . 'L Ir, ' '. .
-.-- - e ., . T ,
, I.IT, , ,? .-,
% , 'i",, ,, , "I -1, - -,,2, ' ' .
t- '--4,
. - -,. IL . . ' %41 ',. __ _ 11 - '- - -, , ,- =-L
'r ' - . 11 . . . - I-,
-
:,i. ,,, . A - I j - ,'. ..-p ' ,", ; .'T ,,-r,-C ; . 1 - , ' j ,- ' , .., .' v, 1,4 , , 1, , , -S' '4 E:.,.-j" , .'T. , ., *' ' ll .- I j, , , I , b 1, r.,- .. . .
i , .. , - ? ,- i, ," I ". .-I,,, rs, 1
. - b't p 4 , L ,r -) ', -%,-
- I , ) , I,_ ",t-,,' r - _ ;: r-' . . ' ) "' ' .. 'i " 'T'il . I- ,
Ir r l
- , . y I . ,-- IL, , ,) ,, %. , . - , , , - j, , -, .
- - . -, , -, _ ' 1, ., r I :
. .. -L. - t -. . ..- . - r, ,r , & - ,,, ,I . . ,. , .1 -z- 5 . ,,
. , I,), . - 1,"..
- - - -, ,-- I1., - - I--I,- 1:1. I , I. II - -, - . L 114;,
I -z
.-- . . ,,.,, r ,, ..,, -, , -, I
, I , , , [I I ,
, lf lk, " , I -- I
-
II
. , , , , ; l'i " , - .. 'L' , , ,, ,,I .r
ll , , 'f n i, -
. - I '. L , .1 , 'r ; 'I I
- . , " -- - .-- ? .1
I '. .i .- , .1
,. ,*u "
, r , r, ., -1. , . I . I I
, -- .; 1, I , . -
'" : ' ' .-- ' - r-. . - ,, ,
, , , ). ,-,
" "
.. , . I I-
--
" " .I
'. ,' .
, ,., ,
I , _(" , i 4 I_- . ;
.", _ _;, ,.
r - . V LI'l
-, , .. .. - I I . ) 74 F ,,, : - ,, ,. -
- ' ' '."", , - i L-' '._- , , I- ', ' ,'
, ,_,, I' j , r I
L' ',- L . 1
., - -
,4.-,-. , , I 1 ", : , -_f , _ , . , " - r ,, , __ " , ;;, - .11 ------ : ' r ' . . 1, - - , . I
. , , ,, . 1-I ; _. - , r , e - 'L:'- . 4 , . - j , -$C ri ZI, l I . J ' Wj, _1k._ -4, 4
,; I '), ,'j 'I.;
-.i - f.. 1,, L, .. r
L'' r :.'
,, : ,' 2, i ; . ,. ) , ". -!, '.,, , ' I_ -1 r .
.-- 11 , '.
L -. , -
, r dI ", .. , , -7 7I" %-lm , " -- - " ,",
.' 'r
7tZ - , , , , - I4, . .." , .,L
, ., -j .- _ ,. , ' -,'.LL ,-.
J ! n- Ill"- 6, , ,.,
, ,
.
- , . ,, . I - 0 "P- f ., e ,,
""',-7-)
", , , ; I - -, -, 2 , ;.F,.., -V
., " - ,.. II 11. 1 , I , -1 .1
. ; .. , ". ' " ;- ,.
_- L.'.'T'
I - - - ,y % .I . . 'I -, - , - . -L I"I 11
.1 .
. I
i' -r I '- ,rL . . - '10. , ,r, j. ,, -, ,- , -.-,'k if , .-- I - a I- , - . -k
, A, - ,;, . . _ ,, ,,r.,,'I. '_!r ,- J, -0'
, -0- .k41
17"', .- - IkT' .-, , -,
- -, I - .. I- , .. .I 11I ,
.i, 'i : - ,plut, - -' j r 1,
,
: ; ,.1,, - " ,-1 , ,,.IL,, ,_
! . rI , - . , I I, . . ",,, . IL . -, ,IL 4 t
II,, ,, . I .. A; 1-i
'L - ",, I
. : -'
I ,- ! ,!,, , j, --- .,It %,",- ,,.. 7 i, , -,, .--9I" i V ... , I I , .... - f " , Ir -
I ' 1, , r : :-I ;, , 1,
". -, - -,.2 ,,, ; --I ? - .I, , ,-"
I 1 . :, ,, V. A lyll- ' ', -
1.114- -I. --f , , - II-1 11 . - , , -, , ,,,
- l - . . I
I I "; '-
1, ; " , ,I 11,. " y".0'. I I " ,,, 4 . ---' 1-- - ---. '4 '! e .. -- i 7,_3 ' P'J, .
I". . , . ,, 11 7f-. '11'11, 'r' - . I4
-

.11 I I. : . - -I'I - * , ",)', -':?.I,- .,, 5'r , L , -

7, - - I- -IL L:1'
Ll I.-
-
. . 'L I 1'I'nI ., , "
_. L. - , - ,
- I I ,
I, _ q "
. ,, . .,oI 111,*- ,-, I- I
'! "-, " " . '. - - -
) , " ' , , %" .1 r
- , -. , . 'Pl 1 A _,r 'J,, I i. ji ,I, . - -IC, :,I'5L , ,,
t-
A ., -jr , ,F,-
X - !, ,,r, " -,I Y :
I - , , -, : ; , - I 1 - !: - , 4r. , ; , ,
.:,,- , .
-r j I.,
1 7 L
. '
I ,,
. , . I II r - ir . I.
,, I I. , .
- - "I ,., . ,- ;l .? wi- :: . ;- , . . le, - i 0 1',,,- - . -", " , , iy,Y.,. -
,, J"17 k- ' 'j i '-!'- - III -. I' ' T"
r, -, , .
--,-I-I'. - I , j,, .,1,
.1 - L:1,
T, - ., ,- ..", I-. -,, ,- ' ,,' , jr, I -4 - _ ,L. 1.I, , L.
Ii 11
.. 7 o .1 - , , ,"I r-, _p,- ,_ ., , -j.. , ,, ,,, , - - c , - , , -
,
I. , I .-: . _,11,'-,."j "," . " _r ' - ., I . . , II.. 'r, ? ,I. , IL7, "' , ,L ,-- .",_,._,- 1-- i L' r _- - , , , , LI --_' ,!
, "' -, ,, - ,! , - - . I_-r,
I - '.1 -- Y - r.-- - ,;, I . .. . , " , r '
. .. I"-."--Y
, -- _ . - -
" . "" 11 i 1 f- %'t , -.'."
II -, . , -- I , f,
-
, J.j
,,
L.,
-f - ,' '- --.-- ' -, 1, , - -. - L-!1 "' - ,
1% ,4, kII
, , i, ' ' -- -
4. ' L ,'L ,.rr .' _,
II11 r I - , j- . - - , ' - ,,, I ,, ,
r . ,
, ."," ;.) . f I'l ' , .. r' -' , 't . , : .- .- , 1 , .- , -1 , r.
- , - , r , .- "6;' -L' . .! I ' !." ,
, , l', , 1. 1_1-1
v , 4 . , , ,. 't, - If 11 ' - L' ,;' - ' .64 - , I' f I-'- , I' ,!, ,, , ,_1'
%. ft - - (-.Ir
I . -
-, i, -P, , ,. " i -7
r.--
' " 1" -' 11 "' , _'._I , ,!._, " . ' ,r ,, . r
, "',
, ;,
; - -, . L , ', j , i - . j', -
? I - r' r . - '' , Ir rl',,1,
-' . - ' r l I
-. I
I : J ,
?,
Z, w, i, , Z )T 1- , .,*
i
L __ L , - :1 - iT,
,
_L II,
;., j .IX
I, ,I -1, -, , .. I r _ -, , J - i
. .1, -
_ ;. - 1,
, 1. _ L"L -. , r . , Ir-I ), " , ' -- -1, ] --
I
, , -, %,
'I , . ! I,, ... -ia
I1- , _, , .,"', " =--' - ", - " I!" r 1 , Z',' - _ 11 -' ' "%4'!
I r,
II I
.
,, I
1, . .-
. fl ",A I J , - " T :. ' - .I _ .
, -- -- . .y W T ;. v -r I .,
' -I 'I 'l ':-'"
- . r' . 11
I--- - - ,-j , i'. ,
. f
11 I- I
7 A.- - , I , ; -1 , . , -) --. j , r I I '-' ."- ' L
r7 - ,,f,
,I . .. "
.
--, _-r . -',-' -_ I . . .,, ,'

. 1L
-- , ,- -.. nL , L I.
, -- - , , ".. I - ., . r-,, 1,

-'' .!LI-- -- .' I .L:' 1. ".1

,, , 1,, , '. , , " - v

,il,..4'";
II , e. ,- i,i::'.,-
I
-, ,- I - - . ,
I I
.
-'r-
I - C,
l I , - , - - - ":
-

'', --y , V* 16
- ';
- - ., - - - 7 -, 1,- ,- -
,;,
, ,,, .
r-_'--- ."
__ - -' '. '
.
"" ' _
I. I L
I

,
, ,.'r _ j r.1 , - , ' I - t -
-.
. - :- - - . ,' ---
r, , I - I, ' , t L_, ".
_ ,, - -- --
. I

" .. _ 7
. , -
-, - -II - v
-
.1
r. , - , " 5-
- , .

." .1 . I I J .jr , 1 '4 '.- _6,:",_ . I. - . - 4- ,j L

 DAVID TAYLOR MODEL BASIN
WASHINGTON, D.C. 20007

A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH

PART I
CONVENTIONAL HULL CONFIGURATIONS
CHAPTER 5
STRESSES IN PRESSURIZED SHELLS OF REVOLUTION

by

Thomas E. Reynolds

Distribution of this report is unlimited.

December 1966 Report 2264

S.F013 03 02
Task 1961
 . 1 , lililYlu i l hIl a hi WniikII61 nNAA1
Iili
hilluIki
g n li

FOREWORD

The results of submarine structural research are now so scattered among

numerous reports and papers that it is difficult for investigators to locate and
evaluate information. Steps have recently been taken to consolidate this
material in the form of a research summary which should be of some value
particularly to those concerned with pressure hull designs. The summary is
intended to be a source of information rather than a textbook and is to in-
clude an explanation of the development and use of important theory as well
as a critical examination of its merit. As presently planned, the first part
of the summary, which deals with conventional hull configurations, will
consist of nine chapters, as follows:

Chapter 1 - Axisymmetric Behavior of Ring-Stiffened Cylindrical

Shells
2 - Asymmetric Behavior of Ring-Stiffened Cylindrical
Shells
3 - Fatigue of Submarine Hull Structures
4-- Interpretation and Significance of Submarine Model
and Full-Scale Experimental Studies
5 - Stresses in Pressurized Shells of Revolution
6 - Pressure Hull Materials
7 - Frame Strength of Ring-Stiffened Cylindrical Shells
8 - Hemispherical Closures
9 - Hull Penetrations
Preliminary versions of Chapters 1 through 4 have been issued as
David Taylor Model Basin Reports C-1569-1-C-1569-4; the remaining
chapters are to be issued in unclassified form.

~ I I II IIIlllI __l-.__l~-arcP
-n~ T*cdC~'
~a-~lk
 TABLE OF CONTENTS
Page

ABSTRACT ........................................................................ ...... . ................ 1

ADMINISTRATIVE INFORMATION ................... ...... ......................... 1

1
INT R O D U C T IO N ............................................................................

THE METHOD OF EDGE COEFFICIENTS..................... . . ................ 2

CYLINDRICAL SHELLS ......................................................................

4
SHORT SHELLS ................................................................... ................... 4
7
L O N G SH E L L S ...........................................................................
SERIES REPRESENTATIONS........................................... 8
EVALUATION .....................................................................
......................... 12

CONICAL SHELLS .................................................................. ....... . ............... 13

13
EXACT SOLUTION (TAYLOR-WENK FORM) ................................................
..................
18
APPROXIMATE SOLUTION (RAETZ-PULOS) ..................................................
................
18
L ong Sh ells ............................................................................
. ................................... 22
Short Sh ells ............................................................................
24
ANOTHER APPROXIMATE FORM (AFTER HETENYI) ................................
................
E V A L U A T IO N ............................................................................
32

.....................................................................................
SP H E R IC A L SHE L L S ..................... 33
34
E X A C T SO L U T IO N S .......................................................................
APPROXIMATE SOLUTIONS .................................................. 35
Geckeler's Solution for the Polar Region ................... ............... 35
Hetenyi's Solution (Polar Region Excluded)....................................... 40
Leckie's Solution (All Regions) ................................................ 45
47
E V A L U A T IO N ............................................................................

OTHER TYPES OF SHELLS............................................. 48

49
.......................................................
STIFFENING RINGS................................

COMPOSITE SHELLS.................................................................................... 51

53
A C K N O W L E D G M E N T S .......................................................................

REFERENCES 64
................................................ ........................................
 W11m10014
119hNII

LIST OF FIGURES
Page
Figure 5- 1 - Composite of Typical Shells Used in Submarine Hulls................................. 54

Figure 5- 2 - Loads and Deformations in Intersecting Shells ........................................... 54

Figure 5- 3 - Cylindrical Shell with Edge Loads ............................................ 54

Figure 5- 4 - Conical Shell with Edge Loads ............................................... 54

Figure 5- 5 - Taylor-Wenk Functions for Conical Shells ............................... .................... 55

Figure 5- 6 - Raetz-Pulos Functions ................... .................................... 57

Figure 5- 7 - Comparison of Raetz-Pulos Solution with Strain Data

for Cone-Cylinder Model ........ ........... ......... .. ............................58

Figure 5- 8 - Spherical Shell Element .................... ....... ......................60

Figure 5- 9 - Spherical Shells Bounded by Small Polar Angles ......................................... 60

Figure 5-10 - Spherical Shells Bounded by Large Polar Angles ......................................... 60

Figure 5-11 - Experimental Strains for Pressurized Clamped Spherical

Cap Compared with Hetenyi's Solution ......................................................... 61

Figure 5-12 - Stiffening Ring with Edge Loads ................................................................... 63

Figure 5-13 - Cross Section of Cone-Cylinder Juncture with Tapered Ring ..................... 63

Figure 5-14 - nth Juncture in a Chain of Connected Shells ........................... .................... 63

Table 5-1 - Edge Coefficients for Spherical Shells Compared ........................................... 47

.11 3 1 I I I I ~ ~.~~
rPa~-r~n~src~
ps~l--r~' *
-, " h, ,,, 1W111, .111 1W
ill, ,, 1,1,11,611III1II1 , ,J1,
WJ1,111ANUM119

NOTATION
Ar Cross-sectional area of ring

...A Arbitrary coefficients

4

...C Arbitrary coefficients

4

2
1 1 ot A (spherical shell parameter)
(1 2v cot OB (spherical shell parameter)

1cot 2p 0, (spherical shell parameter)

Vr(1 -4v)
1+ (conical shell parameter)
eA

1 - (1 4 (conical shell parameter)

E Young's modulus

FA, FB, GA, GB Conical shell parameters

fl ... f4 Special functions for cylindrical shells

H Edge force (perpendicular to axis of revolution)

per unit length

Shell thickness

IZ
A Moment of inertia of ring

J1' J2' KI1 K2 Special functions for spherical shells

K 2 R 2 0/EhL

KA , KB 2 R 2f3/Eh cos a

k 2 R 2 0/EhL (sinh 2 0 - sin 2 0)

L Shell length (also differential operator)

1 Horizontal dimension in ring

M Bending moment per unit length

m C , m d, nC, nd , tc, t Special functions for spherical shells

d
 I11,III
l I111 I I1
li1IlA

NOTATION (Continued)

N Normal stress resultant

p Pressure (positive external)

Q Shear force (normal to shell) per unit length

R Radius

Wc , X C, X d X
, e

XB - X'
isy

,C
Angle between axis of cone and generator

3a(1 - V2)/ cos a/Rh

e- 1x sin p3
3
77 e- x cos Px

A V3 (1 - v2)h tan

Special functions for cylindrical shells

Poisson's ratio

2/xh tana 4/12 (1 - v 2 )

p (3R 2 /h 2 ) (1 _ V2)

or Normal stress

05 Angular coordinate in meridional direction

I IIlrFlo
 NOTATION (Continued)

x Angular rotation in meridional direction

Schleicher functions

. ,..
B Functions for conical shells

A, B Subscripts denoting left- and right-hand edges

of shell

0 Subscript denoting circumferential direction

11
 ABSTRACT
Analytical methods utilizing the linear bending theory of thin shells are
presented which are judged the most useful for determining stresses in subma-
rine pressure hulls. Only axisymmetric deformations are considered, and
emphasis is placed on methods that are sufficiently simple to be applied at
the practical level. Attention is restricted to analytical solutions for
cylindrical, conical, and spherical shells. Efforts have been made to employ
a uniform system of notation and sign conventions wherever possible.

ADMINISTRATIVE INFORMATION
This project is being sponsored by the Naval Ships Systems Command, Code 6442,
under Subproject No. S-F013 03 02, Task 1961.

INTRODUCTION
The industrial uses of shell structures are so extensive today that it is not surprising
to find such industries as petroleum, aerospace, and atomic power represented in the tech-
nical literature on shells. Nor is it unexpected that with such widespread interest, the vol-
ume of this literature should have grown to staggering proportions. One consequence is that
an effective literature search is becoming increasingly important and correspondingly more
difficult. The special needs of individual research or development groups often narrow their
fields of search considerably, but some problems are so frequently encountered and of such
importance that they have universal interest. For instance, during flights a rocket motor case
must withstand certain dynamic loads that would never be experienced by a submarine hull.
Yet both structures are also subjected to static, rotationally symmetric loads, and to the
extent that the configurations of both have basic similarities, the presence of these loads
creates structural problems common to rockets and submarines.
The present chapter is concerned with this common area of interest, i.e., the axially
symmetric deformations and stresses in those shells of revolution frequently used in subma-
rine pressure hulls. Because so much has been written in this general field, the contributions
of submarine structural research can be counted as much by what has been accomplished in
the selection and adaptation of available theory as by what has been done in the way of
original work. It would seem that one of the main purposes of this research should be to
develop methods that can find practical application in improving the structural efficiency of
submarine hulls. A summary of the significant results of such research would therefore em-
phasize the practical over the precise where a choice must be made. The material has thus
been chosen (it is hoped) to reflect the needs of the practicing engineer who must frequently
"make do" with formulas and methods that are somewhat inexact but sufficiently simple
to
permit application at the practical level. It is not expected, therefore, that this chapter will
 III IY IYYIII
11011II191IW1 ~

have much to interest the mathematician or the research worker engaged in developing new
shell theory; in any case, they will have been over most of the ground before.
The main portion of a submarine hull is the easiest to deal with analytically since it is
usually nearly cylindrical in form. Noncylindrical shells such as cones, spheres, and toroids
present problems of considerably greater difficulty. Figure 5-1 shows a composite of some
typical shell elements commonly used in submarine hulls. Spherical or spheroidal domes are
often used as end closures; conical or toroidal sections, or combinations thereof, frequently
serve as the connection between two cylinders of different diameters. Ring stiffeners are
employed both to increase stability and to reduce stresses near junctures of dissimilar shells.
The methods described in this chapter are those considered most useful for analyzing these
structures. All are based on linear bending theory, and their application is restricted to thin
shells having no imperfections. As general reference texts, the author recommends
s ' * 5-2
Timoshenko's "Plates and Shells" - and Flugge's "Stresses in Shells" where most
of the analytical work and much more can be found.
It should be emphasized that only analytical solutions to shell problems are consid-
ered here. This puts definite restrictions on the types of shells to be covered since analyti-
cal solutions to the differential equations of equilibrium are known for only a few simple
geometric shapes. Accurate solutions for shells of arbitrary shape can be accomplished only
through numerical methods. Today this can be done with comparative ease through the use of
high-speed computers, and efforts along these lines have become widespread. The trouble is
that such computer programs usually are written for the special needs of individual groups
and as yet are not of such scope and flexibility that they are readily usable by the general
public. Krause 5 -3 gives an interesting account of the pitfalls that can be encountered when
one group tries to use a program developed by another. His paper also discusses some of the
numerical techniques commonly used in this kind of programming and includes an extensive
bibliography. Perhaps the day is not far off when shell analysis will be performed exclu-
sively by computers, but until it arrives, some engineers will have to go on using the
methods described in this chapter.

THE METHOD OF EDGE COEFFICIENTS

When loaded only by hydrostatic pressure, a shell of revolution responds in what might
be called its natural pattern of deformation, a pattern peculiar to the type of shell in question.
Two dissimilar shells will in general therefore deform naturally in different amounts under the
same pressure load. Should the two shells be joined, they must deform equally at the circle of
contact and it is clear that under such conditions, the deformations of each shell will be dis-
turbed from their "natural" form. The result is a deformation pattern that varies in the axial

*References are listed on page 64.

n I I I ii I I
 _ _ __~~_ ~-

direction (but not circumferentially so long as the loading is axisymmetric) accompanied by

bending as well as membrane stresses. These bending stresses (usually critical at or near
junctures) are of considerable interest to designers, and the problem of their determination
is often not an easy one.
When two dissimilar shells are joined, one can find the stresses developed in each
under hydrostatic pressure by enforcing conditions of compatible deformations and equilibrium
of forces and moments at the juncture. This process is often lengthy and tedious and, so far,
few short cuts have been devised, although the principles involved are not complicated. The
method of edge or influence coefficients has had wide acceptance in the solutions of such
problems, and it is the method to be employed in the development that follows.
Figure 5-2 shows two shells of revolution having arbitrary meridional shapes and sub-
jected to normal pressure. The edge forces and moments represent the interactions of the
shells per unit of circumference at the juncture. It is usually the practice to resolve the
forces into components parallel and perpendicular to the axis of revolution as indicated in
the figure. When normal pressure is the only external load, the axial force is simply that
R
due to the axial component of the pressure, or p - , R being the local radius from the axis
2
of revolution to the midthickness of the shell. For Shell 1, the meridional stress ox l and
the circumferential stress a0, are functions of x, a meridional coordinate and the edge
loads:
xi = OxX (X, H1A , MlA, H1B , M1B, P)
[5-1]
01 = a1 (x, H1A M1A , HlB , MlB , P)

Similar relationships will, of course, hold for Shell 2. To determine the edge forces and
moments, one must introduce the deformations of the shell juncture, and this is usually
done in terms of the deflection w * normal to the axis of revolution and the rotation X of a
shell element in the meridional plane. These quantities can be expressed as functions of
the edge forces and moments by means of edge coefficients:

- WI B dfi B 1l A d 1B WlB
B1 - H1+B ---- M + - HIA + M MlA+
1HB M 1MB 9
1B H1A A M1A A p

[5-21
dX1B 6 0
X1B dX1B X1B IX1B
= - HlB + M M1B + H1A + MlA P
B a e HlBM
or1A wp B Me1 A

*The deflection normal to the shell will be designated by w.

 11111

where WilB and XlB denote the deflection and rotation of Shell 1 at Point B.* Similar relation-
ships hold for Point A and for Shell 2. Conditions of compatibility at Point B require that

lB = W2A

[5-3]
XlB X2A

and for equilibrium at Point B, we must have

HIB = H2A

[5-41
MB = M 2 A

The unknown forces and moments can be determined from these equations and then be sub-
stituted into Equation [5-1]. It is frequently found in shell problems that the deformations
at one edge (e.g., Point B) are insensitive to the forces and moments applied at a remote
IB
edge (Point A). In the present example, this would mean that the coefficients 8HI
aWlB dXlB (XlB 1A
, could be neglected and the analysis would be greatly simplified.
8M1A
-
'H1A 1A

Whether or not such simplifications can be made, the crux of the problem is to find the
edge coefficients appearing in Equations [5-2]. Usually the differential equations for the
shell in question must be solved, and this is where many of the difficulties are encountered.
The value of any analysis is therefore in providing the necessary edge coefficients together
with the corresponding equations ([5-1]) relating stresses and edge loads. Once these rela-
tionships have been established, the remainder of the analysis, although often tedious, is
usually straightforward.

CYLINDRICAL SHELLS

SHORT SHELLS

The equations for a cylindrical shell have already been presented in Chapter 1.
However, in that case they were not arranged in such a way that the edge coefficients could
be easily discerned. If the second-order "beam-column" effect is neglected, the differential
equation for a cylindrical shell under hydrostatic pressure is

*The inclusion of - and - as separate coefficients is somewhat artificial since all quantities are linearly
ap dp
dependent on p, when that is the only external load. What they represent are the membrane deformations (i.e.,
those that would occur in the absence of structural discontinuities) and those that arise from an axial force due
to pressure acting against the ends of the structure.

-~II I I I I I I I r ~I
 M 110
1111IW116 alI
SIIIImYim I

2 )
d4 12 (1 12 (1 - (1 - v2)
- + w= p [5-5]
dx 4 7 2 h2 Eh 3

where w is positive inward and p is the pressure taken positive when external. The general
solution to this equation is

Ox Oz Ox Ox
w= cos - cosh - +C 22 sin - sinh -
1 L L L L
[5-6]
Ox Ox Ox Ox pR 2
+ C 3 cos - sinh - +C 4 sin - cosh - + (1- v/2) -
L L L L Eh

(Reference 5-1, pp. 399 ff.) where the term multiplied by p is a particular solution of Equation
[5-5]. The C's are unknown coefficients whose values are determined by the boundary con-
is/3 (1 - v 2) L
ditions and 0 is as defined in Chapter 1. It should not be confused with the
circumferential coordinate having the same designation. If the shell is considered to be acted
on by the edge forces and moments with directions as shown in Figure 5-3 (Edges A and B are
separated by the distance L), then the coefficients are as follows:

2 R2 0 OMA
C 1= HA (sinh Ocosh O-sin Ocos 0) + - (cosh 2 - cos 2 0)
2 2 LA L
EhL (sinh 0-sin )
20MB
+HB (sin Ocosh O-cos Osinh C) - - L sin Osinh O]

2
2R 0 2 MA
EhL 2

[5-7]
2R2 0 OMA
C3 = E-H A sinh 2 0
- (coshOsinhO+cos Osin )
EhL (sinh 2 C-sin 2 0) L
OMB
-HB sin Osinh 0+ L (cos Osinh 0+sin Ocosh 0)

2R2 0 2
OMA
C4 = [-HA sin 0- - (cosh Osinh 0 + cos Osin 0)
EhL (sinh 2 0-sin 2 ) L

OMB
-HB sin Osinh 0+ - (cos Osinh 0+sin Ocosh 0)
L f
 The stresses in the shell at any location can be written in terms of these coefficients:

pR Ox Ox
L cosh
L
3
Ga,
ox Ox
- sinh L
L
[5-8]
4)cos
ox
+C + -v C3 V2
¥ _C
C' 3 )sin
C Cos
sin
-
L
sinh
v
2 TvCi- C 1
C3)
(C4
2 ox
- 1,
_/1
- cosh
L

pR + E 3 Ox Ox Ox Ox
x 2h I 22 cos -L cosh -L - C sin -L sinh -L

Ox Ox Ox Ox
+ C 4 cos
4 -
L
sinh - - C sin - cosh L
L L

In these equations, the upper and lower signs refer to the outer and inner surfaces, respectively.
Here and throughout this chapter, tensile stresses are considered positive.
9w
Finally, the displacements w and rotations of the edges X = - can be obtained using
coefficients:*
following edge axthe
the following edge coefficients:*

Ow A 8
- k (sinh 0 cosh 0 - sin 0 cos 0) XA kO
=- -CL (cosh 2 - cos 2 )
8
HA HA
[5-91
WA 8 2
kO XA 2kO
- - (cosh 2 -cos 2
G) -- (cosh 0 sinh 0 + cos 0 sin 0)
9M A L MA L2

W 8
A XA
*It will be noted that there are a number of identities among these equations. The relations
aMA 9HA
0
6A WB A @B B B )A O'B XA XB
- =-- , - = - -=-- , -= -,- - follow directly from the
dHB dHA oMB dHA HB dMB dHB MA dMB MA
5-4 4WA )wB (A ()WB (A dwB XA B
theorem of reciprocity. The others, - , =
,- - , - = - ,
kHA dHB dMA OMB dMB OMA HA OHB
8
NA A XB &A GB
=_- = _ _ =- , result from the geometrical symmetry of the problem.
dMA dMB 9HB 3HA 3p dp

- I I I L ~R ;
---------
 WIN il 11WHIMUMN
IIIW1IYII

3WA dXA
- 2kO
2 sin 0 sinh 0
k (sin 0 cosh 0 - cos 0 sinh 0) MB
aHB L

aWA 9XA 2
2kO- 2kO
-- sin 0 sinh 0 - - (cos 0 sinh 6 + sin 0 cosh 6)
dMB L dMB L
2

awA R 2
OXA

Op 2 Eh dp

dWB dXB kO
- k (sinh 6 cosh 6 - sin 6 cos 6) - (cosh 2 6 - cos 2 6)
OM B
HB L
[5- 9]
OWB
k6 dXB 2kO2
O (cosh 2 0 cos 2 - - (cosh 0 sinh 0 + cos 0 sin 6)
dMB
6) 2
OMB L
-k (sin cosh - cos sinh )
dWB 2kO
-k (sin 0 cosh O - cos 0 sinh O) S- - sin 6 sinh 6
HA OHJA L
dXB

MA
dWB 2
2k0
-2kO sin Osinh 0 - - (cosh 0 sinh 0 + sin 0 cosh 0)
dMA dXB
MA L
2

dwB R2 (9 -0
dp (1- ) (9p

where

2R 2 0
k=
EhL (sinh 2 0- sin 2 0)

LONG SHELLS
Equations [5-9] are exact in the sense that they include the effects of the loads at one
edge on the deformations at the other. The magnitudes of these effects are directly depend-
ent on the shell parameter 0. It is necessary to use the equations in their exact form only
when considering short shells (i.e., 0 - 3). The effects are small, for long shells, and one
can make the approximations:

dHBw wAB MB MAB XA XB MB dXB

o--H -7HAB
A A
A B A +.B dMA
-0

---
-----
sl-~I~~
 Furthermore, in such cases, sinh 0 cosh 0 >> sin 0, cos 0 so that the remaining equations
can be greatly simplified:

dWA &XA 2
KO
alA 9H A L

dWA 8
KO XA K 2
= - 2 - 0
OM L dMA L
2
A
[5-10]
daWB dXB KO 2
OHBp -K
B aHB L

dWB KO dXB K
2 - 02
dMB L dMB L
2

where

2R 2 0
K=
EhL

dw A woB
The coefficients - and -are, of course, unaltered. At the edges of a long shell, the
3p dp
stresses given by Equations [5-8] can be closely approximated by

pR 2RO (MAO 6 vMA

h hL L 2
h
6
MA
pR + at Edge A (x = 0)
x 2h

MB O 6 vMB [5-11]
pR 2RO ( H
h
A hL B h2

pR 6MB
x 2h at Edge B (x = L)
h2

SERIES REPRESENTATIONS
It is unfortunate that no such simplifications can be made for short shells. The exact
equations are complicated not only because of their length but because they require frequent
reference to tables for values of the circular and hyperbolic functions-a process that is often

L I I I I I II -i---~--~----
tedious and inconvenient. Raetz has sought to overcome this difficulty by making use of
Maclaurin expansions for these functions.* Such expansions are best applied in the short
shell range where rapid convergence is assured. A few of the expansions that have been
found useful are as follows:

sinh 2 8 + sin 2 3 (26)4 (20)8 (26)12

2
sinh 0 - sin 0
2
02 6! 10! 14! I
cosh 0 sinh 0 + cos 0 sin 0 3 (9)8 (26)12
A2 + (20)4
sinh 2 0 - sin 2 0 03,7 5! 9! 13! J

cosh 0 sinh 0 - cos 0 sin 0 2 (20) 4 (20)8 (20)12

A3 = 1 + 3! + 3! + 3!
sinh 2
0 - sin 2
0 7! 11! 15! 3
[5-12]
4
2 sinh 0 sin 0 3 [ 1 (20) 13 (2)8 1" 5 (20)12
-- + + ...
sinh 2
0 - sin 2 0 027 2 6! 2) 10! 14!

cosh 6 sin 6 - sinh 0 cos 60 1 2 (26Y4 4 (20)8

As
2
= --1-1 3!- + - 3!
2 0 L \2/ 7! 11!
sinh 0 - sin

cosh 0 sin 0 !+sinh 6 cos 60 3 1 \2 (20)4 )4

A6 =
sinh 2
6 - sin 2 0 30 [1- _(2/ 2
5! 21
(29)8
9!

(20)8 (20)12
(26)4 +4! +- 4!
8! 12! 16!

An alternative form is

4 881
1+ - + - +
cosh 0+cos 0 1 4! 8!
S sinh 0 + sin 0 0
0 o8+
1 + -
+- +...
5
S 9!
[5-13]

1+204 +2 11 + -
cosh 0 - cos 0 3 6! 10!
f2 sinh 0 - sin 0 0 04 0
8

1 +6 - +6 - + **
L 7! 11!

*These appear in a paper as yet unpublished.

iAIIJIM110 1 M
1 h

- 04 0 o8
08 --
1+2 - +2 - +
cosh 0-cos 0 0 6! 10!
3= sinh 0 + sin0 2
L 1+
04
-
5!
o8
+ - +
9!

- [5-13]
04 o8

-
1+ -*+ - +"
cosh 0+ cos 0 6 4! 8!
3
4 sinh 0- sin 0 4
08
1+6 -+6 '
-+*
7! 11!

where the functions appearing in Equations [5-12] can be found from the relations

1 = fl f2 4 = f2 3 - f)

/3 + /4 2 - fl
A2 5 [5-141
2 2

fl + f4 - f3
2 6
2

The functions fl through f4 can also be expressed in another series form:

180
m = 1, 3, 5, ...
/l r12
M --- 4 04

2
0
00
m = 1, 2, 3, ...
m 21
=

[5-151
10
+ -
10 m = 1, 2, 3, ...
2 7T4 m =1
Mrr + 4+ 4

47v4 _

- 1603 Z _
0

404
m = 1, 3, 5, ...
M ---
774

II IIII 1111111~1~
1111
 For all of the series expressions listed, convergence is more rapid as 0 is reduced.
The choice of which form is most convenient will usually depend on the nature of the com-
putations to be performed. None appears to have any distinct advantages over the others.
Equations [5-71] and [5-9] can be rewritten in terms of the new functions as follows:

0 0
C 1 =K A3HA + -L MA +5HB - L0- 4 MB

OMA
C2 = K -

[5-7a]
(A 1 + 1) A4 0
C3 = K 2 HA- 2 MA - B + 6MB]

K[(1 -Al) o0 A4 0
C4 = K [ HA 2A 2 H B + -- \6 MB
2A -L

dWA dXA 0
= KA 3
dH A dH A

(9wA
0 dXA
-K -
dMA
L dMA

dWA dXA
= KA 5 -K -0 A
(9H
B L
dHB

dWA dXA
- K 0- A4
S -2 K 2 80 2 2 A) [5-9a]
dMB SR 2 dMB
=-K-
L 4
dwA dXA
dp dp

wB OXB 0
= - KA =-K T )A
dHB 3
aHB

dWB 0 OXB
=K -
L
Al
M11
B MB
 1 I.".

all A -KX O9H =-K -X

L A4
A

2
(9WB 0 dXB )0
- -K - 4 2K X6 [5-9a]
dMA L 4 MA L

dwB v R2 dXB
= 1- - =0
Op 2 Eh Op

EVALUATION
In discussing these equations, one inevitably must turn to the question of reliability.
The best means of evaluation is direct comparison with experimental measurement. The
difficulty with making such a comparison is that there are so much experimental data from
which to choose. One could refer to Chapter 1 of this summary as a place to start looking
for test reports, but there is no need. The equations presented here (if properly transcribed)
are correct to the extent that linear thin shell theory is correct, and submarine hull geometries
are well within the limits of this theory. Agreement with experiment will be found to be as
good or as bad as the specimens tested and the quality of the measurements. Using reliable
strain gages, properly installed on accurately machined shells, one is apt to find only very
small discrepancies between theory and experiment. In such cases, the magnitude of the
discrepancies will depend pretty much on how accurately the gages can be positioned, the
length of the gage over which the measured strain is necessarily averaged, and the magni-
tude of the strain gradient in the region of interest. Positioning errors can be practically
eliminated through photoelasticity studies, but with a considerable sacrifice in sensitivity.
What this comes down to, then, is that the size of the shell relative to the size of the
measuring elements is all-important. In practice, of course, the real hull is extremely large
but full of imperfections unavoidably introduced during construction. Because of them, the
deformations under pressure are to some extent rotationally nonsymmetric-an effect not
covered in the present chapter. It is possible to account analytically for this kind of be-
havior in an approximate way (see, for example, Chapters 2 and 7), but to do it with pre-
cision would require exhaustive measurements and analytical efforts far beyond the scope
of this chapter. Measurements have shown, however, that averages of measured deforma-
tions in such cases conform closely to calculated values for the perfect (symmetric) case.

_ I I I I I ~ ---- -~----rrbbrt~
 CONICAL SHELLS
EXACT SOLUTION (TAYLOR-WENK FORM)
The differential equation of equilibrium for a conical shell of constant thickness under
5s
hydrostatic pressure can be stated in the forms-

3 2
2 d 4-w + 2x -d w - 2 -d w + 12 (1 v2) W 9 (1 v 2 ) p 2
= [5-16]
dx 4 dX3 dX2 h2 tan 2 a Eh 3

Referring to Figure 5-4, x is the distance from the apex measured along a generator,
a is one half the apex angle, and w is the deflection normal to the shell. The "beam-column"
effect which has been discussed in connection with cylinders is not included here.
This equation has been investigated by many authors. 5-6
Dubois appears to have been
the first to obtain a solution to the homogenous form of the equation (pressure term omitted).
He expressed his result in terms of special infinite series and asymptotic formulas. Much of
the subsequent effort on the problem has been devoted to reducing this highly intricate solu-
tion to a more simple, usable form and to its application for specific types of loading.
Flgge (Reference 5-2, pp. 371 ff.), for example, solved the equation directly in terms of
Bessel functions of second order which he expressed as functions tabulated by Schleicher. 5 -7
A solution equivalent to this was obtained by Watts and Burrows s-8 which they then applied
to the stress analysis of the juncture of a cylinder and a truncated cone. There have been
many other papers on the subject, mention of which will have to be omitted because of space
limitations. Attention instead will be confined to two or three papers which are particularly
useful for analyzing submarine structures.
The Watts-Burrows solution to Equation [5-16]* is given as follows:

dw 2 x/3(1- v2) 3pxtan 2 a

x ='= Eh (C 1 kei 2 - C2 ker 2 + C 3 bei 2 - C4 ber 2 + 2 Eh

sina /t
= Eh C ker2' - v ker 2 + C2 kei2 6- v kei2 [5-171

+ C3 ber 2 - v ber 2 + 4 bei2' 6- vbei 2

*Actually their solution is derived from two second-order equations that are equivalent to Equation [5-16]:
3
L(Q) + Eh X cota = - pxtana
2
12 (1 - 1) -
L(X) - Qcota =0
3
Eh
d2 d ()
L()=x- ()+ - ()--
dx 2 dx x
Q
= Qx tan o
where Q is the shear force acting in a direction normal to the shell.
 III IIIII Yi

p2 sina tana
+ - -
Eh 2
[5-171
2 v/23 (1 - 2) RCOS
= 2 x 12(1 - v2=
I tana sin a h

ber2 , bei 2 , ker 2 and kei 2 are Kelvin (Thomson) functions of second order. The primes indicate
differentiation with respect to , and the C's are constants to be determined from the boundary
conditions. In form, this solution is exact, but approximations can be, and frequently are,
made in determining the constants since their determination is a formidable problem. In most
cases, the determination of the constants involves the evaluation of the Kelvin functions at
the edges. The difficulty is that the practical range of interest for submarine structures and
pressure vessels covers large values of the argument e for which tables of these functions
are not readily available. Hence no solution is really useful unless it includes tables or
graphs of the functions in this range.
In Reference 5-5, Taylor and Wenk have provided one of the most complete, yet
practical solutions. To obtain values of the constants, they made use of one approximation-
that deformations near one edge are unaffected by forces and moments acting at the other.
In this sense, the shell is assumed to be long or "single-ended." This often used approxi-
mation greatly simplifies the computation. It is generally considered accurate so long as
-
6B - 6A -r where A and B designate the two edges.* The results of the Taylor-Wenk
analysis are as follows:

2
QA xA h tan a (eA ker2 'A + v ker 2 A) + 2M A U 2 A kei 'A
2
C1
A (FA + 2 vGA)

QA XA h tan a (CA kei2 eA + 2v kei 2 A) - 2 MA U 2 WA ker 2 6A

C 2
h (FA + 2 v GA)
[5-181
QB XB A tan a (eB ber 2 6B + 2v ber 2 SB) + 2 MB 2
U B bei 2 6B
C-
h (FB + 2v GB)

Bh ta r a
QB B (eB bei 2 B + 2v bei 2 B)- 2
MB U2 B ber 2 6B
4
h(FB + 2
GB,)

5-9
*Baltrukonis has the equivalent short or "double-ended" shell solution where this approximation is not
made. Throughout this chapter (except for the case of the cylinder), equations are written out in detail only for
the single-ended case. The corresponding equations for the double-ended case are usually too complicated to be
presented in this way, and it is easier to work them out numerically.

a~ IIL I r II I 'II n~-----~------r-~c---

-~r*aanhe~-~
n~-arn*nr.
 U2 =2 /1(1 _V2)
FA =A (ker 2 6A ker2 6A + kei 2 6A kei2 6A)

2
GA =(ker2 A) + (kei 2 'A) 2 [5-18]

FB = 6B (ber 2 eB ber2 B + bei 2 6B bei 2 B)

2 2
GB= (ber2 6B) + (bei 2 'B)

It should be remembered that in the notation of this report, A refers to the small end and B to
the large end of the truncated cone. In Reference 5-5, some of the symbols and sign conven-
tions are different, and subscripts 1 and 2 are used to denote the large and small ends,
respectively. The coefficients, it will be observed, are written in terms of the shear forces

QA and QB rather than the vertical forces HA and H B . For both ends of the cone, these are
related by:

pR
Q = H cosa + - sina [5-19]
2

The stresses a0 (circumferential) and a x (along a generator) at any point can then be found
from the following relations as given in Reference 5-10:

Eh 2 d3w 1 d2w 1 dw px
ax = tana - + - - - - - - -tana
12 (1 -v 2
) dx3 x dx 2 x2 dx 2h

2 (1 - v 2) dX2 x dx

[5-20]
2
a--0 1 d-(xNx) ± Eh /
(V -d w+ 1-+ dw
2
h dx 2 (1 - v ) dX2 x dx

Eh3 d w 1 d 2w 1 dw px
N = tana - + - - - - 2 tan a
12 (1 - v ) 2
dx3a x dx 2
x dx 2

where, as before, a positive stress indicates tension, and the upper and lower signs in the
expression for ax refer to the Outer and inner surfaces, respectively. The determination of
the stresses using Equations [5-171] to [5-20] is usually laborious and not always necessary
since it is often the deformations and stresses at the boundaries of the shell rather than at
 I .1 .0 MMYIYIlliahiWi
.

an interior point that are required. It is here that the analysis of Taylor and Wenk s5 s is par-
ticularly useful since it presents the necessary edge coefficients in a convenient, yet exact
form. Furthermore, the accuracy of other more approximate solutions is readily assessable
from an examination of these coefficients. Usually such approximations are based on the
condition that 6(x) exceeds a certain value at the small end of the cone. No such approxima-
tion is involved in the Taylor-Wenk coefficients which were obtained from the exact solution.
As previously noted, the only approximation involved is that conditions at both edges were
not considered simultaneously. This was again applied in determining the edge coefficients;
i.e., C 3 and C 4 were neglected when considering deformations at Edge A and C 1 and C 2 were
neglected for Edge B. In the notation of this report, the coefficients are as follows:

dWB
OHA = KA 0 6 cos a =- KB O3 cosa
dgA dHB

dWB
dMA - KA PA 05 - KB PB D2
dMB

dVA RAKA [ d9B RB KB

1 - v/2 1 - v/2
= [ 6 sinc + I sin a +
dp RA PA dp 2 3
RBPB

+ RAPA (A) -
2 stan 2
I1
RB PB
2 tan 2 a]
4(1-v) RA 4(1-v)
[5-211
dXA dXB
= - KA PA 5
(9HA HB 2 KB B 02
2
dXA KA P 2 dXB 2
Kcos 2
dMA cosa 4 MB Cosa 01

dXA KA dXB KB
3 3
dp
- -
2
tana
[RA PA 5 2 RA A dp
- = - -
2
tan a
1
RB 2 - B B

3 (1 + v) sina 3 (1 + v) sina
2 - 2 (RB B 22
B PB)
2 (RA A) 1I

III II I I L
u lilustOillmulIYIIdIs IIIIYIIIIYIYIIIYYI
IIYIgalIYIIm YIY
II IYIIIIIYIIIIIIYIYIIYII

cos a A sina
A
PA 3(1) COa
R4A 6 2 v5R A

cosha B sin a
P = 3 (1 - v2)
RB' h 2 2RB

2 RA2 PA 1
RAA A
KA = Eh cosa
-Eh
A 4T tan
Z11&a
2 RB2 PB RB 6B
KB = Eh cosa -Ehx/rT tana

-4 A GA 6B GB
v2(FA + 2vGA) - (B + 2vGB)

-CA -C B
5 FA ++ 2vGA 02 FB + 2,GB

4 v 2 GA 4v 2
GB
-ADA -A B 'DB
$
A
D6 = 12(F +2V A) 3= 1
2"(FB + 2 VGB)
A

CA 6A (ker2 6A kei 2 6A -kei 2 A ker 2 CA)

DA =(ker2,'A) 2
+ (kei 2-'A)2

CB = 'B (ber 2 '$B bei2 'B - bei ' B ber 'B)

2 2

D B =(ber2 B)2 + (bei 2 6) 2

Using series expansions for the Bessel and Kelvin functions, Taylor and Wenk obtained values
for the O's over a wide range of the argument '. They presented the results in the form of
graphs which are reproduced from Reference 5-5 here in Figdre 5-5. These curves clearly
show how the O-functions tend toward unity for large values of ' and the extent of the error
involved when they are taken to be unity as is done in many of the approximate solutions.

d"-~ -----~-- ---- I~ ~~

~r~ --- ~--^
- IIWNIW Ini

APPROXIMATE SOLUTION (RAETZ-PULOS)

Long Shells

One of the most useful of the simplified solutions is based on an approximation that
has been employed by Geckeler s - 1 1 in dealing with problems of this type. When it is per-
missible to neglect the variations in radius with the meridional coordinate, the governing
differential equations can usually be simplified. In the case of a segment of a conical shell
along which the total change in radius is not large, this is a reasonable approximation.
l0
Furthermore, Wenk and Taylor s5 have shown that for the range of interest of pressure
d3 w d2w
vessel designers, the terms in Equation [5-16] involving - and - can be neglected.
dX3 dX2
With the origin of x taken at one edge of the segment rather than at the cone apex, the differ-
ential equation in its homogeneous form is:

d4w
=4 '4 w =0 [5-22]
4
dx
where
2
3 (1 - v 2 ) cos a
h2 R 2

and R is a constant average radius.

Since this equation is identical in form to Equation [5-5], it is referred to as the
equivalent cylinder approximation.* Its solution in terms of exponential rather than hyper-
bolic functions is

w b = e -,x ( 1 cos fx + C 2 sin 8x) + e8 x (C 3 cos Px + C 4 sin fpx) [5-23]

The subscript "b" indicates that this is the deflection resulting from the bending of the shell
by edge forces and moments and does not include "membrane" deflections (independent of x)
as would be produced by normal pressure. Raetz and Pulos5-1 2 have considered the long
shell case treated by Taylor and Wenk, i.e., where the edges of the shell are sufficiently
separated that deformations near one edge are unaffected by conditions at the other. In that
case when determining the deflection wb near Edge A, Raetz and Pulos have shown that the
coefficients C 3 and C 4 can be neglected (here the use of the exponential form makes this

*Geekeler's approach was to reduce the Love-Meissner differential equations for arbitrary shells of revolution
to a simplified equation of the form [5-22] by making certain approximations. Hence, the equivalent "Geekeler"
equation for any shell, regardless of its shape can always be represented by Equation [5-22], and its solution
by [5-23]. The definitions of w, x, and f can vary according to the type of shell being considered.

4 1 II I I I II I I I II~
 simplification possible) and the shell parameters can be defined in terms of the local
geometry. Their solution is
Near Edge A:

KA -pAx
wb - e [QA cos PA x + PA MA (cos fPAa -sin PAX)] [5-24a]
cosa

where KA and PA have the same meanings as in Equation [5-131. At the other end of the
shell, C1 and C 2 can likewise be neglected, so that:
Near Edge B:

-K
-KB PB (x -xB) cos
Wb -cos eos X
PB [QB cos PB B + PB MB (sin PB XB - cos pB XB)]

[5-24b]
x
+ sin PB [QB sin PB xB -PB MB (sin PB XB + cos PB xB)]l

-K
-KB -PBY
Wb - e [QB cos PB Y + MB PB (sin PB Y - cos PB y)]
cos a

y= B -x

It can be noted that RA and RB are related by:

RB = RA + XB sin a [5-25]

where a > 0 for Edge A at small end,

a < 0 for Edge A at large end, and
XB is the slant length of the cone frustrum.
It is also important to remember that the sign conventions used here are those indicated in
Figure 5-4 and not those followed in Reference 5-12. With the present conventions:

P RA
QA =HA cosa + sina
2

[5-26]
PRB
QB= HB cosa + sina
2

I-~-CP~ -- -~~*,*~ X~ -------

 1, 11
. lIIum iu al olll ,lulllt III

'The reader is cautioned that unless one system of sign conventions is consistently followed,
errors can easily result, especially when several connected shell elements are under
consideration.
To obtain the total stress at any point in the shell, one must determine the stress as-
sociated with the bending deflection wB and add to it the corresponding membrane stress.
To do this, Raetz and Pulos have made use of simplified relations for the stress resultants
consistent with the approximations already made:

pR Eh 2 d3 wb Eh d2w
ax _ tan a +
2h cosa 12 (1 - v2) dw3 2 (1 - v 2 ) dX2

[5-27]
pR (1 - v/2) Ewb
a0 Cosaa
hAcos - - R Cosa + v ax

In each of these equations, the first term is the membrane stress. To get an idea of the kind
of approximations involved, one can compare these relations with the exact Equations [5-20].
By combining Equations [5-27] with [5-24a] and [5-24b], Raetz and Pulos obtained expressions
for the stresses than can be written as follows:
Near Edge A,
PRA tana
U ho s + [QA (A + 2
+ A) MA PA CA]
2 h cosa h
6
+ 6 [QA A +
MAA (A + CA)]
h2 PA
[5-28a]
p RA 2 A
O = (1 - v/2) + cosa QA7A +MA 1A ( A rA)] + vax
h cosa

Near Edge B,
PRB tan a
x = 2 h cosa + ----- [QB (CB - B) - 2 MB B]

6
1[-QB B + MB 1B (rB +
]
2
h PB

[5-28b]
p RB 2
PB
S =- h Cosa
cos a (1 - v/2) + h cos a [ QB CB + MB B (<B - B)

" I I II I I I I II I
where

7A = e A cos 8A X
~A = e-3Ax sin JA X

7 sin 8BY
B = e _B cos PBY 4B eB

Graphs of the functions 71and are reproduced here from Reference 5-10 in Figure 5-5.
To get the corresponding edge coefficients, Raetz and Pulos have employed an approxi-
mation for the radial deflection ii which is discussed in Reference 5-12:

2
pR
= wb cosa + (1 - v/2) [5-29]
EA cosa

The edge coefficients can then be written:

dWA dWB
- KA cosa - KB cosa
aHA dH B

diwA 9WB
- KA PA -KB PB
dMA dMB

dWA KA KA RA dB KB KB RB
(1 - v/ 2 ) + - sina dp - (1 - v/2)- - sina
dp 2 B2
2-KA A [5-30]
dXA =- KA P3 dXB
A
dlH
KB PB
A HOH
B

2 KA P 2 2 2
dXA dXB KB P
dMA cos dMB cosa

dXA KA tana dXB KB tana 3

P2 ( RA PA 2 RA PA (9P 2 (RB 2 R B 8B

dw dw
where XB has the meaning , not - . It can be seen that these coefficients bear a
dx dy
close similarity to those for the cylindrical shell (Equations [5-10]) and that they are almost
identical to those given by Taylor and Wenk (Equations [5-21]) for large arguments of 6 where

the 0-functions approach unity. The exceptions are the third terms in - and - which do
dp dp
not appear in the Raetz-Pulos coefficients. It might also be mentioned that the expressions

21
 aXA aXB
for - and - appearing in Equations [5-12] include within the parentheses the term
ap Op
3
- -2R/ which results from taking the radius R to be a function of x rather than as a constant.
This term is usually negligible compared with the first term.
The foregoing equations can be used with confidence so long as J3 AXB and PB XB are
no smaller than 3.* In this range, conditions at one edge do not significantly affect deforma-
tions at the other. However, it should be realized that near the middle of the shell, condi-
tions at both edges are apt to have influences that cannot be disregarded unless PA XB and
PB xB are greater than 6. In other words, the effects of both discontinuities can be expected
to overlap (and over a considerable portion of the shell if PxB = 3). Raetz and Pulos have
shown that the total effect of both discontinuities can be accurately accounted for by
superposition. To calculate the stress at any point using Equations [5-28], one adds to the
membrane stresses the additional stresses resulting from M and Q at both edges. If PA XB
or PB xB < 3, this procedure will result in inaccuracies because in that range, the effects of
the discontinuities overlap throughout the entire length of the shell so that MA and QA are
influenced by MB and QB, and vice versa. An analysis that takes this interaction into
account is described in the next section.

Short Shells
This case has been covered in a number of papers including, for example, the work of
Linkous and Horvay s5 - 1 3 and that of Raetz.** Equations developed in this latter paper are
convenient for presentation here because they are developed for the particular loading con-
ditions of interest.
For the short shell, it is sufficiently accurate to dispense with PA and PB and use
the single parameter

8=
f3 (1 _ v2) CRSl,[5-311

RA + RB
where R is an average radius given by 22 ; a is still measured from Edge A so that
XB denotes the length of the shell. The "bending" deflection, is again obtained as a
solution to the differential Equation [5-22]. However, in this case, it is convenient to ex-
press wb in terms of hyperbolic instead of exponential functions since it is not permissible

*This is approximately the same as saying 6B - 6A >

3 1- 7r Vz, the limit placed on the Taylor-Wenk
analysis (page 14).
**Unpublished work on short shells.

* I II II I ~___ 111111bm~F-~ll~*Dnrsrrarr
1-.-~
111
_-
to make the approximations leading to Equations [5-24]. The form of wb is then the same as
that for the solution [5-6] of the cylindrical shell equations (excluding the membrane term):

wb = C 1 cos fx cosh fx + C 2 sin fix sinh 8x

[5-32]
[5-321
+ C 3 cos 3 sinh 3 + C 4 sin 3 cosh 3x

If 0 is allowed to represent 3
B, the coefficients can be expressed by Equations [5-7a] with
H replaced by Q:

C 1 = K [A3 QA + +
3 l MA S QB - 44 MB]

C 2 = Kp A
MA
[5-33]
3=K( + 1) B +
2 B +'6 MBI

(1 - \1) A4 QB
3
4 K 2 QA - X2 MA 2 + 3 6 MB]

2R 2 f
Here K represents , and the series relations [5-12]-[5-15] all apply in the present
Eh
case.
The radial deflection wi7is again obtained using the approximation [5-29]1 and the
stresses are given by

pR Eh
+ d 2 Wb
x 2 h cosa
2(1- V2) dX2
[5-341
pR (1 - v/2) Ewb
Oo = - - Cosa + vo
h Acosa R x

It may be noticed that the equation for ax contains one less term than does Equation [5-271].

Raetz has found that this term involving - is much snaller than the others and can
dx
be neglected.
 In terms of the A-functions, the edge coefficients for this case are as follows:

9
wA WB
diH ,,
-= - -K 3
aHB_

diFA Kf3 1 , d9wB Kf3A 1

dMA cos MB cosa

dWA (99B
- - KXAs - =- -KA 5
dHB HA

dWA -K A4 dWB K 3 A4

dMB cosa dMA cosa

dWB K [1-v/2
9A
( +A 5)Rsina 1 -v2 (3 + Xs) R sina
K [1-v/2 3
ap 9p 2 cosa Lp
[5-35]
dXA dXB KOA
Kos 1
HA COS al dHB cosa

dXA 2K13 2 X2 dXB 2K p32 X

dMA cos 2 a 8MB cos2a

dXA K 13 A4 dXB K 13 A4
HB cos a dHA COSa

dXA 2K 32 0
XB 2K 32 A6
6

dMB 2 a MA cos 2
,
cos

dXA KR3 tan a dXB KR3 tana

( 1 + A4 ) S 4)
dp cos a Op cosa

ANOTHER APPROXIMATE FORM (AFTER HETENYI)

A later section of this report presents an approximate solution due to Hetenyi5-14 for
spherical shells which is remarkable for its simplicity and accuracy. The same approach
permits an equivalent solution for conical shells. The author was unable to find such a
solution in the literature, but believes this only to be the sign of an incomplete search since
the method is so readily adaptable.

- I I ill r z I I --- ~-~-

 The two second-order homogeneous equations for the conical shell can be written:

L (Q) + Eh X cot a = 0
[5-36]
()- 12(1 - v 2 ) cot 0
L(X)- Q Coto = 0
3
Eh

in which
Q =x Q tan a

2 d( ) ( )
L( )= - d ( )+
dx 2 dx X

The combination of Equations [5-36] produces the fourth-order equation (see Reference 5-1,
p. 458),

LL(Q) + 4 A4 Q= 0 [5-37a]

which can be expressed as the two second-order equations

L(Q) ± 2 iX2 - 0 [5-37b]

where

3 (1- v2)
)1h tana

with the transformations

-_ q(z)

[5-381
z 2 VT
Equations [5-37b] takes the form

2
d 2q q15qq + 2diA 2 q0 [5-39]
2 4 2
dz

which are easily solved if the term -15 q/4 02 can be neglected. We can expect in general
d 2q
that d 2 q >> q so that the approximation ought to be reasonable except for small values of 2.
2
dz

-11111
- ls Il--~
-U~~~"1*~*lil~_
 - --- Im - IIm-- 11
IN
II

Thus, the apex region is excluded from consideration.* With this approximation, Equations
[5-37a] and [5-37b] are reduced to

d. 2 i X2 q = 0
[5-40a]

q +4 4
q= 0
[5-40b]
4
dz

It will be seen that these are the cylindrical shell equations whose solution is

q = e-X z (C 1 cos Az + C 2 sin AXz)

[5-411
+ eXz (C 3 cos Xz + C 4 sin Xz)

The shear force is thus obtained from the relation

- Z2
Qtana - [e - X z
(C cos Az + C 2 sin Xz)
4 1
[5-421
Xz
+e (C 3 cos Xz + C4 sin AXz)]

The first of Equations [5-28] can be solved for the rotation X:

tana d 2Q 1 dQ 4
X -+
Eh dz 2 z d Z221
[5-43]
tan a d2q 15 q
Eh'/7 d -
4 z2
15q
As before, we neglect the quantity -q and have the results
2
4z

tan a d2q
x=- Eh & [5-441
2
dz

*Hetenyi's solution for the spherical shell excludes the polar region on the same grounds.

4)1(11111 1 I I I I I IIII I 111

Knowing the basic variables Q and X, we can determine all quantities necessary for obtaining
the stresses and deflections. With the aid of Reference 5-1 (pp. 454 ff.) and making allowances
for differences in sign conventions, we get the relations:

2
Q p x tana 4q pz
N - x 2 z2vr- -8 tan a

dQ p R (1 - v/2) 2 (dq q) p32

NO = (1 - v/2) tan a
dx cosa aV dz 2a 4

2Eh dX V) Eh3 d 2
M 1(1 - -
2) (d
(1 _-12
V2) d _ 6 (1 - v
[5-451
+v = VM d2q
Mo = 2
2) dx 3 - 2
1 (1 - v X da

sina d -__ sina [ dq q (1+++v

W =- - =14)
Eh d x 2 Eh dz 2z

2
p
4
sin 2 a (1- v/2) p? (1-v/2)
+ +
16Ehcosa Ehcosa

in which the membrane stress resultants and deflection have been included. For the case of
a long or single-ended shell, near Edge A the solution is

q = e - Az (C 1 cos Az + C 2 sinAA) [5-46a]

and near Edge B

q = eAZ (C 3 cos A + C 4 sinA 2) [5-46b]

The C's can then be written in terms of the edge moments and forces.

ZA )A A A ta n a
C- dA e SzA
zA [ QA A tana (sin AzA + dA cos AS) + MA A tan a sin XAA
4

[5-47]
Za a Az QA OA tana
C2 - dA e A (-cos AhA + dA sin A A) - MA htana cos AhA]
4

-- ~~r~ru~*x
 ZB XIZB z B t an a
AzB B
dB e 4 (-sin kZB +dB cos XZB) +MB Atana sin XZB
4
[5-47]

zB ZB tan a
C4-- dB
B
eXZB [ QB 4
(Cos XzB + d B sin AZB) - M B Xtan a Cos XZB

dA =1+ 21 - 4v
XZA

1 -4v
dB = 1 2 XZB

The corresponding edge coefficients in which the effects of lateral pressure have been included
are as follows:

aWA KA cos a alB KB cosa sin a

aHA dA
[1+ sin a
A OH B dB 4 PB RB
4pAR A
2)
(1 - 16 v 2 ) (1-16 sina 2
+ + 4 )
2 43A RA 2v) B B

dWA KA J A KB JB

9MA dA dB

aWA RA tana adA KA (1 - v/2) awB RB tana e3HB

KB (1 - v/2)
2 3 2 2,8f
ap 2 aHA PA alB B
19X

dXA KA PA aXB KB PB [5-48]

OH A OHB dB

2
KA PA2 2KB P2
aXA
MXB
dMA dA cosa dB cosa

RA KA PA aXB R KB APB
dXA
tan a - - tan a
dp 2 dA Op 2 dB

*The reader may observe that it is possible for dA to be zero, in which case the solution is meaningless. How-
2
XA (4 v- 1) tan a
ever this can happen only if - = , which is extremely close to the apex and therefore well out-
h 16 ( -
side the region in which the solution is intended to be applicable.
28

- r __ I ii II I
The symbols PA' IB, KA, and KB are as defined in the section covering the Wenk-Taylor
analysis. One can see at a glance the close correspondence between these equations and
those obtained by Hetenyi (Equations [5-75]) for the spherical shell.
It is also instructive to express the edge coefficients in terms of the parameter (
employed by Taylor and Wenk. Using the relations

sina _ - sina
2 PA RA A 2 fB RB

[5-491
vf-(1-4v) dB = 1 - (1-4v)
dA =1+
A 2 6 26B

one can write Equations [5-481 as follows:

OHA WB
= KA cosa * o06 9
=-KB cosa 03
HB
MWA
dWB
- KA PA (05 KB PB )2
MA MB

adA RA KA sina KA OWB R B KB sin a (1 - v/2)

9, * 06 +
(1 - v/2) 22 . 3 + KB 2
dp 2 2 PA Op PB

dXA [5-50]
XB
= - KA 8A (05 - KB PB &2
dHA HB

8MA
dXA 2 KA PA2 04 dXB p21co 1
Kg P
22KB
dMA cos a dMB cos a

dXA RA KA PA OXB RB KB PB
- tan a 2
2 2 tana o s5 (3P 2 2
Op

where
1 1
1 -2 = B ; 04 5= dA

e 1 + j; 6
1
dA
1 (_ - 16 2)
dB 6B 8 62 LA 8 eA2
" 111

1 1 1
Expansion of - and - In power series in - results in:
dA dB

2
1 (1 -4v)
=1- -(1 -4v)+
ds
2 Q2

1 _ v)2
-=1+ (1 -4v) (1-4
dA 2A -4)+2

[5-511
2 v ( - 4 )2

6B 4Q
2
2 v (1 - 4 v)
+ +
6A 4(

For the case v = 0.3, the values are

0.02
1 ee =
2 - 10l-- - +
+ 2 - **
0.84853 0.01
=1 - + ....
3

[5-52]
0.02
4 = 5 = 1+ + 01 0
4 510 6
6.2

0.84853 0.01
0)= 1+ +
6A

The first two terms in such series are identical with corresponding terms given by Wenk and
Taylor in their asymptotic expansions of the 0-functions for large (see Figure 5-5).

p= - 1.855
=15 5
10 6B 2

[5-531
3.730
2 = 1-
10 Bf

" I II_ I I I I I __ ~,~ -~--~-~

 0.84853 1.865
=1-

04= 1+ 41.855
10 6A '2

[5-531
0' =1+IOA2 3.730
s10 6A A2

0.84853 1.865
*
6=IA 6A A2

Since the o's tend toward unity for large values of 6 just as the Q's do, it is easily
shown that the edge coefficients given by Equations [5-50] reduce essentially to those given
by Raetz and Pulos (Equations [5-30]). The only differences are the second terms in the
dXA dXB
Raetz-Pulos expressions for - and -
which do not appear in Equations [5-50].
Op Op
The stresses near the edges can be found by substituting the solution [5-41] into the
relations [5-45]:

Near Edge A,

N 4 e- \ z (C 1 cos Az + C 2 sin A z) PZ2

2- 8

2A e -Az1
NO= -C sin Xz+ 1 + - cos za

+C 2 cos )z- 1+ 2 sin Az]

z2
+ - (1 - v/2) tan a [5-54]
4

e - Az1-4v
M X zx/ tana C1 -cos Az + 1 + 2 z sin Az

- C02 sin kz + 1 + 2- )cos A

2(1 - v 2) e-AZ
MO = VM x + (C2 cos Ax - C 1 sin Ax)
A2 z2 tan a

*yr*r~yl~a
c*i*
**rr'*m"J11)(1*(~~l~
 _

Near Edge B,

4 Xz sin
Pz
z)- -tan
2
a
NX = 4 e (C 3 cos X+C 4
2 8

N0 2Xe C -sin Xz + 1 - cos sz

1 3 1-2Xz

+c4 [cosXz+ 1 - 2 sin X]

p2
+ - (1 - v/2) tan a [5-54]
4

e Az { 1 -4 v)
M= zXZtana C3 cos Xz + 1I - sin XZ
x AOzV-tana 3 2X0

+ C4 Sin Asz 1- Cos X 2

2
2 (1 - v ) e Xx
M= vMx - (Ca sin Xx - C4 cos Xx)
,\2 o 2 Vr" tan a

EVALUATION

A good deal of experimentation has been done with conical shells to evaluate the analy-
ses just discussed. It is not practical here to cite all sources of information, and no effort
will be made to do so. Instead mention will be made only of some work carried on at the
Model Basin which may be regarded as typical and provides enough data to draw general
conclusions as to the applicability of theory.
5
Investigations by Borg, s-1s Krenzke, 5- 1 6 and Raetz 5s- 1 7 have produced a large amount
of data that support the general theory. In particular, they have shown the Raetz-Pulos
method to be adequate for values of A as small as 11, and this includes values of ac up to
60 deg. In addition, the data indicate that the rule that superposition of edge effects is
unnecessary when [3 AXB and 83 BXB exceed 6 is a good ope. The general results of these
studies are encouraging since they show that the relatively simple Raetz-Pulos equations
are reliable over a wide range of geometry. Figure 5-7 is a typical plot of strain distribution
taken from Reference 5-15 for a shell consisting of two cylinders of different diameters con-
4
nected by a conical section with a = 60 deg. Since the Hete nyi-type solution was only
recently obtained, there was not time to include it in the numerical comparisons.

I I IPN r I WI
 SPHERICAL SHELLS
Spherical shells present some of the more interesting mathematical problems to be
found in the field of stress analysis. One might suppose that because of the high degree of
symmetry possessed by the spherical shell, its analysis would be mathematically simple.
Unfortunately this is not the case. All known forms of the exact solution are so cumbersome
and comnplex that they are impractical to use except when programmed into a high-speed
computer. Because of this, much effort has been devoted to obtaining approximate solutions
and it is these that will be of particular interest here.
Figure 5-8 is a general representation of a spherical shell element in a pressure vessel.
As drawn, it could depict a transition component between two cylindrical sections of different
diameters; 0, the meridional coordinate, is the angle between the axis of symmetry and the
radius to a point on the shell. If A is small and OB is 90 deg, the case could be that of a
hemispherical closure or end cap with a small crown opening like a hatch. With OA zero and
OB equal to 90 deg, we have a hemisphere. For OB slightly less than 180 deg and OA equal
to zero, the case could be a sphere with a small hatch. Since this chapter covers only axi-
symmetric deformations, noncircular openings and openings whose centers do not lie on the
axis of symmetry are excluded. These and other more difficult cases are taken up in the
chapter on hull penetrations.
The basic theory for spherical shells can be found in References 5-1, 5-2, and 5-4.
In addition, the reader may wish to refer to an excellent summary s5- 1 8 by Tsui and Stern of
the important methods of solution and a discussion by Galletly5-1 9 of the applicability of
some of these methods to open-crown hemispheres.
With Q denoting the shear force per unit length on a shell element (Figure 5-8) and y
the rotation in the meridional plane, the governing differential equations for no surface loading
can be written as follows:

L(Q) + vQ +E
Eh -0
[5-55]
2
12 (1 - v )
L (x) -v- 3 Q=O
Eh

where L is the operator,

L( d 2 ( ) d( ))ot 2

L( ) = - + - cotek-( ) cot2
d It
d92
Equations [5-471 can be combined to yield the equation

LL (Qo) + 4 p 4 Qo = 0 [5-56a]
 401119011101 10111

which is equivalent to the two second-order equations:

L(Qo) +
2ip 2 Q =0 [5-56b1

where

3 R2 p2
P4 (1 V2)
h2 4

2
3R
- (1 - v2) for thin shells
h2

The corresponding set of equations for X are identical in form.

EXACT SOLUTIONS
0
Evidently it was Reissner 5 -2 who first recognized the advantages of taking as the
dependent variables Q and X rather than the displacements u and w, as is a common
practice. Once a solution to the equations has been obtained, one can get the displacements,
rotations, and stresses from the following relations (which include the membrane deforma-
tions due to external pressure):

1
x - Eh [L(Q) + vQ0]

d Q dQ¢
1 (
S( + - coto-Qocot2 +VQ
Eh d02 dh

R sin 5 dQg
0 pR2 (1 V)
dg - vQ' cot o + sin 0 [5-571
Eh .
Eh 2 Eh

N 6M
S-h h2

No 6M O
+ h2
aO h 42

-aC I I IIL 311 1 I I II I

 Yli
1IIYIIIIIII
YIYlill
IIIIIIIIII -
ami
luIII d t ii
lt1 ilii IIIYIIumIII
IIYIY
IY MIY I -mill WIFE

M- M9
1 EA3 dX + v X cot
o

M S Eh 3
=-+7ct4)
dy ) = M+12
Eh3 o
2) R 0+ 1 X ot
12 (1v

pR
N =Q cot ¢-2

N =o do
2

The terms involving p are the membrane components.

Two forms of the exact solution are frequently found in the literature. One, due to
Meissner5- 2 1 and Bolle, 5-22 is expressed in hypergeometrical series; the other, introduced
by Love,5- 4 is in terms of Legendre polynomials. Both forms are too unmanageable* to be
of much use in practice and will not be discussed in detail (they are completely described
in References 5-1, 5-2, 5-18, and 5-19). Furthermore, they exhibit slow convergence for
certain ranges of 0 where the computational task becomes formidable. It has been found
that acceptable accuracy can be obtained using far simpler approximate solutions. The
importance of the exact solutions is that they are the standard against which the accuracy
of all approximate solutions can be assessed. The remainder of this section explains some
of these simpler methods.

APPROXIMATE SOLUTIONS

Various approaches have been used to obtain workable solutions to the differential
Equations [5-56]. So numerous are the papers on this subject that it would be difficult even
to compile a complete bibliography, and to present a detailed discussion of each method
would require a treatise of far greater scope than was intended for this chapter. Instead it
will be the objective here to present a few methods that give results of good accuracy and
can be put to practical use without the aid of high-speed computing machines.

Geckeler's Solution for the Polar Region

When the angle 0 is sufficiently small, certain approximations can be made that greatly
simplify the solution of the differential equations. Cases that fall within this range are illu-
strated in Figure 5-9. One type of solution is based on shallow shell theory; it will not be

*Galletly, 5 -2 3 in a discussion of Reference 5-19, estimated that the desk computation time for one set of edge
coefficients using Love's method is about one month.

r~--- Il~m~rmraxllrl*~
~nlrr
 discussed here since it is covered in some detail by Lomacky in Chapter 9 (Hull Penetration§)
of this summary. Another solution was originally obtained by Geckeler5-24* and has been
carried out in detail by Esslinger 5 - 2 5 and Galletly. s- 1 9 Of the cases represented in Figure
5-9, Case b, that of a spherical shell with a small polar opening, is the one of primary inter-
est here, and the only one for which detailed expressions are presented. Case a, the shallow
cap, is not apt to be encountered frequently in conventional submarine construction. The
same is true of Case c. Moreover, since / is restricted to small angles, the interaction in
that case between Edges A and B would have to be considered-a situation that would lead
to severe analytical complications.
5
If the angle is restricted to small values, its trigonometric functions can be approxi-
mated by the first term of the appropriate series representations. Thus cot 0 can be replaced
1
by - and Equations [5-56b] take the form

d2Q 1 dQ¢ Qq
- + - - 2ip 2 Q =0 [5-58]
d5 2 b dq ¢2

Each of the equations is reducible to the modified Bessel's equation of first order. The
solution can be written in terms of the first order Kelvin functions:

Q = A 1 ber'(6) +,A 2 bei'(6) + A 3 ker'(6) + A 4 kei'(6) [5-59]

where =pV 0$ and the primes indicate differentiation with respect to 6.

These functions can be found tabulated in many references in the literature, among them
Fliigge's "Four-Place Tables of Transcendental Functions."5- 2 6 Several other forms of the
solution are possible. Esslinger has chosen to write it in terms of Schleicher functions:

Q0 = C 10 '(6) + C2202) + C 3 '3() + C 4 q4(4) [5-60]

5 11
*This should not be confused with the well-known "Geckeler approximation'" - previously discussed in the
section on conical shells. The application of this approximation to spherical shells is described in a later
article.

-.~II II IIIIII II I I I IIIIIIIIIIR

To dispel any impression that the Kelvin and Schleicher functions are fundamentally different,
the following relations are noted:*

2
S(x) = ber(x) 3 (x ) = - - kei(x)
77
[5-61]
2
J2 (X) = - bei(x) 0 4 (x) = - -r ker(x)

They can be found in the Jahnke-Emde tables5- 2 7 as real and imaginary parts of Bessel and
Hankel functions:

(( = ReJ (X -7)

2 (X) = lmJ 0 (XArV)

[5-621
113(x)H
H
H' (x NF?.
0 4(X) = Re x )

04 (X = 1 xN

1
Carrying the approximation cot 0 = - over to the relations [5-57], we get the following
expressions:

1 d2 Q 1 dQ 5 Qo
S Eh d€2 € vQt )
do 2

_ Re (dQ
X=- d vQ ) ** 2
IoQ + --
pR
(1 - v) q [5-63]
Eh do € 2Eh
3
Eh
dX vX
S12 (1 - v 2 ) R do+€

*This multiplicity of names and symbols assigned somewhat indiscriminately to the same basic functions can
often lead to needless confusion. One who looks for tables of Kelvin functions and finds Thomson functions in-
stead may face a long search unless he knows that Lord Kelvin was Sir William Thomson.

R sin dQo vQ \
5 19
**Galletly gives this as - - , but this seems inconsistent with the basic
approximatEh
a pproximation. d
 A 11 I ulId 'N'' ll, IIllul" I'h,,

Eh3 y
MO = vMO +
12

N k pR
2

dQO pR
NO
S 2

pR
where --2 is the membrane stress resultant. Unfortunately, from here on, the equations have
been developed by Esslinger only for the case v = 0.25. In Chapter 9 of this summary,
Lomacky has rewritten them in terms of general v, and it is these equations that are presented
here.
For Case b of Figure 5-9 it is assumed that the interaction between Edges A and B
is negligible. This means that C 1 and C 2 can be set equal to zero. The resulting edge co-
efficients are:

OwA Wc 0
XA Xc
aHA EOA aHA EhOA

OWA Xc 0XA Xd
- - [5-64]
dMA EhA dMA Eh2 OA

OWA R2 (1 -) XA
-O A - =0
Op 2Eh Op - 0

where

(' 3 P3b+ ' 4 0 4 )

Xc = A X
Xe

(0 + ) 2 /3 (1 - v2)
Xd = A X
e

• , 1-v "2
Xe =03 04' 2
40 3 + 2 +04]
4

3(02 + 04) - 2 ( 33 4 - 4 ; 3 ) + (1 - v2) A (03 + 4)

Xe 2 3 (1 - v2

CA OAAP

d I I II I I I
 molkh YI II lliIYIYIIIIII

In these equations, the approximation has been made that HA = NA and H B = N .

Finally, using the expressions for a and a0 given in Equations [5-57], these stresses can
be found from the relations

pR2
No = nc (e) J + nd (6) K - 2

K -
pR
No =tc()J + td () -

[5-65]
h J + md (6) K

M0 VM¢ 3 (1 - v 2 )
h h + (0 4 J - 3'K)*

in which

J2M
2A---A c K 2 MA**
J = 1J HA cos OA K = KKKH
HA cos OA ~
4 A 2
6A q4(6A) -(1 - v) 03V(A) J2 =
Xe %
2
4'(A) - 2v3 (1 - 2)
Xe V2'

6A /3 (6A) + (1- v) 4(AVA) 03(Q 3 2 (1 _ v 2)

K Xe V 2
Xe)2

nc (6)
4c
nd M( =
tc (6) V/2 04 (6) - nc(6) td (6) =- )/ 03 (6)- n( M

1- (1 - v
MC ( 6) = ) ( - 4), 4 )]
q[3( +
6 (1 v2)

1 4 (1- ) 03(6)]
6 (1 - v2)

Mo
*Esslinger does not give an expression for - ; Galletly's expression differs from the one presented here,
h
but it appears to contain a sign error.
**Esslinger does not make the approximation cos OA = 1, although this would seem to be called for if the
equations are to be consistent.

39

- ---~-~,~,.,,,.,.~,,,~~,,,
 Notice that J and K are constants involving functions evaluated at Edge A
(eA = P v2-4A) whereas nc, d tI, td, mC, and md are all functions of the variable . The
equations are valid, of course, only over a small range of 6 because of the limits placed on.
However, one usually finds that the bending effects decay rapidly with increasing 6, so that
where the solution is no longer valid the membrane solution alone may be sufficiently accurate.

Hetenyi's Solution (Polar Region Excluded)

Because of its simplicity and accuracy, this solution (already mentioned in the section
on conical shells) is of great practical interest. Its range of validity is not limited to large
values of 0, but it does exclude the polar region in which Geckeler's solution is applicable.
Figure 5-10 therefore represents two general cases for which Hetenyi's solution should be
good. The only restriction in either case is that the opening angles must not be too small.
Hetenyi makes the transformation

q()1 f() [5-66]

1

q(4) now replaces Q as the dependent variable, and Equations [5-56b] take the form

1 [d2q 1
+ q 2 4 cot2 + 2 ip)
/s d4)2 = 0 [5-67]

The advantage in using the transformation [5-66] is that Equations [5-67] have no terms in-
dq
volving - . Equations [5-67] are therefore a good deal less complicated than Equations
d1 3 cot * This is
[5-56b] and can be solved easily if one neglects the term q 2
-2 4cot .* This is

acceptable so long as cot 2 0 is not large (i.e., 0 is not small). The resulting equations to
be solved are:

d 2 q + 2ip2 q = 0 [5-68a]
do
or, the single fourth-order equation:

d 4-+
q
4 p4 q =0 [5-68b]
4
d0

*This means that q -I - - cot 2 ) < < and it is the one basic approximation made in arriving at the
i2 4 d2
solution.

I I L I I L I
 YII -

Its solution is

q = e-P (C 1 cos p 5 + C 2 sin p )

[5-69]
+ eP (C 3 cos p +C 4 sin p[)

It is seen that Equations [5-68b] and [5-69] are identical to those for the cylindrical shell.
The complete solution to the spherical shell is
1

Q = - [ e-P (C cos p + C 2 sin po)

[5-70]
+ePk (C 3 cos p + C 4 sin p)] [50

With the aid of the first of Equations [5-571, one gets for the rotation

1 [d2q f
X I q cot 2 +v [5-71a]
Eh -\/i do2 (2 4

Neglecting the coefficient of q in this equation is consistent with the approximation already
made in arriving at Equation [5-49a]. Hence one can write

1 d2q
X= d 2q [5-71b]
Eh -,n d02

with the assurance that no additional approximations have been introduced. The other quanti-
ties given by Equations [5-57] can then be expressed in terms of q as follows:

R Fsi-o dq
w=- Eh d - q ( 1p2 + , cot 1+ R2
sin q
2 Eh

S R Fd 3 q d2q (1 )cot
M= - 4p4 ),IS- i d -3 - d 2 cotn

R(1-v 2) cot o d2 q
M 0 = VM - 4pvi 2
[5-72]
4p 4 d0

cot 0 pR
N =q /-no 2

N -- 1 dq q cot
cot - pR
.MINIi INIMMIN 110110,~r

For the case of the long or "single-ended" shell, we have near Edge A:

q = e-P O (C 1 cos pO + C 2 sin pq) [5-73a]

and near Edge B:

q=ePI (C 3 cos pO +C 4 sin pC) [5-73b]

The C's can then be written in terms of the edge moments and rotations:

eP A
sin OA MA
] 2 sin p OA
QA [sin POA + CA cos pCA + P -
R
CA

MA
ePCB sin B sin PCA] -2p
[-cos PeA +cA cos p A
QA R

[5-74]

C- MB
[- sin p B + CB cos p B] + 2 p -R? sin p OB
GB QB

e /sin OB
M
[cos
B
GB - P R cOS PO
I Q [Cos PO + 0B sinP0B

(1 -2 ) (I1-2gv)
CB = 1 -
in which cA = 1 + cot A and
c)ot B'.*
2p 2p
The corresponding edge coefficients for the load parameters H, M, and p are as follows:

2 A )]
2 cot OA cot
dWA pR sin 2 OA.
2
1 +2 + (1 - 4 v
HA Eh cA p 8p2
[5-75]
2
aw0B 2 pR sin 2 'B cot OB cot B
(1 - 4 v2)
HB Eh cB 2
[1- 8p

*Here, as in the case of the Hetenyi-type solution for conical shells, we have a denominator term which can
become zero. For the case v = 0.3 and h/R = 0.05, this would happen if kB = 2 deg; for thinner shells, the
angle would be even smaller. Such cases are clearly outside the region in which the solution is applicable.

-I Ill r L I I I I
 2
MwA p2sin A WB 2 p 2 sin 4B
dMA Eh c A dMB Eh cB

oWA
(9W-A R cot (A OwA R2 R cot 4B OWB R2
- + - (1- v)
9p 2 dH
A
2Eh Op 2 dH B 2 Eh

OXA 2p2 sin OA 0

XB 2p 2
sin OB
OHA Eh c [5-751
A
dH B Eh CB

0 3
XA 4p dXB 4p
3

OMA EAR cA MB Eh cB
"

OXA p 2 R cos A 0 p2 R cos B

XB
0p Eh cA 9p Eh cB

In forming these coefficients, use has been made of the relations

pR pR
+ [5-76]
QA =HA sin A + cos 4A QB = HB sin 4B -- cos B

For the moments and stress resultants, we have

Near Edge A:

N e -P cot [C1 p R
sin
Nk o [C cos pS + C2 sin p] - 2

pe-PO cot 4
NO = -no C Cos po4+sin p 4 + Sp cos p4 [5-77a]

cot in pR
+ 2 -cos p + sin p +c sin p
2p 2
 .1

R C- o si (1 - 2v)
M - C [cosp- sin p cot sin pC
p
2p P 2p

+ C2 cos po + sin p + (
2p
cot 0 cos P } [5-77a]

R(1 - v2 ) Cot
MO = vM - eo PC [C 1 sin pC - C 2 cos pC]
2
M
=vM~-
0 2 p ViF

Near Edge B:

eP 0 cot [ pR
Vs [C cos P + sin P 2

peP cot q\
NO = s [10 cos p 9)
(cos p k - sin p -
/I7 K'a
cot . pR
+C4 (cos p+s in p0- sin
2p P I) 2
[5-77b]
R eP (1 - 2 2v)
2p n cos p +sin p - p. cot sin p
1 2 p,

-(1 - v)
- C4 cos p - sin p - cot 'cos p ]}

MO = vM + eP0 [C3 sin po + C4 cos p]

2p2 -

It is interesting to note that all relations obtained from Hete'nyi's solution can be re-
5- 11
duced exactly to those resulting from the application of Geckeler's approximate method
to the case of a spherical shell.* If ' is in the neighborhood of 90 deg, sin ' can be approxi-
mated by 1 and cot ' by zero. In that case, Equation [5-66] is simply

Q =q [5-78]

and Equation [5-68b] reduces to

d4Q
4 [5-79]
+ 4p Q = 0

*This approximation is discussed in References 5-1, 5-18, and 5-19.

3 1 111 I IIII _I~ _~.~la

b-- x.~~~arrs~
rbl~ C
31*
In all subsequent equations, Q0 replaces q, 1 replaces sin S, and all terms multiplied by
cot 0 do not appear. The result thus can be regarded as an approximate form of Hetenyi's
solution. Since this latter solution is itself relatively simple, one may wonder whether the
slight additional simplification gained by going to the Geckeler form is worth the accuracy
lost.

Leckie's Solution (All Regions)

Since it is intended to be applicable for all ranges of the angle 0, this solution is con-
sidered to be the most accurate of the approximate forms described here. Leckie 5 - 2 8 obtained
his solution as a special result of some general theoretical work on the antisymmetric response
of spherical shells to localized loads. The solution has received further discussion in a paper
29
by Leckie and Penny.5- Baker and Clines- 30 independently arrived at the identical solu-
tion in developing theory for the axisymmetric bending of general shells of revolution. Their
paper is perhaps the more informative because they have carried out the solution in greater
detail. A valuable discussion is also given in Tsui and Stern. s - 1 8
With the transformation

Q0 = q(¢) f(€)
[5-801

sine

Equations [5-56b] become

1- 3cot 2 (A
si €
d 2o + -1 +i 2 p2 + 3
+
1
- =0
sin € d02 o2 2 4 [2

[5-811

3 1 3 cot 2
Since the term + - - 3 is small compared with 2 p 2 for all values of (,* it
40 2 2 4
can be omitted. The result is

d2q 1 dq q +2i2
d0 2
de ipq =0 [5-82]
+ € db 02

3 2
3 cot o
*When 0 is small, -
4 40
42

~__IYI __1_~_I__YVd~__~~_mll-(i-i-* ~*u*nurq*'~ rrr-- -LI

 -1,N1 IIlYIlluIhIlIIIIY
l

which is identical in form to Equation [5-58] obtained by Geckeler, hence solvable in terms
of the Kelvin functions. The complete solution is

Q = i [A1 ber'(6) + A 2 bei'(6) + A 3 ker'(6) + A 4 kei'(e)]

[5-83]
= -- [C 1 0I'(6) + C 2 02(6) + C 3 03(6) + C 4 04()]
'sin 0

where, as before, e has the meaning p v2". It can be seen that this solution bears similari-
ties to the solutions of both Geckeler and Hete'nyi. In the polar region
- 1, and the
sin 9
solution becomes exactly that of Geckeler. On the other hand, Leckie has found through
asymptotic expansions of the Kelvin functions that the solution is reducible to the simple
h
Hetenyi form for 6 > 6. This means that when R - 0.01 and v = 0.3, for example, € should
exceed 19 deg. For this case we find:

cot 19 deg = 2.904

1
- 3.016

1
so that the approximation cot = - results in an error of only 4 percent. This indicates
that there may be a range in 4 the extent of which depends on() over which both the
Geckeler and the Hete"nyi solutions are reasonably accurate.
For these reasons, the author has not considered it necessary to carry out the more
cumbersome Leckie equations in detail. The Geckeler equations (which are complicated
enough) should suffice up to a value of 6 where the Het6nyi equations can be used. This
is not to downgrade the value of Leckie's solution, for without it there would be difficulty
in finding a connection between the Geckeler and Hetinyi solutions and a consequent range
of possible overlap. Moreover, Tsui and Stern s -18s have found excellent agreement between
the Leckie edge coefficients as given by Baker and Cline and those obtained from the exact
solution of Meissner5- 2 1 over a wide range of . It is thus indicated that the Leckie solu-
tion can serve as a useful substitute for the unwieldy exact solutions of Meissner and Love
wherever a standard of accuracy is needed. Tsui and Stern have also presented an extensive
series of curves for the edge coefficients obtained from the Meissner solution and as given
by Baker and Cline.

u I IL I I II I I I ~a--- ~--rr
 EVALUATION
One way to get an idea of "how approximate" some of these solutions are is to com-
pare them with an exact solution in terms of edge coefficients. Table 5-1 makes such a
comparison for an open-crown hemisphere with the following properti.s:

R = 181 in.
h = 2 in.
A = 10.5 deg
v = 0.3

TABLE 5-1

Edge Coefficients for Spherical Shells Compared

A W
OtA QyA OXA
Eh- Eh - -Eh- -E-
dHA dMA dHA dMA

Geckeler 147.01 54.50 54.50 40.41

Approximation s 'l (-11.0)* (14.0) (14.0) (11.5)

5-14167.14
4 50.18 50.18 37.24
Hete'nyi (1.5) (5.0) (5.0) (3.0)
Geckeler-Esslinger5- 24 163.45 46.62 46.62 35.35
(Polar Region) (-1.0) (-2.0) (-2.0) (-2.5)

Exact (Love)5-4 164.G66 47.64** 47.71** 36.24

R= 181 in. h= 2 in. OA = 10.5 deg v = 0.3

*Numbers in parentheses are percentage deviations from the

exact solution.

**From reciprocity considerations, these two coefficients should

be identical. Presumably the differences arise from small round-off
error.

Values of the coefficients from the Geckeler approximation, 5-11 from Geckeler's solution for
5 19
the polar region,5- 2 4 and from Love's exact solution5 -4 were found in Galletly's paper. -
The Love coefficients were calculat6d using a hypergeometric series representation. It
can be seen that the Geckeler (polar region)5- 24
coefficients are in excellent agreement with
the exact solution. More striking perhaps is the close agreement shown for the Hetenyi co-
efficients since the polar angle OA is relatively small, and it is for small angles that this
solution is least accurate. Even the Geckeler approximation5- 11
gives reasonably good
results, considering the size of the angle.
Such a comparison, of course, does not give a complete picture. A more thorough
examination might be to present plots of the stress distributions for each solution. In
Reference 5-1 Timoshenko has presented a plot of M and N o versus 0 for the case of a
clamped spherical cap under external normal pressure where
 R = 90 in.
h = 3 in.
B = 35 deg
1
6

In this case, the difference between Hetenyi's solution and the exact solution is scarcely dis-
cernible, and the curve representing the Geckeler approximation 5s - 1 1 is sufficiently close for
engineering purposes.
The author was able to find very little in the way of experimental data (again possibly
because of an inadequate literature search) with which to evaluate theory. One source is a
recent Model Basin study by Nishida s5 -3 1 which is devoted primarily to the strength of layered
spherical shells but which does have some data on monolithic shells. Figure 5-11 shows
experimental strains measured on two 90-deg machined aluminum spherical caps (OB = 45 deg)
having clamped edges and subjected to normal pressure on the convex surface. The strains
are plotted for a unit pressure; the theoretical curves are obtained from Hete'nyi's solution.
In both cases, the points tend generally to follow the shape of the curves, widest disagree-
ment being for the thicker of the two shells. However, it would be unwise to make any
sweeping conclusions from such a limited amount of data.

OTHER TYPES OF SHELLS

The preceding sections have dealt with the three types of shells for which solutions
are sufficiently simple for practical engineering applications. Two other shells, the torus
and the spheroid (or ellipsoid), also have relatively simple geometric shapes and are fre-
quently used in pressure vessels and in submarine hulls. However, the analytical solutions
for these shells are far too complex for practical use and will not be discussed here. In
practice, the differential equations are usually solved numerically using high-speed computers.
The best-known solution for the spheroid is due to Naghdi and de Silva, 5 32
that for the torus
33
is due to Clark.5-
Probably the easiest way to get a workable analytical solution for these shells is to
make use of Geckeler's approximation, s - 1 1 and this can be subject to serious errors. It can
yield useful results, however, if used with care, even in cases where the shell geometries
cannot be simply expressed. What this amounts to is that the section of shell to be investi-
gated is replaced by an equivalent cylinder-usually one whose radius is equal to the average
radius of the actual shell. The analysis then proceeds as described in the section on cylin-
drical shells. It takes a fair amount of engineering experience to know when this approxima-
tion can be safely applied, and no attempt will be made here to set down rules of practice.
Other approximations are, of course, possible. For instance, a shell might be re-
placed by an equivalent conical segment. In such a case, one would do well to use the

irl I I I I I I I I_
AIIIIII
iiilimillumilmilllm idillM 10111061 ilifillll
1111111Ifillim l Y IAlfidil,
ilillolYlYIIIIiIII

simplest possible solutions for the segment since the exact conical shell solutions are so
complicated as to defeat the purpose of the approximation.
Where a long continuous shell is to be analyzed, it can be broken up into a series of
segments. Since such segments often will be relatively short, they must usually be con-
sidered as "double-ended," thus requiring the simultaneous determination of all four co-
efficients for each segment. Because this greatly complicates the analysis, it is well to
treat the element in the simplest possible way. A good discussion of such finite element
techniques together with a useful bibliography can be found in Reference 5-34.

STIFFENING RINGS
A stiffening ring such as the one shown in Figure 5-12 can be dealt with in a
simple manner provided certain assumptions can be made, i.e.:

1. The ring is sufficiently short that bending in the meridional plane can be neglected.
2. The cross section is not so deep that thick ring theory must be employed.
With these conditions, the ring can deflect radially and rotate within the plane of its cross
section but does not undergo any change in cross-sectional shape.
The drawing in Figure 5-12 is supposed to represent a typical stiffening ring to which
two shells are joined. RA is the radius to the mean surface of the shell on the left; R is
B
the corresponding radius of the shell to the right. The appropriate forces and moments are
drawn with the same conventions used for shells. The deformations of the ring may be
described as follows:

WA =0 -A XO

WB = WO + B Xo [5-84]

XA XO
= XB

where i~i and X have the usual meanings and the subscript "0" refers to the center of gravity
of the section.
From simple geometrical relations, we have

Ro (A + £B) (RA + B)
0 EA [HA RA - HB RB + P [5-85]
 .0 minim I,,IYYII

Xo = [- HARAPA - HBRBPB -MARA - MBRB +

P (F pB R
Z
[5-85]
pR pR
+ ---
2 (RA - R°) - 2 (RB - R°).

Ar indicates the cross-sectional area of the ring and I z its moment of inertia about the
z-axis, through the center of gravity. The two expressions contain some slight approximations.
The pressure term in W0 comes from taking the lateral pressure to be applied at an average
RA + RB p
radius equal to 2 . In Xo the term - 2 ( RB - RA) was gotten from the approxi-

mation that the pressure to the left of the center of gravity acts at the radius RA and that to
the right acts at RB. In most cases, this term will be so small that it can be neglected
altogether.
From the relations [5-84] and [5-85], the following edge coefficients are obtained:

dWA RoRA ~2 aWB RoRB RoRBB

1RoRA 0 B B0
-- +
dH A EAr EI z OHB, EAr EI z
9-

dWA eARORA WB EBRORB

dMA El z dMB EI z

OWA RORB AeBRo R B OWB Ro0 RA ABR 0RA

-- +
HB EAr EI z 8H A EAr El z

dwA EA RORB OWB eBRORA

[5-86]
M
MB El z dMA El z

WA R0 (B EAr
+ eB) dp 2EA r (RA + RB) (eA + eB)
dp 2Er (RA + RB) (A
2EA

PAO 2 eBR o
2El z [R (RA - Ro) +2EI[RZ (R A Ro)

- R (RB - R o ) + FR, - 2RA] - R~ (RB - Ro) + eRB - ARA'

dXA EARARo dXB dXA RARO dXB

aHA EI Z oH A dMA El z dMA

a r II' 111 I I I I ---~

~h------
------------ ----.-
 dXA EBRBRo dXB dXA RBRo dXB
dHB EI z =HB dMB EIz dMB
[5-86]
dXA R0 dXB
p Ez [R (RA - R)-B (R B RB -ARA]
dp 2ElZ O

Sometimes use is made of flexible tapered rings like that shown in Figure 5-13 to
reduce stresses at cone-cylinder junctures. Because of their flexibility, these rings must be
treated as tapered shells and their behavior cannot be adequately described by the simple
Equations [5-86]. For a detailed treatment of this case, the reader is referred to an analysis
5 5
by Raetz. -3

COMPOSITE SHELLS
The complete analysis of a composite shell structure such as a submarine hull is a
formidable problem because of the number of junctures involved. Since each discontinuity
must be treated as a juncture, an attempt to analyze the structure all at once could require
the solution of dozens of simultaneous equations-a situation which taxes a computer's
storage capacity and can lead to such large numerical errors that results are sometimes
meaningless. By following such a course, one may find that conditions at Frame 1 have,
numerically, a significant effect on the deformations, say, at Frame 75. There are many
ways around this. One is to use some engineering judgment to limit the number of junctures
to be considered at one time. On either side of a deep frame, for example, it might not be
necessary to carry the analysis beyond three small frame spacings. That is, the third frame
is so little affected by the presence of the deep frame that it deforms essentially like a
typical frame in a uniformly stiffened shell of infinite length. Errors can also be reduced by
a judicious choice of the order in which the equations are solved. The number of equations
to be solved in any structure depends on the number of consecutive "double-ended" elements.
Fortunately only a few unknowns appear in any one of the equations. Consider, for example,
a simple hull configuration containing N junctures in which only two elements meet at each
interior juncture. If we want to eliminate i and X as unknowns so that H and M are to be
found, the system of equations might look something like this: For the nth juncture
(represented by Figure 5-14),
 II
1 IIIIIIII

d nB dinB
0= Hn HnA +n-1 M
1 nA

FdWnB W(n+ 1)A 35nB 02(n+1)A [5-87]

+
n
Hn rH'i B H (n + 1) A I n OMnB OM(n + 1) A

F (n+ 1A ( +l 1
-H+ -Hnl
L 8H n-lMn+l
&(n+)BI B Mn+ 1 L'(n+ 1 ! I +1)B
-- ( +n

[&XnB 1FXnB
0 Hn- 1 L nA n-A1 nA

[ X n
B
3
X(n+1)A M XnB dX(n
+
1)A

n HnB (n + 1) A n LMnB -M(n + 1) A

SX(n + 1) E X(n+ 1)

-Hn+ 1 H(n + I)B 1 (n+ fnB

Here, it has been assumed that the only axial forces are due to pressure; the terms a np and
SnP therefore represent known quantities. The coefficients of the H's and M's can be repre-
sented by a matrix that has a good many zeroes and is banded about the main diagonal:

M1

H2
( ) () ( ) 2

H
( n
() 4) [5-88]

HNi

HN
()() M

Lra~- I I I I I I
Each ( ) indicates a small submatrix of the coefficients appearing in the pair of equations
for each juncture. The band is three elements wide because the equations contain three
unknown H's and three unknown M's at each interior juncture. This is not presented as
the best scheme for solving the equations, but only as an example. It is worth noting, how-
ever, that certain computer routines, are especially well adapted to the handling of banded
matrices of this type.
In addition, some computational effort can be avoided if one remembers that for the
linear systems considered here, the theorem of reciprocity ensures that

donA WnB RB) dXnA WnB ( B)

HnB 8HnA R/A aHnB MfnA A

dWnA XnA dWnB aXnB

dMnA HnA [5-89]
dnB -HnB
nA n

aiWnA dXnB RB aXnA dXnB ( RB)

(MnB dHnA RA dMnB nA A

One notable achievement in the analysis of a complete submarine pressure hull is a

computer program recently developed at the Naval Construction Research Establishment in
Scotland. This program, described in References 5-36, 5-37, and 5-38, is capable of deter-
mining stresses and deformations at every juncture of interest in a typical pressure hull as
well as stress distributions in areas selected by the programmer. The program can handle
virtually any type of shell; the only real limitations are that the number of junctures and the
number of elements that can be entered as input are restricted by the capacity of the
computer. Recent comparisons with independent calculations done at the Model Basin
showed excellent correlation.

ACKNOWLEDGMENTS
The author is greatly indebted to Mr. R.V. Raetz and to Mr. O. Lomacky for their
valuable suggestions and advice.
 Figure 5-1 - Composite of Typical Shells Used in Submarine Hulls

2A

H2 A
B)
X'B

R1
I
1A
XA Deformations
With Sign Conventions
....-. Axis of Revolution-- --

Figure 5-2 - Loads and Deformations in Intersecting Shells

HpR

-2 A L X
pR
MB 2

nA 5-3
Figure
Figure 5-3 -- Cylindrical Shell with
Cylindrical Shell
Edge Loads

SNx + dN
x

PRA
2 -
w MB

H
A

Figure 5-4 - Conical Shell with Edge Loads

1 I III- 1 1 1 111 1 I I ~IC------ 13~~

I------~ll~1 1~)
~
 i
49i111
n li~ ui
IgilYIIY

Figure 5-5 - Taylor-Wenk Functions for Conical Shells

L07 FOR LARGE VALU

S1.
C-0
Lo6 4 .1+

L05 - I+ VT 3.

L041
0.1+--
Lo4 "- 0.84853

L03
Z 6
C1 fil

Figure 5-5a - Edge A (10 < e < 40)

Figure 5-5b - Edge B (10 < C 25)

70 8 90 100 110 120 130 140 150 160 170

Figure 5-5c - Edge B (20 < . < 170)

I I IL I L I Il r --
 _ _ I I _

Figure 5-6 - Raetz-Pulos Functions

Xli~
lllils~-
l( -~- llll(ms~-----
*n ru~axn~ --- --- *
 Figure 5-7 - Comparison of Raetz-Pulos Solution with Strain Data for Cone-Cylinder Model

6
Stiffening Ring
- Theoretical Strains
Experimental Strains
2- o 90 degree Generator, External
0 90 degree Generator, Internal 10
* 180 degree Generator,External
M 180degree Generator,Internal

48

00

00
C
02-

0
*o

Large Cylinder: R 13.5", h 0.127" Cone: h- 0.110" W Small Cylinder: R 8", h,0.092"
a * 60 degrees
Distance Along Shell

Figure 5-7a - Circumferential Strains

 Theoretical Strains Stiffening Ring
- External
-- Internal
24 - Experimental Strains
o 90 degree Generator, External
I 90 degree Generator, Internal
* 180 degree Generator, External
0 M 180 degree Generator, Internal
16
C

zoo °I -

ID I
'l I
-16 I I
\l
E
8-24

1
I1
I

-40
Large Cylinder: Rz 13.5", h0.127" Cone: h0O.110" i'* Small Cylinder: R 8' h=0.092"
a 60 degrees
Distance Along Shell

Figure 5-7b - Longitudinal Strains

 _

Li

Figure 5-8 - Spherical Shell Element

Figure 5-9a - Shallow Cap T

Figure 5-9a - Shallow Cap Figure 5-9b - Complement

Figure 5-9c - Shallow Segment

Figure 5-9 - Spherical Shells Bounded by Small Polar Angles

j
I
I
-I
Figure 5-10 - Spherical Shells Bounded by Large Polar AngleF

-- I I II I I Ill -_ _--~-- -------

b~l^--i -d-~**~*rr
 1111-~~" 1 11 -- ~- - -1~~- --- I

Figure 5-11 - Experimental Strains for Pressurized Clamped Spherical Cap Compared
with Hetenyi's Solution

3.0
OUTSIDE THEORETICAL

2.5 0

0/
INSIDE THEORETICAL

2.0
0 0 OUTSIDE
O INSIDE
-3.
CIRCUMFERENTIAL STRAINS

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
9kB - 0
OB
1.5

-1.0

0.50 INSIDE THEORETICAL

-3.5
-3.0
-2.5

-2.0

-1.5 O OIONL OUTSIDE THEORETICAL

OJe _M DOUTSIDE
0 INSIDE
-1.0
MERIDIONAL STRAINS

-0.5

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
OB 0

Figure 5.11a - h/R = 0.0149, OB = 45 Degrees, v = 0.3

3-crulru*ncuunr~ ~rl ulxn~~

11111141mmom 1

-1.4 I I I I
INSIDE THEORETICAL

-1.2

-1.0

ZOUTSIDE THEORETICAL
-0.8
* OUTSIDE
0 INSIDE
-0.6
S CIRCUMFERENTIAL STRAINS
g sI
-0.4 C

-0.2

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

B -

OB
-1.8

-1.6

1.4 OUTSIDE THEORETICAL

-1.4. gO ,, . .

-1.2

-1.0 -

, INSIDE THEORETICAL
-0.8-0.8

-0.6
SOUTSIDE
O INSIDE
-0.4 MERIDIONAL STRAINS

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
OB - 1
FB
Figure 5-11b - h/R = 0.0296, 9B = 45 Degrees, v = 0.3

I I I I I I
 r4- A --
VA, l'* VB
B

Figure 5-12 - Stiffening Ring with Edge Loads

TAPERED CYLINDRICAL
RING SHELL
CONICAL
TRANSITION
SECTION/

REINFORCING
FRAME

Figure 5-13 - Cross Section of Cone-Cylinder

Juncture with Tapered Ring

n+1

'Hn

Juncture n

SMn-1
Hn-1

Figure 5-14 - nth Juncture in a Chain of Connected Shells

.. 111111
1,16
i *0111m
ol Mi

REFERENCES

5- 1. Timoshenko, S. and Woinowsky-Krieger, S., "Theory of Plates and Shells,"

McGraw-Hill Book Company, Inc., New York (1959).
5- 2. Flugge, W., "Stresses in Shells,"' Springer-Verlag, Berlin (1962).
5- 3. Kraus, H., "A Review and Evaluation of Computer Programs for the Analysis of
Stresses in Pressure Vessels," Welding Research Council Bulletin No. 108 (Sep 1965).
5- 4. Love, A.E.H., "A Treatise on the Mathematical Theory of Elasticity," Fourth
Edition, Cambridge University Press, London (1927).
5- 5. Taylor, C.E. and Wenk, E., Jr., "Analysis of Stress in the Conical Elements of
Shell Structures," Proceedings of the Second U.S. National Congreds of Applied Mechanics,
pp. 323-331 (May 1956); also David Taylor Model Basin Report 981 (May 1956).
5- 6. Dubois, F., "Uber die Festigkeit der Kegelschale," Doctoral Dissertation,
Zurich (1917).
5- 7. Schleicher, F., "Kreisplatten auf Elasticher Unterlage," Berlin (1926).
5- 8. Watts, G.W. and Burrows, W.R., "The Basic Elastic Theory of Vessel Heads under
Internal Pressure," Journal of Applied Mechanics, Vol. 16, pp. 55-73 (1949).

5- 9. Baltrukonis, J.H., "Influence Coefficients for Edge-Loaded Short, Thin, Conical

Frustrums," Journal of Applied Mechanics, Vol. 26, pp. 241-245 (1959).
5-10. Wenk, E., Jr. and Taylor, C.E., "Analysis of Stresses at the Reinforced Intersec-
tion of Conical and Cylindrical Shells," David Taylor Model Basin Report 826 (Mar 1953).
-0

5-11. Geckeler, J.W., "Uber die Festigkeit achsenymmetrischer Schalen," Forschungs-

arbeiten auf dem Gebiete des Ingenieurwesens, No. 276, pp. 1-52, Berlin (1926).
5-12. Raetz, R.V. and Pulos, J.G., "A Procedure for Computing Stresses in a Conical
Shell near Ring Stiffeners or Reinforced Intersections," David Taylor Model Basin Report
1015 (Apr 1958).

5-13. Linkous, C. and Horvay, G., "Analysis of Short Cylindrical and Conical Shell
Sections," Knolls Atomic Power Laboratory Report No. KAPL-912 (Apr 1953).

5-14. Hetenyi, M., "Spherical Shells Subjected to Axial Symmetrical Bending," Publica-
tion 5, International Association for Bridge and Structural Engineering, Zu'rich (1938).

5-15. Borg, M.F., "An Investigation of the Stresses and Strains Near Intersections of
Conical and Cylindrical Shells,"' David Taylor Model Basin Report 911 (Feb 1956).
5-16. Krenzke, M.A., "Hydrostatic Tests of Conical Reducers between Cylinders with
and without Stiffeners at the Cone-Cylinder Junctures," David Taylor Model Basin Report
1187 (Feb 1959).

I I I I II L ~-~-~--a_~
 __ __~_~

5-17. Raetz, R.V., "An Experimental Investigation of the Strength of Small-Scale Conical
Reducer Sections between Cylindrical Shells under External Hydrostatic Pressure," David
Taylor Model Basin Report 1397 (Mar 1960).
5-18. Tsui, E.Y.W. and Stern, P., "A Critical Survey of the Metliods for Analysis of
Spherical Shells," Lockheed Missiles and Space Company Technical Report 6-90-63-41
(Aug 1963).

5-19. Galletly, G.D., "Influence Coefficients for Hemispherical Shells with Small
Openings at the Vertex," Journal of Applied Mechanics, Vol. 22, pp. 20-24 (1955); also
David Taylor Model Basin Report 870 (May 1956).

5-20. Reissner, H., "Spannungen in Kugelschalen," Muller-Breslau-Festschrift, p. 181,

Leipzig (1912).

5-21. Meissner, E., "Das Elastizit'aitsproblem fur dunne Schalen von Ringflachen-,
Kugeloder Kegelform," Physik, Z. 14, pp. 343-349 (1913).

5-22. Bolle, L., "Festigkeitsberechnung von Kugelschalen," Dissertation Zurich (1916).

5-23. Galletly, G.D., Discussion of Reference 5-19, Journal of Applied Mechanics,
Vol. 22 (1955).

5-24. Geckeler, J.W., "Zur Theorie der Elastizititflacher rotations-symmetrischer

Schalen,"' Inginieur-Archiv, pp. 255-270, Berlin (1930).

5-25. Esslinger, M., "Statische Berechnung von kesselbden," Julius Springer, pp. 1-20,
Berlin (1952).

5-26. FlIigge, W., "Four Place Tables of Transcendental Functions," Pergamon Press,
New York (1954).

5-27. Jahnke, E. and Emde, F., "Tables of Functions," Dover Publications, Inc.,
New York (1945).

5-28. Leckie, F.A., "Localized Loads Applied to Spherical Shells," Journal Mechanical
Engineering Science, Vol. 3, No. 2, pp. 111-118 (1961).

5-29. Leckie, F.A. and Penny, R.K., "A Critical Study of the Solutions for the Asymmetric
Bending of Spherical Shells,"' Proceedings, World Conference on Shell Structures,
San Francisco (1962); also Welding Research Council Bulletin No. 90 (1963).
5-30. Baker, B.R. and Cline, G.B., Jr., "Influence Coefficients for Thin Smooth Shells
of Revolution Subjected to Symmetric Edge Loads," Journal of Applied Mechanics, Vol. 29,
Series 3, pp. 335-339 (1962).

5-31. Nishida, K., "Tests of Machined Multilayer Spherical Shells with Clamped
Boundaries under External Hydrostatic Pressure," David Taylor Model Basin Report 2012
(Aug 1965).
-WI , II 11li 11hh
liylj 111
il d i . '

5-32. Naghdi, P.M. and de Silva, C.N., "Deformation of Elastic Ellipsoidal Shells of
Revolution," Proceedings of the Second U.S. National Congress of Applied Mechanics,
ASME p. 333 (1955).

5-33. Clark, R.A., "On the Theory of Thin Elastic Toroidal Shells," Journal of
Mathematics and Physics, Vol. 29, No. 3, pp. 146-178 (1950).

5-34. Zienkiewicz, O.C. and Holister, G.S., "Stress Analysis," John Wiley and Sons,
Ltd., London (1965).

Raetz, R.V., "Analysis of Stresses at Junctures of Axisymmetric Shells with

5-35.
Flexible Insert Rings of Linearly Varying Thickness," David Taylor Model Basin Report
1444 (Jan 1961).

Kendrick, S., "The Deformation of Axisymmetric Shells," Naval Construction

5-36.
Research Establishment Report R.457 (Jul 1961).

5-37. Kendrick, S. and McKeeman, J.L., "Pegasus Computer Specifications-

Axisymmetric Stress Analysis," Naval Construction Research Establishment Report R.452
(Mar 1961).

5-38. Kendrick, S. and McKeeman, J.L., "Pegasus Computer Specifications-

Axisymmetric Stress Analysis - Part II," Naval Construction Research Establishment
Report R.477 (May 1963).

~ I II I I I I ------ -,,------..k.----.nn~nbonrp-
 INITIAL DISTRIBUTION
Copies Copies
3 NAVMAT 2 NAVSHIPYD PTSMH
1 Lab Mgt Div (Code 0331)
2 Special Projects Office (Code 001) 2 CO, USNROTC & NAVADMINU MIT
1 Dept NAME, Dr. A.H. Keil
6 NAVSHIPSYSCOM
2 Tech Info Br (Code 2021) 1 CO, PGSCHOL, Webb
1 Chief Sci for R & D (Code 031)
2 Struc & Ship Protec Sec (Code 03423) 1 CO, PGSCHOL, Monterey
1 Sub Br (Code 525)
20 DDC
7 NAVSEC
1 Prelim Des Br (Code 6420) 1 DIR, APL, Univ of Washington, Seattle
1 Prelim Des Sec (Code 6421)
1 Ship Protec Sec (Code 6423) 1 NAS, Attn: Common Undersea Warfare
1 Hull Des Br (Code 6440)
2 Sci & Res Sec (Code 6442) 1 Prof. M. Hete"nyi
1 Hull Struc Sec (Code 6443) Stanford University, Stanford, California

3 CHONR 1 Dr. E. Wenk, Jr.

1 Res Coord (Code 104) Lib of Congress
1 Struc Mech Br (Code 439)
1 Undersea Prog (Code 466) 1. Dr. R. DeHart, SWRI

4 CNO 1 Prof. J. Kempner

1 Sub Warfare Div (Code 0P31) Polytechnic Institute of Brooklyn
1 Ship Charac Div (Code 0OP366)
1 Tech Anal & Adv Gr (Code OPO7T)
1 Underseas Warfare Dev Div (Code OP71)

1 CO, USNOL

1 CO, USNRL

1 CO & DIR, USNMEL

1 CO, USNEL

1 CDR, USNOTS, China Lake

1 CDR, USNOTS, Pasadena (P-8082)

1 CO & DIR, USNUSL

1 CO, USNUOS

2 NAVSHIPYD SFRAN BAY

**~~__/t~_~~l_____L~_
.911 , , , , , ., 1,
1,1,
, , Iii
iwi11AININI'l
iIHIM
INNIN
ii
I,limillilo
blifill'Ild
pul Al11111
Al
I.,I.Willm
1111110i
W,
11,
1
 111111mill
i
111,11111110
ki,61IM
.,~I~ M m II

UNCLASSIFIED
Security Classification
DOCUMENT CONTROL DATA - R&D
(Security classification of title, body of abstract and indexing annotation must be entered when the overall report is classified)
1 ORIGINATING ACTIVITY (Corporate author) 2a. REPORT SECURITY CLASSIFICATION

David Taylor Model Basin UNCLASSIFIED

Washington, D.C. 20007 2 b GROUP

3 REPORT TITLE

A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. PART I - CONVENTIONAL HULL

CONFIGURATIONS, CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF REVOLUTION
4 DESCRIPTIVE NOTES (Type of report and inclusive dates)

5 AUTHOR(S) (Last name, first name, initial)

Reynolds, Thomas E.

6. REPORT DATE 7a TOTAL NO. OF PAGES 7b. NO OF REFS

December 1966 74 38
8a CONTRACT OR GRANT NO. 9a ORIGINATOR'S REPORT NUMBER(S)

b. PROJECT No. S-F013 03 02 2264

Task 1961
c. 9b. OTHER REPORT No(S) (Any other numbers that may be assigned
this report)

d
10. AVAIL ABILITY/LIMITATION NOTICES

Distribution of this report is unlimited.

11. SUPPLEMENTARY NOTES 12 SPONSORING MILITARY ACTIVITY

Naval Ships Systems Command
Department of the Navy

13 ABSTRACT
Analytical methods utilizing the linear bending theory of thin
shells are presented which are judged the most useful for determining
stresses in submarine pressure hulls. Only axisymmetric deformations
are considered, and emphasis is placed on methods that are sufficiently
simple to be applied at the practical level. Attention is restricted to
analytical solutions for cylindrical, conical, and spherical shells.
Efforts have been made to employ a uniform system of notation and
sign conventions wherever possible.

DD 1 JAN 1473 UNCLASSIFIED

Security Classification
 UNCLASSIFIED
Security Classification
14.
KEY WORDS
Thin Shells
Stress Analysis
Cylindrical Shells
Conical Shells
Spherical Shells
Stiffening Rings
INSTRUCTIONS
1. ORIGINATING ACTIVITY: Enter the name and address
of the contractor, subcontractor, grantee, Department of De-
fense activity or other organization (corporate author) issuing
the report.
2a. REPORT SECUITY CLASSIFICATION; Enter the over-
all security classification of the report. Indicate whether
"Restricted Data" is included. Marking is to be in accord-
ance with appropriate security regulations.
2b. GROUP: Automatic downgrading is specified in DoD Di-
rective 5200. 10 and Armed Forces Industrial Manual. Enter
the group number. Also, when applicable, show that optional
markings have been used for Group 3 and Group 4 as author-
ized.
3. REPORT TITLE: Enter the complete report title in all
capital letters. Titles in all cases should be unclassified.
If a meaningful title cannot be selected without classifica-
tion, show title classification in all capitals in parenthesis
immediately following the title.
4. DESCRIPTIVE NOTES If appropriate, enter the type of
report, e.g., interim, progress, summary, annual, or final.
Give the inclusive dates when a specific reporting period is
covered.
5. AUTHOR(S): Enter the name(s) of author(s) as shown on
or in the report. Enter last name, first name, middle initial.
If military, show rank and branch of service. The name of
the principal author is an absolute minimum requirement.
6. REPORT DATE Enter the date of the report as day,
month, year; or month, year. If more than one date appears
on the report, use date of publication.
7a. TOTAL NUMBER OF PAGES: The total page count
should follow normal pagination procedures, i.e., enter the
number of pages containing information.
7b. NUMBER OF REFERENCES Enter the total number of
references cited in the report.
8a. CONTRACT OR GRANT NUMBER: If appropriate, enter
the applicable number of the contract or grant under which
the report was written.
8b, Sc, & 8d. PROJECT NUMBER: Enter the appropriate
military department identification, such as project number,
subproject number, system numbers, task number, etc.
9a. ORIGINATOR'S REPORT NUMBER(S): Enter the offi-
cial report number by which the document will be identified
and controlled by the originating activity. This number must
be unique to this report.
9b. OTHER REPORT NUMBER(S): If the report has been
assigned any other report numbers (either by the originator
or by the sponsor), also enter this number(s).
10. AVAILABILITY/LIMITATION NOTICES: Enter any lim-
itations on further dissemination of the report, other than those
LINK A LINK B LINK C
ROLE WT ROLE WT ROLE WT
imposed by security classification, using standard statements
such as:
(1) "Qualified requesters may obtain copies of this
report from DDC."
(2) "Foreign announcement and dissemination of this
report by DDC is not authorized."
(3) "U. S. Government agencies may obtain copies of
this report directly from DDC. Other qualified DDC
users shall request through
(4) "U. S. military agencies may obtain copies of this
report directly from DDC. Other qualified users
shall request through
(5) "All distribution of this report is controlled. Qual-
ified DDC users shall request through
If the report has been furnished to the Office of Technical
Services, Department of Commerce, for sale to the public, indi-
cate this fact and enter the price, if known.
1L SUPPLEMENTARY NOTES: Use for additional explana-
tory notes.
12. SPONSORING MILITARY ACTIVITY: Enter the name of
the departmental project office or laboratory sponsoring (pay-
ing for) the research and development. Include address.
13. ABSTRACT: Enter an abstract giving a brief and factual
summary of the document indicative of the report, even though
it may also appear elsewhere in the body of the technical re-
port. If additional space is required, a continuation sheet shall
be attached.
It is highly desirable that the abstract of classified reports
be unclassified. Each paragraph of the abstract shall end with
an indication of the military security classification of the in-
formation in the paragraph, represented as (TS), (S), (C), or (U).
There is no limitation on the length of the abstract. How-
ever, the suggested length is from 150 to 225 words.
14. KEY WORDS: Key words are technically meaningful terms
or short phrases that characterize a report and may be used as
index entries for cataloging the report. Key words must be
selected so that no security classification is required. Identi-
fiers, such as equipment model designation, trade name, military
project code name, geographic location, may be used as key
words but will be followed by an indication of technical con-
text. The assignment of links, roles, and weights is optional.
UNCLASSIFIED
Security Classification
- - I- ---- , , I, , i , 11, ,,11,1111W
Ail'Aliwillildlil
1,1111
1110141111illwil
lllillul 0 = 1IM
11111111111
I'li
IW1111W
Ihilli mikilijil
111,11
, 11111i
David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vil, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E.
II. S-F013 03 02
Task 1961
~-i
David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E.
II. S-F013 03 02
Task 1961
David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vil, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E.
II. S-F013 03 02
Task 1961
David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E.
II. S-F013 03 02
Task 1961
 t

David Taylor Model Basin. Report 2264. David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design-- A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses-- CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)-- illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen- Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)-- mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened) that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses-- conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E. possible. I. Reynolds, Thomas E.
II. S-F013 03 02 II. S-F013 03 02
Task 1961 Task 1961

David Taylor Model Basin. Report 2264. David Taylor Model Basin. Report 2264.
A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design-- A SUMMARY OF SUBMARINE STRUCTURAL RESEARCH. 1. Submarines--Design--
PART I - CONVENTIONAL HULL CONFIGURATIONS, Research PART I - CONVENTIONAL HULL CONFIGURATIONS, Research
CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses-- CHAPTER 5: STRESSES IN PRESSURIZED SHELLS OF 2. Submarines--Stresses--
REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research REVOLUTION, by Thomas E. Reynolds. Dec 1966. vii, 67p. Research
illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)-- illus., graphs, tables, diags., refs. UNCLASSIFIED 3. Thin shells (Stiffened)--
Stresses--Research Stresses--Research
Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen- Analytical methods utilizing the linear bending theory of thin 4. Cylindrical shells (Stiffen-
shells are presented which are judged the most useful for deter- ed)--Stresses--Research shells are presented which are judged the most useful for deter- ed)--Stresses--Research
mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)-- mining stresses in submarine pressure hulls. Only axisymmetric 5. Conical shells (Stiffened)--
deformations are considered, and emphasis is placed on methods Stresses--Research deformations are considered, and emphasis is placed on methods Stresses--Research
that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened) that are sufficiently simple to be applied at the practical level. 6. Spherical shells (Stiffened)
Attention is restricted to analytical solutions for cylindrical, --Stresses--Research Attention is restricted to analytical solutions for cylindrical, --Stresses--Research
conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses-- conical, and spherical shells. Efforts have been made to employ 7. Ring stiffeners--Stresses--
a uniform system of notation and sign conventions wherever Research a uniform system of notation and sign conventions wherever Research
possible. I. Reynolds, Thomas E. possible. I. Reynolds, Thomas E.
II. S-F013 03 02 II. S-F013 03 02
Task 1961 Task 1961
 Date Due
Ti

Lib-26-67
 H. t _.,. , L , - I. , , 7 ,-',, , . , t -. - ', :
_ _.- ; - _,_ , " ___. _
.1 - - - - I-_ - - .7-_ , - I - . I - I- . I- - - -. - - , , . , - - - - - - - -_ ,- . - - - - - - -- - - - -_ -,- - ,_ q -
' . . - I - - - - . - - . ,. - - -
I ,je . . I I : . I I -
,I---,7 1 - . . I I .; - I I .-, . _l I

- I- 1, . I - - I_ I -I I I . . I-
.I I II . - , I . " . I.
..,. _-.. I Iq
'11 . I . I. .
II .. .
I ,
I. -1 ... I.I I . I I I I - I I . . . I
. I .
r ,. I . 1. -. -I . . -, , 111. , . . - - - I I . . . . I II I - II - - I .- . I _.,, - - .
I -.. . I 1. . % I 11 I. I- .I I. I- I I ,I . - I .I -- .1 -
(1 .: ,, , I , . I I - - I .. I I '. I I I I I - I I I I II
I . ,e "
I . . . - I . .1 I. . - I . -1
.. 11 I
I- . - I , I I . 11 .. , .: - , .a
., I .I I- . . . I-
:
II I . 11 -
I- I . 11 I I I _. . I I I I I . I . : . . '.',,', 7.. . -
. I I .- . .I .; .. I - I -. . I . I . I. :, i t . -. ..., I
11 I . . I :. I - .. :. . -:
.1 - 11, I - . I I .. - I I .
. . 1, I 11 I I - . ..: I, . . . I- - . I - . .- I. I . " - .I . ,. I I . .- __
I I. -'I w.,
11- I I . _. . I. . I- ._ . I " .. ,. I ..1.
-
I .- _ . - II I.. . . . -' 'I I I I I . _ .. . I I11 -1 I .
I.. I - . I-1
., 1.,,. I I -. . , :.,,-.
. , I I I -:. , ,I - I -1 I - . 11 . I. - .- I. I
I I . . . .. I . .. I I
i .. .. I : . ,. . ,
-I . I -. . I , 11 I- . I4 I II . I- I I I_-, :. . . I I ''
. 1. I- II . I., I . ,I , .., I 1. I II - _. . "I , I I- - I I I''I I,It .-
..
:. . - I I " . II I I II I I .
.. I. I I II I
I
I . .
I I
I ,.. -' ' . . I- I - .- - .
I I- I . %- .. - I. .. . . : I. .I .! , . I . I I . . .. I . I - . I .I I
; I I ., . . , . . . I - . - I. ,-I ! IT I . ,
I , r I .,, I . I- - r I I I - ._ . I -_ I I 11 I I . . 'lk . . - . .. . I
. - . . I . I ,, . I - , . I
II. .I II -.I . I!,: . -- I I , I I. I . I I. 1.,. .. .-- . .1 II . I. :1I. , . . 11 I
, . . . .I i .I . I I I.-
. - I I . ,- . I - ,- I . . 11 .I . I , . I .1... I I . I
i I I .I I - I.. I .I . I , I I . , I , - . I I . I 1. .
I I 11 I I. "., , . - I 1, : .. I . II . I I. I .1. . . . I ., I
I . - . I .- I I I
I I . . I. 1. . . . ; I .. I - .1I . I . - . I . . .I , . . I I I . .. . .
I . . .1. . .: I i . ,, , .. I I . .. .I , - ."
. .. I I I I . . . .. I I - I- . II I . I ,. l I - . I I. I, . I . . I :- I ,, -. I I
I I.1 I-, - I I I II I I . . I - I. .
. .. I- I . ,. 11I o I I . . . I I . I I ,
. . I - I -I , I- -. . I I . ; .1 . . . . 11 .. I - I I t
: - - .. . -. I . I I I v . - .
__ , q _1 I ; , I I i . I I - ,- . . I I
- I .. I I I
.. I I , I- . . , ) - , I -- , . :.I. - - , I I_. . . .1 - . .- , - . . I I , . I . I. I , I I . I .
,, : , f I . .t _, . , - - - - I . - I .. , I . . . I . . I . . -- . , . I . . - . - I I .. I ; I " I - .
I ;, I F I I . I- . ,.. I_.
_.. I I - .1 1, I. .1 --
- I . . , ,I : , 4. ., -:. - .I - I., I . -:, , , , ., Ii-, - - I. 7 - ., .1 . . I.. .. . : I I .- I . I- " . II I
", .A . :,, 11 - - I
, , .. _. I I .- I . - , , II. I I. .1 -
.
- , I , . . iI
.
_-, , - I . II I. I I . I.: :II- I. I "II , ,. I.I-: '. I; '... , : I. I. , .1 ,1, , -1 . %- , , . ,
, . I. I Y . I . . I
11 L 11".."
. 1
, , -I I .:II - I I I"',, . : I, ... ,I. 1:I . ,." , - II I.,- ,. ,_ , . - " , :, , I , . . , , " I I : i
. - . . - . I
.
11 .. . - 1. . I .
I . I , _ : IJ I q ", - 11 - , -: , , .- - .1 .1 . I .- - . ., I .1 . II
" , I II . . - I . .- . . L I ., .I . -_ - , , - I ., - . . I -1 ,_ . - I I . I. I I I I II
."I.: , ,-,- I . - . I , . - , _:,
_, .'. , . , I., I. I I --,., -I -1,
. - ,r '', - ,- 1, " o. _. , ., I_ .- I ,I -, _. , ', .
, ., I .. .-
., - .
I . - 1: I
I . _Fl . I . , .., , , , ;l . . . -. , " , ,,,,
I " , _ , i .
'- . _':
. . .
_*.,,.
_:.
, I .., : I- - : I
, _ . . - I I .- I
. .1
I.. I .
I - :11- I .. I-- . . I - " ., , I . , r',
. .. : - , '. . -4, .., ,. -, . I . 11 . I, . I ., 11 . I -, d I-_ , :" " . ., . .. .,.,z :. . . I ..- I - I .- . I.
I - ; . , . -.- ..-1 1. : , ;.. . j, , ,-,)I .., ;I, r .,, 't,- , ." 7,.- I - .1. , ,- , ,- -, , I. - . - - I. I
. I : ,., . - . . .. - - . - I - . .", . 1. . .;-_I ., r- I . - I , I I . .. I....
I .- : ..1, ., ,_, . :1.- I. - ;: .. ,I, , . . - - , , - ;, , II-- . -- ,, I- - -1,
*1 - .- - - . .. ,II:: - ,- -- ;, , . " , ,- ; ". . - ,. ! , . I , .., I. , I _. -,,., - ,1, ., -I.,- ,. . I" -,, ,v,.v , , ., . ,- . . , I . I I I I - - I ..
. , I . . , .-
I I . . I ! __ b, I . - , .',, ,. -- , . , , II . '.I -I I 'L-, , -- - , " _ . t .I , , ,: , ;, , , , . . " 4. , :, III . :, I - , I . I.
.. " , , .,( I '. , . - I - .-I.,-, .1;_', ", , , '. , . i _.-.I ,I -,, ,- , 1. . ':, II . , . . - , , - - I , - I -: I .
-II, [" I- . . .. 7 , I. ";. - ,,. , . ,.".
I -- , I - - . .
,,- .. I .- I.- I II L;' I I - , , - : ,I: . , , , . I '* , V - - I..II ., 111-... I,. . _ . , - -,;_-,-., . - , . . . - p. I, rl , . , , ,
i . - - .
-. . : . I - : . ..
I - 1. . : - . I , "; . , ,. .1 I - J!, I " - -- - ",_ ,, I";
I ; - r . .1, , I. I ; , -I *" , 3 . - . , - " _. , ..- . , . I
. Z, ,, ,t, ., -,, " . - - . , - . _'
I I I . I I - , t . . , - , F .1..j ,II -,, , , ,I - , , ... r- % q 14, - -, .,z , _ _; . - -- _ , i ,fi , , I ., . , , I.
-I
. II . I . , . , b .' ..I.- I. j I" _ . .4 , .4" J ._.,". IVP - , 1", , x 1 I- .. , ': 1", ,. F, (1.,. - - , , ;- , ,., , -. . I - . , .,:. .
. I . I . . ,. - I .,. . : I I. I- , .,-I".-). . ,_I- - . ,_ ,I; ., " , ,,.,,-,,i fI ]- , L .. I - V T , v. IP., :,., - , ., . , ,, , , Z, .- . 11 - ".. I , .1 .- . , .. I I
. 4. . I - . , . , j ') ,f' ' , , "," T -1 I ,.-
I
.., - , I I .. ,, , , ...". , , -- 7 ,! . ... , . -,,- . * It , ,"N", , I _. ,
, ', - ,-i11 I- ,! , 4 - .111,
,qH_;_, , ,, - :,;, . . , "J" , ,7'. , -, : ..-: I1.,, - ; I,,-- ,.M.'- ; -,_- , , - . .1 I I I
. I. -
I -, - . -. ,: - i --,.-, - ., -,,.."!). , . I-, I .,,'. -,-
-- ,.*,
. ,7'I, : j, ,, --i-, , ",- ,,I,-, 11 ,, ' .1-, . " -1, i 0 11__Y l ,;L:7,Z. I--__ , ) "I- __ ,;. 1.
- I I I," .1 . __. I . . .I I , I , . I I , .
- .. . - I - JI , - .- - , - -i ._ -,, , - ... . ,.:;.", L, ,_ v - - .., j , j , , " 1),[ , . ", - j. ,,- ,_ ," , ..... . . - 1 I,-
-'I' I-_ ., - I I I . I I -. -
, , . : -- , . ,, - ,:- , . ;... . .1, ..I, - _ ,,,I,'?-F I V ,I . - ,I -". - t .,-3','. " I -1- .,
, . ill", .
, . 1. 7, _. 'i - ,N, I I k-L - I
. ..,,- :11
L '
. "',*,I. - r I. r .* , w I- , , , . .. . , t "
1, .- , , I I 11:, , ,I, -1 II " ' : - ., , -- ,,, - -. 4 ): , , '._. .,.:I,__, , ; i r," , ,.I.., . ,"I- . II" . V 1)P t."4' - - q,_kl . , - ;, , xr ,. ,..'" I- ,, , ,;,-,, .), , ,- ,-..- - - ..-. ') I, . , ,I-, :l I . _. . . I, . I i I I ,,I
ul , ,
- f
, . , .- I- ,-_ F- ,-',._' ,- ' ,I11v ,,.-.L'-' : ,.."'O-j ,'-1_1
". ,_, --4, -- f. .-. it, , Y .,, ', .4 I',,,,,,,I .. - ,- , ,, I. .. -.I " ,I , , -! .r,-,,, . b .
,
, -, -'. .. .. ..- -I . I
, , , I
. . 111.
-
-:
. I . . . J l --- ' , -F l -I ,i - ,f ., -,N - - V. ,-
. - , : .-'. 17 7 -",' _.. ., , - -, _ I - - . -, I : l I1. "
. I : , . , : .- . ,-, ; -,
I _, I.
- , .r. , - I" '. . .. - "- A 11 I , , . , .,V .y- ,z. ... , ." . , I. , . " , " - - " ,- , _. _ . . -
r, - _.Ml -,, ...,, , qt--,P- I ,. - -. 1 .

,
II I
I I I -. .I
. .1 -: .,j I
.. : ,..- 1,:" f, , ,,I-'- , :,':, , , ',
-.3' .
, ,-
,
, - I ,.
.. I, , , ""
, I
:, ,,-,.,-,'_. . '.4 ,., ,"-,
.. -
. .
I . .. " 1, "fE,, _ 12 ..;._AP 4 lp , , _r' 4L_' '_ . ' .,!.; . - , - , - , . - i
.A I
,
A- .
I. -
.
1; ,l . -
-
...:. -.: :;, : I , ,,I
I
. ,,,, - ,

-, -..,, __
.,_ ..i,-,, !',-Lcp
. ,: _,i2 -
.L.,
,- ,.-
I I.; , - ,
,, _, ___
',-,.,I
-il
, : " Y , ,,, .'-, ,iV.,..,,
... ,
-;. .- .",11I j , - - ul-
.I.LL,)
_u-, ,'*F.-..,,..;."7
1*4,
,- "I. !',
I -_-1
I.,1.
1I-110
, ,,I,.,:.
. .,II.,"
. ,",1,', ,_
"I .- !. r, f ,
I", ,,,...,.,I , I , L":, ._..-r , .1 .. J'I.- '' * I' I j " I -. ,- '-','I" "I ; % 1, I I ., I
.''I I ; I

:; d ".
, -
I -
I
..
L-
m,
- ,,
.
-
r:., , .
- "
,_
i
, ,I _
';- , -, '",
, - _. ,
.. ,
1,,
1_ -p = .
.. , .
,: '.
f ,, ,10
.
,,
-,
"r - ,
, -: : -' 'jT..
-4
,
.1 . ,L .,
.., .51, 1111
P, I.
- ., _.
- ''.
=V , 7
1_4*
,, I
L' ., _ ',,'
k I ' . t , ' -r ', -V ,V 1 - ' . -
'
-'
-
"-Fl ,
-
U ,
_,";
- , .,. '
- I . ''I , - . - _.. : I
I
-
-
-
.
. ,. .1
,- , -, j_ ,f .. ,'- ,LL
IL I, ., I - I "I.. ,
_
. ,_ . - . ; , - - 1,1. .., L .- -
, ., *,"'6
" , Is. ILI
". " - . ,. ,.,
- . "L.-.
'. , - - , , -, .1.
. -r
, _1 .'4 .. ,
- . Ii
, I .I,-1-1ll k -1. _
'. -,H, " I?-?_-_-, " "I 'I'll
Ll
. , , -
5 - ,;Y.
,,;
i , , !,t, L-4 y
". _. -
- C,
__
-'J
. -
, I%i-T , ,'.'
"
j ' 11. "-I i.j_ )'- ,
,i-. .. p.; ,...il-,,,
, , '.' ,j ,
, ,-
"', .
.- . .. ,, ; r
I ,
" ,
'-- .-- .
1,
.
.
_t_-.
." _
_ I . . I I
.
.
. ,

I ,, , , -: , %
'., . L.". , i_ I- , __--'.71 I, ( ,.-,-
. , -. 1I,.
I , , ., , - ',I
,
P, -, ;,.-, ,. ;N,4 .l 'I ' , '* '-"-- ' 2 - - .4 r,,, , "
, -Illp .
. ;'r , ="
,
,,, , " -,
.;r
!-.4_, , A hw f , , 1. .E ";4,."_lIi, -C, j _ :, , ' -,
- ,- ,
__ Ih-. ,.11."
,
'. I , -' ,,q
"
'
. .
_., ' .- , ,
it;
,_ L I'. ",-,' :1
I''I
-1
..
--.
1. . , I -. - -_, , ,- , I_, r,., , -, , , . , , " " . -. , . '.'
I -.-I , ,.1., -
I., ,
, .4 o , - I, , ;o , . I .1 ., ir'._ _,.. - , A I . - " ,q .--. . . , - - r . ,
, I , 1, .: ,
. - 11

,,
, k,. .,;, ,, II -. % , -,- - ,I - 6 ,. f , Y , C.- , , I I
, . . . I I
-
I
,., ,.- _ . -
II , ,
!'L, "i -1 - I -, ''y-r " . - -)L, -- -
i
l : ,, , . -
."
It T- ? - ." 7 " , .r , '. , , I , ., - , .., ,. 11 I , I - I I.I. ... - .

1. , W ,,, :§;k,.J
4,i--"O ' ' 'i,!.".- -'.' '
I
1. I - 7". ; - , 1, , .,.r .", I"I'T . '1 -. 1 )
i , .. " ',
.
- -
.,.
.
.;
. I , ,
.
. I ., ..
-4.
,
L,; .,
I- .- .
.
-:-
I ,
.: 'F-: " I
11
. -., ,, ,
,, _ .,
I . ., '.
- -1- L-rp
; -,-%A
.1 .-.
'' , ,,) -r
, $,
? . I..1, ,4, , , , , ., -
" . . ,,,, - -'.I
- , L16. It~ - T-7,-.,j
, ", 'I* .1.
- 11,
. _T" . =
, " -i -"L
y
W1 - .
-1 ."I
-aik-t
., ,.
,j
_ Aj, "I'll,
I. il, . . f., t - ", -
. 6
,
, --
.
, -, -, , ,
, "-. . I.-
,
,
- -
I
I .-
,
1! - I . I I I I
- ,. ; q I . .I I - -, I I "," , ... .11 I.
.: - mL, Kl,,.,S
. . , : -_?, , ,- ,-I . ;! ,,f ,,.],- I .T-A ..;, " ;, "ki
" ,, , - , . ,, li . 9 I It Z, '. ko4:.;, '. I11,-,
01 , - "I.' - e0 f7. J,'
- -r
r. I ,, r ,, _. - 1,4- , . .I'I, , , _,
1, . , ., . . - ;. . . . -, I 7 , .- -
, r,, ,! 7, ,\ -, q-,.vI-" ,,", h, . - , 1) ,,I, i-
p.1 .. 1, .-x " ... . , -, . I I 11
.1 I I . " _, ..1 -,_--1-..: -, - . _. -- ,.,-,, ,,,T -, - ",-. -i-_
..J , . , A , - ', .- '. -1.P _* : ." ,;, , . . --",Tl
. - , ,, . . ., -,,--,; A 1 _' . - .,, t-i., i. :. V; ,_ I I - I I I - 11.
. I
I
-_., ', '. ,
_ , I"," " _ 'j,_
.
j _
, g7 - ", ,"( , -'),I .." ,
_, rf" . . ." ,
, -W :J - , . ,, I "... _11"N 115 , , -_ 1
.''

. L,,,viI
I- -, ,, 14
-
::,4 ri_ I , ..' , "I " . I -11 -- =4
I. ' , ",-i..'F . --!. , ,. , . I . ;; , I . ,
. I - I I . I -:-
II I- I.,:_
".. . ,._" ,, _.
,,," -.-
1',C' . ,j IL!
',, !-, ,, 7",-."
,x O 5. -I , ,- X, T.Jf.,;.]l n l 't I .. , I - I
- ,, .;t- k _4, ir) ,, :,, , llllj'," , '1, .,., , '_
-
L." l ' - T -
v
I . . I ,
,.- .1 - .I I .,,_.
- , - - ,,, , .
- -I

I
1,
I
_. : _, r. _
- D
* , -f , , , - 0" ,
i, , - . 1,"jIr
.-
-. , ,,
-- -..,.4 ' I.,, , q--,,,,r -- ...,,ldh.
.. -
'T't -, ',
- , ,', "-
- , . _
., _.17,-:,;
., %. .
, M il.L
. 1 i4
il
-
lc
,
L
, j
_ , 'A T,.;" T--
---'
i-:7Li-
-
" - .-- .
I.. _. "-o"fI.., ' -1 ':
. I.,
j , , ,,,
-
;
-
, , Y '. %,. ,I
, , -
__ .
: . ,I . . "I
II
I'll,
, ,

t ._j,-.. ,I; ,mz1,I

. - .. . . .
.II . '
.
.
1 -
- , -
,
- ,, i- , ,,,
1111.
-
-_-; ,_.: . - ,
i, rd
IIT,'
- I .Ij , .- , '
4 '! , .--.-. - ,, , I. , , , _.
1 -' -
.- ,lq
;, , , _'_
r, I 11
II A. - A.,",
,-1,-"'. I"-W " ?
11;
* . . " -, T =, -1 .
.
_
-.-- ,. I ,, I - .. . j - ,- LI I I H z '- I, . _1,
I I. ,
, I,
4- ' , N", ,: ,, - ';" '-. - -. . ".X. , , ,, ... j"4" ,

11
, - , " , - , , , ,

,",rl
-,
- ; I _ I " - T . I =_ . Z. 11' , ._g
.. : -_, 1. .- , ,m, , ., ..Z "' - '., , ;, I'. ,--,-"-_. _'l - : - ,. - *., - Io" -. I , " if*,lwl.. _I' .' ,, -,.-,' 2 R ' _4 ,
,q,:c ,"-,-, , " - ,."I . , . " _11' T.- - -- , ... -
J, I I , - . -1
.. . "-:. , : -1 , I _r I -, . I '111.. e 11 .. - I I I -
.I , , . 12 - I . II , - - . 1 , W '. __A ,!_ _ I " ., .. ,, , 1. I . ,
I . . I 12 . 1. , , ,
j
,
, - I. "1 . -
, , . 11 . ,' ,
. I ' 141' '4-1' , " 11 .14
t , I I--- .. q , . . ,j 4" , )", FT ,
. ;7"",. ''.-' jl', , ;4., -1 . ' ' ,.- _ '.i , T
- I I - , ",, _ .1 , - I , - , , II - - I--.- - . , - .11 . -
.
. I I I11
.
I -.. I . I I ,, , ,
. , ,I ., ,I, ,:' -- , _ , 1. j . ,i)j. . _'L: 1-4; :,-I, - :" , ., , , ,.- , , I
- . _.
: . :1 I. , i . 1. , - -.L .11- .I - 1 - - . ". a 01--r;
'j.1c , . , "f- .
_
_ -- - t rljlf 4
v
-, ' '-I. - T,-. ,- -, I'-..
- .!-, -., -, - , 4. -.-I
.,fml-
.:,,,-
. - , I, I-I11,-.- -
I I..--
. .
I - I 1, . I . I , - -, I _ . - t - I ,',' I ,-, 7-, ,I_ . 7- .-
.1 , I . . I., I
- , , .L . I
- III,; , , , 1.
I" I. i -. :_., , . . _ _f , , " .41.11 F '- - - . I :.
j " . - . - , , .'T
' I . -
I ..
-6 " , -,, - ,
, ,' , , .f, , - , - r-., f " -, L'' 'N , F, , . I
I - . . , l . -12 1 ,y r ' Y _ .f, , . ,,. - .
-1 - - " . I "c'
I i I .. I .. , - , I - , L "",III, , ,-. q - , ,
I _ _. - I - ;0, - - ,, " _ - . - I. . I
.. I- -1 . - , ),;, , '_ L I.- - ,7-, 'j[,'jv ', ) , -. , , . ". , li - -, -, , , , " , I --
, " I- .
i :, I I-. I . %_I L
.
I
_ I- I
'' , "* 1, '
. , . " I I .
, '_'-.r- - - . - . " , -
;-IL' " II II(I!, , -.r -
, - , .. .. . . - ". , I , I
, ' ---l j' I __, .,' - 'L. .,, --' I - . -1 e -"I -- I" - - .VT,5 ._. . . ., ,, , , ,- _ .. - :
I.--_ .,-2, '. , ''jII. e - - j,,!._"
, - - '" IL-i ' 41'. I J' ' !,'r ,5', ' , , I, ,
L.
't-I- - , - T'_: __
' - ,
,,',' L''L
L'-
Ij" ;: ,,I
- - .,
, '(. '' -"' I-
_..
-''j.,
_L,
:.
' -
. I
I , L *: - .I .

- ; 7 ''L
, .. I .
- .. _., .
_. ;l . I
, - . i_ -I ,I . . - ql 11_ i . - -. . I I
. .- I I I'I - F.,,, o v . .y .; (' .r;. -, _'-LF,,k,'," ..I, HJ, I - ' -I!'. 1, . _,__ , -, ..- -- . . -
.. -:- , I .
. I . I
. - I :" , .- ,,I I, __
- iI I 11; I . 1,
, . I,. ,, , - - - - - . I- I .- - I - , . , , ., .. I . I I . - . -
'-
_.: I I ..
I
. . I 1,--. I- - I -
-
. - , r I I II , _ - , " . ,1, . ,_4' L'A , 11
- " Z jr - i7,-.- ,,, _
$ -. . _-_,:.-..I. .... It, -. 1.
-,- ,, IIi- ,Lr _'.
- " i, - " . - , ', . , , , -,. o.L' _.". .1 ,' "', : , . ., . - _.. . . I . .
I I . . - , , '' -. - - _. - ''I i.
I'T . , ,' , , - 7 ,L
--- , , 1. - , ,'.
L .', '[
-j:IT
_, L'-, ., - , '.- -t .., '-'jj.,, .'f. . _ -, , 11" ! , . --
.11." l _ ,_ , '_, ,_ .. lw
1- . - , , -. . ,_ "'; . . .
I . - , ; I , , - ,, I
I - I I : 1. I . . - ,o , , . I - t . q .' _, . , , j,

I
.- L . , II -
I
_-
.
I
I :

I -
I
I. -.I - .k , - , - ,
- -
- , I ., I -, -
, . .
.:-,
, _.T- -.
__ -. ,
, ',_'
k:I - ,I ,
,11
..
. ,. -

'1.--,.'7
-- ," -o-- I ,,_--
j , - ,-.
-, ., _ ,
, ,., _ . ,, ,,_-

I
7e,., - -
-Ta, - , , ,L'j,_,_ --
, j. ,

,.,
T_ ;., . iplj"-, .,: .,__,J, 1,
" ,,I.. ''L'. 1- I
,
,, . _.11,'Y _ . 1"..- ,,
__" A
1'1
,
'. : I
-
:1 , t .
I,I 1. _ " *, I I

.,
I
-
- .
I. II.
I
.
I
- I I. - I
I I
, .v I P, -. .. , , " -,- , ''I , ". I T. ., L . , , , .: I . . . .
: q I - , - _, , , , -" , jjj , , * , _ - -1 . ' ., -', - ,
",, . .)
' 'I, " , ' _.."LlIfl. '; V_. I , . . _1
, -I . _.. .,., _ r, _
I
_I. " _ . _,,: , ." I_ _ _' , __'L, " - , ,17, ljw -? - , I' - it%, - , , _- , , ., :, .1
I -I I
. , . I.,I, . - I I , I I I . T. I . ,: I . :. ;,j , ,,, _: , , , r. _-, I . - " * I ' "''. I'- .-- - ) _ 6L" ," ' I' I I ,," -, .. ,- , - - - - , - , . ". :'. I . I.. .Z. ,
. - . -
. .
-I , I - w . ', 1, . , . "_ -11-:. . . , -11 , , , - I.. 71 lL,
. , . - . - -, '' I- .Y " , -- ; _tI. f, ", ,F.J .
! T ,_ - . -, -. , 11
'' 'E
4" s; -.k
,
, .- : , ,) ,W , I "',
- ,
- , ---, ,- - , r, I.."
.
I.-- . ',
,
, ,, I
1: .
, .. .
,., .
-
.
. I
, .,., I: I
,- , ',' '. " " .,, : PI i , j I- - , , , - j , _ . __ 1,
" . . '- -. . , .. - . ''.' ., 14: 1
,-. 7.1, .. I - o I , -,' - j ,-I-, - _ 'r . . , , - ,, .,! , "., W " j2r.-, - j , -,4 j.1111 L
L P , -- v, Y v, I - . . , , , " ", - .r , .
I r,. I. I lj !-", i : . , ,.I . ,., . 1 .- ,,-
_ . . ; - : ., 1 -, - , ,
... - -, 'j ,.. , '. ,,,., -?a , A p- - !. F-- " .!,-j..1.- Z, - ' ' , . .
I %. . I- 7. I__
.' _-, -.
- II -1 II " , .. , j I- I --I -- "," , " - .y. , " ,, I -, , , .
.! , I , . , , , . ', " : f , - , , k.,. _,";,):, . - . - . _ . I I ''
'. ' ,, -._ ? ";7' ' '_-Q ,
. , ;, 7, j -, 11 , - , ), , " p I, -.i, _L_ , ' . , - ''I :
'_ .-
. I
1
.
. -
,. -
-
, . I 1: , : . --;. , . . - -'t1 -'-, " ..- . 'I , " 4Z. _ I -1: ' " 7I -.L L'. '
... -
I, " - - %, - . I - . .. , . ;, '.' - _7 -' ; "', - - '. ' ; ', ,K.-'. I-, '- . y- t tl" , - , ,,,'T, , ,
- . ,.jj ' .%
-q"
.
If, , .- . _-T.11 -- - ,, , . 5_ _ 1, ., I- - t' ' . , - .. '.. ,' "', , . ,.. 1: ... I I I
'. I -_
L
I . :- a, , I ; . , - ."
, -
I, I .: , 14,.,. , ;
- .---., , -. -_ , , J, I- ; - . , , ,I d _T, , " ,iL - 1% %I , - . .1 I . I.
.., I ,- I __ , , . .i1, 7.r .: ,j.._ ,, ,, ,- ril - . ". _ , . .. - [ _'. L , j - I
. ,, r , 11 , - -, . I .
.
I .
' r ."4,7I4,j,,
. .-
, o' .-' I; -- , , _L?I , , _.1 - '
-. -,X .- . ."
- - . -
:_ . . II . - . , 1 I. I ., I 11 II l ' , I , ,,, :-- Z- c --- , , ll I I',SPY ." ,I- - - x - , ., -. , '1:
. -r-. .
_, , I. , . I
II11 - I .:- 7 . . :7 I . , -I, .I ,-l", I. ,.,-- -. I -L'- ,, t_.,,, -, ,I
11 1-
I - ..'' . II. . .-- li- I J_ . .. LI '' - _ rq'_.",_jL'I. ' - ,,, , I - - ". ,',) '. ,' I r . ,, I . "- I i - I . I
. .1 , ... . - , I -, ,' -.: '. , , --- ,,I 4,.' ., . 0F). .N, ,-I ,,I , _. ) , , ._'t, I, 4_ ,p . ,
" ,, , -L .'I i , , rL % .! ,L. -,- . 7... *-, : . L, . , ",
.
I
I - ., , , ,. I
... . - __: .- -, - ., ' '1 ' L" '. , r'l ,f__
_fj
. r ',S, ',V, - ' ' 4 4' L. ',f, 4 ., " , iy , , 1,,,,,, j., , , , . ,,! ., 'El, -I, , 1 . - .., , , L ' .. . , 11, . ., .1 . . _' . - r '.,. I . .-
r, , ", , , , :, :1 , - , -
,- I , ", , ) ' j II I_. . . - I
, ' - _ =4 , - J ' L- - ' ' " I - I,-'ll , ,-.,_, ;. ',_q 4. F_ _ r l , Tj , ' , '" "), , 0 ,' ,, j - ' .1, '
11 I I - -
, . T I - - , : :_ . i- - -, ,. ..I, fl, , 11 '. , Y, - _. "'I -. - 11 . . , I. . .. I. - I . . . .
, .. I- _. &I , II , : 6 .- I - f - " L, " ,. , ,; , ,_ .- '', , I,- -. , 4 ,-, I, _4lk. - " . , -LI.,;, %. L ;i , , - . - - .r . ! . . I
P , I I _. " ,, . -, , . _. . , I, 'L , -1, , " ,1 , , 4 . '. ,V I , 1-1 'T -,,f ,' - .1, ,-, ,I . k "' ., - f.L ' , , z "J
, -1 ' -,,,, ,_ . ; L IF, , , L,,rII .
- ', ,_; I .1 $, -L" I. . , , .. 11 . I . .
-' ' . I I . I , . - I I I - . _ , ,
. ' - _ il (I , I !- , p ,- 1I f, - - - -
. "I I . I I - . . - , - .L . ' ' . 1, , ,- " '- -1, _
- - -, v
t -L, , , , , , - r - I- ., , , - - _', I ,- , d . ,; , '. . . , I I - .
- , I
I11 , I I L I. -. ,..,- , . :, : ; . - I.. - ,." . L Il.. ' 4 I- - -- I i l- " , =
P + -__,,W, ., .1_.. ; ?' 4_ _ ' - _ -' i f.' L4T ' , __- ; I r , , , , , . I I I. I- . I
-1. , -, '. 'i I -- . - - - : . ". ,-, A r- l , _7 - .1;., . I, , . . . 11, . I I . r . I , I I .1 .
-, I . 'r, .1- _.I I' .,- - F. -A - , -. - '1, , -
ll . . I I .. - . - " I - 7,,,-,, t. , , L , 't ff I" c _ : - . , , - .1. - : . , -
,. .... 1, . . , . . . I 1, . - , .. .. , - . _. , , , .' I 1, t ,f, ,
- . - , . . ,U, _.'., .,, ,,- I , I I I
,, , ,. , , I . - - ". ,r,, , , , - ,,-(-"- 1 .,
. I . - I I I . , . 1, , , , 1. I I - , . 71 ;, ... , I, 1, -I., . , ':. -'T ' -., -,', , - . - ?U - - . , . ., C 1.v_. . .,l I.,c., , . , - , _'._f. " I - ,-. .
. I . I . , , ; ,
,_ : t_, , ; ' -: _ - - , , ' - I - . - , i II .- I - . - -, I ,- - .,ii Ii!. %v ,,-'! , P . y, 1, - __, , ,-., - , - - ., . I , ,.. -. ,- , , :. -1
_'L" _'.- fo rb . "."l 1: . ,"-: ,- , ., _.. . ..- -I . , , I I _ - .I I I .. . .I
, I __: I- , L- ., - III .
I4 . 0 7 - I I . , V,,Ir,..,_
, . -' ' I '
- - - - . , I I- :'- . I
, _': , :-. ..., , . '. ''-L : ' .: - :. . - 1. I - , . ;_-. - I, I 1, , -- , . . .. 1', -, -.,. _ :, - , "I c , 11" I I. - - . , i_ ,, ,..- , %; ) " I
. . I., ; I, , _.. ..Ik
.,
L.
- '. . t , : ,, .,:: , I . - - , ! -
, . . , _ ,. r - ' - r' ' . ,-, . 1.
_- .7 Ji .-- t, , .,.L ;_ ,; I r . , . . . I_ , ; , - , '. ..-. , : . I -
. . I
1 I
. , I _. 11 __ : . . . . - I - . r, , II' , I It " . .I- j -I . -1 ll " j . , , , , -, ;t . , z-7 , I- I ,. . I , --, . I . . . L. ' . I . I I I
, I , . : , 7. , . ". . - , I I'I, i , ." I - . .. , I - , -: ,,I- . .r .. -,;( . -- I
I
, , ., . , z,, . , %I I L I , "'% ". , _ I , , -. , - , ., ,. ,_ - , - , I . '. ., , . 11
i,
6 . . - '. , , , _W - -. . . .- ,. : , I .-, l , r. , I ';, , : , 2 -. . w , . , L , I - , . . I . I
I . , , , 'r ' : - - . .-- .1 ,- ,, , . .. - ,- 'i .. I, I- , . j. I 1
I
7 - ' . . t, - , .1 : . : I ,, , ., .. '; ' : , L. . 4 : ' . . - I . ,
I . . I
-
.
L"' ' II . . , . .. - - . I ,, -,: - .j I -.1 -
", L" , ' - I . : :, - - , .- ", , . 4 , .m - : . - I
. . I - I
. . . . I L - L' . - I . .. ! , *. ;,.. - - -1 . "'i , , Z"_I , - _ I . . . . - .1
. - , : I . '.1 , , . I . : , - I . _ ., , . - _p. . -_ , T,- -, -- , -, -- -- - , , ,,, ,- -'j , , -, .". - , , I : -, i . I - II . 1, - . __ _ I II . I . I . _,
.I I - . I L L _: ) I .1. I -. - , . ,I --, ' ' ' l ' " L"I , ' ., In -, .. , . " , 7 . :. : I. j . , I, I. - " I, I . 1 , . . - 1. ,. . I .
-. , , ! -, :rt'll - - . . . . . . I -. _
- . I , . , 1 :
,
I
1- . - I - . . - ".. ,
" - - r -' -,' -, , , . , , . I . I
.; , ... ,,, .." I , , , -,- m , , t - .-, r ,, ' -", , , , . ' I I , I
:I : . , ! I. . I I - . , ; I , - , , .. I - _. I, - ;. I r; , ,., - j, k. ,I_. II-, '_ . . _ ' :, L ,,
." I .
-. . I I -,- . I . I . I . . .. I I ' . - I . . - .. , , - -1 LlL -' 1, - ., .'. ._. " ., I,_ 11 . ,, : ., : :. , f I: I: :% . . - r . - , m
.. , . , - .- .- - - , . "P - ' 'L"'- -, I -
-_I
I ll - , , ., '. :.:, ., -'- -. ., I , r , . , , ;: ,V.-.,4 ., I _; - L'- 1 L , , " . - - : . . , . , 1, I ' ' . ', I I. I I 1 -I I
- ; Ir ' . . I . ,. .- I - - , , . 11"I 4 . %. I . , , , ,, - , __. , , , , : A' "I
,_ , ,. , 'I', ,,,r, - i- . II,, 11. , - , 11, , ."., - , . . -1 . I ,- .. L, -, I.,
; . p I r. . ; I ,. . , : I
- ., r _j 1,11
.- , : . wi, . I - I. I . r ., - : , -.. ,- I 1 - w _: ..- -, lj L'-e"- . -, . - . . I '- II, I . - : , , -. . . I -, - , : ,
_ : I I L, _ , , : .1 . I I- I. . .I .
_r, II I I , - ., -. - ,_ - ' - _' k .-- ,. , , . . , - 4 . _1
:, ' r , ' . I. . : , I; . , I . I . - L' _' , , . I. - ,- ., . ,- , -, . II . I . 1.
,
r. lz _. . , I I.- ., , ,,, I: . i .' , ' .' - .'.: ., , .,; ', . I . .. I - _.
I-
, -, , !.- , -_ . '. :% -:' " ";-I,_.- I , -, - , ,- , , -, , :_ . ' . - I- , . ' __. " I .... , . _- L" ' . . : .. " . , I
. 11 -, ,. I
...
. - - _" - _I I -. I . II .
I. . II I :,- , . . ' , - 11: I, ,, i -. , " - . I. ' , . . , . 11
. , r, . ,., -I
. ".. .. , L'
, . , ": , ,.;' - L'_ : . . ,. I , . - II
I .
.
. L' . ,I : - I . . - . -.
I . - :- . r , 4, ., I- - - r," , .,34: ,.'
--' . .. . -1:.'' I - 1:, -- , . I I.
, . . . .- . I I . , . I : ' '.' " ' - ' a . . . - . '', - - I :. .1 " 2 '. , . . I ,. I I . "
I. - I
. . I . . I . , I , .. . . " L I. . , :- . . , _ I- ' 4 :' : . I - , . I - -- . I . .
. , I . I : ; ,. . II: . , , w -. Z,, , . . - _ , - - . , , ..I ,
p . .11, - I I , , , ,. - I I I . I I
- I -: , , , - -1i , - I .1
: .1, i 1, . I . .
, - . I . L. _.:, - , . I . I -I - I. , . I . I rI
L. . L , . I,
1.I I. I . __ ., 11- - . ., L,. I' , r,:; ,' , -
I . - .q - -. ,:: ._r II . . - .- ! . , , 4 , . ' ' ."_ I ._; .. . I II- 1. . .
I I
'_. . , . .. . I I-. , . I,
- . I I - I I . I ,I ,. .1. i,. .:: - , .: , I
, - I ! : . I,., !' r. _ .", . L . - I. . I . I
11 - .1 t - ., - . . . . I
L i. ., , ,, (:. _ - . I : " - , -, I -
, . . 'I," . , - . , . -
- _r . I - - b . I . .. I I I . I ,
'. _ , - . I . : ,
I. - II I , 7 .- L..' , : " . ,, 1, i , , -, I : 11,_: I - 11
I, - . ! . . - . I -, , , !
. i . . .. " _1
I I .I , . I - . , i , . I I , . I . I - I L I I - ,
. I .
I .
- - r I
, , . . . , 1. . : . I , .. ; - . I:* -- . I . I
- 1 I , ,
- II r I
I , ': . '', - , . . . . 11
. .% . I. .I L' I - . : _ , . , ; , ,.
" . 11,
, - I I I :, q , . , -. , I I II I .
I . , ., , I .1 ..
. - ... . . : .I . - - . - - , . ; .t .
I I- , .
I I . :
" ."':
. . . . : , I - . I. . I I I. . I. I 11 . .. . . - I .. I - . I - . .. I - _ I -- - I I I I
. , I I.
- . . . :: . - I.I - I ,- , . " I
IL I . I - . .1
,
.. " - , - I . -IrI .- . I . I. r - I I. .- , - i, ':j .. ,. - 1, - - : - - . . . . . , 1_ - . I I
.. ,, - , !- ..-
. II . - . I- : . .; ; .j 4 _: - . " - . I , . , . 1, 1,, - .1, - 1, . I - . I I
-": . . , ", . , , I ,
- 1 7i..- __ - I - .
1; ll I L - I : Y . '' , - , - L' - I
, ; , . I ;- . - -, " - , 1; . -,. -1 11 I, , I I
. - : 1-1 -t -: - . ., ; - ' . I:, L , , , I
r , . - - , , , - - . . -_ - -
r f 1., I : - .- - - , , - :', . nII.; _%, I , I
I. I L 'II ir- , , -.. '.L, __ " I - - . ' : '!: L - .. .. . : . r ' ._ I. I -1 . . I I- - .I - , .-
. . , I . . I . . I I
. - . ; . . I - , , , ., . ., . I . - , . I -I I LI I. . ' 1 4 . . ,- " r : II . I. . I - ., I . I . . I
. d . - . I I
:, . : . f L . ; . . . . , , , - . . , 1: - , -, , ,- _'' - - 'r ' - II I - , . . : I I .
. .__ , , ;, . :