2.

(a) DiBOs
Di3Cw Shooting yethod fo ind
The meueal Son cf The B..P,
tous d you detenrine beten
appgoermath on in shaoting method
wSing iraa jrerdation?
Z a))Conside a two- pur boutzdoy
Vale problem, y'- flZ9y)-) ih
conditi ony, y)= cnd
the lbonday
y(B)=y
L -(2)
we set, y'-2,
2': f(,y2) win
(3)
fon Consi degation mo problem as jnitial value
problem we reqine 2=yat=a.
This I5s basicaly the slope of the nunou
y iritiay. ie the initial Slop e.
Bud mo Such yole u 29escribedccodjrg
fhe above pg10b| em (3), 9f we tave o
T0

we 1t toCemate anume Gad silkioh

d Caculated vae
We Con Then Conne the
fon yat 2-b ith he qinen bounday
CMdition y6=y1. Bu odusting aheg
bette
Nave la) onic ves Suppose Salution
afprorjmatior fon the Computed
SInCe fhe queys deivative at a repu uhic
fhe tuketoy of fhe Computed Soution
tasyets to shoot the vae t eb, This
techine is Caled a she Hngmethod
 y9=9, by guesing the initiad slopeae
mmproving Slopes sucesiveky we
able to Shoot he tasyet the 1.e, yl5=Y
Sotion
t is Poss ible to improre tindig
by inea nterpolatjon on M9ot
e9 fhe Selant method y
methods
Nestor Rouphson method ete.

The Sharting techwige depends Solely

or
Onfhe Chare of rhe intial Slope
the problen
Gen Vae ya). Achually,T.vP
oF Bup b Conveted o an belto
o() Linear nderpolation fon finding
pporoxihotion in Shontng method'
The Shoting methed e quioes yood
ritalqueses to. the slope cndthi
both ine
method Can be appied to Consjder
pyobles. we
). and noh- inea poblen
uti fhe 2nd ogde bouday Vaue
as (
f )
y),
wnous theslope
I.v, P, we must
OL ahy
stat with wo itial
0f yo) 8 m. we and then detemine
gueses fu hem
jen
Mumerically Co9oespohdinq Vaues oft
Vae method
ch
iritial
S ) usig any Moand m, cnd the
queses oe
let, the two
ufy) ae obtaired b9 he
Copesponding Vaues symo) and
1alne nethod denoted
ym,1) gespechvely
jneopolati on , we fhen
Bywsinginean capproemet on Mfo
Obta a better
M-m6
Mo -)
DU)-9(o) y|my)- yny )

his M=
gives,Mot(m; ho)

phe Ivpy')=4 (X)9)

we ous sowe

ym,1). we ogan we na
chd obtain
0nteyolatjon wsith miy y(4)) and
yMy) ) to obtan a bette
(Me,
appomati on Ma fos m, nd SUDN.
poces is grepeted Until Coegaee
lhe
Bobtojred e til e voule Of
Dy) agces witn y) toTedesie
aCCay the speed uf Con vegen
cepends ön hou god fneiti al gue
610 Howeve ]n me coye of on
nea problemy e ineo nterpati
yjelds Sometme unsatsfctay gesl
n tis Caye ne Neustn Rophson net
on selaht ves bette 91e3ut, |t
]neesting tomote heoe Hhot, one Co
 e houstne mproved slope 99adly
-ppoah to shoot the taget yealn
iene oundeogy Conditio at he
end
wute ad vcuntoages qnd dsadvantoey
of multistep method. pnediton
(6)Establsh Adem- Bashbth
fonmule fon dy zfl (6)= 9o
) Te ad vantuges of multi step
method
Aecuaey multi- sBep methad
O hghe ho
Can acieve,mefnods,
highen ecly
e3pecily fo
tor
Sinale-step
Jagg en Step size. methedsCun
Efiieney> nuti step Sirge-stp
efhient fhan
be mone poreou3.
s They use
netho ds Value to CoAute ne
)
Corputed
values methas conbe
SstabtyfMutti- Step mefhody,
) Single- step
mogre Stebe thun
dompen oC0|lation cnd
as ney Can
teges of mutti-step
Dsadvan
method
Comple rity> mutti- step mefhods segue
to
moe Comp ex
Statg Cen be siep methed , cuthey
Sirgle-
'mlement than memoyChd Comuttiona
n quine mone eftrt.
(2 stating vaues> muthsBep rethds
Con be
vaes ohich
quine Statng fu ho n
espeCjaly to
difficult to obtjn,
inear pooblems.
paopagaton multtstepmetnodg
(3 E0r SmalUK
C Poroegate egNOAS,
aunwate
Sep8 Can
(n poreviows Ot
and affect the aCeuoray
Subseqent stef%
predictor fonma for
(b) Adam- Bashfortn

( ) be the Neuttor's ipoltng

Wth u=
de9ee m uhich nyolake
Porjal of
4,9) a the (+1) nesh pts. as,
Now, Rewiting the di ffenontialeaton
dy
ahd integpating from to ut we obir
f(,9)) dx
Lnt

nthdn fh
ohene fhe co-effiuent

on theRHSof ()
Tus,fhe expgesi
Motahonay Vntl -ythldbtytdydtyt
N-Xhtuh
t dm tm) X= h d
3)
(sastoth mothod and The L
eongtnts doj Y3ushfarth
di
Caled Adom- SHep
methad.
Constrts. thii js act)finst fewd's
the folloysing ane The
do s')da =J'du=1

.2

du
|.23
S'(u't344u)du
+3422
da
1234

3 ++3)- 25)
20
tohen m=0, then Adans-Byhtoth m
neduces to the Elerts melhod. Metaly
This 15 bayed on tne points lon k4

we eprese how al mofwand difbs

ot te
n he abovefnua /n term

have tno final funnuei

-rfe-frth

Sustrhuing there we gety

24
t s is an opn tupefonmula
3 ) D'sCus Fzoiit and inpiit Nmeria
Scheme? meal
EStablishcan- iCoson
(b) oe dimesional heat
Scheme fuh
Conducting eeuatt n unerioul chome
eo
shpus that this Stade
i3 yncoditionay
2a)Spit imesical Schome:
ohen al Statig yaes oae ineett;
bowdey
aVaable from inital cnd Vaues aie
Con itions and eoch ne ).
fom trhe values thot ane
Obtained
NmeiCu Scheme.
SchnËdt's nefhod. fomula.
omple ) predicton
u) nine's
Scheme'> An mpliat
Inphuto umgicad where the
Scheme, s one
wneyCol on sComputd
Sol
of ne
hes yoleefomun that de pehds on koth
c
wsingCugnent and new alues 0f ne
the means Thut, the he
Solut on. thisme Solution is not epucity
alue. oft s impucty
theequati
bt on.
deirod by he -nicol8on method.
) ranw FMla
Xample CoNct
(i) ihe's
 we b
and ohiNiCrane
chcollad staiitgthe
Small necesoy ohin Thisis evation
expresed by-The.
e
or sepsie he
degivative
iffernce
onile aloulati have jmensioral
Cran-MIca
stble s heot
neduces
or n
one-dimensi
Simpie seen i5
at is Conduction
The schem,Ins fo and rder of
expreKed tne vesy tsme
The
all hgmustKeptbe paubiuG heot on
ent by ne S(194?)
tot to computational ona
tre explicit scheme
desivative
Spuce friteVoulid mal Step
onable
ttanL d=Condu
heotctivity Condcing
ah
Werage by voune
ÞropoRtd
ie, on
Volueg because St= rotthilTne Ot
eaation. yPeiffui fu
tne, fowerd conegent reas k mefnod
tjme ofA Of a cund Oe
neu cont S butlt orErat
meod
u
 () enat engiCollo
M fiyt
-) heeandhe jne-us ea
u
fallousg:
Ctae
algebie at
time-leve
ofu
The
The hosh. of
contan M 9oin con
ae fu along yalueg lGah
Vaues
Theas ench ong of Valwes.lo
of 0 here eqhatË
u tjnevatues
seCond
ne
egos, h
LH.S)f mUnoon
of lrtens
re
jne
hea
discrelisatho, system wousn
(t+) along
jf simteneoy tne
tocotion value e
Bowndaogd
re ievel.
the hNeepoirts Caculat
xhe ng.
eanent
at wnnun
9rou
At me
Unnons ves
R.HS g.P
ctingNiColson gene mesh tineIritialalong
The
dot jtn M Ö=1
ne
ale n of
fofo
 twe moleeuo,
oneCnn hne
Heoe Computational
belofomed.
the
ne
mhyLotin wohee
In Conduetivity
Stable Crank-
he-
each
lerel
rismethod, N, tave we cl
Y
nAm-1
n i- (Hi)
T=, Conside eve
cotesponds
aW>0) t NiColson
N set B
matejal.Thof
e Liss)) NiColy
hei.e h=
b-g bf
2(+9) t hne wsing ())
orijd a nwumerË to- 219) oh
diffeance constnt Heat easly
witehe 12()
pornts ne
4t. = dinection e(itb j) (it,Ö) SCheme
Shoon
enatian, cel above
Um,npquahon
tl h=n19 (my leveÖth
shich Valey)
(wo
seheme
Compuate 3
ty) i3 end
B hee The
 Compl
bemayex.and get
we Bh)
-2ces( zn.
aso Fwny1.
e fmyn
Emij+2() the
of-Sotion
swn t e .
+e54)42(-).
by9eonh [os
Bh
P)2(19y -1s
2l19)
Umn-nn) tae ada
-9A5eißhem), (Bh) Sin(Bhy
|-22Sin'1421(Bhy
-2Rlas
2l+9) fomn
Fmyn}t we. aso
ineassingle 9(=T)
Swce
ays
,Maman, is gives Sideste
Fmunt
eqnatHon,NEm-mt
sutsty
he t ison ans aa
Solti both lies
-Emn
von- eonside
ea egue
he
fowhng nda viing hat,
we
The - Sin
ce, Ci see
SU
 [ub we
Quadratwe DnconJitioray
fomulai
7Gaw Stabieis SHeNce,
(th (n+) The of
taset 1,2,-
f)bd
neplatf;onpoits ottn lnterval Quate
Fonulal.
EStabhih l
0
dpree tae fhe Naley
a fhese 2es nodal cranh-
we of mpi-es 'Wn a goh-
btajn, we [a, (with
0thagohol 0f
pojrts pojrts
0f 6) that, n, pl(y)=noal 9>0:
epedent
palgnori Wnn()=0.
raws N0Coyons
with Thoe the (7 Nicetson'
ji ()=) Con e)as, ahd
oj
polgnomjas (*
Xi, weipoyromyals method
N, ae Gs
n o ght 4)). mpliit
Ln
-(2) l2) Xn funchorn
ae
the n. nafhing
Lago e as mthe wn Legonio SChome
ayintesth l) t a)
I. bt
 Nouo The n=}3,E), th |t
ot auadatune The as
neived
C,e)
othegohal famua
isaus Gaws the
by aothe fumue
wtth)(3) Enn
defired wnouhn. Fn),
he
afpying )=n wth Legerdne apponoimte
raws gord (nt)!
ihy by that
weight
Le guadate
preprtyCRodiguspaynamials
gon hem=n ifL Ofne vae
derays tunct degeejeen (*) (wh
7)
n of B j=9),, ot
Lgee on. on fonmuda wnour the
furmula
dote, fomua]. tne n e
ae
wwity polynag
jntenva ntepa.
palganiny -) and a as
set The n.
to
 and hot So
into
beween whee
S0,and3 AeAll to,ty
Wn)= we (he C,
9(a(tto)
(1-+)-(t)WnH
WhHl(7-t)
t)= to) Cuh
Srots 1)
wte
of bt4
(). Prut)=0 t
Corstont.
(-t) The
)
ralian
getawe nodal
Ae
neal poirts
age
)
and
 iet (6) s(0) n=91;
1t
ghee
Hj=
the method Co Tne Runl)=01
Whoe,
Prrilt) aefoNmuaH$S
|angest HojHy
f)- Ony i3moted
method.
o (a tow
Vaue Ds etfiUnts
on lagestpoweN no Theonee
ews The On dol
is
do uwhee
|h egon bosed
hyou pouCA basic mHs points hat, o0tsgiven
method
2igen on ahd eqnation fhe
of maghitude value to, th e oF by (4-5)
P()
(9
find Undeteminedideat
aVone indeper )3= t), un
atoi fuoL method ond
doemined
of Legerdre
Smallest a~ Of tn
of
inin inSfue tnis dent -n sho 1 to
fas
A a sjrg and us ti,
be
sgare Integutior
lnegaordependof te
Co-¬ificieny paynNg Co-eBficjents
dy he Conputiry
igeh mati a thety
Pouwer tn
(3)
d,, goraty bon
-
 he
Because Sides expsed
an be oyCotoesPon
ho (A'v)n Peuhipng Aywnd
ane
fayte by
donotes e toding
A
Sealan.
gtios HV;=v; cand be
1ate eigen
The V (o-) wgig Lneal
of 4i,i=23, fa ineay
vectos
9th tiry poe-muttpymg
ce. Suf (1=l2,--n), The
-Valeigoe i
conporent ficjentby by in
A, neletions, depen whch
tandn we
lage we
of obta dend ane be
fn 31e get both
v P then
 often det, ofa
Detnine
$9, TICE, 15ugeh
matn. ne
Aof fst Vol e feme
1he
Ieast igenpowen the to 1, rnei
Aah Laongt Hbe met
Vae
Of
A
If eigen eild
dene
laogest A f Ofsane the t a
eiglhvalve a be id non- leastThe 0f
method. 4wlt wevle of
cUUSPONdngeigen
Then mor- CEi gen the ally and
egon provjdedcon whouo
a Sihgwlan to
vole gies The Siindrgula tne lest
eastgrncakst ofmat dekomy
value
detovjnnmo0Cal
e y the vectas find the That,
ot geei eiger gehvalies vatietmmageIgen
mnyconly . 1A| +o
The nvese
poCo ofMveoye
4 if mat.ne
matiz eigenyaey
yecte vale
numejc9,aly emayn valne di
mumejcal
they b-e
Sey,
of matX
leasi
A. (rwnerjcalg) mati of
me
tiy we same)reNins of C t
a B