Scribd Logo
Upload

Welcome to Scribd!

  • 34 views

    Project No. 02: Submitted By: Khalil Ahmad

    Uploaded by

    AarfeenKamal
    This document describes a numerical method for solving the one-dimensional heat equation using a finite difference scheme. It presents two Matlab programs: 1) constantk.m uses an explicit conservative finite difference scheme to solve the heat equation with constant thermal conductivity k. The temperature distribution approaches a linear function over time as expected. 2) heat_variablek.m modifies the previous program to use a non-conservative scheme for a heat equation where k varies with position. The limiting temperature distribution is no longer linear due to the varying k.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd

    Project No. 02: Submitted By: Khalil Ahmad

    Uploaded by

    AarfeenKamal
    34 views7 pages
    This document describes a numerical method for solving the one-dimensional heat equation using a finite difference scheme. It presents two Matlab programs: 1) constantk.m uses an explicit conservative finite difference scheme to solve the heat equation with constant thermal conductivity k. The temperature distribution approaches a linear function over time as expected. 2) heat_variablek.m modifies the previous program to use a non-conservative scheme for a heat equation where k varies with position. The limiting temperature distribution is no longer linear due to the varying k.

    Original Title

    Project 2

    Copyright

    © © All Rights Reserved

    Available Formats

    DOC, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document describes a numerical method for solving the one-dimensional heat equation using a finite difference scheme. It presents two Matlab programs: 1) constantk.m uses an explicit conservative finite difference scheme to solve the heat equation with constant thermal conductivity k. The temperature distribution approaches a linear function over time as expected. 2) heat_variablek.m modifies the previous program to use a non-conservative scheme for a heat equation where k varies with position. The limiting temperature distribution is no longer linear due to the varying k.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    34 views7 pages

    Project No. 02: Submitted By: Khalil Ahmad

    Uploaded by

    AarfeenKamal
    This document describes a numerical method for solving the one-dimensional heat equation using a finite difference scheme. It presents two Matlab programs: 1) constantk.m uses an explicit conservative finite difference scheme to solve the heat equation with constant thermal conductivity k. The temperature distribution approaches a linear function over time as expected. 2) heat_variablek.m modifies the previous program to use a non-conservative scheme for a heat equation where k varies with position. The limiting temperature distribution is no longer linear due to the varying k.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    of 7

    Project No.

    02
    Project Title:
    Submitted By: Khalil Ahmad
    Submitted To: Dr. Khalid Pervaiz

    Dated: 1!0"!201#
    $%&titute o' &(ace Tech%olo)y $&lamabad* Pa+i&ta%
    Conservative FDE Scheme
    The% the (artial di''ere%tial e,uatio% become& 'or co%&ervative &cheme

    u
    t
    k
    u
    x
    =
    2
    2
    -here u.x* t/ i& the tem(erature at time t a di&ta%ce x alo%) the -ire. 0e-riti%) the (artial di''ere%tial
    e,uatio% i% term& o' 'i%ite di''ere%ce a((ro1imatio%& to the derivative&* -e )et
    u u
    t
    k
    u u u
    x
    j
    n
    j
    n
    j
    n
    j
    n
    j
    n +
    +

    =
    +
    1
    1 1
    2
    2

    .
    Thu& i' 'or a (articular n* -e +%o- the value& o' u
    j
    n
    'or all j* -e ca% &olve the e,uatio% above to 'i%d
    u
    j
    n+1
    'or each j:
    ( ) ( )
    ( ) u u
    k t
    x
    u u u s u u s u
    j
    n
    j
    n
    j
    n
    j
    n
    j
    n
    j
    n
    j
    n
    j
    n +
    + +
    = + + = + +
    1
    2 1 1 1 1
    2 1 2

    -here s 2 kt3.x/
    2
    . $% other -ord&* thi& e,uatio% tell& u& ho- to 'i%d the tem(erature di&tributio% at
    time &te( n41 )ive% the tem(erature di&tributio% at time &te( n. .At the e%d(oi%t& j 2 0 a%d j 2 J* thi&
    e,uatio% re'er& to tem(erature& out&ide the (re&cribed ra%)e 'or x* but at the&e (oi%t& -e -ill i)%ore the
    e,uatio% above a%d a((ly the bou%dary co%ditio%& i%&tead./ 5e ca% i%ter(ret thi& e,uatio% a& &ayi%)
    that the tem(erature at a )ive% locatio% at the %e1t time &te( i& a -ei)hted avera)e o' it& tem(erature
    a%d the tem(erature& o' it& %ei)hbor& at the curre%t time &te(. $% other -ord&* i% time t* a )ive%
    &ectio% o' the -ire o' le%)th x tra%&'er& to each o' it& %ei)hbor& a (ortio% s o' it& heat e%er)y a%d
    +ee(& the remai%i%) (ortio% 12s o' it& heat e%er)y. Thu& %umerical im(leme%tatio% o' the heat
    e,uatio% i& a di&cretized ver&io% o' the micro&co(ic de&cri(tio% o' di''u&io% )ave i%itially* that heat
    e%er)y &(read& due to ra%dom i%teractio%& bet-ee% %earby (article&.
    The 'ollo-i%) 6!'ile* -hich -e have %amed cnstantk.m* iterate& the (rocedure de&cribed above.
    Matlab Program:
    function [ output_args ] = constantk( input_args )
    %UNTITLED Summary of tis function go!s !r!
    % D!tai"!# !$p"anation go!s !r!

    % D!tai"!# !$p"anation go!s !r!
    $ = "inspac!(%&'&'())*
    t = "inspac!(+','-))*


    . = "!ngt($)*
    N = "!ngt(t)*
    #$ = m!an(#iff($))*
    #t = m!an(#iff(t))*
    k) = )*
    k( = /*
    k = (k)0k()1( *
    s = k2#t1#$3(*

    u = 4!ros(N'.)*
    u()'5) = sign(pi2$)* %init*

    for n = )5N%)
    u(n0)' (5.%)) = s2(u(n' /5.) 0 u(n' )5.%()) 0 () % (2s)2u(n' (5.%))*

    u(n0)' )) = +* %6#ry())*
    u(n0)' .) = )*% 6#ry(()*
    !n#
    figur!())
    u7a"s = (' t' $' u()'5)' [+ )]*
    surf($' t' u)
    tit"!(8T!mp!ratur! in 9 an# t using cons!r7ati7! sc!m!8)
    $"a6!" 8$ in m!t!rs8* y"a6!" 8tim! in s!c8* 4"a6!" 8u in :;8
    figur!(()
    p"ot($'u)
    tit"!(8T!mp!ratur! in 9 using cons!r7ati7! sc!m!8)
    $"a6!"(89 in m!t!rs8)
    y"a6!"(8T!mp!ratur! in :; 8)
    !n#
    Matlab Plots:
    Result Discussion:
    Thi& loo+& much better7 A& time i%crea&e&* the tem(erature di&tributio% &eem& to a((roach a li%ear
    'u%ctio% o' x. $%deed u.x* t/ i& the limiti%) 8&teady &tate8 'or thi& (roblem9 it &ati&'ie& the bou%dary
    co%ditio%& a%d it yield& 0 o% both &ide& o' the (artial di''ere%tial e,uatio%.
    :e%erally &(ea+i%)* it i& be&t to u%der&ta%d &ome o' the theory o' (artial di''ere%tial e,uatio%& be'ore
    attem(ti%) a %umerical &olutio% li+e -e have do%e here. ;o-ever 'or thi& (articular ca&e at lea&t* the
    &im(le rule o' thumb o' +ee(i%) the coe''icie%t& o' the iteratio% (o&itive yield& reali&tic re&ult&. A
    theoretical e1ami%atio% o' the &tability o' thi& 'i%ite di''ere%ce &cheme 'or the o%e!dime%&io%al heat
    e,uatio% &ho-& that i%deed a%y value o' s bet-ee% 0 a%d 0.< -ill -or+* a%d &u))e&t& that the be&t
    value o' t to u&e 'or a )ive% x i& the o%e that ma+e& s 2 0.2<. Notice that -hile -e ca% )et more
    accurate re&ult& i% thi& ca&e by reduci%) x* i' -e reduce it by a 'actor o' 10 -e mu&t reduce t by a
    'actor o' 100 to com(e%&ate* ma+i%) the com(utatio% ta+e 1000 time& a& lo%) a%d u&e 1000 time& the
    memory7
    Conservative FDE Scheme
    =o%&ider the %o%!co%&ervative o%e!dime%&io%al heat e,uatio% -ith a varyi%) thermal co%ductivity
    k.x/*

    u
    t x
    k x
    u
    x
    k x
    u
    x
    k x
    u
    x
    =

    = + . / . / . /
    2
    2
    .
    >or the 'ir&t derivative& o% the ri)ht!ha%d &ide* -e u&e a &ymmetric 'i%ite di''ere%ce a((ro1imatio%* &o
    that our di&crete a((ro1imatio% to the (artial di''ere%tial e,uatio%& become&
    u u
    t
    k
    u u u
    x
    k k
    x
    u u
    x
    j
    n
    j
    n
    j
    j
    n
    j
    n
    j
    n
    j j j
    n
    j
    n +
    + + +

    =
    +
    +

    1
    1 1
    2
    1 1 1 1
    2
    2 2
    *
    -here
    k
    j 2 k.a 4 jx/. The% the time iteratio% 'or thi& method i&
    ( ) ( ) ( )( )
    u s u u s u s s u u
    j
    n
    j j
    n
    j
    n
    j j
    n
    j j j
    n
    j
    n +
    + + +
    = + + +
    1
    1 1 1 1 1 1
    1 2 02< .
    *
    5here
    s
    j 2
    k
    j t3.x/
    2
    . $% the 'ollo-i%) 6!'ile* heat_variablek.m* modi'y (reviou& 6!'ile
    to i%cor(orate thi& iteratio%.
    Notice that k i& %o- a&&umed to be a vector -ith the &ame le%)th a& x* a%d that a& a re&ult &o i& s. Thi&
    i% tur% re,uire& that -e u&e vectorized multi(licatio% i% the mai% iteratio%* -hich -e have %o- &(lit
    i%to three li%e&.
    Matlab Program:
    %% <!at !=uation u_t=u_$$ % !$p"icit finit! #iff!r!nc! sc!m!

    function !at_7aria6"!>
    % D!tai"!# !$p"anation go!s !r!
    $ = "inspac!(%&'&'())*
    t = "inspac!(+','-))*
    . = "!ngt($)*
    N = "!ngt(t)*
    #$ = m!an(#iff($))*
    #t = m!an(#iff(t))*
    k = () 0 ($1&)?3() *
    s = k2#t1#$3(*
    u = 4!ros(N'.)*
    u()'5) = sign(pi2$)* %init*

    for n = )5N%)
    u(n0)' (5.%)) = s((5.%))?2(u(n' /5.) 0 u(n' )5.%()) 0 ???
    () % (2s((5.%)))?2u(n'(5.%)) 0 ???
    +?(&2(s(/5.) % s()5.%())?2(u(n' /5.) % u(n' )5.%())*
    u(n0)' )) = +* %6#ry())*
    u(n0)' .) = )*% 6#ry(()*
    !n#
    figur!())
    u7a"s = k' t' $' u()'5)' [+ )]*
    surf($' t' u)
    tit"!(8T!mp!ratur! 7ariation @it t an# 9 using Non ;ons!r7ati7! Sc!m!8)
    $"a6!" 8$ in m!t!rs8* y"a6!" 8tim! in s!c8* 4"a6!" 8u in :;8
    figur!(()
    p"ot($'u)
    tit"!(8T!mp!ratur! 7ariation in 9 using Non ;ons!r7ati7! Sc!m!8)
    $"a6!"(89 in m!t!rs8)
    y"a6!"(8T!mp!ratur! in :; 8)

    !n#
    Matlab Plots:
    Result Discussion:
    $% thi& ca&e the limiti%) tem(erature di&tributio% i& %ot li%ear9 it ha& a &tee(er tem(erature )radie%t i%
    the middle* -here the thermal co%ductivity i& lo-er.

    You might also like