- JAbtJJ_X_T2

NASA CONTRACT 9-I043_.5

MAL
3.
3-
_. ._ Prepared fer:
Nattoeal AeronauticsandSpaceA_tntst_'attorr
NannedSpaceeraftCenter
Under Contract NAS9-10435-

--4

J
- Prepared by.."

Envt ronment_l and._hemal. Systems SectJOr_............. !r

ARpt!._.!:Plech_ tcs_.Department

? Concur_nce:
__ _;
F

" i

-- 4
 I_RE_RD

: Thts theme] ne_ork mode121ng handbookwag generat_J-under Task No. 4 _ ._

of-_ASA Contract HA_g-10435 entttled "AdvancedSINDP_ThermalAnalyzer_ : i
0evelopment..".
/ i
IJ_derth_ota] COn_rac_L_three_,taskswere dewted to documentation:
_askZ(o. 1 for: develOp
men}of-l_ze UsertsJ4aOua], TRYtRepo_-No._1469_-HOOt,
RO_QQ,April T971,.Tas_ No. 2-fo_ developmentof the Prog_m HanualL,T_ |

P.epo__N_ 14690-HQOZ_RO.OO,__une 1911, and.T_sk_No.4_for dev.e_]o_nt, of-

ti_e ThermalN_ta_l_kHode]ltn9 Handbnekconta4ned.here_n,

- Contrasted to the User|s Henual and_O_gramManuaiwhich describe the

manuscrtpt describing, a! I _e cemp1extttesof _al: Halfiemattca1 Hode111ng

was far beyondthe scopeand budgetary alloea_ions of this task. Hence,
lil this docu._nt was fnten_ed-te
ComRleteusageend content of serve as .an tnltta] bufldtngablock
the SINDAComputerProgram, fer a
camplete •
_: _ futut_ comprehensive,document. Ustng th4s guideline a three-fold purpose
was est_bltshed: (|) te acquatnt the newuser wtth the terminology and
conceptsused in t_ermal m_themattcalmode3_t"n9,(2) to present .td_j0re
ex_e.rien_edand occasional use_ vrl_b qu_k fo_ulae and methodsfor solving
evez,yda_problems_Ceup_d_wt_lzstudy cases whfch lend t_sfght tn_o the
rela_onshtps-that.extsL-among_J;he-var<ou_solu_lon techniques and p_m-
ete_s, and _3_ te begtn- to__atalogue in an o_derl_-fashto_ oommon
fomu_ae wlt_cb_t_the f_pZt ed t_oau_amatedcnnversat-_o_.al _
languagetechniC. _i1

i
)

$
 /

TABLEOF CONTENTS -)

I 1.0 INTRODUCTION T
-,L,I_
. ............. " ......... ' " " "---"
; i 1.1 Thermal Ma1:hJ_odeitEj as a CognitiVe Proces_.s_.,_..__;._._.:
......... 11....... i
--- I 2.0........
.THERHAL..HATHEMAT_CAL .
MODELING....................................
3 '

-:
! 2.1 Network.
Sotul;ion-._.
, , ...............................
3.......
z _
I 2,_ _ModelI ng £]_emen_ ................... L.......................... 4_
--i' 2 2.1 Nodes 4
I
..l : 2.2.1.1 Concepts ........... , . . 4
2 2 1 2 NedeTypes 5
2.2_I_3 Method of NbdaltZatton ....... . 7
_ 2.2.1.4 Computational Methods-Nodes.... 11

. 2.2.2 Cnnducto_ ................. 12

--l:
)
' 2 2 2 1 Concepts 12
2.2.2.2 Conductor Types............ 12

; i 2.2.2.3 Computatienal Metheds - Conduct'ion

CondUcf,
ors ; _ . . , . ; . . . ,---. 15

i:_: 2.2.2.3;1 Rectangular" Nodes, .... 'f5

_ 2.2.2.3,2-- Cl_c_larSe6ttons , .-.- .......... T6
! 2,2-,2.3.3- Para11_ Conduetars .... ,. T6_
2 2__.__S&r_es
_)dueto_ 17
:- -- 2.2.2.4 C_mpUl_atlonal Me_t_ds -(:onVec_tow
- " Conductors i-8-
I

--_ 2":2.2.4.1- Comb4nedNa_-_u_nd ..

__;_ For('ed:Convec4;To_-, , . . , 19

:-_.' E..2.2.4,2 Natural-Con--on

Eq.uatton¢ . .-.. _ .-.... Z-T-
- i
2.2.2.4.3 Forced Conve_lon Tn Tubes- -
-. and DuCts--,- .... ., .-, -, . 26-

Pla_;es- 31
....... l_.__)
.'..

- - o

I: -".f " : :.... " """_

....... " ......... = ..............
:'-_
" ---- t
 I]

Page
2.2.2._.5 ,.Fo_:+)dConvectien
Cylinders 33 "
J

2.2:.2.4;.6 fenced Coavect.ton iJ

Over Sphe_s.......... 36 -i
2.2.2-.5:_ComputaCtena_Hethods - RadiaCTon

l
l
'_
Condu_.tors
".+--..1._- **..+4+ A)--+ _ • • -_ • e+, p..Im_-- ++ 38 i.-.f
( :1

2_2.2+6..+Comp_tat1+onal
- Hst_odSl--Hass:FLow . .:
Conductors-+_.

_+2.3+..Energy-
SouPces
.or. Sinks,..................... 5B+ _._ .. :

2.2.3._ Concep.ts................ 58 , 1
2.2.3.2 Typesof Heat Sourcesor Sinks. . . 58 ; I
2.2 +3.3 .ComputattonalConstderattons - i
Sourcesor Stnks....... - ; ; _

2.3 NetworkSolu'tlon
.............. 59 :-i

2,3.1 Steady State ............... . 61 :

_ 2 3 2 Transient Analysis 63
2.3.2.1- ForwardlYl_erenctng......... 63 . +!
2.3.2.3_ BackwaPdDifferencing ....... t_5

2.3.3 Summ_ry:
of-Other TeehntqlP.s..... o . . . .... 67 t
++ • _ i++ • • d • • • 6 • • • • .*+,.
2+4-l'4ode
I'I.
ng-Par_Ime-te
rs .__.68 ........ i
3.0 OPERAT:[ON_
-
PAP_'_'ER-
R ELATIONSH
t PS. • • • • 6 • e, *t+ ",b • ++.----.,----
20.-+
'i
3_10ne-DSmenstonal Ba_of- Metal 79 t
3_ Other One_J_n_nstonalCases_
3.3 Two-DtmenstonaJ
P%_teoT:_gCat.. , ......
• • • in-q[-- • • • •

_-.
• *- .

. ,
• /8
78
t ti
3+4 Other Two-Dlme_st0nalCases . ........ ; 79 ]
4 0 REFERENCES - 82 1

V
 I

v_

...... ........ : :---- !-- _ --

 t ¸ _.'_

, ,:_.1 LIST OF TABLES i

:._!_ Table , Page

Z-2 Orde_ of Magnttode:_f CorLvectt_e

Heat T_'an_er-- ' _

,:..._:_
_/t _
:'" 2-3:

2-4.
gm_t_ cat Constants-for Lamlnar-E]o_LOvera C_]/_nde_ ........

E_fectlve EmT .t_mce for Rat_Pl_tes--_nstd_-._ Black

Enclosure............. ,.::............... , ......... , . , .
35

51
:
J

C,' J-

] o - .

:--L'" ".-"

vt-N_

_ -_: = , .... Z _. • F" T -" "-- . = - " -

 NOHENCLATURE

A • APea
C----= Thema] Capjctt.ance
Cp = SpecTftc-hea_c
D - DJameter
DH " H.ydrau11cdtameteP
E - Vo|tage - : -
F_..-: Rad_atCOrLconftJju_tton. (form) fa_¢._tQ_¢
......
G - ThereaT-conductance
h - Convectiveheat transfer coef._tctent-
L - Lengthor runntng leagth
! = Current
k = Themal conduettVlty
- Hass flow rate
n = Arbtt,rary- exponer_t-
*,L_ N - _be_ of Iter_ttons
- : Heat-ra_e
r = Rad|us
R = Resistance

T - Temperature
T_ - SurrounCing-med4aor f_Pee_
stream temperatu_e_
U-.-- Veiocl-ty ....
U® - Free s_eam vetoc|ty
V- r VoI Urns
W _-FlOw-rate
ws- - Sa._ltng f_equeney-.
wc " Haxtmum frequency,coml)_nent
x * ArbCteary d4st_nee-

tx
 NOHENC_TURE
(Concluded)

_T =---Te_eratu_e d4f_erer_e
_" = Scrlpt-F (g.._..V_.b_e__...f_m
fad.or_)

= d_fustvtt_
-__._Thermal;
= Ceefft_ent ofVoTumetrl¢-expansfon .
p = Dens_
o • Stephan-Bo1_Jgaxm constan_
= Stab_1"Jt,
y- fa_'tor-
c = Emlt_n_
e- = Mole
x = Radtatten- llnearlzatlon
faclor
= Convergencecrltevlon-(relaxatlon criterion)
¢ = Oamplngfactor

S_bols, subscripts and units net spec_fically inert'cloned

In the
Nomenclatune._are
explalned at the point of usage-wlthln the te_t.

• a/ " " - "" " .: " "I -:

 - 1.0 INTRODUCTION

1.1 Themal Hath Hedelt,ng-as a Cognitive Process i!; "q

_ A b_tef Introduction to _e-rudtmontaw toch_t_LUes of thema] medeltng )
coupled._tth a-sTmpl_eunde_tandlng of the variOUS-baSleheal t_ansfeP i!_
mechanisms
a_e:_e prerequisites fo_-a begtnnl,_g-thenhalmath.mede_e_. J!

Properly applted_ the_ody o.f concepts,_eciptes and techn_ques-applt_abIe............._!

to: thermal.._th_ modeltrig constJtuCes a va_ d engtnee_ ng-too]. whfch-.canbe. '
t

apl_lfed_¢othe so]u_on of re_t..e_ineertng problems. A good..lumped param- _

eter represe_aL_on.o1=Lthermal svs.temreclPi!res_tnaddtt|on to the baste _i!
pPtncfples-ancLtochnlques, an-elusive mtxt'uroof experience.(wt_h =eal _
system,_ both physical -andmodel) and EngtneerSngJudgementto transfer the )i
end produst tJ_toan aCcurate, vePsaCtTeand cost effective themal math _

medel" il!

_. ,_ •_ GeneraT_y,the
reduce-re an overalTproblemsenceunterod
ob_ec_Of aeht_vtngInthe
develOplnga thennal._ath
greatest accuraw modeL-
fOE the-Teas_.... |'li
cost. CosL._acto_ are rather well d_ftned, and fall tn¢o _o ¢l_sses: _!
4_
(1) the .cost ef developing the nledel, and (2} the cost ef ustng the model._
gevelepmentcosts _an be ba_eda_mostsolely on the actua] engineering man,
-- po_er reqg|red tO de the-_ob wtttdn, the co_¢tratnts ef trtme-and budget.
Ho_ever_the potential costs 1evolved t_ using a mode]am of-ten no$ as
obvtbus-ner as- I-rheas._ F6_-examp;_mostChorea1math-modelsw_11._b_,.n_
l_zed on e eomputerwhteh ts highly W_LorffLt:zed.__ T_ such an env4ronment;
= a-totaL1 ve-ly-small increase t n requ_red-processor time, _perbaps--from5_-
-- minutes--to 15 mtnutes, o#-tenresuits tn a reduced,teb pt'_ort-ty and a ¢orres-
= pending-slowdowntn turnaroundttme_ peehaps__rOm Oneday to oneweek. .
The problemor-achieving accu_ac_fn a mathmodel, wh_3esub_ee_to
:-- varies g_eatly fron_en_ themal math model to anether-,
cos-t-constraints,
-'. GenePaTaccuracyrequtrementsmaybe-as stralght for_a_d as "tL=mperature-
:: accuracyshall- be-compattbte-_t¢_ themocoupteA/I) converter quanttia_foh
error, '=or _,.. thdmostat Wstol*es_s." On the _ther hahd_ accuracy lave-l_

i
 -; I

:_-"_ mtght be Indirectly indicated by:requiring that a model must "be su_J¢tent-
•.: . j
__2-,! ty det_tied to pe_dt meaningful _ra'a_trtc anaiyses-wt-th rasp_act to tnsu-
:--'i latJon thickness variations tn tncremeCs of 114 tnch," Clear_y, theL_-.ts
_-- _ goTng to be a-gre_tde_l o_ Engineering Judgement Involved in 4eveloplng a
-; _.,_ model..that._ls "sufi_cle._y detatled"..to be "meanlngfuI_"

__ Succeeding sections of Zhts _epgrt vdI1. present many of the_basic

- : pNnciNes and.techntques:_volved_tn t_ermal maCh-model-$ng_Exper-lence,
-" " o£-coucse,-carL onlY_be _c_t_.cL from hands.on_famtl.tart.ty wTth-_eal theist
"] systems, a_'d_pattie1 patton :in the modelin& and analyst s--thereof;_ Eng"meertng
"::_. !. JudgemenJ:-cer_
probebiy:be described mo_e_accur_tel_t as__the_._t'es.u'r..t
of zb-
selecting f4_n_the-body of untque femtltar Infoma.tlon, a gestural under-
2:2-' standing wMch can.be-extended to gutde the4nvesttgatton and- comprehension
of new_and_fem,!tar areas. As.such, Eng_eertng .J_dgemen_cannot be
-- p_esented tn a table, or-a £tgure, or even an.e_ctre-boOk. However, It ts
possible to pRsea_t _ collation and dtsc_ton of examples of the_'_&l••math
modeltng whtch wtll serve, not as a source of p-acttcal ezpertence, but as
:_ _ a body o_ unique and famIHar-tnfo_,atton from whtch the reader may abstract
:. . as-mgch E:ngtneertng Oudgemnt a_.:he ts _ble. Such a coTlectton ts tncluded
- :- in-the latter half of this report.
°:.

.... i

: i
I

._ _
i
:.% I

;2

/ "=-- t
7
_JL ......

._.0 THERMAL
MATHE_TICAL
MODELING

2.1_ Network Solut_on

Twosystem-are sa|d to be analogouswhen.the_,both ho_e slmila_-equa.

¢tons and boundarycond_tons; a_l the-eqgaLlonsdescribing tn_.behavto_ of__
one syste_ can be transformed:tnCo__theequaCtons_fcLr_ th-_oth-_r by-s_mply-
changlPg_S3mho]_;o£¢he-vartables, Themal and_elect_ical ,,,ystems-a_e-two
such analogops_.sysCSP_s
_s shown-tntable 2-1 ...............

Quanttty Themal E]ectri,'al

System System__
Eotanti al T E•
...,,, . _

Reststance R :_
]
_ __C°nductance G
Capacttance (; e-
gg
Ol_nJsLaw- __: GT I -

Tab|e 2-i. Thereat,EleCtrical Sy4;_era

Analogy

T_e anatog_betweenthemal-and el;eetHeal_ sTstem"aJlows the engtEneer

ta_-ut;ll_m--tk_-vddel_knownbaste laws-such as Ohn_sLawand Kt_-chhof-f-_s
Lawsuse_f01" balancleg ne._woPJcs,
Nume_cal teehntques used te solve the
pac_"lal d_ffecen_ equ_d_tons descHbt,g suck_;y!;tem ha.vsbeen conveniently
adapted to c0mpu¢ersOl_t4ons, thus--enablingtN_ engJeee_te fead_l-y _r_pu_e
temperature d4st_tbutt ons ana g_'ad_nts-ef cem_lex-phys3ca1 thermal networks.

Thema4-analyze_ co_uteP pPeg_ams-have-bten developedwhtch requt._re

the-useP to deftne _ therma3-netwnp_of-_he-_ stem-anai-ogousto an elec-
trtce_ _rcut¢. The .etwo_ _omponents-_Pef_put tnt.b _ computeran#
pre-progran_edmuttnes perfoPm _the trans_enf o_ steady state-sot_l:tons.....

3
 nume_Lcaltechniques for sol-vtn;-th|S network.

...................
-2.2... ModeTt:ngL.El-ement_ _
2;2-J_ "Nodes
iit TJntssect_FOn-dtscusse._
the development-of-a thez_na-l-netwo_k
and the
2-.2.1.1 _- . i.:_
Zn_o_te_-to-_leve]op-j_:l:hermaT
network:and:applynume_cal:,_echntques-
tO. _#._sotut_on_._l__necessazy, to.stdzd:htt_e the .theme]- system tpJ_o--a
: " - "_i
numbe_o_:.£$n1_ce:.suhVolume_cai1.ed:nodes.---Thethenna]-p_oper--t]es_of-each
....
node are-cons1_ler-ed..to-be--conce_r_ted-ar.-the:_centr_a1:.nodal
polnt of_eech__ ..............
subvolume, EacJ_nocle_repr_sent_, twO._thermaT network eIements,._atempera-
l:ure (potential) and _.capac_anee (theme1 mass) as show_-tn_E_Jul___g.] ............

. .•, , .

• - _ "': II

• ; L o . - :_-•

• __, -. _.._. _,(,_ (node 3

!,!_1 ............... :

Ngure _-I. _lodalt_catIon.=__

• .. The temper_u_e_ __i_tgn_l-_O _qnode-eep_esents-the-average
mass "-

,_: ii_I _-
computed
temp_-_f
rom-th.li..e0_plslcet--propecleso.r
th. subvo_tle._
.-
theubVo,meeter-
T_n.capae4tanOe__._ as_lgned-to-a
_-, .. node
-, e:_a_ate_Lat: the temperatureo"f_the-node_,_ Is aSs_med_)b-e-concentrated
-L_. at th_ _oda_ center-of:-the_subvotume,jecause a-node-represents a-_,_mp1._qg-
;:: -___:=__ _6f parameters :t_:_ll_gTe
. point- tn:-spac_: the temper___ur_d_stHbu_4on
-- ; th_ough-t;hesubvolume-tmpi_cL_: the nodal--tempe eatu_e-ts l_rnear--asshown---
-;=_i--" ......... tq_-;,'tgure/2,2c_a_d-nota stepfunctto_ as tllu_r4_e_-T_ Flgure_ .... [:

_: ":'_-....

-] .
 STEPFUNCTION
ACTUAL EACHNODEAT LUMPEDPARAMETER
DISTRIBUTION UNIFORMTEMPERATURE SPATIALPOINTMASS

IL • •

T . "_" T i/

x x x

(a) (b) (c)

Figure 2.2. TemperatureDistributions

In a homogeneousm_terial, the temperature ai_ a point other than _he
nodal point maybe approximatedby interpolation betweenadjacent nods1
! ") points where f_e-_empe_atuPesare known.
The e_or introduced by dividing, a systeminto finite size nodes rather
than-volumesdW3 whe_e-d_aiLo_oadhes zero is dependenten numerousconsid-
erat-ions:Mater4aLtherma1..p_eper_les_ bound_ry_cond_tlons, nodesize,_
node ce_er Rlacement, and time..In_rement-eftr_nslentca-leulaC4ons.The_
techniques fo_-preper noda_tza_ton-_o_ntmtZe-th_ error-w_ll be discussed"
aterSection._.

2,_J_.Z NodeTypes

To _his po4nt-ont_-nodeF-wh._chrepresent subvo3umes w_-tha- ftn_e

t_emal mass_capaC_tance)have bes_ dts_. In manyTnstan_es=two
other f,vpes of node_ale require_ td define a therma_ network, They are
nodeshevfng a zerQ-eapacltenceor an-tn_nt_e Ca_ci_anee, Themal
analyzers such-a_ _NDA usuall_L_ the th_ types of nodesparticular
namesas follows:

. 5

-.
 NodeT_pe N._____ )
F-tn_te Ca_ct.tance_.... D1ffusten
Zero.Capacltance Arl tl_e_c_
Ip_E_lY:eCapacltance Beundary

The..dtffuston node has a ft,tte_ capa_ttance,and ts- used.._o .repre=

sent noma] matertal.nodallzattor_.._..Zt-is-characterized.by a. g_a!_.n_
o_..
loss_of potentia]-energy whtch._depeed_..on the capacttance value-, the net.
heat fl.ow into the node,,and the _tme over.wbt.ch.the..heat t"s-ftowtn9.....
Hathem_L|.cal.ly,.a._dtf.fus_on:_node.
is:. de_Ined-by-the exp_ss_on_ .........................................
i..............

The artthmet-lc-node (ze_ capacitance)"Is a physlcalTy unreal quanta,

howe.ge_,its effective use wtth numerical soluttens can often be helpful tn
tnLe_pretln9 results 1_ suchappltcaL-fonSas surface t_n_eratures, bondltne
te_eratures, and node cOupltng temperatures. _ else finds use tn repre-
senting themal s_stomelemen_ which have-small eapa¢ttance value_ tn
comparisonto the la_je majority Of- the otlm_-flodes_n .the s_stemwhtch
)
results- in cemputerrun ttme r_ductton with m_norchanges tn preen11
accurac_Y_Examplesof these eould be small compO_sp_s such as bolts,
ft-lms_ or--T_llets; gaseous.contentsof sma11ducts or-tubes; end-]ow mass
tnsul-atfons. Arithmetic nodesshould be few |n numberwhencent_asted to
the fetal numbor-of-:nedest_L_the_k._l_e temperature_of an a_lthmett_ ...........
node respopdstns_ntaneously_t_ surroundings,. Hathema.t_-a_ly,aw
arithmetic nede is de_ned by the expression:

zQ = 0
The beunda_ node{tnftn_te-capacttenee) ts used-to repx_sen_-_enstant-
_emperature_soureesw_thtn _ thermal nel_vorlc. Common uses-are_ deep
space stnk temperatu_, recovery _peya_ce; luna_ suttee t_mpe_ature.
In addition, a-boundar_nodem_y-be-Usedto represent thermal system _om-
ponents suchas th_ bulk"prepellant: _n a Targe-tank _h_t_ hast very--.
l_rge-themai _nass(capa_ee_ Hethem_ttc_11ya-boundarynode_fs
defined as:
T _ _onst_nt
, )

_J _..= - ..... - .: " : I ---

 3

.;I 2-.2.1;_ Methodof-Noda!_atton

i_: -I i The p]acementof dtffuston node centers and-the.chotce-o_,node_shapes 11

) ) dend.t on fa s: (1> pot.
i ::.- I " desired, (_) the-exgected temperatUredtstHbu_on, (3) [ahyS.tcaLreasonable- (

dependenton other-const_erattonsJ (_i) accuracy des_._L, (2) _tructural _t

deslgn._(3) COml_ute_storage_capebillies, and (_¢_comgutertime requl red. 1:
Eact_-factU_, however-_embodiesothe_:_o_stde_a_on_... For example, tO
anttctp_te the_expected:tempe_ature_dts"_tbuttenone.mustdraw heawtly on Ii
Entg__eertng.Ju_dg_ent:asto the effec_s_of-'the-exp,ected boundar__.cond!!;Jons
................. _.
• _- and..assoctalmd
metert al p_p_r_ttes, i
- _"_ In gener-al, the shapeof a dtffuston-nede is chosento .be-a sfiapte )
:-_ geometric t_tgurewhoseames and volume can be easily calculated. Irregular- i-
:: ly shapeds_uctural membersmay-beapp_exlmeted, by__stmple-shapesby era- i
'-:-:-' ploytng as_npttens that are censlstent •with the desired resuE_. Node ,

: _:_ i:_ centers _re assumedto be located at the_mass

first, centeotds
w|th the of thecenter
nodes. In
)-_::_ somecases,-nodal divisions are dectded node locatqens
being thusly defined as a consequence. In these cases, nodal-I_unda_les
wtlT usuaily lie along-structural boundaries, and-structural members _tl1 be
.- divided t__a s3qnmel_Hcand equal--fashfon_ In other _L_es, OUtltUt r_qut_e-
men_sw_Ti dictate the _ocations o/_node centers, with the nodat boundaries
asSigned-as a conseqj_enee......These
two ePl_oa_hesare t'llustrated_n F_gu_e2-3,
:' .... [n the-flrs_ case, _gu_e _-_(__tt ts-des_red.-tOp_.llare a genePatmodel ...................
) _ "_---- of--the-s_tm'e; but tn the secondCase, Fig_e-_.-3(b) _; is dest_d, to-
',_: :._ modelthe response of-the two-the_ocouples loo_d on the bondTifte between
the two members.

_i_
•_---_-"

7
_ i _---- ............................ • _ _q4

f -"

uu___ -......... • ............ m ........

(a)_.CJn_al Node-Boundaries (b) Node-_nters at _llermocoupte Loca_Lons. i

Ftgure 2-3, Altemato Rodal__zat,

ton-_Be.thods ;'
The_aboveexampie,al-ludes to a general -destrah|l-'Lty for_ reGtangular]_
shapednodes,Thts ts true-for, the_s_tmplereasorLthat_tt ts easy to.com_
puto the areas and volumesrequtred for the |nput calculatTons. Such
stmple node1shapesare tn keeptng wtth cu_ren_engineering practice.
By contrast,DuSt_ber_e_(refe_ence1) suggestedthat.nodelfzatlon be per- i
formedtn such a mannerthat th_paths of heat flow assumeda triangular
pattern, as showntn Ftgure 2-4(a), The only drawbackto thts theorett-
calTy soundapproachts tha_ tt remainsfor the engtneer to computethe I )
volumesof the Irregular polygonal nodesv_tch are the consequence
of such
a tact, as shownin Figure 2_4(b).

". _-- --

r._'-t , • Q"

L a, •

(a) _anguta_ Heat. _) Node Shapes ROlu_g, Co) SlI_le ReGtangutar-

Flow Paths from (a) Nodlt44z_tton--

Ftgure 2-4. Pol.vgo_l Nodi_11ZattonVs ReetangularHodal"iza_ton

Note howmuchstmpier tS the-rectanguJe_"nddallzat-_n approachtndfceted tn k

Ftgure 2-4(o). As mtght be expected, to-achieve-the Sames-_l'lcity* of i
ea4ceiat4on, 61r_utar st_uctores at_ node_lzed tn pie-shaped wedges_-_nnu-
l_r rtfigs, Or • coml_inatieno-f the- two, a_ show)_tn Figure 2-5. [--)

".-_ _J - " _ ,__ _ - . .,-- - - "__ I

 (a)- (b) ......
.(_....................................

Figure-2-5. NodaL_at_onof C_rcu)_r-Eloment_-

Bounda_r,odes_arJ___used
,.to _deflne pp.:l_. 1inns_ or sur__aceso_
_nstant _emper.atur_T__ .:.two,or- th_e.dlmens-1ona1_mode1_,resl_e-c=-..
tl.ve4y, The-phys_
caT-1oca_oe.ot:.:a bou_lary_..node
Is- detemlned, solely- by
the conduc_lonpaths connectedto I__ A-slngle boundarynodem_ be used
to model a1_.!_
bou_larles which are at. the sametemperature, Thls- po;nt Is
111ust_ated4n Figure 2--_whlcb shows-that the Indlcated bounda_ nod_.witl
suf_¢e as a modeto_ the ent.1_e, constant temperature edge.of..the. St_cture
(in thts ease, at-IO0°T:).

Ftgu_e_2_-6"_Sample8ounda_ _le
_-1 _.-_odes-have_a-numbe_of-use_h a_-I consequence-of..the-
fact-that_sucl_ nodes:se_'ve-_-anengTnee_|ng-mode3-o_-the: pmYerbt_L1"wafee-
d_-al)prOaches
o; tl_tckness-_lx___hem- zero".--A-¢_p_cat
al_ltca_or_lfes-
_n _he-n!odet:_-ot_exl_.H:or sur-faces__o_._reentry
vehicles _cl_ ars--
often_3.¢_ed:-to severe_ _ap].d_ chang*r,g,,boueda_ cond_Jons.-_ tl_
phy___.al-_]_ten__tl)e.
-s.uPface-ten_eraturs remefnsvery etose-tO r_d4_tt_r_
equ4_4br-[_ltwfth- the su_face=h=eotinf._tna_'Cattnf _h_s_e_n-be
acc_r_tely-s4m_ated, by the-use of:a _urf_e ar-fthn_.te node. Th4s apP_-
ca_l_n 4:d4Ttustrated--_ FYoUre-2:T.
O "'"

g
 dx_ _ 7n°des _.,,thmettc-
node _i -

L _node M"

%_
Ftgure 2-7. Use of Arithmetic Nodes--_.o
Hodel Surfaces
The case where heat ftows_from a surface by conduction ts usually one ,]
In whtch two structures are bondedtogether and a bondltne temperature 4_ i
seugh_. Whenthe structures are homogeneous_abondltfle temperature maybe- )
estab14shedby stmpte l|nea_" tnterpo]A."lon between-the nearesl node centers. _
Whenthe ma_ertais a_1 dissimilar, tt t_more appropriate to use an artth- :
mettC node at the bondline, leav4ng to the compu_.;.'r
the pracess of per-
fomtng a _onductance-we_ghted-a_erage uf l_e _dJotninzj-dJ-ffu_0nnode_
temperature whtCh.,tn-essence, t__the resut_--of_tndt_g-the steady state ._
(heat-t_ = heat Out)_t_mperet_re for_n 8r4thn_c node:

Ar__thmettcnodesmay-a_sobe-usecLadvantageousiy_-4n-ptace-of*-dt f_uston i
nodes=whiehhave a capaQtt_ce-that" i_ small .hen.eomparedto the g_eat-
m_or|_y o_;nodes tn- the sy._te_, Th_ -often-occurs when-modelt no a smal-1-
quan_lt_ el-gas fn • t_beor-ol_evenclesur-,_ or_hen model_9 stall struc_
tubal part_J_au(:h-as-_res, bol-ts, ft_13et;s,ftlms, and-shee_s-,where
detat-tsd temlmra_res are des_ed (whtch precludes lump|ngsueh-4temsaTons
wt-th-lacger flea_ nodes)_ The correct use ryr aHtl_mettc-nodest_ these
Ca_s generally resui_s tn a cons1_ab_e savTng-ef cOmpo_rt_ne whenthe
 2.2.i .4 Comp_!Ll_onA1
Methods- Nodes_

_L_ Zn develeptng_.athemaJ network, _mputattons-vdth respect to nodes---

ere generally 1trotted 1_ calculating the aap_act_anse
of- diffusion nodes-.
The following fomula ts used_

C-p • V . C-p_
where: ......
C = tbemal capacftance._BTU/%---
p - dens_ --LB_FT_
V _ vo]umL-FT _
Cp_T sp__t_tc heat - BTUtEB-_E--

The spectftc heat (Cp)_andtJze density (p) of rnaterfalS m_ vary wtth

temperature. The neceSStt7to ut_e temperaturedependent-prop_er_tes for_
ananetysis dependsen the de9ree wttt_vhtch_the properties ww. and_he
temperature range ever which _ capacttanoeof-the matertal w_l be calcu-
lated. Host theme1 analyzers can accommodate
temperature varying thermal
properties.

_]) ?he use of arl-thne_¢_nodesmayalso requtre somecomputations._Re-

placement.of small capacitance d_ffuston n_deswtth an artthme_c nod_mu_t.
be precededby computationstel vertfy tHat_the capac_:Lance-eonductor-effe_
are such that_Lhe nede-tn qu_ttOnwHl-essenttatly r_acb steady sta_e
tempePaturesdurdng the-time step mq_zed by_the _9e.t _ nodes. The-useof
on a_ltlunet_- node-to p_edtct_su_ace temperatures,where su_ace-radta_o_
o_--ver_hig__:_heaMng_J_tes--_retnvol.ved:v'equt.res.--care_ul-anaTysts
to t_sure--
th_ _tab11-_tyof-the ar4thmet1_-node_--S_ab1_1_-e_Iterla- and sotu_on--
te_hnt"qoes-a_e-dtSeusse_tn-seotl:on 2._. _om-th-_s se,ctten, tt:cen b_-seen
the_-s0_Lton_eehntques usSng:)J4)eaH:zed:"__ast-pass" temperatureValues-
mAY_ use of-analyzer:-cont_o_-c-onstants-t_o:_estHct-t.hemaximum
node temperature_hange" o_-compu_attor_ttmestep. The-engineer must fuPther
be eaut-tonedaga_n_ usln_-coUpTed._:arlthmatA'c
nod_ wfthout_4 _-ompl_6e-
understanding-of the tmp_.atto_ :and:requTre_-a_r-co,_rol e.onz4_m_s
used-_o ti_ure a va11_-_utton.

il
:ii
 _,**
_:_,.___.
_,**** - y "
.
.......

- r_ure _- _onduc_or-
Co__.ve_ton-
_s th_ processof _ner_y_Cransportby"co_btne_ct_o_ o_ hea¢
_onduct_m,-.enea_stePage__nd-mtx4_g_mett_n. Heatidril flow b_ c_nducZfon
_, a-su_fa_e-t_adjacent-partfctesoq fi-utd-_the_ufd particles _11 )

--- i

---" 12
moveto a mg4on e,f lower tea_eratu_e where they w_lt mtx with, and transfer
a part of their energy te, other flutd partteles. The energy ts actuatly
smrecl...tn..thef-lutd parttcle_ andJs ca_ted as a resu',t of their m'_s
mot_. F-_gu_...2_lO_tllustrates the-convection conducto_..............................................

Figure 2-10. ConvectionConductor

ConVeYorswhich represen_ condusttonor convectlen p_ths are _e-

fervid to as ltnear con_stors becausethe heat fl_ rate ts a function of
t_e temperature dt_¢erenca betweennodal te_eratu_e$ to the first power.

Radiation ts the processby which heat flows betweentwo bodte_ when

the bedtes are separated tn spaee. Energy ts transferred through electro--.._
magnetic-wave phenomena.Radiation conductors-.at'e-tenmdnOn-linear be-
causethe heat f.low be_een two-surfaces b_ radiation .ts a functton ef the
difference of-_he fourth powersof the surface- temperatures.

Q-G (Ti_-- Tj _)

Ftgure 2-11 Illustrates the.ra_atiOn c._nducter.

13
 FtguPe 2-11. Radga¢tonConductor

Fluid flow thermal systemsmayalso be simulated by thermal modeling.

Energystored in the thermal mass(capacitance) of a f]utd lump (node) Is
z_ transferred from one point to another by the movementof the flutd mass.
This type of conductor is generally referRd to as a massflow conductor
end ts illustrated in FtguR 2-12. The massflow conductor is also ( )
lInetlr

_- (_- G(Tt - T,t)

; k
I

Ftgu_ 2-i2. MassFlew _onductoPs

t '

-- 14
L
 2.2.2.3 Computational.._.chods- Con.du_ttonconductors
Conduct'toncenduetors are co,uteri from the equation:

kA
S=__
where.*
G = them_l conduc_ance_-BTU/HR@-E
k = the_ma.lconductt.vtty - BTU/HR--.FT-°F" i
A ,_ cross sectional area through which hea_ flews:-. F_2
L = _engl_Ll_e_.en.ad_!_n_ngnodes=_FT I_

The thermal conduc_f,,_y (k) of matertel s ma_v_, _1th _emperatur_ 11

or other tn_uenctng f_ctors wt__hinthe system; the cross-sectional a_ea
through which the heat flew_ (A)_and length betweennode centers (L) are
determined-by the stze and shapeof the adJotn_g:nodes. As wtth the
capacitance calculations, the necessity te use ten_eratu_e dependent-
properties dependson the de.greewtth wh_ehthe conducttvfl_ychangeso_r
"_, the tempera_urorange _xpec1_dduring t_e-analss1._,
2.2.2.3.t Rectangular-Nodes
Yhe length, L, of_ the heat tQowpatio_ usedf_ conducttbnconductance
calcu] at1ons for rec_anguTarnodesis the d_s_._,cebel_een_node-
centers,
and the area_ A, to be _sed-t_ the_cross_sec_tonal a_ea-pe_pendtcu)arto-the
11ne Joi_tng the node Cento_. The conventten-i_ deptcted tn F49u_e2-_%, f:

lr--_vv-4 !
_,kNt

Ftgufe 2-13. Slmp)e _enduc_r-Represen_n9 a Heat

F_0wPath through Hatep_aT

m t
 2.2.2.3.2 Ctrcular Sections
For conductOPc_etweennodesWhichape ci_cula_ sections, the con-
venttons showhin Ftgu_e2-14 should be used:

A= ed

" C ...In(r_o/_'1l

Where:
e radtans
d-FT
ro . consistentunits with r t
_V_ rt consistent units with rO

Figure 2-34. Apea and LengthEquivalents for

Ctrcu]ar Sectton Nodes
)
2.2.2.3.3 Paralle] CondUctor1

Twoor-m_e-papellet conduct|onpaths bet>/ee, nodesm_ybe sun_d to

cpea¢_one:conducCor__va]ue_by-the
followtn_ equation-_

61- =-61 + G2+ _.. 6n

Th_s may:be-helpful tn comp_uttng_an

p_egut-va_ent
cenductor betweentwo nedes
as illustrated t'n Ftgupe2_15.

GT " G1@G2
Figure"2.1B. Parallet ConductoeFlow Oaths

16
n

_ /. .... " :..:- ( _ . :

2.2.2.3.4 Sert.esConductors
Twoor-more sertes, cenduct_onpaths betweennod,_smaybe combtned
crea_.eone conductorvalue by the followtng equa_tr_n:

_2.........
_n..

Tht-sm_Y be helpful__ln comp_Utl

rig_the conductors-be_een--two,dissimt_lar
shapedor dissimilar" mate_|al nodes-as showntn _gure_2.]6_

• , °

A1 &l k2A2

1- G1-_2

Ftgure 2.t6"o Sort_- Co.due_r Paths

17
 ._..__": .._.

-::" The followTng table -illustrates tYplc_1-vaZues of average heat

_-._. transfer coefficients encounf,eped-bLenEInee-rJng ppecidce,

i,

-..: _,. Convecttve ged_um Convectt ve Hea_:.T_ansfer Coef_ctenL

-_ ,:- , . h - BTWH_FT2_°F t

_:. _. Air, Frea;Convec_oR _-5--

' -- AI r,_ Forced Convection 5--5(t
0t 1._ Forced Co.vect_LOR 10-300-

i Water, _Bol]l ng. 500-lO_000

_ _::;::
._. i_te_%
Steam, POrced Cop_vecCt
COndensing_ on 50-2000
IO00-2(_O00--

Table 2-2. Order of Hagnltude of Conveettve Heat Transfer Coefficients

Equations for the computation Of conVeCt.tv_ heat transfer coefftc_n_

_ :: will be divided lhto three categories: natural conveetton, forced sonvectton
_ In tubes or ducts, and forced convection over ex.l_Hor surfaces. The equa-
-:--
--_ _,:._-'- ttons presented for the calculation of cohvectTve film coefficients are

_;;._ (,[) the most ge.erally used expPessTons. Others ape available and are applicable
for maw spectftc thermal problems-Or analyse,. IX, should be remembered
:_ '= that_ the predicted v_tues for h are only approximate. The accuraey of the
_.._-
heat-transfe_ coefficient ealeui-ated from any available equation or graph
.:- :_ may.be no bet'tel_ t-hart 30 percent.
2.2.2.4.1 Co_btned Nattural and Foreed ConVection

• : L__Caseswhepe bott_ natu_l end foeeed convection ape co_blned_-lC-._s

_ - lef-t, to the engineer _o deride whTch-heat-tcansfer phe,omena is stgrd-f-tca_t-
'- - and Uttltze the proper equations to computo
_ the_e_fecttve ftlm coefficient-
:: (h). The-{.ollowt_ deter_r_3ons--sho_ld be helpful.

%:E I) 6elculab.. the G_shef r_n_ber {Gr)

p2 g I_ {T - T®) L3 _ .

P-.: -;__ p2

_ 19
 .>
Gr = G_ashof"number_
--dtmensfonles;
9 =-g_vtty constant- 4.1"7x10__/HPJ
IZ = coefficient o_ volume_rt-_exp_4_ton
@ fllm-_ra_- 1/°R
T:_ _empera_re of the bOd__su_f_-- °F "
T= _ten_er-a_uze of--the surround_ medj._:..._r_.°E
...... :_.................
L.- length-o_:f]ow Ra_IJ--lET
r fluid__vtscosl_'7_--L1/HB-F_T

Z).. Calcula_:e.
theT_eynolds-n_ber_(Re)...il ..........

Re.-U .L

where: :
Re = Reynoldsnumbe_-- dimensionless
U = fluld velocJt3r--FT/HR
p = f_Id denstl_ - LB/FTe

L = lengthof fTowpath -_ ")

= fl_td VtSCOSf_.=.LB]Et_=.__....:

3-} If Gr--<_
0_2_5 Re2- the e_fect_-naturat- co,vec_on er_the-
avecagehea.t transfer co_ftc_ent-fer
p_.-Jr_.d convectto_ fs less than
4). If Gr-> -l_0 Re2 _rce_ convec1_J_n..ha_neglt94ble- e_fect_ ..
o.-_tu_al" convectto_--

5) Zn the _e91enw_ bet% free an_ forced convection-effec_ are

O_ the sam Or_ler-of m_it_de, heat tcansfer ts tnereased by
buoyancye_feets- acting fn: the direct;ton Of_ftow anti,decreased-
whenee_ing In the oppOst-_e dtlm_.t_n.-
6) _n cases Whereit i;_doubtfu_ whethe_foz_l _ ft_# (onvec-_on
_r _ppTtes_the h_a_ transfe_ coeft_icTent-Tt 9eneratl_ caicu-
_d: by using forced and-fme-convect4OnrelationS- Separately
an_ the l_ge_one-ts used. T4_-ae_ur_W-'of-thts _Ute of then_
ts esf.Tm_ed-t_ be about--2-6_;

.° 2O
 2.2.2.4".-2, Natural; ConvectionEquatle_s
Natural convectionoccult whenfluld is set in moron as a result of
density-dlffe_en{es
due _.._m_ture varlatl_Sin the fluid. Ag_at
de_t..o£"materialhas_been published on I_ trans_ coefflclep-ts._ result- ._
ing..._.B,
many.t_chnlque_ w_ a wlde_var.i_ty of resu_ teav_ng_th_.
"occaslonal use_-o.f th_tef_matton east.13t.._oon_sed;_esen.te_ h_et. ........................
• fs-a wide var4et_ of ap_lJcatto_s co_ensed_tnto.a_small numberof ¢Poups=
The intttzl daCe-wetobtained-from the.Oil and Gas:Journal, Eqya_lons .... ::
used are- the. on_s.which..apl_aarto-have the..w.ides¢eccep__.__or-Che_ ............................
particular .set-o__condtt4_o_s_tn_Lved.......

: CQr_elations
of naturalcon_¢tlonheatt_an:;fer
usuallytake....
_e form.-

= K (BrPr)"

where:
h = heat t_nsfer Coefficient
K = an empirical constant
Pr = Prand_ numbe_
G_ = S_ashofnt_ber
n = _n empt.Hc_l expQnen_-

To apply the aboveequat3ons-toan a_tual problem_Solut(an, the

following, steps Shouldbe performed:
1)-. CalcuTate the product of-th_ Prandti and G_aShofnumbers

p2-6 Cp g &TE_
GPPr =
pk
where:

,h
21

-- - ................... " ': 7 -. ..... _-__'_---

.... :" I
 @ fflm temperature- BTU/FT2-HR-°F-FT

p - density@ film te_e_ture -LB/_ 3

= eoeffl(i_ Of v_e_Ic ex_nslon.
@ film_m_ratum - I/°R
• g =-gravt_ constant- 4.1_0 e _/HR2

• Cp -sp_cle_ @Jllm temper_Lt_re - BTUILB-PP

= viscosity_ _L.f.:llmtamper_ture - LB/FT-,HR
L - hetgh_-or length of su_ace _...Flt.

_T = temperature d(fTerence betweenthe. we!l_and and_l u,Id - oF

Film temperature - Twall Z..Tfluld . °F....................

2 .............

:i

Z) )f 10_ < Gr Pr < 109 assumeflow 1_ lamtnar and go to Step 3. :

If Gr Pr .- 109 assumefiow ts tuvt_lent and go te Step 4.
)
3) Laminar Flew

h - C1 _ (Gr pr) 1/_

where:
h -- heat-_ransfe_-coefft clent- - tlTUIFT_-HR_F
k = thermal conducttWty i_-ftlm temperate- BTU/FT-HR-_F
b_= het_ or- length of suPfe_. - FT
GpP_'. Preduct o£-Grashof and _ram_.l numbers-
fre_ ebove-- d_menstoVii_s
¢1 t_ dependenton _e (leometcyof the s_s_.eman(&-c_n
be de_mtned-fro_ F_I7

Film _rn_'a_Um = Twal"l+ tflUtd

_ ._ -°T

* • ¢ •
 4} Turbulent Flow
1/3
h --¢t (_r-Rr)
where:

k = the_la).l, COnducl_Fvit_-_ ft_m tempe._ature-..BTU/_-HR-°F ...... :-

L --hefght-o_-Ieng'_b,.of-csu_ace-- FT
_r Pr -Producto£ Grashofand._randt] numbers_

C| "Is dependenton the geene¢-ryof_the s_stem and can

be-detemtned from-Fi gUre-2-18 .-
I f_om-abeve - dJmen_onless
Z:
j _Im _ratu_ - Tw@II_ITfl_d - _
t

.2-

T "

23-
'_ L_IINAR FLOW

•, 2..
2. ,,,L

-- Cool WarmGas Co01" Warm ilarm. - Cooi GaS ............

:::

Shozrt_Ia11<-Eft:... --Sho_:_Ja1_-<-O.IS.-__ 5horJ:gall:< 2ft ....

- Cz - 0..548 CL _=._0;548
........ Cz = 0.540

: " \_I HahnGas Cool gall

'- w._l cool .,..) Q y///Z/_/

-_-
-- .a11
_ L,q_id //7/7//. (_garm Gas
3
Coo; gall
Short gall <"0.15 ft
Smal_gall < 2 f_ sect Small Idl_11< 2 ft sect
Cz-= 0.55
cz = 0.35 cz = o.71
)
; Ceel Gas _/armgall _'%
•_. _,j j _./////,/ -- garm _as
YJ//;'3/ co_-Tu
: gain gall 8ool GaS .....

_ Sma-llgall < 2 ft sect, :$mall galT < 2-f-t-seer _, Ltqu_d

Smal-1Tube: D _ 1-1n
Cz " 0._I . C_ _ 0.35 CL_ 0.53 SubO for I_ •

gain (;'as (;as Or- -

Ltqu_d
-;
.- _ gem germ Or
:._ u,u_ C6oTSphere
Smai1Tub_ 1 tn < 11< 4 4n Sd_l DCameter
Cz - _.47 Sub D fOP L I_fqU_d! R'.<0.1_ ft
----- Ga_: R < 0.50 +'f, _-_
:. _ - _3 sub_ _orL
•:
T
F_gun__47. EmfFtca-1 Constants for-ItatuP_il Convec_fol)- LaminarFtow
24

--- j --_ -- ...... _" 1975022410-T8C(

 ,I
J

TURBULENT
F OW m m_

_ Cool _ Ham Gas. Coo-1 g_,rm Warm" • Cool.Gas

__! Hall Wall. _td Hal"l gall L'Fgotd ,

1 i,i_
i Longgall > 2._% Longgall • 0.15 ft- Long Hall->-Zft Long-Hat1• 0.t5 fL "
_-_ Cz = 0.t3 C1 _,__O.,}L .... Cz = 0.13 _ Cz _ O,T3 :

Ham Gas-- Cool Ha]-1 Ham-LJqzdd Coot:.Gas:..__, t'i

Cool Hell .......... Ham G_S ...... Cool ga.l.1- Ham.Hail ........... :

Large H¢11 • 3 ft 2 Large Hall • 3 f_-2 Large Hall • 0.25 ff2 Large H_LL> 3 ft 2
Cz - 0.08 Cz._._0;102 Cz = 0.08 Ci ,, 0.162 1

HarmHall .Cool Lt "

(-')
-b(-
/J "/:""//

C-oolOas
L.j
,.-//...-//
Harmgall HarmLtqufd
Co6THaT1

Cool LfqUtd
HarmHall

i _

Large Hal-1 • 3 ft 2 Large Hall-_ 3 ft 2 CX • 0.128 C_- 0.128

CZ " O,OB 6S = 0.162
" -j

CooLGas Cool L-tqutd

H_,t-m
Id_lU_d.

Cz " 0.128 C! - 0_120-- C_ - 0.128

i ii

C gain Gas --_, Ltquld Ham or Gas or

, Cool situs1
| ttqu4& R > 6.15 ft
Large Tube: D > 6 tfi Largo Tub_- IT> 6 tn ] Gai_ R > 0.-50 ft
{ _ CZ " 0.11 SUk 0 for L CI - 0.11 Sub 0 for L I Cz--'O.15-Sub R for

• Ft_ure.2-18. Emptrtcai Constantsfor NeUral Conveetlon- iurbutent F}ow

25-
2_?._4.3 Fo_cedConvectto.-In Tubesand Ducts

TO-apg_ the-empirical relationships for fo_.e_L convection heat

tY_lnsfercoefftcten_,s in tubes:_nd duct3, the fo]}owtng_T_epsshould he
peffenned+_Comments on t3m_4_y_Jcal p.__pe_ty,detemlnatlon and ............................................
i=
entrance effects-a_e_1_ded at the end of-the sectlon...... ,"
1) Calcu|ate the hydraulic dtameterk OH, whtch Ss defTnedas:

wel_zeepe_'lmel;e
r- a i
)
The _draullc diameter-for somecommon
shapedtubes or ducts ts: :,
J
!
,+
Round: OH= 0, where D 1s the diameter-of the tube l+
Square: DH= b where L is the length of a side )

Rectlngle: OH= , whe_ a.and b axe._tholengths

of the stdes ) )i

Elliptical: OH= a , where a and b are major and

m_noraxes

Annular: OH= Oo - Oi_where Do Is the-oute_ tube dlame+_r I!

D]. _ the__aner t_Jbedlameter

Z)... C_o_mpu_
_th.e_._.olds numbe_,Re; from th_followtng equatton_

Re-
where:
OH - hyd_auitc 6tamete_,.defined above- FT
- veio¢tt_ - FTIHR
p * density-- LB/_'_3
_ V_SCOSlty- _/FT,HR

3)- If Re < 2100 the _ov_ ca, he-constdered-tamlnar, go to-_t_F4

IF Re : lO_O00-the-f:tow-ca_be considered_urbuTen_ g_ to _tep S
Zf ZlO0"< f_e> tO,O00 the tl-o. can be eees_e_ed t_ansttl_m_l, )
g_ to _tep

.* _ . . _ . -
0
h " 1.86 k Re P_

h = .3.66_ |s- • mtnJmumfop ]mTnar _Howin 1ong::t_u._b_es:

for very short tubes

(_H <_005 Re_ ::

where:
_i]) v h = convectJveheat transfer-coefPIctent - BTU/FT2-HR-°T
/
R_- Reynoldsnumberfrem step_2 - d4menstonless

OH - hydeaultc-.dtameter- fPora_step1 - FT

L_E_tubeop.ducb Tength_- FT

D - tube..o_..duct:dt_ster - FT.,..

Pr - Prandt] number_ _- - d_menslon_ess

Cp- SpecH_Icheat - BTUILB-OT

=-

K - therm_ ¢ondUcttvtt_ - BTU/TT-HR_°F

_, v|scoStty - LB/FT-HI_-

0
27"
 5) TuPbulent Flow ".] )

h = 0,023 _K Re'8 p.33

where:
h =.heat transfet__coeff|c_ent_-BTU/FT2-HR-°F
k = thermal conducCJvtt_--BTU/FT--HR-°E
OH.=-_hydrau]-tc-dtametep.--:
from step:.]. - -IT
Re = Reyp.
o]ds-number- from Step 2-- dimensionless
Cp.p
Pr-= PrandCl__adPJ"= _ - dJmensfon]ess
Cp.=-spec_l_ heet..- BTU/rB-°F
p = viscosity - LB/FT-HR

6) TPansU_on.Flow

h = C1 Cp U p _ z/3/__b
I_'_"
S ).z_
"

where:
h • heat trancfe__coeffTctent - BTU/FT2-HR-°F
Cp=spectfte heat - BTU/LB-°F
U = velocity - FT/HR
p = dens|ty - LB/FTs P ,. I

Pr = Prand_ number= _ - dtmensqonless

u = _tseostty eval_Wd at the average_._tlm
¢em_rature --LB/FT-HR
:- _b = ¢lsc_stty_ evaluateCat the.bult f]uld
temperature:- LB/FT--HR
-" _s = vts_'°sttY evat__ temperature -
LB/FT-HR.
k - the,_nal cond_cl_Lvtty- BTU/F$-HR-6F-

-I C1 • aFtguce2=19
funetton of_ Re and-can be-detelm_Jfle_L
from

-I
-I
I

-! ¢

28
,_: FORCED
COftVECT|ON
TRANSIT|ON
FLOW/N TOBES
ANDDUCTS

O.006

) o.ool.0 2.0 3.0 4.0 6.0 8.010.0 x 103

Re

Figure 2-19. Empirical Constant for-Forced Convection

Transition Flow tfi TubeSand Duets

T.hemoph,vsfcalPrope_des
Because_thefilm Coetf_tent (h) is a function of the themophysical
properties ef the-fluid, tt ts Important to acceunt fop-any significant
vartatten |n the properties as a result of _he tempet*atu_ef the f_u_d. Fop
liqu_dspgeneeally ol_ly the temperature depe,denC_of the viscosity (_) ts of
ma_orimpOt_tance.Fez-gases, the ether I_rol_erttesdensity (p), condusttvtty
(k) and specific hea_ (Gp) usuaHy va_-ystgniftcalitl_. I¢ts reco_nendedthat
all ef the properties except-Cp be evaluated at the average ftll_ t_emperature
of the flutd Tf defined as:
Tf • 0.5 (Ts + Tb) _he_eTs - s_faoe tempecature_
Tb - fluid bulk (:empereture

! Cp _houl¢the-evaluated at the fluid I_ulk te/_oeratur_.

 "'7

- All of _he aboveequations, wtth the except|on of the one fo_ com.-
-_:- puting h for l_tnar flow _n very short tubes, tgnore entranse effects on
:= the ftlm coefficient. The tmporta,ce of entrance e_fects dependson the
_. ___....... fluid fl_.cendttion and.the length (L)/hEdrJultc diameter (all) of the
: tube or duct. Entrance effects are appreetable-where: .............

_" FOr lamlnar-.flew b-_

--_ F6r tu_oulent_.flow L_ < 10

uN
/: In general, entrance effect& increase the effective film coefficient.
_ The local heat transfer coefficient (hx) dtvtded by the free stream
_- heat transfer coefficient (h®) approachesI as L Increases. If the de_atl
-- of an analysis requtPes the consideration of entrance-effects on the con-
vection coefficient, ]tterery reseaPchmayhe helpful in finding an equa-
- tJon approximating the effect. _ An ex_ple of entrance effect correction
is as ._ollows: consider turbulent flow in short circular tubes wheR
:.- (2 < LID < 60), the effective ftlm coefficient can be-approximatedby:

_:- hE = h.+ h _F "? where :_ < L/O < 20

(__0)

: . +h(p) bere<_Io,.
- -, where::

; t_E = effective ftlm coeff|ctentlCO._rected

entrenee effeCtS sameunits as h" for

i h - turbulent film coefficient calculated from step

; D = tube dtameter- units eonststent with L
-:.
L = tube length - units censistent with i_
• f

- I
• ]
2 !
:2
_ 30
)

:
o
 2.2.2.4.4 FencedConvecti.onOver Fla_ Plate_

___ Emperlcal relattonshtps for forced convec*Jonever the ext_i,ior sur£ace

_ I of a flatplate are gtven-to_ beth-the entire "_JKJth(L_ of:the_pta_e._nd at

:_...! any__nte_med _ate length (x). EquationsfoP- bOth-locations are g_.n.:_n stel_
_--, wlse .fashlen

I) Calculate-the-Re_olds
numbe_Re

- . t_.p X-

•. ReL U_L
where:
U. - free stream velect_ - FT/HR
.:._ p = f_utddensity-

_i') _ = flufd viscosity - LB/FT,HI_

x • tntemedlate plate lengCh- FT

-__. L - total "lplplplplplplplplp_length

- FT- ..........

2) If Re < 5xl(1s fl_. is lamtna_, gO to-.step 3

- I-f-Re > 5xlG5 ftow ts turbulen_,-go to step-4

J 3} Lamina_J_ea_transfer coeff-tdtenC-(h)....

-- Evaluated at x_ h = .332 _ Rex

Average-value fo_ piate wtth "rengthL,

--
: _
;:..__------
k
h = .664 _ ReL
1/2 pr/ 3

where:
h • cenveettve _eat,-tranSfe_ coe_tcten¢- BTWHR-.FT2-°F

- k = then_i condueC_4ty - BTU/H_FT-°F

oo_.
l) L

31
FF

m_
 • - tt ] _ _ _ -- . _ -:.:,,.. -.,_ --_ .'_.......... • ..... i • 1F_! _. _ i_!._..'.._ -1._ ,r_

i " i

•--L !

--- _. )

_°" • i

:-_-t x - tn_ennedtale plate length - FT

_- L. tot_'i plate lengt__._

;-_,:_ Re•_:.Re_no]'ds-nurserfronLstep 1 - dtmenstoMess:_
.....

Pv",. PrancLt_number= _- dfraensiontess

' " costty --LB/._...:-HR

"-_-'_'---_-" 4)_ TurbulenC-he_t::..__an_sfe_

coef_c_en_ (h)

_: EvaluatJ_d._at
X, h,_ 0.0288 k Re_
2_-__ AvePagevalue fop plate wtth length L,

•- h = 0.036 P

:_-. where:
_ h _ convecttve heat transfe_: coefft ctent - BTU/HR-FT
_-°F

: ".._.
-- k - thermal conductlvI_ - BT.U/HR-FT-°F

_.-: x .=_trrtePmed_ate
plate length - FT
_-,-- .

.: . b = totaT plate "_ength-- FT

::.- Re =-Reynslds nunbe_f_m step1-- dt_n_tonless

,: _ P_- Prandtl number, - - dtmenstonlP-ss

i u _ viscosity - LB/I_'-HR

.2.
:a

.2

, __1 i

t
 Forced Convect'Lon,O,
ver..Cyltnder's._

EmptrtcaT reTattonsht.gLtoe_.f]ow over cylinders ape g|ven for the

stagna_on pot_t,.forwarzLporttOrL of the Cy]!nder, end for T_mtnar flow-over
the to tal--cy] .._n_
d_e__._

. Flow..__._
J

NSCagnatJon
F_._nz

Ftggre 2-20. Flow-Ovelr Cylinders

Stagnation Point

when:-
h = ec_._.°eettJ/e"
heat l_ra_sfe_coef#fGlent- BTU/FT_-HR-R-R-R-R-R-R-R-R-R_F

k-= themal cohductt_tt#-= BTU/F'P-HR-°F

U., free st_eeamVeloc]t_ - FT/HR

g ,, Cylinder d1.,a_ter-* FT--

p = denst-_ - LBtl_rs

. - viscosity- L_/FT-eR
tJl = IS e fundttOw o.fL-(_ = -- aitd _:_nDe ap_roxtmted

CO - spec_ftc tieek-- BTU/LB--°F-

33
 !
_ I

• ":_--i- 2.5- , - _ " " " ' '

- --: 2.0

° C_ " _
• l .......

-I -
i -_-: ,i T.5 r'

-:

' 1.0
_: I t.T 1_0 _.G 3.0 4,0 G.O _.0 10.0
P_

_-.- Figure 2-2_. Emph'3catCel_s_antfor C_qlnder

.. Stagnation Potnt
_-- )

.2 For_al_lpo_tton Of _yltnder-where (-0°< e < 80°1 (see-F_gure2-20)

"-
_--i ' h_ - t.14 ___ Pr 1 -. -
_2
:. _. w_

-2 i_= _onvectl._ .heat tr_nsfer-coefflc_Lent - N4J]FT2-HR-_ ..... "

.[

; D = c_11ndevdlameteg_ _ i

_2 -_ Ui • free s-t_am vel.oCJt.V- FT:/HR. . _

I o.= denil_ - LB/_ _

Pr = _r_ndtl _t_-. =-_ - dtmenst.o_less

':- • Cp- specTftc heat- BTU/_B-°F

-- e - afigle meesuredfrOWnthe $tagnat_0_ potnt - degreeg

N
4.. ° . _
° . . -- -- . --.... --
 - :-
i Average ftl.._ coefficient
...................7
i
for lamtnar _low,over a cy!tnde_-

: for a gas h = C D..L__._.

i _;-- _)r_ 11qufd h= 1.1 ck/U_-_D ) Pr"

where:....... ),_
_:
,, _
:._ I ,
h _--convecttveheat_transfer-coefftctent_._.,BTUIFT2-HR-°F i:_
'_

_ : ,_" k = the.real conductlvt_j - BTWFT-HR-°F )!

_'
)_: ::_:-_ D = c.Y!.In_d__e._L
d1._leteP---FI:-: .......

:- U®- free s_:reamveloctty- FT/HR

p - denst_ - LB/FT3 li
p - vtscesf_y - LB/FT-HR i

)_:;-.... PP- Prandtl number- - dimensionless

,.) Cp = specific heat - BTU/LB-°F

: -:.'_. C = emplrtcal constant dote_tned fror,_T_ble 2-3 )
- as a functlen of Reynoldsnumber(Re) - U= p D/; i:!
' _ :- n = emptrfcal exponent del_rmtnedfrom Tabl_ 2,3 I
: :.. as a "_u,,'lC_On
of Reynoldsnumber(Re) = U..p D/_ !

i -- I_

: , .__ ReD C n

! _:-..-- 0 _4-4 O.891 O.330

._-- 4-40 0.821 0.385
_';::,. 40-4,000 O,615 0.466
_:-- 4,900-40,000 0.174 0,618
: 4_,000-400,009 0.023_t 0._05 !

_-_ -"-: Table2-3. E,_pIHcalC)ns*.ants

for Lain,
nat Flow0Vera eyllnder !I
:_ :_:_- Turbulence tnducedtn _he-ate ,_'#st_eam
of the Cylinder-maytncr_Sse h l_V
) a_ mUchas 50 percent.

--. 35
 .)
2.2.2.4.& Forced Convectlo.n Over Spheres

Empt_cal relatfonshtps for_average value heat transfer coeff|c_ents

for f1_.o_e__sphe_s, am-g1_ for gasses and_11quld as- outllned In the
fol towt,ng s_eps_ • " _

1) Calc-ulate Reynolds number Re --U. __D/_ tL

whe_: i'i
U..= fRe.s.t.x, eam velo_t_y --FT/H__ _

p -_ dens_"_Y- LBIFT_ i j,
D -._sphere d_ameter-- FT
p = vtscosH:y-- Lfi/FT-HR

2) For a gas wherL(T < Re < 25) use step 3

For a gas where (25 < Re < 100,000) use step 4

For a 11quid where (1, < Re < 3,000) use step 5 i

3) Gas where 1 < Re < 25 il

) t
h " Gp U® p®(_-+ 0-.48
Re0'5 )

wheR:
h = convective heat trans/_' coefPLeJent - BTU/FT2;HR-°F-_

Cp_= spectf-t¢ he_- BT_J/LB.°E

U= = free- sCream-_e_loett¥ _ I_/HR-
p_ = f_ee stream-density_: LJt/FI_
Re - Reynolclsnumber fro_ step 1 - dimensionless

4_ GaS _eve 25 < Re ._ 100,000-

,.

-. 1975022410-TSD06
 : • °
(L)
_ . where
- .-' , h = convective heat transfer coefficient .- BTU/FT2--HR-_F

__ "I k - thern_ conductlv._ty_.-.BTULFT-He-°F

:: i,_L. D= sphere diameter - FT
: ,!:__..., U= = fl'ee strea___leloc|.t.y ._/HR _
- p= = free stream density - LB/FT3 _ ..........
]_ = VISCOSlt_/...--.LBI_I-HI( .........

m_-_- _- 5) L!qul:d _w.he_m_.].

< .R__.< _2_00

-g--- .97 + 0.68 P

.-" where:
- -- h = convec.ttve hea_ transfer coe#ftcterlt - BTU/FT2-HR-°F.

--
o._/
k = _emal conductivity - BTU/FT-HR-°F

:::
:.- ;:l' PP = PrandtT number = - dlmensionless

- Cp = spectft_ heat- BTU/LB-°F

p = v|scostty - LB/FT-HR

D = sphere dtameteP- I_

:_- U= = free Stream..veiocfty - FT/HR

-_-_ p= = free stream dens|ty - LB/ET3

-_; --: .

--

. .: 37
 .J
2.2.2.G Computat;tonalMethods- Radiation Cendustors

M_em_l analEze_ computor___cograms

11nea_tze.the radtaUon term
prtor ta perfom|ng the. heat balance at each ttme step. Th.t.s_opera_ion
stmpl¥ amountsto £actortng._(T_L;- T:I_l tnto .the..fo)low|ng ¢ompgnents
(Tt +Tt + Tj-Z)(Tt- the-ton(Tt3+
Tt2 T:_i_*Tj_3) ts: evaluated by-the-computer-each.t|me piss ustng the .
J
eu_rent:.va:luesof-T1 andTJ:. Thts-qu_anttty |s. then multi_led bl_the .tn- t.
put-_alue:of the-radiation conductor_:thus .reducing the-radta_on_equa_ion
to a l_.ear- form.. The theme1 engineer need only be_.conc_rned wtth the........ ._
tnput value of the radiation conductorWhfchtakes the following form:

G =o¢Ft_ j At for radta_Lon_to a bl_ch be__

G-_ e_ff.j At for redtat_tonbetweengray surfaces =:

where: :
G = Input-value for ,'adtatton conductors BTU/HI_0R _ ) !
o = Stephan-BoTtzmann constant = .1713xI0-_2 - BTU/HR-FT2-OR
_ "
¢ - era|trance - d_mens|enless
Ft. j - geomtrtc-conftgu_atton fa_tor from surface t to
surface _- dimensionless
At " area of-Surface | -F_ 2

_._[.,_ .g._ body radiation ifactor - dimensionless

The emtttance. ¢_ ts a measure01_-how-_vella body ca_ tad|ate energy _s

comparedw4tha black body. Emtt_SncetS the ra_o of-the totol-em_sstve
pew_r-of-a re_l-surface at temperatureTt_ the tot_l endsst_/epower-of-a-
black surface _t the same£empera_'q_re.The em_ttaneesef vat'tous surfaces -_
are-_ f_4nc_lonof the. mater_a_ Surface _ond_t4on.an_ _emp_*a_re of a
body. Tt_e-surfaCeof a-body, &nd therefore the-em_ctance,ma_bs altered-
by po/_h_ngj roughtng,_p_tnt, fng, etc; Th_ wlues of _ for mest-_mmen "
_ato_al_ and surface cendtttons-h&vebeer_mes_re_ at Varteus tempera-
t_s and are presented _ _al_es or gr._phs_n manyreferenee m_mals.
It ts left to the _ngtneee to, detemfee the value Of-eml_tance to-Ge used
and-_hethe_ the Vartat;ton _f_¢ wtth tempere_ur_ts stg._fcant over the _Jl
temperattlre range expecte_ fo_ _ sut_face-.
 The geometricshape(configuration)
factorfromsurfaceI to sur-
faceJ, Fi-j,is the fractionof totalradiatedenergyfre,_surfacei
whichis directlyincidenton surface_ assumingsurfacei to be emitting
energ__dlf-fuse.ly.
F_.I wouldbe-the.._fraction
of-totalradlant..energ_
....
fromsurfaceJ_whlchi.s-:;intercepte(t
by surface..__.
The conflgut_atlon
• factorsfor finiterogl_s.of.__diff.ps_e-_reas
am relatedby...'

AtFi j = AjFj.t

The configuration
factor,:.Fi_
j, is a function..of---the
ge_omet_of..
the_.s_stem
only. Several"o-:.
,terprogramshavebeen dewloped-tocomputethe shape
factorshelen surfaceswith complexgeemetr_es; however,ferm factors
bei_eensomesurfaceswlth simplegeometries canbe han_ computed.

Figures 2.22 through 2.38 present configuration factors for various

simplegeomei_rles. The use of theseFiguresamt_onfigu_ati_r, factor
aIgebraw.i_lallowth$ engineerto determinefo_m factorsfor many simple
;'.l.i radiatlon problems.
The following
examples
ofconfiguration
factoralgebrashould-be
helpful:

AIF1.3__.A_F3.1
P.1F1.34= AIF|. 3 + A1F1.4

A12F12.3#• A,IF.I..34+ A2F2,..34

A1-2F12=34
" A1F;I-3+ AIF1-4 + A2F2-3 _ _F2_4

A1F1-4 "A3F3-;2 (symme

Z_ cdl ly posttloned)

39
 t ,

'l_t¸ _ . _ ._

I)

03 ii. i t'u i ll.,.u i r It ricH.. .'_

! _- Figure 2-22. Radtan_- I.lerchal!ge Conftgu_-atton
02
I[II
Fac-_rst
I.I II I l:lJll
IIII I III lillll_"
_
Patnt Sources
"

ii1|. i 18f.*_l_"L_I_IF ] I_ -5 I J I I I'' •

005 I.I iJ_._,_n..J_,-_,, :,I0_ _i
I "" "flTI _./3'_Y'_I_ LJ4q'l_'l"t'Trl liB ''PJ'_-I
I UF.,'rillflJ rl I I_1_1111 _ LI i I..I. _JH_I'_

_o_ Illll I':111IL-IIIII II 4

_t ioo.oo_
, IIII IJ !1 LIII ..............
......... .- ,__
oo=. _'_",_-I I::N',; ', ',',', LLH]," k',
'¢_ F_"IVl II].IIIHII I I.II IIIIIILI I
i o.i 02 O_ i0 2 " S ? i0 c,_
Y

Th4s conflgurat|on _;s a pllme po|nt source JAz and a plane _-,ctlngle A3
_- parallel to the plane of d_.z. The nomll to _1 passes, thn_ugh otto-cot-
net of A2. Tnr= curves for this ¢o_flguratton are g|ven above _he_ F_3
.wtth = - o/o and

I)
iili ts plotted ds ,_ function of w and ¥_ - h/o.

o,,, .....
, I I il-llillllili
0.o_l_-I
I-I:::;:.:_
lillLl_

tc 004 Lc 7-:_; :::;:=_- - _ ..............

_. i __: lI_ o_I __rTliHiilnl.

ObO;'
_
. _X_OI
III, '-'
d_ ll _i_l_ i i !111
I_el_ I I,,I,[+-.t-t-_'--]i
i. i i Jill i]I11 i 141 l

i 7' -i-" 8 ooo_P,/;; ;; ;41;; lill"_ l _;-;U ;;; ; l',;l- l- Fi: I

°°°_FI I/llllrllllll- L I IIIllU-I-IIII-Jt"t
oo,.1_ IIIIIllJll[o I 2IJ 345
"'Oj OE030_ lllll Iil11"-IIN
IO 20_0

i =-
_t_
• I
!_ I
/is; tile point sri!t'ce is lOClltel
¢ollnin stde with 13, The hikes
It,l: Otis corlier 6tf I nlctliille l_lit liilt OiIli
o'f the WG re¢l;pngies tntel',illcc It In in..

--
_ " -_ti 4-Adapl_j TN-tI836,. O.C. lTamilton and il. i. t4orga_. ;
,L7-I"

• _ -t ,_ ;
 03 :,T_,_ _ ....I IIIIIlll.l
..........

02 I.ImUill_l I_r_JiNr_l
iiiiiiiii: 'J 'J
•i i i i
0.3
02

0.1
[ I

n.= '"'"'""
_----" IIIIliJlll|
,I ,L,,,_,,
IZ'IIIAII
,,I ,,I] iJLIItlUMkIIt_!_I
, _LIR_JII_ eO.02
&

0.007 0O07
I

-v-'o::::::o.,"&'o_o_ ;o "_"_%'_-"b--_o o, o_ o_ ,o _ _,

02
004 ...,_ ........ ;.;_,..,., ", . _-,',H,,,_
_"_1 i | It]; I Illil I IJ Illl_ Zllll
.._01 00 1 I I ]J I,,.LI 111 I I I Illl,i lllll
tCQ07 -- 002 II'eJ,%l " fill II|lliilll
o.o, _,-"
- I II I1_11iilll
" oo_°°' _ oo,I[llli_lll",L
PI,i_'i'tl:l_l II I[llllllU
0004_ ......... _ .]..........................

"_ _,n_i]llllll.[ll. I i I'%[J] i_lllN %['_J kllL._'%_"d'lJ

°_o o, o_o_
o, _ _,_ _ o_,__,1 0 OI 02 03 05 I_.Jl°_,%_
IO _ _ 45 IO

Thts conftgurat|'o.n Is a plane poirit source _I:Z and o plane reP,tan§_,e A_. ]n
whtch t_e plahes of JA_ ailedAj_ fntersect at aft allg|e _(0 ° < ,) < Z80o). The
conf]guPatlon-fltctor values are 9|Yen tn _h_, c_frve_ pl___'ted _¢*zc fC_ VO_C,U:.
;;_* o;' _, _, aria _, where N - o/b and r.. o/b. When _ - 0% F_2 - 0 for
t/ < /, and o._ for m • _,; wizen # - 18o% Fz8 - 0 for all val_es _f N and z,,

41
 D
Ftgure 2-22 (Continued)

F.-i(1 +._e) ........ i ':

This" confl0urat_on _s a pla_,_point so-r_r 8./,1and an_r_e.f;n|te-plane-AS,

wlth the ptanes of 8,42and A8 lntoroectt,_ at, an angle e. The COnfiguration
factor values mAYbe calcull_i from _ above eqUa_lon.

Thts c_;mf|gu_'o_ton
1i a ptehe potnt source _f and any surface _ generated5¥
an tnflnttoly long 11ne, movtflgporalle_ to ftse_f imd to the pl"ane of _41,

The cbnfiguret_on-fo_tor vel.es _ be co_putodfrom the abo_ equetlon.

: 42
E)
t
 i
O )
Ftgure 2-22 (Continued)
1
i

oe

_-,,4 !

Ol 02 03 05 2 _ 4 _

Thfs _nflge_Uon ts a plane po|nt _eurce dAz and a plane ctr_lar d4sk
A2, The plane of dAi ts parallel to t_e plane of A_; the petal source i-
loceted at a dtsL_Ice _Z frm the normtl to the center of _. The conf4g_
rattan factor _'12Is plet_d abo_ for _'ar/oosvales of E _ D, where
E • r2/d and p = _X" !

I
!
• i

O[ I'I=R_'D L
' F. I'/,-__ ' "_

,,, o)
i

I
TMi configuration ts a plane potnt sourced4X and a pian_ d_sk A2I the
p| ane_of dAX and A2 ,t_rsect at an angie of #0°, Th_ centers of A2 end
_4 11_ fn a plane that 1_ perpendtCulerto the two plane_, The equation
abovets 01yent5 terns of _ and _ for _., r/b _d Z_- _.
 Ftgu_e 2-22 (Continued)

05
0,¢-
03

lOJ I
_ a7

Th|s conf|gurotton ts a plane point source_dA-_ and a right circular c.vllnder

A_ of lengt_ Z. Th_ nomal to &ll passes through the center of one end of the
- cyltnder and is perpendicular to the axts of the cylinder. The configuration
factor |s plottod abo_e as a function of D and & where D ,, d/_ and L ,, i_/p,
• [

i_7 ,

Th4s ¢onft"gurotton cons]sts #f _o concentric Cylinders of r_ltus _ and d

ahd lang_ Z w]_h a point sbi_ce dA; on the thstde o_ th_ ta;$e cyllndeP at
: one end, The coffft_.rattPfl factor P11 ts fro_ the potnt source JA_ on A_ to
At;At _as not toclude the ends Of the ahnulus. The curws for th_'s conflgu-
rst(onara9]w. at_ in teFmnf D and L, w|th D - d_/r and /, ,- Z/_.

i,,[ !]
 .[i)
- Ftgure 2-23. Rad)a.t-lncey.changeCOnft_JurattO_Fa6_rs t
LTNESOURCES

O_ I I II "1 II1_111 I--J I I-I'l'lUll-la

[,j [ I I LIJII__'fl-EI I I I IIII !

_--7" o_ LLtl_uL_I Ill I"'" '-j

02 I IJ I/_N']_ _

/// ',,/ _o, 007

[l__'_ ! • °
II -- _°_"
I I u ON - I,
ll//_
..L.__r
_,.............................................
00_
_rJ_/rMllilL
02 03 0.5 I0
I 2Iit3 4I Illlll
_ _
tl m :;

iii- toTht._
tha plano of &11, wtth
conftgm'a'don ts a 4A1
1theoppostte one
source 4/1 edge
z and a of A_.. The
pTaner_cl_n _'_,_ts .
paranel
a functtod of : ahd _. where : - _/a _d U -_/o.

Thts .ConfigUration ts e-Ttne source_i; _nd _ r_ht ctrcuta_"_ltn_r A#, _

bo_ of" _engt_ _; the nomil.throUgh each-endof the source-pesses_r_ugh-_ ii
Ind_tS'_lomal tO_ the tenor l_ne of the cyTtnde_ at-the eh(_s.The cohftgu-
_h_rm_'at
f onDfdct_r.
_V_'valUeSand_/;
-enL-Fl°t_m_'4b°v_'_/_,
in terns of _e porametei_ D"ana_r., i':

"l'Add_ed_ NACARepm,
e;.-No.-T'N-4_ by _ {;. HaMI_.on and _ R. _lOr_litt. i

i 46--

°'
; --c_R .- Z4 -

--i j Flgure 2-23 (Cont;luded}

• _1
-03 ":'|r: ', .;;;'--",;_,';
................
_ :'." f W,',','Y, ', ',
O2 11111%1 I |llllllillL% I II NtJ_IIIIL LI
i I I It I_l M lilt[ [IILFt_ I%1 I_lf I IHI I I

" - tl _ I I_IIIIltlIIlI_I-I/ILItlil_II II
:_ = ": __ _,,i_iq..[:. i '_'i).[-?.__.:: '_:

_ _ oo_; ;_1..
; ,,;_ '.','i. ',',',;_',_
Va,_li',',:"',",',_Pi
"oo241-_--1_g_14411
_l.-FI4_i_i.FIbl
.LLli'I I I I1i111.%11i_ QI I._'tl,'G'd-I 1

' :: 101
_t_Q
__
..........................

- :_ i .-iii]!']-_:-
ii--i.i;_;:._ii.i]i'7! :: _4 11111111111 II
0.7 I ]_l_qt_JI Illllll

[=.::_
_ i ':_'_o
°z _ "l
_ n: _11 i !'l_ IX.;_H_.
_llll III._I1IIIII I.I!
I IlL

_" Ilo_ i;_,',i , ._*, i-i;:- ----1[ "......

i _ :-'-:' -_ o:
oo__;_;L"t,,_,._
I_IS-",i NIIEI_I _1111 IIIII 1.1 I °°_
(_o___-_.[ _-:]_'_i,'i.': "_:''_' ; :;:1:1_
....................
i:SSl
..........................
• ii.ii_ [ "';""
!__; !!._'.iTi_!. _'. _. i _.i'_.._'._._'--': : ". i_ 0.02 lilli I !il __""'<""1"_'""' '
I _IIIM% I%1-111% _LiiJII I I

•2:;
[ t °o. -t • ., o.,,...........
O.Ol .......
:-'._:. "'" 161LSO,, i_L,_"

-- 0_ II1_11-1. i I-_111 lilit--i " Illi_imli-I I O.I --'';" " "" ...... "":-:- " ..... _"...... "_

0.2 I_-_IIBIB_:j._ I IIIUIIILIIL.I

llllllJllll I ill i iiiii illll [ I IIJilILIIIII. I I

---- _o, __UII!I !illHilllill LI -_oo __,_,,,,,. ..... ,,*,_

I_ 004 ....................... , ..........

-- _0.05 17-i.--
-do7 I-IIII1%.' "ki.q.L_-77._!! ..,,...b_--_;:
' ]ii]i]]]
iiii, _ _"-] ;ll_G.3i_oIJIIII l; L';:IIlliII; 11
" ' '"_'*_"
s ooSI I_; ;: 7,;i'.: ;;.'.-.;W._.;;;;;;.......
_°° _k-NNt_-_
_ o0; II _.-:-::.----'
_N--H
: : .........
_1 iIINII_I_Ill _l_Jlil I]111 I I. ODO I1,... , ! li_lill_ I_i] Illll I I
-= il °°_ t1.1111 i _lili_l%i_liilllll..I I O_:_-........ " ..................
, 7 _oo,N.,.,__ _H-Hi- i_,::iiikl]%:Oi?_
- _[4
_"'1:)
___
OI 02G3 OS IO 8 3_1_, I0 #0 OgOio
__,OI 0203 0_,
.-,_.._.
-.<-..--,.
iO 2 5-45 I0 20
: %:.-'.
- _ .
_1_*120 Jell,_iSO
-.
%;..

Thts configuration is a ltne source 4AI arid a planc- r_engl_ Ae that Inter-
-=" set_ts the plane of--4A1 0% _n Imglii ti. Till! coliftguraU6n foctl)f t_i lilo_il

aboT. Ion i-0% l'll- 0 farl <_, a_d I foi'# _ r.; _e_i,, l_o°i lli,,o

t __.___ ell v41uts of # aiid I,,

--
! 11

. 2

_. • _--

' ":- 47
 Ft_ure 2-24. Radtant.._, lnte_hange Configuration Facets t
PLANE
SOURCES

/_ f-- ,.oL__,! !iF.i=::i: _ : !:::_. ,_

I_ • . L I I I Illll lllll _ I _IL_I--IPrlT f al..I

/-.- L Z I / __._,I I, I,,,,_.,_-_,,,,_",_

i i II[]_BT 1_ i IlJl Jl III J_
F_

, /" . I /7- ILoo5 ......................... ,- L-L_ 1:::]

;/
.
I // _-o_,..........
t J;4 _ _ Illl_rlll
,L............. -_-J:ll!_l__
J_t 1 i ills i elll I_

,_,,,,,,..,.,
"" IllAJ IIRII IIJ_l-- I I IIIIII Illll. I

P----_--'I c_i":-I.,tflll,
Hl:Hlll t ] IIIJIILIIII-lJ/-I I
Or- G203 _5 _0 .2 Z,45, I0 20 ¢o
F

A_Tht$ con£tgurat"lon
and ,_8, consists of tdent|c_l,
The conftOuratt_n par_ltel,
fact Or values directly_bove-ln
are plotte(I opposed _c_.tmg]e$
ter_ of the
parameters ¢ ahd _, where m = b/o and ¥ _ _/o.
I
)

TMs conf_u_ltlon consfsts of'pSr_itlet, .d|rectly Oppose_ljptane-cf-r_Ji_--

dfsk_. ¢ - _e/d '_nd ,P ,, d./_'l,

_Ada_tedfrom I_CA Repo_ No. TN-_B-36,-byD. C. Hant]lto. arid M. R. 14orgim.

4O
 The _r._y-Oo_y shape factor (SGrtpt F)_.j, ts the preduc_of the
geome.t,r_ s):ape fac_er-Fl, j aed-a facet whtdLallows for the departure of- -)
the surface from black body condtltcms, Fo_radtatton enclosure_, the

Jt-j factors are- gene_zl_ly evaluated w11_ a computer p_ogram. The input
for _ne program being the AtFi; j va]ues- for eveW su_faee of the e.closure
- to eve_ other sulfate.anaL the emtt_ance .and area _fo_...ea_. sur_face..__SJmpl
! _fed ..................

Cv_ eqnattons f°__-tt-_j ex_s_o_ _o-c_mponen_g_ay enciosunes.

Paralle_..f.1aJ;.plate$:- FI. E = F_ 1 = I - _,_

t"!

_I = ..l _

Concentric cylinders of tnflntte hefght oe concentric spheres: ...... -:

r_.2 = _, FE.I _ 0 i

-'; f1_2 '= 1 '

leA1 /1 -I _ "

Pot "non-enclosed" su_aces an eff'ec-tive emJl;tance_ _eff' b.., .... _h_

_.j = eel._ t F1._I

The sffe_lve _._t_ttance is -_ f_nctt_n-oP the-eminences of t_ _ _
Su_faces an¢ the- Cont%gu_at-t
on faetors. (_) bebveen--them,._11teer_r induced ! :

_l_h use o_-_eff, tS t_e-result- 8f-tlegle_'t_ilg se_ _efleottot_s ft_m i ;

su_-aces ot_er _han the.t_o fm_whtch tt_-effeottve_eod-ttance w.s determ_ed.
B_-i_c-tflg-_ottel-*s me_hed_'or _h_ee flat plate sur_c_s vd_.h-emissivities t-i
of E1, c_, _d 1, the followt.c_ equ_tton een be _ons_

Table Z-_ p_nt_ Zef f appt_xtm_tons for-wrJmis cT* _' an_

i!-] F1.z * F2._vaiue_
.... (1- - ,
Intermediate Wlues ca. be app_Xt_a_ed t_ t.[e_:polatton. _ J

2 .- ._ 2 -
 FLA T PLAI"I[S |NSZ0.f. A I[ILACk ENCt.0SU_tI[

EFFECTIVE [MIITANCE FOR FII,_F_I ,, 1.00

1

I_ 1.00 1.000 .gO0 .0(_n .T(_n--,,600 .500 .tOl_--.,,]lO0 ._M ;t6_ .6_n ofllfl
N-
! .90 *000 .8_? ,?_1_ *6**T _ *_7-_ *SIll *-Hi *_9_ _ _0_,9 *0Z0

¢
[ _60 *600 *560 *_1? .4?1 *_ ,_166 ,_10_ *_4! .109 *_1_- _0_6 *00g

U ._10 *_06 *200 *Z_ *_ eZ_-:*_li9 ,19_ .101 *Z_I *069----_1-? *001

r ,ZO *ZOO ,19_ ,lll_ ,|Tq .16§ _,lg6 ,t,_| ,iZZ ,6_1_ ,0_,7 ,o]IZ *nOT
A
C *1_ .10Cl *090 *OgS .09_ *Ol_ .el0& *026* _f69 *0_1 *0IT *0ZZ .O_l_
I[
• 0S *090 *049 *0/_(1 *0**T .0.*_'_ *04** *0ill *0ST *0]i;_ *0_;I .015 *00_l
|
r* • 01 *010 *01O *0|0 *0|0 *0¢_ *00_ *OO_ *ooi_ ;60T *6OS *00J .0(11

51
i 5;_"
 + .. _ H i 1
- Table2_4 (Continued)

_-" - FI.AT P_.A/.E$ _qS_DE A 6t.4CK EltCLO._JRE

[ EFFECI+IVI[ ktI,I|TI+ANC£.IFGR PI_#21 • *60

--_- EN_TR_¢I= OF SI_VACg--2

-++
-- 4
1,00 ,eO _..1103--.30+.~-,60 *_0 +101 --.30- .20+1+.*1"0 .O_S ,01 -.

_'_*_ +"_" " "................................. 14E1,00 1*00_---*e00 ,80n ,TO0 ,_0 *500 ,600 -300 ,_OO .._.100-- ,03n +_010
I ,90+. *S100 ,ll_J *T29 ,64+2" .3_3 ,464L- *371----.21133- *tll+l--.*OSl_L-.-*_4_ *nl_
t

+
+. _ __ t *60
A
,IO0 *T_9 *454 ,611.. .SQ_ *&26- *_14S *261---_11"1 *_gG t_&_
t6 *20 ,700 ,63h1 -,581 *S11--*+.!3--1o383---*$16 ,Z40 ,16 +, *084 *042 *00S1

" .+ . F. *60 *600 *563 *_)011 *31_31 ,398 q134.L--,?.00 *_t6 ,169 *016 *0_9 *006

O" *SO *_00 ,464.-_r+26 ,663 ,3131--,Z9_ ,24_. ,190 ,132 ,061i ,033 ,007
._ F
.._ *40 *400 ,31_1 *343 v31_ ,280 ,24_ *ZOt ,160 .113 .039- *030 *006
S
"_":'-- U *30 *300 *.ZS_ *26" ,,Z_ *_L6+--+L_O ,160 +L2T *0_0 ,046 ,0Z3 300_.
_- R
. +-" _ F *_0 *3100 ,1031 ,177 ,2_4 ,169 ,13_ ,11_ *05+0 *065 ,033 ,016 ,00_
A
C ,10 ,100 *oiJ)5 *Oen *0114 ,077 ,068 .039 *048 *033 -019 *010 *0t_2

tO_ + r11
, 0_0 J048 .043 ,06_ ,O_g .033 ,0)0 *023 ,010 ,nlo ,0n3 *0n1
1
-"._+-+ ..Jj *01 ,010 *010 *009 .00g ,003 ,007 ,006 *006 *004 *002 ,001 *000

EJ'FEcTIvE EHITTA#CI_ FOIl P'_'_f2| * ,30

:_ Jq41t_._)l¢l[ OF SURFACE*2

1,00 *4)0 ,30 *?0 - ,1_ ,50 i_O *|0 *|4 *10 ,03 *0_ ...................

- -- I[ 1,(10 1,006 *g00 *SOn ,till ,6_03 ,34_0 640(I .300 *_00 ,100 *030 ,1110
I ,qO *_00 _lr1+_+ .1_-_ .6kn *._1 .43,_ *_J11 *_60 .10t .004- *047 ,009

.__ I" *80 *800 *7116 *_:_ *37T *_00 *4|J _3_0 *3168 *_1f4 *011111.044+ *009
• A
_'-- _ *30 ,10_ +666 ,377 ,613 ,4_1 ,378 ._._ _=_3 ,1S9 *Oil ,041 ,606

E .1_ ,66D .33! ,son ,_+,_ +_I .SS; ,_1_ ,zoO- ,1_33 ,07| ,o_7 -6_?

- - _' 0 ,30 ,300 ,46_- **_1 ,)_I ,_13 ,_I_ ,23E ,Ii2 ,12_ *06_ ,033 ,on*
• F
"- *40 *_0 .371 ;34_i *)Ol *_73 *333 ,133 .13_t ,10!1 ,0_I_ ,021 i006
6
++. u ,]0 ,300 ,_I0 ,_0 ,Ill ,;_6§ dli ,133 ,Ire ,+13 ,0&4 *4_t *+03
• R
F *+0 *200 *II+" *I++ _139 1143 ,123 *lOS *08_ *piE* ,03| ,011 *0++
+
¢ ,10 ,_0 *094 *OIlIP ,011 e035 *063 *033 *04+6 ,031 .017 ,00_ *1102

_---- *01 ,050 ,_7 i04i *061 *033 *O_J ,01_ *022 ,016 *009 ,003 *flOl
1

_- :'+ i _ ,o3 .,OLO ,o03 ,oo+ +o6b ,006_m_t ;on+ *oo3 ,0o3 ,noi ,ooi ,Kon

-+ [vfllm.o
--- .++ ....... i.. • +i u- ",•23.-, .+, "--++++.....+......
 Table 24 (Continued) ..................................................................................
_-_._.__/_

FLAT _LATF._S|NSIDb k _.AC_ ENCt.OSUA[

¢_F1;¢TIVE F.HITTAflCE FC_ F|2¢SF2_ • *40

 i _) Tabl_ 2-4 (Continued)

FLAT PLATES INSIOE _ BLA(_ _NCLO$URE

_" E_f[Ct'IVE ENiTTANCE FOR f12_21 " *10

£NIfRA_¢C--_ $U_fPLE 2

:_ 1.00 .0_ *80 *TO *_0 *_G----.40 *$_ *20 .1_ *0_ *01

E I*00 _*000 .000 .800 tT00 *600 *EO0 *600 *_ 120n .I0_ *06_ _0_
N-
! ,_lO.- *_00 _811 .I_1 *63] .84] ,_6_ *_| *_T| ._!4 *0_1 *GA_ *_
T
T *EO *000 .T_l *648 *_6i ,484 .40_ *]24 .2_3 .163 *08_ .041 *008

[ *60 *000 ,_6_ .68_ ,42_ .36_ ,_0_ *]_0 .16_- .12_ *06_ ¸.081 *006

_-': 0 *SO *_00 .4_ *404 *_9 ,3O6 *Z86 *206 *|_ .10_ *_8_ ,0_6 *OOS
F
_40 *400 *_2 *_6 *_8_ *_66 *206 .166 *1_ *084 *04_ .021 *004
S
U *_0 *]00 *_T2 *26) *_1_ .15_ _|_ *12_ *09_ *06_ *01_ *016 *00_
A
._ fA *20 *ZOO ._81 .16_ *l_§ d]_ *tO_ *084 *004 *0_ _0_ _Ot! .0_
¢ .18 ,|00 *O9| ,01| *0_ *06_ *O_a ._A_ *_] *02_ .011 *_0_ ,nO_

• *0_ _0_0 *04_ *0_1 *081_ *0_1 _026 *02_ .016 *0|| *00_ *00_ *00|
- 1

8I _.rOV_nV_"_T_'O 55
._-" _ :7- ......... -* _'-" _ .. _" -'" _ - "". - " _. -"
 i.

Table 2-4 (ConGluded)

_'_'_ FLAT PLATES IN&I_ A 81_AGK-.

[NCLO_SlJIE

4PFI[CrZVF. EN|TTANC[ FOR FlEeFEl • .0_

l.O0 ..+*-00 *8(t .711- *_:-.+*SO+.. *80 --+30 *20 .10 nO§ .01

E I*OA_ 1,000 *900 *800 *TOn[NITTAf_F.

*400 *-SOG+,40n
OF _lRIrACt *300
2 *200 ,_nrJ_.++OSll.-.++olt_
H

T
T *80 .000 *T21 *661 *56E _2 J402..-.*322+_ ,2_2-1_.161 ++0001 *040 ++008
A
I_ ..74_ ,T00 ,051 *_62-.+.402.--*-42_-*333- .20_--*21_ .+_2 *OT1 ,034 *OOT........
¢
• E +1*60 *_0 ,_r_1 .402 ,423 *301 *$05 ,243 +.188 .122 ,061 .031 *006
I I +090 ,900 *81L4721 ,411--._41 ,4S3-.._+_I ,27L ,1810 *000 *04_+.+.*009
0 *SO *bOO ,451 ,402 ,35_ *)01 *ES| ..20] *1_3 ,102 *0.*'I ,026 *OOS
F
• 40 .400 *$61 +03_2 *_S3 *E40 *20] *108 ,12_ ,002- *041 *021.. *004
$
U *30 .300 +0271 *242 *Z1_ ,105 *_ *_3 *037.--*062 *031 ,016 ++003
A
F .20 *200 ,1111 *-_61 *143 ,12:+ ,102 *082 *062 ,041 *021 *010 *0012
- JL A
C *10 .1_0 ++090 ,081 *071 +0061 *OSl ,041 *0|1 *0_1 *OZO *00[+ *001
E
10S *OEO ++ Od_* _ I +00410 * 0 E4_410_ _* 10 _16. 10_1 10_6 10_0 4100_ i100 _+ 100_

1
•01 *010 .o00 8o08 .0o? .006 ++OOS ,004 *003 *002 *001 ,OOl ,elnl_ _ ;=
! k

PAAT PLATES INSIDE A BLACK ENCLOSURE

EFFEcTIvE EMITTANC£ FUR p|ZiFEI • .01

EMIA'TAflEEOF- SURF_2

1,00 ,30 ,80 *TO *60 *SO *_0 - *_0 ,20 .1_ *OS *0_--..

8 1,00 1,000 ,000 *800 *?00 ,600 *EO_ ,400 _300 ,100 ,100 *0_0 *OtO
14
! ,90 *900 *i10 *?i0 ,6_0 *340 ,4_0 ,3_0 *_70 *_80 *090 ++040 *009
T
f *_0 *_CO ,_U *040 *_0 ._0_ *400 *_EO ,2_ .160" ,_80 *0_0 *005 • ]
I_ ,Io ,10o ,oi0 ,so+ ,40¢ ,+_'+. ,_I ,zs_- ,El9 ,IAO ,0_0 *000,0d?
C
E +.40 ,600 ,EAO ,Av_ ,43| ,gEl ,301 ++_4| ,Ill ,1_+ ,0_ *0_0 ,000

0 ,)0 ,_06 +,4_0 ,400 ,|)l ,101 ,E31 ,+01 ,l_l *|_0 ,000 ,OJ3 +000
r
,,,0 ,+00 ,|00 ,_Z+ ,|01 +_l *I01 ,101 ,111 ;000 ,040 ,+OZO ,004
0
U ,30 ,300 ,+TO ,140 +++10 *lmi *lSl *]|l *030 *050 *_0 *Oi_ ,On_l

+" _10 ++;_00 ,140 *|80 ,140 *AgO +1050 *OliO ,010 *0*0 ,000 *010 ,0_
A
C +10 *|00 0000 *010 +070 ++010 *0S0 ,040 +0|0 * O_rl_ *0|0 _003 *00|
E
• OS *0S0 _,040 +00"+ *035 *OYO +OES +0_10 molt ,0)0 ,00_ *06J ,001
|
001 ,010 ,ts_+ ++O01l *OOT *OOl* *OOS ,0_4 ,01_$ *OOZ ,01_1 *COl ,000 "++_

omc+m,.+
=.
m
+0 "+um l,_(]Al,i,l, _
L)
2.2.2.6 Co_qtatlonal Hethods: Hass Flow ConduCtors
The use of a mass--f]o_conductor in a thermal network |s a convenient
methodof accounttmj for the transfer of energy_from one point to another "i
due to the actual mo)/ement
(flow) of-a fllzId Fromone potnt to another__ i
Mass:f:lo_ conductorsare Computedfrom the equation: ::;

!!
wheL,_-:-..
..... i
=-the massflow rate of the fluid.___LB/HR !
Cp- _]_|d._spe¢tftC heat _BTU/LB--°F _
The massflow rate (e) ts rolated to the fluid velocity by the expression: J

_=pAU I
a

where: i
A -" the cross sectional area throughwhich the
.) fluid flows- R)
U = the fluidvelocl_ - _/HR
p - the fl_Iddensity - LB/ET-
3

The thermophystcalpropec-ties Cp and p should be evaluated _at.the buUc

•t.emper_turoo.f_thefluid, tii_
The mass.Fix)w-conductorwhich is tn addition to afiy_other modeof'-
hea.t transfer s4m_ty aceountsfor the internal energ_ term of-a mass
movtngfrom one-locat_LOn-t_nOthea'.

5?
 )
2.2.3 Ener;y Sourcesor Sinks

z.2.3.
Energysources or stnkso Q, a_e medelingelementswhtch.allow. the
impression of post_ive or negative-heatfllg.rates on the nodesof a thermal-................................
netwoflcIndependentof- conductorpaths to the node.
2.2.3.2 _Typesof Heat Sourcesor Sinks
Con,
nonengineering applications of heat sources in themal models
are:

• Solar andPlanetary_Heating
• _AerodynamicHeating
• Avionic Coldplate Heat Loads
• Changeof State Latent Enet_Jy
• Thermal Control Heaters

Ce,_onapplication for heat s_nks are:

• Changeof StateLatent Energy )
• RadiatorHeat _Jeotion

,_i
_ Z.2.3.3 ComputationalC,
• Aerodynamic onstderatlons
Cooling - Sourc.e_s
or S_inks
Heating rates-may be.impressedon dt-ffus-ton(finte capacitance-).or
arithmetic(zero capacitanCe) nodes, )Costthermalanalyzers provide a
separate entry block for entertng _ea_tng or cooling rates. For example,
the SIROAeomputer-p_'am uses the S_tRCEOATABLOat(for such entries.
In the usual case, heating rates a_e not consideredwheneomputtngthe
time steps for transient analysis, and large_heatln;_eates on low capaci-

__- tahoe nodesm_c create Instability in the network solution. Also the
_-- impression of large heat sourceson aHthmettc nodeswtth radiation (non-
linear) conductor__t_ched often Causeslarge erroneous temperature
oscfllattons in the art-thmettc and ad,iotntngnodes. Both of tJ,ese
dlfflculties 6an _e avoldedwittlthe use of the programcontrolconstantsIn-
cOt)orated tn most thermal network analyzers. Thesecontrol constants are -"
the ti_ step multtplicat_ton factor and the maximum cianqe atlowe61,
temperature il)

58

• . " --: : .... -_ _ . ,- : .. ." • ..- -',. _ "_ . _

 )
2.3 Network SOlution
J1

An area which is of utmost coece_n to theengineer is an-explanation

• ]

of the system of equ_ttunS which are solVe_ in a typical thermal problem,

At this point it must be completely understood that- the phzsical system is
reduced to a lumped-(nodalized} system and thcLLthe-Chotce of the nodali- ;
ZattorL has a far g_eater bear-Lng on the p_oblem than the partteutar 'i
numertcal_lutlo.-techniqlLe and:assumplztons--ffiat are usu_mpteyed in !_,..
the p_eblem so_lution__ Each _ume_ical solution method-fs_bounded by l
accur_cy_constraints-d4ctated-by the_patti cuia__set of assUmpttonstass_- ;_j
_ ctated with the particular problem umP.zxjotngsolutlon, - i

;:: The basic transient heat transfer equ_tton applied to linear- conduct;ton i]
problem is: t

___T
= _V_" T __9_
2 = _+_2 _ +_2 )!!
at ax2 ay2 az2 (i)

: The solution techniques most commonlyapplled-stem from a reduction of

:_ Equation (I). I-tcan be easily seen that any nonlinearterms such as )_,
i

' radiation must be linearized by somemethod if the above equation is to be ;

uSed.
)

Eor example the radiation term G{Tt 4 - Tj_) can be ltnearized by ):i

Gy(Ti - Tj) - G(TI4 - Tj_)_ (21 i_

J i i

J,! faetor. _h_ dalculati_Lof y and the llneart-

where-y is tl_l-iaearlzat_o, (i
i
_I zatien ef the C_,>is perfermed automatically by thermal networJ(-analy_zez_ !-_

I and wlll be explal_edin more detNl In the forwarddlffet'enclngLse_t-loe_ !!

I Tbe-process Of noda-TCzat-tenreduees the-v0lumeC¢_ d4mans_ns and-

_ properties so that equat.to_ (1) _'an be cast tn the f_l-lowtng form:

aT"" -

_ 59
 i )

Each ter_.of the summation,8ij(Tj-Ti)_ represent_._he.

heat rate_,
foTlowlnginto node - i from node - J, .I_grating equation (3) yields:
a

Ti(tnew) = Tl.(told
) + I=__..._ T.Gij.Tj-Tildt(
flew (4} f
....................
ILCLt
d J -

TO actua11_pe4_rorm_be-'E_teg__-aeion.4ndi (4-) on-8-computer

ca1_ iflequation-. /,
it would-be-mos-k-convenient-lf
the_.Jnteg_and
was--nota functlon_of_tlme:_.
 •. r

• ".. L.

T_

.. =_- 2;3.1 S_ead_Jtate

; A thermal Systcu,:lias_reached the stoady state when the net heat t'Low--
._i::_ -= rate t:o each diffusion _n<_arlthnettc node Ts zero. Fromequation-(3),
/: i_-- / this yields:

_ ) %(Tj-ri - 0_ (e)
_i _, Rearranging the terms tn_eq__.on=.(_8)..ytetd¢.

- The subscripts in equation _9) obscure t_eLegance of the steady state !

-_-- _.olutton, which can also be expressed as:

.....
:+: T_- [G] T (lo)

Equation (lO)claarly reveals that finding the steady state temperature

vector simply-amoemts to finding the etgenvectov-of LG] for an eigenvalue-
of 1. The only problem is that-,- for a-neWark with more than a few (say
_. _ 20) nodes,_the-calculation of- the--etgenveetor using an exl_lt¢-tt--algovttl_n,
::. such a_ Gausstan Eltmtnatten, would take a DcohtbJ-ttve ameunC-of time.
-: -- This _ tn addi ttoa to the fact. that-most- of::the elements ef__[G]_are zero_
= :.--_
::
suggests that- an A:teJ_atJ_ve-techntquesuck as expces_ed tn e_ (_1)
..... ;- would ..resUtJL.t.n a.huge reduction- in. requt red compu_eetime.

-_:_ .ft+l. N1 T-t (1t1

- - ..r_ Whenit iS also realized that-equatlew Ol_) ]s ov,ly-va-l-t_-_en the G-ma_rlX

• - is absolote4_ constant With-respeCt to te_pe£eture, t-t becomesclear that

____ Usl_j thts tochn_lue, tt ts simple-rotter to apdate the 6 maCrtx (e_a_u_ts

I equatten (1-1) _ the ol_}yrsotutten t_hntque tha_ ts untvecsally usabie,
_ ([) temperature var_thg G*s and ltneartze rad_ai4on S.'s)-based oh tke-neW-gt,ess
__- tempera_um-veetor T, t-+1-before_proceeding to the next IteratiOn.- In .

il same-1.engt.'%so that(11)_
addl-t-lonL eq_atlon _e-T2...vecto_ can betl_at
does-not reCLUlm extended
the-Tjto+I tnctude_a vector of the
and T1--v_ctors_be
t:empe_atures whl-ch represent the boundary condltlons placed on the network.

p__ytded. The 1"trst-ls slmgly a__xed number, N, whlchrepresents the---

maximum_umbe_of iterations to be perfonmd_ The second is a ¢on_ve._ence_:.
o_,_ as tt ts-mo_e_cQm_nly called,_ B:relaxatton crt_certon, 6., Wht_h tlS-
deftned asz...............................................................
Two cr_t_teria for ten_attng_the-tteratt_e.process a_ generally

In other we_ds, calculations cease when no "Mgniflcant" tmprevement from

: .... one iteration to the next is noted (where "significant" Is deftned tn
I < (12)
terms of _).

/i I Strong oscillation tn the l:emperature vector ts Often noted when I

analyzing a network which ts dominated by radiation effects. To hasten --)
(and in somecases, to enable)the eonvergence of the solutton_ a damping
=- faetor,_¢ . fs often applied to equation {11) to yfeld the more versat(le

:---i solution technique expressed tn equatton (,3,.

-_I+I= _ [G] Tj + (I-_;)_T

'i
CO " C_I) (*_3 l')

To summarize, the fully_general ftMte differencing technique .for.

steady state anal.__t$ 1.S-an tte_attve algorithm eeClutHng three paPe_eCer$,
t_, _, and _, fe_ the cenCrol of co_orgenc_. It ndght be noted here that
aPtthmetto nodes Mways receive a steady state selutten f_hntqUel even
when they _ppezzr in a-n_._twork.wk.i_ ts undef'go-t-nga t_anetent analysts_ L

i
 _ 2.3.2 Transient Anal.vsis J

2.3,2.1 ForwardDtfferenct.n9
Forwarddifferencing dertves tts namefrom the fact that a11 tempera-
tures are-extrapolated forward tn _.tme;t,_for the purposeof evaluating the
expression _n equaCtOn(S). Thts point-Is illustrated tn I_tgure- 2-25.

ton tnew

Ftgure_2-25. ForwardExtrapo3atton Usedtn ForwaPdDifferencing

_ L._) Combiningequations (5) and 1i) uflder the aboveassumptionyields the basic
forward d|ffe_n_ng equation:

Tt (t_t) " Tt(t) + _tt_AtGtj___j(t) - Tt(t)) 114)

As m'Lght,
be expeeted, the temperatueesat Mme-t are used to evaluate-an.v
temperature varyte_g,oP-radtatton conductor_, That ts, fo_ e_le, the
11neartzatton factor, y, _s computed:as:

x - (T-j2(t) + T12{t)) (T_I(I_) (IS)

The b_Stc forwaPde_t_'apolatton assump_ti_has t_ con_equeflcP,

s:
(a) the resulting solutlon equation ts expHctt, and (b)-the_solut]_n
can be unstable. When-_tis st_,_ed that an-_uatlon ts :,xpllct% it t_eans
that a1-1of+the unknownS-ape on the-leer" handst<le-an¢ all of the known
quantities aPe oh the right hand-st_; - bt-nce ai1 g_ the temperaCctPes
II at ttme t are lehown,eq_atto-n (14) deHnes the t*.,n_peratatres
at-t_me

83
.... • +
 )

t + At, eXp]-lcttly. Znst_.btltty results_ from_the fact thai equation (14)

Is, so to speak. "open-ended". Sy_.choostng__a sufftctestl_ large _t, the
new-.tee_ratures can be made--unreasonable .... it has bee_shewn (reference
3_.,.however, that if _L Ts restricted l:o be less.-than..the s_abl_lt.t_
factor, T, deftn,.-din equatton_(16),._then .the Solution nor_el]y_w.l]] be
stable ................... l:!

m_n-_
G--_.=-__=_CS_IN 06)

This_, commonlz
called-"theCSGHIN,)isthe smallesttime constantin-the......
thermalne_ork. The "CSGM)N"does not includeboundarye._. Sharp
gradientscoupledwith highmagnitudeheatingratescouldpossiblycause
Instabl
lIty.

It-is clear, then, the)forward differencing represents-a two-edged

• sword. On the one hand, the solution equation is explicit, which means
that-the volumeof-computations for each at step wtll be a minimum. But )
on the ether han_, the stze of the time step is limited to the smallest
time constant-In _he entire network. However,additional advantages
" aecrue from both of these conditions. Ftrs_, _he user neednot specify )
any convergencecrt_,ria such as were required by the steady state tech-
nique, and seeond,.._for_
normal Cases the user:need no_ speetf7 the ttm step,
At, since thts can"be c_nvenientty computedby the computerusing a simple
atgort-thm such aS given: in equation (17).

At- _ _ 07)

Another modelling po4nt _o be considered is that the maxtmum

A_
used teor a ,transit.hi: analysis is often-not restricted by the network,
But rather by the @requencyspectrumof-the boundary_nditiohs. [n
grOss towns, one cannot use a at of one hour to analy_e a network sub-
Jecte_ t_ a bound_y condition which var_es at tJ_ rate of 50 cycles pe_
hour. In t_TnePterm, Shannon'sSa_pltng TheorOm(referenct,_4) dtctates
that: _ I

.!
- 64 _
-" - / .- .::" : i. ::-... .......... " " ..... i
\
 • .:': •

-: "s > 2=c (18)

:i.2"-

•::: max_t--_ (19)

=. where=_ ts theJ_t_hes_ frequency...component In. the drt_tng .stgnal, Of _.

:: course, the 1lmt.t _xpres_e_-In equation_(1._) ts: theoz_ttcal .tn _hat t t ,"
_-. assumes,that tbe_s:ystem_Ts_a_per_ect:low-p__ass flltee, Thts_is t_ue, tn_
- practice, only at" nodal peters _atJ_J'.far awayfrom the bounda_. Hence,
4.: analyses-o.f_the temperature._response•near rap_ll.V, vat_! _g boundarte_.
:3"::_ wtll requtre a.At muchsmaller than the maxlmum expressedtn equation (19).
-_- It |s no act=talent,then, that forward differencing fomuJaCtonsenjoy
:. the _¢tdest and most frequent use tn engfneeNng prec_ce_ ..(l_er femula-
.*.o

-_- tlonshavepractical valueonly whentrmchotceof _ _s not restricted

_.._ _ by boundarycondtttons_
--;- 2.3.Z.Z BackwardDifferencing

.: Backwarddifferencing takes Its namefrom the fact that ali tempera-

at tlne t_ewam extrapolated In ttm In o1_lerto evaluate
backwj____.dd
: " ttte-_ tems per equation-l-15, Thts point ts Illustrated tn Ftgure 2-26.

to1 d tnL_ _I

I ¢tgUre 2-26. Backwa_ Extrapolation Usedtn BackwardOtff_n_tng

.... -S--- . , .. ' .- :'.-'" . ". " - ." .__.... , ".:: "?... • "; -- -
 ;_-"!_-i Using thls assumptlon-to canb_ne-_quatlons (5) and (7), and stad:Ing f:he
_, " :--_ result in mat_x fore yields:

:!_: T- (t+At) - T (:L:)__+-_t-[61C:] T, (t+At) (20)

..... ".r-I The_olutlon to equation (2) obvlousiy requ_es _a.t _..mat_1_.Inverslon

_::_:i [ be perfomed, bu_Tt.also requlres, sttlc1_..that the.G/_Cmatrlx b_ '" i

F._. _:-i tempe_ture I nvarl an_... (Otheml-se_he-bas I c assumptl
on-would require ,':i
_1_":_- th_t:LG/C] be__eva_uat_d,for.:[- T_t_At)__and_._
of course,_these.tempera- i
_, .
_l_, _ t_res are-.oLy_" known,) In. practice,, then, equ._t1_L(20)- Is rewor_ke_

:._.... into a form wht_h lends _el_:l:o .t_erat_ve_calculattons .......................... J

• T_.+i . _'_
Ct Tl(t) ,
+ I Gl-iT:l_'(t+_t) " _
.;. _--'E + GtJ )

-_: ]? S_nc_T(t+At) appearSon both s_desnl_-tJneequation, thts formulation ts

it _tl1 be noted that _hen _t appreachestnftnt_, equation (21) re-

:,_" ;- t duces to the s_mad_State equation dlscusse_In Section 2.3.1, Hence.,
can be deduca_ has p_oven
_f • _t been '_n reference 3) that hadcwa_J
- . dtffemnetng ts stoble for any _t. I_ addition, backwarddifferenetng ts
_: _ sub,eat to the-sam _e_-of 1todaY-tonteradnatton.¢etterta,_N and _, a:_wer_-
- applle_ to the s_eady state setutton: Stability for large AI_. howew,r, does

_,.
-_--- net pm©lude
faotOt___, _sthe
alsepossibility
pvovtd_d toof ensure
an oscillating $olUtten_-so a damping
the even_uatcenvergence_o the
:_ :_ so'iutton,
I:

.:- The g_eat-aavantage of backwarddifferencing ltes tn the effect-lye- i

;_- selection of lit, Ourlltg peNods of rapidly va_g boundary cond_ttons_
:_ ; _t m_ be m_cedto as smail a value as de_tred, and then, durtng peH6ds
ot_ slowl_ _L_t_g bounda_con_t_tons,it can be enlarged _o a compaf.|ble
L}
va_ue, wtthout regard for_the mtntmumtt_e constant, T, of the network.
It als_ happens that when a large At ts app_opr_to,_the number-of
Iterations necessary to satisfy _he relaxation crtl_rIo,, 6, is relatively
smU. On the other_hand,.when At ts_compresse_ to less than the CSr_ZN _
(_), the.number 6f tterat|on_per time step will still_ be large-when COrn- L
pared .to fo_arddtfferenctng, I,

2.3. 3 Suwa_ of Othe_ TechnlqUe_ i

:

Numerousother approaches to fOrmu]attng: f_tntte, dtfferenct,g so] utt ons...................

i_
are _va1Iable-(reference 5J. However, al]..of these..g_ral ly- amount-to ;'!

modiflcatlons, o_. co.fo_natlons of bas_c..for_ard.and backw&rd dl f_erenclng ........ i I

Central differencing, for example, t__a_ tmpllc|t ter_,ntque which computes _:_
the current-Iteration tenperatures as the arithmetic average of the forward I
and backward dt ffe_enGtng predlct!ons...

predJctlonuses the forward differencing

As anether example'; exponenttal

equatlon but-provides stability

i:
i
for all at by exponentially "derating" the terms according to the j
,:) ratlo of _t to At. _:

iI
{!
/ i

L.._

67
 2_4:._Hodel
tn9 Parameters
+_+-+ + The solutions-to the-heat transfer equatt_n develepedin. the pre-
_- •
_.-:_. ceding sections- required that con_ctnuous var_.ablas-bequanttzed. Spatial
_+ + vat1ables wn_e quontt zed th rough.+tbo
-artt fl ces.of nodes and conductoPs,
and time was quanttzed into "time +tops" o_ size at. By assuming,_oPthe
F_--_ sake-of discussion, that:_all-nodes are..cubtcal._wtth s_le.AX_ 1_e_L_.this 'I
z--.
++: - +x-c;n be usedas..a.g.eneral,spal:ta! qua_um,.._md-it_can_be.related to the ....... ,:

_ time_quantum,.Z_t_._
For.example, the forwa_d_dtfferonctng stability
;_ _ criterion, _ (and.therefore_th_maxlmum_tlme..s_ep)is tel:arealto Ax i_
_-_-, as shown:te eq_tfon (21): _
!
_X 2
--"-t '+ +tmax (21)

..... ,+ one dimensional: m - 2

._ :=- _ dtmonstonaT: m = 4
; -- three dtmensiomat: m = 6 )

=+: :'_" Since the finite differencing solution approachesthe exact soluttnn
: i as _x and-_tapproach ze_, it is logical to ask if anything imposesa
__ ii T minimumon these values. The-answerts yes: cost and computerco_e memory ..
..... space....
Cle_l_+ the Iafterconst_alni_s tPlc_ ex to-anexplI_it non-
) !
_-: '-+- zero ndnimum+..._That-is,
z smelt ax meansa la_g_ numberof nedes and
_:"
!- conducted, and the computer'smemorymust-contain enough-spaceto hold
p+_::_ all of _he pa__moto Ps (_apactt_nce_temperetul_, conductance,etc. )- ass_ II
_+_ ctated with these modelingelements, In addition to ustng muchmere-corn- _
purer _lme for analyses, a large modelalso eosts mopeto develop than a
i_ _:_:-:- small one.

_.._-- Tile time stop can he Chosenas small as desired with a consequent
thcPease in the eon_uter run ttme required fop analysts. The relatt0nsiltp
:--J betweenrun ti_ and timestep stze is linear fop forward differencing
+_ - + becauseexectl_ one '+iteration" ts _qut_N for each tim step, FOr
) +_--, implicit mothod_, liewevet_, the rela¢lOnslitp t_ hot _S predictable, 5eomise

..... i

+ 68
r_ '

i+_:
_ / - ++_ " -: +-+ :-: :..... : _ ++ "': = +-:-+:- :_:_ +:
 the numberof-Iterations ts controlled by-the convergencecriterta, N
and 6. In addition, the greater the numherof Jterattons, and the greater
the numberof nodes processed-duringeach iteration, then the greater
wlll be_the susceptibility of the answersto computerround-off error.

Looking at the problemfrom-.theo_her direction, Ax may also be..chosen

large enoughto lne-lude_theentire _hermal sysJ;e_tnone, node_ The-time.
step, however, at least.for _orwa_l dtf-ferencing, .!.s limtted in stze to the ................
_. CSGMIN.Thls_1_mlt...Is not imposedon.the Impliclt solu_lon technlo,ue_-,
although a..dam_e_d,
oscillator. R sponseJaayresult.when At. ts too large.
Stnce_uch_an oscillating response can be crt-Clc_ll£._dampedby ustn9 the •
damptn9factor, T, there,..ts no deflnl1_ve equation for..the.maxtmumAt.
"* that maybe usedsucoessfully with implicit solutions. To further compli-
cate thts understanding, there is no definitive equatton for the value
of Gwhich wt]l yield the mosteffective damping, although a _tQl_we_ of 0.5
has been routinely usedwith good results.
To s,Jn_lartze, the thernlal math-nlodelerts faced with the task of
• t 'J

designing a model and selectir_g a solution technique which wtll yield

good, stable answersfor the least cost. To do this, the modelermust
choosevalues for the following paramters:

Ax - nodesize.
At-- ttJnestep
N.- t teratt ens
6 --convergenGec_ttert on
- dampingfaotOr

FOrWarddtfferenetng, which doesnet requt_ N-, 6, or ¢, and which deftnes

a_ leas_ the maximumAt, Is often chosenas the solut4on technique stmply
becauseit reduces-the numberof parameterswhich mustbe J_jgled about.
In the hopes of leadtng to a mute logtcal rationale for selecting model
and solutton paranleter_, a Gase study using variOUSmethedsis presented
in Sectten 3.0.

69

• L'
- -T. :.-', "..--- - : .. i
71
 i
: analyzing a model-. Fo_*a given compQter, the numberof node-lteratlons
which can be run In a g_ven mount of tlme Is falrly equal rogardless of
the solution technique. ,_ctually, the comparison of node-lteratlons for r

a one-dlmenSlonaLmOdel ve_us node-lteraClons for a three dlmenslonal

model ts..not stk_lctly reasol_able since _the.latter model wt]l contain more
conductors per node... However, loP.order.--of-magnitude comparisons, node- i
t terottons represents a stmple and useful ,_osurs of cost.

Figure 3-2 shows___herosults for x =. 1 t:nch and-_ =-25.seconds. The_

e_:fect_ of-time-s_o v.aHattonS and solut;on techntq_u_e-ona 12 node mudel
_ are sho_n_ A_ might: be-expected, tt appears tl_t the smaller the time
step_ the groafer the accurac3r and the 3reater the cost. Since ahost-ef-
minimax hypotheses could be formulated from thts ftgu_, al1 of which or
; none of which might be Crue, no further hypotheses wtll be stated hero.
I As indicated t, the Introduction, tt remains fo_ the reader to abstract from _-
-- _ this figure as muchmeaningful information as he ts able.
I
i Figure 3-3 ts stmtlar to Figure 3-2, except that the effects of node I
I stze variations are emphasized. Since the At and 6 fOr the tmpltctt routines
were held constant over the various cases, t_o additional cases a_e plotted
for a smaller Atand 6. Again, as might be expected, it appears that the
smaller the node size, _x, a_Lhence, the grater the number of nedes,
then the greater the accuracy and the greater the cOst. Further conclu-
sions wtll be-teft for the reader _o fomulata.

Figure 3-4 shows the.e_fects oF ttmu step variations-and solution

techntque:_n the results for x =-..6inches _nd t - 100-seconds. It-_tll
be-nottGed that..the temperature dispersion ts not as g_at as for x - 1
tnCh and-t-- 28 seconds.-Thts iS to be expected since the mass of the bar
from x = 0 to x - 6 serve_ as s muchmore effe_ttve lov¢ pass f_lte_,
which tmpltes g_eater accuracy for _a gtven sa_pllng froqueney (i.e. _).

FtguR 3-5 'is stmllar to Ftgure 3-4 except thet tt htghltght_ the _:_
effects of node stze variations. For a gtveh solution technique, tt was
e_Pect_d that _11 points wouid 1te on a Smoothcurve. Thts d_d n_t hold

72

/-
for backwarddlfferenc1ng with 24 nodes....To dlscover why, two addltlonal
cases weTe run wtth ft _st, the time step,_ar.d then the t_mestep and i/
relaxation crtt_Jon-_edu_ed. The results tndlcate that:the roll-off at
la_Je node-tterat|ons was due to the accumulatedrelaxatton error. :J

73
 ..-. :_J Figure 3-3. The Effecf,- of a Variation in NodeStze at _P.5 seconds
:_ -- and x,,t inch ..

=_ -.:- _ 0
I-I -FORWARO
(At=t)
"_"
- .: --" . .... -CgNTRAL
(At'=3"r)
<C_-8ACKWARD (At.=_3t)
•: 'r=CSGMIN .........
::i -2_: 6
::"_. +_oo - '12)/NO...OFNODES
-
:__ 24J i'
i:

' I

.- _ '-,- L. 24

•= _ _-!

- Ti ip,
:"
F'- ' • 24A_t s,,O.O5
=; :--. -zo _ _ 24 At=_,6=_O .

-. 24
i--,

- :LF_ 6
: - -zoo ...... []6 12

_.;? .....
6

..... -1000 ..... ""-

_" -,-_-- . .,_-d IL - -- -- - ] _ j [I I-IlilllI J LJ
1111': 114111 ...._-- lI iI I11b -- *- .-

.... "------ ----I ..... l jii."

lllll -- LI I III
-4-LIllE II IkllI [ _.- ),-I
l]-l_lI II
i
-- " iBI I I lllll 0 - I [till ¢_ -- I I II _ .... IIIII Cl
]. .4 a
Ill
= o
%

"" -- NODE-
tTERATION_
,:" °
z _
.g-
75
-i :
 3.2 p_her One-DimensionalCases

i
_
-
In the discussion of Section 3J, dimensionswere retained_so that_e- ....
::_2[ reader would_havesomaintut tt ve understandtng of the relattve magnttudes
of the vaNous _arameters. However,since the basic heat transfer equations
_-- can.be castin a dimensionless fore, .the results.el_ Secti_..3.1 can_be......
applied to other one-dimensionalconfigurations, as required.

:_: 3.3_.._]_xo-Dt
menstonal P1ate of-_tal __-
This groupof cases treatsa plate of alumina, ntne inches on each
;= st de and-onetnch thick....Bounda_J condtttons are imposed•as.._t
em _,__ed t n _ ._
_-- Figure 3-6, andthe temp_e_t_u_at the c_enl;erof the plate,. 100
seconds, is desired.

.:E Ftguma3,6. Boundary_Conditions

on Twe-DtmenstonalPlate

'_- The _esulte of a variety of analyses of this plate are presented tn

Figure 3-7 and 3-8. The results in Figure 3-7 appea_to be tnco_ectly
•!;'_ plotted since they appear to convergeto e potnt 10 degrees f_om the

,_ "assumed"answer. However,the results Shownin Figure 3_8 confirm that

lt_ the solutton must be eloseP to 290°F than 280°F. The difference oecurred
)_i__ becausethe modelused fbt" _e cases dtsplayed in Flgure 3-7 cofltatned
_ only 36 nodes, and this num_r was evtdeetly tnsufftetent to accurately
_)_ represent the character of the heat fle_ tn the pla_e. Thts problemdtd
!_ not arise for the one-dtnenstonal cases tn Section 3.1 becausettle heat

t_ 78
z
 flow, by-deffnttton, alv_ys foltowed the_Gonduel:o_.pathsexactly. In
selecttno_ finer and finer-rashes_for-the 1_o dimensional_case, _t is
ev.tdento_that_the.modelledheatflow paths (t.e_.-the_conductorS) wtll 1te
eloSer and.closer to:..the actual paths, wtth • consequent__tncrease
tn
temperatu_r_,
predtc_ On aGcura_,, i
3.4 i_Other Two-OJn_.n.s(onal
Case._
¢

The prtncip_les of st_a_Lty and non-dimensional analysis m_y.__ i

applle_ t_. the resul_ts'Jatr.Sectton3.3 for other comb|nationsof mat_lal
J

ty_e and-dtr_nstons, _ i

../

m_ I
 ¢,

F|gure 3-7. The Effect of Variation tn T4meStepfor Center ef Plate at

t_100 seconds
...... l
ezo00 t
0- FORHARD
D- CE.N'r_L (_,,1.0)
O" BACKtfARD(6-T.O)
t = CSGHIN
_5 = NUHBER OF NODES
TTHESTEPINDICATED
+100

+1o
_-'C_

3 0.53
0.1'I
_ aeqL __
-10 _ _ v
I

0 o
33 3

O
10_
t
-100 l _" 3z T

_.lO0_ iI ii 1.i|1..
i ii _-- [ iL Li
il _ i I I ill,
!!![].
lill
, lill4--.
I-IIIII- , IJl
I I 1[" i_
I- i i llli- k.I II 4 I I liiii

j.¢ Q o

0
_J.?
NODE-ITERATIONS
3

80,

..... S--; _ :;- "- " : T- - _ .... "_:l ,_

 >* 81
_ 0 81(_t-.]_)
. _oa ±1

_: , .j_ _ 81 81 (,t,_.,=.05)

_ 36 81 (At=%6--.05)
-'- - 10
: 36 81

0
9 36
: 09 [3
. -.100
_- 9

_-., -1000 _-