Category : Science and Education
Archive : SSIIM11.ZIP
Filename : MANUAL

Output of file : MANUAL contained in archive : SSIIM11.ZIP

ÿWPCÐq
ûÿ2^ÿÿ BN¤ÿÿZTimes Roman3|Z-êfô\ PÂC&Pþþþþþþþÿþÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ06/01-93 20.46 manual manual ÿÉ.ÿ==·==========\3QQQQQ£zzQpQpQpQpQ=3=3=3=3\\\\\\\\\\Q\\\\\f\QQQzQzQzQzQ\pQpQpQpQ\\\\\\\\=3=3=3=3G\p3p3p3p3p3\\\\\\£z=z=z=fGfGfGfGp3p3p3\\\\\\\pQpQpQ\p3\z=fGp3\\\\\N@.S\=Q\\\\\39\7\7!QQ\·==\\´ff·=Q·7Ñggee3g< ÿÿHÿÿØÿÿ`ÚÿÿX: ÓUSNOÓÓ .,,. ÓÐ÷3Ø'3Ø'StandardÐ6Â&Ð6Â&StandardÿÿrJet IIISi PostSÖæõ*£°°°°÷ÐÐÐÑ#oô\ PÂCb×XP#Ñ*£"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^2CRdd§CCCdq2C28dddddddddd88qqqY¸zoCNz²ooz½zC8C^dCYdYdYCdd88d8ddddCN8ddddY`(`léC2CCÈ!CCCCCCCCCCd8YYYYY²YzYzYzYzYC8C8C8C8ddddddddddYdddddodYYYYYYYdzYzYzYzYddddddddC8C8C8C8Ndz8z8z8z8z8dddddd²CCCoNoNoNoNz8z8z8dddddd½dzYzYzYdz8dCoNz8dddddNF2[dCYddddd7>d<d<$YYdÈCCddÄooÈCYÈ<Ñqqnn8!qBBnnnqÅÅyyÐPq7¥c1RÅÅÅyyÐX¥yycnnnÎ~nyRzczXzcyhCBnn²dhcnnonvyXzXshn~XyBBnss~|y~~~~~~~~~~~~~~~~~~~XXXXXXXyyyyyyyyyyyyyyyyyyyyBBBBBBBBBBBBnnnnnnnsssssssssssszC:2f=.°Z-êfô\ PÂC&Pþþþþþþþÿþÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ7oC2°b×oô\ PÂCXPþþþþþþþÿþÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿC28ddddddddddCCrrrdºzNd½zoÈC8CtdCdoYoYCdo8Co8§odooYNCodddYO,OhÿÿÿÿCÈ!CCCCCCCCCCo8dddddÈYYYYYN8N8N8N8oddddooooddoddûÿ2éÿÿÄ
ÿÿÔ
0Y`3|06/01-93 20.46 manual manual ÿÉ.ÿTimes RomanTimes Roman BoldTimes Roman Italic"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^2CTdd§CCCd2C28ddddddddddCCd¸zzzzCYo§zzdoz§zooN8NTdCddYdY8dd88Y8ddddNN8dYYYNP7PléC2CC²!CCCCCCCCCCd8zdzdzdzdzd²YzYzYzYzYC8C8C8C8dddddddddoYzddddoYdzdzdzdzdYYYYdzYzYzYzYddddddddC8C8C8C8YYo8o8o8o8o8dddddd½zNzNzNdNdNdNdNo8o8o8dddddd§oYoNoNoNdo8dzNdNo8oYoYdddNF2idNdddddd7>d<d<+ood²CCddÄdd²CoÈ<Ñnn8!BBnnnÅÅyyÐP7¥c1RÅÅÅyyÐX¥yycnnnÎ~nyRzczXzcyhCBnn²dhcnnonvyXzXshn~XyBBnss~|y~~~~~~~~~~~~~~~~~~~XXXXXXXyyyyyyyyyyyyyyyyyyyyBBBBBBBBBBBBnnnnnnnsssssssssssszCûÿ2 ``{`Û Z;"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^(5APP|555PZ(5(,PPPPPPPPPP,,ZZZGtkktbYtt5>tbttYtkYbttttb5,5KP5GPGPG5PP,,P,|PPPP5>,PPtPPGM MWé5(55 5555555555P,tGtGtGtGtGkkGbGbGbGbG5,5,5,5,tPtPtPtPtPtPtPtPtPtPtGtPtPtPtPtPYPtGtGtGkGkGkGkGtPbGbGbGbGtPtPtPtPtPtPtPtP5,5,5,5,>tPb,b,b,b,b,tPtPtPtPtPtPtk5k5k5Y>Y>Y>Y>b,b,b,tPtPtPtPtPtPttPbGbGbGtPb,tPk5Y>b,tPtPtPtPtPN8(HP5GPPPPP,2xxP0zzPx0GGP 55PPYY 5G 0ÑZZXXr,Z55XXXr{rrZ``§@Z,rO(Bn``§{rrrrF{{{``iOXXXrrrÎtekX`BbObFbOt`wS55tXnXPtSgOtX{XYX_`FbFn\zStXn{neF`55X\\nec`reeeeeeeeeeeeeeeeeeeFFFFFFF````````````````````555555555555XXXXXXX\\\\\\\\\\\\nnnnnnnnnnnnnnnnnnnntbt5tn{"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^.=f\\·===\i.=.3\\\\\\\\\\==iii\«zzpG\zpfz·z=3=k\=\fQfQ=\f3=f3f\ffQG=f\\\QH(H_é=.==·==========f3\\\\\·QzQzQzQzQG3G3G3G3f\\\\ffff\\f\\\\pf\\\QQQQfzQzQzQzQ\\\\\\ffG3G3G3G3\fz3z3z3z3z3ffff\\·QQQfGfGfGfGz=z=z=ffffff·\zQzQzQfz3fQfGz=\\f\fN@.c\=\\\\\\7=\7\73\\\·==\\·ff·=\·7Ñiiee3i<d<d<$YYdÈCCddÄooÈCYÈ<Ñqqnn8!qBBnnnqÅÅyyÐPq7¥c1RÅÅÅyyÐX¥yycnnnÎ~nyRzczXzcyhCBnn²dhcnnonvyXzXshn~XyBBnss~|y~~~~~~~~~~~~~~~~~~~XXXXXXXyyyyyyyyyyyyyyyyyyyyBBBBBBBBBBBBnnnnnnnsssssssssssszCÓUSNOÓÓ .,,. ÓÐ÷3Ø'3Ø'StandardÐ6Â&Ð6Â&StandardG-1)HPLAII21.PRSêfô\ *£°°°°÷ÐÐÐÑ#oô\ PÂCb×XP#Ña8DocumentgÁDocument Style Styleº ÂXÂÂX` ` ÂÆÐ ` Æ

ûÿ2!pkkô¥_ a4DocumentgDocument Style StyleØªÖÁÁÃ ÃÃÃ.ÄÄ Ä Äa6DocumentgÁDocument Style Style½³GÂXÂÆÐ Æ

a5DocumentgÁDocument Style Style}±ÂXÂÆ(#Æ

a2DocumentgDocument Style Style<ñoÇ
Ô ?È ÔÃ ÃØA.Ø ÃÃ××××

Ä ÄÄÄûÿ2]#v6!t¬! " £º"a7DocumentgÁDocument Style StyleyýÂXÂÂX` ` ÂÆ(#` Æ

BibliogrphyÁBibliography¬:ÂXÂÁ°°
ÁÆ(#Æ

a1Right ParÁRight-Aligned Paragraph Numbers:`SÁ@ ÁØI.Ø Á°°
ÁÂXÂÆ(#Æ

a2Right ParÁRight-Aligned Paragraph Numbers Cýö ÁÁÁ@ø` ÁØA.Ø ÁÁÂ°` ` ÂÆ(#` Æ

ûÿ2V&
§#¬6$ µâ$
¿%a3DocumentgDocument Style Style
B½
ºb
Ô ?È ÔÁÁÃ ÃØ1.Ø ××××
Ä Äa3Right ParÁRight-Aligned Paragraph NumbersLã!·
ÁÁÁ` ` ÁÁ@P
¸ÁØ1.Ø Á` ` ÁÂ¸¸ÂÆ(#¸Æ

a4Right ParÁRight-Aligned Paragraph Numbers UjÁÁÁ` ` ÁÁ¸¸ÁÁ@¨ ÁØa.Ø Á¸¸ÁÂ` ÂÆ(#Æ

a5Right ParÁRight-Aligned Paragraph Numbers
_oãÁÁÁ` ` ÁÁ¸¸ÁÁÁÁ@hÁØ(1)Ø ÁÁÂ¸hh#ÂÆ(#hÆ

ûÿ2Å)È&ÐP'Ù (Ìù(a6Right ParÁRight-Aligned Paragraph NumbershÎÁÁÁ` ` ÁÁ¸¸ÁÁÁÁhh#ÁÁ@àÀ$ÁØ(a)Ø Áhh#ÁÂÀÀ(ÂÆ(#ÀÆ

a7Right ParÁRight-Aligned Paragraph NumberspfJÁÁÁ` ` ÁÁ¸¸ÁÁÁÁhh#ÁÁÀÀ(ÁÁ@°*ÁØi)Ø ÁÀÀ(ÁÂh-ÂÆ(#Æ

a8Right ParÁRight-Aligned Paragraph NumbersyW"3!ÁÁÁ` ` ÁÁ¸¸ÁÁÁÁhh#ÁÁÀÀ(ÁÁ-ÁÁ@p/ÁØa)Ø Á-ÁÂÀpp2ÂÆ(#pÆ

a1DocumentgDocument Style StyleXqq
ÙÙ
Ô álÈ ÔÁà^ì)ÁÃÃÃ Ã××ØI.Ø ÙÙÙÙ××

ÄÄÄ Äûÿ20-+÷)"+",©,Doc InitInitialize Document StyleÿÅ×

×Ô0*0*°Ô Ò I. A. 1. a.(1)(a) i) a) I. 1. A. a.(1)(a) i) a)DocumentgÒÙÙTech InitInitialize Technical Style¨.
kÒ I. A. 1. a.(1)(a) i) a) 1 .1 .1 .1 .1 .1 .1 .1 þTechnicalÒÙÙa5TechnicalTechnical Document Style)WÙD¾ÁÁÃ ÃØ(1)Ø . Ä Äa6TechnicalTechnical Document Style)àDÂÁÁÃ ÃØ(a)Ø . Ä Äûÿ20®b-§.¦·.Â]/a2TechnicalTechnical Document Style<6óèí
Ô ?È ÔÃ ÃØA.Ø ÃÃ××××ÄÄ

ÁÁÄ Äa3TechnicalTechnical Document Style9W¢g
Ô 2È ÔÃ ÃØ1.Ø ××××
ÁÁÄ Äa4TechnicalTechnical Document Style8bv¢{Ô 2 ÔÃ ÃØa.Ø ××××
ÁÁÄ Äa1TechnicalTechnical Document StyleF²è!<
Ô ?È ÔÃÃÃ Ã××ØI.Ø ÙÙÙÙ××

ÄÄÁÁÄ Äûÿ2þ;Q0×0]1`5a7TechnicalTechnical Document Style(@DÆÁÁÃ ÃØi)Ø . Ä Äa8TechnicalTechnical Document Style(àDËÁÁÃ ÃØa)Ø . Ä ÄPleadingHeader for numbered pleading paper«P@n¹ÐÈÐÐ °° è ÐÕÈ$]ÐÈÐÐ °°X° ÐÐÐX°` ¸hÀpÈ xÐ (#%Ø'0*,à.813è5@8:ð
ÁHHÀÁ1

ÁHHÀÁ2

ÁHHÀÁ3

ÁHHÀÁ4

ÁHHÀÁ5

ÁHHÀÁ6

ÁHHÀÁ7

ÁHHÀÁ8

ÁHHÀÁ9

ÁHÐÀÁ10

ÁHÐÀÁ11

ÁHÐÀÁ12

ÁHÐÀÁ13

ÁHÐÀÁ14

ÁHÐÀÁ15

ÁHÐÀÁ16

ÁHÐÀÁ17

ÁHÐÀÁ18

ÁHÐÀÁ19

ÁHÐÀÁ20

ÁHÐÀÁ21

ÁHÐÀÁ22

ÁHÐÀÁ23

ÁHÐÀÁ24

ÁHÐÀÁ25

ÁHÐÀÁ26

ÁHÐÀÁ27

ÁHÐÀÁ28Ó À+ ÓÕ"âÿÿÿÿ¸ÿÿXÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^d£ÈÈM7ÈâddoÈÈÈÈÈÈÈÈÈÈooâââ²p!!ôÞ!!!ôd!!Þ!Þô!!z!!ôo¼È²È²È²ÈÈooÈo7ÈÈÈÈoÈÈ!ÈÈ²ÀPÀØédCÈo!²!²!²!²!²d²ô²ô²ô²ô²oooo!È!È!È!È!È!È!È!È!È!È!²!È!È!È!È!ÈÞÈ!²!²!²²²²²!Èô²ô²ô²ô²!È!È!È!È!È!È!È!Èoooo!Èôoôoôoôoôo!È!È!È!È!È!Èd!ÞÞÞÞôoôoôo!È!È!È!È!È!Èz!!Èô²ô²ô²!Èôo!ÈÞôo!È!È!È!È!ÈNdµÈ²ÈÈÈÈÈn|,,Èx00È,xH²²ÈÈÈÞÞ²xÑââÜÜoCâÜÜÜ3âññ¡ ânIÆc¤>ññ¡3°I333ññÆÜYÜÜÎ!üÜñ¤õÆô°ôÆ!ñ(Ð"ÜÜdÈ!ÐÅ!Ü3ÜÞÜíñ°ô°æ1Ð!Ü>3ü°ñÜææüøñüüüüüüüüüüüüüüüüüüü°°°°°°°ññññññññññññññññññññÜÜÜÜÜÜÜææææææææææææ!ô!!3ûÿ2jS0<ÎBlI`
P"âÿÿÿÿ¸ÿÿXÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^dÞÈÈMÈäddoÈÈÈÈÈÈÈÈÈÈäääÈt!!!ô77È7z!7ô7!Þ!!!!oèÈÈÞ²Þ²ÈÞoÞoMÞÈÞÞ²ÞÈ!ÈÈ²XÐédCÞo!È!È!È!È!È!!²²²²²oooo!Þ7È7È7È7È!Þ!Þ!Þ!Þ!È!È!Þ7È7È!È!ÈôÞ!È!È!È!²!²!²!²!Þ²²²²7È7È7È7È7È7È7Þ7ÞooooÈ7Þooooo!Þ!Þ!Þ!Þ7È7È!!²!²!²ÞÞÞÞ!Þ!Þ!Þ!Þ!Þ!Þ!!È²²²!Þo!Þ!²Þ!È!È!Þ7È!ÞNdØÈÈÈÈÈÈÈx,,Èx++È,xoÈÈÈÈÈÞÞÈxÑääÜÜoCäÜÜÜ3äññ¡ änIÆc¤>ññ¡3°I333ññÆÜYÜÜÎ!üÜñ¤õÆô°ôÆ!ñ(Ð"ÜÜdÞ!ÐÅ!Ü3ÜÞÜíñ°ô°æ1Ð!Ü>3ü°ñÜææüøñüüüüüüüüüüüüüüüüüüü°°°°°°°ññññññññññññññññññññÜÜÜÜÜÜÜææææææææææææ!ô!!3"âÿÿÿÿ¸ÿÿXÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^Kdzúéddd©KdKSSS©©©ÙÈÈÙ·§ÙÙduÙ·ÙÙ§ÙÈ§·ÙÙÙÙ·dSdddSSSéduSÙ<¢édKdd,2ddddddddddSÙÙÙÙÙÈÈ····dSdSdSdSÙÙÙÙÙÙÙÙÙÙÙÙÙÙÙÙ§ÙÙÙÈÈÈÈÙ····ÙÙÙÙÙÙÙÙdSdSdSdSuÙ·S·S·S·S·SÙÙÙÙÙÙÙÈdÈdÈd§u§u§u§u·S·S·SÙÙÙÙÙÙÙÙ···Ù·SÙÈd§u·SÙÙÙÙÙNiKdS]ááZääáZ6,dd&§§,d,ZÑ©©¥¥ÖS2©cc¥¥¥ÖæÖÖ©((µµ9x©R÷ÖJ{Îï(((µµ9æÖÖÖÖ÷æææµµÅ¥¥¥ÖÖÖÎÙ½È¥µ{¸··ÙµÞdcÙ¥Î¥ÙÂÙ¥æ¥§¥²µ·ÏåÙ¥ïÎæÎ½µcc¥Î½ºµÖ½½½½½½½½½½½½½½½½½½½µµµµµµµµµµµµµµµµµµµµcccccccccccc¥¥¥¥¥¥¥ÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÙ·ÙdÙÏæ"âÿÿÿÿ¸ÿÿXÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^Kd§,úddd«KdKSdd«««ÙÈÙÙÈ·ééuéÈÙé·éÙ§ÈÙÙ,ÙÙÈdSd®d§§d§Sd§Sú§§§ud§ÙvBvédKdd,2dddddddddd§SÙÙÙÙÙ,ÙÙÈÈÈÈuSuSuSuSÙ§ééééÙ§Ù§Ù§Ù§ÙÙÙ§ééÙÙ·§ÙÙÙÙÙÙÙÙ§ÈÈÈÈééééééé§é§uSuSuSuSé§ÈSÈSÈSÈSÈSÙ§Ù§Ù§Ù§éé,ÙÙÙÙ§u§u§u§uÈdÈdÈdÙ§Ù§Ù§Ù§Ù§Ù§,ÙÙÈÈÈÙ§ÈSÙ§Ù§uÈdÙÙÙ§éÙ§NiK¢dZcááZààáZS,dd,§§,d,ZÑ««¥¥ÖS2«cc¥¥¥ÖæÖÖ«((µµ9x«R÷ÖJ{Îï(((µµ9æÖÖÖÖ÷æææµµÅ¥¥¥ÖÖÖÎÙ½È¥µ{¸··ÙµÞdcÙ¥Î¥§ÙÂÙ¥æ¥§¥²µ·ÏåÙ¥ïÎæÎ½µcc¥Î½ºµÖ½½½½½½½½½½½½½½½½½½½µµµµµµµµµµµµµµµµµµµµcccccccccccc¥¥¥¥¥¥¥ÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÎÙ·ÙdÙÏæ"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^2CoddÈ§CCCdr2C28ddddddddddCCrrrdºzNd½zoÈC8CtdCdoYoYCdo8Co8§odooYNCodddYO,OhéC2CCÈ!CCCCCCCCCCo8dddddÈYYYYYN8N8N8N8oddddooooddoddddzodddYYYYoYYYYddddddooN8N8N8N8do88888ooooddÈYYYoNoNoNoNCCCooooooÈdYYYo8oYoNCddodoNF2ldCdddddd `üV^\ZÿÿÛ)"ÿÿÿÿmÿÿ½ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ^(1<P|kt|b|tYktt ttk5,5]P5PYGYG5PY,5Y,YPYYG>5YPtPPG?#?Sé5(55 5555555555Y,tPtPtPtPtP ttGkGkGkGkG>,>,>,>,tY|P|P|P|PtYtYtYtYtPtPtY|P|PtPtPbYtPtPtPtGtGtGtGtYkGkGkGkG|P|P|P|P|P|P|Y|Y>,>,>,>,P|Yk,k,k,k,k,tYtYtYtY|P|P ttGtGtGY>Y>Y>Y>k5k5k5tYtYtYtYtYtY ttPkGkGkGtYk,tYtGY>k5tPtPtY|PtYN8(VP5PPPPPP05xxP0xxPx0,PPP 55PP YY 5P 0Ñ[[XXr,[55XXXr{rr[``§@[,rO(Bn``§{rrrrF{{{``iOXXXrrrÎtekX`BbObFbOt`wS55tXnXYtSgOtX{XYX_`FbFn\zStXn{neF`55X\\nec`reeeeeeeeeeeeeeeeeeeFFFFFFF````````````````````555555555555XXXXXXX\\\\\\\\\\\\nnnnnnnnnnnnnnnnnnnntbt5tn{2f=.°Z-êfô\ PÂC&Pþþþþþþþÿþÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ7oC2°b×oô\ PÂCXPþ þÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ#mÞd°.Þô\ PÂC°PþþþþþþþÿþÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÀ#mèd° .èô4 p(AC°þþþþþþþÿþÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÚR§dK°ù#§ô\ PÂCPþþþþþþþÿþÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÀÚR®dK° è#®ô4 p(ACþþþþþþþÿþÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÀ7tC2° vÏtô4 p(ACXþþþþþþþÿþÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿW!C(°"ACô\ PÂChPþþþþþþþÿþÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÀ2j=.° Î>êjô4 p(AC&þþþþþþþÿþÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿt,Y5(°B,¬Yô\ PÂCàPþþþþþþþÿþÿÿÿ
ÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÀÿÿt,]5(° Ö;¬]ô4 p(ACàþþþþþþþÿþÿÿÿÿÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ<ÿÿ5nC2°ånô*f9 xCXXþþþþþþþÿÿÿÿÿþÿÿþÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿN8(HP5GPPPPP,2xxP0zzPx0ûÿ2ì^ø´ú;$,ÿ%ÿÿÿfµÿp Q#ppÿÓ fÿ& D
úÑ§z ¯!
j.IÙ.á$d þ$×"D
Éÿ» +¨":¡.:ÿ¼"¼¨Äz ÿG þ$!:¯!
ÿò k¼~ 7¾.ÿL ÔMc\!t.=
ß*P ÿÀ £(2,2) ÿô<1,1> ÿ%:!(4,5) ÿ+/ì&<4,5> ÿ <1,5> ÿ&C (4,2) ÿxStreamwise direction ÿ1<cross ÿ 1stream ÿ
1ïwise ÿ 1È directionÿ
×#Å_,OÿÅ-;d,©Z,õÿ
Ð3nè3Yÿê3À3ZA4W ÿ 0Âinflowÿ
6ÿpÝÿ
®".8ÿÃÜ]ÿ
81á$¯0é$ÿG/þ$¨0$³0B%ÿ
9S¬@ÿ
»æ ÿ½<3,3>ÿ
r/ #ÿ¶ gh×Ýÿ
Lÿ
Äxÿ
Ó±±ÿò ± X ÿ6BX ÿú< Y ÿåØ(xnumber = 4 ÿØÍ&ynumber = 5 ÿ?B*<,> : grid intersection ÿS((,) : calculation node ÿº'N. ÿ&ò". ÿð
@. ÿ². ÿâ. ÿ;. ÿ!:!. ÿP. ÿø. ÿ³a. ÿg'. ÿÝ&l.§%ÿ<4¨¨¨¨¨¨¨¨T¨¨¨TTTTTüTüTTüüüTTüTüüüTüüüÿ¥ %1E .â%1¥ %1ÿ
Í °,
°,ÿ
W+
+ÿ
Ë.Ëuÿ
^.^ÿ
º©Y©ÿ
Gàæàÿ
ï'Àï'ÿ
äu¼uÿ
%uÿ

u
Xÿ
6£ X
X
eÝP
ÿ Ã
W 0.15 hÿ
©u©ñÿ©Ð1ñ!ñ ÿöÉ0.4 hÿ
JuJ(ÿJÓ(Â( ÿù±0.6 hÿ
tut_&ÿt=(ü_&ì_& ÿü6*0.8 h ÿ$K7G 3 0 15 40 60 80 100ÿ
s±#sÿ
É9.v"9. ÿ â";Vertical grid distributionO>¼!ÿ%ÿÿÿfµÿp Q#ppÿÓ f00)
Î
P2ðÿ
y<°)
°ÿ
y Dÿ
y<:)
:ÿ
y<>)
>ÿ
¶Î
¶¼ ÿ
_Î
_¼ ÿ
7Î
7¼ ÿ
>$Î
>$¼ ÿ
t+Î
t+¼ ÿ
Ù2Î
Ù2¼ ÿ
68Î
68¼ ÿ
)
ÔÔÿ"Ôo:ÿ
X
X
ÿXq¿
ó

~ÿP ÿsj ÿm$üi ÿ¤+sa2,b2^;} ÿ6»a1,b1)
Î
ð6Î
CnÚ2; h8Í*ÿ%ÿÿÿfµÿp Q#ppÿÓ f&3*ÿ
¾6 ÿ
¾6ññ[êú!ÿ
2 |&2 ÿ
|&ÿ
[¥!D%¥!ÿ
[D%ÿ
[LD%Lÿ
ÿ¬ÿÿ
¬ÿ
;*¬;*ÿ
0¬0
ß*P ÿ Tó2,2 ÿ 9¼2,4 ÿ h×5,4 ÿ ! 7,4 ÿ
72¼10,4 ÿ
72½10,2 ÿ
ó w11,2 ÿ
¦11,4 ÿ
¦¨$14,4 ÿ
!¨$14,2ÿ
íäÿLäíä:ÿ
44
ÿ4qÚ

ÿ´j ÿ³iÿ
' 'Ý$ÿ'D&À&Ý$t'Ý$ ÿI(E&iÿ
_%("(ÿº(".("á( ÿ¦õ(j ÿ wÔ"xnumber = 14 ÿw® ynumber = 4 ÿ wêMBSkjot = 10 ÿwÄMBStart = 5ÿ
}
¦ÿÿ¬ÈãLhÿ
ÈFÓÿ¬H¶ÿ
äD8ÿ[Jÿþrë<ß+ÿ%ÿÿÿfµÿp Q#ppÿÓ f ÿ
Ê,% ;,%ÿ
Ê ;ÿ
Êç ;çÿ
Ê ;ÿ
J*ÿ
/J/*ÿ
-J-*ÿ
8J8*
ÿUP ÿ Z 4,5 ÿ W" 5,5ÿ
uY ¯$Y ÿI&Y ¯$À ¯$ó ÿ'÷ iÿ
á¶ÿ/Þ Êü]nÿ
k$ÿ/ê¨Àÿ
@
Q_ÿeÍ¸1ÿ
@
¬¤ÿu
ø^aé ÿ3@posneg = -1 ÿ
ß
posneg = 1 ÿ W"¬5,4 ÿÃ înode = 4 ÿ Z4,2 ÿ

dir = 1+8Æ+ÿ%ÿÿÿfµÿp Q#ppÿÓ f¦}0Á
ß
+ßñX¦õlõ vnõ]Rô9~vj 9Úd*'Ý dôÄÞAÊ²ÝSÄÞ/ªÄÝ/
ºøP ÿ ¾DUpdate ÿ
´MB-Flow ÿCÿSediment ÿ
<UWrite X ÿF
eRead Res ÿ HñMap ÿ
³ ìLong ÿ/ÕCross ÿ ¾3-D
ºøP ÿ UWrite Res
ÀP ÿ
#æTextÖC¥
ú ^¦
z¦
ß*P ÿ
øucontrol ÿ
Zkoordina ÿ ?resultç¦
þ!G
Ï ÿ àô$SPIDER ÿ-Ó(XCYC.DAT ÿ
»
SSIIMç%Båo
ÿ S(+%Mapç%ååo
ç%åo
ÿ ú'éLongitudinal ÿ
'Ã Cross-section ÿ
)æ)Graphicsÿ2 b(Å ?)1L+i-'0»W1P24!Ó/#â-"`,¾!Z)l!b(Å ÿ®)l!S*s ù*E0+iÿ`,£!!-s -Eâ-iÿ6.+"/â Ó/! z0Øÿ¶(éo*s ê,Å //O!±0µ"ÿí(óù*|u-´ð/! Æ1Ù!ÿ*
(»á1»á1'0 «-yÉ)y(i'¨i'»(»ÿÉ)yÉ)»o*Á*:0+L+ÿ,y,»-Ä>-Ç-°â-yÿ0 0»z0±0ð±0 ÿ*
w)Íá1Íá1ñ0¦n.o-$*\~(Þ~(Íw)Íÿ
-Ch+Cÿ*Ch+êh+ÿ
w)º'*óÿ+åk*-ä)·ÿ
E,T¸-ÿ÷.º-eâ-Äÿ
^B^¥
ÿ^= ·¥
¥
ÿ
t^tÀ
ÿtX ÎÀ
À
ÿ
3y3¥
ÿ3= ¥
Ú¥
ÿ
Ã= ÃÿÃyiÿ
D!
ÿBHÏ¼ÿ
ç®#«®#ÿD®#«T#«$ÿÕ ÞÆ"ÞÆ"$Õ $ÿn$Õ Á#Õ u$ÿ
Õ Ç$Tÿç%v$¬$úÿÕ ·H$\H$N$Nÿç%N$¨$ôÿÕ 66« 26¦ ÿÊ4 66L ,6!Ð °°°° ÐÐ÷Ð6Â&Ð6Â&StandardG-1)HPLAII21.PRSêfô\ Ð6Â&Ð6Â&Standardl Document Style*£°°°°÷Ð
Ñ#oô\ PÂCb×XP#Ñ
Áàñ
aÁUNIVERSITY OF TRONDHEIM
Áà
aÁTHE NORWEGIAN INSTITUTE OF TECHNOLOGY
ÁàÑaÁDIVISION OF HYDRAULIC ENGINEERING

Ô # ÔÃÃÃ ÃÁà´aÁA THREE©DIMENSIONAL NUMERICAL
ÁàaÁ MODEL FOR SIMULATION OF
ÁàZaÁ SEDIMENT MOVEMENTS IN
Áà! aÁ WATER INTAKES
ÁàaÁWITH MULTIBLOCK OPTION

Áàa ÁVersion 1.1

Ô # ÔÁà+aÁUser's guideÄ ÄÄÄ

Ô Úsp ÔÁàa¤ÁÃÃÃ ÃBY
Ô Ús½ ÔÄÄÁà aÁÃÃÄ ÄÃ ÃNILS REIDAR B. OLSENÄ ÄÄÄ

ÁàaÁ16. OCTOBER 1993
Ô$*p-p-p-°°Ô Ô Ús ÔÃ ÃÃÃForewordÄÄÄ Ä

The SSII model was developed in 1990©91 during the work with my dr. ing. degree at the
Division of Hydraulic Engineering at the Norwegian Institute of Technology. SSII is an
abbreviation for Sediment Simulation In Intakes. The model was originally build around
the numerical model Spider, which was made by dr. M. Melaaen during the work on his
dissertation in 1989©90. Spider solves a general flow problem for a three©dimensional
geometry. SSII was made up of sediment calculation routines, communications with Spider
and a graphical user interface made in OS/2.

The main motivation for making SSII was the difficulty to simulate fine sediments in
physical models. The fine sediments, often under 0.2 mm, are be important for wear on
turbines. It was also an advantage to be able to simulate other problems as for example
sediment filling of reservoirs and waterways.

At the time SSII was made I had limited funding for computer equipment. This, together
with lack of knowledge of UNIX made it necessary to develop the models on a PC. Then
a problem arouse, the 640 kB limit of DOS. The arrays that the model uses is often one
order of magnitude larger than the DOS limit. Therefore, the operating system OS/2 was
used. Compared to UNIX, OS/2 is much more user©friendly, and this has been a major
advantage during the development process.

In 1992/1993 a multiblock module for calculating water flow was made. This was added
to SSII, and the resulting model was called SSIIM. Version 1.0 was uploaded on the net
17th of June 1993.

Several people have provided me with insight into the various problems I have
encountered in the development of this program. I have benefitted greatly from the
knowledge of Prof. Melaaen at the Norwegian Institute of Technology in the subject of
computational fluid dynamics. In the topics of hydraulics and sedimentation engineering I
have learned from Prof. Lysne at Division of Hydraulic Engineering at the Norwegian
Institute of Technology, and from Prof. Julien, Prof, Gessler, Prof. Wohl and Prof.
Bienkjewicz at Colorado State University. Mr. Alfredsen at the Norwegian Hydrotechnical
Laboratory has helped me during problems with software and hardware. The following
people have helped me in testing the code: M. Skoglund, O. Jimenez, A. LÀQÀvoll, L.
Abrahamsen and S. Stokseth.

I want to thank you all for your contributions.

Trondheim, 16th of September 1993

Nils Reidar BÀQÀe Olsen

ÔB*p-p-p-°°Ô Ô Ús ÔÃ ÃÃÃTable of contentÄÄÄ Ä

Disclaimer and legal mattersÁp®!"!Á1

Limitations of programÁp®!"!Á1

Advise for new users of SSIIMÁp®!"!Á1

Structure of SSIIMÁpk!" Á 2

Theoretical basis for water flow calculationÁpk!" Á 3

Theoretical basis for sediment flow calculationÁpk!" Á 5

File structureÁpk!" Á 6

GraphicsÁpJ!" Á24

Experience with convergence for water flow calculationÁpJ!" Á25

Advise for interpretation of resultsÁpJ!" Á26

LiteratureÁpJ!" Á27

Appendix A, Control input fileÁpJ!" Á29

Appendix B, Flow chart for filesÁp!"Á 30

Ômp-p-p-°°Ô Ô Ús ÔÐ
Èé
ÐÓ

ÓÃ ÃÃÃDisclaimer and legal mattersÄÄÄ Ä

I disclaim all warranties with regard to this software and the information in this document,
whether expressed or implied, including without limitation, warranties of fitness and
merchantability. In no event shall I or my employer, SINTEF NHL, be liable for any
special, indirect or consequential damages or any damages whatsoever resulting from loss
of use, data or profits, whether in an action of contract, negligence or other tortuous
action, arising out of or in connection with the use or performance of this software. It is
therefore not recommended that the program be used for solving a problem whose
incorrect solution could lead to injury to a person or loss of property. If you do use the
program in such a manner, it is at your own risk. It is necessary to know that to
understand and interpret the program results properly it is required that the user have
knowledge and experience in computational fluid dynamics and hydraulic engineering.

Provided the user complies with the above statements, the program can be used freely.
The program can be distributed freely on condition and that an unchanged copy of this
manual is distributed with the program.

Nils Reidar B. Olsen

Ô Ús ÔÐ
éé
ÐÓ

ÓÃ ÃÃÃLimitations of programÄÄÄ Ä

Some of the most important limitations of the program are listed below. Note that some
combinations of different options may not have been tested, and then there is a risk of a
bug in the program.

* The program neglects non©orthogonal diffusive terms.
* The grid lines in the vertical direction have to be completely vertical.
* Internal walls cannot be used within two cells from a multi©block connection.

Ô Úsè ÔÃ ÃÃÃAdvise for new users of SSIIMÄÄÄ Ä

Generally, it is advisable to start with reading this manual. It is important to understand
that the model is made up of several sub©models. The main dialog box for the program
has different push©buttons to start the different models. Some models can be run
simultaneously.

The hardware requirements for running the program are mainly focused on having enough
Ô X% ÔRAM. In the beginning of the Ã ÃboogieÄ Ä file it is printed how much RAM the program
allocates for the arrays. This can be added to the RAM requirement for the program itself,
about 600 k, plus what the operating system requires. An estimate for the amount of RAM
is thereby obtained. OS/2 will use the harddisk as extra memory if there are not sufficient
RAM. The penalty is that the program runs ÃÃveryÄÄ much slower. This situation can be
detected by observing if the system swaps to the harddisk while running only the SSIIMÔ!*p-¿+¿+°°Ôprogram. The water flow module may take from some hours to some days to converge for
a typical case, when there is enough RAM. However, under OS/2 it is not any problem to
let the program run in the background while doing other tasks on the computer.

An advice for the first©time user is to run one of the example cases first. Then, try to
modify some of the parameters and run it again. Often the user wants to simulate a
particular case. It is then advisable to try to find a similar example case and modify this
step by step.

Some words about crashing the program and input control. There are various controls for
input and checking of intermediate results. If any of these controls finds something wrong,
Ô X
Ôan error message is written to the Ã ÃboogieÄ Ä file, and the program terminates. Therefore, if
Ô Xì
Ôthe program suddenly stops, and the main dialog box disappears, check the Ã ÃboogieÄ Ä file for
possible error messages.

If a fairly regular geometry is present, like a canal, it is advisable to use a spreadsheet to
Ô X Ôgenerate the Ã ÃkoordinaÄ Ä file. If a natural river or a reservoir is simulate, a map is needed
and the grid is drawn on the map. Then digitizing equipment is used togehter with a
Ô Xb Ôspreadsheet to make the Ã ÃkoordinaÄ Ä file.

Some words about bugs. The models 3©D Bed and Anim were developed several years ago
and have not been maintained since then. There are many bugs in these modules, and my
advice is not to use these modules. Also, the more advanced options have not been tested
so well as the more simples options. Combining different advanced options can lead to
errors. For example, using the SOU scheme with the multi©block option. There is some
more information about this in the following pages of this manual. Note that most
programs have bugs, and SSIIM is probably no exception to this rule.

Finally, the results from the program ought to interpreted according to:

© Possibilities of bugs in the program making errors
© Previous cases where the results have been compared with measurements
© Numerical errors, like false diffusion, grid independency, etc.
© Accuracy of boundary conditions

Knowledge and experience in computational fluid dynamics and hydraulic engineering are
essential for the assessment of the validity and accuracy of the results. There is a seperate
chapter in this manual for assistance in interpreting the results.

Ô Úsh$ ÔÃ ÃÃÃStructure of SSIIMÄÄÄ Ä

SSIIM is a numerical model for calculating sediment movements in water intakes. A flow
chart of the model is given in Appendix B. The model contains a pre and a post processor
that communicate with the model Spider. Spider is a model that calculates flow in a
general three©dimensional geometry. In SSIIM there is also a module for calculating
sediment movements, and a module for calculating bed changes. The sediment flowÔB*p-¿+¿+°°Ômodule is based on the convection©diffusion equation for the sediment concentration. The
module that calculates bed changes is based on continuity for the sediments in the cells
close to the bed. This is further descried in [4].

Ô X¤ ÔWhen SSIIM is started, the two files Ã Ãcontrol Ä Äand Ã ÃkoordinaÄ Ä are read and the grid is
initialized based on the data in the files. The water surface elevations are calculated using
a standard one©dimensional backwater routine.

After initialization, it is possible to write or read result files and also to see graphics and
to start the MB©flow routine or the sediment transport routine. This can be done by
pushing the buttons on the user interface. It is also possible to read/write files and start
Ô X
Ôprograms automatically after the initialization, by specifying this in the Ã ÃcontrolÄ Ä file, on the
Ô Xì
ÔÃ ÃF 2Ä Ä data set.

The communication between SSIIM and Spider is by exchanging information in ASCII
files. SSIIM writes the boundary conditions, grid and some other parameter to a file called
"XCYC.DAT". Spider reads this file, and calculates the result and writes the result to a
file named "RESULT". This is read by SSIIM for further calculation of sediment
movements and presentation of results. Because it is possible to transfer files between
computers, it is possible to run SSII on the PC and Spider on another computer, for
example the CRAY. However, the files must be transferred manually, for example with
the ftp©program.

If the bedlevels change significantly, it is possible to make SSIIM write the geometry to
Spider, and take the results to calculate a new geometry. The new data can be sent to
Spider for a new calculation and so on. This can be iterated several times to compute bed
changes over a time period. Instead of using Spider, this can be done automatically by the
use of the MB©flow routine.

Ô ÚsN ÔÃ ÃÃÃTheoretical basis for water flow calculationÄÄÄ Ä

The Navier©Stokes equations for turbulent flow in a general three©dimensional geometry
are solved to obtain the water velocity. The k©ÀÀ model is used for calculating the turbulent
shear stress.

The Navier©Stokes equations for non©transient non©compressible and constant density flow
can be modelled as:
ÚX!Ð°ª'ddddddY 4ddZ-ÐÿÿßÛ~U_j {PARTIAL U_i} OVER {PARTIAL x_j} ~ = ~ 1 OVER rho PARTIAL
OVER {PARTIAL x_j} ( © P delta_ij ~ © ~ rho overline {u_i u_j}
)êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··+ûUzz³Áj··OUzz×Ri··hlxzz½2j··=lxzz2j··rûPzzSÁij··ûuzzyÁi··ûuzzûÁjø·ä&ù·)&·Ò&Ò···÷,··l,··\ûä·· `,··ål,··ûã··çûã··H`1··Çû(·· û)·· Ô"!°+ÔÁ`(!"Á(1)
Ô"!°+Ô
Ô$!°+Ô
Ô++""&°°""ÔThe term on the left side of the equation is the convective term. The first term on the
right-hand side is the pressure term. The second term on the left side of the equation is the
Reynolds stress term. To evaluate this term, a turbulence model is required.

Ô XB* ÔÃ ÃThe k©ÀÀ turbulence modelÄ ÄÔ0B*p-¿+¿+°°°ª'+Ý*!0ÔThe eddy©viscosity concept with the k©ÀÀ model is used to model the Reynolds stress term:
ÚÐA0m
°dddddd¥ 9ddZ-béÿÿÿÿßS© overline{u_i u_j} ~ = ~ nu_T ~ left( {partial U_i} over
{partial x_j} + { partial U_j } over { partial x_i } right ) ~
© ~ 2 over 3 k delta_ijêfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··ûã··Þûä··÷,··l,·· ûâ··,··µl,··BûãÒ···6h··îj··i··±6o··±îq··±pø·ä&ø·&Î·&··ûuzzêÁi·· ûuzzlÁjzzÁT··OUzz×Ri··hlxzz½2j··ôUzz|Rj··
lxzzb2i··ûkzzS Áij··û··âû ··"`2··"l3SßÐÚÔ$°°""é""!A°$Ô
Ô"A°Ô
Ô"A°ÔÁ`(!"Á(2)
Ô$A°Ô
Ô""°°""ÔThe first two terms on the right side of the equation forms the diffusive term in the
Navier©Stokes equation.

The eddy©viscosity is modeled in the k©ÀÀ model as:
Úa0Â°á ddddddÃúddZ-bÈÿÿß$nu_T = c_mu { k^2 OVER varepsilon } êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··Ïzz··@=zzT··2Ïc··ö4k··ÄÏäQ·ãúzz^|2ßÚÔ$°°""1""!a°$Ô
Ô"a°ÔÁ`(!"Á(3)
Ô"a°Ô
Ô$a°Ô
Ô""Õ°°""Ôk is turbulent kinetic energy, defined by:
Ú¢0F¤°ndddddd~ÜddZ-bÈÿÿß%$k == {1 OVER 2} OVERLINE {u_i u_i} êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··Ïk··eÏuzzÅi··çÏuzzGi··fÏÎ·ãúÒ·el··ö41··ö@2%ß¢ÚÔ$°°""¾ YY""!°Y$Ô
Ô"°YÔÁ`(!"Á(4)
Ô"°YÔ
Ô$°YÔ
ÔYY""b°°""Ôk is modeled as:
Ú¡ ü°ûdddddd4ddZ-ÈßU_j ( {PARTIAL k} OVER {PARTIAL x_j} ) = PARTIAL OVER {PARTIAL
x_j} ( nu_T OVER sigma_k {PARTIAL k} OVER {PARTIAL x_j} ) +
P_k © varepsilon êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··3ûUzz»Áj··§`k··lxzzé2j··Clxzz2jzzRTzz¢2k··{`k··hlxzz½2j··ûPzzÁk··Ýû(··û)··Íû(··òû)·)&·Ø&c·&·ý&··O`,·· Ô"¡°#ÔÁ`(!"Á(5)
Ô"¡°#Ô
Ô$¡°#Ô
Ô Xï ÔÔ##""ï°°""ÔPÃÃkÄÄ is given by:
Ú*Á0Fü°dddddd~4ddZ-bÈßsP_k = nu_T {PARTIAL U_i} OVER {PARTIAL x_j} ( {PARTIAL U_j}
OVER {PARTIAL x_i} + {PARTIAL U_i} OVER {PARTIAL x_j} )êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··+ûPzzÁkzz®ÁT··lUzzôRi··lxzzÚ2j··àUzzhRj··ùlxzzN2i··Uzz
Ri··lxzzó2j··Õûä··,··-l,··,··¡l,··ûâ··-,··Fl,··Cûø·&ø·u&ø·&··)û(··Bû)ß*ÚÔ$°°""ØY Y ""!Á°Y $Ô
Ô"Á°Y ÔÁ`(!"Á(6)
Ô"Á°Y Ô
Ô$Á°Y Ô
ÔY Y ""|°°""ÔÀÀ is modelled as:
Úá hü°dddddd
4ddZ-hÈßßU_j {PARTIAL varepsilon} OVER {PARTIAL x_j} = PARTIAL OVER
{PARTIAL x_j} ( nu_t OVER sigma_varepsilon {PARTIAL
varepsilon} OVER {PARTIAL x_j} ) + C_{varepsilon 1} varepsilon
OVER k P_k © C_{varepsilon 2} varepsilon^2 OVER kêfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··BûUzzÊÁj··flxzz»2j··Ølxzz-2jzzARt··lxzzZ2j··:ûC··_lk··ÍûPzzAÁk··åûC··)
lk·û&·m&q·®&·&Ü·C&J·î &··`,··l,··ðûä··»`,··l,··¶`,··l,··Ìûâ··wûã··o`=··Ö··Ál%zz52=··`=zz´Á=··V`=zz_ Á=··
`=··bû(··û)zz÷Á1zz¢ Á2zze
¨2ßÚÔ$°°""e{{""!á°{$Ô
Ô"á°{Ô
Ô"á°{ÔÁ`(!"Á(7)
Ô$á°{Ô
Ô{{"" °°""ÔIn the above equations, the c are different constants.

The equations are discretized with a control©volume approach. An implicit solves is used,
also for the multi©block option. The SIMPLE method is used for pressure©correction. The
power©law scheme or the second©order upwind scheme is used in the discretization of the
convective terms. The numerical methods are further described in [3] and [4].

Ô Úsõ( ÔÃ ÃÃÃTheoretical basis for sediment flow calculationÄÄÄ Ä
ÔB*p-¿+¿+°°a°ýA°á a°nYu°û#Z¡°Y çÁ°{t"áÔSediment transport is traditionally divided in bedload and suspended load. The suspended
load can be calculated with the convection©diffusion equation for the sediment
concentration, c:
Ú0h
Ð°kdddddd ddZ-bÈß0¢U_j {PARTIAL c} OVER {PARTIAL x_j} ~ + ~ w ( {partial c} OVER
{ PARTIAL x_z} ) ~ = ~ÃÃÄÄ © ~ PARTIAL OVER {PARTIAL x_j} ( GAMMA
{PARTIAL c} OVER {PARTIAL x_j} )êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··3ûUzz»Áj··j`c··Wlxzz¬2j··ïûw··9`c·· lxzzu2z··:lxzz2j··õ`c··âlxzz7 2j·ì&³·µ&·Ï&·w&··`,··ÿl,··1ûâ··á`,··Èl,··Dûä··ûã··`,··âl,··`,··l,··iû(···û)··Äû(··l û)··û0ßÚÔ$°°""»{{""!°{$Ô
Ô"°{ÔÁ`(!"Á(8)
Ô"°{Ô
Ô$°{Ô
Ô{{""_°°""ÔThe fall velocity of the sediment particles is denoted w. The diffusion coefficient, ÀÀ, is
taken from the k©ÀÀ model:
ÚT!0«Ð°á ddddddãddZ-béÿÿß×GAMMA = nu_T OVER Scêfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··Ï··`··Ïäc·úzz~&T··@Sc×ßTÚÔ$°°""1¾¾""!!°¾$Ô
Ô"!°¾ÔÁ`(!"Á(9)
Ô"!°¾Ô
Ô$!°¾Ô
Ô¾¾""Õ°°""ÔSc is the Schmidt number, and it is around unity.

The first term in Equation 8 is the convection of sediments, which is the sediments that
are transported through the walls of the finite volume because of the velocity of the water
at the wall. The right-hand term is the diffusion of sediments. For the case described here,
ÀÀ is the diffusion coefficient due to the mixing by turbulence in the water. The term
therefore tells how much sediments are transported trough the wall of the finite volume
because of turbulence and the difference in concentration between the two sides of the
wall. The second term on the left side of Equation 8 is due to the fall velocity of the
sediments. This is treated as an extra convective term, and added to the velocities in the
vertical direction.

For a three©dimensional flow situation, van Rijn [5] developed a formula for the
equilibrium sediment concentration close to the bed. This gives the possibility of
simulating the interaction between the sediment that moves close to the bed and in
suspension. Van Rijn's formula for bed concentration is given as
ÚàA0´
d°ddddddì ddZ-bÈßc©c_bed ~ = ~ 0.015 ~ d_50^{2/3} OVER a ~ {left [ {tau_0 ©
tau_{critical}} over {tau_{critical}} right ] ^{1.5}} OVER {
left [ {(rho_s © rho_w) g} over nu^2 right ]^{0.1}}êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··$/czzyõbed··GÀd·· azz¦criticalzzcriticalzzéDszzúDw··~g··y/ä··Æàã··~ã··7/0··/.··Á/015zzº!2zz÷!/zz!3zz£50zz¦0zz: 1zzw .zz 5··8~(··K~)zz¥2zz& Ê0zzc Ê.zz Ê1·4ZH·ÈZ··Ûv··ÛBx··Ûiw·· }·· B·· i~:·z··î6v··îîx··îw··÷6}··÷î··÷~ô·%··%à)··4à)··¬À)··u~#··~#··:@cßàÚÔ$°°""eÇÇ""!A°Ç$Ô
Ô"A°ÇÔ
Ô"A°ÇÔ
Ô"A°ÇÔ Á`Ä "Á(10)
Ô"A°ÇÔ
Ô"A°ÇÔ
Ô$A°ÇÔ
Ô XÄ ÔÔÇÇ""Ä °°""ÔThe sediment particle diameter is denoted dÃÃ50ÄÄ. a is a reference level, set to
Ô X! Ô1.5 % of the water depth. À)ÀÃÃ0ÄÄ is the bed shear stress, À)ÀÃÃcriticalÄÄ is the critical bed shear stress
Ô X" Ôfor movement of sediment particles, À#ÀÃÃwÄÄ and À#ÀÃÃsÄÄ are the density of water and sediment, ÀÀ is
the viscosity of the water and g is the acceleration of gravity.

The concentration given by Eq. 10 is "forced" on the bed boundary finite volumes. This
means that the concentrations for these points are not changed during the solution of the
Ô X#' Ôconvection©diffusion equation for the sedimentsÃÃÄÄ. Sediment continuity for the bed boundary
finite volumes is therefore usually not satisfied. The discrepancy in continuity can be used
to calculate changes in the bed levels.

ÔPÇ*p-¿+¿+°°1°k{
°á ¾!°ÇÜ$APÔÔ Ús ÔÃ ÃÃÃFile structureÄÄÄ Ä
ÐÐX°` ¸hÀpÈ xÐ (#%Ø'0*,à.813è5@8:ð ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ°ÿÿÐÐ
A flowchart describing the various files are given in Appendix B.

Ô X ÔThe two main input files are the Ã ÃcontrolÄ Ä file and the Ã ÃkoordinaÄ Ä file. These must be made
Ô Xñ Ôbefore the program is started. All the files except the Ã ÃcontÄ Ä file are ASCII files.

Ô X¬ ÔÃÃThe Ã ÃkoordinaÄ Ä fileÄÄÚË!txÕÈÈ¡±;$,ddGRID.CGM é$t)Ù·ËRÁàMMzÁ
Ô SØ ÔÃ ÃÁàKMgÁÄ ÄFig. 1. Example of grid seen from aboveËÚ
Ô$°°""°°""Á!°"$Ô
This is the input file where the bed of the geometry is described. An example is show in
Fig. 1. The grid can be made using a map, and a spreadsheet can also be used.

The necessary input data is the x,y and z coordinates of the points where the grid lines
Ô Xã# Ômeet. The format of the data is given below.Ã Ã

Ô Xµ% Ôi j x y z Ä Ä

An example:

1 1 0.34 0.54 0.11Ô0B* p-¿+¿+°°°
"é#±! 0Ô1 2 0.35 0.66 0.12
...

The first two numbers are integers, while the following three are floats. The numbers are
read in a free format, which means that the distance between them does not matter. The
sequence of the points are not important, as long as all points are included. This is not
controlled by the model, so the used must do this by looking at the grid in the graphic
modules of the program.

The last line in the file must be a sequence where the first integer is above 500. This
marks the end of the file. The program will stop reading the file when this line is
encountered.

Some words about indexing and numbering of grid lines and cells. The variable names for
the number of grid lines in the three directions are:

xnumber : streamwise direction
ynumber : cross©streamwise direction
znumber : vertical direction

The numbering of the grid lines goes from 1 to xnumber in the streamwise direction, and
similarly for the other two directions.

However, the grid lines define cells between the grid lines. The variables are calculated in
the center of each cell. This means that a numbering for cells also is required. The word
node is often used for the center of a cell.

From a geometrical view of the grid, it is observed that the number of lines always
exceeds the number of cells by one in each direction. When the arrays are defined, it is
therefore a choice for the programmer to start the numbering of the cells on one or two.
The choice that is made in this program is that the numbering starts on two. This ÃÃÄÄmeans
that the cell that is defined by grid lines i=1 and i=2 and j=1 and j=2 has the number
(2,2). Cell number (1,1) does not exist. The numbering of the cells is also shown in Fig. 1.
The numbering of the grid lines is shown with the <,> sign, while the numbering of the
calculation nodes is shown with the (,) sign. The grid is non©staggered.

Ô X! ÔÃÃThe Ã ÃcontrolÄ Ä fileÄÄ

This file contains all other data that are necessary for the program. An example is given in
Appendix A. SSII reads each character of the file one by one, and stops if a capital
T,F,G,I,S,N,B or W is encountered. Then a data set is read, depending on the letter. A
data set is here defined as one or more numbers or letters that the program uses. This can
for example be the water discharge, or the Manning's friction coefficient. It is possible to
use lower©case letters between the data sets, and it is possible to have more than one data
set on each line. Not all data sets are required, but some are. Default values are given
Ô XÞ) Ôwhen a non©required data set is missing. SSII controls the data sets in the Ã ÃcontrolÄ Ä file to a
Ô XÇ* Ôcertain degree, and if an error is found, a message is written to the Ã ÃboogieÄ Ä file and theÔÇ*
p-¿+¿+°°Ôprogram is terminated.

For the initialization and graphics, the following data sets are used:

Ô X¤ ÔRequired: Ã ÃG 1, G 3, W 1, W 2Ä Ä
Ô X ÔOptional: Ã ÃT, F 1, F 2, F 7, G 8, P 2©4Ä Ä

For the sediment calculation, the following data sets are used:

Ô X1 ÔRequired: Ã ÃS, N, B, IÄ Ä
Ô X ÔOptional: Ã ÃF 4, F 6, F 8©12, F 23Ä Ä

For the water flow calculation by using MB©Flow, the following data sets are used:

Required: None
Ô X§
ÔOptional: Ã ÃF 15©17, F 22, F 24©25, G 6©8, G 11, W 3©4, K 1©10 Ä Ä

Ô Xy ÔIn the following the data sets for the Ã ÃcontrolÄ Ä file is described:

Ô XK ÔÃ ÃTÄ ÄÁQ Q OÁTitle field. The following 30 characters are used in the graphics programs.

Ô X ÔÃ ÃF 1Ä ÄÁQ Q OÁDebugging possibility. If the character that follows is a D, one will get a
Ô X ÔÁQ Q OÁmore extensive print©out to the Ã ÃboogieÄ Ä file. If the character is a C, the
Ô Xï ÔÁQ Q OÁcoefficients in the discretized equations will be printed to the Ã ÃboogieÄ Ä file.

Ô XÁ ÔÃ ÃF 2Ä ÄÁQ Q OÁAutomatic execution possibility. Some parts of the program will be executed
ÁQ Q OÁdirectly after the initialization if a character is placed in this field.
ÁQ Q OÁThe sub©programs will be executed in the order they are given. The
ÁQ Q OÁpossibilities are:

Ô XN ÔÁQ Q OÁR Áò
ò
wÁRead the Ã ÃresultÄ Ä file
Ô X7 ÔÁQ Q OÁIÁò
ò
wÁInitialize the sediment calculations
Ô X ÔÁQ Q OÁSÁò
ò
wÁCalculate sediment concentration
Ô X ÔÁQ Q OÁWÁò
ò
wÁStart the multi©block water flow routine
Ô Xò ÔÁQ Q OÁBÁò
ò
wÁChange the bed according to sediment calculation
Ô XÛ ÔÁQ Q OÁM Áò
ò
wÁWrite result file
Ô XÄ ÔÁQ Q OÁVÁò
ò
wÁInitialize water surface level

Ô X# ÔÃ ÃF 4Ä ÄÁQ Q OÁRelaxation factor, maximum iterations for bed changes and depth of active
ÁQ Q OÁbed sediment layer. These three parameters are used when calculating the
ÁQ Q OÁchange in bed grain size distribution. Maximum iterations give the
ÁQ Q OÁmaximum times the suspended load is recalculated during one run.

ÁQ Q OÁDefaults: relaxation: 0.8, iterations: 15

Ô XÞ) ÔÃ ÃÄ ÄÃ ÃF 6Ä ÄÁQ Q OÁCoefficients for formula for bed concentration. Default is van Rijn's
ÁQ Q OÁcoefficients: 0.015, 1.5 and 0.3. If one uses this option, the sedimentÔÇ*p-¿+¿+°°ÔÁQ Q OÁtransport formula given in dataset F 10 must be R, which is van Rijn's
ÁQ Q OÁformula is used as basis.

Ô X» ÔÃ ÃF 7Ä ÄÁQ Q OÁRun options. read 10 characters. If the following capital letters are
ÁQ Q OÁincluded this will mean:

Ô Xv ÔÁQ Q OÁD: Áò
ò
wÁDouble the number of grid cells in streamwise direction
Ô X_ ÔÁQ Q OÁÁò
ò
wÁin comparison to what is given in the Ã ÃkoordinaÄ Ä file. Each
ÁQ Q OÁÁò
ò
wÁcell is divided in two equal parts.
Ô X1 ÔÁQ Q OÁJ: Áò
ò
wÁDouble the number of grid cells in the cross©streamwise
ÁQ Q OÁÁò
ò
wÁdirection.
Ô X
ÔÁQ Q OÁI: Áò
ò
wÁInflowing velocities in the y©direction are set to zero.
Ô Xì
ÔÁQ Q OÁA:Áò
ò
wÁDiffusion for sediment calculations in non©vertical direction is
ÁQ Q OÁÁò
ò
wÁset to zero.
Ô X¾ ÔÁQ Q OÁN:Áò
ò
wÁConcentration gradient at the bed is not set to zero. Default is
ÁQ Q OÁÁò
ò
wÁwhen the gradient is set to zero. Then the concentration at the
ÁQ Q OÁÁò
ò
wÁbed cell will not be lower than the cell above.
Ô Xy ÔÁQ Q OÁB:Áò
ò
wÁCorrection for sloping bed is used when calculating bed
ÁQ Q OÁÁò
ò
wÁsediment concentration.
Ô XK ÔÁQ Q OÁG:Áò
ò
wÁCell walls at outblocked area is not changed when there
ÁQ Q OÁÁò
ò
wÁare changes in the cells outside the block.
Ô X ÔÁQ Q OÁV:Áò
ò
wÁ90 degree turning of the plot seen from above (map).
Ô X ÔÁQ Q OÁZ:Áò
ò
wÁVertical distribution of inflowing sediment is uniform.
Ô Xï ÔÁQ Q OÁX:Áò
ò
wÁGrid is read in from the "XCYC" file. This is only used in
ÁQ Q OÁÁò
ò
wÁthe post©processor, and with presentation of results
ÁQ Q OÁÁò
ò
wÁfrom Spider where the lines in the k©direction are not vertical.
Ô Xª ÔÁQ Q OÁC:Áò
ò
wÁInflowing and ouflowing water in default walls is set to zero.
ÁQ Q OÁÁò
ò
wÁThis means that the water flow must be specified on the G 7
ÁQ Q OÁÁò
ò
wÁdata sets.

Ô XN ÔÃ ÃF 8Ä ÄÁQ Q OÁMaximum bed level change relative to water depth. This is controlled for all
ÁQ Q OÁthe cells. Default: 0.1. This parameter is used to compute the time step
ÁQ Q OÁfor the bed changes.
Ã Ã
Ô Xò ÔF 9Ä ÄÁQ Q OÁFactor that is used to change the turbulent viscosity of the inflowing water.
ÁQ Q OÁThe factor is proportional to the turbulent viscosity. Default: 1.0.

Ô X! ÔÃ ÃF 10Ä ÄÁQ Q OÁWhich sediment transport formula is used to calculate the concentration at
ÁQ Q OÁthe bed. The following options are given:

Ô Xh$ ÔÁQ Q OÁRÁò
ò
wÁvan Rijn's formula
Ô XQ% ÔÁQ Q OÁEÁò
ò
wÁEngelund/Hanssen's formula
Ô X:& ÔÁQ Q OÁAÁò
ò
wÁAckers/White's formula
Ô X#' ÔÁQ Q OÁYÁò
ò
wÁYang's streampower formula
Ô X ( ÔÁQ Q OÁIÁò
ò
wÁEinstein's bedload formula
Ô Xõ( ÔÁQ Q OÁMÁò
ò
wÁMayer©Peter/Mueller's formula
Ô XÞ) ÔÁQ Q OÁSÁò
ò
wÁShen/Hung's formula
ÔÇ* p-¿+¿+°°ÔÁQ Q OÁDefault: R.

ÁQ Q OÁNote that only the option R is fully tested.

Ô X¤ ÔÃ ÃF 11Ä ÄÁQ Q OÁDensity of sediments and Shield's coefficient. Default: 2.65 and 0.047.

Ô Xv ÔÃ ÃF 12Ä ÄÁQ Q OÁSchmidt's coefficient. Default: 1.0

Ô XH ÔÃ ÃÄ ÄÃ ÃF 15Ä ÄÁQ Q OÁAn integer that determines how the law walls will be used in the cells
ÁQ Q OÁwhich borders both the wall and the bed. A value of 0 will make the
ÁQ Q OÁprogram use wall laws on both walls. A value of 1 will make the program
ÁQ Q OÁonly use wall laws on the bed wall. Default: 0.

Ô XÕ ÔÃ ÃF 16Ä ÄÁQ Q OÁRoughness coefficient which is used on the side walls and the bed. If not
ÁQ Q OÁset, the coefficient is calculated from the Manning's friction coefficient.
Ô X§
ÔÁQ Q OÁThe file Ã ÃbedroughÄ Ä overrides this value for the bed cells.

Ô Xy ÔÃ ÃF 17 Ä ÄÁQ Q OÁTime step in seconds. When this is above 10ÃÃ©8ÄÄ a transient term is included.
ÁQ Q OÁThis is however not tested, and it is not certain that the procedure will give
ÁQ Q OÁcorrect results.

Ô X ÔÃ ÃF 20Ä ÄÁQ Q OÁRepeated calculation option. An integer is read, and the calculation sequence
Ô X ÔÁQ Q OÁon the Ã ÃF 2Ä Ä data set will be repeated this many times. Note that the graphical
ÁQ Q OÁview of the bedlevel changes will only appear on the last iteration when
ÁQ Q OÁsediment calculations are done. Also note that if a result file is read in the
Ô XÁ ÔÁQ Q OÁÃ ÃF 2Ä Ä data set, it is olnly read during the first iteration.

Ô X ÔÃ ÃF 22Ä ÄÁQ Q OÁMinimum porosity and relaxation factor for porosity calculations. Two
ÁQ Q OÁfloats. Default 0.2 and 2.0.

Ô XN ÔÃ ÃF 23Ä ÄÁQ Q OÁAccelerated deposition routine. Two floats are read. The routine fills a
ÁQ Q OÁreservoir to a certain percentage of the total volume. This volume fraction is
ÁQ Q OÁgiven on the first float. The filling is based on only ÃÃoneÄÄ water flow
ÁQ Q OÁcalculation. Therefore the sediments may fill over the water surface in some
ÁQ Q OÁlocations. The routine then moves away some sediments so that a
ÁQ Q OÁcertain water depth is kept. This depth is given by the second parameter, in
ÁQ Q OÁmeters. The surplus sediment is distributed to neighboring cells according to
ÁQ Q OÁthe one water flow calculation. The redistribution is iterated so that there are
ÁQ Q OÁno filling above the minimum water depth.

ÁQ Q OÁNote that the bed changes are calculated in the center of the cells, and that
ÁQ Q OÁchanges in the grid therefore are interpolated from four surrounding cells.
ÁQ Q OÁThis means that even if the four surrounding elements are filled to the
ÁQ Q OÁgiven criteria, the grid line may not be exactly on this level. This may cause
ÁQ Q OÁbedlevels to rise above the waterlevel. The used should examine the grid
ÁQ Q OÁafter the bed changes to observe whether this has occurred. Choosing a
ÁQ Q OÁhigher value of the minimum water depth will decrease the chance for such
ÁQ Q OÁa phenomena to occur. ÔÇ*
p-¿+¿+°°ÔÁQ Q OÁNote that this routine has not been tested yet.

Ô XÒ ÔÃ ÃF 24Ä ÄÁQ Q OÁTurbulence model. An integer is read, which corresponds to the following
ÁQ Q OÁmodels:

ÁQ Q OÁ0 : standard k©ÀÀ model (default)
ÁQ Q OÁ1 : standard RNG model, see reference [7]
ÁQ Q OÁ2 : constant eddy©viscosity model
ÁQ Q OÁ3 : LES RNG model, see reference [7]

ÁQ Q OÁNote that only option 0 has been implemented.

Ô Xì
ÔÃ ÃF 25Ä ÄÁQ Q OÁPorosity parameters. Four floats and one integer. The two first floats are
Ô XÕ ÔÁQ Q OÁidentical to the ones on the Ã ÃF 22Ä Ä data set. The following two floats give
ÁQ Q OÁthe porosity on the second and third level above the ground. These have
ÁQ Q OÁdefault values 0.5 and 0.8. These are used if the roughness height is larger
ÁQ Q OÁthan the height of the bed cell. They are also used if the roughness height is
Ô Xy ÔÁQ Q OÁlarger than the levels of the porosity in the Ã ÃporosityÄ Ä file. The effective
ÁQ Q OÁporosity height is set to maximum of bed cell height and roughness height.
ÁQ Q OÁThe last integer determines the procedure for finding particle diameter in the
ÁQ Q OÁporosity formula. The following options are given: (default 0)

ÁQ Q OÁ0 : Maximum of roughness height and porosity height
ÁQ Q OÁ1 : Maximum of roughness height and 0.33 * porosity height
ÁQ Q OÁ2 : Equal to height of bed cell
ÁQ Q OÁ3 : Maximum of height of bed cell and porosity height

Ô X ÔÃ ÃF 26Ä ÄÁQ Q OÁVolume fraction of sediments in deposits. One float is read. Default 0.3.
ÁQ Q OÁIf the water content is 51 % in a fully saturated sample, the volume fraction
ÁQ Q OÁwill be 0.49.

Ô X7 ÔÃ ÃG 1Ä ÄÁQ Q OÁxnumber, ynumber, znumber and lnumber. There are four integers that show
ÁQ Q OÁthe number of grid lines in the streamwise, cross©streamwise and vertical
ÁQ Q OÁdirection. lnumber is the number of sediment sizes. This data set must be
Ô Xò ÔÁQ Q OÁpresent in the Ã ÃcontrolÄ Ä file. The program will read these values and allocate
ÁQ Q OÁspace for the arrays accordingly. If the space (RAM) on the computer is not
ÁQ Q OÁsufficient, the program will crash.
Ã ÃÄ Ä
Ô X" ÔÃ ÃG 3Ä ÄÁQ Q OÁVertical distribution of grid cells. This is further explained on Fig. 2,
ÁQ Q OÁwhere an example is given. This dataset must be present in the file.
ÚA þx°È¡§%ÿ<ddVEGRDI.CGMé$þ)ÙH4
Ô SØ ÔÁà&
MwÁÚÃ Ã
Ô XQ% ÔG 6Ä Ä ÁQ Q OÁData set for calculating water surface location with an adaptive grid. Three
ÁQ Q OÁintegers and two floats:

ÁQ Q OÁÁò
ò
wÁiSurf:
ÁQ Q OÁÁò
ò
wÁjSurf:
ÁQ Q OÁÁò
ò
wÁkSurf:
Ô0Ç*p-¿+¿+°°°°"NA0ÔÔ$°°""°°""NA°"$ÔÁQ Q OÁThese are three integers that indicate three grid lines. This point is a
ÁQ Q OÁreference point, and it is not moved. In the present implementation, Ksurf
ÁQ Q OÁhave to be equal to znumber + 1. If not, a warning message is sent to the
ÁQ Q OÁboogie file, and Ksurf is set to znumber + 1. The computations continue
ÁQ Q OÁafterwards.

ÁQ Q OÁÁò
ò
wÁRelaxSurface:

ÁQ Q OÁThis is a float that relaxes the estimation of the increment to the new
ÁQ Q OÁrecalculated water surface. Recommended values are between 0.5 and 0.95.

ÁQ Q OÁÁò
ò
wÁConvSurface:

ÁQ Q OÁThis float sets the limit for when the water surface should be recalculated.
ÁQ Q OÁThe water surface will be updated when the maximum residual of the
ÁQ Q OÁequations are below this parameter. Recommended value: 0.01 © 1.0

Ô XÇ* ÔÃ ÃG 7Ä ÄÁQ Q OÁThis data set specifies water inflow on geometry sides, bed or top.Ô0Ç*p-¿+¿+°°°°"NA0ÔÔ X ÔÁQ Q OÁEach surface is given on Ã ÃoneÄ Ä G 7 dataset. It is possible to have up to 19
ÁQ Q OÁG 7 datasets.

ÁQ Q OÁOn each dataset, seven integers and four floats are read. The names of these
ÁQ Q OÁvariables are:

ÁQ Q OÁG 7 type, side, a1, a2, b1, b2, parallel, update, discharge, Xdir, Ydir, Zdir

ÁQ Q OÁEach variable is explained in the following:

ÁQ Q OÁ© type: 1: outflow, 0: inflow.
ÁQ Q OÁ© side: 1: plane i=1, ©1: plane i=xnumber,
ÁQ Q OÁ (cross©streamwise plane)
ÁQ Q OÁ 2: plane j=1, ©2: plane j=ynumber,
ÁQ Q OÁ (streamwise plane)
ÁQ Q OÁ 3: plane k=1, ©3: plane k=znumber
ÁQ Q OÁ (horizontal plane)
ÁQ Q OÁ© a1,a2,b1,b2: four integers that determine the limits of the
ÁQ Q OÁ surface. An example is shown in Fig. 3.
Úa: xûÈÉO>¼! ddAREA.CGM: )·ËÔ S ÔÚÔ$°°""K°°""Õa°"$ÔÁQ Q OÁ© parallel: direction of the flow:
ÁQ Q OÁ 0: normal to surface
ÁQ Q OÁ 1: parallel to grid lines normal to surface
ÁQ Q OÁ 2: direction is specified (vector directions)

ÁQ Q OÁ© update: 0 for not update, 1 for update.
ÁQ Q OÁ (Not implemented by Jan©93)
ÁQ Q OÁ© discharge: discharge in qm/s. Note that the sign of the
ÁQ Q OÁ discharge must correspond with the direction of the
ÁQ Q OÁ desired flow velocity. Positive discharges indicate
ÁQ Q OÁ discharges in positive directions.
ÁQ Q OÁ© Xdir: direction vector in x©direction
ÁQ Q OÁ© Ydir: direction vector in y©direction
ÁQ Q OÁ© Zdir: direction vector in z©direction
ÁQ Q OÁ
ÁQ Q OÁNote that the update option is not implemented by March©93. Still, thereÔ0Ç*p-¿+¿+°°°û"ý a0ÔÁQ Q OÁmust be an integer present for this data in the data set.

ÁQ Q OÁExample: G 7 0 1 2 11 2 11 0 0 32.0 1.0 0.0 0.0

ÁQ Q OÁThis example specifies inflow in the most upstream cross©section. The
ÁQ Q OÁinflow area is from cell no. 2 to cell no. 11 in both cross©
ÁQ Q OÁstreamwise and vertical direction. The flow direction is normal to the
ÁQ Q OÁcross©section. The discharge is 32 cubic meters/second.
ÁQ Q OÁ
ÁQ Q OÁThe parameter "side" can be used to specify flux on sections that have been
ÁQ Q OÁ"amputated" by the multi©block procedure. The parameter side is then
ÁQ Q OÁevaluated as the number of the block plus 10. Example: A geometry with
ÁQ Q OÁone multi©block that starts at node i=30. To specify flux on wall i=29, use
ÁQ Q OÁthe G7 data set with the parameter "side" set to 11 (10+1).

ÁQ Q OÁRemember to define the walls of the boundary that when this dataset is
ÁQ Q OÁused. This must be done on the W 4 data set.

Ô Xb ÔÃ ÃG 8Ä ÄÁQ Q OÁValues for initial velocities. Up to 19 data G 8 data sets can be used. Six
ÁQ Q OÁintegers are read first to specify the volume that is being set. Then three
ÁQ Q OÁfloats are read, which gives the velocities in the three directions.

ÁQ Q OÁG 8 i1 i2 j1 j2 k1 k2 U V W

Ô XØ ÔÃ ÃG 11Ä ÄÁQ Q OÁSource terms for the velocity equations. Six integers and two floats.

ÁQ Q OÁi1,i2,j1,j2,k1,k2, source, relax

ÁQ Q OÁThe first six integers give the cells that are influenced by the source
ÁQ Q OÁterm. The source variable is the form factor times a diameter of a cylinder
ÁQ Q OÁin the cell. The relaxation variable is recommended set between 1.0 and
ÁQ Q OÁ2.0

Ô X ÔÃ ÃG 12Ä ÄÁQ Q OÁSediment source for multi©block border. This is used where there is inflow
ÁQ Q OÁof sediments in a branch of a block that is cut of. An integer is first read,
ÁQ Q OÁwhich tells the number of the block. Then lnumber floats are read, which is
ÁQ Q OÁinflow of sediments for each size. This is given in kg/s. The option is not
ÁQ Q OÁfully tested yet.

Ô X# ÔÃ ÃG 13Ä ÄÁQ Q OÁOutblocking option that is used when a region of the geometry is blocked
ÁQ Q OÁout by a solid object. An integer is read first, which determines which sides
ÁQ Q OÁthe wall laws will be applied on. The following options are possible:

ÁQ Q OÁ0: No wall laws are specified
ÁQ Q OÁ1: Wall laws are used on the sides of the block
ÁQ Q OÁ2: Wall laws are used on the sides and the top of the block
ÁQ Q OÁ3: Wall laws are used on the sides, the top and the bottom of the block
ÔÇ*p-¿+¿+°°ÔÁQ Q OÁSix integers are then read, i1,i2,j1,j2,k1,k2. These integers define the block.

Ô XÒ ÔÁQ Q OÁUp to 19 Ã ÃG 13Ä Ä data sets can be used.

Ô X¤ ÔÃ ÃP 2Ä ÄÁQ Q OÁFive floating points that give scaling for the graphical presentation. The
ÁQ Q OÁfirst three gives scales in streamwise, cross©streamwise and vertical
ÁQ Q OÁdirection. The fourth and fifth give movements in left©right and vertical
ÁQ Q OÁdirection. Defaults: 1.0 for the scales, and 0.0 for the movements.

Ô X1 ÔÃ ÃP 3Ä ÄÁQ Q OÁFour integers that give initial location of the graphical plots in streamwise,
ÁQ Q OÁcross©streamwise and vertical direction, and sediment fraction number.

Ô Xì
ÔÃ ÃÄ ÄÃ ÃP 4Ä ÄÁQ Q OÁA character that indicates initial type of plot. "g" means grid, "v" means
ÁQ Q OÁvelocity vectors, "c" means concentration.

Ô X§
ÔÃ ÃÄ ÄÃ ÃW 1Ä ÄÁQ Q OÁManning's number, discharge and downstream waterlevel. This dataset must
ÁQ Q OÁbe present in the file. The parameters given here are used to generate the
ÁQ Q OÁwaterlevel for the calculations. A routine like HEC©2 is used.

Ô XK ÔÃ ÃW 2Ä ÄÁQ Q OÁWater surface initialization array of integers. The first integer tells how
ÁQ Q OÁmany numbers there are in the array. The next numbers tell which cross©
ÁQ Q OÁsections are going to be used in the initialization of the water surface of the
ÁQ Q OÁgrid. The integers must be given in rising order, and start with 1. This
ÁQ Q OÁdataset must be present in the file.

Ô XÁ ÔÃ ÃW 3Ä ÄÁQ Q OÁSpecification of multiple blocks for the multi©block water flow module.
ÁQ Q OÁFirst an integer that indicates the number of extra blocks is given. The
ÁQ Q OÁmaximum value is 9. Then two integers for each extra block are given. The
ÁQ Q OÁfirst integer (MBSkjot) tells where the block is cut off. The second
ÁQ Q OÁinteger (MBStart) tells where the block is added. If the second integer is
ÁQ Q OÁnegative, the block is added on the left side of the main block. Otherwise
ÁQ Q OÁit is added on the right side of the block. This is explained with an
Ú ë x°ÈÉh8Í*ddMULBLOK.CGMë )·ËFig. 4ÚÁQ Q OÁexample on Fig. 4. The corresponding dataset would be: W 3 1 10 ©5

Ô Xò ÔÃ ÃW 4Ä ÄÁQ Q OÁSpecification of extra walls for the multi©block water flow module. Seven
ÁQ Q OÁintegers have to be given for each wall. There can be up to 29 walls, and
Ô XÄ ÔÁQ Q OÁeach wall is described on Ã ÃoneÄ Ä W 4 data set.

ÁQ Q OÁThe variable names are:

ÁQ Q OÁW 4 dir,posneg,node,a1,a2,b1,b2

Ô X:& Ô ÁQ Q OÁThe first integer, dir, indicates the plane. 1 is the j©k plane (cross©section),
ÁQ Q OÁ2 is the i©k plane (longitudinal section) and 3 is the i©j plane (seen from
ÁQ Q OÁabove).

ÁQ Q OÁThe second integer, posneg, indicates if the wall is in the
ÁQ Q OÁpositive or negative direction of the node. The coordinates are given forÔ0Ç*p-¿+¿+°°°°"c0ÔÔ$°°""°°""y°"$ÔÁQ Q OÁnodes. 1 or ©1 is given. If 0 is given, a previously set wall is deleted.

ÁQ Q OÁThe third integer is the number of the node plane.

ÁQ Q OÁAn example is given in Fig. 5 below. The figure shows the i©j plane.
ÁQ Q OÁThe wall is to be given on node i=4. If the second integer, posneg, is 1,
ÁQ Q OÁthen wall laws are applied on the wall upstream of node 4, in the negative
ÁQ Q OÁi©direction. If posneg = ©1, then the wall laws are applied on node 4
ÁQ Q OÁif the cell in the downstream i©direction (line i=3) is a wall.
Ú¡ÃxZÈÉë<ß+ddMULWAL.CGMÃ)·ËFig. 5ÚÔ$°°""ª°°""K¡°"$Ô
ÁQ Q OÁThe four following integers are indexes a1,a2,b1,b2, which gives the
ÁQ Q OÁtwo©dimensional coordinates for the corner points of the part of the planeÔ@Ç*p-¿+¿+°°!°°"c°Z"å,¡@ÔÁQ Q OÁthat is described. The four integers are further explained on Fig. 3.

ÁQ Q OÁNote that this option has not been tested for internal walls yet.

Ô X¤ ÔÃ ÃW 5Ä ÄÁQ Q OÁDifferent Manning's values than the default value for cross©sections. An
ÁQ Q OÁinteger is first read, which tells how many cross©sections are read. Then
ÁQ Q OÁan integer and a float is read for each cross©section. The integer tells which
ÁQ Q OÁcross©section is changed, and the float tells the Manning's value. Several
Ô XH ÔÁQ Q OÁÃ ÃW 5Ä Ä data sets can be used.

Ô X ÔÃ ÃSÄ ÄÁQ Q OÁInteger that gives the number of the sediment fraction, float that gives the
ÁQ Q OÁsize of the sediments for these fraction, float that gives the fall velocity for
ÁQ Q OÁthe fraction.

ÁQ Q OÁThere must be lnumber datasets like this in the file, as long as
ÁQ Q OÁsediment concentration is to be calculated. All numbers are given in SI
ÁQ Q OÁunits, that is, the grain size is given in meters and the fall velocity is given
ÁQ Q OÁin meters/second. This data set must be present when sediment transport is
ÁQ Q OÁcalculated.

ÁQ Q OÁNote that this dataset defines the grain sizes that are used in the program.
ÁQ Q OÁThis goes for both the grain sizes in the bed and in the suspended sediment
ÁQ Q OÁcalculation.

ÁQ Q OÁThe grain sizes should be numbered from 1 to lnumber, where size 1 is the
ÁQ Q OÁcoarsest size, and the following sizes have increasingly smaller diameter.

Ã ÃÄ ÄÃ ÃNÄ ÄÁQ Q OÁThese data sets define different grain size distributions for the sediments
ÁQ Q OÁthat are in the bed when the calculation starts.

ÁQ Q OÁMaximum 10 different samples can be used.

ÁQ Q OÁEach sample has its own number. This starts with zero, and increases
ÁQ Q OÁsequentially to the total number of samples © 1.

ÁQ Q OÁFirst, an integer for the number of the sample is given. The second integer
ÁQ Q OÁshows which size is described. The third number is a float, which gives the
ÁQ Q OÁfraction of the size in the sample. It is required to have lnumber N data sets
ÁQ Q OÁfor each sample. These data sets must be present when sediment transport is
ÁQ Q OÁcalculated.

ÁQ Q OÁThe number of N datasets that is required is:

ÁQ Q OÁ (number of grain size distributions) x (number of sediment sizes)
Ã Ã
Ô XÇ* ÔBÄ ÄÁQ Q OÁThis data set gives where the samples (from the N © data sets) are placed inÔÇ*p-¿+¿+°°ÔÁQ Q OÁthe geometry before the calculation starts.

ÁQ Q OÁThe first integer indicates the number of the sample. The four following
ÁQ Q OÁintegers give the number of the corner cells on a rectangle of the bed.
ÁQ Q OÁThe sample is placed on the bed from i=second integer to i= third integer
ÁQ Q OÁand j=fourth integer to j=fifth integer. A B data set overwrite previous B
ÁQ Q OÁdatasets.

ÁQ Q OÁThis data set must be present when sediment transport is
ÁQ Q OÁcalculated.

ÁQ Q OÁThe dataset B 0 0 0 0 0 tells that sample no. 0 covers the whole
ÁQ Q OÁbed.

Ô X¾ ÔÃ ÃIÄ ÄÁQ Q OÁInflowing sediments. First integer shows which sediment fraction is
ÁQ Q OÁsimulated. Second number is a float that is the amount of inflowing
ÁQ Q OÁsediment of this size in kg/s. lnumber of these data sets must be present
ÁQ Q OÁwhen sediment concentrations are calculated.

ÁQ Q OÁNote that the I dataset follows the sizes on the S data set, but it has nothing
ÁQ Q OÁto do with the N and B data sets.
Ã ÃÄ Ä
Ô X ÔÃ ÃÄ ÄÃ ÃK 1Ä ÄÁQ Q OÁNumber of iterations for flow procedure and number that determines the
ÁQ Q OÁminimum iterations between updates of water surface. Two integers.
ÁQ Q OÁDefault: 10 9000

Ô Xª ÔÃ ÃK 2Ä ÄÁQ Q OÁTwo integers that indicate of laws of the wall or symmetry conditions are
ÁQ Q OÁbeing used. The first number applies to the side walls. The second number
ÁQ Q OÁapplies for the surface. 0 is used for wall laws, and 1 for free
ÁQ Q OÁsurface/symmetry. Wall laws are always used for the bed, if not changed by
ÁQ Q OÁthe W 4 data set. Default: 0 1.

Ô X ÔÃ ÃK 3Ä ÄÁQ Q OÁRelaxation factors. Six floats. For the three velocity equations, the pressure
ÁQ Q OÁcorrection equation and the k and ÀÀ equation. For further description of the
ÁQ Q OÁrelaxation factors, see a following chapter. Default: 0.8 0.8 0.8 0.2 0.5 0.5

Ô XÄ ÔÃ ÃK 4Ä ÄÁQ Q OÁNumber of iteration for each equation. Six integers. Works for both Spider
ÁQ Q OÁand the multi©block flow module. Default: 1 1 1 5 1 1

Ô X# ÔÃ ÃK 5Ä ÄÁQ Q OÁBlock©correction index for each equation. Six integers. If 0, no block©
ÁQ Q OÁcorrection. If 1, the block©correction is used. Default: 0 0 0 0 0 0

ÁQ Q OÁNote that the block©correction will not work with the SOU scheme.

Ô X ( ÔÃ ÃK 6Ä ÄÁQ Q OÁSix integers are read which determines wheter the SOU or POW scheme is
ÁQ Q OÁused. If 0 POW is used, if 1, SOU is used. Note that presently the multi©
ÁQ Q OÁblock flow module may not converge completely when the SOU scheme is
ÁQ Q OÁused with multiple blocks. This may be due to a bug. SOU will work ifÔÇ*p-¿+¿+°°ÔÁQ Q OÁthere is only one block.

ÁQ Q OÁNote that the multi©block flow module will always use the POW scheme
ÁQ Q OÁfor the pressure©correction equation. Default: 0 0 0 0 0 0

Ô X ÔÃ ÃK 7Ä ÄÁQ Q OÁCorrection of negative sources if 1. Six integers. This must not be used for
ÁQ Q OÁthe pressure or velocity equations. Only works for Spider.
ÁQ Q OÁDefault: 0 0 0 0 0 0

Ô X1 ÔÃ ÃK 8Ä ÄÁQ Q OÁA character. If N the calculation stars from default. If Y, the Ã ÃcontÄ Ä file is
ÁQ Q OÁread by Spider. This data set only works with Spider. The same effect can
ÁQ Q OÁbe achived with the multi©block flow module by using the F 2 data set.
ÁQ Q OÁDefault: N

Ô X¾ ÔÃ ÃK 9Ä ÄÁQ Q OÁA character deciding if Spider should used SIMPLE or SIMPLEC.
ÁQ Q OÁY= SIMPLE, N = SIMPLEC. Only works with Spider. Default: Y

Ô Xy ÔÃ ÃK 10Ä ÄÁQ Q OÁA character that decides if a vector solver is used with Spider. The vector
ÁQ Q OÁsolver is used on the Cray. Y = vector solver, N = TDMA solver. The
ÁQ Q OÁmulti©block flow module also has two solvers, a Gauss©Seidel solver and a
ÁQ Q OÁTDMA solver. Then Y indicates the Gauss©Seidel solver. For use on a PC,
ÁQ Q OÁthe TDMA solver is recommended. Default: Y

Remember that the order of the data sets may be important. For example the G 1 data set
should be early in the file. If the order of the data sets follows the description given here,
this should not be a problem.

Ô X| ÔÃÃThe Ã ÃXCYC.DATÄ Ä fileÄÄ

This file is the input file for Spider, and contains the grid for the calculation, inflow
velocity and turbulence, number of iterations, relaxation coefficients, block©correction
Ô X Ôpresence etc. The file is written from SSII as Ã ÃXCYC.LASÄ Ä and has to be renamed or
Ô X Ôcopied to the Ã ÃXCYC.DATÄ Ä file. Remember that the CRAY and computers running UNIX
require that capital letters are used for this file.

Ô X! ÔÃÃThe Ã ÃloggfilÄÄÄ ÄÃÃ fileÄÄ

This is a file that is used to log bed changes between each time the bed is changed. The
bed changes are written with the "append" mode in C, so that for the file to be updated, it
must exist before SSII writes to it. What is written to the file will be appended to what is
present in the file.

What is written to the file is a time step for each change, together with the bed changes
for each cell in meters. The bed sediment grain size distribution for each time step is also
written to this file.
ÔÇ*p-¿+¿+°°ÔÔ X ÔWhen SSII starts, it looks for a file called Ã Ãloggfil.preÄ Ä, and if it exist, it uses this file to
change the bed and give initial bed sediment grain size distribution. If one decides to start
Ô XÒ Ôfrom a previous time step, one must have copied Ã ÃloggfilÄ Ä to Ã Ãloggfil.preÄ Ä before starting SSII.
Ã Ã
Ô X¤ ÔÄ ÄÃ ÃloggfilÄ Ä is written from the "bedchange" module in SSII, which is activated in the "3D bed"
graphics window.

Ô X_ ÔThe Ã ÃloggfilÄ Ä file is never overwritten by the program. New bed changes are appended to the
end of the file. The user must therefore take care so that only the relevant data exist in the
Ô X1 ÔÃ ÃloggfilÄ Ä and Ã Ãloggfil.preÄ Ä files.

Ô X
ÔNote that if the program does not find the Ã Ãloggfil.preÄ Ä file, a warning is written to the
Ô Xì
ÔÃ ÃboogieÄ Ä file, and the program proceeds normally.

Ô X§
ÔÃÃThe Ã ÃboogieÄ Ä fileÄÄÃ Ã Ä Ä

This is a file that shows a print©out of intermediate results from the calculations. It also
shows parameters as average water velocity, shear stress and water depth in the
initialization. Trap efficiency and sediment grain size distribution is also written here. If
errors occur, an explanation is also often written to this file before the program stops. The
file contains the data that is normally written to the screen in a DOS program.

Ô Xï ÔThe option D on the Ã ÃF 1Ä Ä data set will give additional print©out to the file.

Initially in the file it is written how much memory that is occupied by the arrays that are
dynamically allocated. To estimate the total memory requirement, add 1 MB to this value.

A table then follows, which shows the cross©sectional area, hydraulic radius, average
velocity and water level at the cross©sections that have been used for initializing the water
Ô XN Ôsurface. If the option D on the Ã ÃF 1Ä Ä data set is used, this information is written for all the
cross©sections additionally. Then a table of waterlevels for all cross©sections follows.

If the MB©flow module is used, the residual norms are written. Then follows a sequence
of two lines for each iteration of MB©flow. An example with four iterations is shown
below:

Ô V! ÔÃÃIter: 5, Resid: 1.69e-05 4.10e-06 2.73e-05 1.17e-04 1.38e-02 1.13e-02
Cont: 9.23e-08, DefMax: 1.65e-03, U,V,W(96,7,20): 6.40e-01 -5.14e-03 5.76e-02
Iter: 6, Resid: 1.62e-05 3.85e-06 2.62e-05 1.10e-04 1.31e-02 1.08e-02
Cont: 9.23e-08, DefMax: 1.56e-03, U,V,W(96,7,20): 6.40e-01 -5.14e-03 5.76e-02
Iter: 7, Resid: 1.57e-05 3.65e-06 2.50e-05 1.04e-04 1.25e-02 1.03e-02
Cont: 9.23e-08, DefMax: 1.48e-03, U,V,W(96,7,20): 6.40e-01 -5.14e-03 5.76e-02
Iter: 8, Resid: 1.51e-05 3.46e-06 2.38e-05 9.86e-05 1.18e-02 9.77e-03
Ô V ( ÔCont: 9.23e-08, DefMax: 1.41e-03, U,V,W(96,7,20): 6.40e-01 -5.14e-03 5.76e-02ÄÄ

The first line has the word "Iter" at first. Then an integer follows, which shows the
number of the iteration. In the example above this runs from iteration number 5 to 9. ThenÔÇ*p-¿+¿+°°Ôthe residuals for the six equations are shown. The x,y and z velocity equations are first,
Ô Xé Ôthen the pressure equation and the k and ÀÀ equation follow. All these must be under 10ÃÃ©3ÄÄ
before the solution has converged.

The second line starts with the word "Cont:". Then a floating point value is shown. This is
the sum of all the inflow and outflow in the geometry. This should be a very low value,
Ô Xv Ôtypically under 10ÃÃ©7ÄÄ. If a larger value is given, check the boundary conditions. Then the
word "DefMax" is written. The residual for the cell with largest water continuity defect is
then written. The indexes for this cell are then written, with the velocities in the three
directions for this cell. In iteration 9 for the example above, the maximum water
continuity defect was 1.41e©3 kg/s for cell i=96, j=7, k=20. The velocity in the x©direction
for this cell was 0.64 m/s, the velocity in the y direction was ©5.14 mm/s and the velocity
in the vertical direction was 5.76 cm/s.

If the sedimentation calculation is used, the values of the sub©micro time steps are written
for each recalculation of the bed sediment grain size distribution. The fluxes through the
walls and the trapped sediments are also written for each size. An example is shown
below:

Ô VK ÔÃÃTrap efficiency calculation: all values in kg/s
l=1: FluxI1: 2000, FluxI2: 0.00401221, FluxJ2: 0 Trapped: 1999.84
l=2: FluxI1: 2000, FluxI2: 89.5709, FluxJ2: 0 Trapped: 1909.75
ÄÄ
Two sizes are given in the example, and the inflow is 2000 kg/s for each size. Size 1 has
an outflow of 0.004 kg/s and 1999.84 kg is trapped. For size 2 the outflow is 89.57 kg/s
and 1909.75 kg is trapped. The continuity defect should be under the equivalent residual
norm for the solution of the convection©diffusion equation.

After the sedimentation calculation has finished, the sum of sub©micro time steps is
written. This is called the micro time step. A further explanation of the time steps is given
in [4], on page 39.

If the bed changes are calculated this is also written to the file. The bed changes are given
in meters. An example is given below:

Ô VÛ ÔÃÃBedMove(10,8) = 1.009495e-03 meters
BedMove(10,9) = 1.030140e-03 meters
BedMove(10,10) = 1.019934e-03 meters
ÄÄ
In the example, the deposition in cell i=10, j=10 causes the bed to rise 1.02 mm during
the given micro time step.

Ô X#' ÔÃÃThe Ã ÃconresÄ Ä fileÄÄ

This file is written after the sediment concentration calculation has finished. Each line in
Ô XÞ) Ôthe file contains three indexes for the node numbers, and then Ã ÃlnumberÄ Ä floats that give the
concentration for the sizes. An example is given below:ÔÇ*p-¿+¿+°°ÔÔ V ÔÃÃ1 1 21 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00
1 2 2 5.075079e-04 0.000000e+00 1.026459e-04 1.030081e-04 1.032351e-04
1 2 3 4.064470e-04 0.000000e+00 1.011788e-04 1.015358e-04 1.017596e-04
1 2 4 4.004061e-04 0.000000e+00 9.967497e-05 1.000267e-04 1.002471e-04
ÄÄ

Ô Xv ÔÃÃThe Ã ÃinnflowÄ Ä fileÄÄ

This file is used to read velocities in three directions for the upstream boundary condition.
The program searches for this file, and uses the data on the file if it exists. If the file does
Ô X Ônot exist, a warning message is written to the Ã ÃboogieÄ Ä file, and the program proceeds
normally.

On each line the velocities in a cell of the upstream cross©section are given. First, the
character E is written. Then the indexes j and k (horizontal and vertical) are given. Then
the velocity components in the x,y and z directions are given. An example is given below:

Ô Vy ÔÃÃE 2 2 0.299115 0.023009 0ÁÐÐ~Á
Ô Vb ÔE 2 3 1.79469 0.138055 0Á¯¯}Á
Ô VK ÔE 2 4 1.9941 0.153394 0Á|Á
Ô V4 ÔE 2 5 2.19351 0.168733 0Á¯¯}Á
ÄÄ

Ô Xï ÔÃÃThe Ã ÃporosityÄ Ä fileÄÄ

This file is used when the bed of the river is covered by stones, and a porosity term is
used in some of the cells. This file describes the location and magnitude of the porosity in
the geometry. Note that the porosity routine is not finished. An example is given below:

Ô Ve ÔÃÃP 17 6 3.349774 3.399189 3.450101 3.499517 0.000000 0.700000 0.833333 1.000000
P 17 7 3.358273 3.413603 3.470610 3.525940 0.000000 0.653846 0.807692 1.000000
P 17 8 3.403323 3.426084 3.449536 3.472297 0.000000 0.642857 0.785714 1.000000
ÄÄ
First, the character P is read. Then two indexes for the i and j number of the bed cell is
read. Then four vertical levels are read, which have the same zero reference as the
Ô XÛ ÔÃ ÃkoordinaÄ Ä file. The porosities in each of these levels are then read.

Ô X" ÔÃÃThe Ã ÃinterpolÄ Ä fileÄÄ

Vertical profiles of velocity or concentration are sometimes needed. Coordinates for the
Ô XQ% Ôlocations where the profiles are wanted are given in this file. When the Ã Ãwrite resultsÄ Ä
routine is activated, it will search for this file. If this file is not found, it will proceed
Ô X#' Ônormally and write the Ã Ãresult Ä Äfile. If the Ã ÃinterpolÄ Ä file is found, the program will Ã ÃnotÄ Ä write
Ô X ( Ôto the Ã Ãresult Ä Äfile, but write the interpolated vertical velocities to a file named Ã ÃinterresÄ Ä.
Ô Xõ( ÔAn example for an Ã ÃinterpolÄ Ä file is given below:

Ô VÇ* ÔÃÃM 2.03 0.5ÔÇ*p-¿+¿+°°ÔM 4.06 0.39
M 4.06 0.5
ÄÄ
The character M is read first, and then the x and y coordinates for the point one wishes to
interpolate the vertical profile to. If the concentration has been calculated recently
Ô X Ô(concentration in cell [2,2,0,lnumber] is above 10ÃÃ©8ÄÄ), then the concentrations are
interpolated. Otherwise, the velocities are interpolated. The results are written to a file
Ô X_ Ôcalled Ã ÃinterresÄ Ä.

Ô X ÔÃÃThe Ã ÃbedroughÄ Ä fileÄÄ

This file is used to give a roughness height to individual bed cells. Values in this file
Ô XÕ Ôoverrides the value calculated by Manning's coefficient, and the value given on the Ã ÃF 16Ä Ä
data set. On each line a character, two integer and a float are given. The first character is
a B, and the two following integers are indexes for the bed cell. The float is the roughness
in meters. An example is given below:

Ô Vb ÔÃÃB 19 2 0.001
B 19 3 0.001
B 19 4 0.001
Ô V Ô ÄÄ

ÃÃPlotting and graphics filesÄÄ

Ô XÁ ÔAll the Ã Ãsurf*.datÄ Ä and the Ã Ã*.blnÄ Ä files are intended to be used with the SURFER graphic
program, which draws threedimensional plots. The files are written from the "writespider"
Ô X Ômodule in SSII, and are therefore automatically generated every time the Ã ÃXCYC.LASÄ Ä file
is written from SSII.

Ô X7 ÔÃÃThe Ã Ãsurfbed.dat Ä ÄfileÄÄ

This file writes the bed geometry.

Ô XÄ ÔÃÃTheÃ Ã surfmap.datÄ Ä fileÄÄ

This file writes the shear stress, bed velocity and depth©averaged velocity as seen from
above.

Ô X:& ÔÃÃThe Ã Ãsurfcro.datÄ Ä fileÄÄ

This file writes velocities in a cross©section. The number of the cross©section is "xnode",
Ô Xõ( Ôwhich can be chosen from the menu or is written in the P 3 dataset in the Ã ÃcontrolÄ Ä file.

ÔÇ*p-¿+¿+°°ÔÔ X ÔÃÃThe Ã Ãcontour.blnÄ Ä fileÄÄ

This file describes the line around the bed plane of the grid. It is used with the
Ô X» ÔÃ Ãsurfmap.datÄ Ä file by the Surfer graphics program.

Ô Xv ÔÃÃThe Ã ÃRESULTÄ Ä fileÄÄ

This file contains the results from the calculations from Spider. Spider writes this file after
converging. The results are velocities in three dimensions, k and ÀÀ, pressure, and the
fluxes on all the walls of the cells.

The file is an ASCII file and is read by SSII before starting on the sediment calculations.

Ô Ús§
ÔÃ ÃÃÃGraphicsÄÄÄ Ä

In SSII there are four graphics modules for presentation of results. These can be invoked
any time during the calculation or afterwards. More than one module can run
simultaneously. The four modules are:

Ô X Ô© Ã ÃMapÄ Ä
Ô Xj Ô© Ã ÃLongprofÄ Ä
Ô XS Ô© Ã ÃCrossÄ Ä
Ô X< Ô© Ã ÃBedÄ Ä

The modules use the standard GUI for OS/2, and have approximately the same system of
menus. The modules show a plot, which the used can change dimension of or choose
different parameters for the presentation.

Ô X² ÔÃ ÃMapÄ Ä presents the geometry seen from above. It is possible to get velocity vector plots and
plots and bar plots of concentration, diffusivity, k, etc. It is also possible to plot the grid
,and change between different vertical levels.

Ô XV ÔÃ ÃLongprofÄ Ä presents a longitudinal profile of the geometry. Graphs with different parameters
as a function of depth along the longitudinal profile is obtained. It is also possible to view
the grid or the velocity vectors. It is possible to change between different longitudinal
profiles.

Ô Xã# ÔÃ ÃCrossÄ Ä presents a cross©section of the profile. It is only possible to see a velocity vector
profile. It is possible to change between different cross©sections.

Ô X& ÔÃ ÃBedÄ Ä presents a three©dimensional view of the bed level. Bed changes are marked in
different colors. It is possible to scale and rotate the plot. The bed change module can be
invoked from this window. This window emerges automatically when the sediment
calculation has finished. Note that this routine is not maintained and have several bugs
that will not be fixed.ÔB*p-¿+¿+°°ÔÔ X ÔIt is possible to store the plot in an OS/2 metafile, with the menu command Ã Ãsystem©saveÄ Ä.
It can then be printed or plotted on paper by the standard utilities that comes with OS/2.
Ô XÒ ÔNote that the name of this file will be Ã Ãmapfile.metÄ Ä,Ã Ã longfile.metÄ Ä or Ã Ãcrosfil.metÄ Ä, depending
on which of the graphics procedures that produced the file. If there is an existing file with
the same name, this will ÃÃnotÄÄ be overwritten. The command then has no effect. It is
therefore recommended to rename the file right after it is made.

Ô ÚsH ÔÃ ÃÃÃExperience with convergence for water flow calculationÄÄÄ Ä

In computational fluid mechanics cases it is often a problem to get the solution to
converge. Two factors are important for this problem:

© A good grid.
© Good relaxation coefficients.

It is presumed that the boundary conditions are correct, which is always a good thing to
check first.

Experience shows that the degree of non©orthogonality of the grid will affect the
convergence. A higher degree of non©orthogonality will give slower convergence. A
slower convergence will also be experienced where strong gradients are present. This
applies for example at the inflow of a jet from a wall.

For the convergence of the k and ÀÀ equations for river problems, the size of the cell
closest to the bed is important. This can be changed by changing the first number in the G
Ô X Ô3 data set in the Ã ÃcontrolÄ Ä file. This parameter also depends on the roughness of the bed.
The size of the bed cell should increase with increasing roughness. The height of the bed
cell must however be greater than the roughness of the bed. The following formula can be
used to determine the roughness of the bed, given the friction coefficient M of Manning's
formula:
Úa0(³°K!dddddd`ëddZ-béÿÿÿÿß k_s ~ = ~ {(26 over M)} ^ 6êfô\ PÂC&Pêfô\ PÂC&Pêfô\ PÂC&P··Ïkzzjs··@M··éÏä··§Ï(··426··ÑÏ)zzm6s·óú ßÚÔ$°°""; ; ""!a°; $Ô
Ô"a°; Ô
Ô"a°; Ô
Ô$a°; Ô
Ô; ; ""? °°""ÔThe relaxation coefficients are set to the default values of 0.8 for the velocity equations
and 0.2 for the pressure©correction equation. For k and ÀÀ a value of 0.5 is set. The values
of the relaxation coefficients for the velocity equations and the pressure©correction
equations are presumed to give an optimum convergence for the average flow case. Values
of 0.5 for all equations will often give a slower convergence, but with less probability of
blowing up the solution. Theoretically, the sum of the relaxation coefficient for the
velocity equations and the relaxation coefficient for the pressure©correction equation
should be unity for optimum convergence. However, for some difficult cases, this rule has
to be abandoned.

If the solution blows up after the first few iterations, it is possible to set the relaxation
coefficients very low, and then increase the coefficients for the following iterations.Ô0B*p-¿+¿+°°°K!; a$a0Ô
For some cases where the equation for k is slowest in convergence, a more rapid
convergence has been archived when changing the relaxation coefficient for k from 0.5 to
1.0 after some iterations.

On general it can be said that lower relaxation coefficients will give less instabilities
during the convergence, but a slower convergence. Higher relaxation coefficients will give
more rapid convergence if there are no instabilities. Instabilities can be observed during
the iterations when the residual or the velocities increase and decrease periodically.

Use of block©correction will lead to a more rapid convergence. There is however a
possibility of this procedure leading to negative values of k. Spider will then crash
because in the wall laws the square root of k is calculated. In the version of Spider that is
used in connection with SSII, it is therefore a control in the wall laws that checks if k at
Ô X¾ Ôthe wall is under 10ÃÃ©8ÄÄ. Then a warning is printed, k is set to 10ÃÃ©8ÄÄ and the program then
carries on with the calculations. This procedure does not cause errors as long as there is
no correction in the final iteration just before the solution converges.

Ô Xb ÔThe multi©block flow module avoids this problem never using a k value below 10ÃÃ©9ÄÄ. This
Ô XK Ôvalue is equivalent bed shear stress of 3.0x10ÃÃ©7ÄÄ N/mÃÃ2ÄÄ, and an interpretation of this
Ô X4 Ôminimum value is that the bed shear stress also has a minimum value of 3.0x10ÃÃ©7ÄÄ N/mÃÃ2ÄÄ.

For some cases there have been problems getting the solution to converge if there are
parts of the geometry that has relatively low total velocity. Experience has shown that the
initial conditions then may be important. If such a situation is present it is important to
start the iterations with very low initial velocities.

Ô Ús| ÔÃ ÃÃÃAdvise for interpretation of resultsÄÄÄ Ä

As mentioned earlier, it is advisable to have experience in computational fluid dynamics
when proper interpretation of the results is required. Some guidance is given below.

An important numerical effect that can deteriorate the results is false diffusion. This effect
is most profound for first©order schemes, including the POW scheme. The effect depends
on how well the flow velocity vectors are aligned with the grid lines. For small alignment
angles, the effect is small. Maximum false diffusion will happen when the grid lines are
aligned 45 degrees with the flow. The amount of false diffusion also depends on the size
of the grid cells.

There are three methods to decrease the amount of false diffusion:

1. Decrease the size of the grid cells == increase the number of cells
2. Align the grid with the flow field
3. Use the second©order upwind scheme

Point 2 may be difficult in a practical situation. However, the calculations should beÔB*p-¿+¿+°°Ôcarried out using approach 1 and/or 3 to assess the effect of false diffusion.

Another important characteristic is the boundary condition. This especially applies to the
inflowing boundary. If the velocity field at the inflowing boundary is not known exactly,
one should try different velocity distributions to try to assess the effect of this parameter.
For a river running into a reservoir, it is possible to model a part of the river upstream of
the reservoir, and thereby obtaining a better estimate for the velocity distribution where
the river enters the reservoir. The upstream boundary condition is also important for
sediment calculations, where both the total amount of sediment inflow and the sediment
grain size distribution can be varied. For the bed boundary, it is possible to vary the
roughness to investigate the effect of this parameter. It can also sometimes be
advantageous to make variations for the formula for sediment concentration close to the
bed. This especially applies for sediment particles outside the range for which the formula
is applied for.

When interpreting the results from the model it is also important to keep the accuracy of
model in mind. The k©ÀÀ turbulence model has limitations in how accurate the turbulence
field is predicted. This will also affect the velocity field. For example, when calculating
the recirculation zone for a step case, the length of the recirculation zone can often not be
predicted with any better accuracy than 10 © 30 %.

In some situations the water flow will be time©dependent. An example can be oscillations
behind a cylinder or in an expansion. The equations that are solved by the program do not
have time©dependent terms. Therefore the converged solution will be steady. It is possible
to obtain a steady solution from the model although the physical problem is unsteady. This
must be considered when interpreting the results. The effects of an unsteady solution
compared to the given solution should be assessed if this is probable. When an unsteady
case is solved with a steady method there can be convergence problems. If the relaxation
factors have to be fairly low to get the solution to converge, this can be an indication that
the flow field may be unsteady.

Another topic of interpretation is the resolution of the calculated flow field compared with
the size of the grid cells. Several cells are required to dissolve a recirculating zone. Flow
field characteristics smaller than about 4©7 cells does often not show up in the solution.

Ô ÚsÄ ÔÃ ÃÃÃLiteratureÄÄÄ Ä

Ô Xã# Ô[1]ÁQ Q OÁEngelund, F., and Hansen, E., "A monograph on sediment transport in
ÁQ Q OÁalluvial streams", Teknisk Forlag, Copenhagen, Denmark, 1967.

Ô X& Ô[2]ÁQ Q OÁMayer©Peter, E. and Mueller, R., "Formulas for bed load transport", Report
Ô X' Ô ÁQ Q OÁon Second Meeting of International Association for Hydraulic Research,
ÁQ Q OÁStockholm, Sweden, 1948.

Ô XB* Ô[3]ÁQ Q OÁMelaaen, M. C., "Analysis of curvilinear non©orthogonal coordinates for ÔB*p-¿+¿+°°ÔÁQ Q OÁnumerical calculation of fluid flow in complex geometries", Dr. Ing. Thesis,
ÁQ Q OÁThe Norwegian Institute of Technology, Trondheim, 1990.

Ô X» Ô[4]ÁQ Q OÁOlsen, N. R., "A numerical model for simulation of sediment movements in
ÁQ Q OÁwater intakes", Dr. Ing. Dissertation, The Norwegian Institute of
ÁQ Q OÁTechnology, Trondheim, 1991.

Ô X_ Ô[5]ÁQ Q OÁVan Rijn, L. C., "Mathematical modeling of morphological processes in the
ÁQ Q OÁcase of suspended sediment transport", Ph.D Thesis, Delft University of
ÁQ Q OÁTechnology, 1987.

Ô X
Ô[6]ÁQ Q OÁVanoni, V., et al, "Sedimentation Engineering", ASCE Manuals and reports
ÁQ Q OÁon engineering practice © No54, 1975.

Ô X¾ Ô[7]ÁQ Q OÁAshworth, R., "Renormalisation Group Turbulence Model",
ÁQ Q OÁEight International Conference on Numerical Methods in Laminar and
ÁQ Q OÁTurbulent Flow, Swansea, July, 1993.

Ô Xb Ô[8]ÁQ Q OÁOlsen, N. R. B., and Melaaen, M. C., "Numerical Modeling of Erosion
ÁQ Q OÁaround a Cylinder and Sediment Deposition in a Hydropower Reservoir",
ÁQ Q OÁEight International Conference on Numerical Methods in Laminar and
ÁQ Q OÁTurbulent Flow, Swansea, July 1993.

Ô Xï Ô[9]ÁQ Q OÁOlsen, N. R. B., and Melaaen, M. C., "Three©dimensional numerical
ÁQ Q OÁmodeling of scour around cylinders", ASCE Journal of Hydraulic
ÁQ Q OÁEngineering, Vol. 119, No. 9, September 1993.

Ôp-¿+¿+°°Ô Ô X ÔÃ ÃÃÃAppendix A: control input fileÄÄÄ Ä

this is the control input file for the ssiim program for the mae tian case.
only one sediment size is calculated : 0.125 mm.

T Mae Tian Reservoir title field
Ô XH ÔF 2 RISÁQ Q OÁÁò
ò
wÁÁò
ò
wÁ run-choice
Ô X1 ÔF 8 0.8ÁQ Q OÁ max bedlevelchange
F 10 M sediment formula, possibilities: s,e,m,y,r,a,i
F 11 2.65 0.047 density and shields's coefficient
Ô Xì
ÔG 1 24 12 11 1Áò
ò
wÁ number of nodes in x,y and z direction+sed.size
Ô XÕ ÔG 3 0 3 7 12 20 30 40 60 80 90 100ÁÁÁÁdistribution of vertical grids
G 4 1 6 6 10.0
Ô X§
ÔP 2 0.7 1.0 0.4 0.0Áò
ò
wÁ graphics scaling parameters
Ô X ÔP 3 6 4 6 1ÁQ Q OÁÁò
ò
wÁ graphics nodes, cross-sections and part size no.
Ô Xy ÔP 5 1 9 3 11 2 10 5 13 6 12 ÁxÁ color indexes, do not use four
P 7 H
Ô XK ÔI 1 0.00000000000Áò
ò
wÁinflowing sediments in kg/s
Ô X4 ÔS 1 0.000125 0.009Áò
ò
wÁÁò
ò
wÁ sediment fraction nr, size, fallvelo,
Ô X ÔN 0 1 1.0ÁQ Q OÁÁò
ò
wÁ sediment sample
Ô X ÔB 0 0 0 0 0ÁQ Q OÁÁò
ò
wÁÁò
ò
wÁ bed coordinates, composed of sediment
Ô Xï Ô ÁQ Q OÁ fraction, part of bed covered with this
Ô XØ ÔW 1 50 8.3 6.0Áò
ò
wÁ mannings M, discharge and downst. w. lev.
Ô XÁ ÔW 2 13 1 2 3 4 5 7 9 11 13 15 17 19 21ÁòòÁ initialization cross-section numbers
Ô Xª ÔK 1 10 1000ÁQ Q OÁÁò
ò
wÁÁò
ò
wÁÁò
ò
wÁnumber of iterations for flow procedure
Ô X ÔK 2 0 1ÁQ Q OÁÁò
ò
wÁÁò
ò
wÁ coeff. for influence of banks
Ô X| ÔK 3 0.8 0.8 0.8 0.2 0.5 0.5Áò
ò
wÁ relaxation coefficients
K 4 1 1 1 5 1 1 number of sweeps for each equation
Ô XN ÔK 5 1 1 1 1 1 1 Áò
ò
wÁÁò
ò
wÁblock-correction for eack equation,1=yes
Ô X7 ÔK 6 0 0 0 0 0 0 Áò
ò
wÁ use sou-scheme (1) or pow-scheme (0)
Ô X ÔK 7 0 0 0 0 1 1 Áò
ò
wÁ correction of non-negative variables 1=yes
Ô X ÔK 8 NÁQ Q OÁÁò
ò
wÁÁò
ò
wÁ do not continue (c) from previous iteration
K 9 Y use simple procedure instead of simplec
Ô XÛ ÔK 10 YÁQ Q OÁÁò
ò
wÁÁò
ò
wÁÁò
ò
wÁÁò
ò
wÁ use cray-vector-procedure
ÔÄ p-¿+¿+°°Ô Ô X ÔÃ ÃAppendix B, Flowchart for SSIIMÄ Ä

ÚyÁé$xkÈÈ+8Æ+ZnPSSIIM.CGM!¯é$yÚÔ$°°""»°°"":&Á°"$Ô

3 Responses to “Category : Science and Education
Archive : SSIIM11.ZIP
Filename : MANUAL
”

Daniel says:

January 27, 2013 at 3:59 pm

Very nice! Thank you for this wonderful archive. I wonder why I found it only now. Long live the BBS file archives!
Joshie says:

March 18, 2014 at 4:57 pm

This is so awesome! 😀 I’d be cool if you could download an entire archive of this at once, though.
DiskingRound says:

January 14, 2015 at 10:57 pm

But one thing that puzzles me is the “mtswslnkmcjklsdlsbdmMICROSOFT” string. There is an article about it here. It is definitely worth a read: http://www.os2museum.com/wp/mtswslnk/

3 Responses to “Category : Science and EducationArchive : SSIIM11.ZIPFilename : MANUAL”

3 Responses to “Category : Science and Education
Archive : SSIIM11.ZIP
Filename : MANUAL
”