# ﻿Finally, to obtarun the transfermation :re'latlonsbip' between the glebal XYZ and the actual local xyz coordinate

systems, we substitute Eq.(5.33')~nto Eq, (5,3S} and carry out the required matrix multiplication, This yields

vf! xX + ?xZ coslLl' -\IF ~ + ? xl sin4J

(536) By ,comparing Eqs.{5.29) and(5,36), we can see thattbe member rotation matrix r can be

. xpressed as,

rxX

-v¥ xx -I- ? xl coslV

- ~ 1tX + ~ xl. swn'4J

_ ate {hat the rotation matrix depends only on the global coordinates of the member ends and its angle of fOB I.V. Based on the faregoing derivation, the ang]e of ron 41 is defined as e angie. measured clockwise positive when ~ooll::in,g .in the: negative x direction, through hicb the local xyz coordinate system must be retated around its x axis. so that the xy pl iiIIne becomes vertical with the y axis poi:nt~ng upvlfard (1. e .. , in the posi tive direction of the global Y axis).

The expression of the rotation matrix r, as given by Eq . .{5.37).can be used to determine

e transformation matrices T for the members of space names oriented in any a[1Di trary _ ire tions, except for vertical members. TIlls, is because for such members rxx andrltZ are zero, causing some elemeots ofr in Eq.(5_.37), to become undefined, Tlus situation can be remedied. by defimng the angle of ron differently t70r vertical members, as follews. For

e special case ofvertical members (i.e., members WIDth centroidal or local x axis parallel o the global Y axis). the angle of roll qJ is defined as the angle, measured clockwise pc sitive when loolcing in the negative x d~rection,. througfl wmcb the local Kyz ,coord~naDte

(51)

(5.37)

system mustbe rotated around its x axis, so that me local zaxis becomes parallel to, and po~nts ~n the positive directi.on of,. ~he global] Z .ax~s (Fwg.5. 7 (b»).

The expression of t!he mtal!won matrix r for vertical members can be derived using a procedure similar to that used prev.~ous1y for members widli other orientatiens, We imagine that the vertical member's desired 't or actUiru design) OrlefJ!Ul~h:m: is reached in two steps. as shown in Figs":5 .. 7(a)and(b}Jn the fir.st step, while the members xaxis ]S oriented in. the desiredt vertical) direction, its y and z axes are orie-nted. so that tbe localz axi s is parallel to' the global Z axis, as sho~ In IFig~5.1(a).. As wndicated therei,tbe- member's principal axes in. th~s (Jmaiginary) erientstion are designed asy andz (instead of'y and z, respectively). The dlrect~on of the Jecal xaxis (known from the global coerdmases of the memee« ends) is represented by the unit vector i~ = IxY Iy, while the ruI<l::ctionoftbe z axis ~s given by tfue uni t vector i;z; = I;!; . Thel!nli~ vecto.r if,. directed along~heyaxi.s., can therefore be oonveniently esta:b~ished. uslngme: cross product iz xix.

" Ix Ty h
I
I
ty "" iz-x Ix. = dd 0 o 1 = .;rl{YIx
0 r~y 0 (5.38)

Thus, the transfermationrelationship.betweentbe g~.obal XYZ and the auxiliary local ~'f . coord~l1I~le system isgjvenby

n [0 rx'i" 0 ] ~k ]
= I;}!Y 0 0 h
my
:tz 0 ~ 1I'z I (539)

(52)

(a) OrienlaJion ,0/ VeNictJl Metnber with z Axis par-ailel to Global Z ...b;is

'.1",

"

... ,

~

~ I

,

,

" 1

~

:~

:~

i!

'~

!

\ 11

I ' ,

,I i 111t- 1

~~Z . z ,_ i.i

---,-1j, ..._ ---- I: i~

'f ,. -~,iZ:."

\."'lI . z.

i-', "i

'~

(O) Actua.l Ol'ienla.(ion QfV€Hicai Menzie:r

(53)

Inthe next step'. we rotate the aJW{Hiaqr xyz, coomi:nate system aootf jtsK~axis!>in a cooo®er,eloc\nyjse sense, by the angle 'of roll !,f, mil the :memibe~sprinci,pa~ axes are: in their d!es~red orieetations, This final ,oniemltmn.on (!1,ftihe ,member is de~icted in Fig.,:5_1(b ).jn which thememrer pr,~n<:ipa] axes are .now deswgtIDatedas the y :aJnd a-axes, From tms figure, we CM1 see thaD! thetr.msformation relationship bet'weeD, the auxiliary x,JZ"®dU'Jie actual X]lz.coordma.te systems Is the same as gI1Ve:FiI pNVJ'01!lS~y ~n Eq.(5_3.5) Thus, the

(Fig. 5~7)

desired transfurma.tion. frem jbe globel XYZ coordinate system to ~he' local xyz coordifID!ilIte sys~em C~[) be, obminedby substiwltin_g Eq.,(:5.39) ~m;of:q"(S.3.S}.3ind perronnillJJg the requsted DIa!trix multiplication. This yie1dS

I[ij [0

~v I "'" I -'rxYOOS~

I L = r",ysirru'fl

o

o -J.[~Jl]

S1D'" my

CO&~" lz

(SAIO)

{ ...

.f .._ rwoos,!\IoI xy.sinlf

o ]

si_n~1

oo.s.~

(5,41)

In m.OSl spw:iefirames~ members are USu.Uy' ,onenited so that their angles 'of ron CUI, be mund by inspectioo, There are s1:nJcture.s, ~lIlweVter.~n which.tmle orientations oif some members may be slilchthat their angles, mroU cannot be conviendy determined The oJie1:WUlt~0J11 of such 3. member can alm.mauvely 'be speei fled by means of the: g1dlal cooedinates of a re!ferencepo.int: that lays in one .Qffhepni:liIcipa~ (xy OIXZ) planes o:f~be memilerj, but nM on. its centroidal (x) axis,

To discuss the process of determining the :member mm1trnon matrix r using such a reference po,in!, considertlle spa.oe~f[ame Member sh.mw ilil F.~i,5i,8, ,and ],etXp" Yp• and Zp deno~e the global coerdinates of .!ID ,a!l"bitmri~y chosen reference pcint P" whichis loca.ted wn themembers ]'ocarn xyplane~ but not 0:0 its x axis. Smeetbe g~.oba~ ecerdinases of themember end b are X~, Yb• and. 4,. the pos.iti(lll. vector p~ direcled.&omJ} member end b tereference POll1JJitt P, am. be wnnenas

(SA2)

N@te th:am bo~h poim~' sb and. r are loceted il. thelocal xy planeand, ti1~refoIe:, VeC~{'IF p aJ:sol hesin till__atplane.. Since the dJ~re.d]j0n. cosines of~he~oca~ x axis are already kna:wn

from the global ceosdinates of the member ends, the d~r,ection cosines of the local z axis can beconviently established using the follewing relationship

Ix x P

IZ= ---

(5.43)

I Ix x P I

In which li~ x pi represents the magnitude> ofthe vector that results from the cross product of the vectors ix and P Wjth boih i~. and i.z now known, the direction cosines. of the local y-axis. can be obtained via the cross product.

. ~ ..

~y = J.~ x ~:>:

(5.44)

IY

I .

n

II

~

-·---------------~-----------------------------x

(Fig. .. 5.8,)

In the case that the reference pain,! Pis specified in the local xz plane of the member) the direction eosmes ofthe local y axis need to be determined first using the relationship

P xix

ly=

IP x lx I

(5.45)

(55)

(5.46)

.h should 00 wa~ized.that:~he proooolBe descri.bed.by Bq.(S_42)thro.1!]gh (S,46) enables !!IS toobta~n the member rom!tiQ:~. matrix. r direc!lJly by means ofa. n~fere!lloopa:~nt, withcut .invulvif1lgfhe an,gle 'of ron ,o,fdl.e member. HfJweve:r. ifdesLred.th!e; ang~e ()f[01~ 'can also be ,obta~11Ied from the g:J.obarn coordinalWS of a l'efieJ1ence point To estabHshthe re.~atiOil1!Smp between the .angme of mn l!,I;I and :a:r,eler,eooe po]n~ r of a.memlber,. we firs~ det:ennine the oo:mpon~nts ,o(i the posi'~i~EiI. vector p in.th.e :auxHiaryxyz coordinate sys:tem .• b,y ap~lyin.8 t:h.etmnsIbrmaHo.n. re:lation~bip given in. Eq.(S.3,3).as

f:liY

-(Xp- Xt,) I (Yp § Vb)

n::"\-

r;rl -

Pol _ I.

(Zp - Z~I)

(5.4Th)

In which Px. Pfj and ·p.rrepreSeNlt, respectiv:ely~ me compone:fii~s of~heposjtion ve.aor Pin t_he directions of the ~Qca~x axis ,and. th.e ,and z· pes of tml.e auiHary ~ coordinate system" Now. if the refereucepoint P nes.~nthe xy p~ane ,of the mem~er as sbO'VIIl1~n Fig, 5.'9. ~hem,..ve CD. see fr('1mtJhis fiiguredlat the arng1e of ran '4r"and. t!D.e components o,f p are related br dueioUoVl'i.ns equations:

(56)

,~_ - --- - _,_

In 'me case: that. the reference pownt P is specified in the local xz 'phme~then. froen Frug.5.W we can see that the relatiomsbjp.s oonveem. the sine and (losme of !oJ and the oompon:e!l1tsp can be expressed as

ODS. =" .' ..... ~. =

V :+Pz'j

(5.49')

Equations (5.48) and. (5049) are valid. for :space-fromememoors oriel1~edm any aJbittary dirl::ctioDS.inclllldIDg vem.cw mem'beu. Howeveif~ since I\:X"'" .[~ ~ 0 for vertical memJbe:rs.me ex,ress~.ons for pyand ~ ugiv1en imI.lBqs5.47 (b) and. (0). ,ammot 'be used; appropriate exPWSS]ons for the oomponents, !oftbe pos:~ti.OJl. vectorp Ln the awcHiary;(y,z coordinate ,system.lm!lls1l be d~ri.ved. 'by appbing Eq.(S:.,39').as

o o

(51)

Px =rxY{Yp - YI:!} W ="rx,(Xp~· Xb), .Pr "" Zr ·~Zb

(S,'Oa} 1(550b) (5,50c)

Jtis ~mportant to reaHze ti1lat,forv,erticalllemoors" m3qs.S.50 (b) and. (c) shGIl~d beused t(!l evaluate .P'Y and .pZ" whereas for memms'W'lith other ,orien.btio.ns, these ,eomponeIWlttS are obtained from, Eqs.5.47{b) and {c).Afterpy' and p~ have been evahJl1Ited ,the sine and cosine of the members :Ingle of ron " can be de4enDltn;ed. ewther by ~_,( j _ 4(8)~f the reference point lies in the xy plane, or via :JBqJ.(S .. 49) if the referencepoiat is located in the xz plane, Once .si~."fand oos4<' areb.to~tih.emeD1llberrotation marltrix r can. be determined by Eq.(S . .37> if the member is not one[lJ~ed~n the vertical d~rection.. or via. lSq . .( :'t41 )1 if the member Is "meal.

5.8 PROCEDURE Ji10K ANALYSIS

Us,lng the concepts discussed. ~n the previous sections" we can now deve],op the following step-by-step precedure for the arnaly,s,i:s of space frames 'by die matrix stiffness method.

I. PrepM"e an .. anslytical m.odel,of the :Stl'itlcrufe.identifyWg Hs degrees of DeedJ~m snd restrained coordinates (as diseussed in .secno,n 5.1). Recall thaJlt fer borizonta~ members,.tl1e ooommatetransmDl1lanoml cam be avoided 'by se]ectl.mg U\le, left-end joint. .of'the member as thek:girmingjoinrt.

2. Evallmate the stmCitae stiffrness .matrix S(NOOf' x NDOF)and.fi~ed,,:joint fowe vector Pt{NDOf x 1). for eachm~mibeF (If the ,sttucmw, ped"orm the' foHowing operat~oru;,:

a. Calculatethe len,gt:h ,and. dirootios cosines O. e. ,roseland. sm0)of the mereber,

b, CompUltefue memoor stiffnessm2iItrix in. the glQbal coordinate sys~em, K. using its eXJp~ici~ form glven in Eq.(5 .3). Them.ember global stiffu~s matrix

a]temat~vely ean be obta~ned by first fonningthe mem:ber ]oca~ st~:ffiless matrix k and the llanisfom1lationmatnx T~ _d then evalliilating~he maru:ixtJriple product, K.=TTkT. The matrix K. must be s-ymmettic.

c. If the memcer is subjectedto external loads. then evaluatethe me-mberfixed~ end force vector .in thegJloha~ coor"natt system. F f. Tine member gJ.oba[ MX!edend. force vector can be obmjnedby first fomtrnng the member loeal fixed-end foree vector Qi'(Eq.{ 5Ai).and.dlenlUl~ llierelauMonsmp Fr=T if Qr.

d. Ide~mrnfythe rne:Dilbelf oooe rr1!umlbmand, s®ow the pertinent ,e:I!emel.its of K andF f in their proper posit~ons in the struiCtwe stiffness matrix S snd the flxed-joine for()e vec:torP(, :respectiive,!y.,

The !Complete strucmre stiffness mamx. S. iIlbtaiE!looby a8S1emblmg tie ,st~:ffness coefficients of aU the meD1l.ber1io.f the strectare, mUSlt be sy.mm.emc.

3 If the struenne ~s soubj ectedjo ,jownt loads,then form the joint ~oad vec~o.r P' (NDOFxl).

.:!L Determiae the j,oinl! dis:pl3(iemem.t8 d. SWsblmt:e: p~ P~Jand S inJ10 the :s:lrucrure stiffness relationship, f-Pr= Sl:!, ood solve the .re.sul~lng system cf simultaneous eqw~~~J]jS. for die ool!!lo'!iiV11 jomtdisp~ace:men,ts d, To checkdIDe selution f(!lf correcmess, :substrtl!J~e the naarerieal values of the jo.~nt displacements d back irJ~o the stiflh.es:8relatWmsmpsholilld. besatis;fied. Note 'fuat joint translatwoEis are censideredpositive when in thepositi~e directions of the global X and Y axea.and joi rn{rotations are ooMldemd.pw;itiw;; when eeanter c~ockwise .

5. Compute member end dis.p·~acemela and end force~. and suppost reactions. for each member of the strucmre, cmryout th.e foUolMn,g ste,ps~

(59')

a. Obtain member end di:splacemet1!m in the gldbal c>oordina.tesys~em. v, from the j~Wll!t OJ sp~a0emern~s, dl.us:~ng the .memlOOir eode numbers,

b, Frorn fhe memlbermms(ormM~()n matrix T (Eq.(5:'18)). and. determine the

member End dilspJ;acements~n the I.()ca~. ooordi11late system~ WI~ using the

ttam;forma~lon Re1ationslUp, u= 'I v.

c. From the member local stiffaess matrix k (Eq.(S.3» and ~~al fixed~n.dfon~;e vector ,Ot (Eq_(S.4 »)~ t~um. c~dcublikllihe me-moor end. forces In the ]oca~ eoordiaate sy.stem~ Q, JJSiEg d~e mffiIess ndationShip' Q"" kn + Qr.

d. Determine the memoor end forces i11l~he gb)!bal ceordinase system • F, us~ng the rransfoanaticn relationmip' F = l'TQ .

e. If' rhemember is attached t'O a support j'omm. then use 'the memher code mimbers \$0 store tllilepe-rt~nent ,eieDl1.emlts of F ~n tl1e~r proper posiHons wn the support reaction vector R

6 Check the calcu~:M~o:n of member end forces and support reaeticas by appl yi~gthe ,equ~tiblT~um. ,eqw!li(ll.ns ( V~ ~I~, Lfy= o~ ana I)\.1: = 10) to the free 'body of the mtlK: stttlcrure. litlle ,ca1cIJJatm(Jru. have been ,earned out oorrec~IDy ,then the equiHbnwn equatious should. be satisfied,

Ins~ead of following steps, 5(c) and (d) Df thispR)cedlUm;:'.~helnember end ferces al~ematively cam. be obta~ned b~ first ,ca].clllaJHngtheg1()ba~ forees P us~ngtbeg~ohal s~iffness .re]liIlti.~rWlIip F ~ Ky +f'f ~ mdthen. ,evah.ting the; local farces Q from the transformation relationship Q "" T f. It slouJd a~so be nosed thart i~ ~susuruly not necessaryto determ~ne tmJe global end ['Owes fbr aU the members 'of the structare , beeaese such forces are net used for design fJWll'OSe5" Ho'weveli" • F vecters for the members that Me &tlached to SUipiPORs aw al.ways: eval.ooted. so that they em. be used to f(l:l1]]}. the SU,i!lO" reaction vector R.

FORM DETERMINATION OF'

_- ?~

LAMELLA DOME

IMPORTANCE

Analysis" Design, Erection, Laycut etc, of any space strucjure depends on the determination of its form. of geometry i.e, coordinates of nodes, length of members. ccnnectivity of member, etc. accurately.

MATHEMATICAL MODEL

Here) we would be analyzed the complete Dome and the retractable Dome. For which we . consider the span of 30 m and a Rise of 1.5 m In the development of geometry of Closed Lamella dome and retractable lamella dome, the l8ilitude and longitude range is fixed as shoV!lil1. ~n Frug.6.1. Now ,assuming a interva~. ie. in. Z direction as 21.625° and ,P interval in X - Y plane, as 30° at j oints are staggered. Reier Dra wing No_ I; the latitudes are ranging from 0 to 86 . .:5° ... so i~ can be conviently to amlilyzed Retractable dome.

Now the Complete Dome is divided in, 5 Rings (4 parts) run the lati rude directwon (1.direction) and 12 parts in the longimdinal direction. Which shows that, fO'l the purpose of analysts, the, Dome frameis composed of 84 jo~nts and 228; members each with constant cross-sectional properties a~ong its length, The coordinate of 1/12di pan of dome is calenlatedas gjven. in the table below,

In 'the case of Re,tracmble; Dome, we will 3l1laiy.ze pa'['t~.cu].a.r 11 I 21b part of dome. from which we win design the complete structure. Thautpart w.iIl deploy (open) mn the zdirection by 1/3 parton 3-hinge; Cilllm'WC,t~.on. This, upper 113 part VlBill rest on the below part. Thus it wiU Dav,e lI2joints and 23, members mcludin.g 3-hinge connection .. For the purpose of ana~.ysis the: structw"e M.n rotate &1150• 30°. 4i5C1 and. 540• At 54(1. the structure win ex.act~y in the vertical cond~tion. which bas maximum axiat forces and. stresses. now the figure wHI give tbe clear idea . This struenere will open by cable 'wire a~ the joint as shown mn the figure, The coordinate ofthe slt:ruCrure is calculated as shownin table below.

(61)

I ,I :11 'I

t

NUMBERING SCHEME

To, prepare t~le input data, we M(V,e' given proper numbers to nodes and members containing jin .m!. The numbering scheme adopted for the Closed Lamella dome as discussed wn the previous chapter is given in Fig. 6.2 ..

COORDINATE GENERATION

The parametric equanons of sphere are:

X = R x cos p, x sin a Y=R x s.infJ x sinu Z = R« cos a

COOR_DINATE OF lind! PART OF DOME IN CLOSED POSITION

~oillt '0: D, X Y Z
Al 90 15; 14.489 3,.88 'I)
A2 14,489' -3.88 e
Bl 68,.375 IS 13.469' 3.,61 5..53
I 13.469 I 0 553
B2 0 I
I
BJ '15 I 13.469 ·3.61 5.53
I Cl 46:75 15 W55 2.821 10.217
(;2' 10 .. 55 -2.82:7 10.277 I
, I
Dl 25.125 ~5 6.152 1.648 13.58
D1 0 6.152 0 13.58
I
.D2 I 15 6.152 -1.648 13.58
El 3.5 II 15 , 0.,g,85 0.237 ~4 ... 9r;
E2 I 0.885 ,·0 . .237 W.4.97 ~151rnl@>r1I]l ~1~G1\n n\'jrY]l51M nr;]fU]1 U~~lo.llWJ . U\!J~U8\!JlWJciJ~8\!J~

,COORnINATE OF' 1/12'111 PART OF .DOm

o 16.14

IN 19 Q RiOTATlON POSITION

Cl I 11.83'9 2.827
cz J 1..839 ~2.a27
D1 8A.M3 t648
DI 8.4143 0
.D~2 8.4,l43 ~t648
£1 3.,615; 0.237
£2 3.615 -0.237 10.,86

iJoimt X Y Z

A,l. 14.48,9 3.8,8 0

Al 14.48,9 -3.88 0

.Bl 13.469 I 3.61. :5.53

I .B2 0.469' 0 553

83 '~3.469 ~3.61 5.5] 10 .. 86

..42 14.489

y Z
3.8,8 0
-3..1:18 (II
3J.n 5.53
0 5.53
-1.6W 5 .. 53
2.827' i 1.1 I .BI W.3A69'

1.5.161

112 13.469

'~ :5.167

B:J' 13.46:9
.
Cl B.237
C2 ~3.231'
Dl 11.021
D'I 11.021
D'2' 11.021. -2Jt27 1 t.r

1.648 I :~ 6" 14

l1.859,1

-1.'1)4,8, 16.14

I1J~,S9

El I. B2 0.237 19 .. 965 :£2 I '1 .. 132 -0.231' 19'.'96,5

.IN 45'''R',QTAITION POSI,TION

IN ,54.,lJ62j'O ROTA TlON POSITION

~oiDt X I Y Z ~O~Dt' X Y Z
Al 14.489 3.8,8 0 A.l 1.4.489 3.8:8 0
A2 141.4:89 ~3.88 [) All' l4,.489' -3.88, 0
Bl 13,469 3.61, 553 Bl B.469 3.161 .5_53
B2 BA69' 0 5.53 B2 13.46,9 o \$.,53
IJ3 1J..469 I ~3.,61 5.53 BJ' 13.469 -3,.61 .5.53
Cl 14.6052 2,827 1.0.9-16 CI 15.473 2.82-1 10.731
cr 14.652 -2.827 IOJn6 C2 I 15.41.3 -2.,821 10.731
Dl 13.7'95 I 1.648 16.414 DJ 15.473 1.648 1,6.236
,
.BI 13.795 0 1.6.414, D'l 15.413 I 0 16.236
D2 13 .. 795 -1,648, H5'.414 I D~ 15.473 -1.,648 16.236
E1 1 i.on I 0.237 21.1.0.5 I .E1 13.469 .0 .. 237 2t304
£2 1 ~.O21 ~.237 21.105 I E:1' B.469' -0.237 2t304 (64)

1

L~

y

L~

~t::( SUPORT Fz bJ;t. M'!l' Mz <!> smORT F:[F, Fz C1BXElDmOO

3D VIEW

Figure: 15° Rotation Condition

(6S}

1

PLAN

L~'

x

L~SlJPOltr Ft 1ttih My Mz <!> SmOll::r Px Fl Fz .6.FD::ED JOINT

Figure: 30° Rotation Condition

(66)

(67}

Figure: 450 Rotation Condition

{68)

Figure: 54°' Rotation Condition