Structural

Engineering

ITEEATIVE GRAVITY LOAD HETHOD FOR P-DEIXA ANALYSfS

by Bryan SraEford S m i t h and

Regina G a i o r t i

June 1988
Structural Engineering S e r i e s Reporr No. 8 8 4

Department of Civil Engineering

: and Applied Mechanics
3

8- 4 McGiII University
Montreal
-s R q h r q R q m r t No. M-4
 authors wish to r e m d their appreciation o f t h slprport givm to

this by the Na- Sckres ad w-itq F(esearch Cumzfl of

GuXda.
 .......................................................
m. ii
................................................ iii
mOF m.. ............................................. iv

E x OF nGURES................................................. V

ZJST OF m..... ............................................. v

 Gravity I m d i q mlied to Ceflected Shaps ..................
G m v i Q L o % h q Applied to F i c t i t i m s C o l m ................

a33 F1emxa.l .......................................

Pie Shear ..........................................
T?ie Wormed Shape of a L m - ard U m - R i S e Rigid F Y a m ....

m e T k Z t S h e a r W a l l 5 ...............................
~

T k a lkzt Rigid Elam - S ..............................

?he Shear Wall ........................................
Pte Rigid Elmas Model .......................................

I . m f l d o n s by -lying m v i t y lPads to k f l d Skip of

S k 3 r M1 .................................................. 13

2 . by mlying Gravity Lmds to &fleck& shsp of

shear Wau .................................................. 14

3 . I l e f l A by App1yirqC;raVity Lm& to sen- Shape of

Rigid ................................................. 15

4 . by Applying Wity W to S h a p of

Rigid Frame ................................................. 16

In heavily clad 1- --rise -, the PWta effects
aD2 4 1 emqh 'a be neq1eZtd. w,
with the tc@r& toGmrds
taller ard l i g b l y clad
the PDalta effects
Ww,having qzeate~lateral
mre signifi-. In - -,
flexibiliky,

the P

Delta effects are large etmqh to

m n k r sizes. In an extrew - recflire an increase in the

of a very M l e
large gravity loadirq, the *Delta effects d d , if m- a
desiqned

with a
for,

be severe to initiate a l l a p s e . Tus, in the &sign of any

r , it is iiqx?bnt to assess whet3er these m-
order effects ane siqnificant m d , if so, to ammnt for thw in the
a r d y s i s and &sign.
As an alternative, the authors, in this report, a sinple ard

6Y uSbq a fram zrdysis p q m m , t h e

irvJamental lateral load rpethDd

-
ram realistic iterative m&mi of Pmta analysis for

c=an be -,
regults

with
w.

a s by the itexative
1-
-btional effcat.
After a first order horizDntal lwd analysis of Me structure using a
fmme analysis pmqmm, the gravity loads are applied to the unloaded

stmcmre, specified as df&L- by W e first mder d i s p l a m , A,,

to obtain an in displamm-k, k , , F i g . la. The gravity
loads are then -lid to the smwbxe spzified as deflected tiy this
to &#in amthfcc m t in dkpl-, qZ, F i g . lb.

me p- is -ted until the &ti& in

displarmmwt is negligible, F i g . LC. ?be final d e f l % c t i ~ , incluling

the PMta affect, at storey i, q, is the sum o f the first o m

displaosmarh ard aLl the 6 t i a - d d k @ h m m d s ,

In practFca, the mthd can be simplified by adding a full-hiqht,

a x i a l l y rigid fiditiars mlunm with a flexural stif* @Val& to
 A; = A f dfl f 6F2 + Si3+ .*.

Figure 1 : Gravity h d h g Applied to Deflected Shape

4
z e m , & e i n g i t to the s m by axially rigid l i n k , F i g .
2. Then the total qravity 1mclb-q at each f l m r is applied only to
rnrrespodirq levels of the de
n - fictitims m l m in the sm-e

iterative n*anner as W e . In t h i s way, only t k mrdinates

f i c t i t i m s m l m . a d not the e n t k struchnre, r e d to be

of t h ~

altered.

Rigid w ,braced -, s-e shaar walls and mupled valls are

u m k z s of a family of cantilever stmdmxs which, subjected to

la- loam, w i t h different d h t i m s of f l W a d

shear n d e responses. lh flexlaaL -, P i g . 3, is a d i n a t i o n

of overall berdirg of the omplete stmzhm a d ~ v i d u a lflexrrre of

the vertical mabra's. We skar resprmse, Fig. 4 , js a f-

acticn.

A s*le shear wall deform in a plrsly flexural loode. A bmced f m n ~

2dqt-s a @rhxcily flexural with d y a shear
f r m mdring. A 1a.1- or medimrise rigid f-, on the other h r d .
a primarily sl-tea~ o f def- ti^ w i t h m l y a mall

bzidrg m,
Fig. 5. Hrklwer, in a tall rigid frarae with a large

heiqht-tp-uidth ratio, the bendirg tmre signifirant.

The mupled wall is an irrtermadiate struchrre udeqoing a carbination

of flexlaal ard shear behaviour. Tm w e r e chosen to

f i l b t e the iterati= gravity load methad of P-Mta a n a l y s k . At

one extrem, a w l e shear w a l l smmcbre defmdsq in a flermraL &

oonfiguratia, ard, at another e x t ~ ~a ~r i g, i d frmne structlae
5
 a it + BF1 + Bt2 + di3 + .
*.

Figure 2 : Gravity Loading Applied t o Fictitious Colum.

Figure 3 : The PtexuraI Response

Figure 4 : The Shear Response

 -- B e n i l n g Component

-- - Shear Ca7onenr

Xrsulcax

Figure 5 : 'The Ilefonned Shape o t a la- and 14cdiu-Hise R i g i d Frame

 --
Graviep Load
11-90 RNfn
2
990 X S l s t o r s y

- -
Figurc 6 : Thc T e s t Shear H a l I Structure
Figure 7 : The Tczt Kip,id Frame S t r u c t u r e

1I
Figure 8 : The Shear Wall Model

Figure 9 : The Rigid Frame Model

O N -
m nm- z
n
---
\cr m
TOR EY TOTAL DEVIATIOH
I tlDnEN'T FROM
APPLIEU G . L. ITWTIVE
- (Lst ITER.) HKl'1100

70
19
18
I1
16
15
14
13
12
II
10
9
0
7
6
5
4
3
2
I
-
Table 4 ; bhlonerrs by Anplying G r a v i t y Loads t o Deflected Shape o f Higid Pclme
Wles 3 ard 4 -
It t% ~eflectiward mclcnts, m s p z t i v e l y , for
the rhid fzmbz &hind by applying t
! q a v i t y loads to t!! deformed

s h p iteratively. With m x h a c k v i a t i m frrrm the iterative 3etfioci

o f 0.103% ard 0.014% for the defieztions ard rmwnts, w i v e l y , it

can be said that the method wark equally w e l l for rigid frames a i d for

shear walls. L? f a c t , grrm the very mall deviatims -, it

be &uded that the results cbtausEd by -lying the qravity loads to
the d e f l e s d simp as p w are equivalent to tr,ose rhtained by the

i t e r a t i v e metkcd.

Ey applying the w - t y loads to the deflecCz shape of the structure,

the aM-Lysis was execam more r a p i d l y t k n by the i t e r a t i v e new,
since all that it was n- to do rras to M e the mrdina- of
the s k n c b x e h the w t e r kpk. Zhat is, no h a d calcllations, as

psr the iterative m z t h d , were p e r f o n e l f o r Bach iteratian.

analysis by the pmposed wti-4 ws d a m i

n -third the time it tmk
to perfom the iterative analysis. Moraner, the m e w m he further
simplified, in the mse of a larger strucbue, by ad- a Ml-height

Eictitiars w l m as desaibed p r w i a s l y .
1. Maim, P. F . , me Desim of Steel Waiw3luws, &r&im
of S t e d Cot=tmctim, W i l l m e , Ontario, Om*. 4, m.4-21,
1974.

2. Gaiotti, R . and W i p , A., A S W of P-Wta Methcds nf

fnr T a l l mild-, projer;t report presented to Mill University,
m, m, in k c . 1385, in p r t i a l fulfil-t of the
reEprirsments for the &qree of Sacrelor of c i v i l B-qimerirq.