7/26/2019 Masonry infill analysis
1/25
INFILLS
IN
SEISMIC
RESISTANT BUILDING
3
By Vitelmo Bertero,
1
F. ASCE and Steven Bro kken
2
ABSTRACT:
This paper summarizes studies in which the effects of masonry and
lightweight concrete infills onR/Cmoment existing frame buildings were stud
ied experimentally and analytically. The experimental investigation consisted
of a series of quasi-static cyclic and monotonic load tests on
1/3-scale
models
of the lower 3-1/2 stories of an 11 story-three bay reinforced concrete frame
infilled in the outer two bays. Different panel material and reinforcement com
binations were tested. For reasons of economy, ease of construction, favorable
mechanical properties, and efficiency of different types of masonry infill, it was
concluded that the most promising panel configuration consisted of solid brick
laid in mortar reinforced with two mats of welded wire fabric, one bonded to
each side of the wall in a layer of cement stucco (mortar). The implications of
these experimentally obtained results are analyzed by investigating how the
infills affect the dynamic response of
R/C
moment resisting frame buildings,
as well as considering the effect of these implications on design of new build
ings, and retrofitting of existing buildings located in regions with differing seis
mic risk.
INTRODUCTION
Recognit ion that
the
dynamic characteristics
of the
bare basic struc
tural system
are
significantly changed
by the
incorpo rat ion
of
infills
has
led to the formulation of two building design phi losophies in seismic
resistant design.
One
philosophy requires that
the
infills
be
effectively
isolated structurally from
the
structural system
so
that their structural
effects cancorrectlybeneglected. The second considers theinfillsto be
tightly placed,
and,
therefore, their interaction with
the
structural sys
tem
to
resist
the
effect
of all
k inds
of
excitations sh ou ld
be
proper ly con
sidered
in the
design, deta i l ing,
and
construction.
The authors bel ieve that
the
second philosophy offers more concep
tualand practical advantages, particularlyif the basic structural system
is moment resisting frame. This
is
because
a
main principle
for
seismic-
resis tan t des ign i s : A void unnecessary ma sses ,
and, If a
m a s s
is
nec
essary,use itstructurally to resist seismic effects (3).T h us if wallsand
partit ions
are
needed
and the
economical material
is
m a s o n ry
or
con
crete, attempts should
be
m a d e
to use
these infills
as
structural ele
men t s .Thep rope ruse of infill elementscan be of great practical value
This paper
is
dedicated
to Dr.
Bruno Thurliman
on his
60th anniversary
as a
tributeto histeachingandresearchin theareaofinelastic behavior.
'Prof, ofCiv. Engrg., Univ.of California, Berkeley,Calif.
2
Design Engr., URS/JohnA.B lume & A ssoc, Engrs.
Note.Discussion open until November
1, 1983. To
extend
the
closing date
one month,
a
written request m ust be filed with
the
ASCE Manager
of
Technical
and Professional Publications.Them anuscriptfor this paperwassubm ittedfor
review
and
possible publication
on
February 18, 1982. T his paper
is
part
of the
Journal
of
Structural Engineering,
Vol. 109, No. 6,
June,
1983.
ASCE, ISSN
0733-9445/83/0006-1337/$01.00. Paper No. 18059.
1337
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromasc
elibrary.orgbyIndianInstitute
ofTechnology,Delhion04/2
6/16.CopyrightASCE.Forpe
rsonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
2/25
in strengthening and stiffening the usually very f lexible moment resist
ing bare frame. The connection details between infi l l and frame are also
simplified, bu t beca use , of the inte rac ting effects, the infills ca n be s ub
jected to deformations and stress beyond their elastic resistance and pro
duce semibr i t t le types of fai lure when unreinforced masonry and con
crete panels are used. This is not a ser ious disadvantage with respect to
isolated panels, however, because i t is recommended that the infi l l panels
contain adequate reinforcement even in this case (10).
When the panel inf i l ls are t ightly placed in the frame, the problem of
avoiding prem ature failure raises the qu est ions : (1) H ow sho uld these
panels be reinforced; and (2) how should they be connected to their sur
roundings? A comprehensive review of the l i terature avai lable on these
problems to 1974 (6) revealed the need for further research, and so an
integrated experimental investigation was init iated in 1974 at the Uni
versity of California, Berkeley.
Resul ts obtained to 1978 have been repor ted in Refs . 2 , 5 , and 6 . A
second series of experiments on a 3-1/2 story and 1-1/2 bay subassem-
blage of an 11-story apartment building (Fig. 1) have been recently com
pleted. Eighteen tests were conducted to investigate the relative perfor
mance of various types of infi l l ing materials and construction techniques
(4). The effects of infills on the seismic resistant
R/C
cons t ruc t ion were
studied analytically and have been reported in detail (4).
T his pa pe r is pre se nte d w ith the following objectives: (1) T o su m
marize the experimental investigation and the results obtained; (2) to
evaluate these results and to assess the practical use of infills in sites
located in regions with differing seismic risk; (3) to formulate recom
mendat ions for the design of new seismic-res is tant bui ld ings with in
filled frame structural systems, and for the retrofitting of existing build
ings having R/C moment res is t ing f rames as a s t ructural system.
DESCRIPTION OF EXPERIMENTAL INVESTIGATION AND RESULTS
Specimens.The specimens were s imilar to those used in the f i r s t
series of stud ies (5,6) a nd are illustra ted in Fig. 1 an d 2. Four different
types of infi l ls were used. Two infil ls consisted of hollow-unit masonry:
clay (Fig.
2(b))
an d co ncrete block. T he third type of m as on ry infil ls w ere
split brick with exterior welded wire fabric (WWF) reinforcement (Fig.
3). The wires of the WWF mat were spliced to dowels lef t anchored in
the confined regions of the bounding frame members (Fig. 3) so that the
panel was f i rmly at tached to the bounding f rame. The four th type of
infi l l was l ightweight concrete panels.
REPAIR, STRENGTHENING AND RETROFITTING OF SPECIMENS
R epair M ethod.A fter an infil led f rame loading prog ram w as com
pleted, i t was found that severe panel damage was general ly conf ined
to one level and so panel replacement was necessary at only one level .
The damaged panel was removed , wi th care taken to r e ta in the r e in
forcing steel (or WWF) protruding from the frame which had.been cast
in p lace for panel reinforcement anchorage. Cracks in the beams and
columns were repaired by epoxy injection. If crushing of concrete had
occurred, al l loose concrete was removed from the frame members, leav-
1338
J Struct Eng 1983 109(6): 1337 1361
Downloadedfromasce
library.orgbyIndianInstituteofTechnology,Delhion04/26
/16.CopyrightASCE.Forper
sonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
3/25
( e )
COLUMNS IB X18
BEAMS = 12 X 24
); and
{b)
Reinforced Hollow Brick Infill
1339
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromas
celibrary.orgbyIndianInstitut
eofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
4/25
]
FIG. 3.Third Type of Test Specimens: Solid Brick Infills Reinforced with WWF
at Each Face
Strengthening Method.
During some tests the spiral transverse steel
was observed to fracture in critical inelastic regions of the columns in
the first story, causing immediate shear failure at that location in the
column. A ny type of repair becam e difficult an d re nd ere d this story level
useless in subsequent tes t ing. I t was , therefore, decided to s t rengthen
this story so that panels in other stories could be tested. Strengthening
was achieved by placing a rather substantial amount of reinforcing steel
in the pane l op enin g a nd casting this story solid (5 in. thick) in c oncre te.
Retrofitting Method.To retrofit inf i l l panels into an existing bare
frame, this frame was dril led to attach an anchorage system for the panel
reinforcement. This anchorage system consisted of steel plates attached
to the beam s with anch or bolts at 8 in. O .C. (200 m m ) an d to the col
umns with bol ts a t 4 in . O.C. (100 mm). Wedge anchors were used in
the columns and the third-story beams. The f irst- and second-story beams
were dr i l led completely through, threaded rods were inser ted to secure,
by means of nuts, on both sides of the beam, anchorage plates for welded
wire fabric reinforcement anchorage (see Fig. 4).
TESTING OF SPECIMENS
The models were loaded as shown in Fig . 2(a) (6) . The rat io between
the lateral force and corresponding over turning moment was calculated
by a dyna m ic elastic analysis of the entire fram e. A nalyse s we re co n
ducted on both the bare frame and the infi l led frame. Overturning mo
ment from stories above the subassemblage, as calculated from analysis,
was applied automatically using a preset transfer between the lateral and
axial jacks through a servocontrol system.
In the first series of studies (5,6), four tests Were conducted. In the
second series, a total of 18 tests we re perform ed. M ain results are sum
m arized in T able 1 an d som e typical load-deform ation relatio nsh ips for
the specimens tested are i l lustrated in Figs. 5-10.
1340
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromascelibrary.orgbyIndianInstitute
ofTechnology,Delhion04/26/16.CopyrightASCE.Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
5/25
b) e)
FIG. 4.Details of WWF Reinforced Infi l l Used to Retrofit ExistingR/C Bare Frame:
(a) Frame-Panel Anchorage System;(b) Deta i l B (See Fig . (a)) , Threaded Rod P ro
vid ing Posi t ive Anchorage Bol t ing Complete ly through Beam; (c) Deta i l C (See
Fig. (a)) , Wedge Anchor Fastening WWF to Column;
(d)
Detail A (See Fig. (a)),
Wedge Anchor Fastening WWF to Beam; and (e) Sect ion X-X (See Fig . (d))
E V A L U A T I O N O F T E S T R E S U L T S A N D T H E I M P L IC A T I O N S O N D E S I G N
A N D R E T R O F IT T I N G O F S E I S M I C - R E S I S T A N T B U I L D I N G S
Infills no t on ly m odify th e available (supp lied) stiffness, str en gt h
(yielding and ul t imate) , damping, hysteret ic behavior and deformation
1341
J Struct Eng 1983 109(6): 1337-1361
Downloadedfromasce
library.orgbyIndianInstitute
ofTechnology,Delhion04/26
/16.CopyrightASCE.Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
6/25
T A i L E 1.Summary o f Specime ns Tested and Their
Test
specimen
number
D
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Model
number
(2)
1
1 ,R1
3
2
1, R2
1, R3
2 , R l
3 , R 1
3, R 2
3,
R 3
3,
R 4
1,R4
2, R2
2 , R3
4
5
5, Rl
4 , R l
Loading
program
(3)
Monotonic
Cyclic
Monotonic
Cyclic
Monotonic
Cyclic
Cyclic
Cyclic
Cyclic
Cyclic
Monotonic
Monotonic
Cyclic
Monotonic
Cyclic
Cyclic
Monotonic
Cyclic
First-story panel
(4)
Clay brick p = 0%
Clay brick p = 0%
Con crete brick p = 0.6%
Clay brick p = 0.6%
6 in . R /C
6 in . R /C
Clay brick p = 0.15%
Con crete brick p = 0.6%
No panel
6 in . R /C
6 in . R /C
6 in . R /C
6 i n . R / C
6 in . R /C
No panel
Split brick 90 = 0.4%
Split brick 90 = 0.4%
Split brick 45 = 0.4%
Second-story panel
(5)
Clay brick p = 0.6%
Clay brick p = 0.6%
LWC p = 0.6%
Clay brick p = 0.6%
Clay brick p = 0.6%
Clay brick p = 0.6%
Clay brick p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
Clay brick p = 0.15%
Clay brick p = 0.6%
Clay brick p = 0.6%
No panel
Split brick 90 = 0.4%
Split brick 90 = 0.4%
Split brick 45 = 0.4%
a
l kip = 4.45 kN.
Factored by 2.0 in. /2 .5 in.
Note: +1 K/ in . = 0 .175 kN /m m .
capacity of the building structure, but these changes also introduce mod
ifications in the demands of these same response parameters to any given
ear thquake ground motion.
The addition of infi l ls br ings an increase in the building mass. This
increase in mass has two main effects : (1) The reactive mass, M, is in
creased; an d (2) the p erio d,
T,
of the structure is increased. Furthermore,
while the addition of the infills by virtue of its mass increases the period
T, i t also introduces an increase in st iffness and thus decreases the T.
EFFECTS OF INFILL ON THE SUPPLIED LATERAL STIFFNESS, K
AND ON THE PERIOD, T
The lateral st iffness of the subassemblage tested, based on the inter-
story drift, is given in T able 1. B ecause th e initial tang en tial stiffness
deteriorates very quickly at the service lateral load, an effective inters-
tory lateral stiffness, K*, at service load level has been evaluated and
introduced. In interpreting the signif icance of these values regarding the
lateral stiffness of the prototype frame, K?, it has to be considered that
the interstory lateral stiffness of the model frame Kf can be considered
as twice that measured in the tests of the subassemblage and that the
K1
is equal to the
Kf
multiplied by the length scale L, i .e. ,
1342
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromasc
elibrary.orgbyIndianInstitute
ofTechnology,Delhion04/26/16.CopyrightASCE.Forpersonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
7/25
Maximum Resistance and interstory Lateral Stiffness
-~-~ ...
Third-story panel
(6)
Clay brick p = 0.6%
Clay brick p = 0.6%
LWC p = 0.6%
Clay brick p = 0.6%
Clay brick p = 0.6%
RC p = 0.6%
Clay brick p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
LWC p = 0.6%
RC p = 0.6%
Clay brick p = 0.6%
RC p = 0.6%
No panel
Split brick 90 = 0.4%
Split brick 90 = 0.4%
Split brick 45 = 0.4%
Max
load
H. in
thousands
of
pounds
3
(7)
55.2
35.3
67.9
54.5
68.6
80.0
39.2
46.7
27.4
92.7
100.0
63.2
76.0
83.0
12.5
70.7 56.6
b
61.3 49.0
b
57.3 45.8
b
Location
of failure
(8)
First story
First story
First story
First story
Third story
Second story
First story
First story
First story
Second story
Second story
Second story
Third story
Second story
To ta l mechan ism
First story
First story
Combined
mechan ism
Maximum
Initial
tangent
(K/ln.)+
(9)
1,090
1,090
585
920
195
271
780
725
103
990
1,500
494
178
203
65
1,250
834
960
Interstory Lateral Stiffness
Effective
at
service
K? (K/ln.)+
(10)
206
236
212
187
195
238
195
250
60
358
409
167
176
210
35
292 (234)
b
118 (94)
b
203 (162)
b
Relative
Kf/Klf
(11)
5.89
6.74
6.06
5.34
5.57
6.80
5.57
7.14
1.71
10.23
11.69
4.77
5.03
6.00
1.00
8.34 (6.69)
b
3.37 (2.69)
b
5.80 (4.63)
b
K? = K?L
S
= 2KfL, (1)
T his inters tory lateral stiffness K? will be u se d as rep res en tativ e of the
lateral stiffness of the prototype.
Lateral Stiffness of Infilled
Frames versus Bare Frame.Compar ing
the values given in Table 1, i t can be seen that considering average of
Kf for infills of the same type, the smallest of all lateral stiffness of in
filled frame, (Kf)^, (obtained for the solid brick panels reinforced with
w elded w ire fabric) was 4.66 that of the bare frame (Kf)^ . T he largest
of all the (K^)ifcorrespo nding to the re inforced l ightw eight concrete w as
10.94 times the (Kf)fc/ and in the average the (Kf),/ was 6.31 times the
(Kf)v-
Effect of (K^)if on Period, T, of Building.
Although in genera l the
addit ion of an infi l l decreases the period,
T,
the specific amount of de
crease depends upon how the tota l mass of the bui lding, M, changes
relative to the stiffness with the addition of infill . Depending on the
assumpt ion of how the M changes , the di fferent resul ts are summarized
in Table 2, where two bounds regarding the changes in M have been
evaluated: Upperbound,all 11 fram es of bu ildin gs of Fig. 1 are infilled;
and lowerbound,only 4 of the 11 frames are infilled. F or ea ch of th ese
two bou nd s tw o cases w ere c ons idered, one -in w hich th e
M
is assumed
the same as when the s t ructure i s cons idered as a bare f rame, and the
1343
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromas
celibrary.orgbyIndianInstituteofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
8/25
TABLE 2.Effects of Infills on the Period,T
if
, of the Prototype Building
Degree of
changes
and type
of infill
(D
Lowest (so l id
b r i ck w i t h
W W F )
Average (hol
l ow
masonry)
H ighest ( l igh t
we igh t
concrete)
A L L 1
UP P E R B O UND
FRAMES ARE INFILLED
SAME MASS
(2)
0.46
0.40
0.30
n
in Seconds,
f o r 7 V ,
in Seconds
1.30
(3)
0.60
0.52
0.39
1.01
(4)
0.46
0.40
0.30
Infill
Adds
Mass
(5)
0.49
0.42
0.32
L O W E R B O UND
ON LY 4 OF 11 FRAME S ARE
SAME MASS
r
r
-/r
(6)
0.65
0.58
0.47
in Seconds,
for TV,
in Seconds
1.30
(7)
0.84
0.75
0.61
1.01
(8)
0.66
0.59
0.47
INFILLED
Infill
Adds
Mass
(9)
0.66
0.59
0.48
Tyis the period of the prototype building with bare frame structure.
other in w hich th e infill add s m as s. A nalysis of the resu lts obtaine d re
veals that any of the infill, even the softest, will produce significant change
in the
T
of the bui lding. Furthermore, the effect of the added mass due
to infills on the
T,
is very small and can be neglected.
Per iod of the Prototype B ui ld ing , T .To hav e the values of T in sec
for the prototype building, i t is necessary to est imate i ts period where
a bare frame structure building is used, T
bf
. T h i s Ty can be analytically
computed or est imated from the experimental results . The analytical ly
computed value was 1.30 sec (6). Using the experimental s t iffness of the
subassemblage and applying Eq. 1 , cons ider ing as the prototype mass
the estimated one of 23,144 kips (102,945 kN), the T
bf
resul ts to be equ al
to 1.01 sec. Using these two values as an est imation of the period of the
bare frame building, i t is possible to compute the period for the infi l led
frame building,
T
if
.
These values are given in Table 2.
EFFECTS OF INFILL ON THE SUPPLIED STRENGTH TO THE BUILDING
These effects are again evaluated on the basis of the results obtained
in the tes t specimens , making di fferent assumpt ions regarding the num
ber of frames that are infi l led in the real building. The evaluation of the
strength is based on the est imation of the base shear s trength,
V
n
, that
the model of the bui lding could have res is ted. This es t imat ion in turn
will be based on the measured lateral resis tance of the specimen tested,
(V)
s
,w hich is equ al to the m axim um lateral force
H
plotted in the dia
grams of Figs. 5-9 and summarized in Table 1.
Base Shear St rength of Bare Frame,
{V)
hl
.
Considering tha t the
maximum measured la tera l res is tance H of the specimen in Tes t 15 and
12.5 kips (55.6 kN ), the total lateral resistan ce of the m od el of the co m
plete building, i f the only resis t ing structural element were the 11 bare
frames, would amount to 275 kips (1,224 kN).
1344
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromasc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
9/25
TABLE 3.Effects of Infills on Supplied Maximum Strength of the Prototype
Building,(V)
if
Typa of
Infill and
reinforcement
D
Unreinforced
masonry
Reinforced
hollow
masonry
Solid brick
reinforced
withWWF
Reinforced
lightweight
concrete
p . as
a per
centage)
(2)
0
0.15
0.60
0.40
0.60
Upper B ound
All 11 Frames are Infilled
V ) h
In
thousand
pounds
(3)
Lower 35,3
Lower 39.0
Lowest 46.7
Average 65.0
Highest 83.0
Lowest 57.3
A verage 63.1
Highest 70.7
Lower 92.7
Higher 100.0
V ) ? / /
V ) f r
(4)
2.82
3.14
3.74
5.20
6.64
4.58
5.05
5.65
7.42
8.00
7/26/2019 Masonry infill analysis
10/25
4 0
2 0
C L
x 0
= VY
H
y x 2]L
S
2
(2)
W hen o nly 4 of the 11 frames are infil led, the d eterm ina tion of Vre
quires analysis of the load-de form ation relat io nsh ip of the infi lled frames
and that of the bare f rame, (Figs . 5-9) , and an assumpt ion regarding
the in-plane flexibility of the floor system (diaphragm). To simplify the
discuss ion, i t wi l l be assumed that the diaphragm is r igid and that no
tors ion is developed.
DISPLACEMENT, A ( IN)
(a)
DISPLACEMENT, A ( IN)
(b)
FIG.
7.Load-Deflection Relationship for Unreinforced Clay Brick Infill
1346
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
sc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
11/25
DISPLACEMENT, A (IN)
DISPLACEMENT, A (IN)
(a)
(b)
FIG.
8.Load-Deflection Relationship for Reinforced Concrete Block
Infill:
(a)
Monotonlc Test: Specimen 3;(b) Cyclic Test: Specimen 8
A s il lustrated in Fig. 10, th e infilled frame re ach es its pe ak ela stic
strength at a displacement ( interstory drif t ) somewhat smaller than the
one a t w hich the bare frame reaches it s m axim um la tera l s t ren gth . T hus
the elast ic st rength of the bui lding cannot be obtained adding the peak
stren gth of th e ba re frame to that of the infilled frame . F or each different
type of infi l l i t would be necessary to analyze the load-deformation of
the infil led frame together with that of the bare frame. From inspection
of the resul ts obtained, i t has been concluded that a lower bound of the
strength can be obtained by considering that when the infi l led frame
80
60
40
,
H
K
I
P
S
)
o
-J
-20
40
-60
_
1
L
^
+ H
. S P UT
BRIM.
p.0.*
y
iv
i i
/f^^V
f
1 1
y
Y
i i i i
DISPLACEMENT, &0NI
DISPLACEMENT, a (IN)
(b)
FIG. 9.Load-Deflection Relationship for WWF Reinforced Brick Infill
1347
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
scelibrary.orgbyIndianInstituteofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
12/25
H ( KIPS )
0 A y en?- | 2 3
DISPLACEMENT A
IM T
(IN)
INT
FIG.10.Lateral Load-lnterstory Drift Diagrams for Some Specimens Tested
reached i ts peak elas t ic s t reng th , the bare f rame had develope d equa l
to half of i ts maximum strength, i .e. , that the [(V
n
)ff]{h
if
)
ma>
. = 2 x [1/2
(V)l
f
] =
12.5 kips (55.6 kN ). A s sho w n in T able 3 , a l thou gh the un -
reinforced masonry infill resulted in the lowest lateral resistance, it still
w as 2.82 t imes the resistance of the bare frame w he n 11 frames w ere
infilled, an d 1.34 times w h e n o nly 4 of th e 11 frames w ere infilled. T he
largest increase in lateral resistance w as obtain ed for the reinforced l ight
weight concrete infi l l , amounting to on the average, 672 and 212 percent
increases, depe nd ing on w he the r 11 or only 4 of the frames w ere infilled.
ESTIMATION OF DEMANDS: EFFECTS OF CHANGE IN T
The dynamic response depends no t on ly on the dynamic charac ter
istics of the building
(T ,
,
V
n
a n d
\i),
but also on the dynamic charac
ter is tics of the g rou nd m otion s . A n easy wa y to obtain an idea of the
potential effects of the changes in
T
over the response is by analyzing
the response spectra of the cr i t ical ground motions. In doing so the fol
lowing two cases have to be dist inguished: l inear elastic and inelastic
response. Before discussing these two cases, i t is necessary to define the
following:
Mass, M, of the Building.Because the two main effects of the change
in mass are small for this particular building, i t will be assumed that the
mass is the same 23,144
k/g
(102,990 kN /g ) w he th er the structure of the
building is considered as bare frame or infilled frame.
Per iod, T, of the B are Frame B ui lding. T o i l lus t rate how the in i tia l
stiffness of the bare frame can affect the influence of infills, the two
following periods of the bare frame will be considered: the
Ty
est imated
1348
J Struct Eng 1983 109(6): 1337 1361
Downloadedfromasce
library.orgbyIndianInstituteofTechnology,Delhion04/26
/16.CopyrightASCE.Forpersonaluseonly;allrightsreserv
ed.
7/26/2019 Masonry infill analysis
13/25
T, PERIOD (SECS.I
FIG.
11.Smooth Linear Elastic Response Spectra for an Effective Peak Ground
Acceleration of 0.5g and a Damping Ratio = 5% after Newmark and Hall (8).
Illustration of Effects of Changes inTon Force and Deformation Demands
from test results equals 1.01 sec, and the one obtained analytical ly, i .e . ,
1.30 sec.
D a m p i n g R a t i o , .Although the addit ion of infi l ls may introduce
considerable change in , usually increasing i t for large deformations,
(values of = 12% hav e bee n m eas ure d) for s im plici ty 's sak e, the for
the infi l led frame building is assumed to be the same as for the bare
frame bui lding un de r s t rong gro un d m ot ion s , i .e . , = 5%.
Linear Elas t ic Response.A l inear e las t ic response spect ra as sug
gested by Newmark and Hall (8) , for a maximum effective peak accel
eration of 0.5
g
(Fig. 11), has been selected for discussion.
Effect of Changes in T on Se i smic Force Demands , V.Table 4
summarizes this effect . Because of the decrease in
T
ind uc ed by th e ef
fect of the infills from 1.30 to 0.39 sec (in the case of the largest de
crease), when al l the frames are infi l led the demands in design seismic
forces increase ab ou t 1 41% . Figure 11 i l lustrates th is increa se. For s im
plici ty i t is assumed that the total seismic force demand is direct ly given
by the f i rs t mode response, i .e . , the response of the s t ructure i s cons id
ered as that of a s ingle degree of freedom having the total mass M of
the bui lding and the per iods computed in Table 3. In the case that
T
bf
= 1.01 sec, the addition of infills changes this value to 0.40, 0.46, and
0.30 sec for the average, lowest, and highest decreases. This change causes
an increase in seismic force demands of 86%. Table 4 shows the est i
m ated increase w h en only 4 of the 11 frames are infi lled; the m in im um
increase is 56% . Inc rease s in seismic forces of the o rd er of 56% to 141%
are very significant and cannot be neglected. I t is clear that for the type
of ground mot ions represented in the se lected e las t ic response spect ra ,
the more flexible the bare frame, the larger the increase in the seismic
1349
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
sc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
14/25
TABLE 4.Increase In Linear Elastic Seismic Force Demands,V, Due to
Con
sideration of Infills as Structural Elements
D e g r e e s
of change
in T, and
type of infill
(D
Lowest (solid
brick with
WWF)
A verage (hol
low
masonry)
Highest (light
weight
concrete)
Note: Compa
T
v
, in
seconds
(2)
1.30/1.01
1.30/1.01
1.30/1.01
rison of V
Upper Bound
All 11 Frames are Infilled
T j . i n
seconds
(3)
0.60/0.46
0.52/0.40
0.39/0.30
? with seis
v?/v
(4)
2.41/1.86
2.41/1.86
2.41/1.86
Increase,
as a per
centage
(5)
141/86
141/86
141/86
Lower Bound
Only 4 of 11 Frames are Infilled
TjF,In
seconds
(6)
0.84/0.66
0.75/0.54
0.61/0.47
vim
(7)
1.56/1.57
1.76/1.86
2.41/1.86
inic force demands based on the building
Increase,
as a per
centage
(8)
5 6 / 5 7
76/86
141/86
>are frame
structure,
Vy,
and for same mass.
forces attracted by the addition of the infill .
Effect of Ch ang es in T on D eform at ion D em and s . Figu re 11 i llus
t ra tes how the maximum disp lacement decreases 82% when the
T
bf
of
1.30 sec is reduced by the addition of the infill to a
T,f
of 0.39 sec. Table
5 summ ar izes the decrease in d i sp lacement de m an ds . I t should be noted
that even when only four frames are infi l led, the decreases vary from
33% to 60%. These decreases in deformation are very significant and
have beneficial effects : The smaller the deformation the smaller the dam
age, ei ther to the s tructural or nonstructural components , and the smaller
the P-A effects , which are two of the main drawbacks in the use of just
bare moment res is t ing frame.
TABLE 5.-Decrease In Linear Elastic Displacement Demands,
slderatlon of Infills as Structural Elements
V
Due to Con-
Degrees
of
change
i n r ,
and type
of infills
(D
Lowest (so l id
br ick wi th
WWF)
Average
(ho l
l o w
masonry)
Highest ( l ight
we igh t
concrete)
T ^ , i n
seconds
(2)
1.30/1.01
1.30/1.01
1.30/1.01
Upper Bound
All 11 Frames are Infilled
Tjf, in
seconds
(3)
0 .60 /0 .46
0 .52 /0 .40
0 .39 /0 .30
8/B
(4)
0 .44 /0 .34
0 .34 /0 .24
0 .18 /0 .15
Decrease,
as a per
centage
(5)
5 6 /6 6
6 6 /76
8 2 /8 5
Lower Bound
Only 4 of 11 Frames are Infilled
T>,
in
seconds
(6)
0 .84 /0 .60
0 .75 /0 . 5 4
0 .61 /0 .47
(7)
0 .6 6 /0 . 6 7
0 .6 0 /0 . 4 9
0 .4 6 /0 . 4 0
Decrease,
as a per
centage
(8)
3 4 /3 3
4 0 / 5 1
5 4 / 6 0
Note: Comparison of 5 with the displacement demands based on the building bare frame
structure, S.
1350
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
sc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
15/25
Overal l Effect of Inf i l l s on Strength Demand and Strength Supply:
Intens i ty of M ot ion s tha t Inf i l led Fram e B ui lding C an R es is t Elas t i-
cally.B ased on the resul ts sum m arized in T ables 3 and 4, the following
observations can be m ad e reg ard ing th e overall effect of infill o n stren gth s,
w he n the behav ior rem ains in the e las t ic rang e: (1) W hen all the bare
frames of the building are infi l led, the increase in supplied strength con
s iderably exceeds the increase in s t rength demands; (2) in cases where
only 4 of the 11 frames are inf il led w i th panels hav ing a p > 0.4%, the
increase in suppl ied s t rength is larger than the increase in demanded
st rength.
From the s tand po int of e las t ic s t reng th, i t app ears that the use of
al l types of infi l ls (considered in the Berkeley invest igation), when prop
erly re inforced w i th p & 0.4%, is adva ntag eo us , in com parison to the
behavior of bare f rame bui ldings . The only thing remaining is to es t i
mate what intens i ty of ground mot ions the suppl ied e las t ic s t rength wi l l
be capable of resis t ing. The main results of this est imation are sum
marized in Table 6. From comparison of resul ts obta ined between in
fi l led frames and bare frame buildings, the fol lowing observations can
be made .
Case W here A l l Fram es are Inf i lled. Un reinforced m aso nry infills
could be used advantageously ( i .e . , e las t ic s t rength suppl ied larger than
elast ic s trength demands) in seismic regions in which the peak effective
acceleration
a
ep
is < 0.12
g,
whi ch , a cco rd i ng t o t he A T C recommenda
tions (1), is for most of the U.S. (areas 1, 2, and 3). In the case of rein
forced lightweight concrete infills, these infills could be used in seismic
regions in wh icha
ep
< 0.32g , wh ich m eans they could be used in reg ions
of very severe ea r thquake ground mot ions . The maximum va lue spec i
fied by A T C (1) for a
ep
is 0.40 g.
Case W here O nly 4 of the 11 Fram es are Inf i l led. Un reinforced m a-
TABLE 6.Building Seismic Resistant Coefficient, C = (V)/Wand Effective Peak
Acceleration,
a
That it can Resist Elastically
Type of
infill and
reinforcement
D
None bare
frame
Unreinforced
masonry
Reinforced
hollow
masonry
Solid brick
reinforced
with WWF
Reinforced
lightweight
concrete
p, as
a per
centage
(2)
0.%
0.15%
0.6%
0.4%
0.6%
Upper
Bound
All 11 Frames are Infilled
V, in
thousands
of
pounds
3
(3)
2,475
6,989
(L)
7,762
(L)
12,870
(Av)
12,494
(Av)
19,107
(Av)
C
(4)
0.11
0.30
0.34
0.56
0.54
0.83
T, in
seconds
(5)
1.30
0.52
0.52
0.52
0.60
0.39
a
f/
(6)
0.10
0.12
0.13
0.22
0.21
0.32
Lower B ound
Only 4 of 11 Frames are Infilled
V,in
thousands
of
pounds
8
(7)
2,475
3,329
3,610
5,468
5,330
7,735
C
(8)
0.11
0.14
0.16
0.24
0.23
0.33
T, in
seconds
0)
1.30
0.75
0.75
0.75
0.84
0.61
q>/g
(10)
0.10
0.07
0.08
0.13
0.14
0.17
*1
kip =
4.45
kN.
1351
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
scelibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/26
/16.
CopyrightASCE.
Forper
sonaluseonly;allrightsreserv
ed.
7/26/2019 Masonry infill analysis
16/25
sonry could be used in se ismic regions where the
a
ep
0.07
g,
i.e. , in
regions located in the U .S. area classif ied by A T C (1) as 1 an d 2 . T he
solid split bricks reinforced with welded wire fabric could be used ad
vantageously with respect to bare f rame in regions where a
ep
0.14 g
(i.e., for a l l 1 , 2 , and 3 areas according to ATC map) without danger of
suffering serious damage. Similarly, reinforced lightweight concrete in
fil l could be used in areas where a
ep
0.17g, i.e . , ATC areas 1-4.
I t can be concluded that inf i l l ing moment resis t ing f rames with prop
erly reinforced panels offers advantages when designed so that the frames
would remain in the e last ic range dur ing the most severe ear thquake
ground motion that can occur . But what would happen if these inf i l ls
were subjected to deformations larger than those corresponding to i ts
max imum elastic streng th? Ca n th e infilled frame survive suc h defor
mations without severe damage? In a t temping to answer i t is necessary
to analyze the inelastic behavior of infills in the infilled frames, and how
this behavior affects the performance of the frames.
Effect of Infil l on the Inelas tic R esp on se of th e B uildin g. In the
analysis of this effect is is convenient to distinguish the following cases:
DuctileMoment ResistingFrame Infilled with UnreinforcedM asonry.
Un
der cyclic loading, (Fig. 7(b)) as soon as the panel reaches i ts maximum
strength (which occurs with very small amounts of inelast ic deforma
t ions, approximately 1.5 times that which will correspond to linear elas
tic behav ior, given a disp lacem ent ductil ity ratio, (JL
8
, of about 2.5), there
is a reduction in strength to a value of about 23 kips (102 kN) that is
c lose but somewhat higher (10%) than that observed in the exper iments
conducted with a first soft story frame (Specimen 9, Fig. 6), and then
an increase up to a value of about 30 kips (133 kN) up to a |x
6
of about
39.
It should be noted that after a
(JL
S
of 2.5, som e po rtion s of the un
reinforced infill started to spall out. If an analysis using an inelastic re
sponse spectra derived from the linear elastic spectra of
Fig.
11 according
to the rules given in Ref. 8 for a (x
6
= 2.5 is conducted, the increase in
strength de m and du e to the decrease in T from 1.30 sec to 0.52 sec is
found to be 138%, wh ile exper iments sho w that the increase in the s up
plied strength is 182% for |A
6
up to 2.5. T herefore, rega rding s treng th,
it appears that ductile moment resistant frame with unreinforced infil ls
can be used advantageously in regions where
a
ep
is =0.26
g
if all the 11
frames are infilled, or
a^
0.22
g
if only 4 of the 11 frames are infilled.
T he real pro blem w ith th is kind of infill is no t initial stiffness or str en gt h,
but that with panels having large dimensions, as those under s tudy, as
soon as maximum strength is reached the masonry uni ts can shat ter and
large portions of the infil l spall out. In earthquake response, this is l ike
an explosive failure with shedding of large portions of unreinforced ma
sonry all around. This type of explosive failure of unreinforced masonry
infills has been typically observed after moderate to severe earthquake
ground m otion. In general it is inadvisable to use u nreinforced m aso nry
infills except in cases where the response demands will not exceed the
elastic range, and where out-of-plane failure of the infills can be restrained.
Nonductile Moment ResistingFrameInfilled with UnreinforcedMasonry.
This case is s imilar to the previous one but even more dangerous be
cause the explosive type of failure of the infill leads the infilled frame
to behave like one soft story frame with very large demands in shear
1352
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromasce
library.orgbyIndianInstitute
ofTechnology,Delhion04/26/16.CopyrightASCE.Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
17/25
and plas tic ro ta tions in the columns and /o r the beam s or beam -column
joints adjacent to the failed infilled pa ne l. A s thes e elem en ts ha ve no t
been designed to resist such demands, the explosive fai lure of the un-
reinforced m aso nry us ual ly will lead to the collapse of the frame. T hu s
this system should not be used except for cases where the bui lding can
resist e lastically th e effect of the m ost sev ere earth qu ak e gr ou nd m otion ,
i .e . , should be l imited to regions where a < 0.12g if all the frames are
infilled or a s 0.07 g if only 4 of the 11 frames are infilled.
Properly Designed Ductile Mom ent R esistantFrameInfilled with Reinforced
Masonry orConcrete Panels. 1)
Reinforced masonry infi l ls . Experiments
show that |x
s
at the average peak strength of the reinforced masonry
infill, V)f, is at least equal to two. Therefore, the reinforced masonry
infil led frame b uilding on th e average can resist seismic gr ou nd m otio ns
(of the types given a design response spectra as that of Fig. 11) having
the following peak accelerations: If all 11 frames are infilled
a
ep
= 0.40
g
for T = 0.52 sec and a
ep
= 0.38 g for T = 0.40 sec; if 4 of the 11 frames
are infilled a
ep
= 0.26 g for T = 0.75 sec and a = 0.18 g for T = 0.54
sec.
In the case where the infil l consisted of solid split bricks reinforced
with two layers of WWFsince the infil led frame can develop a |A
8
=
4.2 with a reduction of only 14% in strength (Figs. 9 and 10), i t becomes
evident that this type of st ructural system can resist earthquake ground
motions have the fol lowing a : If all 11 frames are infilled a
ey
= 0.77 g
for T = 0.60 sec and a
ep
= 0.59 g for T = 0.46 sec; if 4 of the 11 frames
are infilled
a^
= 0.55
g
for
T =
0.84 sec and
a
ep
= 0.44
g
for
T
= 0.66
sec.
In the case of a building with bare ductile framefor a
T
if
= 1.30 sec
i t would require developing a (x
s
s 6.1 to be able to resist a g ro un d
mot ion wi th an a^ = 0.55g, and for a Ty = 1.01 sec it would require a
|i,
6
5.6 to resist an
a
ep
= 0.44
g.
Since exper iments have shown tha t
the bare frame structure can develop a JJL
S
= 6.1 w ith ou t an y significant
loss in strength, i t would appear that there is no advantage in using
infills except when the majority of the frames are infil led. However, i t
should be recognized that for a bare frame structure to develop a jx
s
=
6.1, i t would have to undergo lateral displacements considerably larger
than that needed for an infi l led frame building to develop JJL
S
= 4.2. Fur
thermore, while in the case of the infil led frame, most of the damage
wil l be developed in just one or two stories where the inelast ic defor
mations are concentrated; in the case of the bare duct i le moment re
sist ing frame, the damage wil l spread throughout the whole height .
In the case of solid split bricks reinforced with WWF, the specimens
were deflected, producing an interstory drift of 2.4 in. at the story where
inelast ic deformation was concentrated. This drif t , which means an in
terstory drift ratio of 0.07, was achieved without any significant spalling
of debris. This interstory drif t , when translated in duct i l i ty displace
ment , means a |x
8
= 14 w hich w as at tained w ith a redu ct ion of st ren gth
of 32 pe rce nt (Figs. 9 an d 10). T herefo re, this sp ecim en cou ld resist th e
following a w ith ou t da n ge r of failure (collapse): If all 11 frame s are in
filled
a
ep
= 2.05
g
for
T
= 0.60 sec and
a^ =
1.54
g
for T = 0.46 sec; if 4
of 11 frames are infilled
a
ep
= 1.62
g
for
T
= 0.84 sec anda
ep
=
1.31
g
for
T
= 0.66 sec.
1353
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
scelibrary.orgbyIndianInstitu
teofTechnology,
Delhion04/
26/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
18/25
In conclusion i t can be stated that the use of specially designed mo
ment resistant frame infi l led with reinforced masonry, particularly solid
spl i t br icks with WWF, can be used advantageously for even the most
severe seismic regions of the U.S.; provided the number of stories is
limited to about
11.
This l imitation is necessary because the inelastic de
formation in this type of structure is usually concentrated in one or two
stories, the larger this number of stories of a building the larger will be
the demand in the story in which this inelastic deformation is concen
trated. Fur thermore, the f rame has to have very duct i le members be
cause the inelastic demands at the story in which the inelastic defor
mat ions concentrate , would be very large. This problem has been
discussed by Park and Paulay (9) , who show that the required column
curvature ductility factor
U
ti/ycii can t>
e
typically expressed as
U
ci/yci
= 12.54r - 3.2 where r is the number of the story to the top of which
the deflections are to be measured.
(2) R einforced Lig htw eigh t C onc rete Infills. T his type of infilled frame
is capable of dissipating energy with a ducti l i ty somewhat larger than
two without any loss in strength. However, for a |x
g
just larger th an
three the s t rength reduces rapidly to a value somewhat h igher than the
strength corresponding to the soft story frame. Considering a (x
s
= 2, it
has been estimated that buildings with this type of infi l led frame can
resist ground motions with the following a
ep
: If all 11 fram es are infilled
a
ep
< 0.54
g
for
T =
0.39 sec and
a^
< 0.54
g
for
T
= 0.30 sec; if 4 of 11
frames are infilled
a^
< 0.31
g
for T = 0.61 sec and
a^
s 0.25
g
for T =
0.47 sec. Considering the value at which strength appears to be stabi
l ized, 42 kips (187 kN), which is considerably higher than the 27.4 kips
(122 kN) w hic h is th e m ax im um lateral resistan ce of a ba re fram e soft
story, and that the inelastic deformation at this level gives a m = 6.6,
the following values of
a
ep
can be ob tain ed : If all 11 frames are infilled
tie,,
< 0.64
g
for T = 0.39 sec and
a^
< 0.48
g
for
T =
0.30 sec; if 4 of the
11 frames are infilled
Ugp
^ 0.37
g
for T
0.61 sec and u
e
p ^ 0.28
g
for
T
= 0.47 sec.
From analysis of the above results i t can be concluded that
R/C
bare
frame buildings of the type investigated can be advantageously infi l led
with reinforced l ightweight concrete for even the most severe seismic
regions of the U.S. if all the frames are infilled, and for the ATC Map
areas 1, 2, 3, 4, and 5 if only 4 of the 11 frames are infilled.
Nonductile Moment Resistant Fram e Infilled with Reinforced Pane ls.In
general this type of construction is not advisable if significant inelastic
deformation is expected. In infilled frames the inelastic deformation is
concentrated within a few stories, usually the lower ones, so ducti l i ty
demands on the frame members of these stories can be very large, con
sequently these members should be ducti le. Because of this type of be
havior a designer could be tempted to design as duct i le only the mem
bers of the story or stories in which inelastic behavior of the infill is
expected. To design in this manner appears logical and economical,
however , the designer must be aware that the resul ts obtained in th is
investigation, as well as in others, clearly show that for such a design
to work i t must be assured that the inelastic deformations will actually
concentrate in the weakest spot, i .e. , the story that is designed as duc
tile.
This is not an easy task. The uncertainties involved in predicting
1354
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromascelibrary.orgbyIndianInstitute
ofTechnology,Delhion04/26/16.CopyrightASCE.Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
19/25
the cri t ical seismic response of buildings are so large that conservative
precaut ions should a lways be taken. Fu rtherm ore, the s t reng th, s tif fness
and deformation capacity of masonry infi l ls are very sensit ive to quali ty
control of the materials and workmanship. To believe that i t is possible
to control exac tly w he re inelast ic defo rm ations can occur in a real
building is too optimistic.
CONCLUSIONS
In view of the relat ively small amount of experimental data on which
the fol lowing conclusions are based, and the idealizat ions, s implifica
t ions,
and assumpt ions made in the numerical analys is conducted, i t i s
convenient to c lear ly recognize the cons t ra ints surrounding the val idi ty
of the conclusions so that they wil l not be misused. These l imitat ions
are summarized regarding the fol lowing parameters :
1. Type of Frame. A special ly designed R/C moment res i s t ing space
frame and 3 bays and 11 stories .
2.
Type of Infi l ls . Unreinforced and reinforced masonry units (hollow
and solid bricks, and concrete blocks) and lightweight reinforced concrete.
3. Qual i ty control of M ater ia ls . A l thou gh the m aso nry un i ts use d in
construction were carefully selected and the grout , mortar, and concrete
carefully designed, mixed, placed, and cured, considerable variat ions in
the mechanical characteris t ics of these materials were observed. The re
sults indicated that the behavior of the infill is very sensitive to vari
at ions in the quali ty of material and, therefore, good quali ty control of
all material is a must for infills, particularly masonry infills.
4. Workmanship. Some weaker , s t i f fer , and premature types of ine
las t ic behavior and pat tern of cracking and/or crushing were a t t r ibuted
to lack of uniform workmanship in laying the masonry uni ts and in the
anchorage of the infi l l to the frame; thus excellent workmanship is
required.
5.
Infill Panel A rrang em ent . T he two external bays of the 3 bay frames
were ful ly infi l led, i .e . , without any opening, and formed what could
be called a cou pled infil led fram e.
6. T ype of B ui lding Co nsidered in the A ssessm ent of the Im pl ications
of Resul ts Obtained. Regular bui ldings having a rectangular plan con
sisting of 11 frames of 3 ba ys a nd of 11 stories hig h w h er e th e fram es
are fully infilled, as described in item five, and the locations of these
infi l led frames are such that no significant torsional forces are induced
during the se ismic response of the bui lding. The importance of this l im
i ta t ion cannot be overemphasized.
7. Idealization of the A ctual Lateral Load -Deform ation R elat ionsh ips
of the Bare and Infi l led Frames. The analytical assessment of the impli
cat ions of the experimental resul ts regarding behavior of the bui lding
have been made ideal iz ing the actual experimental re la t ionship by a l in
ear elast ic-perfectly plast ic model using different yielding strengths and
ductility levels.
8 . Dynamic Character is t ics of Bui lding Si te and of Ground Motions .
I t is assum ed that the bui lding is on firm gro un d an d a r igid found a
t ion can be cons t ructed, a nd tha t a ll the gro un d m ot ions that can occur
1355
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
sc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
20/25
have dynamic characterist ics similar to those included in the derivation
of the smoothed linear elastic (Fig. 11) and inelastic design response
spectra suggested by Newmark and Hal l (8) . The impor tance of the l im
i tat ions imposed by these assumptions in conjunct ion with the ideal i
zation pointed out in i tem seve n shou ld be em phasized , particularly w he re
significant inelastic behavior is involved in the response. The effects of
ground motions containing severe acceleration pulses (higha
ep
) of long
duration should be investigated before the conclusions from these re
sul ts are appl ied to the design of new bui ld ings and/or to ret rof i t t ing
of existing buildings. The interacting effects of the observed significant
deformation softening after reaching peak lateral resistance, with long
accelerat ion pulses inpu t , can lead to deformation de m an ds considerably
higher than those predicted by a l inear elasti-perfectly plastic idealiza
tion (7).
9. R eliabil ity of the A nalytical R esults . In view of all th e as su m ptio ns ,
ideal izat ions , and uncer taint ies involved in the conducted analyses , the
numerical values obtained should be considered as approximate and in
dicating trends, rather than an exact representation of what can be ex
pected in specific cases.
CONCLUSIONS
Conclusions Regarding Overall Behavior of the Infilled Specimen
Tested.
1.
The addition of either unreinforced or reinforced infi l l to moment
resisting frame increases significantly the lateral stiffness and lateral re
sistance of the frame.
2.
As soon as cracking occurs , which happens very ear ly , a t service
lateral load level, the initial tangential lateral stiffness decreases
signif
icantly, up to 80 percent, to a value that remains practically constant for
a long range of lateral load. To represent this behavior an effective in-
terstory stiffness at lateral service load has been defined.
3.
The ins tantaneous lateral s t i f fness and s t rength depends on the
previous loading history. Under monotonically increasing load these two
characterist ics depend on the type of infi l l . These characterist ics do not
depend upon how the panel is reinforced but they are sensi t ive to the
quali ty control of the materials and to how well the infi l l is made, par
t icularly to the workmanship along the interfaces of the infi l ls and the
boundary f rame elements .
4.
Hysteret ic behavior depends upon the type of inf i l l , the amount
and ar rangement of reinforcement , the way that the panel is a t tached
(anchored) to the frame, and the loading history. The cyclic loading,
including force reversals of unreinforced infills, leads to considerable de
ter ioration in st iffness and strength when compared with the values ob
served under monotonic loading; the largest deter iorat ion occurred un
der cyclic loading with full reversal of deformations. This deterioration
is due to propagation of infi l l damage that usually concentrates in one
story. The peak strength under cyclic loading, which is smaller than that
obtained under monotonical ly increasing load, deter iorates as the se
verity of deformation and number of cycles increases, but remains some-
1356
J Struct Eng 1983 109(6): 1337-1361
Downloadedfromasce
library.orgbyIndianInstitute
ofTechnology,Delhion04/26
/16.CopyrightASCE.Forper
sonaluseonly;allrightsreserved.
7/26/2019 Masonry infill analysis
21/25
what larger than the s t rength of a f rame wi th a sof t s tory corresponding
to the s tory in which damage of the infi l l concentrates . Excellent hys-
tere t ic behavior has been obta ined wi th the use of sol id br ick masonry
infi l ls externally reinforced with welded wire fabric covered with cement
mortar.
5. A l thoug h the inters tory displacem ent ducti li ty un de r pea k s t reng th
is smal l , about 2 , large values are obta ined under reduced s t rength. In
the case of solid brick externally reinforced with welded wire fabric, this
duct i l i ty was 4.2 under 86% of the peak s t rength, and reached the value
of 14 un de r 68% of pe ak stren gth .
6. Except for very few specim ens (Specimen 18 an d o ne rep orte d in
R ef. 6) w ho se failure m ech anis m s invo lved tw o stories , in all oth er spec
imens the damage concentra tes in one s tory, consequent ly the f inal
m echanism of failure is w ha t can be defined as a som ew hat s t rength
ene d soft s tory fram e. T hu s the energy diss ipated by an inf il led
R/C
frame should be larger than a bare soft s tory frame.
7. Fai lure of unreinforced masonry inf i l l s was accompanied by pro
duction of substantial debris containing hazardously large pieces of ma
sonry. The amount of debris in reinforced infi l ls was smaller and most
was contained in the plane of the infi l l , part icularly in solid brick ma
sonry reinforced externally with welded wire fabric.
8. The initial effective viscous damping coefficient of the virgin spec
imen s is smal ler than 2 % . A s soon as cracking dev elops th e value of this
damping coefficient increases up to 12%.
Conclusions from Comparison of Behaviors of Infilled Frames and
Bare Frame.
1. The initial tangential interstory lateral stiffness of the virgin infilled
frames wa s m or e th an 10 tim es the similar stiffness of th e ba re fram e.
2.
The effective interstory lateral stiffness of virgin infilled frames was
5.3 to 11.7 t imes the lateral s tiffness of the bare frame de pe n di n g o n the
type of infill. In case of repaired infills and retrofitting of repaired frames,
this effective lateral stiffness w as a t least 3.4 time s th at of the virg in ba re
frame.
3.
The maximum lateral resis tance of virgin infi l led frames was 4.8 to
5.8 t imes that obtained for the bare frame. For cases of repaired infi l ls
and retrofi t t ing of repaired frames the maximum lateral resis tance was
2.8 to 8.0 times that of the bare frame.
4.
The interstory displacement ducti l i ty rat io of the infi l led frame is
smaller than that of a bar e frame b u t larger th an th at of a ba re soft s tory
frame. For what can be considered a maximum acceptable interstory drift
index, say 0.02 or even for values of this index up to 0.07, the hysteret ic
behavior of the solid brick masonry externally reinforced with welded
wire fabric was superior ( large energy absorption and energy dissipation
capacities) to that of the bare frame.
5. T he add it ion of infil ls introd uce s s ignificant ch ang es in the dy nam ic
characteristic of the bare moment resisting frame. In the linear elastic
range the fundamental per iod is decreased more than 54%, whi le the
m ass is increased in not m ore th an 10%. T he effective viscuou s da m pin g
coefficient is increased considerably, up to 500%. In the inelastic range
1357
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromas
celibrary.orgbyIndianInstituteofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
22/25
the pat tern of la teral deformations changed fundamental ly because most
of the significant inelastic deformations concentrate in one, or at the most,
two stories.
CONCLUSIONS DRAWN FROM ASSESSMENT OF THE IMPLICATION
OF
EXPERIMENTAL RESULTS OBTAINED REGARDING
TH E
SEISMIC RESISTANT DESIGN OF BUILDINGS
1. The addition of infil l into the moment resisting frames of a build
ing introduces significant changes in the dynamic characteristics of the
bui ld ing which should be considered in i t s design. These changes de
pend upon the number of frames that are infi l led as well as the locat ion
of these frames.
2.
T he m ass is increased; how ever, eve n w he n all the transverse frames
of the building under consideration (Fig. 1) are infil led, the increase with
respect to a bare frame building is only about 10%, and those two main
effects of this increase are negligible.
3. The stiffness of the building is increased significantly in the case
where all the frames are infil led, the increase varies from 366% to 994%.
If only four of the frames are infilled the increase varies from 136% to
353%.
4. If the 11 frames are infil led the d ecre ases in the fun da m en tal pe
riod varies from 54% to 70%. If only four frames are infilled, the decrease
varies from 35% to 53%.
5. The value of the effective viscous damping ratio for the whole
building increases w he n com pare d w ith a bare frame structu re.
6. Stre ng th Sup ply . A dd ition of infills to the frames in creas es th e
available (supplied) strength of the bare frame building significantly. If
all the 11 frames are infil led the lateral stren gth in the tra nsv erse direc
t ion of the bui lding is increased in 182% up to 700%, depending upon
the typ e of infills. In the ca se w h e re on ly 4 of the 11 fram es are infilled,
the increase varies from 34% to 255%.
7. Strength Demands. For l inear elast ic behavior the addit ion of in
fills to the bare frame increases the stre ng th de m an ds in 86% u p to 141%
w he n all the frames are infilled, a n d in 56% to 141% w h e n on ly 4 of th e
11 frames are infilled.
8. Supplied Strength vs. Demanded Strength in the Case of Elast ic
Behavior. From comparison of values given in the above conclusions 6
and 7, i t can be con clud ed that, excep t for cases of unreinfo rced infills
in which only 4 of the 11 frames are infil led, the increase in supplied
st rength is larger than the increase in the d em an de d s t rength , thu s f rom
the viewpoint of strength it is beneficial to add infil ls.
9. Deformation Demands in the Case of Elast ic Behavior. The addi
tion of the infil ls decreases the demands on maximum displacement with
respect to that corresponding to the bare frame building. The decreases
vary from 56% to 85% in cases where all the frames are infil led, and
33% to 60% in cases where only 4 of the 11 frames are infil led. This
decrease in displacement demand is a significant advantage in the use
of infills.
10. From conclusions 8 and 9 it is obvious that if i t is possible to de-
1358
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfromascelibrary.orgbyIndianInstituteofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forp
ersonaluseonly;allrightsrese
rved.
7/26/2019 Masonry infill analysis
23/25
s ign the bui ld ing to rem ain in the e las t ic rang e , then it i s adv anta
geous to add any of the types of infi l ls , even unreinforced masonry, i f
all the frames are infilled and
a
ep
^ 0.12
g.
In cases where only 4 of the
11 frames are infilled, i t is ad va nta ge ou s to a d d an y ty pe of infills re in
forced with p 2: 0.4% that have been considered in this study. While a
bare frame bu ilding can resist e lastical ly gr ou nd m otion s similar to th ose
considered in the d erivat ion of the res po ns e spectra of Fig. 11 w ith an
effective peak acceleration of a = 0.10 g, the addition of infills of solid
bricks reinforced external ly with wire welded fabric al lows the bui lding
to resist an
a
ep
= 0.21
g,
i.e. , an increase of 110% in intensity of ground
m otio ns if all th e frame s are infilled. If only 4 of th e 11 fram es ar e infilled
it can resist an
a
ep
=
0.14
g,
i.e. , an increase of 40%. By infil l ing all the
frames with reinforced l ightweight concrete i t is possible to resist e las
t ical ly ground motions with an a
ep
s 0.32g, which means tha t they can
be used in all the seismic regions of the U.S. except those classified as
area 7 in the A T C m ap area classification.
11.
For bui ld ings which can res is t the ext reme ground mot ion ex
pected at the site through large inelastic deformations, the use of infil ls
like that of solid bricks reinforced externally with welded wire fabric of
fers considerable advantage over the use of just bare frame. Because these
infi l led frames can develop an interstory displacement duct i l i ty jx
s
= 4.2
with a reduct ion in strength of only 14%, the bui lding can resist ground
mot ions wi th an
a^
s 0.44
g
ev en if on ly 4 of th e 11 fram es are infil led.
To be able to resist a similar ground motion the bare frame building wil l
need to develop a
JJL
S
> 5.6 with significant ly larger disp lacem ent , an d
consequent ly more damage throughout the whole bui ld ing.
CONCLUSIONS D R A W N FROM ASSESSMENT OF THE IMPLICATION
OF EXPERIMENTAL RESULTS OBTAINED REGARDING THE
REPAIR
AND RETROFITTING OF EXISTING BUILDINGS
1. For bare frames that have been damaged (cracking and spal l ing of
unconfined concrete) due to considerable yielding, developing interstory
displacement duct i l i ty of four, the fol lowing repair technique gives good
resul t : removal of any crushed and loose concrete and recast ing of i t ,
and injection of cracks with epoxy.
2. Undamaged, or damaged bare frames after their repair , can be ef
fectively retrofit ted for seismic resista nt pu rp os es b y th e ad dit ion of rein
forced infil ls that are properly attached (anchored) to the frame. Of all
the infil ls studied, the one that offers the greatest potential to retrofit
st i ffness, st rength and energy dissipat ion capaci ty to exist ing bui ldings
is the one based on use of sol id bricks reinforced external ly with welded
wire fabric covered with cement mortar and anchored to the frame, as
illustrated in Fig. 4.
ACKNOWLEDGMENTS
This paper is dedicated to
Prof.
Dr . B runo Thur l iman on h i s 60 th an
niversary as a tribute to his teaching and research in the area of inelastic
behavior.
1359
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
scelibrary.orgbyIndianInstitu
teofTechnology,
Delhion04/26/16.
CopyrightASCE.
Forp
ersonaluseonly;allrightsrese
rved.
7/26/2019 Masonry infill analysis
24/25
APPENDIX I .REFERENCES
1. T entative Provisions for the D evelo pm ent of Seismic R egulations for B uild
i ngs , A pplied T echnology Counci l Publ ica tion A T C 3-06, Nat ional B ureau
of Standards, June, 1978.
2.
A xley, J. W ., an d B ertero, V. V., Infill Pan els: T heir Influence o n Seismic
R esponse of B ui ldings , Report
No. EERC 79-28,
Ear thquake Engineer ing Re
search Center, Un iversity of California, B erkeley, Calif., Sept., 1979.
3.
B ertero, V. V., Seismic Performance of R einforced Con crete Stru ctures ,
Amales
de la
Academia
Nacional de
Ciencia
Exactas,
Fisicas
y Naturales, Buenos A i re s ,
A rgentina, Vol. 31, 1979, pp . 75 -144.
4.
B rokken, S. T ., an d B ertero, V. V., Stu die s on Effects of Infills in Seismic
R esi stan t R /C Cons t ruc t ion , Report No. EER C 81-12,Ear thquak e Engineer
ing Research Center, University of California, Berkeley, Calif., Oct., 1981.
5.
Klingner, R. E., an d B ertero, V. V., Earthq uake R esistance of Infilled Fr am es,
Proceedings,
ASCE, Vol. 104, No. ST6, June, 1978.
6. Klingner, R. E., an d B ertero, V. V., Infil led Fram es in Earth qua ke R esistant
Construc t ion, Report No. EERC 76-32, Ear thquak e Engineer ing R esearch
Center, University of California, Berkeley, Calif., Dec, 1976.
7.
M ahin, S. A . , an d B ertero, V. V., A n Evaluation of Inelastic Seismic Design
Spectra, Proceedings, ASCE, Vol. 107, No. ST9, Sept. , 1981.
8. Ne wm ark, N . M ., and Hall , W. I . , Pro ced ures an d Criteria for E arthqu ake
Resis tant Design, Building
Standards
for
Disaster
Mitigation, Nat ional Bureau
of Standards, Building Science Series 46, Feb., 1973.
9. Park, R. , and Paulay, T. ,
Reinforced Concrete
Structures,John Wiley and Son s ,
New York, N.Y., 1975.
10.
Priestley, M. J . N . , M aso nry , Design ofEarthquake Resistant Structures, E.
Rosenblueth, ed. , John Wiley and Sons, New York, N.Y., 1980, pp. 195-222.
APPENDIX I I .NOTATION
The following symbols are used in this paper:
a = a c c e l e r a t i on ;
K = la te ra l s t i f fness ;
L = l e n g t h ;
M = m a s s ;
P = axia l lo ad ;
r = n u m b e r o f t h e s t o r y t o t h e t o p ;
T = p e r i o d ;
V = b a s e s h e a r s t r e n g t h ;
p = pe r c e n t a g e o f m a i n r e in f o r c ing s t e e l ;
= c u r v a t u r e ;
t, = e ff ec ti ve v i s c o u s d a m p i n g r a t i o ;
J U L ,= duc t i l i t y r a t i o ; a n d
A = l a te r a l d i s p l a c e m e n t .
S u b s c r i p t s
bf = ba r e f r a m e ;
ep =
effec tive p ea k ;
I = i n t e r s t o r y ;
; / = inf i l led f ram e;
rif = re info rced inf i l led f ram e;
s = sca le ;
uci = u l t im a t e c u r v a tu r e a t s e c t i o n /;
1360
J. Struct. Eng., 1983, 109(6): 1337-1361
Downloadedfroma
scelibrary.orgbyIndianInstituteofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpe
rsonaluseonly;allrightsreser
ved.
7/26/2019 Masonry infill analysis
25/25
yd = yielding curv ature at sect ion i; and
8 = displacement .
Superscripts
D
=
demands;
m
= m odel frame;
p
= p rototy p e frame; and
s specimen.
Downloadedfromasc
elibrary.orgbyIndianInstitute
ofTechnology,
Delhion04/2
6/16.
CopyrightASCE.
Forpersonaluseonly;allrightsreserved.