Uniaxial and Biaxial Bending of Reinforced Brickwork Columns

S R Davies
E A El t.r:üi'.v

AllSTRACT

In a previous pape r lhe authors des c ri't- C'd 0 _'1 i ~erati v c 301uti 'J 11 fo r the
anal ysie of re ctangular re info rc ed bri ckwork cLlhllnJ :S ~;ubjec ted to bi axial
bending. In this pape r the method is de vel op~d f'Jr~h f;r ar.d in teraction diagrame
inc luded which enabl e the designer to c:o"1s ider ~,j;'2 f'f reet of different
eccentriclties of loading abou t b oth axes. ri~lC curves are sui tab le f o r solvi ng
for both uniaxial and b ia.xi a l bending .

A brief de sc ripti on of laboratory test~ on colu!l!ns i9 also included and a

compariaon made of experimental and theoretic.::.l \"alucs .

Depa rtment of Civil Enginet' ring 3nd lluilding: Sc-i .:,]UI.:,

Un lversity of tdinbur gh,
Klng's Bul1dlngs .
England

843
 INTRODUCTION
,
Coluou1s of reinforced concrete loaded eccentrically about one axis have
been studied extensively and some recommendati ons have also been put forward
fo r uniaxial bending af reinforced masonry. In the case of biaxial bending
of reinforced concrete columns it ia usual to consider the two axes separately and
to use some combination of twa separate uni axial cases .

Very li ttle work appears to have been done for the case of biaxial bending
in co lumns of reinforced brickwork and a procedure for the design cf such
columns ia outlined in this paper. Only one type of rectangular section ia
considered and although t he number of interaction diagrama i8 limited to three
cases it will be apparent that the method can be ex t ended to other sections and
additional load cases .

Biaxial bending can arise directly from an applied load which is eccentric
to both axes ar indirectly from forces introduced by wind ar earthquake and it
ia important that the columns are adequately designed to allow for these moments.

The problem is non-linear in that the bending produces a deformation which

introduces secondary bending and the procedure used for obtaining a solution is
based on an iterative method similar to that used by Farah and Huggins (1) for
columns of r einfor ced concrete. The columns are first subdivided into sections
and a triple iteration procedure used 1) across the section 2) between adjacent
cross-sections 3) over the column l ength. ( 2) .

BASIS OF THE METHOD

Since the eccentricities at any sec t ion will con sis t of t he algebraic sum
of the applied eccent ri city and a component of the deflection it follows that
a knowledge of the varia t ion of the eccentricity implies a knowledge of the
deflected formo

(a) I nitially the following assump tions are made with respect to the central
!Section.
1. The eccentricities
2. The co rner strains

The forces and moments at the central section can now be calculated
independently on the basis of the two assumptions and any differences corrected
fo r, by changing the corner strains incrementally. 'llhe calculat ion of the
force s and moments from the assumed comer strains is based on an i ntegrat ion
of the stresses over the sec t ion.

(b) Once the t wo s et s of for ces and moments are wi th in acceptable limits the
analysis proceeds to the next section by first ca1.culating the value of the
eccentricity at the second section as a function of the pre vious value and the
curvature. This requires a second iteration since the curvature is itself a
f'unct ion of the hfO eccentrici ties.

The comer strain at the second section are now assumed and the first
iteration repea t ed at s ection two until t he t wo se ts of forces and. moments are
in agreement.

(c) Finally the process reaches the top of the column and the calculated
eccentricity compared with the known applled eccentricity. If the difference
ia not within acceptable l~mits the whole process ~s repeat ed by incremen t ing
the assumed values of the eccentrlclties at the central section.

844
 This would be very tedious for hand calculation and a computer program has
been developed in Fortran to carry out the iterative processes. A flow chart
fo r this program is shown in Fig. 1. In this chart the second iteration
between adjaeent cross- seetions has been omit t ed sinee it was found that, for the
seetione eonsidered, the inclusion of this iteration did not greatly affeet the
re eul t s.

MATERIAL PROPERTIES AND GEOMETRICAL DATA

Stress-strain relationships must be defined for the three materials used in
the construction of the eolumns and for thi s present work the relationships used
are as shown in Figs . 2 and 3.

Although the computer programme is wri tten in non-dimensional form it is

based on columns of rectangular section similar to that shown in Fig . 4. The
modification required to analyse other sectional forms would present no
diff i eulties.

TREORETICAL RESULTS
Interac t ion diagrams are shown in Figs . 5, 6 and 7 for a typical reetangular
section for which dl = 0 .9T and d2 = 0 . 85B . The diagrams are based on values
of the axial l oad P of 0 . 2, 0 . 3 and 0.4.
where P = N/f BT
m

For these interaetion diagrams i t has been as sumed that the grout and the
brickwork have iden tical stress-strain charac teris ti cs.

A set of these charts could be produced for othe r load cases and other
rectangular sections with different values of d1 and d2 . ~~e charts sho~m
have been derived using factors of safety of 2.5 for brickwork and 1.15 for steel
so that ultimate values af the mornents are indicated .

The charts can be used in two ways.

1. Knowing the applied axial load and the moments about both axes, it is first
necessary to select the appropriate chart for the axial load (or interpolate
between two charts). Then by locating the interaction point using the known
values of moments, the required amount of steel can be ascertained .

2. Knowing the axial load and the area of s teel used , suitable values of the
ratio of M to M can be determined.
x y
Example 1:

Consider a rectangular column with B = 1OOmm , T ~ 200mm , d1 = O.9T and

d2 = 0 . 85B . Assuming that the axial load (N) is 300 kN and the IDoments are
Mx = 7 .2kNm and My = 2.4kNm . The required area of steel can be deteImined 2
as shown below . The design stresses are taken as f m = 15N/mm 2 and Fy = J..j.60N/mm .

Solution
300 x 100
P = N/fmBT 15 x 100 x 200 = O.JO
2 7 .2 x 1000 x 1000
M)fmBT 0.12
15 x 100 x 200 x 200

845
 2 2 .4 x 1000 x 1000 0 . 08
Mjfm TB 15 x 100 x 100 x 200
From chart (Fig . 6)

A If m BT 1~ x 10- 4
s
11 x 15 x 100 x 200
i. e . A
s 1000

Example 2:

Using the same section and design stresses as given in Example 1 and
assuming that the steel area ~s known to be 390mm 2 , then safe combínation of
M and M can be determined fram the charts as shown below,
x y
Calculate N 300 0·3
f BT 15 x 100 x 200
m
Therefore use ~

390
100 x 200 x 15

The broken line lying behleen A IB'T f values of 15 x 10- 4 and 10 x 10-
4
s rn
shown in Fig. 6 represents sa':e combinations of Mx and My'

EXPERIMENTAL INVESTIGATION
A test rig was constructeà for application ofaxial loads and biaxial
bending to columns constructed of half s~ale brickwork . Details of the test
rig are shown in Fig. 8 .

Initial tests indicated that there ',;ere inherent weaknesses in the rig
construction which required modification. 1hese changes wer e concerned mainly
wi th the support condi tioos at ei ther eod and the final arrangement the axial
load was applied through steel balls placed at both end of the column.

The results obtained from three subsequent tests are shown below.

Test Ult. axial load Measured Values (kNrn) Theoretical Values (kNm)
No. kN Mu Mu Mu Nu
x Y x y

235 . 8 10 . 95 4.07 10 . 08 4 . 48

2 235 . 8 9 · 12 4 . 98 8 . 19 4.7 1

3 294.8 7.08 3,01 8.65 4.27

The deflection profiles detennined using the computer programme, f~r
constant axial load P and different combinations of Mx and My are shown in Figs .
9 and 10. The measured values obtained from the tests are also shown on the
aame diagrama.

846
CONCLUSIONS
The limited number of practical tests completed to date shows that the
theor etical resul ts are conservative . Additio llUl testing ia required before
firm conclusion can be drawn but the results indicate that the theoretical
approach gives results which are in elose agreement wi th the measure values.

REFERENCES

1. FARAH , A. and HIGGINS, M."' . ' Analysis af Reinforced concrete columns

subjected to Longitudinal Load and biaxial
bending '. ACI Journal, July 1969. pp. 569 -
575 .
2. DAVIES , S.R. and EL1~IFY , E. A. ' Biaxial Bending of Reinforced Masonry
Columns' . 5th International Masonry
Conference, Washington, 1979.

ACKNOWLEDGEMENTS
The authors wish to thank the Londen Brick Company for donating bricks for
the continuation of this project and the British Council for financiaI support
given to the second author .

847
"'-

.,.
00
00
(.OMMEN T!:J
rES
~III LvA.E

I?e.~ pa,..",mc..f<L,r"s .
ND
.4N·{>;4/ V'A/,.c.,s 01 é(I)' O. O , E(2.) • 0.0, CALL FI (I'F . 8F)( • tJFY ) C..J«<J.1L I."..L
S-fro.in . é (3) • O. O ~ ç • " , 0000, ;n!C.NJ4.t -/.rGCS

Pc • P'::
c",..,.r&. ;".*",4..L
4"&/ u 1~.. / +OI'"'CU
C4/~/,f<. i",,,".. 1 ,
~f>f'I"&Á
6uu:l/"l.J
,"c""CJtfs .
,,,<C r ) : ... C q /"""-I,t4IL 04f·tlA.~"~
• c.u:. .. ,,+r--i(.if,'cs .j.r
Ex ( J""'J.) ~ . . ' Il(..t:f s wicn
Cc.lt;.,I,f./..,tc. ;"tt.tT'lt4.(
,r.....:as 4'" fo ;"i",.( CAL.L FI (PC, l3oX. so y)
1
~ut.. of ~f".,""
I :3 1,3
c-.,..,... c...,.~II·CJ/. ,,,J
'-/G44/ .... 'ftuI ~·cilrft:s
""t -Iof tJtt/ -I ~,;.,.
élI) .ECI)U.e
1 Ol .. ~ .. 4(.u.+'·/4'<h"tZ..!
M"'" ( PF•• FM..aF'(, tel) •• Cl) .6(1»
c",J'-' ... tf. /;,*,"""""",,,'tt 1
....t """;,, ,c.c:I, ..... i.f-
'rc.c.s. eI ... Co -t-~ CALL FI (PrCI).Brx(I),B7YII)
nt.'.. ss ... '-'1
1
inc.f"II."'PCJft A"
Y&.S
C4/c..J47< /",rli.t
clf!,·i"'A.~/·Y"ZS

~ -lo. ;1<U'OAM.d& CAL" Fl (lElI). GHt).G~(J)

$6(J). 'ttC')'oI,CX) L

H I) , HJ)tlfC.,
EC'} .f(a)flf(;
H3)'UJ)t&'El', t
I
C?N0 J
hG../
 .--r_ _f.;:....:..
/ _f"'''---_..:7.:.EN:;:",' DIV

" ,(7
-"'~=- ____,, ____~~+.fy/ZC!{)--
CO~PI:GSS.IO",

,, ,
I

,,
STI1ESS

1
1 ./
I ,
I 1
'I

Fie. 2. Sh",-t _ ie.rm de.siJn stf"e$S- SfM.//t. f'(lledi"A....

for r<ú,f.rc~t: C fy ;n ,v/",,,,,-)

-- - --_ .,-,........... ~ !
jm

Flfi. s. S+r~$S _ sfra,,, rd.:fio. SÁ" !.r f,.icJ;:wol"!<..

«.ud'5"/(~(/I),

849
 T
.t =ZS'3 . t

~i
(j:l) C I"OSS
CI)
-..5t2 c tIOrl

f"
~
0.0>

0. ..
"'.O.9T
<I. ,0. &S S
0. .. 11 • 4-'c N/1ft",. 1-
)
f : D. Z.

0..-

0.04

0.00
0.00 0.04 0. 12 0.t8 0.., o...
rtXIFI8T2
F, 8" S 1'O(-f'fY INTEKTtIJí OI~.

850
 ... t 8
dI. ~ 1~
10.11
L"

LI'
"-
...
-1'.,-

"-
~.>-
.x;
~
",[O] d, =O·~T
"- -:;..
LI.
'-. ... -".,O.t5"S
0.10 "-
......0 'R
Fi, 4-60 "1-'
L. - - - - - - - - +?(,,:--r,
#0 "' "- P ="·3
7'i "-
L.
I "-
-5'+
10M ~O ,
,. "- ,
L. \
\
L.
... .... ... .. li .... ..111
ltXIFIIII2
. ..
f, 9' (, I.HI\" INTEJR:TIOII DIAG.

1$ l
0.1. i hl 1
!
~
0.1.

0."

0.12
-9".
/
<f;>o
,r
,,~

[O]
d"".'!T
.. lO
'" ~
/S- ch, 0-I!5"8
+/., . . ~ p/ _ 46. N("'''"
.... I' = O.'f
.... '" + .....O-~

"OI
S-...
'2>'~
0.02

O.DO
"DO .... 0.01 O." .... O.tO 0.12 .... ",. O.la
lIlVfIIITl!
. .,
f'g. Y. ~-tfY INTERfl:TION OIAG.

851
 G,«"vr,rZJ _I......-- .
~/C"'_~k ~~
CD.tVMA/

r I

852
 ,I!l
t

o 4
"l

~ o
~ Q

~ ...
II

: o

.'"
O

"-

"'"
'I-

.1),,//ecr/on .-"'/"0/.//"-
/./7 ~/qn~ a/ 4..-vb,/
axl.s {~5é./".1

2. • ~ S ,., '1. ' .. AI#1

l! r -. (€ .•::t..,. ~cl) i
. i.DAD
S7''''44
IV (IQi)
_.
~( i<N",) M; (I"" Ir> 1
~
I -t 23S-· 8 Z.OI /. 00

2 o 23S . i 4·DO 2 ·00

3 A 2JS'8 '·9. 3 ·cO

it O 2J~·8 9·00 3· q,.- .

853
 - -'-

()

~ I

o "-

,
()

~ '"
't
"~
~
~
I>
I "

,
\I

~ ()
t
Cl
O)
~ I
A LJ<?//rc;:'" ,., /'/"0.*/'"
//,7/7 r 0/ /?evQd/
I~

II c7X/J (~.$"<"2)

O
, 1/
~

,JJQ' "tj/'Cr( (,'"

, -4
" OI'o ' .. I

e r = ~"-~V')'Í
L (JÃD
N (kll) "1. (.«,<I,.,) 1tr(.I.II,., )
srAt;.
I + 294· .1\ 2 ·oS I· di
2 o 294tJ 3·04 2·0~
H
3 A 294·8 4-02 3·02

854