UDC 624.072.2.012.45:69.

022

Structural idealization for deep beams with

*

web openings
F. K. Kong" and G. R.Sharpt

U N I V E R S I T Y O F C A M B R I D G E : E N G I N E E R I N GD E P A R T M E N TA N D
U N I V E R S I T Y O F N O T T I N G H A M : C I V I LE N G I N E E R I N GD E P A R T M E N T

SYNOPSIS measured ultimate load

The exactanalysis of reinforcedconcrete deep ultimate load computed from equation 2
beams with web openings presents formidable prob- ultimate load computed from equation 3
lems. However, the ultimate shear strengths of such clear-shear-span distance (Figure 1)
beams can be predicted with reasonable accuracy by depth at which a typical bar intersects the
using a simple structural idealization. The idealization potential critical diagonalcrack in a solid
was based on pilot tests on 24 beams, but has now deep beam, which is approximately the line
been checkedagainst afirther 32 tests covering a much joining the loading and reaction points
wider range of web openings and web-reinforcement Y1 depth at which a typical bar intersects a po-
patterns. A n improved shear strength equation is pre- tential critical diagonal crackin a deep beam
sented and design hints are given. with openings, idealized as the EA lineor CB
in Figure 6
Notation CY angle of intersection between a typical bar
A area of an individualwebbar(for thepur- and the potentialcritical diagonal crack des-
pose of equations l to 3, the main longitud- cribed in the definition of y above
inal bars are also regarded as web bars) angle of intersection between a typical bar
a,,a, coefficientsdefiningthedimensions of an and a potential critical diagonal crack (line
opening (Figure 2) EA or CB in Figure 6)
b breadth
(thickness) of beam empirical coefficient, equal to 1 5 for web
C, empiricalcoefficient in equations 1 to 3 (for bars and 1.0 for main bars
angles defining the directions of the poten-
normal-weightconcrete, C, = 1.40;for
tial critical diagonal cracks (lines
E A and CB
lightweight concrete, C, = 1.35 where the
in Figure 6)
cylinder-splitting strength5 is determined in
accordance with ASTM Standard C330, but
C, = 1.0 where5 is determined in accord- Introduction
ance with BS 1881) The designof reinforced concrete deep beams with
Cz empiricalcoefficient in equations 1 to 3 (for web openings is not yet covered by the major codes
deformed bars, C, = 300 N/mm2; for plain of practice.('-3) In Great Britain, the Construction
round bars, C, = 130 N/mm2) Industry Research and Information Association has
D over-alldepth of beam (Figure 1) been supervising the preparation of a design guide(4)
ft cylinder-splitting
tensile
strength of concrete forpractisingengineers,butextensive s~rveys(~-~)
k,, k l , coefficients defining the positionof an open- have shown that I'ittle information is available in the
k,', kp' ing (Figures 2 and 5) literature on the effects of web openings upon the
L simple
span of beam (Figure 1) ultimate load behaviourof concrete deep beams.Re-
Quit ultimate shear strength (QU,, = W,/2 orW3/2) cently, the authors proposed(8) a design method for
deep beams with openings,using a simple structural
*University of Cambridge: Engineering Department. idealization based on 24 pilot tests('). The pilot tests
tAllott and Lomax, Consulting Engineers, Manchester. were necessarily of an exploratory nature; covering

81
Magazine of Concrete Research : Vol. 29, No. 99 : June 1977

only one type of web reinforcement and one span/

depth ratio LID, they cannot be expected to provide
adequate answers to questions regarding the validity
of the structural idealization under various circum-
stances. A further 32 tests have therefore been car-
ried ouf and are reported here; these tests were spec-
ially designed to test the validity of the structural
idealization and t o supply additional information to
fill gaps leftby the pilot tests. These 32tests, together
with the 24pilot test^'^', covered (a) sparddepth ratios
LID of 1,1.5 and 2, (b) clear-shear-sparddepth ratios
xID of 0.2,0.25,0.3 and 0.4 and (c) a wide range of L
-~
= 1125 or 7 5 0
~
-f
web openings and web-reinforcement patterns. To (a) Reinforcement details of group 0 beams

m
the authors'
these
most
tests
form
the
comprehensivestudycarriedout todateonthe
subject. Type
W1

Test programme Y Y

T E S TS P E C I M E N S
The test specimens were designed to complement W3
those in the pilot ~ t u d y ' ~ ' a nconsisted
d of 32 simply
supported lightweight-concrete deep beams (Figure
1 and Table1) of over-all depthD 750 mm and width

a t
b 100 mm. Span lengthsL of 1125 mm and 7.50 mm
were used, giving LiD ratios of 1 . 5 and 1 respec-
tively.Similarly,twoclear-shear-spandistances x
W5
IIH
were used, giving xlD ratios of 0.3 and 0.2 respec-
tively.
The test beams were divided into two groups: the W7
group 0 beams had no web reinforcement, whilst the
group W beams incorporated seven types of web
reinforcement. Deformed bars of 10 mm (3/t3 in. nom- (b) Web reinforcement of group W beams (consisting of 10 mm
inal) diameter andof 460 N/mmz yield strength were dia. stirrups, web steel ratio = 1 . 2 %).
used for the web reinforcement and the web steel Figure 1: Dimensions and reinforcement details.
ratio* was asnear aspossible constantat1.2 %
(Table 1)so that theweight of web steelin each group
W beam was very nearly the same. Thelongitudinal TESTING
steel consisted of one 20 mm (3/4 in. nominal) diam- In the pilot study'", all the beams were testedunder
eterdeformedbar of 430Nlmm2 yield strength, two-point loading. In the present investigation, 28 of
anchored to external steelblocks at the ends (Figure the beamsweretestedundertwo-pointloading
1).Lytag sinteredfly-ash lightweight aggregateswere (Figure 1) and 4 under four-pointloading (Figure 3).
used in concrete making, the materials and propor- Loads and reactions were applied through rollers and
tions being the same as b e f ~ r e ' ~aggregatekement
': 100 X 100 X 30 mm steel bearing blocks bedded to
ratio = 2.8 and (total water)/cement ratio = 0.8. the concrete with Crystacal-D quick-setting plaster.
Details of concrete strength are given in Table 1. The semi-circular rollers at the reaction points were
The positions and sizes of the web openings com- specially mounted to permit horizontal translation.
plemented those in the pilot study(9), and are indi- The four-point loadingwas used as a crudebut con-
cated by reference numbers ranging from 0 to 16 as venient approximation for the distributed-load con-
explained in Figure 2. Briefly, the size of arl opening dition, to give some idea of whether the conclusions
is given by ( q x ) x ( a 2 D ) ,where the height factora2 drawnfromtests using two-pointloading were
was kept constant at 0.2 but the breadth factor a , broadly applicable to the distributed-load condition.
varied by increments from 0.3 to 1.5. (The relative
position of the openings can be seen clearly in Figure
4.) T E S TR E S U L T S
The ultimate loads of the test beams are shownin
*The web steel ratio was calculated as the ratio (volume of web Table 2, the modes of failure in Figure 4, and crack
steel)/(volume of concrete). widths in Figure 7.

82
 Structural idealization for deep beams with web openings

T A B LE 1: Properties of the test beams.

Ref. No. Ref. No. of

of beam* web opening tt
(N/mm2) (N1mm2) I (Nlmm2)

GROUP 0 BEAMS

0-0.3/0 0 39-0 37.0 2.69

0-0.311 1 40-4 35.6 2.61
0-0.312 2 41.3 36.9 3-06
0-0.313 3 41.7 35 5 2-69
0-0.3/4 4 40.8 34-7 2.69
0-0.315 5 39.2 35.0 2.74
0-0.316 6 33.4 33.3 2-89
0-0.317 7 43.7 39 2 3.04
0-0.318 1.5 0-3 8 - 33.0 31.8 2-61
0-0-319 9 45.0 38.1 2-80
0-0.3110 IO 36.0 33-6 2.85
0-0.311 1 11 30.8 33.3 2.78
0-0.3112 12 36.7 33- 1 3.11
0-0-3113 13 41.3 37-8 2.92
0-0.3114 14 33.2 30.2 2.76
0-0.3115 15 35.2 33.6 2.92
0-0.3116 16 43.4 37.6 3.07
~

0-0.2/0 0.3 0 39.6 37.4 2.93

0-0.214 0.2 4 - 42.0 39.6 3.19
1.0
0-0.2/13 0.2 13 38 5 39.5 2.85
0-0.2116 0.2 16 40.4 38.9 2.76
I

GROUP W BEAMS
~

w1-0.314 1.19 34.2 39.5 2.93

W2-0.314 40-5 1.19 34.6 2.96
W3-0.314 1.19 33-7 40.9 2.87
w4-0.314 1.5 0.3 4 39.1 1.24 33.3 2-89
W5-0.314 1.11 35.3 36.8 2.93
w6-0.314 37-8 1.25 31.9 2-91
w7-0.314 1.13 33.0 37.4 3.03

W1 1.19 31 8 34.5 2.82

34.3 1.19 33.6 3.04
1.5 0.3 4
W4 1-24 32-5 35.2 2.89
W7 1.13 37.7 31.9 3-04

*Beam notation: A letter 0 before the hyphen indicates that there is no web reinforcement whilst a letter W indicates the presence
of web reinforcement; the xlD ratio is given after the hyphen, followed by the web-opening reference No. Thus W1-0.314 refers to a
beam with web reinforcement type W1 (see Figure l ) , having an x1D ratio of 0.3 and a web opening of type 4.
?The four beams with a suffix A were tested under four-point loading (see Figure 3); otherwise beam Wl(A) was identical With beam
W1-0.314, beam W3(A) identical with beam W3-0.314 and so on.
tt Details of web openings are given in Figures 2 and 4.
$feu = 100 mm strength cube.
Sfc' = cylinder compressive strength (300 x 150 mm)
**A = cylinder-splitting tensile strength (300 X 150 mm) in accordance with ASTM C330.

The structural idealization web openings may be calculated by using the struc-
The arguments that follow are based on the sum tural idealization of Figure 5 , which shows that the
total of the evidence obtained from the present and applied loadis transmitted to the supportmainly by a
the pilot tests(9), studied in the light of the evidence lower path ABC and partly by an upper path AEC.
obtained earlieT(-'o-12);they supersede the arguments The structural idealization suggests that the effective-
previously(9) presented. ness of the lower path should increase with the angle
The ultimate shear strength of a deep beam with 6,whilst that of the upper path should increasewith
83
Magazine of Concrete Research : Vol. 29, No. 99 :June 1977

E 4'. Consider, for the time being, that the opening

occurs at a fixed level, i.e. the dimensions k2D and
k2'Dare fixed. Then, if 4' is kept constantby keeping
the dimension k,'x constant, and 4 is progressively
l ,
l reduced by increasing the dimension k,x, it would be
reasonable to expect aprogressive reduction in ulti-
mate strength. Beams 0-0.317 to 0-0.3/10 (Figure
4a) were designed to test this argument, and the W ,
E values in Table 2 show that ultimate loads were in-
0
0
deed progressively reduced: from 420 kN for beam
.r
I1 0-0.3/7 through 380 kN and 280 kN to 210 kN for
Q beam 0-0.3/10.Ontheotherhand, in beams
0-0,3/1 to 0-0.3/6(Figure 4a), the angle 4 was kept
constant while 4' was progressively reduced; again,
the W , values in Table 2 confirm that the general
trend was a reduction fn the ultimate load.
In the absenceof the opening, the upper and lower
T paths in Figure 5 become one, which is the natural
load path joining the loading and reaction points; for
tX = 150 mm for xld = 0.2 such a solid beam,it has been shown(lo) that the ulti-
225 mm for xld = 0 . 3
mate shear strength Quit could be predicted by equa-

Ref No.
Size T Position tion l :

-
f,bD + C, X A --m2,
Y .
D
. .(l)
I .00
1 .00
I .oo 0.40 where the notationis as given at thebeginning of this
I .00
I .00 paper.
I .oo The structuralidealizationsuggests that, if the
7 0 30 0.30
8 0.50 0-20 0.50 0-40 opening is small or is so located as not to interfere
9 0.70 0.70 significantly with the natural load path, a reasonable
IO I .30 I .30
I1 0.30 I .oo 0.666
estimate of the ultimate strength should be obtain-
I2 0.30 0.30 0.666 able from equation l . This was indeed supportedby
13 1-00 I .00 0.666
14
15
16
0.30
0-30
I .00
0.20 I .00
0.30
I .oo

Figure 2: Reference Nos., sizes and positions of the openings in

the beams tested.
I 0.134
0.134
0.134
thetest results. Forexample, in beams0-0.310,
0-0.3/1,0-0.3112,0-0.3/14and0-0.2/0,the
openings did not interfere significantly with the nat-
ural load paths; hence in Table 2 the ultimate loads
W 2 (and W,) for these five beams have been com-
puted from equation 1. The W,/W,(and W,/W,) ratios
were reasonably close to unity, showing that equation
1 was applicable.
% G If the opening interrupts the natural load path, the
f~veat2 2 5 = 1125 mm
j ultimate strength equation takes themodified form:

7 f,bk,D + C A C, Assin2a,
D
= w,/2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 )
where again the notationis as explainedat thebegin-
ning of the paper. Note that y , is now the depth at
which a typical reinforcing bar intersects the strut
EA of the upper path or the strut CB of the lower
path as the case may be (Figure6), and a,is the angle

7 between the typical bar and the strutEA or CB; the

lines EA and CB represent potential

proposed in the pilot for

the
critical diagonal
cracks in the beam. Equation2 is different from that
reasons tobe
Figure 3: Four-point loading -for beams W1 (A), W3(A),
W4(A) and W7(A). given later in the "Discussion".

84
 Structural idealization for deep beams with web openings

T A B L E 2 : Measured and computed ultimate loads.

Ref. No.
of beam*

GROUP 0 BEAMS

0 - 0 , 310 595 t651 0.91 t651 0.91

0-0.311 460 t637 0.72 t637 0.72
0-0.312 390 295 1.32 346 1.13
0-0.313 280 275 1.02 325 0.86
0-0.314 260 275 0.95 325 0.80
0-0.315 200 278 0.72 327 0.61
0-0.3/6 250 287 0.83 336 0.74
0-0.317 420 396 1.06 391 1.07
0-0.318 380 341 1.11 347 1.10
0-0.319 280 325 0.86 349 0.80
0-0.3110 210 243 0.86 316 0.66
0-0.311 1 360 467 0.77 480 0.75
@0.3112 560 t 707 0.79 t 707 0.79
0-0.3113 300 483 0.62 495 0.61
0-0.3114 560 t 664 0.84 t 664 0.84
0-0.3115 260 183 1 43 224 1.16
0-0.3116 195 48 4.00 181 1.08

0-0.210 655 t 720 0.90 t 720 0.90

0-0.214 360 356 1.01 379 0.95
0-0.2113 500 507 0.99 515 0.95
0-0.2116 340 92 3.70 207 1.64

GROUP W BEAMS
l
w1-0.314 400 ** **
W2-0.314 490 ** **
W3-0.3/4 560 557 1.01 722 0-78
w4-0.314 660 791 0.83 862 0.76
w5-0.314 370 ** **
W6-0.314 825 798 1.03 734 1.12
W7-0.314 630 536 1.18 481 1.31

475 *L **
500 424 1.17 527 0.95
650 797 0.82 867 0-75
670 542 1.24 487 1.38

*Beam notation as in Table 1 .

t W , and W , values for the five beams so marked were computed from equation 1.
**The web reinforcements in these beams did not satisfy the detailing requirements that both the diagonal-cracking regions, above
and below an opening, must be protected.
ttWith C , pre-multiplied by A .

In the first term on the right-hand side of equation The second term on the right-hand side of equation
2, the quantity Cf,bk,D (= C & b m sin +) is a mea- 2 represents the contributionof the reinforcement to
sure of the load-carrying capacity of the strut CB of the shear strength of the beam; the reinforcement
the lower path in Figure5 , and the factor [ 1 - 0.35 restrains the propagation and widening of any critical
k , x / k 2 D ]allows for the experimental observation of diagonal cracks along EA and CB. Unless arrested,
the way in which the load capacity varied with cot 4, such propagation and wideningwill result in the end
where 4 is the inclination of the strut CB to the portion of the beam moving outwards in a predom-
horizontal. The first term is therefore a semi-empir- inantly rotational motion about the loading point.
ical expression for the capacity of the lower path; The structural idealization (Figure 6) explains why
when this capacity is reached, the strut CBfails in a the ability of a reinforcing bar to restrain such rota-
splitting mode (hence thesplitting strengthf t is used) tion increases with the distance y,.
resulting in the formation of a so-called critical diag- In Table 2, the computed ultimate loads are com-
onal crack along CB. pared with measured values, and it can be seen that,
85
Magazine of Concrete Research : Vol. 29, Np. 99 :June 1977

Figure 4: Crack patterns of the beams at failure. Beams

numbered as in Table 1. The circled numbers on the beams
themselves show the se uence in which the cracks were observed;
the other numbers on &e beams show the load, in units of 10
kN, at which the extent of each crack was as marked.

(a) The first nine type 0 beams

0-0*3/0 0 - 0*3/1 0-0.312

86
 Structural idealization for deep beams with web openings

(b) The remaining type 0 beams

0-0*3/9 0-0 - 3/ 10 0-0*3/II

I
0-0-3112 0 - 0.3/13 0 -0*3/
l4

0-0.3/15 0 - 0 - 3 / 16 0-0.2/0

0 - 0 - 2/4 0 - 0-2/13 0-0*2/16

87
Magazine of Concrete Research : Vol. 29, No. 99 :June 1977

Figurz 4 continued:
(c) The type W beams

~ ~ ~~

W I- 0.3 /4 W2-0*3/4 W3- 0.314

W4 -0,314 W 5- 0.3 / 4 W6-0.3/4

w7-0.314

88
 Structural idealization for deep beams with web openings

W e

Figure 5: The structural idealization. Figure 6: Meanings of symbols.

except for beams0-0.3/16 and 0-0.2/16 (see item 3 stances,however,thelowerpathmightbemuch

in "Discussion"), the agreement is generally good. weakerthantheupperpath.Thishappened,for
The exact analysisof reinforced concrete deep beams example, in beams 0-0.3/16 and 0-0.2/16, which
with openings presents formidable problem^"^', but were designed as extremecases to test the structural
for practical design purposes equation 2 could be a idealization. It is clear from Figure4b that the lower
useful tool. paths in these beams were weak relative to the upper
paths; Table 2 shows that forsuch beams equation2
is grossly conservative. In any event, however, the
Discussion absolute shear strengthof a beamis likely to be low if
(l.) Equation 2 presented herediffers from that pro- (see Figure 5) the values of k , and the angle 4 =
posed in the pilotstudy ( 9 ) ; in the pilot study, the cot-' (k,x/k,D)are low. It is suggested that, in design,
equation took the form: k , should be kept notless than, say,0.2 and the angle
4 not less than about 30".
f,bk,D + Y
C, X A-sin2u
D
(4) The web reinforcement must be so arranged as to
protect both the diagonal-crackingregions,above
= w3/2 ................................ 43) and below the opening. The measured loads W1 in
Table 2, studied in conjunction with the web-rein-
where both y and a are measured with reference to forcement patternsin Figure 1 and thecrack patterns
the natural load path, w h c h often bears little relation in Figure 4c, show that:
to the critical diagonal cracks in a beam with open- (a) where the web reinforcement protected only the
ings. This anomaly has now been corrected in equa- lower diagonal-cracking region (Figure 1: type
tion 2. W1) or only the upper diagonal-cracking region
(2) In equation 2 , the coefficient C , in the second (Figure 1: type W2), or where itis used to trim
term is pre-multiplied by a factor h . The upper and the opening (type WS), the ultimate loads were
lowerpaths of a beam withwebopenings are low (Table 2): beam W1-0.3/4 (400 kN), beam
less efficient in carrying loads than the natural load W2-0.3/4 (490 kN), beam W5-0.3/4 (370 kN).
path of a solidbeam, and are more sensitive to imper- (b) where the web reinforcement protected both the
fections - such as diagonal cracks. Hence the effect upper and the lower regions, as in beams W4-
of web reinforcement is more pronouncedin a beam 0,3/4 (660 kN), W6-0.314 (825 kN) and W7-
with openings, but proper detailing is critically im- 0,3/4 (630 kN), theultimate loads were high.
portant (see item 4 below). (c) web steelwas most efficiently used inthe formof
(3) In equation 2, the concrete contribution, as re- type W6 inclined web reinforcement.The group
presented by the first term on the right-hand side,is W beams each had the same amount of web steel,
based on the capacity of the lower path, which is but the ultimate load of beamW6-0.3/4 was
normally the main load path. Underspecial circum- much higher than those of the others.
89

.
Magazine of Concrete Research : Vol. 29, No. 99 :June 1977

( 5 ) In Table2, W., values arenot shownagainst there are restrictions on the over-all dimensions of
beams W1-0.3/4, W2-0.314, W5-0.314 and Wl(A). the beam, and an adequate ultimate strength is the
As explained in 4(a) above, the web reinforcements main concern, type W6 may be the best choice.
in these beams did not protect both the regions above(6) In the design of shallow beams andof the major-
and below the opening.If, for example, equation2 is ity of solid deep beams"','*), it is usually necessary
applied to beam W1-0.314, the computedW , will be to consider shear for the ultimate limit state only. In
over 1000 kN; this artificially high computed load the design of deep beams with openings, however,
arises from the fact that the web-reinforcement de- shear may also be animportant consideration for the
tailing (Figure l : type W l ) was such as to leave the serviceability limit state of cracking (see Appendix
upper region relatively weak and hence the potential 1).
capacity of thelowerpath couldnot be realized
before the collapse of the beam occurred.
(6) Beams Wl(A), W3(A), W4(A) and W7(A) were ACKNOWLEDGEMENT
tested under four-point loading to simulate thedis-
The experiments were carried out by the second
tributed-load condition, as shown in Figure 3. The
results in Table 2 show that equation 2 may also be author at the University of Nottingham, using the deep-
beam facilities at the civil engineering laboratory. Sin-
used for this loading condition. In equation 2, the
cere thanks are due to Professor R. C. Coates, who
dimensions k,x and k,D are independentof the load-
supervisedtheproject, for hisencouragementand
ing condition. To define all they, and c y I values, it is
support.
only necessary to choose a reasonable line to repre-
sent the strutE A in Figure 6; it is suggested that, for
this purpose, the appliedloading may be replaced by
two statically equivalent point loads.
APPENDIX 1
(7) With reference to equation 2, it is reasonable t o
expect that, for agiven beam, there is an upper limit Crack control
to the shear strength, irrespectiveof how much web Test results are summarized in Figure 7. Several observations
steel is used. In the tests, that limit was not reached are useful in design.
at a web steel ratio of 1.2 % , which already repre- (1) The test results have confirmed that maximum crack widths
increased with theextentto which anopeninginterruptsthe
sented a rather heavy web reinforcement for a deep naturalloadpathjoiningthe loading and reactionpoints. For
beam. example, Figure 7a illustrated clearly how progressive increasesin
theextentof such interruption led to progressive increases in
Design hints maximum crack widths.
(2) The crack-width curves in Figure 7 are drawn against a grid
(1) Wherever possible, web openings should bekept mesh of 0 3 mm unit width. This represents a maximum crack
clear of the natural load pathjoining the loading and width limit that is commonly accepted in design. For beams with
reaction points. If the opening is reasonably clear of no web reinforcement, Figures 7a, b and c show that this limiting
the natural load path, the ultimate shear strength may width wasin many instances exceeded at loads below 50 o/o of
be calculated from equation 1 as though there were the ultimate.
(3) Web reinforcement could be highlyeffective in controlling
no openings. crack widths, but properdetailing is an important requirement: the
(2) If the opening intercepts the natural load path, web reinforcement must protectboththe diagonal-cracking
the designer should ensure that the factor k , is not less regions, above and below the opening. Web reinforcement types
than about 0.2 and the angle cot-' (k,x/k,D) not less W4, W6 and W7(Figure 1) satisfied this requirement, and Figure
than about30". The ultimate shear strengthmay then 7d shows that the maximum crack widths in beams W4-0.314,
W6-0-314 and W7-0 3/4 remained small even at comparatively
be calculated from equation 2. high loads. On the other hand, web reinforcement types W1, W2
(3) It is possible to increase substantially the ulti- and W5 didnot satisfy the above-mentioned detailing require-
mate shear strengthby providing web reinforcement. ment, and Figure 7d shows that they were inefficient in crack
In detailing the web reinforcement, however, the de- control; in beams W1-0 314, W2-0.314 and W5-0.314, the 0.3 mm
signer should ensure that both the regions above and crack width limit was reached at comparatively low loads. Web
reinforcement type W3 was designed as an intermediate case to
below the opening ate protected. Web reinforcementtest the detailing criterion, and the test resultsshow that it had an
not meeting this requirement should be disregarded intermediate efficiency.
when equation 2 is being used.
(4) Trimming web openings with loops of reinforce- REFERENCES
ment has little beneficial effect upon ultimate shear 1. BRITISHSTANDARDSINSTITUTION. CP 110: Part 1: 1972. The
strengths (or upon crack control - see Appendix 1). ~tructuraluse of concrete. London. pp. 154.
( 5 ) Inclined web reinforcement of type W6 (Figure 2. COMIT!? EUROPEEN DLJ BETON AND FEDERATIONINTERNATIONALE
1) is very effective for increasing the ultimate shear DE LA PRECONTRAINTE. CEB-FIP International recommenda-
tions for the design and construction of concrete strucmres.
strength (and for crack control - see Appendix 1). pp. 80. Appendix 3: Internutionul recommendations for the
This typeof reinforcement is likely to be mote expen- design and construction of deep beams. London, Cement and
sive to bend and fix than others. However, where Concrete Association, 1970. pp. 17-24.

90
 Structural idealization for-deep beams with web openings

3. ACI COMMITTEE 318. Bui/ding code requirementsfor reinforced

M03mm concrete: ACI 318-71. Section 1 1 . 9 - Special provision for
deep beams. Detroit, AmericanConcreteInstitute, 1971.
pp. 78.
4. OVE ARUP & PARTNERS. The design of deep beams inrein-
forced concrete. London, Constiuction Industry Research and
Information Association, January1977.pp.131.CIRIA
Guide 2.
5. STEVENS, A . and KONG, F . K. Reinforced concrete deepbeams and
CIRIA's draftdesignguide. Paper presented at Cambridge
No web steel; xld = 0.3; openings 0 to 6 University Mechanics Colloquium, 8 May 1975.
6. ALBRITTON, G. E. Review of literature pertaining tothe analysis
of deep beams. Vicksburg, U S . Army Engineer Waterways
ExperimentStation,November1965.pp. 80. Technical
Report No. 1-701.
7. ANON. Bibliography on deep beams. London,Cement and
Concrete Association, 1969. pp. 8.Library Bibliography No.
Ch. 71.
8. KONG, F. K . , ROBINS, P. J . and SHARP, G. R. The design of rein-
forced concretedeep beams in current practice. TheStructural
Engineer. Vol. 53, No. 4. April 1975. pp. 173-180.
9. KONG, F. K . , and SHARP, G . R. Shear strength of lightweight
(b) No web steel; xld = 0.3; openings 7 to 13 reinforcedconcrete deep beams with web openings. The
Structural Engineer. Vol, 51, No. 8. August 1973.pp.
267-275.
10. KONG, F. K . , ROBINS, P. I . , SINGH, A. and SHARP,G. R . Shear
analysis and design of reinforced concrete deep beams. The
Structural Engineer. Vol. 50, No. 10.October1972.pp.
405-409.
11. KONG, F. K . and SINGH, A. Diagonal cracking and ultimate loads
of lightweight concrete deep beams. Journal of the American
Concrete Institute. Proceedings Vol. 69, No. 8. August 1972.
pp. 513-521.
12. KONG, F, K . and SINGH, A. Shear strength of lightweight con-
(c) No web steel; xld = 0.3 or 0.2; openings 0, 4, 13 to 16 crete deep beams subjected to repeated loading. Shear in
reinforced concrete. Detroit, AmericanConcreteInstitute,
700
1974.ACI-ASCE Special Publication SP-42. Vol. 2. pp.
600 - 0 3rnm 461-476.
13. KONG, F. K . and KUBIK, L. A. Discussion of: Collapse load of
deep reinforced concrete beams by P. KUMAR. Magazine of
Concrete Research. Vol. 29, No. 98. March 1977. pp. 42-43.

(d) Web steel as in Figure l ; xld = 0.3; opening 4

in the handsof
Contributions discussingthe above paper should be
Figure 7: Maximum crack widths- beam notation as in TableI. the Editor not later than 31 December 1977.

91