Proceedings of the Institution of

Civil Engineers
Structures & Buildings 156
August 2003 Issue SB3
Pages 319331

Paper 12985 N. K. Subedi P. S. Baglin

Reader, Department of Civil Research Fellow, Department
Received 08/07/2002 Engineering, University of of Civil Engineering,
Accepted 25/05/2002 Dundee University of Dundee
Keywords:
columns/concrete structures/design
methods & aids/slabs & plates

Design of slabcolumn junctions in flat slabs

N. K. Subedi and P. S. Baglin

In a flat-slab type construction the most vulnerable areas
number of arms of crossbar/shearhead
from the point of view of safety of the structure are the  strength reduction factor for flexure ( 09)
junctions between the slabs and the columns. As
architects and engineers strive to achieve increasingly 1. INTRODUCTION
slender forms of construction the vulnerability of the Flat slabs are a popular form of construction in concrete
slabs to punching increases. The existing methods of buildings combining aesthetics and economy. The form of
detailing for punching shear, which range from stirrups, construction provides large uninterrupted floor areas with
proprietary prefabricated systems of shear ladders and minimum construction depth at each floor level. In particular,
shear hoops to stud rails, do not provide absolute the flat-plate type of construction has considerable appeal to
guarantees of safety. In this paper a new concept for the both architects and structural engineers as thinner slabs are
design of slabcolumn junctions in flat slabs is put employed without the column heads providing clear spaces
forward. The new system, which utilises solid steel along the column lines for service ducts. There are two
sections and plate and is described as the NUUL system, structural problems associated with this form of construction:
aims to eliminate the possibility of punching altogether. excessive deflection at the mid-span, and increased punching
The preliminary study shows that the new composite shear force at the slabcolumn junctions. The problem of
system is robust and has the potential for developing excessive deflection can be overcome by using methods such
into a practical method of detailing against punching as prestressing, but a suitable form of reinforcement for
forces in flat slabs. punching resistance is still a major problem.

NOTATION Over the last few decades various methods of detailing for
Ac area of column section punching resistance have evolved around the world, and they
Asc area of column reinforcement have been incorporated in standards and codes of practice such
1 2
bo critical perimeter of slab as ACI and British Standards. The most common forms are as
3,4
c1 dimension of square column follows.
d effective depth of slab
Ec , E s moduli of elasticity: concrete and steel (a) Shear stirrups: conventional reinforcement for shear at
f c9 concrete cylinder compressive strength critical and successive perimeters as recommended in BS
2
f cu characteristic strength of concrete 8110, or in the form of a cross, T or L shape as
1
f cd design value of cylinder strength recommended by the ACI. Other prefabricated proprietary
f ck characteristic cylinder strength systems in this category are shear ladders and shear hoops.
fy characteristic strength of reinforcement (b) Stud rails: prefabricated metal studs welded onto a metal
h1 , t 1 , t 2 , w dimensions of effective section strip and arranged in radial configuration in plan from the
hv depth of steel crossbar plus plate composite centre of the column have also been used as shear
section reinforcement. This method of detailing is popular in other
I c, I s second moments of area: concrete and steel parts of Europe.
lv length of crossbar/shearhead (c) Shearheads: a third category of reinforcement at slab
Mp plastic moment of section column junctions uses structural steel sections welded
m modular ratio together into a grid placed around the column or through
py characteristic strength of steel section the column itself. ACI is the only code that gives a design
1
T1 , T2 , T3 forces in section method for shearheads.
U0 , U1 perimeters
VRd,max maximum punching shear stress In general, there is a considerable amount of literature on the
VuACI ultimate shear according to ACI subject of punching reinforcement, and it is not the intention
Vu test , Vu analysis ultimate shear forces: test and analysis here to go through previous work.
y1 , y2 dimensions
v relative flexural stiffness: steel/concrete Despite the vast amount of study on the punching problem,

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 319
 occasional failure of flat slabs attributed directly to lack of
3 2 1 1 2 3
punching resistance still occurs, such as the failure of a multi-
storey car park in Wolverhampton, England, in 1997. Failure
due to punching shear is often catastrophic, causing loss of
lives and property, and such failure must be avoided as far as 1.5d 1.5d
possible. The choice of a particular system of detailing in
design appears to rest on cost-effectiveness and not on the
structural efficiency of one system against another. The current (a)
3,4
Best Practice Guides recommend the use of prefabricated
systems, but do not appear to address the main structural issue,
Separation of top layer
which is: Do the recommended systems provide extra safety to ---dome effect
avoid the vulnerability of slabcolumn junctions with respect
to punching?

In view of the proliferation of numerous proprietary systems,

and in the light of some unpleasant collapses of flat-slab Initial inclined crack
Separation along horizontal plane
structures still being reported from time to time, it is
considered that there is a lack of fundamental concepts with
regard to the design of slabcolumn junctions. (b)

The design of junctions between the slab and the columns in a

Fig. 1. Modes of failure: (a) conventionally reinforced slab; (b)
flat-slab construction must be robust enough to avoid dome effect
punching failure. This research focuses on the fundamental
concept of the design of slabcolumn junctions in flat slabs.
The main objective of the research is to develop a design
concept that will eliminate punching failure at slabcolumn
junctions.
junctions to reinforce against punching force. The concept is
2. DESIGN CONCEPT called the NUUL system. It is not a proprietary system. The
In flat-slab structures the possibility of failure due to punching system (Fig. 2) consists of crossbars welded together in a
shear should be completely eliminated. If a detailing system cruciform shape, which again is welded onto a bottom plate.
could be put forward to eliminate punching shear failure, then The plate can be square or rectangular (with or without corners
the ultimate failure of the structure would depend on the trimmed), or circular. The plate forms the soffit of the slab
flexural capacity of the slabs, and thus the design would be around the column. The plate has a hole cut out at the centre
based on the flexural mode of failure. The flexural mode of to allow the column bars to pass through. The main component
failure is a desirable mode from the point of view of safety. of the NUUL system is shown in Fig. 2(b). The level of the plate
can be lifted to accommodate a fire protection layer flush with
2.1. Failure modes and effectiveness of conventional the soffit. The steel crossbars forming the cruciform will pass
reinforcement over the column. The plate provides the anchorage for the
The typical mode of failure for a conventionally reinforced slab cruciform and confinement to the soffit of the slab, resisting
is shown in Fig. 1(a). Punching can occur owing to the local crushing around the column. The inverted U-shaped
5
formation of any of the three types of crack 11, 22 or 33. staples (Fig. 2(b)) tie the top main reinforcement of the slab,
A further possible mode of failure is caused by an inclined preventing separation. The NUUL system was developed in
crack initiated near the column perimeter, which, when it hits stages. Earlier versions were considered without the staples, but
the top reinforcement, spreads horizontally along the found to be prone to failure by separating the top layer in a
reinforcement and causes separation of the top layer of dome effect. The current system was devised for an internal
mat/concrete in a dome effect (Fig. 1(b)). Some variations of column of square section supporting approximately square slab
Fig. 1 have also been observed in other forms of reinforcement. panels in all directions.

Experimental results for conventional reinforcement such as 3. EXPERIMENTAL PROGRAMME

stirrups, shear hoops and various forms of stud rail have shown
that the reinforcement does not often reach the limit of its 3.1. Test specimens
capacity at failure. Although most designs are based on The experimental programme consisted of six specimens,
reinforcement reaching the ultimate strength, this is often not including one conventionally reinforced control, as described
the case. The overall appraisal from the existing detailing in Table 1. The slabs were either 1300 mm or 1500 mm square
method suggests that: (a) punching could happen anywhere with 150 mm 3 150 mm column stubs cast as an integral part
within the vicinity of the column; (b) the existing system of of the specimen. In both cases the span of the slab was
reinforcement may not behave as effectively as expected. maintained at 1150 mm. The column stubs projected above and
below the slab. The column was reinforced with 4T16 bars and
2.2. New concept R6 ties at 50 mm centres. Reinforcement for the slab was as
A new concept of detailing that utilises steel plate, solid steel detailed in Table 1. Cover to all faces was 20 mm. The
crossbars and staples was developed to be used at slabcolumn cruciform and plate assemblies of various specimens are shown

320 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 welded lap bars in order to simplify the anchorage problem for
the bottom bars onto the plate. The spacing between these bars
Top flexural
main bars represents the design spacing of the bottom bars. Short length
Slab intermittent welding was used to fix these bond bars on the
150 mm
deep
plate. A typical detail is shown in Fig. 4(a). In the case of SP4
Bottom steel plate Cruciform and SP5 no such bars were provided. All crossbars forming the
Column
cruciform were 16 mm 3 90 mm with the configuration shown
in Fig. 3. The configuration of the crossbars for SP5, and the
inverted U-shaped staples for tying down the top bars, are
shown in Fig. 4(b).

(a)
3.2. Testing detail
The test was carried out by applying load to the column from
the underside of the slab (Fig. 5), and effecting punching to the
slab. The slab was supported on roller supports at a square
perimeter of side 1150 mm by means of rigid beams tied down
to the strong floor of the laboratory. Typically the response of
the slab under the punching load started with radial cracks
from the perimeter of the column at the top tension surface.
The cracks continued to progress radially while cracks parallel
to the column perimeter also started to form. At any stage the
radial cracks were more prominent than the perimeter ones.
The upward deformation of the slab was monitored using an
array of transducers. The progress of cracks in terms of both
width and length continued until spalling of the concrete was
observed around the perimeter of the plate on the soffit of the
slab. The spalling and further deformation of the slab
(b) continued until full punching and the formation of a large
crack around the plate limited the load-carrying capacity of
Fig. 2. NUUL system: (a) slabcolumn junction detail; (b) the slab. At this stage the failure was completed. Views of the
tapered cruciform, plate and staples failure of the slabs are shown in Fig. 6. The test results and the
observed modes of failure are given in Table 2.

in Fig. 3. For the experimental purpose five different details 4. DESIGN CALCULATIONS
were chosen for the base assembly consisting of crossbars, the
plate and the staples. 4.1. ACI provisions for shearhead design
Currently the ACI has provisions for the use of steel I- or
An additional feature of specimens SP1SP3 was the channel-shaped sections as shear reinforcement in flat slabs.
incorporation of short, 12 mm diameter, high-yield bars welded They are identified under the heading shearheads. There are no
2 6
onto the plate surface for the anchorage of the bottom bars in similar recommendations in either BS 8110 or EC2 for the
the slab. The idea of these bond bars was to provide pre- use of structural sections or solid sections as shear

Specimen Dimensions: mm Flexural reinforcement Shear reinforcement

Slab Column Top Bottom

(downstand)

Control, C 1300 square 320 T16 at 100 km m c=c T12 at 200 mm c=c Conventional links: BS 8110
138 deep Both ways Both ways Shear stress 5 N=mm2 around
column perimeter
SP1 1300 square 320 T16 at 100 c=c T12 at 100 c=c 800 mm square plate
150 deep Both ways Both ways 800 mm cruciform
SP2 1500 square 150 T16 at 100 c=c T12 at 100 c=c Trimmed plate
150 deep Both ways Both ways 400 mm cruciform
SP3 1500 square 150 T16 at 100 c=c T12 at 100 c=c Trimmed plate
150 deep Both ways Both ways 700 mm cruciform
SP4 1500 square 150 T16 at 100 c=c T12 at 100 c=c Trimmed plate, staple tie
150 deep Both ways Both ways 400 mm cruciform
SP5 1500 square 150 T16 at 100 c=c T12 at 100 c=c Trimmed plate, staple tie
150 deep Both ways Both ways 400 mm tapered cruciform
kicker cut-out (profiled)

Table 1. Test specimens

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 321
 12 dia. @ 100 c/c bars
welded onto the plate

Crossbar
90 16

Central cut-out
150 150 for 800 800 SP1
column square plate

Bottom bars
12 dia. @ 100 c/c

Crossbars
Crossbars 700 long
400 long

148
5
10
Bond bars Edge-stiffening

0
39
for lapping bar
with bottom bars

0
60
5
10

SP2

SP3
700

Crossbars
400 long
60

Crossbars
0

400 long
148

Staples
Staples

SP4 SP5

Cruciform detail Cruciform detail

Fig. 3. Plan views of crossbars, cruciform and plate assembly (dimensions in mm)

322 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 (c) Once requirements (a) and (b) are satisfied, negative slab
reinforcement may be reduced in proportion to the moment
Bottom bars
12 dia. @ 100 c/c contribution of the shearhead at the design section.

/
Edge-stiffening

////
/

////
Point (c) is not a requirement, and since it works for safer

/
/
bar

////
structures, it is not considered important at this stage. Two of

////
////
/
/
/
/
the possible arrangements for interior shearheads are shown in
////

////

////

////
/

Fig. 7 together with the assumed critical sections for the cases
/
/
/

////
with and without the shearheads.
////
////

Crossbars
////

////

400 long
The derivation of the required plastic moment, M p , for each
Bond bars 7
for connecting arm is based on the following assumptions.
bottom bars
(a)
(a) According to tests, embedded structural sections have the
effect of moving the critical section out away from the
Cover to column (Fig. 7). The critical section is assumed to be
staple 10 mm
Staples located at a distance equal to three-quarters of the
projection beyond the face of the supporting column.
h 150 mm

(b) Throughout most of the length of a shearhead arm the

d 114 mm
Average

shear is constant.
(c) The part of the total shear carried by the shearhead arm is
Staple dia. 10 mm proportional to v , its relative flexural stiffness, compared
with that of the surrounding concrete section
(b)

Fig. 4. (a) Bond bars positions and welding (dimensions in Es Is

mm); (b) staples arrangement 1 v
Ec Ic

For the calculation of I c the effective width of the concrete

section is taken as c1 d: that is, the width at the critical
1150 section. Properties are calculated for the cracked,
transformed section, including the shearhead. The
minimum value of v required is equal to 015.
(d) If the total shear at the support is Vc , the constant shear
150
force in each arm (v Vc )=
.
Test specimen: slab
(e) Any load above the concrete cracking shear, Vc , is carried
175 Column 175
1500 by the steel shearhead.
150 150
( f ) The shearhead arm within a distance from the column face
equal to hv (the depth of the arm) assumes a nearly
constant value greater than (v Vc )=
. This increased value
Strong floor is equal to total shear force per arm, Vu =
, minus the shear
carried by the partially cracked concrete: that is

Fig. 5. Test arrangement: 1500 mm square slab (dimensions in Vu Vc

mm)
(1
v )

(g) Also, according to tests, the cracking shear force, Vc , is

reinforcement in concrete structures. ACI suggests that three about one-half of the ultimate shear, Vu : that is
basic criteria must be considered in the design of shearhead
reinforcement for connections transferring shear due to gravity
load. The following three points are directly quoted from the Vu
2 Vc
ACI.1 2

(a) The shearhead must have a minimum flexural strength so

that the required shear strength of the slab is reached
before the flexural strength of the shearhead is exceeded. Based on the above assumptions, the shear force and moment
(b) The shear stress in the slab at the end of the shearhead diagram for the shearhead arm is constructed (Fig. 8). The
p
reinforcement (critical perimeter) must be limited to f c9 =3. plastic moment required for each arm is computed by

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 323
 Fig. 6. Slab failures: (a) SP1: failure due to column bursting; (b) SP2=SP3: top surface dome effect; (c) SP2=SP3: separation of top mat
(part concrete removed); (d) SP4=SP5: typical failure mode, punching around the plate perimeter (view of the underside of the
slabcolumn junction)

Specimen Concrete, Failure load: Failure mode Shear stress: N=mm2 Column
fcu : N=mm2 kN capacity :
At column At plate kN
perimeter perimeter

Control, C 72 395 Shear failure around column perimeter 645 1416

SP1 62 800 Bursting of punching column 1170 219 1271
SP2 74 767 Separation of top mat followed by 1121 313 1445
failure of cruciform/plate weld
SP3 67 861 Separation of top mat followed by 1259 351 1343
shear failure around plate
SP4 73 879 Shear failure around plate perimeter 1285 358 1431
SP5 59 878 Shear failure around plate perimeter 1284 358 1227
 Column capacity 0:67 f (A
A ) A f :
cu c sc sc y

Table 2. Test results



Vu c1 other geometrical parameters will be known. Hence equation
3 M p hv v lv
(3) can be used to calculate the ultimate capacity of the shear
2
2
force, Vu .

in which  is the strength reduction factor for flexure,

normally assumed to be equal to 09. Equation (3) is the same 4.2. ACI recommendations: upper limits
1
as equation (1137) in the ACI document. For a given section, ACI recommends maximum limits on punching shear forces as
and for the given configuration of the shearhead, M p and the follows

324 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 p
c1 d (a) without shear reinforcement a value of 0:33 f c9 bo d
p
(b) with bars or wires a value of ( f c9 =2)bo d.

The critical sections for (a) and (b) will be located at d=2 from
Critical section
c1 the face of the column (Fig. 7(a))

p
(c) with shearheads a value of ( f c9 =3)bo d on an assumed
(a) critical perimeter located inside the end of shearhead
reinforcement: that is, at 34(lv
c=2) (Fig. 7(b)).

4.3. Extension of ACI recommendations to the present

lv c1/2 test specimens
d/2 In this paper the ACI recommendations have been extended to
the test specimens described here. Since the ACI
d/2 Critical section recommendations are based on experiments carried out on
I- and channel-shaped sections, strictly, the recommendations
apply to those sections only. Therefore, in extending the
application to the plate and crossbars composite sections, the
following assumptions have been made.

(a) For the calculation of the plastic moment strength the

(lv c1/2)
effective section is taken as an inverted T-section, the
width of the bottom flange being the same as the depth of
(b)
the crossbar (Fig. 9(a)).
(b) For the calculation of the relative stiffness, , the concrete
section considered is of width (c1 d), as shown in Fig.
9(b). This is the same as the width that ACI recommends.

(lv c1/2) 4.4. Results of analysis

Critical section
The analysis of the test specimens based on the above
assumptions and using equations (1) and (3) is given in Table
3. The results from the analysis are compared with the test. The
upper limits for the punching resistance according to ACI were
also computed. These are presented in Table 4. As an example
of the various steps involved, the detailed calculations for SP4
lv c1/2
are shown in Appendix 1.

5. DISCUSSION
(c)

5.1. Modes of failure

Fig. 7. Critical section for punching shear according to ACI: The test results presented in Table 2 show that even within a
(a) no shearhead, critical section at d=2 from the column face; limited number of test specimens a variety of failure modes
(b) small interior shearhead with four arms; (c) large interior were observed. The conventionally reinforced control specimen
shearhead with four arms
C failed at the weakest section around the column. The cracks
formed at an angle of 35408 with the horizontal. The links for
this slab were designed using BS 8110 for a maximum
punching shear stress of 5 N=mm2 around the column
Vu/4 Column face perimeter. The result indicated that punching is not a typical
hv
mode of failure at the notional critical section (a) d=2 from the
Mp (lv c1/2) (a)
column edge (ACI) or (b) 1:5d from the column edge (BS 8110).
Any weak section in the vicinity of the punching column is
vulnerable. Among the composite systems, SP1 failed owing to
Vu/ Vc(1 v)/ (vVc)/ ( vVu)/2 the bursting of the column itself (Fig. 6(a)). This shows two
hv
(b) things: (a) that the first specimen tested with the new detailing
revealed an enormous resistance of the slab in punching; and
Mp (b) that the column detailing was crucial. In this specimen it
(c) was found that the loaded end of the column did not distribute
the load uniformly on the section. Instead the load was
transferred directly to the column reinforcement. This was
Fig. 8. Forces in shearhead arms: (a) shearhead arm; (b) shear
force diagram; (c) moment diagram corrected in subsequent specimens, and the columns behaved
satisfactorily.

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 325
 t1

h1

t2
w h1
(a)

Cover plus
orthogonal layers
of main bars
h

d/2 Column size c1 c1 d/2

c1 d

(b)

Fig. 9. Effective sections: (a) section for calculation of plastic strength; (b) section for calculating the relative stiffness. For all test
specimens in this study, w h1 90 mm; t1 16 mm; t2 8 mm; c1 150 mm; d 114 mm; c1 d 264 mm: All main bars
16 mm dia. @ 100 mm c=c

Specimen Concrete, Concrete, Ec : Stiffness ratio, v Ultimate strength: kN Comparison

f cu : N=mm2 kN=mm2 (equation (1))
Vu analysis Vu test Vu analysis =Vu test Vu test =Vu test (control)

Control, C 72 344 395 100

SP1 62 324 0318 584 800 073 202
SP2 74 348 0300 871 767 114 194
SP3 67 334 0310 644 861 075 218
SP4 73 346 0298 871 879 099 222
SP5 59 318 0325 851 878 097 222

Is 2128:56 3 103 mm4 ; Es 205 kN=mm2 ; Ic 42 279 3 103 mm4 ; Mp 16 385 kN:mm;  0:9:

Table 3. Results and comparison

Specimens SP2 and SP3 showed that, in the absence of tie- around the plate perimeter. The bottom plate acting
down members, the top mat of the main reinforcement could compositely with the crossbars alleviated local spalling of the
separate along the horizontal plane and deform as a dome concrete around the column, thus achieving a robust behaviour
towards the later stages of loading (Figs 6(b) and (c)). However, of the joint. The failure plane moved away from the vicinity of
the punching resistance of these specimens was again the column to around the plateslab interface.
remarkable. Having observed the behaviour of SP2 and SP3, it
was decided to introduce the staples in the form of inverted The behaviour of SP4 and SP5 verified clearly that, with the
U-bars. Staples were incorporated in SP4 and SP5. Failures of composite action of the plate, crossbars and staples the joint
SP2 and SP3 were characterised by the punching of the slab was robust with no excessive cracking at the top tension face.

326 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 recommended that future test specimens should be larger,
Specimen ACI upper limit: Vu test : kN Vu test =Vu ACI
kN incorporating a range of configurations.

The comparison of results is presented in Table 3. The analysis

Control, C 390 395 101
SP1 490 800 163 results, Vu analysis , represent the extension of the ACI to the
SP2 320 767 240 plate and crossbars composite sections not previously covered
SP3 444 861 194 by any codes.
SP4 318 879 276
SP5 286 878 307 The results from ACI compare well with the test. The ratio
Vu analysis =Vu test , except for SP2, shows safe, but accurate
Table 4. Punching shear force: upper limits according to ACI predictions. The results for SP4 and SP5, in particular,
represent an excellent correlation between the analysis and
test. The ratio Vu test =Vu test (control) is a measure of the
The dome effect was controlled, and the failure was located enhancement of the punching shear resistance of the new
around the plate perimeter (Fig. 6(d)). Specimens SP4 and SP5 concept. It shows consistent enhancement and, in the case of
showed a very satisfactory behaviour overall. These two SP4 and SP5, of the order of 222 times the control value. Thus
specimens represent the NUUL system for application in the new concept is a robust system with a potential for
practice. However, more tests on similar systems but with practical application. The degree of resistance could also mean
different configurations will be required to confirm consistent that this system could eliminate the vulnerability of failure at
behaviour. the slabcolumn junction.

It is clear from the modes of failure of SP4 and SP5 that, in a In Table 4 the upper limits for punching shear forces according
well-detailed system, the plate helps to move the mode of to ACI are safe but rather conservative. It appears that for the
failure to its perimeter. The ACI does state that the shearhead new concept utilising crossbars and plate composite sections
helps to shift the critical failure plane from d=2 from the face the upper limit could be increased. This of course should be
of the column to a perimeter near the tip of the section. confirmed with more tests.
Therefore the critical failure is assumed to lie at a distance
8
three-quarters of the shearhead length from the column face. 5.3. Punching shear resistance: status of EC2
In particular cases of I- and channel-shaped sections this Eurocode 2 for the design of concrete structures is in its final
recommendation appears to have worked satisfactorily. The draft form. It is due to be published in the final Euronorm (EN)
inclusion of the bottom plate, however, brings more composite form in the first half of 2003. In the current version, EC2 does
action between the crossbars and the concrete. It has the not include specific rules for structural sections or solid bars as
positive effect of protecting any local damage in the slab possible shear reinforcement. It does however suggest, under
around the column on the compression face and also moving punching resistance of slabs, that: For types of shear
the failure plane to the plate perimeter. reinforcement other than links, bent up-bars or mesh, the
design shear resistance may be determined by tests.

5.2. Comparison of analysis and test results For the current study there are two points of interest from EC2:
The failure load and the computed shear stresses for test (a) the basic control parameter or the critical perimeter for
specimens SP1SP5 presented in Table 2 show a considerable punching shear check is suggested at a distance 2:0d from the
robustness of the reinforcing system. Around the column loaded area (edge of column); (b) adjacent to the column, the
perimeter, at failure, the shear stresses are between 1121 and punching shear resistance is limited to a maximum value given
12:85 N=mm2 . These are very high values for the punching by Rd,max 0:5f cd , where 0:6(1
f ck =250), f cd is the
shear stresses. Around the plate perimeter where the failure design value of concrete cylinder compressive strength, and f ck
occurred in SP3SP5, the shear stress is about 3:5 N=mm2 . In is the characteristic cylinder strength at 28 days. In Table 5 a
the current test set-up the edges of the plates were close to the comparison is made between the punching shear stress at the
perimeter supports and hence some enhancement is expected column perimeter calculated from the experiments with those
owing to the close proximity of the supports. It is from EC2. For this comparison both f ck and f cd are taken as 0:8f cu .

For the control specimen, C,

Specimen Concrete, fcu : Assumed f cd : Assumed fck : Shear stress at column the maximum shear stress of
N=mm2 N=mm2 N=mm2 perimeter: N=mm2
13:3 N=mm2 permitted by
Test EC2, Rd,max EC2 appears to be much
higher than the experiment.
Control, C 72 576 576 645 133 Bearing in mind that the
SP1 62 496 496 1170 119 control specimen was
SP2 74 592 592 1121 136 designed for a shear stress of
SP3 67 536 536 1259 123 5 N=mm2 around the column
SP4 73 584 584 1285 134
perimeter (approximately
SP5 59 472 472 1284 115
2:12 N=mm2 at the critical
Table 5. Shear stress at column perimeter: experiment and EC2 perimeter 1:5d), the shear
stress of 6:45 N=mm2 at

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 327
 failure compares well. However, the maximum value of uniformly and efficiently. Verification consists in making sure
13:3 N=mm2 appears to be too high to achieve in a that the unit is placed symmetrically over the column and that
conventionally reinforced slab. For the other specimens utilising the bars are threaded through the staple ties.
the NUUL system, the maximum shear stresses appear to be
reasonable. Since in these cases the failure occurred far away
5.6. Further research
from the column perimeter, the maximum values permitted,
Further tests are essential to verify the application of the
11:513:6 N=mm2 , appear to be justified.
system over a range of configurations and boundary conditions
along the slab edges. It is also important to verify the strain
5.4. Shear strength of crossbars: simplified view
development in crossbars, plate and staples and the
A simplified method of assessing the shear strength of slab
deformation characteristics (deflections, ductility and crack
column junctions using the new concept is to estimate the
9 widths) that will determine both the serviceability and the
shear capacity of the crossbars. According to BS 5950, the
ultimate limit states. The next stage of research will address
shear capacity of a rectangular section is given by
those concerns.
Pv 0:6 py Av , in which py is the design strength and Av is
the area of the section. The crossbars used in our case were
Grade 43, py 275 N=mm2 , 90 mm deep 3 16 mm thick 6. CONCLUSIONS
sections. BS 5950 also suggests that for solid sections the area
should be multiplied by 09. Therefore the capacity of the (a) A new detailing concept for slabcolumn junctions of flat
crossbars in shear is given by slabs has been proposed, comprising a steel plate, solid
steel crossbars and staples.
Pv 4(0:6 py 0:9Av ) 4(0:6 3 275 3 0:9 3 16 3 90) N (b) From the limited number of tests carried out, the system
appears to be robust and has a high resistance to punching
855 kN
force.
(c) The extension of the ACI approach to the design of
shearheads applied to the new system gave good results.
Provided the structure works well, the junction should have at (d) By adopting the ACI approach and the simple concept of
least 855 kN capacity in punching shear. Three of the the resistance of bars in shear, a simplified design method
specimens (SP3, SP4 and SP5) demonstrated that the system could evolve.
had over 855 kN. (e) More tests over a range of configurations are required to
verify the structural behaviour and robustness of the
The NUUL system works with the crossbars passing right system.
through the column. Therefore this is a positive connection, (f ) For the plate and crossbar composite details the upper limit
ensuring the continuity of the shear reinforcement right on punching resistance suggested by the ACI appears to be
through and around the column. The structural system is conservative, bearing in mind that the ACI
simple to understand. By welding the crossbars to the plate and recommendations are primarily for detailing consisting of
by passing the main bars at the top through the staples, the structural sections only.
possibility of spalling and local damage is prevented. The plate
provides containment and full composite action with the
crossbars and mobilises their full capacity in shear. It is
considered that this simplified view will lead to a simple and 6. APPENDIX 1. CALCULATIONS FOR SLAB
clear understanding of the design method for a complicated SPECIMEN SP4
element of structure such as the slabcolumn junction. The specimen SP4 consists of trimmed plate, with staple ties
and crossbars cruciform of length 400 mm (see Fig. 10).
5.5. Cost-effectiveness and practicality of the system
The main component used in the system is the NUUL unit, 6.1. Relative stiffness
consisting of a base plate, crossbars and inverted U-bar staples. The section considered for relative stiffness calculation is at
The unit is well suited for shop fabrication. The cost of the d=2 from the edge of the column. The relative stiffness is given
materials is minimal; labour is involved in cutting the parts by
into shape and in a small amount of welding. In a building, the
size of the columns and the locationwhether internal, edge or
cornerwill govern the variation in design. The main cost of Es Is
4 v
the system is probably in the operation to lift the units into Ec Ic
position. The smaller units can be manoeuvred in position by
two people, but larger units will require lifting equipment.
However, once the units are in position, the site work is
minimal. Only the main bars require threading through the
6.1.1. Calculation of Is
staple ties. Clearly the cost-effectiveness has to be judged by
See Fig. 11. Position of centroid, y 37 mm: Is about the
comparing the time and effort required to fix other proprietary
centroidal axis
systems, some of which require a considerable time on site.

The system therefore appears to be very practical, and it I plate (90 3 83 )=12I crossbar (16 3 903 )=12
simplifies the operation of preparing slabcolumn joints

328 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 6.1.2. Calculation of Ic (Fig. 12).

5
10
(a) Locate the N-axis of the cracked section: Position of the

0
60
Critical perimeter N-axis from the soffit of the slab is y1 . Area of main bars:
according to ACI

0
16 mm dia. @ 100 mm c=c gives 2011 mm2 per 1000 mm.

39
Therefore area over 264 mm 531 mm2 .
(b) Moments of area about the soffit of the slab
5
10

148
y1 [531 3 6 (90 3 16) 6 264( y1
8)
# # # # #
Opening for (lv c1/2) reinf: m crossbar m uncracked concrete
column 93.75
c1 150 (8 3 264) 6 ]
Plan # #
lv 200 plate m
16
16 (531 3 6 3 114) (90 3 16 3 6 3 53)
90 hv 98
8 [264( y1
8)f( y1
8)=2 8g] (8 3 264 3 6 3 4)
700
Assumed effective
Plate and crossbars section for the
calculation of Mp
which yields
elevation

Fig. 10. Slab specimen SP4 (dimensions in mm). Data: 132 y21 22 386 y1
863 364 0
fcu 73 N=mm2 ; Ec 34:6 kN=mm2 ; modular ratio,
m Es =Ec 204=34:6 5:92  6
giving, for a positive root, y1 32:4 mm.
(c) Second moment of area of cracked concrete section, I c ,
16 about xx

I for crossbar (1=12) 16 3 903 972:0 3 103 mm4

90 98
I for plate (1=12) 264 3 83 11:264 3 103 mm4
53
y
I c about xx (1=3) 264 3 24:43 531 3 6 3 81:62
Uncracked conc: reinforcement
8
90
972 3 103 3 6 90 3 16 3 6 3 20:62
crossbar crossbar
Effective section
11:264 3 103 3 6
Fig. 11. Calculation of Is (dimensions in mm) plate

264 3 8 3 6 3 28:42
plate
Therefore
I c 1278:36 3 103 21 214:17 3 103

I s 972 3 103 [1440 (5337)2 ] 3840 720 (374)2 5832:0 3 103 3666:47 3 103
2128:56 3 103 mm4 67:58 3 103 10 220:73 3 103

16 dia. @ 100 c/c

Slab depth 150

16
d 114

81.6
y1 32.4

90

x N-axis x
24.4

Plate t k 8
c1 d 150 114 264

Fig. 12. Calculation of Ic (dimensions in mm)

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 329
 Therefore I c 42 279:31 3 103 mm4 . 6.3. Calculate the ultimate punching shear force, Vu ,
(d) Relative stiffness from equation (3)
SP4 data:
4, lv 200 mm, c1 150, hv 98 mm,
v 0:3,  0:9
205 3 2128:56 3 103
 0:3
34:6 3 42 279:31 3 103
V h  c 1 i
5 M p hv v I v

2
2

6.2. Plastic moment for the section

(a) Determine the equal area axis, xx (see Fig. 13).
0:9 3 16 384:5 Vu [98 0:3(200
150=2)]

A1 A2 A3
from which
16(90
y2 ) 16 y2 (90 3 8)

Vu 871 kN
Therefore y2 22:5 mm.
(b) Calculate M p (see Fig. 14).
Therefore Vu Vu analysis 871 kN.

T1 (16 3 67:5) py 6.4. Comparisons

From test, Vu test 879 kN. Therefore
T2 (16 3 22:5) py
T3 (8 3 90) py Vu analysis
0:99
Vu test

Grade 43, py 275 N=mm2 .

Take moments about the equal area axis 6.4.1 Shear stresses from test results.

(a) At the column perimeter:

M p T1 (67:5=2) T2 (22:5=2) T3 (22:5 4) Perimeter, Uo 4 3 150 600 mm
Shear stress
Vu test =(Uo d) 879 3 103 =(600 3 114) 12:85 N=mm2
On substitution, M p 16 384:5 kN:m: (b) At the plate perimeter:
Plate perimeter, U1 (4 3 390) (4 3 148) 2152 mm
16
Shear stress Vu test =(U1 d) 879 3 103 =(92 152 3 114)
3:58 N=mm2
(c) Capacity of column (see Fig. 15). The capacity of the
90 y2 A1 90 column at failure is calculated using the equation
98
x x Capacity 0:67f cu (Ac
Asc ) Asc f y
y2 A2 Data: f cu 73 N=mm2 ; Asc 4 3 201 804 mm2 ;
A3 Ac 150 3 150 22 500 mm2 :
8 Capacity of column
90
0:67 3 73(22 500
804) 804 3 460 N 1431 kN

Fig. 13. Determination of equal area axis (dimensions in mm) 6.5. ACI: upper limit for punching shear
The upper limit for punching shear using shearhead design
p
according to ACI is ( f 9c b o d)=3, in which f c9 is the concrete
cylinder strength 0:8f cu ; bo is the critical perimeter after
16 introducing the shearhead; and d is the effective depth of the
T1 slab.
67.5

x x
22.5 T2

8 T3 150

90
150

Fig. 14. Calculation of Mp (dimensions in mm) Fig. 15. Cross-section of column. Main bars 16 mm dia. HY

330 Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin
 Data: c1 150 mm; d 114 mm; lv 200 mm;
B f c9 0:8 3 73 58:4 N=mm2 ; AC 137 mm.

Critical perimeter
Therefore the perimeter, bo 8 3 AC 1096 mm:

()(lv c1/2)
p
Limiting shear force ( 58:4)(1096 3 114)=3 318 kN:
A

Pu test =Pu ACI 879=318 2:76:

d/
2
Column
REFERENCES
Crossbar 1. AMERICAN CONCRETE INSTITUTE. Building Code Requirements

c1
/
for Reinforced Concrete and Commentary. ACI, Farmington
2 Hills, MI, 2002, ACI 31802 and ACI 318R-02.
2. BRITISH STANDARDS INSTITUTION. Structural Use of Concrete
C D Part 1: Code of Practice for Design and Construction. BSI,
O Milton Keynes, 1997, BS 8110: Part 1.
3. BRITISH CEMENT ASSOCIATION. Best Practice Guides for In-Situ
()(lv c1/6) Concrete Frame Buildings: Rationalisation of Flat Slab
Reinforcement. BCA, Crowthorne, 2000.
()(lv c1/2)
4. BRITISH CEMENT ASSOCIATION. Best Practice Guides for In-Situ
Concrete Frame Buildings: Prefabricated Punching Shear
lv
Reinforcement for Reinforced Concrete Flat Slabs. BCA,
Crowthorne, 2001.
5. CHANA P. S. A prefabricated shear reinforcement system for
Fig. 16. Location of critical perimeter
flat slabs. Proceedings of the Institution of Civil Engineers
Structures and Buildings, 1993, 99, 345358.
The critical perimeter is located at a distance 34(lv
c1 =2) from 6. BRITISH STANDARDS INSTITUTION. Eurocode 2: Design of
the face of the column (points B and C in Fig. 16). Concrete Structures, Part 1: General Rules and Rules for
Buildings. BSI, Milton Keynes, 1992, DD ENV 199211:
The lengths AB or AC can be calculated from the geometry 1992.
7. WINTER G. and NILSON A. H. Design of Concrete Structures
(ninth edition). McGraw-Hill, New York, 1981.
DO AD 12(c1 d)
8. Eurocode 2: Design of Concrete Structures, Part 1: General
CO 34(lv
c1 =2) c1 =2 34(lv c1 =6) Rules and Rules for Buildings. European Standard, prEN
199211 (Draft for Stage 49).
CD CO
DO 34(lv
c1 =2
2=3d)
9. BRITISH STANDARDS INSTITUTION. Structural Use of Steelwork
p
AC [(CD)2 (AD)2 ] in Building, Part 1: Code of Practice for DesignRolled and
Welded Sections. BSI, Milton Keynes, BS 5950-1: 2000.

Please email, fax or post your discussion contributions to the secretary by 1 February 2004: email: [email protected];
fax: 44 (0)20 7799 1325; or post to Daniela Wong, Journals Department, Institution of Civil Engineers, 17 Great George Street,
London SW1P 3AA.

Structures & Buildings 156 Issue SB3 Design of slabcolumn junctions Subedi Baglin 331