1

2
3
 Two-way slabs transmit loads in two
directions (compare with one-way slabs)
 They are efficient, economical, and widely
used structural system
 In practice two-way slabs take various forms
 For relatively light loads (appt bldgs), flat
plates are used.
 For longer spans, waffle slabs (or two way
joist system) are used (formed with
fiberglass or metal “dome” forms)

4
 For heavy industrial loads, the flat slab
system shown in Figure (c) may be used
 Shear transfer to the column is accomplished
by thickening the slab near the column with
drop panels or flaring the top of the column
top to form a column capital
 Slab systems may incorporate beams
between some or all of the columns. The
resulting structure is referred to as two-way
slabs with beams.

5
 Elastic Analysis of Slabs
 Slabs are 2D structures
 The concepts involved in the elastic analysis
is discussed in chapter 2 → Action is
proportional to action effect (F=k z)
 The same principle holds for linear elastic
analysis of slabs, bearing in mind that the
analysis is much more complicated than for
linear elements

6
 Slabs may be subdivided into:
 Thick slabs → thickness greater than about
1/10th of the span (500 mm for a 5000 mm span)
 Thick slab transmit a portion of the loads as a
flat arch and have significant in-plane-
compressive forces, with the result that the
internal resisting compressive force C is larger
than the internal tensile force T.
 Thin slabs transmit a portion of the loads by
acting as a tension membrane; hence T is larger
than C
 A medium thick slab does not exhibit either arch
action or membrane action and thus T=C

7
 Figure (next slide) shows an element cut
from a medium thick, two-way slab.
 This element is acted on by the moments
shown in Figure (a) and by shears and loads
shown in Figure (b) (Figures are separated
for clarity)
 Two types of moments mx and my about axes
parallel to the edges, and twisting moments
mxy and myx about axes ⊥ to the edges.

8
 dx
dy
wdxdy

9
 NB: mx, my, mxy, and myx are moments and
twisting moments per meter width
 Vy, and Vx are forces per meter width
 δmx is change in mx over a distance of dx →
 ∂mx 
δ mx =  dx
 ∂x 
 Similarly δVy is change in Vy over a distance
of dy →
 ∂V y 
δ V y =  dy
 ∂y 
 and so on

10
  Summing vertical forces→
  ∂V y     ∂Vx  

− wdx dy + V y dx − V y +  
 dy  dx + Vx dy − Vx +  dx  dy = 0
  ∂y     ∂x  

 ∂V y   ∂Vx 
→ − wdx dy − 
 
 dydx −  dxdy = 0
 ∂y   ∂x 

 ∂V y   ∂Vx 
→   +   = −w ……..(4.1)
 ∂y   ∂x 

11
 Summing moments about lines parallel to the
x and y axes and neglecting higher order
terms gives:

 ∂m y   ∂mxy 
  +   = V y and
 ∂y   ∂x  ....(4.2)
 ∂mx   ∂m yx 
  +   = Vx
 ∂x   ∂y 

 Itcan be shown that mxy = myx (theory of

elasticity)

12
 Differentiating(4.2) and substituting in (4.1)
gives the basic equilibrium equation for
medium thick slabs:

 ∂ mx 
2  ∂ 2
m   ∂ 2
my 
 2  + 2 xy
+  = − w ...(4.3)
∂  ∂x∂y   ∂y 2 
 x     
 Thisis purely an equation of statics and
applies regardless of the behavior of the
plate material. (discuss interpretation)

13
 For an elastic plate, the deflection, z, can be
related to the applied load by means of:
 ∂4z   ∂4z   ∂4z  w
 4  + 2 2 2  +  4  = − …..….(4.4)
 ∂x   ∂x ∂y   ∂y  D
w
or ∇ z=−
4

D
Et 3
 where the plate rigidity is : D = ..(4.5)
(
12 1 − υ 2
)
 and ν is Poisson's ratio
14
D is comparable to the EI value of a unit
width of the slab
d 4z w
 Recall that for linear elements →
4
=−
dx EI
 Numerical solution of the 4th order PDE for a
UDL as solved by Navier’s method is:
1 ∞ ∞ amn mπx nπy
z = 4 ∑∑ sin sin
π D m =1 n =1  m n 
2 2 Lx Ly
 2 + 2
L 
 x Ly 
Lx L y
4 mπ x nπ y
where amn =
Lx L y ∫ ∫ w sin
0 0
Lx
sin
Ly
dy dx
15
 Soin an elastic plate analysis, Eqn. (4.4) is
solved to determine the deflection, z, and
the moments are calculated from:
∂2 z  ∂ 2 z 
mx = − D  2 + υ  2 
 ∂x  ∂y 
 ∂2 z  ∂ 2 z 
m y = − D  2 + υ  2  …..(4.6)
 ∂y  ∂x 
 ∂2z 
mxy = − D(1 − υ ) 
 ∂x∂y 
16
 Discussion about closed form solution of the
governing PDE of elastic plates
 Discussion about non-linear material
 Distribution of moments in slabs (qualitative
discussion)

17
18
 The distributions of moments will be
presented in one of two graphical treatments
 The distribution of the negative moments,
mA, or of the positive moments, mB, along
lines across the slab will be depicted as
shown in Figure (b)
 These distributions may be shown as
continuous curves, as shown by the solid
lines and shaded areas, or as a series of
steps, as shown by the dashed line.

19
 The height of the curve at any point
indicates the magnitude of the moment at
that point
 Discussion why the moments mA and mB
decrease towards the support
 Occasionally, the distribution of BMs in a
strip A-B-C across the slab will be plotted as
shown in Figure (c)
 The moments will be expressed in terms of
CwLx2, where Lx is the short dimension of the
panel. The unit is kNm/m
20
mij = α ij wd L2x

21
22
23
24
 Four or more stages:
i. Before cracking the slab acts as an elastic
plate, and for short time loads, the
deformations, stresses and strains can be
predicted from an elastic analysis.
ii. After cracking and before yielding of the
reinforcement, the slab no longer has a
constant stiffness, because the cracked
regions have a lower flexural stiffness, EI
than the uncracked regions and the slab is
no longer isotropic because the crack
pattern may differ in the two directions.

25
  Although these conditions violate the
assumptions in elastic theory, tests indicate that
the elastic theory still predicts the moments
adequately. Generally normal building slabs are
partially cracked under service loads.
iii. yielding of reinforcement eventually starts
in one or more region of high moment and
spreads through the slab as the moments
are redistributed from yielded regions to
areas that are still elastic. The progression
of yielding through a slab fixed on four
edges is illustrated in Figure (next slide)
26
27
  With further load, the regions of yielding known
as yield lines, divide the slab into a series of
trapezoidal and triangular elastic plates as shown
in Figure (d) above. The loads corresponding to
this stage of loading can be estimated by using
yield-line analysis (plastic method analysis)
iv. Although the yield lines divide the slab to
form a mechanism, the hinges jam with
increased deformation, and the slab forms
a very flat compression arch as shown in
Figure (next slide)(avail stiff support). This
stage of loading usually is not considered in
design

28
29
 Figure (next slide) shows a floor made of
simply supported planks supported by simply
supported beams. The floor carries a load of
q kN/m2.
 The moment per meter width in the planks
at section A-A is: m = ql12/8 kNm/m
 The total moment in the entire width of the
floor is: MA-A = (ql2)l12/8 kNm
 This is the familiar equation for the
maximum moment in a simply supported
floor of width l2 and span l1.
30
 l1

l2

31
 The planks apply a uniform load of ql1/2
kN/m on each beam.
 The moment at section B-B in one beam is
thus: M1b = (ql1/2)l22/8 kNm/m
 The total moment in both beams is:
MB-B = (ql1)l22/8
 It is important to note that the full load was
transferred east and west by the planks,
causing a moment equivalent to wl12/8 in the
planks where w = ql2. Then the full load was
transferred north and south by the beams,
causing a similar moment in the beams.

32
 Exactly the same thing happens in the two way
slab shown in Figure (next slide).
 The total moments required along sections A-A
and B-B are: MA-A = (ql2)l12/8 and MB-B = (ql1)l22/8
 Again, the full load was transferred east and
west and then the full load was transferred
north and south- this time by the slab in both
cases.
 This, of course always must be true regardless of
whether the structure has one-way slabs and
beams, two-way slabs or some other system

33
 l1
l2

34
 To emphasize load transfer mechanism in two
way slabs using the column supported two-way
slabs in Figure (next slide)
 If a surface load is applied, it is shared between
imaginary slab strips la in the short direction and
lb in the longer direction.
 Note that the portion of the load that is carried
by the long strips lb is delivered to the beams B1,
which in turn carries it in the short direction.
That portion of the load plus that directly
carried in the short direction by the slab strips
la, sum up to 100% of the load applied to the
panel. The same is true in the other direction

35
36
A similar situation is obtained in the flat
plate floor where broad strips of the slab
centered on the column lines in each
direction serve the same function as the
beams
 Therefore, for column supported
construction (one-way or two-way), 100% of
the applied load must be carried in each
direction, in the case of two-way beam
supported slabs, jointly by the slab and its
supporting beams

37
 The analysis used to derive the moments in
two way slabs was 1st published by Nichol in
1914.
 The derivation using rectangular columns
(instead of the original circular columns by
Nichol) will be shown.
 Assume : (1) A typical rectangular, interior
panel in a large structure and (2) that all the
panels in the structure are uniformly loaded
with the same load.

38
 The two assumptions ensure that the lines of
maximum moment, and hence the lines on
which the shears and twisting moments are
equal to zero, will be lines of symmetry in
the structure.
 This allows one to isolate the portion of the
slab shown shaded in Figure (next slide). This
portion is bounded by line of symmetry.

39
40
41
 The reactions to the vertical loads are
transmitted to the slab by shear around the
face of the columns. It is necessary to know,
or assume, the distribution of this shear to
compute the moments in this slab panel
 The maximum shear transfer occurs at the
corners of the column, with lesser amounts
transferred in the middle of the sides of the
column. For this reason we shall assume that
(3) the column reactions are concentrated at
the four corners of each column

42
 Figure (b) shows a FBD, a side view of the
slab element with the forces and moments
acting on it
 The applied load is (wl1l2/2) at the center of
the shaded panel, minus the load on the area
occupied by the column (wc1c2/2) (∴) shown
upward in the FBD is equilibrated by the
upward reaction at the corners of the
columns (wl1l2/2 – wc1c2/2).
 The total statical moment, Mo, is the sum of
the negative moment, M1, and the negative
moment, M2.

43
 The
magnitude of Mo may be obtained by
summing moments about axis A-A. →

 wl1l2  l1  wc1c2  c1
M o = M1 + M 2 =   − 
 2 4  2  4
 wl1l2 wc1c2  c1
− − 
 2 2 2
 NB: 1st term from slab load, 2nd term from -ve
load on column, 3rd term from reaction at edges
of column. After simplifications →

44
   c1 c2  c1  
 2
 wl2  2 
Mo =   l1 1 − 2 +  2  
 8    l1 l2  l1  

wl2l12
 Note that this is almost equal to →
8
full load is transmitted in the design
direction l1 by a strip width equal to the
width of the panel, i.e. l2

45
 The ACI Code has simplified this expression
slightly by replacing the term in the square
bracket with ln2, where ln is the clear span
between the faces of the columns, given by
 c c 2

ln = l1 − c1 because ln = l1 1 − 2 + 2 
2 2 1 1

 l1 l1 
differs only slightly from the terms in the
square bracket
 The statical moment Mo = wl2ln2/8 (ACI) (A)

46
 If the equilibrium of the element shown in
Figure (c) were studied, a similar equation
for Mo would result, but one having l1 and l2
interchanged and c1 and c2 interchanged
 This indicates once again that the slab in flat
plates and the slabs and supporting beams in
beam supported two-way slabs must be good
for 100% of the loading in both directions.
 Analysis of moments according to the ACI is a
unified approach that is applicable to both
flat slabs and beam-supported two-way slabs

47
 In a plate the slab is supported directly on
the columns w/o any beams. Here the
stiffest portions of the slab are those running
from column to column along the four sides
of a panel. As a result, the moments are
largest in these parts of the slab.
 Figure (next slide) illustrates the moments in
a typical interior panel of a very large slab in
which all panels are uniformly loaded with
equal loads. The slab is supported on circular
columns with a diameter c = 0.1l

48
Fig 4.5.1
49
 The largest negative and positive moments
occur in the strips spanning from column to
column in Figures 4.5.1(b) and 4.5.1(c).
 The curvatures and moment diagrams are
shown for strips along lines A-A and B-B.
 Both strips have –ve moments adjacent to
the columns and +ve moments at mid-span.
 In Figure 4.5.1(d) the moment diagram from
4.5.1(a) is re-plotted to show the average
moments over the width of the middle and
column strips
50
 The total static moment, Mo, accounted for
here is (NB: Factor× qln2 gives moment per meter
width)
 Mo = qln2[(0.122×0.5l2) + (0.041×0.5l2) +
(0.053×0.5l2) + (0.034×0.5l2)] = 0.125ql2ln2
 The distribution of moments given in Figure
(next slide) for a square slab supported on
rigid beams is shown in (a) with the moments
averaged over column-strip and middle-strip
bands in the same way as the flat-plate
moments shown earlier

51
Fig 4.5.2
52
 In addition, the sum of the beam moments
and the column-strip slab moments has been
divided by the width of the column strip and
plotted as the total column-strip moment.
 The distribution of moments in Figure
4.5.1(d) of the flat plates closely resembles
the distribution of middle-strip and total
column-strip moments in Figure 4.5.2 (a).
 An intermediate case in which the beam
stiffness, Ib, equal the stiffness, Is, of a slab
of width, l2, is shown in Figure 4.5.2 (b).

53
 Although the division of moments b/n slab
and beams differs, the distribution of the
total moments is again similar to that shown
in Figures (d) and (a)
 The slab design procedures in the ACI Code
take advantage of this similarity in the
distributions of the total moments by
presenting a unified design procedure for the
whole spectrum of slab and edge-beam
stiffness from slabs supported on isolated
columns to slabs supported on stiff beams in
two directions

54
 Two slab design procedures are allowed by
the ACI (EBCS-2). These are the direct design
method and the equivalent frame design
method. The two methods differ primarily in
the way in which the slab moments are
computed. (We will deal with the 1st only).
 The calculation of the moments in the direct
design method is based on the statical
moment Mo. (Mo = wl2ln2/8 (ACI) (A))
 In this method, the slab is considered panel
by panel, and Eq. (A) is used to compute the
total moment in each panel

55
 The statical moment is then divided up b/n
positive and negative moments, and these
are divided b/n middle strip and column
strips.
 In the equivalent frame method, the slab is
divided into a series of two-dimensional
frames, and the positive and negative
moments are computed via an elastic frame
analysis. Once the +ve and –ve moments are
known, they are divided up b/n middle strips
and column strips in exactly the same way as
in the direct design methods.

56
 Slabsare frequently built with beams from
column to column around the perimeter of
the building. These beams act to stiffen the
edge of the slab and help to reduce the
deflections of the exterior panels of the
slabs. (Very heavily loaded slabs and long-
span waffle slabs sometimes have beams
joining all columns in the structure)

57
 The effects of beam stiffness on deflections
and the distribution of moments are
expressed as a function of αf, defined as the
flexural stiffness, 4EI/l, of the beam divided
by the flexural stiffness of a width of slab
bounded by the centerlines of the adjacent
panels on each side of the beam.
 αf = (4EcbIb/l)/(4EcsIs/l)
 Since the length, l, of the beam and the slab
are equal, this quantity is simplified and
expressed in the Code (ACI) as:
 αf = (EcbIb)/(EcsIs)

58
 If there is no beam, αf = 0. (mostly the case
except at the edges where beams are provided
for stiffening edge panels)
 The sections considered in computing Ib and Is
are shown in Figure (next slide). (NB. Span
direction is l1)
 ACI, Section 14.2.4 defines a beam in monolithic
or fully composite construction as the beam
stem plus a portion of the slab on each side of
the beam extending a distance equal to the
projection of the beam above or below the slab
whichever is greater , but not greater than four
times the slab thickness (next next slide).

59
Fig. Beam and slab
sections for
calculations of αf

60
Governing
projection

61
 Example: Calculation of α for an edge beam
 A 200 mm-thick slab is provided with an edge
beam that has a total depth of 400 mm and a
width of 300 mm as shown in Figure (next
slide). The slab and beam were cast
monolithically and have the same concrete
strength and the same Ec. Compute αf.

62
 6m
200mm
400mm

300mm

300mm 200mm

200mm
200mm

3150mm
Fig: αf calculation

63
 Solution:
 αf = Ib/Is
 (1) Compute Ib: The cross section of the
beam is as shown in Figure (slide above). The
centroid of the beam is located 175 mm from
the top of the slab. → moment of inertia of
the beam is: Ib = (300×4003/12)
+(300×400)×252+ (200×2003)/12) +
(200×200)×752 = 2.0333×109 mm4
 (2) Compute Is: Is = 3150×2003/12 = 2.1×109
mm4
 (3) compute αf = 2.0333×109/2.1×109 = 0.968

64
 ACI code defines minimum thicknesses that
are generally sufficient to limit slab
deflections to acceptable values (same as in
EBCS-2). Thinner slabs can be used if it can
be shown that the computed slab deflections
will not be excessive.
 Slabs without beams between interior
columns……. (SI Version)
 Slabs with beams between the interior
supports…….. (SI Version)

65
66
 Limitations on the use of the DDM
1) there must be a minimum of 3 continuous
spans in each direction. Thus a nine-panel
structure (3 by 3) is the smallest that can
be divided.
2) rectangular panels must have a long-
span/short-span ratio not greater than 2.
one-way action predominates as the span
ratio reaches and exceeds 2
3) successive span lengths in each direction
shall not differ by more than one-third of
the longer span

67
4) columns may be offset from the basic
rectangular grid of the building by up to 0.1
times the span parallel to the offset
5) all loads must be due to gravity only. The
direct design method can not be used for
unbraced laterally loaded frames, foundation
mats, or prestressed slabs.
6) the service live load shall not exceed two
times the service dead load.
7) for a panel with beams b/n supports on all
sides, the relative stiffness of the beams in the
two ⊥ directions given by (αf1l22)/(αf2l12) shall
not be less than 0.2 or greater than 5. (α is the
beam-to-slab stiffness ratio defined earlier

68
 For design, the slab is considered to be a
series of frames in the two directions, as
shown in Figure (next slide). These frames
extend to the middle of the panels on each
side of the column
 In each span of each of the frames, it is
necessary to compute the total statical
moment Mo: Mo = qul2ln2/8; where qu =
factored load; l2 = transverse width of the
strip; ln = clear span between columns

69
Fig: Division of
slab into frames
for design

70
 Example: Compute the statical moment, Mo,
in the slab panels in Figure (next 2 slides).
The slab is 200 mm thick and supports a live
load of 4.53 kN/m2
 Sol: (1) Compute the design load: qd =
1.3×0.2×25 + 1.6×4.54 = 14.76 kN/m2
 (2) Consider panel A spanning from column 1
to column 2. Slab panel A is shown shaded in
Figure (next slide). The moments computed
here would be used to design the
reinforcement parallel to lines 1-2 in this panel

71
 6500 mm

600 mm 500 mm
l2
1 2

ln
6000 mm

6500 mm

72
 Now Mo = (qdl2)ln2/8; where ln = clear span of
slab panel = 6.5-1/2(0.5)-1/2(0.6) = 5.95m ;
l2 = width of panel = 6.5/2 + 6.0/2 = 6.25m
→ Mo = (14.76×6.25×5.952)/8 = 381 kNm
 Consider panel B, spanning from column 1 to
column 4 (next slide). The moments
computed here would be used to design the
reinforcement parallel to lines 1-4 in this
panel. For the purpose of computing ln, the
circular supports are replaced by equivalent
square columns having a side length c1 =
0.886dc.

73
300 mm 1

ln
6000 mm
l2

4
600 mm
5800 mm 6500 mm

74
 → ln = 6.0-1/2(0.3)-1/2(0.886×0.6) = 5.59m;
l2 = 5.8/2 + 6.5/2 = 6.15m; Mo = (14.76 × 6.15
× 5.592)/8 = 331 kNm
 Now the total statical moment will be
divided between the negative and positive
sections of the panel

75
 In the DDM, the total factored statical
moment Mo is divided into +ve and –ve
factored moments according to the rules
given in ACI Code, Section 14.6.30.
 These are illustrated in the Figure (next
slide)
 In interior spans, 65% of Mo is assigned to the
negative moment region and 35% to the +ve
moment region
 The exterior end of an exterior span has
considerably less fixity than the end at the
interior support.

76
0.35 to 0.63 Mo
0.35 Mo

Mo
Mo
0 to 0.65 Mo
0.65 to 0.75 Mo
0.65 Mo

Exterior span Interior span

Assignment of positive- and negative-moment regions

77
 The division of Mo in an end span into +ve and –
ve moment regions is given in Table 14.2 (next
slide). In this table, “exterior edge
unrestrained” refers to a slab whose edge rests
on, but is not attached to, for example, a
masonry wall. “Exterior edge fully restrained”
refers to a slab whose exterior edge is supported
by, and is continuous with, a concrete wall with
a flexural stiffness as large or larger than that of
the slab. If the computed –ve moments on two
sides of an interior support are different, the –ve
moment section of the slab is designed for the
larger of the two.
78
79
 The moments vary continuously across the
width of the slab panels. To aid in steel
placement, the design moments are
averaged over the width of column strips
over the columns and middle strips between
the column strips → define column and
middle strips
 Column strips in both directions extend one-
fourth of the smaller span, lmin, each way
from the column line.
 Middle strips are the strips between the
column strips.

80
 lmax

lmin

81
 ACI Section 14.6.4 defines the fraction of the
negative and positive moments assigned to
the columns strips. The remaining amount of
negative and positive moment is assigned to
the adjacent half-middle strips. Table (next
slide) gives the percentage distribution of –ve
factored moment to the column strip at all
interior supports
 The division is a function of (αf1l2/l1),which
depends on the aspect ratio of the panel,
l2/l1, and the relative stiffness, αf1, of the
beams (if any) spanning parallel to an within
the column strip
82
83
 For floor systems w/o interior beams,
(αf1l2/l1) is taken equal to zero, since αf1 = 0.
In this case 75% of the negative moment is
distributed in the column strip, and the
remaining 25% is divided equally b/n the two
adjacent half middle strips
 For cases where a beam is present in a
column strip (spanning in the direction of l1)
and (αf1l2/l1)≥ 1.0, the second row in table
14.3 applies.
 For 0≤(αf1l2/l1)≤1.0 →use linear interpolation

84
 Table 13-4 gives the percentage distribution of
+ve factored moment to the column strip at mid
span for both interior an exterior spans.
 For floor systems w/o interior beams, 60% of the
+ve moment is assigned to the column strip and
the remaining 40% is divided equally b/n the
adjacent half middle strips.
 If a beam is present in the column strip
(spanning in the direction of l1), either the
percentages in the 2nd row or a linear
interpolation b/n the percentages given in the
1st or 2nd row in Table 13-4 will apply

85
86
 Atan exterior edge, the division of the
exterior-end factored negative moment
distributed to the column and middle strips
spanning ⊥ to the edge also depends on the
torsional stiffness of the edge beam,
calculated as the shear modulus, G, times
the torsional constant of the edge beam, C,
divided by the flexural stiffness of the slab
spanning ⊥ to the edge beam (i.e., EI for a
slab having a width equal to the length of
the edge beam from the center of one span
to the center of the other span) designated
by βt (see next slide)
87
 Edge beam

Width of slab for the calculation of relative torsional

stiffness βt of edge beam

88
 Assuming that ν = 0 → G = E/2 so that βt =
(EcbC/2EcsIs)
 The term C is the torsional constant of the
edge beam which is calculated by subdividing
the cross section into rectangles and carrying
out the summation: C=Σ[(1-0.63x/y)x3y/3];
where x = shorter side of a rectangle and y =
longer side (NB: Several possible combination
of rectangles have to be tried to get the
maximum value of C. To do so wide
rectangles should be made as large as
possible. See Slide 90)

89
90
 Table 13-5 gives percentage distribution of
negative factored moment to column strip at
exterior supports. The set up of this table is
similar to the previous ones (tables 14.3 and
14.4) with the addition of two rows to account
for presence or absence of an edge beam
working in torsion to transfer some of the slab
negative moment into the column.
 When there is no edge beam (βt = 0), all of the
negative moment is assigned to the column
strips. This is reasonable because there is no
torsional edge member to transfer moment from
the middle strips all the way back to the
columns.
91
92
 Ifa stiff beam is present (βt ≥ 2.5), table
gives specific percentages to be assigned to
the column strip, depending on the value of
αf1 and the l2/l1 ratio, as was done in the
previous tables.
 For values of βt between 2.5 and 0.0 and
values of (αf1l2/l1) b/n 1.0 and 0.0, two or
three levels of linear interpolation may be
required to determine the percentage
distribution of negative moment assigned to
the column strip.

93
 If a beam is present in the column strip
(spanning in the direction of l1), a portion of
the column-strip moment is assigned to the
beam (ACI Code, Section 14.6.5).
 If the beam has (αf1l2/l1) > 1, 85% of the
column-strip moment is assigned to the beam
and 15% to the slab.

94
 Calculation of moments in an exterior
panel of a flat plate
 The slab is 200 mm thick and supports a
superimposed service dead load of 1.2 kN/m2
and a service live load of 3 kN/m2. the beam
is 300 mm wide by 400 mm in overall depth
and is cast monolithically with the slab.
 (1) Compute the factored loads: Let qd = 12
kN/m2

95
 A 0.35m D
0.40m

6.0m
1.5m

1.5m B E l2=5.75 m
1.375m

1.375m

5.5m

C F

l1=6.5 m

96
 (2) Compute the moments in span BE.
 → (a) Compute ln and l2 and divide the slab
into middle and column strips. → ln = 6.5-
1/2(0.35)-1/2(0.4) = 6.125m; l2 = 5.75m. The
column strip extends the smaller of l2/4 or
l1/4 on each side of the column centerline.
→ The column strip extends 6/4 = 1.5 m
toward AD and 5.5/4 = 1.275 m toward CF
from line BE as shown in Slide 96. → The
total width of the column strip is 2.875 m.
The half middle strip b/n BE and CF has a
width of 1.375 m, and the other one is 1.5 m

97
→ (b) Compute Mo: Mo = qdl2ln2/8 =
12×5.85×6.1252/8 = 324.6 kNm
 → (c) Divide Mo into positive and negative
moments. The distribution of the total
factored moment to the negative and the
positive moment regions is as given in Table
13-2 under the column “slabs w/o beams b/n
interior supports with edge beam”

98
 From Table 13-2, the total moment is divided
as follows:
 →Interior negative: Mu= 0.70Mo = -226.5 KNm
 →Positive: Mu= 0.50Mo = +161.8 KNm
 →Exterior negative: Mu= 0.30Mo = -97.1 KNm
 (d)
Divide the moments b/n the column and
middle strips
 Interior negative moments (Table 13-3): This
division is a function of αf1l2/l1, which is equal to
zero, since there are no beams to BE
 → Interior column-strip negative moment: 0.75× -
226.5 = -169.9 kNm = -59.1 kNm/m width of column
strip

99
  →Interior middle-strip negative moment = -56.6 kNm.
Half of this goes to each of the half middle strips
 Positive moments: (Table 13-4) :
 →Column-strip positive moment: 0.60×161.8 = 97.1
kNm→ 34.8 kNm/m
 →Middle-strip positive moment = 64.7 kNm. Half of
this goes to each half-middle strip.
 Exterior negative moment: From ACI Section
14.6.4.2, the exterior negative moment is
divided as a function of αf1l2/l1 (again equal to
zero, since there is no beam to l1) and βt. See
next slide for attached torsional member for
which βt will be calculated

100
 500 mm
300 mm 200 mm

200 mm
400 mm

200 mm

200 mm

101
 For Fig (a): C=[(1-0.63×300/400)3003×400/3+(1-
0.63×200/200)2003×200/3] = 2096.3×106 mm4
 For Fig (b): C = 1461.3×106 mm4. The larger of
the values is used; → C = 2096.3×106 mm4
 Is the moment of inertia of the strip of slab being
designed, which has b=5.75m and h=200mm.
 → Is=5750×2003/12 = 3834.3×106 mm4
 Since fck is the same in the slab and beam, Ecb=Ecs and
βt= 2096.3×106 /(2× 3834.3×106) = 0.273
 Interpolating in Table 13-5, we have:
 For βt=0 → 100% to column strip
 For βt=2.5 → 75% to column strip
 → for βt=0.273 → 97.3% to column strip and we have:
102
 Exterior column-strip negative moment: 0.973(-97.1)=
-94.5 kNm = -32.9 kNm/m
 Exterior middle-strip negative moment: -2.6 kNm

103
 Exterior Columns: When design is carried out by
the DDM, ACI specifies that the moment that is
transferred from a slab w/o interior beams to an
edge column is 0.26 to 0.30 Mo, as given in Table
13-2.
 This moment is used to compute the shear
stresses due to moment transfer to the edge
column (discussed later)
 The exterior negative moment from the DDM
calculation is divided b/n the columns above and
below the slab in proportion to the column
stiffness, 4EI/l. the resulting column moments
are used in the design of the columns

104
 InteriorColumns: At interior supports the
column moments are determined from
unbalanced moment resulting from an
uneven distribution of live load.
 The unbalanced moment is computed by
assuming that the longer span adjacent to
the column is loaded with the factored dead
load and half the factored live load, while
the shorter span carries only the factored
dead load
 The total unbalanced negative moment at
the joint is thus: M = 0.65(1/8){(wd+0.5wl)l2ln2
– w’dl’2(l’n)2}; where wd and wl refer to the
factored dead and live loads on the longer
span and w’d, l’2, and l’n refer to the shorter
span adjacent to the column
 A portion of the unbalanced moment is
distributed to the slabs and the rest goes to
the columns. →ACI gives Mcol = 0.07{(wd+
0.5wl) l2ln2 – w’dl’2(l’n)2}
 When a slab panel contains a beam, either
an edge beam or an interior beam b/n the
columns, the moments in the panel are
divided b/n the slab and the beam the same
way the moments are divided b/n the slab
and the beam for interior beams (refer to
literature (Macgregor) as alternative to
design of two-way beam supported slabs
using the coefficients in EBCS-2).

107
A shear failure in a beam results from an
inclined crack caused by flexural and
shearing stresses. This crack starts at the
tensile face of a beam and extends
diagonally to the compression zone.
 In the case of a two-way slab or footing, the
two shear-failure mechanisms shown in
Figure (next slide) are possible.
 One-way shear or beam-action shear (Fig a)
involves an inclined crack extending across
the entire width of the structure.

108
 Two-way shear or punching shear involves a
truncated cone or pyramid-shaped surface
around the column as shown in Fig b.
Generally, the punching-shear capacity of a
slab or footing will be considerably less than
the one-way shear capacity.
 This section is limited to footings and slabs
w/o beams. Refer to literature (Macgregor)
for shear strength of slabs with beams.

109
Inclined crack Pyramid-shaped
failure surface

(a) One-way shear (b)Two -way shear

110
 Behavior of slabs failing in two way shear
 As discussed in DDM, the maximum moments in a
uniformly loaded plate occur around the columns
and lead to a circular crack around each column.
After additional loading, the cracks necessary to
form a fan yield-line mechanism develop (see
next slide), and at about the same time, inclined
or shear cracks form on the truncated conical
surface shown in Fig b. These cracks can be shown
in Fig (slide after next), which shows a slab that has
been sawn through along two sides of the column
after the slab had failed in two-way shear

111
Fan yield line at a column in a flat plate

112
Inclined cracks in
a slab after shear
failure

113
 Alexander and Simonds used the truss model
in Fig (see next slide) to analyze punching-
shear failures. Prior to the formation of the
inclined cracks shown in Fig 13-53, the shear
is transferred by shear stresses in the
concrete. Once the cracks have formed, only
relatively small shear stresses can be
transferred across them. Now the majority of
the vertical shear is transferred by inclined
struts A-B and C-D extending from the
compression zone at the bottom of the slab
to the reinforcement at the top of the slab.
114
 B
A
C
D

115
 Similar struts exist on all four sides of the
column. The horizontal component of the
force in the struts causes a change in the
force in the reinforcement at A and D, and
the vertical pushes up on the bar and is
resisted by the tensile stresses in the
concrete b/n the bars. Eventually, this
concrete cracks in the plane of the bars, and
a punching failure results.
 Such a failure occurs suddenly, with little, if
any, warning.

116
 We will consider the case of shear transfer
w/o appreciable moment transfer. The case
when both shear and moment are transferred
from the slab to the column is discussed in
subsequent sections.
 Location of critical perimeters
 Two-way shear is assumed critical on a vertical
section through the slab or footing extending
around the column. According to the ACI at d/2
1.5d according to EBCS-2. See Figs (next slides)

117
 d/2 (1.5d)

Critical shear
perimeter

Fig. Location of critical

shear perimeter

Critical shear
perimeter

118
 Critical sections for slabs with drop panels
 When high shear forces are being transferred at a
slab-column connection, the slab shear strength
can be increased locally by using a drop panel to
locally increase the thickness of the slab. ACI
requires that the total thickness of the slab and
drop panel to be at least 1.25 times the thickness
of the slab adjacent to the drop panel.
 In slab with drop panels, two critical sections
should be considered, as shown in Figure (next
slide)

119
 d2
(a) Section through a drop panel d1

Edge of drop
panel
d1/2

Fig. Critical sections in a slab

with drop panels d2/2

(b) Critical sections

120
 If a drop panel is also used to control deflections
or reduce the amount of flexural reinforcement
required in the slab, the drop panel must satisfy
the length requirements given in ACI Code,
Section 14.2.5.
 Critical sections near holes and at edges
 When openings are located at less than 10 times
the slab thickness from a column, ACI Code
Section 11.11.6 requires that the critical
perimeter be reduced as shown in Figure (next
slide)

121
 ineffective
Regard as
free edge
(a) openings
Edge of slab

(b) Critical perimeter if A and B do not

exceed the greater of 4h or 2ld
Edge of slab
Fig. Effects of
openings and edges on
the critical shear
perimeter
(c) Critical perimeter if A exceeds the
greater of 4h or 2ld but B does not
122
 Tributary areas for shear in two-way slabs
 For uniformly loaded two-way slabs, the
tributary areas used to calculate Vu are bounded
by lines of zero shear. For interior panels, these
lines can be assumed to pass through the center
of the panel. For edge panels, lines of zero shear
are approximately at 0.42l to 0.45l from the
center of the exterior column, where l is the
span measured from center-to-center of the cols.

123
 However to be conservative in design, ACI Code
Section 8.4.3 requires that the exterior supports
must resist a shear force due to loads acting on
half of the span (0.5l). Also to account for the
larger tributary area for the 1st interior support,
ACI Code, Section 8.4.3 requires that the shear
force from loads acting on half the span must be
increased by 15%. See Figure (next slide)

124
 Tributary area for
two-way shear- col 2 Critical section
1 2 4 for one way
shear- col 4

0.5l

0.5(×1.15)l
Tributary area for
3
two-way shear-col 3

Tributary area for

one-way shear-col 5
5
Critical section

Fig. Critical sections and tributary areas

for shear in a flat plate
125
 Design Equations: Two-way shear with
negligible moment transfer
 Lateral loads and unbalanced floor loads, on a
flat-plate building require that both moments
and shears be transferred from the slab to the
columns. In the case of interior columns in a
braced flat-plate building, the worst loading case
for shear generally corresponds to a negligible
moment transfer from the slab to the column.
Similarly, columns generally transfer little or no
moment to footings
 Design for two-way shear w/o moment transfer is
carried out by using EBCS-2 Eq (…)
126
 6.4 Punching
 6.4.1 General
 (1)P The rules in this section complement those
given in 6.2 (Shear) and cover punching shear in
solid slabs, waffle slabs with solid areas over
columns, and foundations.
 (2)P Punching shear can result from a
concentrated load or reaction acting on a
relatively small area, called the loaded area Aload
of a slab or foundation.
 (3) An appropriate verification model for
checking punching failure at the ULS is shown in
Figure (next slide)
127
128
 (4) The shear resistance should be checked
along defined control perimeters
 (5) The rules in 6.4 are principally
formulated for the case of uniformly
distributed loading. In special case, such as
footings, the load within the control
perimeter adds to the resistance of the
structural system, and may be subtracted
when determining the design punching shear
stress

129
 6.4.2 Load distribution and basic control
perimeter
 (1) The basic control perimeter u1 may
normally be taken to be at a distance 2.0d
from the loaded area and should be
constructed so as to minimize its length (see
Figure next slide). The effective depth of the
slab is assumed constant and may normally
be taken as: deff = (dy + dz)/2; where dy and
dz are the effective depths of the
reinforcement in two orthogonal directions

130
Figure 6.13

131
 (2) Control perimeters at a distance less than 2d
should be considered where the concentrated
force is opposed by a high distributed pressure
(e.g. soil pressure in a base), or by the effects of
a load or reaction within a distance 2d of the
periphery of area of application of the force
 (3) For loaded areas situated near openings, if
the shortest distance b/n the perimeter of the
loaded area and the edge of the opening does
not exceed 6d, that part of the control
perimeter contained b/n two tangents drawn to
the outline of the opening from the center of the
loaded area is considered to be ineffective (SNS)

132
133
 (4)For a loaded area situated near an edge
or a corner, the control perimeter should be
taken as shown in Figure (SNS), if this gives a
perimeter (excluding the unsupported edges)
smaller than that obtained from (1) and (2)
above.

134
Figure 6.15

135
 (5) For loaded areas situated near or on an
edge or corner, i.e. at a distance smaller
than d, special edge reinforcement should
always be provided, see 9.4.1.4
 (6) The control section is that which follows
the control perimeter and extends over the
effective depth d. For slabs of constant
depth, the control section is ⊥ to the middle
plane of the slab. For slabs or footings of
variable depth, the effective depth may be
assumed to be the depth at the perimeter of
the loaded area

136
 (7) Further perimeters, ui, inside and outside
the control area should have the same shape
as the basic control perimeter.
 (8) For slabs with circular column heads for
which lh ≤ 2hH (see Fig next slide) a check of
the punching shear stresses according to
6.4.3 is only required on the control section
outside the column head. The distance of
this section from the centroid of the column
rcont may be taken as: rcont = 2d + lH + 0.5c;
where lH is the distance from the column
face to the edge of the column head and c is
the diameter of a circular column
137
138
 For a rectangular column with a rectangular
head with lh ≤ 2d and overall dimensions l1
and l2 (l1 = c1 +2lH1, l2 = c2 +2lH2, l1 ≤ l2), the
value rcont may be taken as the lesser of: rcont
= 2d + 0.56√(l1 l2) and rcont = 2d + 0.69 l1
 (9) For slabs with enlarged column heads
where lh ≥ 2hH (see Figure NS) the critical
sections both within the head and in the slab
should be checked.
 (10) The provisions of 6.4.2 and 6.4.3 also
apply for checks within the column head with
d taken as dH according to Figure in NS.

139
140
 (11) For circular columns the distances from
the centroid of the column to the control
sections in Figure (SPS) may be taken as:
rcont,ext = lH + 2d + 0.5c; rcont,int = 2(d + hH) +
0.5c
 6.4.3 Punching shear calculation
 (1)P The design procedure for punching shear
is based on checks at a series of control
sections, which have a similar shape as the
basic control section. The following design
shear stresses, per unit area along the
control sections, are defined:

141
 vRd,c is the design value of the punching shear
resistance of a slab w/o punching shear
reinforcement along the control section
considered
 vRd,cs is the design value of the punching
shear resistance of a slab with punching
shear reinforcement along the control
section considered
 vRd,max is the design value of the maximum
punching shear resistance along the control
section considered

142
 (2)
The following checks should be carried
out;
 (a) At the column perimeter, or the perimeter of
the loaded area, the maximum punching shear
stress should not be exceeded: → vEd < vRd,max
 (b) Punching shear reinforcement is not
necessary if: vEd < vRd,c
 (c) Where vEd exceeds the value vRd,c for the
control section considered, punching shear
reinforcement should be provided according to
6.4.5

143
 (3) where the support reaction is eccentric with
regard to the control perimeter, the maximum
shear stress should be taken as: vEd = β(VEd/uid);
where d is mean effective depth of slab, taken
as (dy + dz)/2; ui is length of control perimeter
being considered; and β is given by: β = 1 +
k(MEd/VEd)(u1/W1) Eq.(6.39); where u1 is the
length of the basic control perimeter; k is a
coefficient dependent on the ratios b/n the
column dimensions c1 and c2: its value is a
function of the proportions of the unbalanced
moment transmitted by uneven shear and by
bending and torsion (see Table 6.2)
144
 W1 corresponds to a distribution of shear as
shown in Figure (SNS) and is a function of the
basic control perimeter u1: W1 = ∫0uedl;
where dl is the a length increment of the
perimeter; and e is the distance of dl from
the axis about which the moment MEd acts.
 Table 6.2: Values of k for rectangular
loaded areas
c1/c2 ≤ 0.5 1.0 2.0 ≥ 4.0
k 0.45 0.60 0.70 0.80

145
Fig: Shear distribution due to an unbalanced moment at
a slab-internal column connection

146
 For a rectangular column: W1 = (c12/2) + c1c2
+ 4c2d + 16d2 + 2πdc1; where c1 is the column
dimension parallel to the eccentricity of the
load; and c2 is the column dimension ⊥ to the
eccentricity of the load.
 For internal columns β follows from: β = 1 +
0.6π(e/(D+4d))
 For an internal rectangular column where the
loading is eccentric to both axes, the
following approximate expression for β may
be used: β = 1 + 1.8(√(ey/bz)2 + (ez/by)2; where
147
 eyand ez are the eccentricities MEd/VEd along
y and z axes respectively; by and bz is the
dimensions of the control perimeter (see
figure 6.13); D is the diameter of the circular
column. (Note: ey results from a moment about the z
axis and ez from a moment about the y axis)
 (4)
For edge column connections, where the
eccentricity ⊥ to the slab edge (resulting
from a moment about an axis  to the slab
edge) is toward the interior and there is no
ecc to the edge, the punching force may
be considered to be uniformly distributed
along the control perimeter u1* (See NS).
148
149
 Where there are ecc in both orthogonal
directions, β may be determined using the
following expression: β = (u1/u1*) +
k(u1/W1)epar; where u1 is the full control
perimeter (see Fig 6.15); u1* is the reduced
control perimeter (See Fig 6.20 (a)); epar is
the ecc to the slab edge resulting from a
moment about an axis ⊥ to the slab edge; k
may be determined from Table 6.2 with the
ratio c1/c2 replaced by c1/2c2; and W1 is
calculated for the full perimeter (see fig 6.13

150
 For a rectangular column as shown in Figure
6.20(a): W1 = (c22/4) + c1c2 + 4c1d + 8d2 +
πdc2
 If the ecc ⊥ to the slab edge is not toward
the interior, Expression (6.39) applies. When
calculating W1 the ecc e should be measured
from the centroid of the control perimeter.
 (5) For corner column connections, where
the ecc is toward the interior of the slab, it
is assumed that the punching force is
uniformly distributed along the reduced
control perimeter u1*, as defined in Fig 6.20b

151
 The β-value may then be considered as: β =
u1/u1*. If the ecc is toward the exterior,
Expression (6.39) applies
 (6) For structures where the lateral stability
does not depend on frame action b/n the
slabs and the columns, and where the
adjacent spans do not differ in length by
more than 25%, approximate values for β
may be used (β = 1.15 for internal columns; β
= 1.4 for edge columns).

152
 (7) Where a concentrated load is applied
close to a flat slab column support, the
resistance enhancement according to
6.2.2(5) is not valid and should not be
included.
 (8)The punching shear force VEd in a
foundation slab may be reduced due to the
favorable action of the soil pressure.
 (9) The vertical component Vpd resulting from
inclined prestressing tendons crossing the
control section may be taken into account as
a favorable action where relevant.

153
 6.4.4 Punching shear resistance for slabs or
column bases w/o shear reinforcement
 (1) The punching shear resistance of a slab
should be assessed for the basic control section
according to 6.4.2. The design punching stress
(resistance) is given by: vRd,c = CRd,ck(100ρlfck)1/3
+ 0.10σcp ≥ (vmin + 0.10σcp); where fck is n MPa; k
= 1 + √(200/d) ≤ 2.0 d in mm; ρl = √(ρly× ρlz) ≤
0.02; ρly, ρlz relate to the bonded tension steel in
the y- and z- directions respectively. The values
ρly, ρlz) should be calculated as mean values
taking into account a slab width equal to the
column width plus 3d each side.

154
 σcp=(σcy + σcz)/2; where σcy, σcz are the
normal concrete stresses in the critical
section in y- and z-directions (MPa, positive
if compression): σcy= NEd,y/Acy and σcy=
NEd,z/Acz where NEd,y, NEd,z/ are the
longitudinal forces across the full bay for
internal columns and the longitudinal forces
across the control section for edge columns.
The force may be from a load or prestressing
action; Ac is the area of concrete according
to the definition of NEd(Note: the values of
CRd,c and vmin for use in a Country may…
155
 The recommended value for CRd,c is 0.18/γc
and that for vmin is given by Expression (6.3N)
 (2) The punching resistance of column bases
should be verified at control perimeters
within 2d from the periphery of the column.
The lowest value of resistance found in this
way should control the design. For concertric
loading the net applied force is: VEd,red = VEd -
∆VEd; where VEd is the column load; and ∆Ved
is the net upward force within the control
perimeter considered, i.e., upward pressure
from soil minus self weight of base

156
 vEd = VEd,red/ud
 vRd = CRd,ck(100ρfck)1/3 + 2d/a ≥ vmin × (2d/a) ;
where a is the distance from the periphery of
the column to the control perimeter
considered; CRd,c defined in 6.4.4(1); vmin
defined in 6.4.4(1)
 For eccentric loading: vEd = (VEd,red/ud)[1 +
k(MEdu/VEd,redW)]; where k is defined in
6.4.3(4)

157
 6.4.5 Punching shear resistance of slabs or
column bases with shear reinforcement
 Read!

158
 4.10.0 Equivalent-Frame Methods
 The ACI Code presents two general methods
for calculating the longitudinal distribution
of moments in two-way slab systems. These
are the direct-design method and the
equivalent-frame methods.
 Equivalent-frame methods are intended for
use in analyzing moments in any practical
slab-column frame. Their scope is thus wider
than the direct-design method, which is
subject to the limitations presented in
Section 13-7.

159
 In the direct-design method, the statical
moment M0, is calculated for each slab span.
This moment is then distributed b/n positive-
and negative- moment regions using arbitrary
moment coefficients, which are adjusted to
reflect pattern loadings.
 For equivalent-frame methods, a stiffness
analyses of a slab-column frame is used to
determine the longitudinal distribution of
bending moments, including possible pattern
loadings. The transverse distribution of
moments to column and middle strips, is the
same for both methods
160
 4.10.1 Classic Equivalent-Frame Analysis
of Slab Systems for Vertical Loads
 The slab is divided into a series of equivalent
frames running in two directions of the building
as shown in Figure (SNS).
 These frames consist of the slab, any beams that
are present, and columns above and below the
slab.
 For gravity load analysis, the code allows
analysis of entire equivalent frame extending
over the height of the building, or each floor can
be considered separately with the far ends of
the columns being fixed.
161
162
 The original derivation of the classic equivalent-
frame method assumed that the moment
distribution would be the calculation procedure
used to analyze the continuous-slab system, so
some of the concepts in the method are
awkward to adapt to other methods of analysis.
 (i) Calculation of Stiffness, Carryover, and
Fixed-End Moments
 In the moment distribution method, it is
necessary to compute flexural stiffnesses, K;
carry-over factors, COF; distribution factors, DF;
and fixed-end moments, for each members in
the structure (read..)

163
 In the equivalent-frame method, the increased
stiffness of members within the column-slab
joint region is accounted for, as is the variation
in cross section at drop panels. As a result, all
members have a stiffer section at each end, as
shown in Figure (SNS)
 (ii) Properties of Slab-Beams
 The horizontal members in the equivalent frame
are referred to as slab-beams. These consist of
either only a slab, or a slab and a drop panel, or
a slab with a beam running parallel to the
equivalent frame

164
 (b) Distribution of EI along slab
Fig. 13-31 Variation in stiffness along a span
165
 ACI Code Section 14.7.3 explains how these
nonprismatic beams are to be modeled for
analysis: (Read)
 The application of the approach is illustrated
in Figures (SNS). Tables A-14 through A-16 etc
present moment-distribution constants for
flat plates and for slabs with drop panels.
 Example- Calculation of the Moment-
Distribution Constants for Flat-Plate Floor
(Read pp 670, and 672)

166
 EI1 EI2

l2

(c) Cross section used to

l2 compute I1-Section A-A
Fig. 13-32 EI values for a
slab with a drop panel

(d) Cross section used to

compute I2-Section B-B
167
 EI2 EI1

(b) Variation of EI along

slab beam l2

h1 h2
(c) Cross section used to
compute I1-Section C-C l2
h3

(d) Cross section used to compute I1-Section D-D

168
 (iii)Properties of Columns
 In computing the stiffnesses and carryover
factors for columns, ACI Code Section 14.7.4
states the following:
 1. The moment of inertia of columns at any
section outside of the joints or column
capitals may be based on the gross area of
the concrete, allowing for variations in the
actual moment of inertia due to changes in
the column cross section along the length of
the column (SNS) (read)

169
 I=∞ I=∞
lc EccIc EccIc

I=∞ I=∞

Column stiffness Column stiffness

diagram diagram

Fig. 13-37 Sections

for the calculations
of column stiffness

170
 (iv) Torsional Members and Equivalent
Columns
 When the beam and column frame shown in
Figure (SNS) is loaded, the ends of the
column and beam undergo equal rotations
where they meet at the joint. If the flexural
stiffness, K=M/θ, is known for the two
members, it is possible to calculate the joint
rotations and the end moments in the
members. Similarly, in the case shown in
Figure (b), the ends of the slab and the wall
both undergo equal end rotations when the
slab is loaded
171
 θ
θ

(a) Beam and

column frame

(b) Slab and wall

A’
frame C B’
A
D
(c) Slab column B
frame
Attached torsional θA
member θC
Fig. 13-38 Frame
θA
action and twisting
of edge member
172
 When a flat plate is connected to a column,
as shown in Fig(c), the end rotation of the
column is equal to the end rotation of the
strip of slab C-D, which is attached to the
column.
 The rotation at A of the strip A-B is greater
than the rotation at point C, however,
because there is less restraint to the rotation
of the slab at this point
 In effect the edge of the slab is twisted, as
shown in Fig (d)

173
 As a result, the average rotation of the edge
of the slab is greater than the rotation of the
end of the column
 To account for this effect in slab analysis,
the column is assumed to be attached to the
slab-beam by the transverse torsional
members A-C and C-A’. One way of including
these members in the analysis is by use of
the concept of an equivalent column, which
is a single element consisting of the columns
above and below the floor and attached
torsional members, as shown in Figure (d).

174
 The stiffness of the equivalent column, Kec,
represents the combined stiffness of the columns
and attached torsional members:
 Kec = M/(average rotation of the edge beam)
 The flexibility of the equivalent column, 1/Kec, is
equal to the average rotation of the joint b/n
the “edge beam” and the rest of the slab when a
unit moment is transferred from the slab to the
equivalent column.
 This average rotation is the rotation of the end
of the columns, θc, plus the average twist of the
beam, θt,avg, with both computed for a unit
moment
 θec = θc + θt,avg

175
 The value of θc for a unit moment is 1/ΣKc,
where ΣKc refers to the sum of the flexural
stiffnesses of the columns above and below
the slab.
 Similarly, the value θt,avg for a unit moment is
1/Kt, where Kt is the torsional stiffness of the
attached torsional members. Substituting:
1 1 1
= +
K ec ΣK c K t

176
 If the torsional stiffness of the attached torsional
members is small, Kec will be much smaller than
ΣKc
 The derivation of the torsional stiffness of the
torsional members (or edge beams) is illustrated
in Figure (SNS).
 Figure (a) shows an equivalent column with
attached torsional members that extend halfway
to the next column in each direction.
 A unit torque, T=1, is applied to the equivalent
column with half going to each arm. Linear
torque distribution t, per unit length is assumed
as shown in Figure (b)

177
 l2

2/l2

(b) Distribution of torque per unit

length
T=1/2

(c) Torque diagram

Fig. 13-39 Calculation of Kt θ=

(d) Angle change per unit length

178
 The applied torques give rise to the twisting-
moment diagram shown in Figure (c). Because
half of the torque is applied to each arm, the
maximum twisting moment is ½.
 The twist angle per unit length is shown in Figure
(d). This is calculated by dividing the twisting
moment at any point by CG, the product of the
torsional constant, C (similar to a polar moment
of inertia), and the modulus of rigidity, G.
 The total twist of the end of an arm relative to
the column is the summation of the twists per
unit length and is equal to the area of the
diagram in Figure (d) (diagram is parabolic) →

179
 Area equals 1/3 of the height times the length of
the diagram →
1 (1 − c2 / l2 ) 2  l2  c2  
θ t ,end = 
1 − 

3 2CG  2 l 2 
 Replacing G with E/2

l2 (1 − c2 / l2 )3
θ t ,end =
6CE
 This is the rotation of the end of the arm. The
rotation required for use in Eqn above is the
average rotation of the arm, which is assumed to
be a third of the end rotations.
180
→ l2 (1 −c 2 / l2 )3
θ t ,avg =
18CE

 Finally,the torsional stiffness of one arm is

calculated as Kt=M/θt,avg, where the moment
resisted by one arm is taken as ½, giving:
9 EC
K t (one − arm) =
l2 (1 − c2 / l2 )3
 ACIexpresses the torsional stiffness of the
two arms as →

181
→ Kt = Σ
9 Ecs C
l2 (1 − c2 / l2 )3

 For a corner column there is only one term in

the summation.
 The cross section of the torsional members is
defined in ACI Code Section 13.7.5 and is
illustrated in Figure (SNS)

182
183
 Theconstant C in Eqns above is calculated by
subdividing the cross section into rectangles
and carrying the out the summation
 3 
∑  (1 − 0.63 ) y 
x x
C=
 y 3 
 
 Where x is the shorter side of a rectangle and y is the
longer side.
 Read example 13-7 and 13-8 –Calculation of
Kt, ΣKc, and Kec

184
 If a beam parallel to the l1 direction (a beam
along C-D in Figure 13-38) frames into the
column, a major fraction of the exterior
negative moment is transferred directly to
the column w/o involving the attached
torsional member. In such a case, Kec
underestimates the stiffness of the column.
 This is allowed for empirically by multiplying
Kt by the ratio Isb/Is, where Isb is the moment
of inertia of the slab and beam together and
Is is the moment of inertia of the slab
neglecting the beam stem (ACI Code
Section13.7.5.2)
185
 Arrangement of live loads for structural analysis and
moments at face of supports → See example 13-9 for
Analysis of a Flat-Plate using the Classic Equivalent-
Frame Method
 Distribution of Moments to Column Strips, Middle
Strips, and Beams
 Once the negative and positive moments have been
determined for each equivalent frame, these are
distributed to column and middle strips in the same way
as in the DDM.
 For panels with beams b/n the columns on all sides, the
distribution of moments to the column and middle strips
according to ACI Code Sections 13.6.4 and 13.6.6 is valid
only if αf1l22/αf2l12 falls b/n 0.2 and 5.0. Cases falling
outside of this range tend to approach one-way action,
and other methods of slab analysis are required

186
 4.10.2Use of Computers for an Equivalent-
Frame Analysis
 The classic EFM was derived by assuming that
the structural analysis would be carried out by
hand using the moment-distribution method.
 Thus tables were developed to evaluate fixed-
end moments, stiffnesses, and equivalent-
column stiffnesses for use in such analysis
 If standard frame analysis software based on the
stiffness method is to be used, the torsional
member (and the resulting equivalent-column
stiffness) defined in the classic EFM will need to
be incorporated into the stiffness of either the
slab-beam or column elements.

187
 The general research direction has been to
modify the stiffness of the slab-beam element by
defining an effective slab width to reduce the
element stiffness, particularly at connections.
 The frame analysis results for gravity loading,
obtained using the modified slab-beam
elements, should be in reasonable agreement
with those obtained from the classic EFM.
 Several researchers have worked on the
development of effective slab width models that
could be used to define the stiffness of an
equivalent beam in a standard frame analysis
program for the analysis of slab-column frame
subjected to combined vertical and lateral
loading. (based on plate theory and exp results)

188
 Hueste and Wight→ the 1st step in building a
slab-column frame analysis model is to select
an effective slab width that is a fraction, α,
of the total slab width, l2 (avg) as shown in
Figs. 13-34 and 13-35 (SNS).
 A wide range of range of α values have been
suggested by various researchers, but Wight
prefers to simply use α = 0.5 for all positive-
bending regions and for negative-bending
regions at interior supports.

189
190
191
 For negative-bending regions at exterior
supports, the effective slab width depends
on the torsional stiffness at the edge of the
slab.
 If no edge beam is present, then an α value
of 0.2 is recommended.
 If an edge beam is present and has a
torsional stiffness such that βt, as defined in
Eqn (13-12), is greater than or equal to 2.5,
then the recommended α value is 0.5

192
 Ifthe value of βt is b/n 0.0 and 2.5, a linear
interpolation can be used to find an α value
b/n 0.2 and 0.5.
 For low values of α, the effective slab width
should not be taken to be less than the
column width, c2, plus one-half of the
column total depth, c1, on each side of the
column (Fig. 13-48).
 For slab-column frame along a column line at
the edge of a floor plan, the effective slab
widths are reduced accordingly.

193
 The resulting models for one exterior and one
interior column line is shown in Fig. 13-49
 As indicated in Fig. 13-49, the negative-bending
region at the exterior connection is assumed to
extend over 20 percent (0.2l1) of the span. The
authors recommend that the same assumption be
used for negative-bending regions at all interior
and exterior connections.
 This assumption essentially creates extra node
points within the span and becomes important
when assigning cracked-stiffness values to the
positive and negative moment regions

194
 After the effective slab width, αl2, has been
established, the gross moment of inertia for
the slab-beam can be calculated using either
a section similar to Fig. 13-32 c (if no beam
is present) or a section similar to that in Fig
23-33c (if a beam is present). For both cases,
the effective slab width, αl2, is to be used in
place of the l2 value shown in those figures.
 If a drop panel is present in the negative-
bending region, then a section similar to that
used in Fig. 13-32d (with αl2, in place of l2) is
to be used. Read more.

195
 A final modification is to be made to the slab-
beam stiffness to account for flexural cracking.
In general, the cracked moment of inertia for a
slab-beam section, Icr, is some fraction of the
gross moment of inertia for that section.
Because slabs normally have lower
reinforcement ratios than beams, their cracked
moment of inertia is usually a smaller fraction of
the gross moment of inertia than for a typical
beam section. However, because large portions
along the slab-beam will remain uncracked and
the flexural cracks that do occur usually will not
propagate over the entire width of the slab, an
effective moment of inertia, Ie, needs to be
defined for different portions of the slab-beam
span.
196
 Commonly, a factor β is used to define the
effective moment of inertia as some fraction
of the gross moment of inertia (Ie = βIg). For
all positive-bending regions of the slab, the
author recommends β=0.5.
 Because larger moments typically occur near
interior connections, and in order to not
overestimate the slab-to-edge beam-to-
column stiffness at an exterior connection,
whether or not an edge beam is present, the
author recommends a β factor of 0.33 for all
negative bending regions. (see summary SNS)

197
198
 For analysis of post-tensioned slabs, wight etal
have recommended the use of a β value equal to
0.67 because of the reduced flexural cracking
expected in a post tensioned slab.
 For a gravity load analysis, the slab-beam
elements can be assembled with column
elements that extend one story above and one
story below the floor system (F9g 13-50), as
permitted by ACI Code Section 13.7.2.5
 The column lengths should be set equal to the
center-to-center dimensions from one floor level
to the next, and the gross moment of inertia of
the column sections can be used as input to the
structural analysis software. (moment at face of
support etc. Read)

199
200
 Analysis
of Slab-Column Frames for
Combined Gravity and Lateral Loads
A frame consisting of columns and either flat
plates or flat slabs but lacking shear walls or
other bracing elements is inefficient in
resisting lateral loads and may be subject to
significant lateral drift deflections.
 As a result, slab-column frame structures of
more than two or three stories are generally
braced by shear walls.

201
 When unbraced slab-column frames are used, it
is necessary to analyze equivalent frame
structures for both gravity and lateral loads.
 The general equivalent-frame analysis method
discussed previously can be used by simply
extending the slab-column frame over the full
height of the structure, as shown in Fig. 13-51.
 In order not to overestimate the lateral stiffness
of the slab-column frame (and thus
underestimate the lateral deflections), the
author recommends that the effective moment
of inertia of the column sections should be taken
as 70 percent of the gross moment of inertia, as
required in ACI Code Section 10.10.4.1 for
lateral stability analysis.

202
203