Yield Line Analysis of Reinforced Concrete Slabs with

Openings by Dip and Strike Angles Method

by

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

Reprinted from

Advances in Structural Engineering

Volume 16 No. 7 2013

MULTI-SCIENCE PUBLISHING CO. LTD.

5 Wates Way, Brentwood, Essex CM15 9TB, United Kingdom
Yield Line Analysis of Reinforced Concrete Slabs with
Openings by Dip and Strike Angles Method

A.K.H. Kwan1, J.Y.K. Lam2 and P.L. Ng3, *

Department of Civil Engineering, The University of Hong Kong, Hong Kong, China

(Received: 24 March 2012; Received revised form: 21 April 2013; Accepted: 23 April 2013)

Abstract: Using the dip and strike angles method of representing the geometry of
yield line pattern in terms of the dip and strike angles of the deflected slab segments,
the yield line analysis of reinforced concrete slabs had been successfully
computerized. In this study, the dip and strike angles method is generalized for
applications to slabs with any number of arbitrary-shaped openings subjected to point,
line, patch and uniformly distributed loads. As before, the external work done by
applied loads and the internal energy dissipation along yield lines are evaluated as
functions of the dip and strike angles, but to allow for the presence of openings, the
external work done and internal energy dissipation within the openings are discounted.
Based on the principle of virtual work, the load factor is evaluated in terms of the dip
and strike angles and the yield line pattern is determined by minimizing the load factor
with respect to the dip and strike angles. Examples of slabs with openings under
various boundary conditions and loading patterns are presented to illustrate the
applicability, efficiency and accuracy of the generalized method.

Key words: opening, reinforced concrete slabs, yield line analysis.

1. INTRODUCTION to simple slabs where hand calculation is still possible.

The methods for designing reinforced concrete slabs In actual fact, the yield line analysis method has been
may be broadly classified into elastic and plastic design developed for many decades (Ingerslev 1923;
methods. The elastic design methods are based on either Hognestad 1953; Johansen 1962; Jones and Wood
the finite element analysis method or the grillage 1967) and yet up to now its computerization for
analysis method while the plastic design methods are applications to arbitrary-shaped slabs has remained a
based on either the yield line analysis method or the formidable task. The development of computational
strip analysis method (Park and Gamble 2000). Among procedures for the yield line analysis of arbitrary-shaped
them, the elastic design method based on finite element slabs has attracted wide attention (Dickens and Jones
analysis and the plastic design method based on yield 1988; Johnson 1994, 1995; Thavalingam et al. 1998,
line analysis are the most popular. Since the elastic 1999; Middleton 1997, 2008; Jackson and Middleton
design methods are not capable of taking advantage of 2013) but has been progressing rather slowly.
the moment redistribution after yielding, they are often In a building structure, it is often necessary to make
overly conservative and uneconomical. Comparatively, openings in the floor slabs to allow the passage of
the plastic design methods are preferred because they pipe lines, electrical cables and other utilities. The
tend to produce more economical designs (Kennedy presence of these openings would weaken the slabs and
and Goodchild 2004). However, whilst the finite therefore must be taken into account in the analysis and
element analysis method can handle arbitrary-shaped design. At the elastic state, stress concentration occurs at
slabs, the yield line analysis method has been confined the corners of openings and consequently spurious

*Corresponding author. Email address: [email protected]; Fax: +852-2559-5337; Tel: +852-9587-5310.

Associate Editor: J.G. Dai.

Advances in Structural Engineering Vol. 16 No. 7 2013 1223

 Yield Line Analysis of Reinforced Concrete Slabs with Openings by Dip and Strike Angles Method

bending moments arise near the openings. Such stress original method, the extended dip and strike angles
concentration and spurious bending moments would method also represents the whole set of kinematically
diminish after yielding due to moment redistribution but admissible yield line patterns by a consistent set of dip
have been causing numerical difficulties when the and strike angles. The major difference is that in the
elastic design methods are applied. For circumventing extended method, the external work done and internal
this numerical problem, the plastic design methods energy dissipation within the openings are discounted in
based on yield line analysis are particularly useful. the computations.
However, the identification of yield line patterns in
slabs with openings is not straight forward. After 2. DIP AND STRIKE ANGLES
decades of research, only investigations into simple REPRESENTATION OF YIELD LINE
cases of rectangular slabs with rectangular openings PATTERNS
have been undertaken (Jones 1962; Lash and Banerjee 2.1. Dip and Strike Angles
1967; Zaslavsky 1967; Islam and Park 1971; Kumar and Generally, yield line analysis begins with the
Roy 1991; Kumar and Prakash 2001). Generalization construction of presumed yield line patterns and then
for applications to arbitrary-shaped slabs with arbitrary- proceeds by evaluating the ultimate load of each
shaped openings has been rather limited. presumed yield line pattern. All the yield line patterns
Even without opening, the yield line analysis of must be kinematically admissible. While ascertaining
arbitrary-shaped slabs with any combinations of clamped, whether a yield line pattern is kinematically admissible,
simply−supported or free edges is much more difficult it should be noted that a yield line is actually an
than most researchers would have anticipated. The main intersection line between two deflected slab segments,
difficulty lies in the identification and representation of each assumed to be rigid and rotating about an axis of
the whole set of kinematically admissible yield line rotation. For a yield line pattern to be kinematically
patterns. Traditionally, the presumed yield line patterns admissible, it has to satisfy the following kinematic
are drawn manually by intuition and the geometry of a conditions:
yield line pattern is defined in terms of the unknown (1) Every yield line or its extension must pass
bearing angles and lengths of the yield lines. Certain through the intersection point of the axes of
ingenuity is needed in drawing the yield line patterns. rotation of the two deflected slab segments,
Furthermore, since there may be more than one which intersect to form the yield line.
kinematically admissible yield line patterns, it is not easy (2) For a slab segment associated with a supported
to represent the whole set of yield line patterns to be edge, its axis of rotation is the supported edge;
considered by one consistent set of geometric parameters. while for a slab segment associated with a
For this reason, computerization of the yield line analysis supported corner, its axis of rotation must pass
method has been limited to slabs with simple geometry. through the supported corner.
To overcome such difficulty, the first author has (3) Every yield line is a straight line ending either at
developed the dip and strike angles method for yield line a nodal point where other yield lines meet or at
analysis, which represents the whole set of kinematically a free edge (an unsupported edge). Yield lines
admissible yield line patterns by a consistent set of dip are not allowed to intersect each other, but their
and strike angles of the slab segments deflecting under ends may meet at common nodal points.
load (Kwan 2004). With the yield line patterns defined in The main hurdle in the computerization of the yield
terms of dip and strike angles, the external work done and line analysis method is the representation of all
internal energy dissipation are evaluated as functions of kinematically admissible yield line patterns using a
the dip and strike angles and the load factor is determined single set of geometric parameters. To overcome this
by minimizing the load factor with respect to the dip and hurdle, the dip and strike angles method (Kwan 2004) of
strike angles. This method has been computerized fully treating every yield line as an intersection line between
for applications to polygonal-shaped slabs with any two deflected slab segments, measuring the deflection
combinations of clamped, simply-supported or free edges of each deflected slab segment in terms of its dip and
and subjected to complicated loadings. strike angles, and representing the whole set of
So far, the dip and strike angles method has only been kinematically admissible yield line patterns by a
applied to slabs without openings. Since there is still no consistent set of dip and strike angles of the deflected
computational procedure applicable to arbitrary-shaped slab segments is adopted herein. Each deflected slab
slabs with arbitrary-shaped openings, an attempt is segment is assumed to have negligible elastic
made herein to extend the dip and strike angles method deformation so that it deflects as a rigid body rotating
for applications to slabs with openings. As in the about an axis of rotation. The angle of rotation of the

1224 Advances in Structural Engineering Vol. 16 No. 7 2013

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

deflected slab segment is called the dip angle while the For a deflected slab segment, which rotates as a rigid
bearing angle of the axis of rotation is called the strike body and therefore remains a plane surface after
angle. Furthermore, for each supported edge (a clamped deflection, its deflection equation is given by:
or simply-supported edge) and each supported corner
between unsupported edges (a corner column between z = ax + by + c (1)
free edges), there is a slab segment associated with it.
For a slab segment associated with a supported edge, its in which x and y are the coordinates of any point within
strike angle is the same as that of the supported edge and the slab segment, z is the vertical deflection at (x, y), and
is thus a fixed parameter while for a slab segment a, b and c are the coefficients of the deflection equation.
associated with a supported corner between unsupported If the slab segment rotates by a dip angle of θ about an
edges, its strike angle can vary and is thus a variable axis of rotation with a strike angle of φ and the axis of
parameter. The dip angles of the slab segments added rotation passes through the point (x0, y0), then:
with the variable strike angles provide a consistent set of a = −θ sinφ (2a)
geometric parameters defining the whole set of
kinematically admissible yield line patterns.
b = θ cosφ (2b)
2.2. Slab Segments Associated with Different
Supports c = θ (x0 sinφ −y0 cosφ) (2c)
Details of how slab segments associated with different
supports are dealt with have been presented before To drive the deflection equation of a deflected slab
(Kwan 2004). Herein, further explanations are given by segment, what we need to do is just to substitute the
means of an example. Figure 1 depicts a five-sided slab corresponding values of θ, φ x0 and y0 into the above
ABCDE supported on two supported edges AB and BC, formulas.
and two supported corners D and E. For this slab, there In this particular example, the slab segments ABFGJ,
are altogether four slab segments, viz. ABFGJ, BCHF, BCHF, DIGFH and EJGI are numbered as the 1-st, 2-
DIGFH and EJGI, which are associated with the nd, 3-rd and 4-th slab segments. Their dip and strike
supported edge AB, supported edge BC, supported angles are denoted by θ1 and φ1, θ2 and φ2, θ3 and φ3,
corner D, and supported corner E, respectively. The and θ4 and φ4, respectively, and the coefficients a, b and
equations for the deflections of the four slab segments c of their deflection equations are denoted by a1, b1 and
are derived in the following. c1, a2, b2 and c2, a3, b3 and c3, and a4, b4 and c4,
respectively. For the slab segment ABFGJ,
a1 = −θ1 sinφ1 (3a)
D Axis of
rotation
I
E b1 = θ1 cosφ1 (3b)
Slab segment H
Slab
segment c1 = θ1 (xA sinφ1 − yA cosφ1 ) (3c)
G F C
Slab segment
Axis of J
where xA and yA are the x- and y-coordinates of the point
rotation A on the axis of rotation. For the slab segment BCHF,
Slab segment
a2 = −θ2 sinφ2 (4a)
A B
b2 = θ2 cosφ2 (4b)
Legend Free edge
Simply supported edge
c2 = θ2 (xB sinφ2 − yB cosφ2) (4c)
Clamped edge
Column
where xB and yB are the x- and y-coordinates of the point
Positive yield line
B on the axis of rotation. For the slab segment DIGFH,
Negative yield line
Axis of rotation a3 = −θ3 sinφ3 (5a)
Figure 1. Kinematically admissible yield line pattern for a
five-sided slab b3 = θ3 cosφ3 (5b)

Advances in Structural Engineering Vol. 16 No. 7 2013 1225

 Yield Line Analysis of Reinforced Concrete Slabs with Openings by Dip and Strike Angles Method

c3 = θ3 (xD sinφ3 − yD cosφ3) (5c) (ai − aj)x + (bi − bj)y + (ci − cj) = 0 (7)
This equation gives an intersection line that is infinitely
where xD and yD are the x- and y-coordinates of the point
long. Since the yield line is bounded by the slab
D on the axis of rotation. For the slab segment EJGI,
boundary and other yield lines, it is derived from the
a4 = −θ4 sinφ4 (6a) intersection line using the following steps. As a first
step, the points at which the intersection line crosses the
slab boundary, marked as P and Q in Figure 2, are
b4 = θ4 cosφ4 (6b) determined and the intersection line is trimmed to have
a finite length with end points P and Q. Then, as a
c4 = θ4 (xE sinφ4 − yE cosφ4) (6c) second step, the intersection line PQ is checked against
each of the intersection lines formed by the intersection
where xE and yE are the x- and y-coordinates of the point of the i-th or j-th slab segment with other slab segments.
E on the axis of rotation. Since yield lines are not allowed to cross each other, if
For the slab segments ABFGJ and BCHF, since their the intersection line PQ is found to cross other
axes of rotation align with the supported edges AB and intersection lines formed by the intersection of the i-th
BC, their strike angles φ1 and φ2 can be pre-determined or j-th slab segment with other slab segments, the
from the coordinates of the end points of their respective intersection line PQ is trimmed off to avoid crossing
supported edges. Hence, the strike angles φ1 and φ2 are other intersection lines. The remaining length of the
not variables. Overall, for this particular example of a intersection line, marked as RS in Figure 2, would be the
five-side slab ABCDE supported on two supported yield line formed. By repeating the above trimming
edges AB and BC, and two supported corners D and E, procedures for each intersection line, the whole yield
the whole set of kinematically admissible yield line line pattern can be generated.
patterns may be represented by the dip angles θ1, θ2, θ3 The above steps, which have been presented before
and θ4, and the strike angles φ3 and φ4. (Kwan 2004), are applicable only to slabs with no
openings. For slabs with openings, an additional step of
2.3. Generation of Yield Line Pattern for Slab trimming the parts of each yield line, which happen to
with Opening fall within an opening, is needed because there should
A yield line is actually part of an intersection line be no yield line within an opening. As illustrated in
between two deflected slab segments. Referring to Figure 2, after trimming the part of the yield line RS
Figure 2, let the two deflected slab segments within an opening, the yield line becomes RT. An
intersecting each other to form the yield line be the i-th automatic procedure for trimming the parts of the yield
and j-th slab segments. Equating the vertical deflections lines within the openings has been developed and
of the two deflected slab surfaces, the equation of the implemented in the computer program for yield line
intersection line may be derived as: analysis.

3. COMPUTATION OF LOAD FACTOR FOR

Opening SLAB WITH OPENING
The principle of virtual work is employed to evaluate
the load factor of the slab with or without opening under
the external loads. For this purpose, the equations for the
Q
j -th slab external work done by point, line, patch and uniformly
segment S distributed loads and for the internal energy dissipation
T along the yield lines are derived in the followings.
R
P
i-th slab 3.1. Evaluation of External Work Done
segment The total external work done Wext is equal to the sum of
the external work done by the point, line, patch and
uniformly distributed loads, as given by:

Wext = WextF + WextL + WextP + WextU (8)

Figure 2. Generation of a yield line between adjacent deflected in which WextF is the external work done by the point
slab segments loads, WextL is the external work done by the line loads,

1226 Advances in Structural Engineering Vol. 16 No. 7 2013

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

WextP is the external work done by the patch loads and of Wext is first evaluated as per Eqns 8 and 9 regardless
WextU is the external work done by the uniformly of whether there are openings. Then, the work done by
distributed load. The external work done by the point, the external loads acting on areas within the openings
line, patch and uniformly distributed loads may be are discounted. For every point load, line load and patch
evaluated using the following formulas: load, its action point, action line or action area is
checked against the openings to find out whether the
NF action point or any part of the action line or action area
Wext F = ∑ Fi Zi (9a) is within an opening. Having checked whether the
i =1 action points or any parts of the action lines or action
areas are within the openings, the work done by the
NL  p
loads within the openings is deducted from the value of
 Wext. Finally, for the uniformly distributed load, its work
Wext L = ∑  ∑ ηk H i Li z k  (9b)
i =1  k =1  done within the openings is also deducted from the
value of Wext.

NP  p q  3.2. Evaluation of Internal Energy Dissipation

P
Wext = ∑  ∑ ∑ η jηk Pi Ai z jk  (9c) The total internal energy dissipation Wint is equal to the
i =1 
 j =1 k =1 
sum of the internal energy dissipation at every
positive yield line (a positive yield line is a sagging
yield line) and the internal energy dissipation at every
N  p q 
WextU = ∑  ∑ ∑ η jηkUAi z jk  negative yield line (a negative yield line is a hogging
(9d)
i =1   yield line), as given by:
 j =1 k =1
where NF, NL and NP are the numbers of point loads, line NS  Ni  N H  Ni 
loads and patch loads, respectively; N is the number of Wint = ∑  ∑ M iψ i ∆
k  + ∑  ∑ M iψ i ∆
k  (10)
slab segments; Fi is the magnitude of the point load; Hi i =1 
 k =1  i =1  k =1 
is the intensity per unit length of the line load; Pi is the
intensity per unit area of the patch load; U is the in which NS is the number of positive yield lines; NH is
intensity per unit area of the uniformly distributed load; the number of negative yield lines; Ni is the number of
Li is the length along the line load being considered, Ai divisions of the yield line being considered; Mi is the
is the area of the patch load or uniformly distributed plastic moment per unit length along the yield line; ψi is
load, p × q is the total number of Gauss points for the plastic rotation of the yield line; and ∆
k is the length
numerical integration; and η is the weighting coefficient increment along the yield line for numerical integration.
of the Gauss point. To evaluate the energy dissipation at a yield line, it
To evaluate the work done by an external load, it is first is first necessary to calculate the plastic moment and
necessary to find out which slab segment the point load is rotation of the yield line. The plastic moment of a
acting on and which slab segment(s) the line load or patch yield line is just the resisting moment of the slab
load is acting on. For a point load, after identifying the slab section. Its value is dependent on the effective depth,
segment on which the load is acting, the vertical deflection sectional area and orientation of the reinforcing bars,
zi is evaluated using the deflection equation of the slab and can be determined using Johansen’s stepped yield
segment. For a line load or patch load acting on two or line criterion (Johansen 1962). For yield lines formed
more slab segments and for a uniformly distributed load, within the slab, the plastic rotation of each yield line
the load is separated into several parts, each acting on one can be determined simply from the dip and strike
slab segment. After identifying each slab segment on angles of the two slab segments intersecting to form
which the load is acting, the vertical deflection zk or zjk is the yield line. Let the two slab segments intersecting
evaluated using the deflection equation of the to form the yield line be the i-th and j-th slab segments
corresponding slab segment. Subsequently, the work done and the two end points of the yield line be R and T.
on each slab segment is obtained by numerical integration The plastic rotation of the yield line may be evaluated
using the Gaussian method. Finally, the work done by using coordinate geometry as:
each load is determined by summing the work done on
each slab segment. (ai − a j )( yR − yT ) − (bi − b j )( xR − xT )
If the slab contains one or more openings, special ψ= (11)
2 2
procedures are adopted, as described below. The value ( xR − xT ) + ( yR − yT )

Advances in Structural Engineering Vol. 16 No. 7 2013 1227

 Yield Line Analysis of Reinforced Concrete Slabs with Openings by Dip and Strike Angles Method

in which (xR, yR) are the coordinates of R, and (xT, yT) minimization is carried out successively with respect to
are the coordinates of T. For yield lines formed on the each independent variable by determining the values of
supported edges (note that only negative yield lines can the target function at three different values of the
be formed on supported edges), the plastic rotation of independent variable, treating the target function as a
each yield line is just the same as the dip angle of the parabolic function of the independent variable and
slab segment associated with the supported edge. finding the value of the independent variable that will
The above procedures and equations are applicable to give the minimum value of the target function. This
both slabs with no openings and slabs with openings. It numerical technique is called parabolic minimization. It
is only that when the slab being considered contains one has been proven to be effective and reliable and has
or more openings, each yield line is trimmed or even been used by the first author for years (Kwan 2004).
divided into several discontinuous lengths to remove the
part(s) of the yield line within the opening(s) before the 4. APPLICATION TO SIMPLE CASES FOR
internal energy dissipation at the yield line is evaluated. VERIFICATION
The yield line analysis method developed herein has
3.3. Determination of Load Factor been fully computerized and implemented in the form of
According to the principle of virtual work, the load a computer program with interactive data input for
factor λ may be determined as the ratio of the total practical applications. To verify the applicability and
internal energy dissipation to the total external work accuracy of the computer program, a number of
done (Johansen 1962), as given by: examples are analyzed and the results are compared
with the load factors given by other researchers or
λ = Wint/ Wext (12) obtained by manual calculation.

Both the terms Wext and Wint are functions of the dip and 4.1. Example 1: Square Slab with Central Square
strike angles of the deflected slab segments. For a slab Opening
with m supported edges and n supported corners between Example 1 is an isotropically reinforced square slab
unsupported edges, the variables to be considered are (m with a central square opening simply supported on all
+ n) dip angles and n strike angles. In other words, Wext four sides and subjected to a uniformly distributed load
and Wint are functions of the following form: of intensity U, as depicted in Figure 3. The slab is of size
6.0 m × 6.0 m while the opening is of size 3.0 m × 3.0
Wext = Wext (θ1, θ2,..., θm+n, φ1, φ2,..., φn) (13a) m. Reinforcement is provided such that the plastic
moment of the slab is everywhere equal to M. This
Wint = Wint (θ1, θ2,..., θm+n, φ1, φ2,..., φn) (13b) example has been analyzed using the conventional

By assigning different values to the (m + n) dip angles 6000

and n strike angles, the whole set of kinematically
admissible yield line patterns can be generated.
However, among the kinemetaically admissible yield
line patterns, only the one giving the smallest load factor
would actually occur. It is therefore necessary to find
3000
the set of dip and strike angles that would lead to the
smallest load factor. This is done by minimizing the
load factor λ with respect to the dip and strike angles
using the numerical procedure developed previously by
the first author (Kwan 2004).
The numerical procedure adopted to minimize the Note: All dimensions in mm.
load factor with respect to the dip and strike angles has
the special feature that it does not rely on the calculation
of any gradient (rate of change of the target function Plastic moments Loading applied
with an independent variable). It has to have this feature
M
because as the dip and strike angles vary, the yield line Uniformly distributed
load of intensity U
pattern can change abruptly from one pattern to a very
M
different pattern and as a result the gradient can become
undefined. Without relying on the gradients, the Figure 3. Example 1: square slab with central square opening

1228 Advances in Structural Engineering Vol. 16 No. 7 2013

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

method by Kumar and Prakash (2001). The computer line pattern produced by the computer program is
program produces a load factor of 0.667 M/U, whereas presented in Figure 4.
the value of load factor obtained by Kumar and Prakash
is also 0.667 M/U, indicating exact agreement. The yield 4.3. Example 3: Rectangular Slab with
line pattern produced by the computer program is drawn Rectangular Opening Offset from Centre
in Figure 3. This example is the same as Example 2 except that the
position of the opening is offset by 1.0 m to the left hand
4.2. Example 2: Rectangular Slab with Central side, as shown in Figure 5. The computer program
Rectangular Opening produces a load factor of 5.381, which is slightly higher
Example 2 is an orthotropically reinforced rectangular than the load factor of 5.177 for the case of the same
slab with a central rectangular opening simply slab with the opening located at the centre (i.e. the slab
supported on all four sides and subjected to a uniformly analyzed in Example 2). Hence, the location of the
distributed load, as depicted in Figure 4. The slab is of opening has certain effect on the load factor of the slab.
size 6.0 m × 4.0 m while the opening is of size 1.5 m × The yield line pattern produced by the computer
1.0 m (note that the aspect ratios of the slab and the program is presented in Figure 5. No numerical result
opening are the same). This example is a practical case obtained by other researcher is available for checking
analyzed by Islam and Park (1971). The design because this example is not extracted from the literature.
parameters of the slab are given in Table 1. The To verify the accuracy of the load factor obtained,
computer program yields a load factor of 5.177, whereas manual calculation following the conventional method
the value of load factor obtained by Islam and Park is has been carried out and the load factor so obtained is
also 5.177, indicating again exact agreement. The yield also 5.381, thus verifying the accuracy of the computer
analysis result.
6000
4.4. Example 4: Rectangular Slab with Central
Circular Opening
This example is the same as Example 2 except that the
rectangular opening is replaced by a circular opening of
4000 1000 diameter 1.5 m, as shown in Figure 6. In the computer
analysis, the curved edge of the circular opening is input
as a series of short straight lines. The computer program

2130 1500 2130 6000

Note: All dimensions in mm.

1250 1500 3250

Plastic moments Loading applied

16.50 kNm/m
Uniformly distributed
load of intensity 4 kN/m2 4000
24.12 kNm/m 1000

Figure 4. Example 2: rectangular slab with central rectangular

opening

1680 2110
Table 1. Design parameters of the slab analyzed in
Example 2 Note: All dimensions in mm.

Properties Design parameters Loading applied

Plastic moments
Thickness of slab (mm) 125
16.50 kNm/m
Effective depth of slab (mm) 115 Uniformly distributed
Compressive strength of concrete (MPa) 30 load of intensity 4 kN/m2
24.12 kNm/m
Yield strength of reinforcement (MPa) 460
Reinforcement in short span φ12 at 225 mm spacing Figure 5. Example 3: rectangular slab with rectangular opening
Reinforcement in long span φ12 at 350 mm spacing
offset from the centre

Advances in Structural Engineering Vol. 16 No. 7 2013 1229

 Yield Line Analysis of Reinforced Concrete Slabs with Openings by Dip and Strike Angles Method

6.0 m
6000

0.5 m
2250 1500 2250
5 5.5
5.

9
5.
5.7 6.1 0.75 m

4.0 m 5.3
5.5
4000 1000
5.7 5.7
5.5

6.
5.5

5.
1
9
1920 1920
(a) Rectangular opening of 1.5 m × 1.0 m
Note: All dimensions in mm.

Plastic moments Loading applied

0.75 m
16.50 kNm/m
Uniformly distributed
load of intensity 4 kN/m2
24.12 kNm/m 0.75 m
5.5
Figure 6. Example 4: rectangular slab with central circular opening 5.6
5.7
5.8

5.9
produces a load factor of 6.697, which is quite different
from the load factor of 5.177 for the case of the same
slab with a rectangular opening (i.e. the slab analyzed in
Example 2). Hence, the shape of the opening has (b) Circular opening of 1.5 m diameter
significant effect on the load factor of the slab. The yield
Figure 7. Variation of load factor with location of opening
line pattern produced by the computer program is
presented in Figure 6, from which it can be seen that the
yield line pattern in this example is identical to the yield be used for such purpose, Example 2 and Example 4
line pattern in Example 2. As the yield line patterns are are re-analyzed with the opening located at different
identical, the internal energy dissipation in Example 4 is places within the slab. During the computer analysis,
the same as that in Example 2. It is the smaller external the coordinates of the centre of the opening are varied
work done by the uniformly distributed load acting on a within their permissible ranges and the corresponding
smaller slab area in Example 4 that causes a larger load load factor for each location of the opening is
factor compared to that of Example 2. To verify the evaluated. From the analysis, the load factors
accuracy of the load factor obtained, manual calculation corresponding to different locations of the opening are
following the conventional method has been carried out obtained and plotted for the case of rectangular
and the load factor so obtained is also 6.697, thus opening and the case of circular opening in Figures 7(a)
verifying the accuracy of the computer analysis result. and 7(b), respectively. It is seen from the plots of load
factor that the load factor is generally the lowest when
5. APPLICATION TO PARAMETRIC STUDY the opening, whether rectangular or circular, is located
ON EFFECT OF LOCATION OF OPENING near the centre of the slab. Hence, for a rectangular
When making an opening in a floor slab for the slab simply supported on all four sides, the opening
passage of utility lines, there is often the question of should not be made near the centre of the slab.
where the opening should be strategically located to However, for a rectangular slab clamped at all four
minimize the adverse effect of the opening on the sides, or for a more general case, no simple conclusion
structural capacity, i.e. the load factor, of the slab. can be drawn and a detailed evaluation is needed.
Conventionally, the effect of the location of opening is
evaluated by manual calculation, which is very 6. APPLICATION TO COMPLICATED CASES
laborious and time consuming. With the use of the FOR DEMONSTRATION OF VERSATILITY
computer program, the effect of the location of To demonstrate the versatility of the computer program,
opening can be evaluated automatically within two more complicated examples are analyzed, as
seconds. To illustrate how the computer program can presented below.

1230 Advances in Structural Engineering Vol. 16 No. 7 2013

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

6.1. Example 5: Polygonal Slab with Square 8 4

Opening
Example 5 is a five-sided polygonal slab with a square
opening, as shown in Figure 8. Among the five edges, 2
4
two are clamped edges, one is a simply-supported edge
and two are free edges. There is a column at the corner
between the two free edges. On the whole, the slab is
2
supported on two clamped edges, one simply-supported 8 6
edge and one supported corner between two free edges.
6
The slab is orthotropically reinforced with a plastic
4
moment of M about the y-axis and a plastic moment of
0.5 M about the x-axis under both sagging and hogging
conditions. A point load of magnitude 2 F and a patch
load of magnitude 5 F evenly distributed over an area of
10
1.0 m × 1.0 m are applied to the slab. The yield line
pattern produced by the computer program is drawn in Note: All dimensions in meters.
Figure 8 while the load factor produced by the computer
program is 0.2655 M/F. This complicated example has Plastic moments Loading applied
not been analyzed by any other researcher and therefore M Uniformly distributed
there is no independently obtained numerical result load of intensity U
available for checking the correctness of the above
M
computer analysis results. Nevertheless, based on the
yield line pattern shown in Figure 8, manual calculation Figure 9. Example 6: polygonal slab with circular opening
following the conventional method has been carried out
and the load factor so obtained is also 0.2655 M/F, thus two are simply-supported edges and three are free
verifying the correctness of the load factor calculation edges. There is a column at each corner between two
for a given yield line pattern. free edges. On the whole, the slab is supported on two
simply-supported edges and two supported corners
6.2. Example 6: Polygonal Slab with Circular between free edges. The slab is isotropically reinforced
Opening with a plastic moment of M under both sagging and
Example 6 is five-sided polygonal slab with a circular hogging conditions and is subjected to a uniformly
opening, as shown in Figure 9. Among the five edges, distributed load of intensity U. The yield line pattern
produced by the computer program is drawn in Figure 9
5 5 while the load factor produced by the computer program
is 0.1401 M/U. This complicated example has not been
analyzed by any other researcher and therefore there is
2 no independently obtained numerical result available
1 4 for checking the correctness of the above computer
1
analysis results. Nevertheless, based on the yield line
Patch load
pattern shown in Figure 9, manual calculation following
9
Point load 2 the conventional method has been carried out and the
5
load factor so obtained is also 0.1401 M/U, thus
3
4
2 verifying the correctness of the load factor calculation
for a given yield line pattern.

7. CONCLUSIONS
10
The numerical procedures for the yield line analysis of
Note: All dimensions in meters. slabs by dip and strike angles method, including the
Plastic moments Loading applied
generation of all kinematically admissible yield line
patterns from a single set of dip and strike angles,
M
Magnitude of point load = 2F evaluation of the external work done and internal
Total patch load = 5F energy dissipation, and determination of the load
0.5 M Total uniformly distributed load = F
factor as a function of the dip and strike angles, have
Figure 8. Example 5: polygonal slab with square opening been generalized for applications to slabs with

Advances in Structural Engineering Vol. 16 No. 7 2013 1231

 Yield Line Analysis of Reinforced Concrete Slabs with Openings by Dip and Strike Angles Method

arbitrary-shaped openings. Basically, the generalized Johnson, D. (1995). “Yield-line analysis by sequential linear
numerical procedures are the same as those of the programming”, Solids and Structures, Vol. 32, No. 10,
original dip and strike angles method except that the pp. 1395–1404.
external work done and internal energy dissipation Kennedy, G. and Goodchild, C.H. (2004). Practical Yield Line
within the openings are discounted. This generalized Design, The Concrete Centre, Surrey, UK.
dip and strike angles method is currently the only Kumar, V. and Roy, B.N. (1991). “Limit analysis of reinforced
computerized yield line analysis method applicable to concrete slabs”, Journal of Structural Engineering (Madras),
arbitrary-shaped slabs with arbitrary-shaped openings. Vol. 17, No. 4, pp. 127–136.
The generalized dip and strike angles method for Kumar, V. and Prakash, D. (2001). “Yield line analysis of two way
yield line analysis of slabs with openings was first reinforced concrete slabs with central opening”, Journal of
applied to some simple cases for verification. Exact Structural Engineering (Madras), Vol. 27, No. 4, pp. 251–262.
agreement between the numerical results of the Kwan, A.K.H. (2004). “Dip and strike angles method for yield line
generalized analysis method and the results obtained by analysis of reinforced concrete slabs”, Magazine of Concrete
previous researchers or by manual calculation has been Research, Vol. 56, No. 8, pp. 487–498.
achieved, thus verifying the correctness and accuracy of Lash, S.D. and Banerjee, A. (1967). “Strength of simply supported
the generalized analysis method. It was then applied to square plates with square openings”, Transactions of the
conduct a parametric study on the effect of the location Engineering Institute of Canada, Vol. 10, No. 5, pp. 3–11.
of opening. The parametric study showed that for a Middleton, C.R. (1997). “Concrete bridge assessment: an alternative
simply supported rectangular slab, regardless of whether approach”, The Structural Engineer, Vol. 75, No. 23–24,
the opening is rectangular or circular, the opening pp. 403–409.
should not be made near the centre of the slab. Lastly, it Middleton, C.R. (2008). “Generalised collapse analysis of concrete
was applied to some complicated cases to demonstrate bridges”, Magazine of Concrete Research, Vol. 60, No. 8,
the versatility of the generalized method. All these pp. 575-585.
results showed that the generalized analysis method is a Park, R. and Gamble, W.L. (2000). Reinforced Concrete Slabs, 2nd
useful tool for the design of slabs with openings. Edition, John Wiley & Sons, New York, USA.
Thavalingam, A., Jennings, A., McKeown, J.J. and Sloan, D.
REFERENCES (1998). “A computerised method for rigid-plastic yield-line
Dickens, J.G. and Jones, L.L. (1988). “A general computer program analysis of slabs”, Computers and Structures, Vol. 68, No. 6,
for the yield-line solution of edge supported slabs”, Computers pp. 601–612.
and Structures, Vol. 30, No. 3, pp. 465–476. Thavalingam, A., Jennings, A., Sloan, D. and McKeown, J.J. (1999).
Hognestad, E. (1953). “Yield-line theory for the ultimate flexural “Computer-assisted generation of yield-line patterns for
strength of reinforced concrete slabs”, ACI Journal, Vol. 49, uniformly loaded isotropic slabs using an optimisation strategy”,
No. 3, pp. 637–656. Engineering Structures, Vol. 21, No. 6, pp. 488–496.
Ingerslev, A. (1923). “The strength of rectangular slabs”, Journal of Zaslavsky, A. (1967). “Yield-line analysis of rectangular slabs with
Institution of Structural Engineers, Vol. 1, No. 1, pp. 3–14. central openings”, ACI Journal, Vol. 64, No. 2, pp. 838–844.
Islam, S. and Park, R. (1971). “Yield-line analysis of two way
reinforced concrete slabs with openings”, The Structural NOTATION
Engineer, Vol. 49, No. 6, pp. 269–276. A area of patch load or uniformly distributed
Jackson, A.M. and Middleton, C.R. (2013). “Closely correlating load
lower and upper bound plastic analysis of real slabs”, The a, b, c coefficients of deflection equation
Structural Engineer, Vol. 91, No. 1, pp. 34–40. F magnitude of point load
Johansen, K.W. (1962). Brudlinieteorier, Copenhagen, Denmark. (in H intensity per unit length of line load
Danish) (English translation: Yield Line Theory, Cement and L length of line load
Concrete Association, London, UK.) M plastic moment
Jones, L.L. (1962). “Recent British advances in yield line analysis by N number of slab segments
the equilibrium method”, Proceedings of the International NF number of point loads
Symposium on Flexural Mechanics of Reinforced Concrete, NH number of negative yield lines
Sawyer, Miami, Florida, USA, pp. 295–381. NL number of line loads
Jones, L.L. and Wood, R.H. (1967). Yield-Line Analysis of Slabs, NP number of patch loads
Thames and Hudson, London, UK. NS number of positive yield lines
Johnson, D. (1994). “Mechanism determination by automated yield- P intensity per unit area of patch load
line analysis”, The Structural Engineer, Vol. 72, No. 19, p×q total number of Gauss points for numerical
pp. 323–327. integration

1232 Advances in Structural Engineering Vol. 16 No. 7 2013

A.K.H. Kwan, J.Y.K. Lam and P.L. Ng

U intensity per unit area of uniformly x, y coordinates of a point within a slab segment
distributed load z vertical deflection
Wext external work done ∆
length increment along yield line for
WextF external work done by point loads numerical integration
WextL external work done by line loads η weighting coefficient of Gauss point
WextP external work done by patch loads θ dip angle
WextU external work done by uniformly distributed λ load factor
load φ strike angle
Wint internal energy dissipation ψ plastic rotation of yield line

Advances in Structural Engineering Vol. 16 No. 7 2013 1233