African Concrete Code symposium- 2005

DESIGN OF GROUND BEARING

CONCRETE SLABS
JOHN KNAPTON
John Knapton Consulting Engineers Ltd.
85 Monkseaton Drive
Whitley Bay NE26 3DQ
UK
www.john-knapton.com

SYNOPSIS: A series of ground bearing slab design methods is presented in this

Paper. The methods are all similar in that they model the loading regime
numerically, they assign elastic properties to the ground and they use the
resulting information to produce data which allows a slab to be understood and
installed. They differ in the way in which they transform the loading regime and
ground properties into design output data. The appropriate design procedure
depends upon the nature of the loading, upon the engineering properties of the
materials from which the slab is to be constructed and upon the end use of the
slab. For each design method, a worked example is presented. The paper also
includes information on loads likely to be applied to industrial floors or hard
standings and gives guidance on the provision of movement joints.

156
 African Concrete Code symposium- 2005

INTRODUCTION

The design of ground bearing concrete slabs poses several

problems for the structural engineer. Design methods for ground
bearing slabs are not usually included in codes of practice because
they differ radically from those design methods adopted for
conventional reinforced or prestressed concrete elements. Yet there
are usually far greater savings to be made in accurate design of
ground bearing concrete than there are in any other structural
element. The floor to a distribution warehouse is often of area
12,000m2 and thickness 200mm. Bringing this thickness down by
25mm can save 300m3 of reinforced concrete. The design methods
presented in this paper ensure that ground bearing floor slabs are
designed optimally.

Also, the load induced tensile flexural stresses in ground bearing

slabs are accommodated by the concrete, not by reinforcement.
Usually, ground bearing slabs are reinforced, but the reinforcement
is selected on the basis of controlling cracking resulting from
restraint to shrinkage. Shrinkage occurs as the concrete dries and
results in linear shrinkage and curling. The friction between the
ground and the shrinking concrete develops tensile stress within the
concrete and the self-weight of the concrete develops flexural stress
as it attempts to reposition the curled slab on the ground.

Additionally, the load regime applied to ground bearing slabs is

more complex than is often the case for conventional structural
elements. Whereas loads can frequently be represented as
distributed or point loads, in the case of ground bearing slabs, the
effect of either static or rolling patch loads needs to be considered.
Furthermore, the effect of the interaction of nearby patch loads has
to be assessed. Stresses resulting from restraint to shrinkage or
usually secondary but in the case of ground bearing concrete slabs,
they are of primary importance.

The design methods presented in this paper address these issues.

Data is included which can form the basis for the development of a

157
 African Concrete Code symposium- 2005

design code for ground bearing slabs. The failure of a ground

bearing floor is one of the most feared structural problems. The
disruption to the user is often total, involving decanting the
operation to a temporary industrial unit while the floor is replaced.
But perhaps the crucial matter is the fact that as Africa begins to
prosper, there will be an overwhelming demand for new industrial
buildings of all types and limited funding will create an imperative
for a cost effective engineered ground bearing floor slab. The
design methods described in this Paper are:

1. Using Hetenyi's equations for a slab subjected to a uniformly

distributed load over part of its area.
2. Using Westergaard's equations for patch loads.
3. Using Meyerhof's equations for patch loads
4. Using the Eurocode 2 method for punching shear
5. Using design charts developed from Westergaard equations for a
slab subjected to patch loads

Table 1 shows the situations where each of the 5 design methods is

appropriate.
Additionally, this paper provides design guidance on the selection
of an arrangement of joints so as to eliminate cracking.

Table 1: Design methods reviewed in this Paper

Design Type of slab Type of load Output
Method
Hetenyi Floor slabs Distributed Slab thickness

Modified All ground bearing Patch load Slab thickness &

Westergaard slabs elastic deflexion

Meyerhof Floor slab where Patch load/Point load Floor Slab thickness
cracking is acceptable
Eurocode 2 All ground bearing Patch load Confirmation that slab
slabs will not suffer shear
failure
Westergaard All ground bearing Patch load/Point load Slab thickness
design charts slabs

158
 African Concrete Code symposium- 2005

In all of the design methods, concrete characteristic flexural

strength is used as the basis of stress assessment and load induced
stresses are computed using classical methods. See Table 2 for
concrete strength values.

The following factors have to be taken into account in ground floor

design.
a) Loading Regime
b) Strength of Concrete
c) Strength of existing ground and effect of the sub-base

STRENGTH OF CONCRETE

Concrete slabs are frequently constructed from C30, C35 or C40

concrete with a minimum cement content of 300kg/m3 with a
slump of 50mm or less. Design is based upon comparing concrete
characteristic flexural strength with calculated flexural stresses
whereas specification is by characteristic compressive strength.
Table 2 shows flexural strength values for a range of commonly
used concretes.

LOADING

Loading on ground bearing floors comprises distributed loads and

patch or point loads, typically applied by vehicles and materials
handling equipment. Industrial floors are commonly required to
accommodate distributed loads of 37.5kN/m2. This is a
Table 2: Mean and characteristic 28-day flexural strength values
for various concrete mixes. See Table 1.1 for other properties of
concrete.

159
 African Concrete Code symposium- 2005

Flexural Strength (N/mm2)

Mean Characteristic
___________________________________________________________
_____
Plain C30 concrete 2.0 1.8
Microsilica C30 concrete 2.4 1.9
3
C30 concrete 20kg/m ZC 60/1.00 steel fibre 2.8 2.0
3
C30 concrete 30kg/m ZC 60/1.00 steel fibre 3.2 2.2
3
C30 concrete 40kg/m ZC 60/1.00 steel fibre 3.8 2.7
___________________________________________________________
_____
Plain C35 concrete 2.2 1.95
Microsilica C35 concrete 2.6
2.0
C35 concrete 20kg/m3 ZC 60/1.00 steel fibre 3.0 2.1
3
C35 concrete 30kg/m ZC 60/1.00 steel fibre 3.4 2.3
3
C35 concrete 40kg/m ZC 60/1.00 steel fibre 4.0` 2.9
__________________________________________________________
Plain C40 Concrete 2.4 2.1
Microsilica C40 Concrete 2.8 2.15
3
C40 concrete20kg/m ZC 60/1.00 steel fibre 3.2 2.2
3
C40 concrete 30kg/m ZC 60/1.00 steel fibre 3.6 2.5
3
C40 concrete 40kg/m ZC 60/1.00 steel fibre 4.2 3.2
Plain C45 Concrete 2.7 2.3
Microsilica C45 Concrete 3.1 2.4
C45 concrete20kg/m3 ZC 60/1.00 steel fibre 3.5 2.5
3
C45 concrete 30kg/m ZC 60/1.00 steel fibre 4.0 2.8
3
C45 concrete 40kg/m ZC 60/1.00 steel fibre 4.8 3.6
Plain C50 Concrete 3.0 2.5
Microsilica C50 Concrete 3.4 2.6
C50 concrete20kg/m3 ZC 60/1.00 steel fibre 3.8 2.7
C50 concrete 30kg/m3 ZC 60/1.00 steel fibre 4.2 3.0
C50 concrete 40kg/m3 ZC 60/1.00 steel fibre 5.0 3.8

Plain C55 Concrete 3.4 2.7

160
 African Concrete Code symposium- 2005

Microsilica C55 Concrete 3.8 2.8

C55 concrete20kg/m3 ZC 60/1.00 steel fibre 4.2 2.9
C55 concrete 30kg/m3 ZC 60/1.00 steel fibre 4.6 3.2
C55 concrete 40kg/m3 ZC 60/1.00 steel fibre 5.4 4.1

ZC 60/1.00 refers to a commonly used anchored bright wire fibre

of total length 60mm and wire diameter 1.00mm.
particularly high load which is rarely applied. It is specified in
order that an industrial floor can be guaranteed to accommodate
any future loading which a floor might need to sustain. It is in fact
rare for a floor/hardstanding to be loaded by a distributed load.
However, it has become a simple form of common currency which
is used by the developers of industrial premises to assess the
suitability of industrial premises for different end users. Even when
a floor/hardstanding is specified in terms of distributed load, the
designer should nonetheless consider the likely point or patch
loads. This is because they often produce much higher stress levels
than do distributed loads.

Floors subjected to loads applied by storage racking are often

designed to accommodate 10 tonne or 12 tonne end frames. End
frames each have two legs so the patch load is half of these values.
However, it is common for storage arrangements to cause two such
frames to be positioned side by side. Therefore, the floor is
commonly designed for point or patch loads of 10 tonne or 12
tonne.

External hard standings may need to accommodate highway

vehicles which will typically have patch loads of up to 6tonne. In
container handling areas, such as ports, Reach Stackers or Front
Lift Trucks handling 40ft containers have front axle loads of up to
110 tonne so the half-axle patch load can be up to 55 tonne.
Dynamics may need to be applied to account for braking, turning
and general dynamics. These factors may add 25% to the load so
the total design patch load may be 70t.

161
 African Concrete Code symposium- 2005

Container stacking areas need to be designed to accommodate the

patch loads applied by the corner castings of container feet. A
typical loaded 40ft container weighs 26t so the corner casting load
is 6.5t. Often containers are stored in blocks with four corner
castings in close proximity. This applies an effective patch load of
26t to the slab.

Containers are often stacked three high so the load applied to the
slab is 78 tonne, although a reduction is sometimes made of say
20% to account for the fact that it is unlikely that all stacked
containers will be loaded to their capacity. Sometimes containers
are stacked five high which increases the loading further.

Sometimes, the repetition of loads needs to be taken into account.

A simple expedient is to increase the load by 50% and assume that
it can be applied an infinite number of times. Alternatively, a full
fatigue analysis will need to be undertaken which can prove
mathematically difficult if the loads vary in intensity and duration
of application.

Load and strength partial safety factors of 1 are usually applied in

the case of industrial floors or hard standings. This is because their
failure is unlikely to lead to severe consequences such as death or
injury. It is accepted that if the slab is overloaded or is under
strength then cracking will develop. The alternative would be to
provide an uneconomic solution.

STRENGTH OF EXISTING GROUND

AND EFFECT ON SUB-BASE

The design methods require a value for the modulus of subgrade

reaction (K) which defines the stiffness of the material beneath the
slab. The following four values of K are used in the design
procedures:

K = 0.013 N/mm3 - very poor ground

162
 African Concrete Code symposium- 2005

K = 0.027 N/mm3 - poor ground

K = 0.054 N/mm3 - good ground
K = 0.082 N/mm3 - very good ground (no sub-
base needed)

A subgrade with a K value of 0.027 N/mm3 will deflect vertically

by 1mm when a vertical pressure of 0.027 N/mm2 or 27kN/m2 is
applied through a standard disk. The beneficial effect of a granular
sub-base may be taken into account by increasing K according to
the thickness and strength of the sub-base. However, the above
figures can be used conservatively.

STRESS IN CONCRETE

The stress in a floor slab depends upon:

a) The properties of the subgrade
b) The loading regime:
i) Uniformly Distributed Load (UDL)
ii) Patch/Point loads
c) The thickness of the floor slab
d) The strength of the sub-base

HETENYI DESIGN METHOD FOR UNIFORMLY

DISTRIBUTED LOADING

The common loading system comprises alternate unloaded aisles

and loaded storage areas. The maximum negative bending moment
(hogging) occurs within the centre of the aisles and is given by:

q
M hog = (Ba ' Bb ' )
22

where a = width of the aisle

b = width of the loaded area

163
 African Concrete Code symposium- 2005

a' = a/2
b' = a/2 + b
q = UDL(characteristic load x load factor)(N/mm2)
= (3K/Eh)1/4
K = modulus of subgrade reaction
E = concrete modulus (10,000N/mm2 for sustained load)
h = slab thickness

The maximum positive bending moment (sagging) occurs beneath

the centre of the loaded area (loading from adjacent blocks is
ignored ):

Msag = q/2 2.B(1/2) b

where
Bx = e-x .sin x

The two moment equations can be simplified into a single

conservative equation for any combination of aisle width and
stacking zone width:

Mmax = -0.168q/ 2

The corresponding maximum flexural stress is given by:

max = 6.Mmax /h2

max=1008 q/( 2 h2) (N/mm2)

where the maximum flexural strength cannot exceed the relevant

characteristic value from Table 2. To ease calculation, Table 3
shows values of 2h2 for common combinations of modulus of
subgrade reaction (K) and slab thickness.

164
 African Concrete Code symposium- 2005

Table 3 Values of 2h2 for combinations of slab thickness and

modulus of sub grade reaction.
______________________________________________________
__________
2h2 for slab thickness (mm):
150 175 200 225
250
Modulus of subgrade
reaction K (N/mm3)
0.082 0.061 0.066 0.070 0.074 0.078
0.054 0.049 0.053 0.057 0.060
0.063
0.027 0.035 0.038 0.040 0.043
0.045
0.013 0.024 0.026 0.028 0.029 0.031

Uniformly Distributed Load Example

Consider a 150mm thick floor slab carrying a uniformly distributed

load of 50kN/m2 between aisles on poor ground.

1.008q
max = N / mm 2
.h
2 2

q = 0.05 N / mm 2 x 2 = 0.10 N / mm 2

(In the above equation, 2 is the fatigue factor. A value of 2

means a load of 0.05N/mm2 can be repeated indefinitely)

For poor ground K = 0.027N/mm3

Therefore:
0.10
max = 1.008 = 2.88 N / mm
2

0.035

165
 African Concrete Code symposium- 2005

From the characteristic strength values in Table 2, this stress can be

withstood by a C40 concrete incorporating 40kg/m2 of 60mm long
1mm diameter anchored steel fibre. This design solution can be
used for any combination of aisle and stacking zone width.

MODIFIED WESTERGAARD DESIGN

METHOD FOR PATCH LOADS

Highway vehicles and handling equipment apply loads to the

surface of a floor as patch loads. In most cases, sufficient accuracy
is gained by assuming the patch to be circular and to apply uniform
stress throughout the patch. These assumption lead to minor errors
- the contact patch shape in the case of a commercial vehicle is
nearer to a rectangle than to a circle but to assume accurate shaped
patch loads would preclude the use of Westergaard equations. The
error in assuming constant stress circular loading is very small and
is conservative i.e. true analysis would lead to a slightly thinner
floor. The maximum flexural tensile stress occurs at the bottom (or
top, in the case of corner loading) of the slab under the heaviest
wheel load. The maximum stress under a patch load can be
calculated by the elastic Westergaard4.1 and Timoshenko
equations, sometimes referred to as the modified Westergaard
equations:

a) patch load in mid-slab (i.e. more than 0.5m from slab edge)

0.275(1 + ) 0.36 Eh 3
max = P. log .........................1
h2 Kb 4

b) patch load at edge of slab

P 0.20 Eh3
max = 0.529(1 + 0.54 ) log ................................2
h2 Kb 4

166
 African Concrete Code symposium- 2005

c) patch load at slab corner

0 .6


3P 1.41b
max = 2 1 0.25 ........................................3
h Eh 3


( )
12 1 2 K

where
max = flexural stress (N/mm2)
P = point load (N) i.e. characteristic wheel load x
fatigue factor
= Poisson's ratio, usually 0.15
h = slab thickness (mm)
E = elastic modulus, usually 20,000N/mm2
K = modulus of sub grade reaction (N/mm3)
b = radius of tyre contact zone (mm)
= (P/ .p)1/2
p = contact stress between wheel and floor (N/mm2)

Twin wheels bolted side by side are assumed to be one wheel

transmitting half of the axle load to the floor. In certain cases wheel
loads at one end of an axle magnify the stress beneath wheels at the
opposite end of the axle, distance S away (S is measured between
load patch centres). To calculate the stress magnification, the
characteristic length (radius of relative stiffness, l) has to be found
from Equation 4.

0.25
Eh 3
l = ........................4
(
12 1 2
K)

Values of l are shown in Table 4. Once Equation 4 has been

evaluated, the ratio S/l can be determined so that Fig. 1 can be used

167
 African Concrete Code symposium- 2005

to find Mt/P. Where Mt is the tangential moment. The stress under

the heaviest wheel is to be increased to account for the other wheel.
The stress to be added is calculated from the following equation.

Mt 6
add = P2 .............................5
P h2

where:
P = greater patch load,
P2 = other patch load.
Sum the stresses and verify that the characteristic flexural strength
in Table 4.1 has not been exceeded for the appropriate concrete
mix.

Figure 1: Relationship between ratio S/l and ratio Mt/P used in

assessing the influence of load proximity, Highway Vehicle
Example using Westergaard Equations

168
 African Concrete Code symposium- 2005

Consider a highway vehicle with a rear axle load of 8000kg and

assume a slab thickness of 200mm on good ground
(K=0.054N/mm3). The fatigue factor is 2.0 so the slab can
withstand an infinite number of such repetitions. The design axle
load is 160,000N (i.e.80,000N x fatigue factor(2.0)) so the design
wheel load is 80,000N

p = contact stress between wheel and floor = 0.7N/mm2

b = radius of contact area = (W/ p)1/2
= (80,000/0.7)1/2
=191mm

Substituting known values into Equation 2:

80000 20000 x 2003
max = 0.529(1 + 0.54 x0.15) log 0.2
200 2 0.054 x1914
max = 3.03 N / mm 2

beneath one wheel.

Assume that the wheel at the other end of the axle is 2.7m away (S
= 2700mm)
The radius of relative stiffness (l), using Equation 4 is:

0.25
20000 x 2003
l =
(
12 1 0.15 )
2
0.054

l = 709mm

S 2700
= = 3.8
l 709

From Figure 1
Mt
= 0.005
P

169
 African Concrete Code symposium- 2005

Therefore, using Equation 5

Mt 6
add = P2
P h2

6
= 0.005 2
.80000
200
= 0.06 N / mm2

Total Stress = 3.03 + 0.06N/mm2

= 3.09N/mm2

By comparing this stress with the characteristic strength of C40

concrete reinforced with 40kg/m3 steel fibres(3.2N/mm2), it can be
seen that the proposed mix is satisfactory. A C40 concrete
incorporating 40kg/m2 60mm long 1mm diameter anchored steel
fibres is inadequate for this design. However, the inclusion of a
250mm thick granular sub-base would enhance K from 0.054 to
0.073 which would reduce the stress to below the characteristic
value.

MEYERHOF DESIGN METHOD

FOR POINT & PATCH LOADS

The Meyerhof design method is based upon yield line analysis and
provides a relatively simple way of selecting the thickness of a
concrete slab for a single patch load and for some combinations of
similar patch loads. Meyerhof developed his equations by
considering how a concrete slab would fail. His initial work related
to suspended slabs but he later applied the method to ground
bearing slabs. He considered the way in which a slab would
ultimately fail and equated the potential energy lost when the load
moved downwards to the strain energy absorbed along the cracks
which form at the time of failure.

170
 African Concrete Code symposium- 2005

Consider a patch load applied to a ground bearing slab gradually

increasing in magnitude. The slab directly beneath the load sags
and develops tension towards the underside of the slab. Further
away from the patch load, the tension develops in the upper part of
the slab as the bending transforms from sagging to hogging. For
failure to occur, the concrete has to crack in these tension zones.
The yield line method assumes that at failure, the slab breaks into a
series of plates and that sufficient cracks (or "plastic hinges") need
to form to generate a collection of hinged plates. Figure 2 shows
typical patterns of hinges generated by one, two and four patch
loads.

SINGLE PATCH LOAD YIELD LINE PATTERN

Twin point load yield line pattern

FOUR POINT LOAD YIELD LINE PATTERN

Figure 2: Yield lines or plastic hinges resulting from single, twin

and four patch loads.

171
 African Concrete Code symposium- 2005

These are the Meyerhof equations for point loads:

i) Single point load applied within the body of the slab:

[
Pu = 2 M p + Mn ]
ii) Single point load applied along a free edge of a slab:

([ ] )
Pu = Mp + Mn / 2 + 2Mn

iii) Single point load applied at a corner of a slab:

Pu = 2M n

iv) Two similar point loads spaced x apart and away from edges or
corners of the slab:

1.8x
Pu = 2 +
l
[
M p + Mn ]
v) Four similar point loads applied at corners of a rectangle of side
lengths x and y, all four loads away from edges or corners of the
slab.

1.8( x + y )

Pu = 2 +
l
[
M p + Mn ]

The corresponding equations for patch loading are:

i) Single patch load applied within the body of the slab:

a
[ ]
Pu = 4 M p + Mn / 1
3l

ii) Single patch load applied along a free edge of a slab:

172
 African Concrete Code symposium- 2005

[ ]
Pu = ( M p + M n + 4M n ) / 1
2a
3l

iii) Single patch load applied at a corner of a slab:

4a
Pu = 2 1 + M n
l

iv) Two similar patch loads of radius a spaced x apart and away
from edges or corners of the slab:

4 1.8 x
Pu = + [M p + M n ]
1 a l a
3l 2

v) Four similar patch loads applied at corners of a rectangle of side

lengths x and y, all four loads away from edges or corners of the
slab.

4 1.8( x + y )
Pu = + [
M p + Mn ]
a a
1 3l l
2

Where:
Pu = Ultimate load
Mp = Ultimate positive (sagging) moment of resistance of the slab
Mn = Ultimate negative (hogging) moment of resistance of the slab
The above two values can be obtained from the equation:
h2
M p ,n = f
6

173
 African Concrete Code symposium- 2005

where: f = characteristic flexural strength of concrete - see Table 2

h = slab thickness
a = radius of patch load
x = spacing of two point of patch loads (centre to centre of
load)
y = spacing of four point loads forming corners of a
rectangle - y measured at right angles to x.
l = radius of relative stiffness:

0.25
Eh3
l =
( )
12 1 K
2

(See Table 4.4 for values of l. Table 4.4 reproduced here for
convenience)

Table 4: Radius of relative stiffness values for different slab

thicknesses and support conditions

Modulus of subgrade reaction (K) (N/mm3)

Slab thickness (mm) 0.013 0.027 0.054
0.082
150 816 679 571
515
175 916 763 641
578
200 1012 843 709
639
225 1106 921 774
698
250 1196 997 838
755
275 1285 1071 900
811
300 1372 1143 961

E = Elastic modulus = 20000 N/mm2

= Poissons Ratio = 0.15

174
 African Concrete Code symposium- 2005

Note that the Meyerhof patch load equations apply when :

(a l ) 0.2
For values of a/l between 0.2 & zero, interpolate between the patch
load and the point load Mu values.

Highway Vehicle Example using Meyerhof Equations

Consider a same highway vehicle with a rear axle load of 8000kg.

Assume a slab thickness of 200mm on good ground (K =
0.054N/mm3). The load is dynamic so a factor of 1.6 can be
applied. In the case of a highway vehicle, this will take care of
both dynamics and fatigue. The design axle load is 128,000N
(i.e.80,000N x dynamic/fatigue factor(1.6)) so the design wheel
load is 64,000N (64kN).
Consider a slab of thickness 200mm and try C40 concrete
containing 40kg/m3 of anchored steel fibres.

Radius of patch load, a = (64,000/0.7)1/2

= 170mm

From Table 2, characteristic strength of concrete = 3.2N/mm2.

Apply the Material Partial Safety Factor of 1.3 to obtain design
strength: 3.2/1.3=2.5N/mm2.

h2
Now use: M p ,n = f to obtain the ultimate positive and
6
negative moments of resistance of the proposed slab:

Mp = Mn = 2.5.2002/6 = 16,667N-mm per mm width of slab.

From Table 4, the Radius of Relative Stiffness is 709mm. Note
that this is the distance from the centre of the load to where the
bending moment in the slab changes from sagging to hogging.

175
 African Concrete Code symposium- 2005

Consider the situation where this single patch load is applied away
from the slab edges or corners.

[ ]
a
Pu = 4 M p + M n / 1
3l

Pu = 455,312N = 455kN (64kN required)

Consider the slab edge condition:

[ ]
Pu = ( M p + M n + 4 M n ) / 1
2a
3l

Pu = 204kN (64kN required)

Consider the load applied at the corner of the slab:

4a
Pu = 21 + M n
l

Pu= 65,304N = 65kN (64kN required)

Consider also both wheels separated by 2.7m.
Introduce all of the now known parameters into the equation:

4 1 .8 x
Pu = + [
M p + Mn ]
1 a l a
3l 2

(x is the spacing of the rear wheels, which in this case is 2700mm.)

4 4860
Pu = + [16,667 + 16,667 ]
170 170
1 2127 2127 2

176
 African Concrete Code symposium- 2005

12.56 4860
Pu = + 33,334
0.92 2042

= 534.344N = 534kN
i.e. each wheel load = 534/2 = 267kN (64kN required)

This result indicates that a 200mm thick slab comprising C40 steel
fibre reinforced concrete can sustain between 455kN and 65kN
depending upon the position of the load in relation to the slab. The
Westergaard solution suggested that such a slab could sustain
80kN. Both methods used similar levels of safety factor, although
they were applied in different ways.

The reason for the difference between Westergaard (or other elastic
methods set out in this Paper) and Meyerhof solutions is the
redundancy of a concrete slab. The Westergaard solution considers
the critical stress point and ensures that the actual stress remains
below the strength of the concrete by the safety factor at that point.
Meyerhof recognises that the slab can crack only when sufficient
plastic hinges have formed to develop one of the mechanisms
shown in Figure 2. This means that there is a significant reserve of
strength beyond the condition when stress reaches strength at the
critical point i.e. there is a significant reserve of strength beyond
the Westergaard elastic failure criterion. Loading beyond this
condition will not lead to failure, rather the stress pattern is
redistributed until stress has reached concrete strength along all of
the plastic hinges which form the collapse mechanism.

As can be appreciated from the Meyerhof example, the load can

grow to several times the Westergaard limiting value. However,
there remains the possibility that cracks will develop in the critical
stress location once the Westergaard elastic condition has been
surpassed. In the case of an unreinforced slab, this could lead to
the slab's breaking into two which would be unacceptable in most
circumstances. However, in the case of steel mesh or steel fibre
reinforced concrete, the crack would be bridged by the
reinforcement. The Author's tests have shown that steel fibre

177
 African Concrete Code symposium- 2005

reinforced concrete has a residual flexural strength after the

development of the first crack. Therefore, it is likely that steel fibre
reinforced concrete slabs will crack before the Meyerhof condition
is reached but that the cracks will be controlled by the steel fibres.
Also, it should be recognised that the benefit of redundancy is less
marked in the situation where a single patch load is applied at the
corner of a slab. For this reason, Meyerhof is most beneficial in
jointless slabs where this condition can be omitted.

Which should be used, Westergaard (or the other elastic methods

set out in this Paper) or Meyerhof? The Author has been involved
in full scale tests in which a steel fibre reinforced slab (C40
concrete, 30kg/m3 steel fibre content) of thickness 140mm
withstood a patch load of 290kN, suggesting that the Meyerhof
method is not unconservative. For several years, floor construction
contractors have reported that slabs much thinner than those which
Westergaard's equations would require appear to function
satisfactorily. Therefore, it would seem sensible to use the
Meyerhof equations for ground bearing floors subjected to point or
patch loads in situations where cracks are acceptable. However, in
situations where any degree of cracking would be unacceptable,
then Westergaard equations or the other elastic methods presented
in this Paper should be used. This will be the case generally for
external slabs subjected to the effects of weather. The Meyerhof
equations should be used for steel mesh or steel fibre reinforced
concrete floor slabs where some controlled cracking is acceptable.
However, it should be recognised that stresses induced by patch or
point loads may represent a small part of the stress regime.
Restraint to thermal and moisture induced movement may be the
principal structural action. In the past, it is likely that the
conservative Westergaard equations provided the requisite factor of
ignorance which allowed those additional stresses to be
accommodated. If Meyerhof equations are to be used in design,
then a full understanding of the behaviour of the slab will be
required. In particular, the possibility of punching shear failure
must be investigated and the stresses arising from restraint to
shrinkage and curling must be evaluated. The slab must then be

178
 African Concrete Code symposium- 2005

proportioned on the basis of the most adverse combination of all of

the stresses which are likely to occur, not just those resulting from
patch or point loads.

EUROCODE 2 DESIGN METHOD FOR PUNCHING

SHEAR CHECK OF PATCH LOADS

Even when elastic and plastic design methods confirm that a slab is
adequate, there is the possibility that the load can punch directly
through the slab as a result of shear stresses exceeding the shear
strength of the slab around the perimeter of the patch load. A
check should be made as follows. Firstly, calculate the area over
which the shear stress acts. This is the depth of the slab multiplied
by the perimeter of the load patch. Divide the factored load by this
area to obtain the shear stress. This stress should not exceed the
shear strength of the concrete vm which is calculated from:

vm = 0.5vf cd

where fcd= design cylinder strength (see Table 1.1 - divide the
characteristic cylinder strength by 1.3)

and v = 0.6 1
f ck
250

(fck = characteristic cylinder strength - see Table 1.1)

Consider a 175mm thick C35 slab loaded by a circular 90kN patch
load of diameter 250mm. The area over which the vertical shear
force acts is: x load patch diameter x slab thickness = .250.175
= 137,444mm2. The shear stress is 90,000/137,444 = 0.65N/mm2.
Now calculate the allowable shear stress as follows:

v = 0.6 1
f ck
250

179
 African Concrete Code symposium- 2005

[
v = 0.6 1 28.5
250
]
(see Table 1 for the fck value of 28.5)
Therefore, v = 0.53
Substitute in: vm = 0.5vf cd where the cylinder design strength fcd is
28.5/1.3 = 22N/mm2. (1.3 is the Material Partial Safety Factor for
concrete)

Therefore, the allowable shear stress vm = 0.5 x 0.53 x 22 =

5.83N/mm2. By comparing this figure with the applied shear stress
of 0.65N/mm2, it can be concluded that there is no likelihood of
punching shear failure in this instance.
Note that for suspended slabs, a further shear check is made at a
distance of twice the slab depth from the perimeter of the patch
load. Because of the support offered to a ground bearing slab, this
is rarely necessary.

DESIGN METHOD FOR PATCH LOADS USING DESIGN

CHARTS DERIVED FROM WESTERGAARD EQUATIONS

The above example illustrates the complexity of undertaking

calculations in the case of patch loads. A simplified version of the
elastic Westergaard method has been developed by the Author
specifically for designing industrial floors/hardstandings. In order
to eliminate much of the effort, a series of 10 design charts has
been developed using Westergaard calculations. These charts are
reproduced at the end of this Paper.
The procedure is as follows:

1. Assess the existing conditions:

2. Determine the Actual Point Load (APL) and Modulus of
Subgrade Reaction (K) values to confirm the category of
subgrade.
3. Calculate the additional stress generated by point or patch loads
or wheels in close proximity.

180
 African Concrete Code symposium- 2005

If the distance between loads (S) is greater than 3m, the APL
can be used directly (depending on the radius of contact zone)
to calculate the thickness of the slab using the relevant Design
Chart. In this case, go directly to Stage 6.
If the distance between loads is less than 3m, the radius of
relative stiffness (l) has to be determined. Table 4.4 shows
values of radius of relative stiffness for different K values and
slab thicknesses.

4. From Figure 1, determine Mt/P, the ratio of the tangential

moment to the greater point load by calculating S/l. Then use
the following equation to determine the stress to add.

Mt 6
add = P2 ........................5
P h2
where P = greater patch load (N)
P2 = other patch load (N)
5. From Table 2, select a proposed concrete mix, hence
characteristic strength, char.
6. When two patch loads are acting in close proximity (i.e. less
than 3m apart), the greater patch load (P) produces a flexural
strength max directly beneath it's point of application. The
nearby smaller patch load (P2) produces additional stress add
beneath the larger load. Calculate max ,

max = char add ...........................6

7. Calculate the Single Point Load (SPL) which, acting alone

would generate the same flexural stress as the actual loading
configuration.

SPL = APL char .........................7
max

181
 African Concrete Code symposium- 2005

where APL = Actual Point Load

8. Prior to using the Design Charts, it is necessary to modify the

SPL (Single Point Load) to account for contact area as well as
wheel proximity to obtain the ESPL (Equivalent Single Point
Load). Design Charts 1 to 10 apply directly when patch loads
have a radius of contact between 150mm and 250mm. Some
racking systems and pallet transporters have a contact radius of
less than 150mm and some vehicles have a contact radius
greater than 250mm. In these cases multiply the patch load by a
factor in Table 5 prior to use in the Design Chart.
9. Use the corresponding Design Chart for the mix selected in 3)
to determine slab thickness and return to 3) if an alternative
concrete mix is required.

Table 5: Patch load multiplication factors for loads with a radius of

contact outside the range 150mm to 250mm

Modulus of Subgrade Reaction (K) N/mm3

Radius of contact (mm) 0.013 0.027 0.054
0.082
50 1.5 1.6 1.7
1.7
100 1.2 1.2 1.3
1.3
150 1.0 1.0 1.0
1.0
200 1.0 1.0 1.0
1.0
250 1.0 1.0 1.0
1.0
300 0.9 0.9 0.9
0.9

Design Example for multiple patch loading using design chart

A concrete floor is subjected to two patch loads. A 60kN patch load

is applied 1m away from a 50kN patch load. The 60kN patch load
has a contact zone radius of 100mm and the 50kN patch load has a

182
 African Concrete Code symposium- 2005

300mm radius. The existing ground conditions are poor

(K=0.027N/mm3).
Assume thickness of slab = 225mm
The radius of relative stiffness, l is given by Table 4
From Table 4
l = 921mm
distance apart = 1m (i.e. S = 1000mm)

S 1000
= = 1.086
l 921
From Figure 1

Mt
= 0.053
P

so from Equation 5,

6
add = 0.053 2
50000
200

add = 0.4N/mm2
Try steel fibre reinforced C30 concrete with a characteristic
strength of 2.2N/mm2 (30kg/m3 steel fibre dosage - see Table 2)

i.e. char = 2.2N/mm2

max = char add
= 2.2 - 0.4
= 1.8N/mm2

This is the maximum flexural stress which the 60kN load can be
allowed to develop. Calculate the Single Point Load using
Equation7:

183
 African Concrete Code symposium- 2005

SPL = APL char
max

2.2
= 60 = 73kN
1.8

From Table 5, the modified factor to be applied to the SPL to

obtain the ESPL is 1.2:
73 x1.2 = 88kN
o.
From Design Chart N 4, thickness of slab required = 225mm.

A 225mm thick C30 concrete slab incorporating 30kg/m3 steel

fibre is adequate for this design.

SPACING OF JOINTS

Both construction and movement joints are required in ground

bearing industrial slabs. Construction joint spacings will depend
upon the type of equipment used to install the floor. Recent
developments in large pour technology allow floors/external
hardstandings of up to 3,000m2 to be constructed without
construction joints in one day. If no movement joints were included
in such a floor/hardstanding, it would crack. This is because the
friction between the ground and the shrinking concrete develops
tensile stress within the concrete and the self-weight of the concrete
develops flexural stress as it attempts to reposition the curled slab
on the ground. This cracking is controlled by including movement
joints at spacings which experience indicates cracking would
otherwise occur. The crack spacings are rarely calculated because
it is difficult at design stage to predict the factors which will govern
those spacings.

Table 6 shows joint spacings for various types of concrete. The

joint spacings in the Table have been used successfully for many

184
 African Concrete Code symposium- 2005

years in the UK. Some consider that spacings can be greater than
12m for steel fibre reinforced floors/hardstandings. Whilst 14m
joint spacings will probably be acceptable for most applications,
the additional movement which would occur at joints might lead to
loss of aggregate interlock and joint degradation. Joints spacings
are frequently designed to coincide with column spacings or other
features such as recesses or floor width changes. In any project, it is
necessary to develop a joint layout identifying all practical issues
prior to the placing of concrete.

Table 6: Common Concretes and suggested joint spacings.

Concrete type Joint spacing (m)

Plain C30 concrete 6
Microsilica C30 concrete 6
20kg/m3 ZC 60/1.00 steel fibre reinforcement C30 concrete 6
30kg/m3 ZC 60/1.00 steel fibre reinforcement C30 concrete 8
40kg/m3 ZC 60/1.00 steel fibre reinforcement C30 concrete 10
Plain C40 concrete 6
Microsilica C40 concrete 6
20kg/m3 ZC 60/1.00 steel fibre reinforcement C40 concrete 6
30kg/m3 ZC 60/1.00 steel fibre reinforcement C40 concrete 10
40kg/m3 ZC 60/1.00 steel fibre reinforcement C40 concrete 12

Joint spacings can be increased beyond the figures shown in Table

6 by the use of steel mesh reinforcement. Table 7 shows
commonly available steel mesh sizes. Steel mesh is assumed to
carry the tensile forces developed in the concrete resulting from
restraint to shrinkage caused by loss of moisture or temperature.
Consequently mesh allows greater joint spacings. Steel mesh is
often used in long strip floor construction as it can be placed
conveniently without the need for cutting. Where necessary, mesh
sheets should be lapped at their edges and ends by 450mm.
Overlapping can result in an unacceptable build up of thickness of
reinforcement.

There is often debate regarding the positioning of the mesh in

relation to the depth of the concrete. From an installation point of
view, it is simpler to position the mesh near the bottom of the
concrete. The benefit of positioning the mesh near the upper

185
 African Concrete Code symposium- 2005

surface is that it can control any surface cracking which might

develop as a result of installation defects.

Table 7: Commonly available steel mesh sizes.

Longitudinal wires Cross wires

Mesh Nominal
Nominal Pitch Area Pitch Area Mass
reference wire size
size (mm) (mm) (mm2/m) (mm) (mm2/m) (kg/m2)
(mm)
Square mesh
A393 10 200 393 10 200 393 6.16
A252 8 200 252 8 200 252 3.95
A193 7 200 193 7 200 193 3.02
A142 6 200 142 6 200 142 2.22
A98 5 200 98 5 200 98 1.54
Structural
mesh 12 100 1131 8 200 252 10.9
B1131 10 100 785 8 200 252 8.14
B785 8 100 503 8 200 252 5.93
B503 7 100 385 7 200 193 4.53
B385 6 100 283 7 200 193 3.73
B283 5 100 196 7 200 193 3.05
B196
Long mesh
C785 10 100 785 6 400 70.8 6.72
C636 9 100 636 6 400 70.8 5.55
C503 8 100 503 5 400 49 4.34
C385 7 100 385 5 400 49 3.41
C283 6 100 283 5 400 49 2.61
Wrapping
mesh 5 200 98 5 200 98 1.54
D98 2.5 100 49 2.5 100 49 0.77
D49
Stock sheet Longitudinal wires Cross wires Sheet area
size Length 4.8m Width 2.4m 11.52m2

Pitch : The centre-to-centre spacing of wires in a sheet of mesh.

Width : When specifying mesh the width is the overall dimension measured in
the direction of the cross wires.
Wrapping mesh: Wire usually of Grade 250 for use in wrapping mesh.
Sheet size: Mesh types "A" and "B" is delivered in standard sheets of 4.8 x
2.4m, or in scheduled size sheets. Mesh type "C" is available in sheets or rolls.

186
 African Concrete Code symposium- 2005

It is common for long mesh to be used to enhance joint spacings.

Table 8 shows joint spacings for four commonly used structural
mesh sizes.

Table 8: Enhanced joint spacing with Long Mesh reinforcement

Weight of
Structural Mesh Wire Sizes Joint Spacing
Reinforcement
Name (Diameter in mm) (m)
(kg/m2)
6mm main wires
C283 2.61 16
5mm cross wires
7mm main wires
C385 3.41 21
5mm cross wires
8mm main wires
C503 4.34 27
5mm cross wires
9mm main wires
C636 5.55 35
6mm cross wires

CONCLUSIONS

The paper presents a series of simple design tools which can be

used to proportion the thickness of a concrete ground bearing
industrial floor or external hard standing and which can be used to
select an arrangement of movement joints. The methods have
proved successful in many parts of the world and the ground
bearing floors and hard standings which they produce can be
expected to have a design life in excess of 40 years. The designer
has a choice between no reinforcement, using polypropylene fibres
which provide a small level of reinforcement, steel fibres which
provide a significant level of reinforcement or steel mesh which
provides an enhanced level of reinforcement. In the case of ground
bearing industrial concrete floors and hard standings, the
reinforcement is not provided to accommodate flexural and/or
tensile stresses as is the case with conventional reinforced concrete.
Rather, it controls the cracking resulting from restraint to
shrinkage. Therefore, the provision of progressively higher levels
of reinforcement permits wider joint spacing without the risk of
cracking, or rather with a series of narrow cracks which are of no
consequence.

187
 African Concrete Code symposium- 2005

The information included in the Paper has been selected to include

that which would be of relevance to an African Concrete Code. It
is a self-contained document and would allow designers to
proportion ground bearing concrete slabs with confidence.

188