Proceedings of the Institution of

Civil Engineers
Structures & Buildings 146
November 2001 Issue 4
Pages 363^370

Paper 12378
Received 09/06/2000
Accepted 02/04/2001

Keywords:
H. J. Pam J. C. M. Ho
columns/concrete structures/
Associate Professor, Research Student,
design methods & aids
Department of Civil Department of Civil
Engineering, The University Engineering, The University
of Hong Kong of Hong Kong

Flexural strength enhancement of confined reinforced concrete

columns
H. J. Pam and J. C. M. Ho
As part of a continuing research study, this paper h height of cross section
proposes a new design aid to calculate the actual ke confinement effectiveness coefficient
moment capacity of confined reinforced concrete M moment
columns. Up to now the moment capacity of a reinforced MBS design moment capacity evaluated using BS 8110
concrete column is calculated based on the code’s equivalent rectangular concrete stress block taking
guidelines for an unconfined section. As most reinforced into account partial safety factors for strength of
concrete columns contain transverse or confining materials
reinforcement, which will enhance the column moment Mc maximum moment capacity evaluated from the
capacity, the actual moment capacity will be much moment-curvature curve based on the modified
higher than the unconfined moment capacity. This Mander et al. concrete stress-strain model
additional flexural strength will increase the shear force Mref ultimate moment capacity evaluated using BS 8110
demand in the column, and if it is not accounted for in equivalent rectangular concrete stress block without
the design will jeopardise the column to fail in shear. partial safety factors for the strength of materials
In this study the actual moment capacity of a confined P compressive axial load on columns
reinforced concrete column is obtained by multiplying R ratio of strain increase to stress increase due to
the moment capacity calculated using the BS 8110 confinement
method with the proposed flexural strength enhance- R2 coefficient of correlation
ment factor. By using regression analysis, an equation r Ec/(Ec – Esec)
for the flexural strength enhancement factor is derived s vertical spacing of confining steel (centre-to-centre)
as the function of all the parameters that have effects s’ vertical clear spacing between confining steel
on the moment capacity. An example is presented to wi clear distance between each longitudinal steel bar
show the accuracy of the proposed method. supported by the corner of a hoop or cross tie
x e/ecc
a, b coefficients of regression line
NOTATIONS e strain in concrete corresponding to stress f
Ac area of concrete core measured to outside of perimeter ecc strain corresponding to peak stress in confined
hoops concrete
Ag gross area of cross-section eco strain corresponding to peak stress in unconfined
As area of longitudinal steel concrete
b width of cross section r ratio of longitudinal steel area to gross-section area
cx concrete core dimension parallel to x-axis (measured rc ratio of longitudinal steel area to nominal concrete
to centre line of confining steel) core area (measured to centre line of confining steel)
cy concrete core dimension parallel to y-axis (measured rs volumetric ratio of confining steel (measured to
to centre line of confining steel) outside of confining steel)
dc diameter of circular confining steel (centre-to-centre) o analytical flexural strength enhancement factor
Ec initial tangent modulus of concrete od design flexural strength enhancement factor
Esec secant modulus of concrete at peak stress
f stress in concrete corresponding to strain e
fcc confined concrete compressive strength 1. INTRODUCTION AND RESEARCH SIGNIFICANCE
fco peak stress of unconfined concrete stress-strain Columns play a very important role in building structures in
curve = 0?75fcu resisting the gravity loads as well as lateral loads caused by
fcu unconfined concrete compressive cube strength wind or earthquake. In resisting the gravity and lateral loads,
fl effective lateral confining pressure columns will be subjected to high axial compressive and shear
fy yield strength of longitudinal steel forces. Therefore, a significant amount of transverse reinforce-
fys yield strength of confining steel ment, which has a double function in resisting shear forces as

Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho 363

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 well as providing confinement to the concrete core, must be concrete stress block recommended by BS 8110,2 whereas the
provided in reinforced concrete columns. The presence of confined moment capacity is calculated based on the confined
transverse reinforcement and axial compressive load to a concrete stress-strain relationship by Mander et al.,3 which is
certain extent will cause enhancement in the flexural strength modified to facilitate the comparison with the reference
capacity of the reinforced concrete column. The effect of axial moment. Moment-curvature analysis is used to obtain the
load towards strength enhancement has been implicitly taken maximum value for the confined moment capacity. The cases
into account in the calculation of the column section considered include all column sections that have various
unconfined moment capacity. However, the effects of trans- combinations of the concerned variables, they are
verse or confining reinforcement towards the enhancement of
the section moment capacity have never been considered in the . concrete cube strength, ranging from 25 to 60 MPa
design of reinforced concrete columns. . longitudinal steel ratio, ranging from 1 to 6%
. volumetric ratio of transverse or confining steel ranging
Transverse reinforcement in reinforced concrete columns has from 1 to 3?5%
several functions, they are (a) to resist shear forces, (b) to restrain . yield strength of transverse or confining steel, either 250 or
longitudinal reinforcement from buckling and (c) to confine 460 MPa
concrete core. All of these functions will contribute in . axial compressive load level, ranging from 0?1 to 0?6.
enhancing the flexural ductility and strength of the reinforced
concrete column. Therefore, the ductility and moment capacity The proposed method in this study to calculate the flexural
of a reinforced concrete column will increase with increasing strength enhancement factor is expected to become a powerful
amount of the confining steel. The increase in ductility is design aid for structural design engineers in calculating more
always beneficial to a column regardless of any implications exactly the actual moment capacity of confined reinforced
as the result of it. However, the increase in moment capacity concrete columns.
from the original design value, will underestimate the shear
demand in the column. Depending on the amount of confining
2. ANALYTICAL STRESS-STRAIN CURVES OF
steel and the magnitude of axial compressive load, the increase
CONFINED REINFORCED CONCRETE SECTIONS
in moment capacity can be as high as 40% of the unconfined
One of the available stress-strain models for confined
moment capacity. The increase in moment capacity will
reinforced concrete sections is chosen and modified in this
automatically increase the shear demand. It has been widely
study to be used in generating the moment-curvature curves
known that shear failure is much more brittle than bending
for reinforced concrete columns.
failure, and thus should be prevented. It has to be noted that as
the result of underestimation in the shear capacity, shear failure
may happen prior to bending failure. To prevent this from The chosen model was proposed by Mander et al.3 The major
happening, the actual confined moment capacity that takes parameters in the concrete stress-strain curve are effective
into account the confining steel must be used instead of lateral confining pressure, fl, and confined concrete strength,
the unconfined moment capacity in determining the shear fcc, which are expressed in equations (1) and (2), respectively.
reinforcement in the column.
1 fl ¼ 05ke rs fys
1, 2
Most of the structural design codes require that reinforced
concrete columns be provided with a certain amount of sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !
794fl fl
transverse or confining steel. However, up until now, they 2 fcc ¼ fco 1254 þ 2254 1 þ 2
fco fco
do not provide guidelines as to how to determine the
actual moment capacity of confined reinforced concrete
columns. where ke = confinement effectiveness coefficient, which will be
explained further
This paper proposes a method to estimate the actual moment rs = volumetric ratio of confining steel (measured to
capacity of confined reinforced concrete columns. The major outside of confining steel)
parameter investigated is the flexural strength enhancement fys = yield strength of confining steel
factor, which is defined as the ratio of confined moment fco = peak stress of unconfined concrete stress-strain
capacity to unconfined moment capacity. The flexural strength curve.
enhancement factor is a function of a series of variables,
i.e. amount and yield strength of confining steel, amount of In the original model of Mander et al.3 stress-strain curve, the
longitudinal steel, concrete compressive strength and magni- value of fco is taken to be the concrete compressive cylinder
tude of axial compressive load. The amount of confining steel strength. In order to obtain a fair comparison between the
is expressed in terms of volumetric ratio with respect to the unconfined concrete stress-strain curves of Mander et al. and
volume of the concrete core, the amount of longitudinal steel is BS 8110, the value of fco is modified and replaced by 0?75fcu ,
expressed in terms of area ratio with respect to the gross cross- where fcu is concrete compressive cube strength. This modified
section area, and the axial compressive load is expressed in value is obtained by approximating (within 5% error) the first
terms of axial load level with respect to the full axial load moment of area about its origin of both the stress-strain models
capacity. for strain up to 0?0035 and for concrete cube strength ranging
from 25 to 60 MPa. The first moment of area is used instead of
The unconfined moment capacity, which is served as the the area, as the former is closely proportional to the section
reference, is calculated based on the equivalent rectangular moment capacity.

364 Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 In equations (1) and (2), it should be noted that: (a) for an fcc xr
unconfined section, the value of fl becomes zero and fcc 5 f ¼
r  1 þ xr
becomes equal to fco or 0?75fcu; and (b) it is presented here only
for circular or square confinement, in which the confining where x = e/ecc
pressure is similar in all directions. e = strain in concrete corresponding to f
r = Ec/(Ec – Esec)
The parameter ke in equation (1) is the ratio of the smallest Ec = initial tangent modulus of concrete
effectively confined concrete core area, which is located Esec = secant modulus of concrete at peak stress, equal
midway between two layers of confining steel, to the nominal to fcc/ecc for confined concrete and fco/eco for
concrete core area, which is measured to the centre line of the unconfined concrete.
confining steel. The effectively confined concrete core area is
obtained by assuming that the core concrete spalls in the form In equation (5) the confined concrete strength, fcc , and thus
of a series of second-degree parabolas between the clear the effective lateral confining pressure, fl , control the shape
distance of the longitudinal steel bars supported by the corner of the descending branch of the stress-strain curve, or in other
of a hoop or cross tie in the horizontal plane, and between the words the ductility of the concrete.
clear spacing of the confining steel in the vertical plane.
The reasons for adopting the modified Mander et al. concrete
Mander et al. assumed that the angles between the slope of the
stress-strain model are: (a) only one single equation defines
tangents at the ends of the parabola and its base line to be 458.
both the ascending and descending branches of the stress-strain
Based on all these assumptions, ke was derived. Equations (3a)
curve; (b) the model can also be used for unconfined concrete
and (3b) express ke respectively for circular and rectangular
sections; (c) the model can be applied to any shape of concrete
confinement.
member section confined by any kind of transverse reinforce-
ment (spirals, cross ties, circular or rectangular hoops); and
s0
1  05 (d) there is no residue stress at the tail end of the stress-strain
3a ke ¼
dc
1  rc curve. The last phenomenon resembles more the behaviour of
confined concrete under compression at large deformation.

 P 2    The concrete stress-strain curves for various volumetric ratios

wi s0 s0
1 1 1 of confining steel together with the concrete stress-strain curve
3b 6cx cy 2cx 2cy
ke ¼ recommended in BS 81102 are shown in Fig. 1. All the stress-
1  rc
strain curves were obtained based on the assumed reinforced
concrete section described in Fig. 2.
where s’ = clear spacing between confining steel
dc = diameter of circular confining steel (measured to the
Comparison of the stress-strain curves in Fig. 1 reveals that:
centre line of the steel)
rc = ratio of longitudinal steel area to nominal concrete . The ductility increases with the volumetric ratio of confining
core area (measured to centre line of confining steel.
steel) . The maximum stress increases with the volumetric ratio of
wi = clear distance between each longitudinal steel bar confining steel.
supported by the corner of a hoop or cross tie . The strain corresponding to the peak stress increases with the
cx = concrete core dimension parallel to x-axis volumetric ratio of confining steel.
(measured to centre line of confining steel) . Compared to the unconfined modified stress-strain curve of
cy = concrete core dimension parallel to y-axis Mander et al., the stress-strain curve of BS 8110 under-
(measured to centre line of confining steel). estimates the ductility and value of maximum stress. It
should be noted that the first moment of area up to the strain
Note that for a square confinement, the value of cx is equal of 0?0035 of these two curves about the origin are
to cy. approximately the same. The error is only about 0?4% for
fcu = 50 MPa.
The strain corresponding to the peak stress is expressed as
60

  
fcc 50
4 ecc ¼ eco 1 þ R 1
fco
40
BS 8110
Stress: MPa

Modified
where eco which is equal to 0?002, is the strain corresponding to 30 Mander et al.:
ρs = 0%
the peak stress for unconfined concrete and R, which is an 20 ρs = 1%
ρs = 2%
empirical value, is the ratio of the strain increase to stress ρs = 3%
10
increase due to confinement. According to Mander et al.,3 the
value of R can vary from 3 to 6, and a value of 5 is 0
0 0·002 0·004 0·006 0·008 0·01
recommended for analytical modelling purposes. Strain

Fig. 1. Concrete stress-strain curves of BS 81102 and modified

The entire confined concrete stress-strain relationship is Mander et al.3 for fcu = 50 MPa
expressed following the Popovics’ model4 in the form of:

Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho 365

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 (b) Assume a value of neutral axis depth measured from the
400 mm
extreme concrete compressive fibre (normally starting with
Section properties: one-half of the cross-section height).
Longitudinal bar diameter = 20 mm
Transverse bar diameter = 6–12 mm (c) Calculate the strain and the corresponding stress at the

400 mm
Concrete cover to stirrups = 20 mm
Concrete compressive strength, fcu = 50 MPa centroid of each longitudinal reinforcement bar.
Steel yield stress: Longitudinal = 460 MPa
Transverse = 250 MPa (d ) Determine the stress distribution in the concrete com-
Transverse steel spacing: s = 60 mm
pressive region based on the modified Mander et al.3
stress-strain model for the given volumetric ratio of
confining steel. The resultant concrete compressive force is
Fig. 2. Properties of the assumed reinforced concrete section
for stress-strain curves then obtained by numerical integration of the stress over
the entire compression region.
(e) Calculate the axial force from equilibrium and compare
3. ANALYTICAL MOMENT-CURVATURE AND AXIAL with the applied axial load. If the difference lies within the
LOAD ^ MOMENT INTERACTION CURVES FOR specified tolerance, the assumed neutral axis depth is
CONFINED REINFORCED CONCRETE COLUMNS adopted. The moment capacity and the corresponding
Using the previously discussed stress-strain curves for concrete curvature of the section are then calculated. Otherwise, a
and the elasto-plastic stress-strain model for longitudinal new neutral axis depth is determined from the iteration
reinforcement as shown in Fig. 3, for a specific ratio of (using secant method) and steps (c) to (e) are repeated until
confining steel, a series of analytical moment-curvature curves it converges.
can be generated for various axial compressive load levels. The ( f ) Assign the next value, which is larger than the previous
assumptions and procedure used in generating the moment- one, to the extreme concrete compressive fibre strain and
curvature curves are outlined in the following. repeat steps (b) to (e).
(g) Repeat the whole procedure until the complete moment-
curvature curve is obtained.
Assumptions.

Figure 4 shows a series of moment-curvature curves for different

(a) The strain is linear across the height of the section,
levels of axial compressive load that simulate the column axial
that is, plane bending.
loads from high to low storeys. Note that axial load level is
(b) The tensile strength of the concrete is ignored.
defined as P/fcuAg, where P is compressive axial load acting on
(c) The concrete spalls at a strain of 0?004.
columns and Ag is gross area of cross section. The moment-
(d ) The initial tangent modulus of the concrete, Ec , is adopted
pffiffiffiffiffiffi curvature curves in Fig. 4 were derived using the reinforced
from ACI 318-99,1 which is equivalent to 4230 fcu .
concrete section described in Fig. 2 for a volumetric ratio of
(e) For the elasto-plastic stress-strain curve of the longitudinal
confining steel of 1%. It is observed in Fig. 4 that the ductility
reinforcement, the yield stress and Young’s modulus are
of the moment-curvature curves decreases when the axial load
460 and 200 000 MPa, respectively.
level increases, and the moment beyond the maximum point
( f ) In determining the location of the neutral axis,
degrades more rapidly at higher axial load levels. For the low
convergence is assumed to have been reached when the
axial load levels of 0 and 0?1, the moment-curvature curves
difference between the calculated axial load and the
closely resemble an elasto-plastic shape. The maximum
applied axial load is within 1% of the latter value.
moment increases when the axial load level is increased to
about 0?4, which is considered in this study as the ‘balanced’
Procedure.
level. The rate of moment increase is higher at low axial load
levels, and the highest moment increase occurs when the axial
(a) Assign a value to the extreme concrete compressive fibre
load level is increased from 0?1 to 0?2. For axial load levels
strain (normally starting with a very small value).
higher than the balanced axial load level, the maximum
moment decreases as the axial load level is increased. It has to
Stress

be noted in Fig. 4 that the small drop of moment immediately

after the peak is due to the spalling of the cover concrete when
fy the strain reaches larger than 0?004.

All of the above behaviours of the moment-curvature curves

Tension
are consistent with those of the corresponding axial load–
moment interaction curve, which is shown in Fig. 5. Shown
also in Fig. 5 are the axial load–moment interaction curves for
Strain
higher volumetric ratios of confining steel (rs = 2% and 3%)
Compression and for unconfined columns based on BS 8110.2

Yield strain
Both the axial load and moment in Fig. 5 are non-dimensionalised.
– fy
The former is equal to axial load level. Values of the maximum
axial load capacity were obtained by assuming the concrete
Fig. 3. Stress-strain relationship (symmetrical in tension and had reached the peak stress and the longitudinal steel had
compression) for longitudinal reinforcement yielded, which can be expressed respectively for unconfined
and confined column section as equations (6) and (7).

366 Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 fcc = peak stress of confined
450 concrete
fy = yield strength of longi-
400
tudinal steel
P/Agfcu: Ag = gross area of cross
350
0 section
300 Ac = area of concrete core
0·1 measured to outside of
Moment: kNm

250 0·2 perimeter hoops

As = area of longitudinal
200 0·3
steel.
0·4
150
Equation (7a) is valid when
0·5
the strain, which is uniform
100
0·6 across the height of the cross
50 section, is less than or equal
to the spalling strain and
0 equation (7b) is valid when
0 0·02 0·04 0·06 0·08 0·1 the strain is larger than the
Curvature: rad/m spalling strain, beyond
which the concrete cover is
Fig. 4. Moment-curvature curves for rs = 1% of different axial load levels (based on modified considered to have spalled.
Mander et al. concrete stress-strain model3)
The axial load–moment
interaction curves were also
1·2 calculated based on the
BS 8110
column properties described
Modified
Mander et al.: in Fig. 2. It has to be noted
1·0 ρs = 0%
that for the confined column
sections, the moment and the
ρs = 1% extreme compressive fibre
ρs = 2% strain are taken at the peak
0·8 point of the corresponding
ρs = 3%
moment-curvature curve.
P/(fcubh)

0·6 It is obvious from the axial

load–moment interaction
curves that the increase in
moment happens from zero
0·4 to the balanced axial load
level, which ranges from
about 0?32 for unconfined
0·2 sections to about 0?38 for
confined sections. Beyond the
balanced axial load level,
the moment decreases with
0 the increase in axial load. It
0 0·025 0·05 0·075 0·1 0·125 0·15
2
M/(fcubh ) is very important to observe
in Fig. 5 that the moment
increases with the increase of
Fig. 5. Axial load^moment interaction curves
the confining steel content,
and the largest rate of
increase occurs when the confining steel content is increased
6 Punconfined = fcoAg + fyAs
from 0 to 1%. It is also worth noting that for the interaction
curves derived using the modified Mander et al.3 stress-strain
7a Pconfined = fccAg + fyAs model with rs = 2%, the maximum axial capacity is lower than
that with rs = 1%. This is because the effect of concrete strength
7b Pconfined = fccAc + fyAs enhancement in the concrete core due to confinement is offset
by spalling of the concrete cover when the strain (ecc) reaches
where fco = peak stress of unconfined concrete, which is equal to greater than the spalling strain, which is equal to 0?004. Table 1
0?67fcu and 0?75fcu for BS 8110 and modified Mander lists the values of concrete strain corresponding to the concrete
et al.3 models, respectively peak stress and the maximum axial load level for fcu = 50 MPa.

Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho 367

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 rs (%) Model fcc (MPa) ecc P/fcuAg without applying the partial safety factors for the concrete and
longitudinal steel.
0 BS 8110 33?5 0?0035 0?81
0 Modified Mander et al.3 37?5 0?0020 0?89 It is proposed in this study to relate the flexural strength
1 Modified Mander et al.3 43?2 0?0035 1?01 enhancement factor with confinement effectiveness factor,
2 Modified Mander et al.3 48?5 0?0049 0?93
3 Modified Mander et al.3 53?4 0?0063 1?01 which is expressed as rs fys/fcu , using a linear regression
analysis. This confinement effectiveness factor is almost similar
to the one proposed by Hoshikuma et al.,5 except here the
Table 1. Concrete strain corresponding to concrete peak
stress and maximum P/fcuAg for fcu = 50 MPa cylinder strength is replaced by cube strength. For a reinforced
concrete column section, by varying the parameters rs/fys
(part of the confinement effectiveness factor) for a specific
combination of fcu , r (ratio of longitudinal steel area to gross
4. ANALYTICAL FLEXURAL STRENGTH
section area) and P/Ag fcu , a number of moment-curvature
ENHANCEMENT IN CONFINED REINFORCED
curves can be generated, each of which will give a value of
CONCRETE COLUMNS
maximum moment capacity (Mc). Each of these maximum
From the previous moment-curvature and axial load–moment
moments is then compared to its corresponding unconfined
analyses, it has been shown that the flexural strength increases
moment capacity (Mref), from which the flexural strength
with the increase of the confining steel content. In design
enhancement factor (o) is obtained. Each coordinate point,
practice, the flexural capacity of reinforced concrete columns,
(o, rs fys/fcu), consisting of a pair of values of flexural strength
which always contain transverse steel, are always calculated
enhancement factor and confinement effectiveness factor
based on the code’s recommended concrete stress-strain
respectively, is plotted on the graph of o against rs fys/fcu
relationship,1, 2 in which the effect of confining steel is ignored.
(see Fig. 6). A straight line is later fitted in the graph by linear
regression analysis. The best fitting line is expressed in the
In addition to strength enhancement, confinement also results form of
in higher ductility. Ductility always gives beneficial effects to
structural members. However, flexural strength enhancement Mc r fys
must be treated with caution, as it will cause an increase in the 8 ¼a s þb
Mref fcu
shear demand, which might cause a brittle failure. Therefore, it
is essential for practising structural engineers to estimate the where a and b are coefficients of the regression line.
flexural strength enhancement of confined reinforced concrete
columns, so that premature shear failure will be prevented. The above procedure is repeated for other combinations of
fcu (from 25 to 60 MPa), r (from 1 to 6%) and P/Ag fcu
In addition to confinement, flexural strength enhancement (from 0?1 to 0?6). Fig. 6 shows some examples of the graphs.
depends on many more parameters, such as concrete com- It can be observed from the figure that the flexural strength
pressive strength, amount and distribution of longitudinal enhancement factor can reach as high as 1?4.
reinforcement, as well as
yield stress and shape of
confining steel. These
1·5
1·4
effects on the flexural y = 0·2995x + 1·177 y = 0·8775x + 1·0829
strength enhancement will be R 2 = 0·9251 R 2 = 0·9083
1·4
1·3
investigated and quantified
1·3
Mc/Mref

by means of the moment-

Mc/Mref

1·2
curvature curves. P/A f P/Agfcu = 0·5
1·2 g cu = 0·4
fcu = 25 MPa fcu = 60 MPa
ρ = 1% ρ = 3%
1·1 1·1
The flexural strength
enhancement factor, o, is 1·0 1·0
defined in this study as the 0 0·2 0·4 0·6 0·8 1·0 0 0·1 0·2 0·3 0·4
ρsfys/fcu ρsfys/fcu
ratio of the actual moment (a) (b)
1·20
capacity, Mc , of a confined 1·5

reinforced concrete column to y = 0·2749x + 1·0445 y = 0·7168x + 1·1254

1·15 R 2 = 0·9367 1·4 R 2 = 0·9536
a reference moment capacity,
Mref. The former (Mc) is the 1·3
Mc/Mref

Mc/Mref

maximum moment value 1·10

obtained from the moment- P/Agfcu = 0·3 1·2 P/Agfcu = 0·6

fcu = 40 MPa fcu = 40 MPa
curvature curve based on the 1·05 ρ = 3% ρ = 4%
1·1
modified Mander et al. stress-
strain model,3 and the latter 1·0 1·0
0 0·1 0·2 0·3 0·4 0·5 0 0·1 0·2 0·3 0·4 0·5
(Mref) is the ultimate moment ρsfys/fcu ρsfys/fcu
capacity calculated using (c) (d)

the equivalent rectangular

concrete stress block Fig. 6. Graphs of Mc/Mref against rs fys/fcu
recommended by BS 81102

368 Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 By performing regression analyses on all the obtained values of 400 mm
a and b, it is determined that they can be represented with good
accuracy by
Variation of parameters:
 075  2 Axial load level: 0·1 to 0·6

400 mm
f P ρ: 1–6%
9 a ¼ 00269 cu05 ρs: 1–3·5%
r Ag fcu fcu: 25–60 MPa
fys = 250 or 460 MPa
Concrete cover to main bar = 25 mm
 2
P
10 Ag fcu
b ¼ 00678 025 075 þ 10109 Fig. 9. Properties of the hypothetical column section
fcu r

Figures 7 and 8 show the relationships of a and b as expressed It has to be noted that all of the above analyses and calculation
in equations (9) and (10), respectively. were performed using a hypothetical reinforced concrete
column section described in Fig. 9.
Shown also in Figs 6 to 8 are the coefficients of correlations
(R2), each of which is very close to 1. 5. FLEXURAL STRENGTH ENHANCEMENT FACTOR
USED IN DESIGN PRACTICE
By combining equations (8) to (10), the following equation The above proposed flexural strength enhancement model is
can be derived to obtain the analytical flexural strength based on the comparison of the moment capacity calculated
enhancement factor for reinforced concrete columns by the modified concrete stress-strain model of Mander et al.3
and that calculated according to BS 81102 without taking
 2 2  2 3 into account the partial safety factors for the concrete and
P P
Ag fcu 6 Ag fcu 7 longitudinal steel. However, in design practice, these partial
11 6 7
o ¼ 00269rs fys þ 600678 025 075 þ 101097 safety factors must always be considered when calculating the
025
fcu r 05 4 fcu r 5
design moment capacity, MBS, according to BS 8110. As the
consequence, the actual moment capacity obtained by multi-
plying MBS with the flexural strength enhancement factor (o)
obtained from equation (11) will be smaller than the actual
moment capacity calculated by the modified concrete stress-
1·8
strain model of Mander et al.3 To compensate for the shortfall,
1·5
the analytical flexural strength enhancement factor, o, is
adjusted to become design flexural strength enhancement
1·2 factor, od , which is expressed as

α 0·9 Mref
12 od ¼ o
MBS
0·6 y = 0·0269x
R 2 = 0·9479
where the value of od will always be greater than 1.
0·3

0 6. CALCULATION EXAMPLE
0 10 20 30 40 50 60 The following calculation example demonstrates the use of the
(P/Agfcu)2fcu0·75ρ–0·5
proposed flexural strength enhancement model. The cross-
section details and material properties are shown in Fig. 10. It
Fig. 7. Graph of a against (f 0?75 0?5
cu /r ) (P/Ag fcu) will be clear from the example that the designer has to
calculate two different moment capacities based on BS 8110.2
The first moment capacity (MBS) takes into account the partial
safety factors for the material strengths, whereas the second
1·4
moment capacity (Mref) is without partial safety factors for the
material strengths.
1·3

1·2 500 mm
β
1·1 y = 0·0678x + 1·0109 Section properties:
Longitudinal steel bar diameter = 25 mm
R 2 = 0·9203 Transverse steel bar diameter = 12 mm
500 mm

Clear cover to transverse steel: 25 mm

1·0 Concrete cube strength, fcu = 40 MPa
Steel yield stress: Longitudinal steel: 460 MPa
Transverse steel: 250 MPa
0·9 Transverse steel spacing: s = 100 mm
0 1 2 3 4 5 6 Axial load level: P/(fcuAg) = 0·4

(P/Agfcu)2fcu–0·25ρ–0·75
Fig. 10. Cross-section details and material properties for
Fig. 8. Graph of b against (P/Agfcu)2/f 0?25
cu r
0?75
calculation example

Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho 369

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.
 The calculation procedure is as follows considered as the actual moment capacity of the confined
reinforced concrete column. This actual moment capacity was
(a) Longitudinal steel ratio compared with that obtained using the code’s method2 by the
analytical flexural strength enhancement factor.
025p252  12
r¼ ¼ 00236
5002
The analytical flexural strength enhancement factor was
(b) Volumetric ratio of confining steel developed using regression analysis in which the coefficients of
correlation were all greater than 0?9. This analytical flexural
025p122  ð450  12Þ  8 strength enhancement factor was subsequently modified to
rs ¼ ¼ 00196
4502  100 become design flexural strength enhancement factor, so that it
(c) Analytical flexural strength enhancement factor is ready for use by structural design engineers to obtain the
actual moment capacity of confined reinforced concrete
  columns. It is shown that the flexural strength enhancement
042
o ¼ 00269  00196  250 factor can reach as high as 1?4.
40025  0023605
 2

04
þ 00678 025 þ 10109
40  00236075 An example of how to apply the proposed method is given. It is
¼ 1137 shown in the example that the method gives very good results.
(d ) Factored and unfactored moment capacity calculated
according to BS 81102 The proposed method of design aid in this research will enable
practising structural design engineers to determine the actual
MBS ¼ 6554 kNm moment capacity of confined reinforced concrete columns.
Mref ¼ 7943 kNm

(e) Design flexural strength enhancement factor 8. ACKNOWLEDGEMENTS

The work described in this paper was carried out with generous
7943 financial assistance provided by the Research Grants Council of
od ¼ 1137  ¼ 1378
6554 Hong Kong (RGC Project No. HKU 7012/98E).
( f ) Actual moment capacity
REFERENCES
Mc ¼ od MBS ¼ 1378  6554 ¼ 9031 kNm
1. ACI COMMITTEE 318. Building Code Requirements for
Structural Concrete (ACI 318-99) and Commentary
(ACI 318R-99), American Concrete Institute, Farmington
It should be noted that if the actual moment capacity is Hills, 1999.
calculated from the first principle using the theoretical 2. BRITISH STANDARDS INSTITUTION. BS 8110: Part 1: 1997,
moment-curvature curve derived from the modified concrete Structural use of concrete, Code of practice for design and
stress-strain model of Mander et al.,3 the result is 922?9 kNm. construction, British Standards Institution, London, 1997.
Hence, the difference is about 2?1%, which is considered good 3. MANDER J. B., PRIESTLEY M. J. N. and PARK R. Theoretical
for engineering design purposes. stress-strain model for confined concrete. Journal of
Structural Engineering, ASCE, 1988, 114, No. 8,
7. SUMMARY AND CONCLUSIONS 1804–1825.
The confined concrete stress-strain relationship proposed by 4. POPOVICS S. A numerical approach to the complete stress-
Mander et al.3 was modified and then used to generate a large strain curves of concrete. Cement and Concrete Research,
number of moment-curvature curves of reinforced concrete 1973, 3, No. 5, 583–599.
columns with varying parameters ( fcu, r, rs , fy and P/fcuAg). 5. HOSHIKUMA J., KAWASHIMA K., NAGAYA K. and TAYLOR A. W.
The parameters in the analysis cover a wide range of values Stress-strain model for confined reinforced concrete in
that are commonly found in the design practice. Each of the bridge piers. Journal of Structural Engineering, ASCE,
moment-curvature curves gives a maximum moment, which is 1997, 123, No. 5, 624–633.

Please email, fax or post your discussion contributions to the secretary: email: lyn.richards@ice.org.uk; fax: +44 (0)20 7799 1325;
or post to Lyn Richards, Journals Department, Institution of Civil Engineers, 1^7 Great George Street, London SW1P 3AA.

370 Structures & Buildings 146 Issue 4 Flexural strength enhancement Pam 1 Ho

Downloaded by [ University of Queensland - Central Library] on [23/12/15]. Copyright © ICE Publishing, all rights reserved.