Engineering Structures 31 (2009) 455–465

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Shear truss analogy for concrete members of solid and hollow circular
cross section
J. Turmo a,∗ , G. Ramos b,1 , A.C. Aparicio b,2
a
Department of Civil Engineering, University of Castilla- La Mancha, Avda. Camilo José Cela s/n, 13071 - Ciudad Real, Spain
b
Department of Construction Engineering, Technical University of Catalonia, C/ Jordi Girona 1-3, 08034 - Barcelona, Spain

article info a b s t r a c t

Article history: Solid circular columns are very popular for bridge pier design, due to simplicity of construction and
Received 23 October 2007 because its strength characteristics under wind and seismic loads are similar in any direction. Circular
Received in revised form elements are also used extensively as columns in buildings or as piles for foundations. Hollow core circular
22 May 2008
concrete members are much less structurally used than solid circular cross sections. However, these
Accepted 1 September 2008
Available online 10 October 2008
can be found in concrete chimneys, concrete pipes and elevated water tanks, as well as in large bridge
columns and offshore platforms. Codes do not usually propose specific formulations for evaluating the
Keywords:
shear strength of such structural types or if they do, they do so in a very simplified manner.
Shear strength Hence, an analytical model for evaluating the contribution of the transverse reinforcement in
Truss analogy concrete members of solid and hollow circular cross section is presented in this paper. After a thorough
Circular columns bibliographic research on the previous studies on the matter, the shortage of work accomplished on the
Spiral reinforcement topic has been assessed and the unsolved problems identified. As far as the authors are aware, an accurate
formula for evaluating the shear transferred by spiral reinforcement in solid members is presented for the
first time. An accurate formula for the calculation of hollow core circular columns with both vertical and
spiral reinforcement is also deduced for the first time, leading to innovative results. Assumptions made
for the calculation of this formulation have been deduced theoretically, and then checked empirically.
Hence, shear tests on circular hollow core specimens are also presented. A practical efficiency factor of
circular and spiral shear reinforcement for solid and hollow core circular concrete members to be used
for design purposes is also presented.
© 2008 Elsevier Ltd. All rights reserved.

1. Introduction
Shear strength of concrete members is a topic that focuses the
effort of multiple researchers all around the world since the end
of XIX century. Even though shear transfer mechanisms are qual-
itatively well known, there is no agreement on the quantification
of the shear strength of concrete members. Qualitatively, most of
the researchers and codes [1–4] state that the shear strength VRd
of a beam is the combination of the contribution of the concrete
(VcRd ) and the contribution of the shear reinforcement if present
(VsRd ). VcRd takes into account the shear stress transferred by the
compressed zone of the element, the dowel action, the aggregate
interlock and the arch effect. Its actual value depends on many fac-
tors, such as the tensile longitudinal steel ratio, concrete grade,
∗
Corresponding author. Tel.: +34 926 295423; fax: +34 926 295391.
E-mail addresses: jose.turmo@uclm.es (J. Turmo), gonzalo.ramos@upc.edu
(G. Ramos), angel.carlos.aparicio@upc.edu (A.C. Aparicio).
1 Tel.: +34 93 4017356; fax: +34 93 401 1036.
2 Tel.: +34 93 4016514; fax: +34 93 401 1036.
0141-0296/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engstruct.2008.09.002
aggregate size or shear span, among others. Its evaluation is very
controversial and always relies, somehow, on empirical methods.
As the value of VsRd can be easily calculated with rational models,
such as the truss analogy proposed by Ritter (1889) and Moersch
(1902), research focuses on the elaboration of methods for an accu-
rate evaluation of VcRd . Present shear research topics range from the
use of self-training algorithms to obtain accurate formulas to pre-
dict VcRd for conventional concrete [5–8]; to the evaluation of the
value of VcRd for new concrete types [9–11]; to the development
and application of equilibrium methods, such as the Strut and Tie
Method [12–14]; or to the development and application of equilib-
rium and compatibility methods, such as the Modified Compres-
sion Field Theory [15]. No improvement has been made recently in
the evaluation of VsRd , however.
Traditionally, codes [1–4,16] have proposed empirical formulae
for evaluating VcRd , drawn from tests performed mainly on
beams with rectangular webs and these results are extrapolated
to members with arbitrary cross section. This is also true for the
analytical formulae proposed for evaluating VsRd , which have been
developed for rectangular, T and I beams or box girder bridges.
VsRd is evaluated throughout the codes following the scheme
456 J. Turmo et al. / Engineering Structures 31 (2009) 455–465

Notations
Ast Area per unit length of the transverse reinforcement
Aα Area per unit length of each of the reinforcement
group with an inclination α
AΦ Area of a single rod of diameter Φ
b Width of cross-section
d Distance from extreme compression fibre to the
centroid of longitudinal tension reinforcement
D Diameter of a circular section
D′ Diameter of circular hoop or spiral reinforcement
d′ Concrete cover of the compressed reinforcement
fyt ,d Design strength of the transverse reinforcement
fyα,d Design strength of the reinforcement with an Fig. 1. Shear truss analogy.

inclination α
Iz Is the second moment of area of the whole cross where z is the lever arm; α is the inclination of vertical stirrups
section around a z axis crossing the centroid (90 deg) or bent-up bars (Fig. 1); θ is the inclination of the concrete
p Pitch of the helix struts; Aα is the area per unit length of each of the reinforcement
q Shear flow group with an inclination α ; and fyα,d is the design strength of
R Radius of a circular section the reinforcement. Some variations of this formula can be found.
R′ Radius of the circular or helical transverse reinforce- ACI 318-05 [1] substitutes z for d; d being the distance from the
ment extreme compression fibre to the centroid of longitudinal tension
st Spacing of the stirrups reinforcement. BS 8110 [2] substitutes z for the value of the length
Sz Statical moment of area about a z axis crossing the of the stirrups’ leg d − d′ ; where d′ is the concrete cover of the
centroid compressed reinforcement. In Eurocode 2 EC2 [16], Model Code
t Thickness of a hollow core circular section MC-90 [4] and Spanish standard EHE [3] a value of z = 0.9 · d
V Shear force is suggested.
VcRd Contribution of the concrete to the shear strength of None of the aforementioned codes addresses specifically the
a beam evaluation of circular elements, but the American ACI 318-05
VRd Shear strength of a beam does. The American code [1] and the New Zealand one (according
VsRd Contribution of the shear reinforcement to the shear to [17]) explicitly refer to the application of the shear design
strength of a beam formulae to circular concrete members, but with simplified
xn Depth of the compression zone methods. These recommend the calculation of shear carried by
z Lever arm or distance between the resultant force of truss mechanisms involving transverse reinforcement in circular
compressive and tensile stresses columns, by representing the section as an equivalent rectangle of
z0 Distance from the centroid of the tensile forces to width diameter D and effective depth d of 0.8 · D. A new draft of
the centre of masses of the prism the Spanish standard EHE [18] proposes a lever arm z to be used in
α Angle defining the longitudinal inclination of verti- Eq. (1) equal to 0.8 · D for circular sections.
cal stirrups The little attention paid to circular members in the codes
θ Angle defining the longitudinal inclination of the contrasts with their widespread use. In fact, circular columns
concrete struts are very popular for bridge pier design, due to simplicity of
κ Efficiency factor for spiral reinforcement on solid construction and because its strength characteristics under wind
circular members and seismic loads are similar in any direction. Circular elements
ϕ Angle formed between a vertical axis and a tangent are also used extensively as columns in buildings, or as piles for
to a circular vertical stirrup at the point where a foundations, or used in the form of secant piling to form diaphragm
shear crack is present walls.
ϕ0 Using a cylindrical coordinate system with an axis Feeling that the extrapolation of tests on rectangular beams
containing the longitudinal axis of the specimen ϕ0 , may not be accurate for evaluating circular member shear strength,
is the azimuth of the stirrup position at a height z0 some researchers performed different test campaigns in such
ϕf Using a cylindrical coordinate system with an axis structural elements.
containing the longitudinal axis of the specimen ϕf , Previous to the description of their own work, Clarke and
is the azimuth of the stirrup position at the level of Bijandi [19] reported test results from Capon and de Cossia (1965),
the centroid of the compressive forces Khalifa and Collins (1981) and Nagato.
λ Efficiency factor, which takes into account the According to [19], Capon and de Cossia tested 21 members,
inclination of the spiral regarding the longitudinal mostly 250 mm in diameter. Most had only longitudinal steel, but
axis four had stirrups. Eleven of the specimens are reported as failing in
τ Shear stresses shear, two with stirrups and nine without.
Φ Diameter of reinforcement Also, according to [19], Khalifa and Collins tested five columns,
χ Efficiency factor considering the average inclination all 445 mm in diameter. These were subjected to an axial load of
of the stirrups crossing a given crack 1000 kN, equivalent to a stress of 6.4 MPa. All specimens failed in
shear.
Nagato’s tests are also described in [19]. Test programmes
shown in Eq. (1). involved 16 columns, all with a diameter of 300 mm, with and
X without transverse reinforcement, and with and without axial
VsRd = z · sin α · (cot α + cot θ ) · Aα · fyα,d (1) stress. All specimens are reported as failing in shear.
 J. Turmo et al. / Engineering Structures 31 (2009) 455–465
Ang, Priestley and Paulay [17] tested 25 circular columns under
axial load and cyclic lateral inelastic displacements, 24 of them
failing in shear. The columns had a diameter of 400 mm. Variables
in the test programme included axial load level, longitudinal
reinforcement ratio, transverse reinforcement ratio and aspect
ratio. Transverse spiral reinforcement was used with different
pitch p over diameter D ratio ranging from p/D = 0.075 to p/D =
0.550. They addressed, for the first time, the efficiency of circular
loops and deduced the formula for calculating VsRd presuming a 45-
deg diagonal tension crack and vertical reinforcement, resulting in
the expression
π variable along the length of the column. Furthermore, these
VsRd = D′ · A90 · fy90,d · . (2a)
4
D′ being the diameter of the hoop, this formula is similar
to Eq. (1) using an effective height of the stirrups equal to D′
instead of z, and using a factor of the effectiveness of the stirrups
equal to 0.785. They extended their formulation directly to spiral
bars, without any modification accounting for the longitudinal
inclination of helical reinforcement.
The work of Ang et al. was completed subsequently by Priestley,
Verma and Xiao [20]; and by Priestley, Seible and Calvi [21].
They considered the potential development of steeper angles of
cracking other than 45-deg proposing the following expression for
evaluating VsRd
π circular columns, it can also be found in large bridge columns and
VsRd = D′ · A90 · fy90,d · · cot θ. (2b)
4 piles, as well as offshore platforms. The economical convenience
The angle of the critical inclined flexure shear cracking to the
column axis, according to these authors, can be taken as θ =
30-deg, unless limited to larger angles by the potential corner
to corner crack. These authors also addressed the influence of
axial load and flexural ductility on the shear strength of circular
columns.
Clarke and Bijandi [19] tested both sides of 50 specimens,
with diameters ranging from 152 mm to 500 mm, reporting
79 shear failures. In their research not only diameter, but
concrete strength, longitudinal and transverse steel content, and
compressive longitudinal stresses varied. Their effort was put on
defining appropriate methods to evaluate VcRd for circular sections.
Unaware of the previous work of Ang, Priestley and Paulay,
they directly used the formula for calculating VsRd in rectangular
sections proposed in BS 8110 [2] for predicting the shear strength
of their tests. They addressed, though, the effectiveness of the
shear reinforcement made of helical transverse reinforcement.
According to them, a spiral may be taken as a special form of
inclined link. That is to say, one leg on one side of the unit crosses a
potential shear crack, while the leg on the other side runs broadly
parallel with the potential crack. The two legs will be at angles α
and (180 − α), respectively, and hence the total efficiency of the
spiral, compared with two legs of a vertical link, simplifies to sin α .
Depending on the ratio between the pitch p and the diameter D′ ,
the value of sin α can be deduced from Eq. (3):
 p i
h
sin α = sin ar cos . (3)
4 · R′
In the derivation of the truss mechanism equation of Ang
et al. [17], it was assumed that a diagonal crack is able to mobilize
transverse reinforcement along a crack length extending the full
height of the unit. In [22] Kowalsky and Priestley revised Eq. (2b),
suggesting that in the compression zone of the elements any
crack is closed, preventing any shear being transferred across it
by tension strain in the transverse reinforcement. They proposed a
slightly more complex formulation:
p A practical efficiency factor of the circular and spiral shear
R′
R′ 2 − x2
Z


VsRd = A90 · fy90,d · · dx · cot θ (4)
x=−(R−xn ) R′
457
where xn is the depth of the compression zone, R is the radius of the
column, R′ is the radius of the circular transverse reinforcement.
They immediately simplified this into a simpler formula:
π


VsRd = R + R′ − xn · cot θ · A90 · fy90,d · (5)
4
which is formally equal to Eq. (2b), substituting the total height
of the stirrups by the effective height of the stirrup present in the
cracked depth of the section. Both Eqs. (4) and (5) are dependent on
xn , and, hence, on the actual bending moment acting on the cross
section, which makes the contribution of the shear reinforcement
authors link the effectiveness of shear reinforcement with flexural
cracking, and not with shear cracking. They assumed that VsRd is
equal to zero in a fully compressed section. Application of their
theory would imply that shear reinforcement of simply supported
pre-stressed I beams would not be effective close to the supports,
whereas tests show how inclined cracking appears in the webs in
absence of flexural cracking and how shear reinforcement is able
to carry loads [23].
Regarding the design of hollow core circular concrete members,
it must be said that none of the codes addressed it. Certainly much
less structurally used than circular cross sections, it is functionally
required in concrete chimneys, concrete pipes or elevated water
tanks. As some savings can be obtained when using hollow core
in the use of hollow columns is due to the cost saving afforded
by the reduced section area (up to 70%). Furthermore, hollow
columns are more efficient than solid ones from a structural point
of view. When the weight of the vertical members is relevant in the
performance of the entire structure, a significant reduction in the
seismic mass may be attained by using this structural type.
Very few tests on such elements can be found in literature.
Whitaker et al. (according to [24]) and Zahn et al. [25] investigated
the ductility capacity of slender circular members under flexure.
Some other flexural tests are reported in [24] that were carried out
in Japan.
Three more tests were carried out by Ranzo and Priestley [24],
two of which failed in shear. Units were tested with a cantilever
scheme, under axial force and pseudo-static cyclic lateral loading.
Specimens failing in shear had a diameter D of 1524 mm, a
height of 3581 mm and a thickness t of 139 mm. Transverse
reinforcement consisted of a continuous spiral of pitch p of 70 mm.
Ratio p/D was equal to 0.05. They calculated the shear contribution
of the spiral reinforcement directly, using Eq. (5), for solid circular
members; not considering any effect on VsRd due either to the
slight inclination of the shear reinforcement or to the existence of
a hollow web.
Bibliographic research shows how a scientific solution for the
contribution of spiral reinforcement on VsRd is lacking. This is
similar for the contribution of transverse reinforcement on hollow
core circular elements. Hence, an analytical model for evaluating
the contribution of the transverse reinforcement is presented in
this paper. An accurate formula for evaluating the shear transferred
by spiral reinforcement in solid members is presented for the
first time. An accurate formula for the calculation of hollow core
circular columns with both vertical and spiral reinforcement is
also deduced for the first time, leading to innovative results.
Assumptions made for the calculation of this formulation have
been deduced theoretically, and then checked empirically. Hence,
shear tests on circular hollow core specimens are also presented.
reinforcement for solid and hollow core circular concrete members
to be used in the design is also presented.
458 J. Turmo et al. / Engineering Structures 31 (2009) 455–465
2. Shear stresses on cracked and uncracked circular and
annular cross-sections
For a given solid cross-section of a beam, as is stated in any
complete book of strength of materials [26], the theoretical value of
the vertical shear flow q at a horizontal layer located at a distance y
from the centre of mass can be derived from the shear flow formula
Sz (y) · Vy
qxy (y) =
Iz
where q is the vertical shear flow, Sz is the first moment of the area
above the horizontal layer considered (or statical moment of area
regarding the z axis crossing the centroid), Iz is the second moment
of area of the whole cross section around the z axis crossing the
centroid and Vy is the vertical shear force.
Vertical shear stresses τxy for solid sections at a given height
can be evaluated from Eq. (6), assuming those constant through
the width b of the horizontal layer where stresses are evaluated,
giving the following approximation
Sz (y) · Vy
τxy (y) = . (7)
Iz · b(y)
For an uncracked solid circular section, we can deduce from
this formula that maximum vertical shear stresses would appear
at the level of centre of masses. An estimation of its maximum
value, according to Eq. (7), is 4 · Vy /(3 · π · R2 ). A sketch of such
shear stress distribution can be found in Fig. 2a. If flexural cracking
occurs in the section, Eq. (7) will not be valid any longer and vertical
shear stress distribution will be modified. A sketch of such shear
stress distribution can be found in Fig. 2b, for concrete and steel
behaving as linear materials. This shear stress distribution will
mainly induce a crack pattern, where the crack will be contained in
a plane forming an angle with the horizontal axis that will depend
on the level of compressive stresses. This crack pattern is the one
found in the experimental tests [16,18] and shown graphically in
Fig. 3a.
For determining shear stresses on closed thin-walled members,
some assumptions should also be made. Therefore, shear stresses
are assumed to be uniformly distributed across the thickness t
of the wall and to be acting parallel with the boundaries of the
section. The problem of finding the real shear stress distribution
being an indeterminate one, Eq. (6) should be complemented
with a compatibility equation or with an equation obtained by
minimising the strain energy due to shear stresses [27]. For
symmetric sections, the condition of antisymmetry of the shear
stresses can also be used to find the solution.
For an uncracked hollow core circular section, shear stresses can
be estimated by Eq. (8)
Vy · sin ϕ
τs (s) =
π ·R·t
ϕ being explained graphically in Fig. 2c. Maximum vertical shear
stresses would appear at the level of the centre of masses, with
an estimated value of Vy /(PI · R · t ). A sketch of such shear stress
distribution can be found in Fig. 2c. If flexural cracking occurs in
the section, vertical shear stress distribution will be modified. A
sketch of such shear stress distribution can be found in Fig. 2d.
This shear stress distribution will mainly induce a shear crack
pattern where the crack will not be contained in a plane, but will
form a helix of constant pitch in the space. The crack will form a
constant angle with the horizontal axis that will depend on the
level of compressive stresses at every location. The crack will form
a straight line in the surface of the unfolded cylinder, forming a
constant angle with the horizontal axis. This angle will be roughly
45-deg for specimens with no axial loads. This crack pattern is
showed graphically in Fig. 3b.
3. Formulation for solid circular members
3.1. Circular stirrups
A formulation for evaluating VsRd in solid circular members
provided with circular stirrups was deduced by Ang et al. [17] and
transformed by Kowalsky et al. [22] in Eq. (4). This formulation
can be regarded as the product of the traditional formula for
(6) vertical stirrups Eq. (1), times an efficiency factor χ , considering
the average inclination of the stirrups crossing a given crack, as
shown in Eq. (9).
VsRd = z · cot θ · A90 · fy90,d · χ. (9)
This efficiency factor can be rewritten. Ang used an efficiency
factor of π /4 = 0.785, assuming that the stirrups were fully
effective through their whole height. Kowalsky defined a different
efficiency factor, assuming that the stirrups were only effective on
the cracked depth of the concrete. Eq. (9) provides an efficiency
factor for stirrups being effective on a height z, starting their
effectiveness at a distance z0 from the centre of masses of the
cylinder. This formula Eq. (10), which can be easily deduced from
Fig. 4, can be applied directly to the calculation of VsRd according to
most of the codes [1,3,4,16], and leads to a more general expression
than Eq. (4).
p
z ·cot θ
R′ 2 − (z0 − x · tan θ )2
Z
χ= dx. (10)
0 R′ · z · cot θ
Efficiency factor χ only depends on the ratios z /R′ and z0 /R′ ,
and not on the value of angle θ . This can be proven by applying to
Eq. (10) the change of variable proposed in Eq. (11)
x = z · cot θ · X . (11)
Resulting in Eq. (12)
s
1
 2
z0 − z · X
Z
χ= 1− dX . (12)
0 R′
Eq. (12) can be solved numerically, giving different values to the
ratios z /D′ and z0 /D′ . Values of efficiency factor χ are presented in
Table 1.
Spurious results have been removed from this table. Mathe-
matically, it is not consistent to have ratios z0 /D′ smaller than
(z − 0.5 · D′ )/D′ . Nevertheless, some values which are hard to get
in practice (e.g. z0 /D′ = 1.0, z0 /D′ = 0.5), have been included for
the sake of convenience.
The value of χ is equal to π /4 = 0.785 Eq. (2), if the height
of the stirrups is fully effective, (this is to say, for a value of z /D′
equal to 1 and a value of z0 /D′ equal to 0.5). Eq. (12), though,
lacks simplicity and it depends on values of z or z0 , which are not
(8)
constant throughout the length of the column. Hence, even though
the solution is presented and tabulated here, and it can be used for
academic purposes or for very specific design problems, a design
oriented formula would be more suitable.
A simpler formula for the efficiency factor can be obtained,
making some design assumptions. Lever arm z can be assumed
constant. Its value can be evaluated as 0.8 · D, along with the
recommendations for design described in [1,18]. A minimum value
of D′ for a circular column can be taken as 0.9 · D. The resultant
force of the compressive stresses can be assumed to be at the same
distance from the centre of the cylinder as the resultant force of
the tensile stresses is from the same point. These assumptions
lead to values of z /D′ of 0.9 D′ and of z0 /D′ of 0.45. The value of
the efficiency factor in such conditions is 0.85. This leads to the
following equation
VsRd = z · cot θ · Aα · fyα,d · 0.85. (13)
 J. Turmo et al. / Engineering Structures 31 (2009) 455–465 459

Fig. 2. Shear stresses in (a) uncracked solid circular cross-section, (b) cracked solid circular cross-section, (c) uncracked hollow core circular cross-section and (d) cracked
hollow core circular cross-section with approximation-values for maximum shear stresses.

This simplification can be adopted for design purposes, and it
has the great advantage of making the contribution of the steel
constant throughout the whole length of the column. From an
engineering point of view, this approach would only be inaccurate
and unconventional for columns with low z /D′ ratios and high
z0 /D′ ratios. In this case, Table 1 or Eq. (12) can be used for the
design, even though is rather difficult to get those values in real
structures.
3.2. Spiral shear reinforcement
In previous investigations, efficiency of spiral reinforcement has
been addressed in a simplified manner [19]. For calculating the
contribution of the spiral reinforcement, this should be considered
as describing a curve within the space. A solution for the evaluation
of VsRd can be easily deduced from Fig. 5, only accounting for
the projection of the force developed in the reinforcement on the
vertical axis. In this way, the following formula can be deduced
460 J. Turmo et al. / Engineering Structures 31 (2009) 455–465

Table 1
Value of efficiency factor χ depending on ratios z /D′ and z0 /D′
z /D′
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

z0 /D′ 0.05 0.998 0.988 0.964 0.922 0.856 – – – – –

0.10 0.993 0.993 0.980 0.951 0.902 0.820 – – – –
0.15 0.978 0.988 0.985 0.967 0.933 0.876 – – – –
0.20 0.952 0.973 0.980 0.973 0.951 0.911 0.839 – – –
0.25 0.914 0.946 0.964 0.967 0.957 0.930 0.882 – – –
0.30 0.863 0.908 0.936 0.951 0.951 0.936 0.904 0.842 – –
0.35 0.797 0.856 0.896 0.922 0.933 0.930 0.911 0.871 – –
0.40 0.709 0.786 0.842 0.880 0.902 0.911 0.904 0.880 0.827 –
0.45 0.592 0.694 0.768 0.820 0.856 0.876 0.882 0.871 0.840 –
0.50 0.409 0.559 0.661 0.733 0.785 0.820 0.839 0.842 0.827 0.785

Eq. (15)
2·π·R′
p
z ·cot θ
R′ 2 − (z0 − x · tan θ)2
Z
p
κ= · dx · r 2 . (15)
0 R′ · z · cot θ· 
2·π·R′
1− p

Comparing Eq. (15) with Eq. (10), factor κ can be regarded as

the product of two different factors, χ and λ, as is stated in Eq. (16)
κ = χ · λ. (16)
Efficiency factor χ takes into account the inclination of the
spiral regarding the transverse axis. Thus, χ depends on the ratios
z /R′ and z0 /R′ . λ takes into account the inclination of the spiral
regarding the longitudinal axis. Hence, λ depends on the ratio p/D′ ,
between the pitch p and the diameter of the spiral 2 · R′ , as is shown
in Eq. (17)

1
Fig. 3. Shear crack pattern in (a) solid circular cross-section and (b) hollow core λ= q . (17)
circular cross-section. p

2
2·π·R′
+1
Aφ
VsRd = z · cot θ · · fyt ,d · κ (14) For a nil value of pitch p, the value of efficiency factor λ is one
p /2 and the value of efficiency factor κ is equal to the value of efficiency
where AΦ is the area of the single rod of diameter Φ and κ is factor χ , given by Eq. (10).
the efficiency factor accounting for the tri-dimensionality of the Again, the general expression for the efficiency factor κ lacks
spiral. Efficiency factor κ is directly deduced from Fig. 5, leading to simplicity, as it depends on the value of efficiency factor χ , and

Fig. 4. Efficiency of circular stirrups on solid circular members. Three stirrups crossing the crack. Individual efficiency factor for each stirrup i = 1, 2, 3; sin ϕ1 , sin ϕ2, sin ϕ3 .

Fig. 5. Efficiency of spiral stirrups on solid circular members.

 J. Turmo et al. / Engineering Structures 31 (2009) 455–465 461

Table 2 level of compressive stresses. This angle would be 45 deg for an

Value of sin α for the spiral reinforcement according to Clarke et al. [18] and value uncompressed member.
of efficiency factor λ according to Eq. (17)
VsRd can be calculated from Fig. 7, as the product of the number
p/D′ Efficiency of stirrups crossing the crack, times the mechanical capacity of
Eq. (3) Eq. (17) the stirrup, times the average inclination of the stirrups at the
sin α λ intersection with the crack, as is stated in Eq. (19).


0.0 1.000 1.000 ϕ0 − ϕf · R · cot θ
0.1 0.999 0.999 VsRd = · 2 · Aφ · fy90,d
0.2 0.995 0.998 st
0.3 0.989 0.995 R (ϕ0 −ϕf )·R·cot θ
sin ϕ0 − x·tan θ


0.4 0.980 0.992 0 R
· dx
0.5 0.968 0.988 ×
. (19)
0.6 0.954 0.982 ϕ0 − ϕf · R · cot θ
0.7 0.937 0.976 This equation can be easily simplified into Eq. (20)
0.8 0.917 0.969
0.9 0.893 0.961 VsRd = z · cot θ · A90 · fy90,d . (20)
1.0 0.866 0.953 This apparently surprising result implies that the general
formula for evaluating VsRd in concrete members Eq. (1) can be
directly applied to the calculation of the steel contribution on
hence on ratios z /R′ and z0 /R′ , which are not constant throughout
annular members with closed circular stirrups. This contrasts
the length of the column; and on the value of ratio p/R′ , which
with the results obtained for solid circular members. For these,
is generally constant. Hence, a design oriented formula would be
efficiency factor χ is always less than one (Table 1), whereas
more suitable. Efficiency factor χ can be taken as 0.85, making the annular sections can be regarded as having an efficiency factor χ
same assumptions for evaluating as the ones made in Section 3.1. of one. That is to say, that transverse reinforcement is less effective
The value of efficiency factor λ can now be calculated depending on in members with more robust sections, as in solid circular cross-
ratio p/D′ (Table 2). This value can be compared with the values of sections, than in tubular members, even though it has the same
the simplified approach proposed in [19]. Values proposed in [19] shape in both.
can be obtained from Eq. (3). All these values are shown in Table 2. This astonishing affirmation can be easily explained with
Values of efficiency factor have been calculated for a maximum the examination of Fig. 2. According to this, shear stresses in
ratio p/D′ of 1. This will give an average fictitious angle α of 45-deg solid circular members are mainly vertical; whereas in annular
according to Clarke. This can be regarded as the maximum limit members, shear stresses have the same orientation as the stirrups
never to be surpassed in European practice [16]. As the pitch should have. Hence, in annular members circular stirrups are more
not exceed 3 in. (75.4 mm) according to [1], this limit could be even effective as their geometry follows the orientation of the shear
smaller. stresses provided by the Theory of Elasticity.
It can be seen that values proposed by Clarke and Birjandi
are slightly conservative for high values of ratio p/D′ , and quite 4.2. Spiral shear reinforcement
accurate for lower values. Regarding the solution presented here,
A formulation for evaluating VsRd in hollow core circular mem-
it is to say that the difference between the circular hoops and the
bers provided with spiral stirrups is presented here. Deduction of
spiral ones is irrelevant for low values of ratio p/D′ (e.g. smaller
this formula relies again on the theoretically proven and exper-
than 0.5), as this difference is small for higher values. As practical
imentally checked hypothesis that the cracking pattern in such
values of ratio p/D′ are usually small, the following simplified structures is a helix of constant pitch, as is shown in Fig. 8.
expression is proposed for design purposes. In order to deduce the formula for VsRd , it has proven successful
to study the contribution of the reinforcement on the geometry of
Aφ
VsRd = z · cot θ · · fyt ,d · 0.85. (18) the unfolded cylinder and use cylinder coordinates, as is presented
p/2 in Fig. 9. Here, the crack is represented as a straight line and the
This expression is formally equal to Eq. (14), assuming an stirrups are represented as a zig-zag line, forming an angle α and
efficiency coefficient κ constant and equal to 0.85; that is to say, 180 −α with the horizontal axis. Angle α can be defined by Eq. (21)
efficiency factor χ equal to 0.85 and efficiency factor λ equal to p
cot α = (21)
1.00. Indeed, there is no practical difference between the circular 2 · π · R′
stirrups and spiral reinforcement for usual values of p/D′ ratio. VsRd can be calculated from Fig. 9, as the product of the number
of stirrups crossing the crack, times the mechanical capacity of
the stirrup, times the average inclination of the stirrups at the
4. Formulation for hollow core circular members
intersection with the crack, as stated in Eq. (22).

R ϕf ′
4.1. Circular stirrups ϕ0 − ϕf · R′ · cot θ ϕ0 −R sin ϕ · dϕ
VsRd = · Aφ · fyt ,d ·
. (22)
p/2 ϕ0 − ϕf · R′ · cot α
A formulation for evaluating VsRd in hollow core circular
members provided with circular stirrups has not been previously It must be highlighted that, as half of the legs of the spiral
reinforcement form an angle α with the horizontal axis and the
deduced, and it is presented below. Deduction of this formula relies
rest form an angle π − α , the average number of legs crossing the
on the hypothesis that the cracking pattern in such structures
crack is related with the distance (π − ϕ0 − ϕf )R′ cot θ and not
is a double helix of constant pitch, as is shown in Fig. 6. This
with the distance (π − ϕ0 − ϕf ) · R′ · (cot θ + cot α). Eq. (22) can
assumption has been proven theoretically above and checked be transformed and simplified into Eq. (23)
experimentally.
In order to deduce the formula for VsRd , it has been proven Aφ
VsRd = · fyt ,d · z · cot θ · λ (23)
successful to study the contribution of the reinforcement on the p/2
geometry of the unfolded cylinder and use cylinder coordinates, where efficiency factor λ has already been defined in Eq. (17). In
as presented in Fig. 7. Here, crack is represented as a straight line, the same way as was done in Section 3.2, as practical values of
forming an angle θ with the horizontal axis that depends on the ratio p/D′ are usually small, efficiency factor λ can be taken as one
462 J. Turmo et al. / Engineering Structures 31 (2009) 455–465

Fig. 6. Lateral view of shear cracking of hollow core circular members with circular stirrups.

Fig. 7. Sketch of shear cracking of hollow core circular members with circular stirrups on the unfolded cylinder geometry.

Fig. 8. Lateral view of shear cracking of hollow core circular members with spiral reinforcement.

Fig. 9. Sketch of shear cracking of hollow core circular members with spiral reinforcement on the unfolded cylinder geometry.

for design purposes. Hence this formula can be transformed into 5. Experimental results
Eq. (24)
Results presented in Section 4 are based on the hypothesis that
Aφ shear cracking of concrete in hollow core circular members forms
VsRd = · fyt ,d · z · cot θ. (24)
p/2 a helix of constant pitch. This fact has been theoretically proven to
be correct in Section 2. This fact was also checked experimentally
Again, comparing this formula with Eq. (18), it can be seen that within a test programme carried out at the Structures Laboratory of
spiral reinforcement is more effective in members with annular Technical University of Catalonia. Objectives and a full description
cross-section than in members with a solid circular one. of this test programme surpass the aim of this paper. Hence, only
 J. Turmo et al. / Engineering Structures 31 (2009) 455–465 463

Table 3
Concrete composition and compressive strength
TESTS 1–2 TESTS 3–4

Ciment CEM I/42.5 R (kg/m3 ) 397 397

Sand 0/2 (kg/m3 ) 588 588
Gravel 2/5 (kg/m3 ) 408 408
Aggregate 5/12 (kg/m3 ) 683 683
Water (l/m3 ) 238 238
GleniumC303SCC 0.4% (l/m3 ) 1.59
Rheomac 890F 0.5% (l/m3 ) 1.99
w/c ratio 0.6 0.6
fcm,28 (MPa) 31.93 26.10
Fig. 11. Test setup.
fcm,test (MPa) 32.15 24.72

Table 4
Ultimate loads obtained in the tests
Qult (kN)

TEST 1 466
TEST 2 478
TEST 3 473
TEST 4 433

Fig. 10. General view of the reinforcement, internal formwork and spacers for the
test specimens.

the relevant aspects regarding the scope of this paper will be

described. More detailed information can be found in [28,29].
Shear tests involved a total of four hollow core circular
specimens. All of them were geometrically identical, with a
length of 3000 mm and a span of 2800 mm. External diameter
was 600 mm. Hollow core was achieved introducing a 400 mm
diameter cylinder made of expanded polystyrene. Diaphragms Fig. 12. Cracking at failure on test 1.
of 250 mm width were provided at the bearings and a single
diaphragm of 100 mm width was provided at the load point. The
tests were divided into two series. The first series involved two
specimens made of ordinary concrete (test 1 and test 2). The second
one was made with self-compacting concrete (test 3 and test 4).
Some superplasticizer and viscosity modifier agents were added to
the concrete mix. Concrete composition and mechanical properties
for both concrete types are presented in Table 3.
Reinforcement was identical for the four specimens. Nine circu-
lar stirrups made of 6 mm nominal diameter bars were placed with
300 mm between them. Twelve 20 mm nominal diameter bars
were placed as longitudinal reinforcement, uniformly distributed
along the circumference of the stirrup. Hot-rolled deformed steel
bars of B500S quality were used as passive reinforcement. This
steel grade has a minimum yield strength of 500 MPa, a minimum
tensile strength of 550 MPa and a minimum elongation of 12% for
a five diameter gauge length. Cut and bent bars were supplied, and
afterwards, placed in the forms. Bars and internal formwork were Fig. 13. Measured and estimated crack geometry on the unfolded cylinder.
kept in place with plastic spacers. A general view of the reinforce-
ment, internal formwork and spacers is presented in Fig. 10. cracking pattern implies the formation of one three-dimensional
A picture of the test setup is presented in Fig. 11. Tests were strut carrying the external load from the actuator to the bearings.
performed by applying a punctual load at mid-span with an Failure loads for all the tests are summarized in Table 4.
actuator. Specimens were simply supported on mechanical hinges. Crack geometry was carefully measured at the end of every
Special steel pieces with a curved surface were designed to get a test. Points of Fig. 13 represent the position of the different cracks
more uniform distribution of the bearing and load stresses. These on the unfolded cylinder. Only half of the length of the unfolded
were interposed between the bearings and the test specimens, and cylinder is presented in this figure. Data has been processed with
between these and the actuator. With the same purpose, neoprene- statistical methods [30]. Linear, quadratic and cubic regression has
teflon sheets were also fitted at the same locations. been applied to the data in order to identify the mathematical
The behaviour of all the tests was very similar; all of them expression that better approximates the geometry of the crack.
failing, with a huge crack developing in one of the sides of the Results are presented in Table 5. Determination coefficient R2 ,
specimen after the shear reinforcement yielded. The crack shape which can be seen as the part of the dependent variable that
was a helix of constant pitch, as can be observed in Fig. 12. This can be explained by the independent variables, is already close
464 J. Turmo et al. / Engineering Structures 31 (2009) 455–465
Table 5
Regression of the crack geometry (dimensions of x in mm)
Coefficients
x0 x1
Linear 448.9 −0.602
Quadratic 445.0 −0.582
Cubic 447.9 −0.618
to one for the linear regression. Its value increases only slightly
in the quadratic and cubic regressions. Expected values for
the crack geometry, according to the different approaches, are
also presented in Fig. 13. Here, the small differences on the
expected crack geometry that the presence of higher terms in the
regression implies, can be seen graphically. Actually, application of
hypotheses test methodology to these results cannot reject the null
hypothesis that the linear and quadratic regressions are the same.
This is also true for linear and cubic regressions. Moreover, null
hypothesis that linear regression is not significant can be rejected
with a level of significance higher than 0.05. Hence, crack geometry
for hollow core circular piles can be described as a straight line on
the unfolded cylinder or as helix of constant pitch on the actual
cylinder.
6. Conclusions
Shear stress distributions on solid and hollow core circular
cross-sections are very different. Whereas in the first ones, shear
stresses are mainly vertical; in the second ones, shear stresses are
assumed to act parallel with the boundaries of the section. In the
former, shear crack pattern will be contained in a plane. In the
latter, the crack pattern will not be contained in a plane but will
form a helix of constant pitch in the space, and will form a straight
line in the surface of the unfolded cylinder. This statement has been
proven theoretically and is experimentally correct.
The fact that shear stresses have a different orientation in both
section types, linked to the fact that the shear reinforcement does
have the same geometry for both structural types, leads to the
evidence that the stirrups or spiral reinforcement are not equally
effective in solid and hollow core circular sections. Transverse
reinforcement will be more efficient in those section types whose
shear stresses better align with the shear reinforcement. Circular
or spiral stirrups will be more efficient in hollow core circular
specimens than in solid ones.
A theoretical formulation following the Eurocode 2 approach
is presented, to evaluate the transverse steel contribution on the
shear strength of hollow core and solid circular concrete members,
reinforced with circular and spiral transverse reinforcement.
VsRd = z · cot θ · Ast · fyt ,d · χ · λ
where z is the lever arm; θ is the angle between the inclined strut
and the longitudinal axis of the member; Ast is the area per unit
length of the transverse reinforcement defined as 2 · AΦ /st for
circular stirrups and as 2 · AΦ /p for spiral reinforcement; fyt ,d is the
design strength of the reinforcement; efficiency factor χ ’s value is
defined by Eq. (12) for solid circular members and is one for hollow
core members; efficiency factor λ’s value is one for members with
circular reinforcement and is defined by Eq. (17) for members with
spiral reinforcement.
This accurate formulation can be easily adapted to other
approaches used for evaluating VsRd which are different to the one
proposed by Eurocode 2. Hypotheses, such as that which states that
the stirrups are only effective on the flexurally cracked depth of the
member [19], or as that which assumes that the full height of the
stirrups is able to transfer shear [2], can be used to easily transform
the formulation accordingly.
x2 x3 R2
– – 0.9293
−1.506 × 10−5 – 0.9294
5.522 × 10−5 −3.403 × 10−8 0.9295
The lower the ratio p/D′ , the higher the efficiency of spiral
reinforcement. Small reinforcement diameters with low p/D′
ratios would be preferable to bigger reinforcement diameters with
higher p/D′ ratios.
Complexity of the formulation proposed in Eqs. (12) and (17)
can be circumvent for design purposes, substituting the product
of efficiency factors χ · λ of Eq. (25) by 0.85 when designing solid
circular elements, or by 1.00 when designing hollow core circular
members. These values will safely cover most design situations.
Full formulation can be used for research purposes or when high
accuracy is sought.
Acknowledgements
This work was carried out at the Laboratory of Struc-
tures Technology of the Technical University of Catalonia in
Barcelona, with financial support from the Spanish Ministry of
Science and Technology through the research projects BIA2005-
09229-C02-01/02, BIA2006-15471-C02-01/02, PSE 11-2005/PSS-
380000-2005-12/2007-1 and MAT2002-00849 (Experimental and
Theoretical Study of the Shear Capacity of Structural Reinforced Con-
crete Members and Verification of Formulae Provided by National and
European Codes).
The authors wish to thank Civil Engineer Gabriel Ruiz de
Gordejuela for his great contribution to the progress of the research
during the experimental phase. They would also like to thank the
staff of the Structures Laboratory for their enthusiastic help with
the project and particularly the head of this laboratory, Mr. Tomàs
Garcia.
References
[1] American Concrete Institute Committee 318 Building Code Requirements for
Structural Concrete (ACI 318-05). 2005.
[2] British Standards Institution. BS 8 100 Structural Use of Concrete: Part I: Code
of Practice for Design and Construction, London, 1985.
[3] Comisión Permanente del Hormigón. EHE Instrucción del Hormigón Estruc-
tural. Ministerio de Fomento. Madrid, 1998.
[4] CEB-FIP Model Code 1990. London: Thomas Telford Ltd; 1993.
[5] Oreta AWC. Simulating size effect on shear strength of RC beams without
stirrups using neural networks. Eng Struct 2004;26(5):681–91.
[6] Mansour MY, Dicleli M, Lee JY, et al. Predicting the shear strength of reinforced
concrete beams using artificial neural networks. Eng Struct 2004;26(6):
(25) 781–99.
[7] Cladera A, Mari AR. Shear design procedure for reinforced normal and high-
strength concrete beams using artificial neural networks. Part I: Beams
without stirrups. Eng Struct 2004;26(7):917–26.
[8] Cladera A, Mari AR. Shear design procedure for reinforced normal and high-
strength concrete beams using artificial neural networks. Part II: Beams with
stirrups. Eng Struct 2004;26(7):927–36.
[9] Etxeberria M, Mari AR, Vazquez E. Recycled aggregate concrete as structural
material. Mater Struct 2007;40(5):529–41.
[10] Barragán BE, Gettu R, Zerbino R, Agulló L. On the shear failure of steel fiber
reinforced concrete based on push-off tests. ACI Mater J 2006;103:251–7.
[11] Cladera A, Mari AR. Experimental study on high-strength concrete beams
failing in shear. Eng Struct 2005;27(10):1519–27.
[12] Shahawy MA, Cai CS. A new approach to shear design of pre-stressed concrete
members. PCI J 1999;44(4):92–117.
[13] Tan KH, Cheng GH. Size effect on shear strength of deep beams: Investigating
with strut-and-tie model. J Struct Eng-ASCE 2006;132(5):673–85.
[14] Zhang N, Tan KH. Direct strut-and-tie model for single span
and continuous deep beams. Eng Struct 2007;29(11):2987–3001.
doi:10.1016/j.engstruct.2007.02.004.
[15] Bentz EC, Vecchio FJ, Collins MR. Simplified modified compression field theory
for calculating shear strength of reinforced concrete elements. ACI Struct J
2006;103(4):614–24.
 J. Turmo et al. / Engineering Structures 31 (2009) 455–465
[16] European Committee for Standardization. UNE-ENV 1992-1-1. Eurocode 2.
Design of Concrete Structures. Part 1-1. General Rules and Rules for Buildings.
Brussels, 2004.
[17] Ang BG, Priestley MJN, Paulay T. Seismic shear strength of circular concrete
columns. ACI Struct J 1989;86(1):45–59.
[18] Comisión Permanente del Hormigón. EHE Instrucción del Hormigón Estruc-
tural. Draft. Ministerio de Fomento. Madrid. 2007. http://www.fomento.
es/MFOM/LANG_CASTELLANO/DIRECCIONES_GENERALES/ORGANOS_
COLEGIADOS/CPH/instrucciones/norma_ehe/.
[19] Clarke JL, Birjandi FK. The behaviour of reinforced concrete circular sections in
shear. Struct Eng 1993;71(5):73–81.
[20] Priestley MJN, Verma R, Xiao Y. Seismic shear strength of reinforced concrete
columns. ASCE J Struct Eng 1994;120(8):2310–29.
[21] Priestley MJN, Seible F, Calvi M. Seismic design and retrofit of bridges. John
Wiley & Sons; 1996. p. 333–45.
[22] Kowalsky MJ, Priestley MJN. Improved analytical model for shear strength of
circular reinforced concrete columns in seismic regions. ACI Struct J 2000;
97(3):388–96.
[23] Lin TY, Burns NH. Design of prestressed concrete structures. John Wiley and
Sons; 1984.
465
[24] Ranzo G, Priestley MJN. Seismic performance of large RC circular hollow
columns. In: Proceedings of the 12th world conference on earthquake
engineering. 2000 [on CD, Paper No. 0250].
[25] Zahn FA, Park R, Priestley MJN. Flexural strength and ductility of circular
hollow reinforced concrete column without confinement on inside face. ACI
Struct J 1990;87(2):156–66.
[26] Popov PP, Balan TA. Engineering mechanics of solids. NJ (USA): Prentice Hall;
1998.
[27] Canet MJ. Cálculo de Estructuras. Vol 1: Fundamentos y Estudio de Secciones
Edicions UPC, Barcelona, 2000.
[28] Ruiz de Gordejuela G. Shear behaviour of hollow core cylindrical bridge piles.
M.S. thesis. Directed by Ramos G and Turmo J. Spain: Technical University of
Catalonia; December 2004.
[29] Ruiz de Gordejuela G, Ramos G, Turmo J. Experimental study on shear
behaviour of hollow core cylindrical bridge piles. In: III ACHE conference on
bridges and structures. 2005.
[30] Montgomery CD, Runger CG, Hubele NF. Engineering statistics. John Wiley and
Sons; 1998.