Professional Documents
Culture Documents
Title Simplified model for compressive response of RC column footing with square
cross-section
Abstract
In this paper a simplified calculus model for the prediction of the compressive response of RC column footing with
square cross-section is presented. As it is well known RC concrete footing are generally designed adopting uniform
contact pressures on substrate and assuming a strut and tie model in deep members and a cantilever beam or slab
model in flexible members. Deep and flexible members are distinguished in literature only based on the tangent of the
angle expressed as the ratio between the depth and the shear span of the footing. If the angle is higher than 45° the
strut and tie model is adopted; while, if the angle is lower than 45° the beam or the slab models are adopted. In this
paper several subgrade contact pressures distribution for column footings (rigid or soft soils) were considered in
developing a mechanical model able to derive the complete load displacement curves of RC deep and flexible footing
in compression. Aim of the paper is the choice of the best model (strut and tie or cantilever beam model) to predict the
compressive response of single column footing. Numerical results and available experimental results were utilised to
verify the model. Effects of main parameters such as geometry (depth, width) and shape of footing (single or shaped),
mechanical ratio of longitudinal steel and type of soil are investigated both numerically and analytically. The
comparison between analytical and numerical results allows one to validate the proposed model and to define the
range of application of the strut and tie or beam model depending on the footing geometry, mechanical ratio of
longitudinal reinforcements and type of soil.
highlingth.doc [Highlights]
To view all the submission files, including those not included in the PDF, click on the manuscript title on your EVISE
Homepage, then click 'Download zip file'.
Ref: ENGSTRUCT_2016_831_R1
Title: Simplified model for compressive response of RC column footing with square cross-
section
Journal: Engineering Structures
Thank you to the editor and to the reviewers for the time and the effort spent in review the paper.
All suggestions corrections were made and are in yellow during the text.
-Reviewer 1
The article was essentially improved as many errors were eliminated. Some important references
are added too.
1. A comparative analysis of the results obtained using model for the calculation of flexible
foundation was carried out, with some experimental results. However, it is not clearly
indicated which method of calculation was used in terms of characteristics of the ground
(rigid, flexible, and this soil-Winkler).
It was better specified that Strut and tie and cantilever beam model with punching shear
were here considered. In the strut and tie model concrete crushing of compressed struts
and yielding of steel bars are considered; while in the cantilever or the slab models for
flexible footing the shear failure due to punching failure is also considered with flexural
failure.
2. The results were compared with the calculation of punching according to EN 1992 (EC2)
and ACI 318. Figure 11 has been added to this comparative analysis. The results of the
model with shear deviate significantly from the calculated results of the proposed model,
and it should be reviewed, and eventually modified the image 11.
All corrections and suggestions were added and in particular Fig. 11 was redrawn and a
new section referred to punching shear failure was added and examined.
3. It is shown that tie and strut model suits the samples, although they belong to the flexible
foundations, so the mathematical model can be also applied to flexible foundations.
It was made
After correction of numbered suggestions, the article can be accepted for publication.
-Reviewer 3
- I recommend the paper with corrections given in uploaded file.
For column footing, the minimum percentage of reinforcement for crack control,
according to Eurocode 2 [9], is equal to:
Comparison of the proposed beam (slab) model for calculation of flexible foundations is
done with results of Siburg and Hegger [18], and Simões et al. [19], which is correct
regarding that all examined foundations belong to flexible foundations (angle towards
horizontal is lower than 45°). However, comparison of the examined foundations from
literature is done also with expressions for rigid foundations, for whose calculation the strut
and tie model is proposed (angle towards horizontal is higher than 45°), which is not in
agreement with the proposed classification, because all examined foundations were flexible.
Examined foundations are also compared with calculation by EC2 and ACI. All comparisons
are given in Fig. 11.
Obtained results, according to Fig. 11, show that for the examined foundations the best
fitting calculation is the calculation according to the proposed strut and tie model (although
the foundations are flexible), than by EC2 and ACI standard, while the calculation according
to the beam (slab) model is very unsafe. This shows that the proposed procedure for
calculation of flexible foundations does not express good agreement with experimental
results, which is the consequence of the fact that experimental foundations are fractured due
to punching, and not bending. For the experimental foundations, good results are obtained by
the calculation procedure proposed for rigid foundations. Considering that for the calculation
of flexible foundations, besides the calculation of punching, the calculation according to
maximal moments is needed, in Fig. 11. the calculation of flexible foundations according to
maximal moments calculated by EC2 and ACI standard should be added.
Thus the proposed procedure will be compared with EC2 and ACI standards.
All corrections and suggestions were added and in particular Fig. 11 was redrawn and a
new section referred to punching shear failure was added and examined.
I suggest that the diagram in Fig. 11 be modified such that the ordinate contents the
ratio Pu, calc / Pu,exp, while the abscissa may remain with the failure force measured in
experiments (Pu,exp). In that way, the safety factor (Pu, calc / Pu,exp) for the proposed calculation
procedures can be better observed.
It was made
According to the obtained results the manuscript text and the conclusions should be
revised.
Summary
In this paper, a simplified calculus model for the prediction of the compressive response of RC
column footing with a square cross-section is presented. As it is well-known RC concrete footing
are designed adopting uniform contact pressures on the substrate and assuming a strut and tie model
in deep members and a cantilever beam or slab model in flexible members. Deep and flexible
members are distinguished in literature only based on the tangent of the angle expressed as the ratio
between the depth and the shear span of the footing.
In this paper, several subgrade contact pressures distribution for column footings (rigid or soft
soils) were considered in developing a mechanical model able to derive the complete load
displacement curves of RC deep and flexible footing in compression. The objective of the research
was the more appropriate choice of the calculus model to predict the compressive response of single
column footing. In particular, if the angle, expressed as the ratio between the depth and the shear
span of the footing, is higher than 45° the strut and tie model was suggested for best prediction,
while, if the angle is lower than 45°, the beam or the slab models with punching shear were
adopted. For both cases simplified loading soil profiles corresponding to rigid, flexible or winker
model were adopted. Effects of main parameters such as geometry (depth, width) and shape of
footing (single or shaped), the mechanical ratio of longitudinal reinforcement and type of soil are
investigated both numerically and analytically. Numerical results and available experimental results
were utilised to verify the model. The comparison between analytical and numerical results allows
one to validate the proposed model and to define the range of application of the strut and tie or
beam model depending on the footing geometry, the mechanical ratio of longitudinal reinforcement
and type of soil. Finally, the comparison between analytical and experimental results available in
the literature gives a further confirmation on the reliability of the proposed model.
Keywords: Footing; strut and tie model; beam model; elastic soil; Winkler model.
2
1. INTRODUCTION
Foundation connects the structure with the subgrade. Characteristics of subgrade are very different
from those of structure, especially in term of deformability; in their calculation have not been
considered both in the literature and neither in the technical regulations. The study of soil-structure
interaction is essential to determine the soil-foundation pressures necessary for the design of the
foundation in respect of the limit state and to evaluate the settlements of the foundation. Evaluation
of stresses and relative design of foundation needs the determination of subgrade contact pressures,
and its distribution depends on its deformations and reactions transferred from the building. From
the geotechnical point of view, these aspects are clear, and it is clear that the determination of
pressures of the ground floor is a very complex problem concerning soil, foundation and building
with different behaviours, as discussed in Albert [1]. It depends on: - type of soil; - the intensity of
stresses transferred from the foundation; - nonlinear characteristics of the stress-strain curve of soil;
- short and long term phenomena due to concrete and soil (cracking, creep, consolidation);
foundation geometry.
Among the direct foundations is recurring the use of column footing especially in the case of
structural elements such as reinforced concrete columns and soils having high carrying capacity: in
this case, it is possible to have supports in a limited area. In the design of column footings, besides
geotechnical aspects, assumes an important role the limit state of concrete members, with a control
of allowable settlements (Viggiani [2], Toniolo [3] and Migliacci and Mola [4]).
From a structural point of view, the simplified calculation is adopted in the design of foundations,
which are considered as rigid, with a linear pressures distribution on subgrade. In the case of single
column footing, it is necessary to distinguish between deep and flexible member. As posed in the
literature, strut and tie model is generally adopted for deep elements, while cantilever beam model
is generally adopted for the flexible members. Column footing can be considered deep when,
referring to Fig.1, > 45°, whereas can be considered flexible when < 45°. In both cases,
3
important phenomenon determining columns footing failure is the punching shear failure.
Considerable significance is given to the control of column punching through floor slab with or
without shear reinforcement especially based on experimental research. Therefore, due to the small
number of experiments, this analysis is often based on theoretical and empirical equations. Some
interesting models, such as that of Simões et al. [5], have been developed.
Elastic analyses of soil-structure interaction as those of Terzaghi [6] and Heteny [7], based on
continuum models have shown that for deep footing the shape of contact pressures is that of Fig. 2
a) or Fig. 2 b) for rigid or soft soils, respectively; otherwise if footing is flexible, the shape of
contact pressures are those of Fig. 2 c) or Fig. 2 d) for rigid or soft soils, respectively. So contact
P
pressures distribution on the substrate could be different from the constant value p m , that is
B2
usually used.
If the footing is deep and the soil is rigid the expression of contact pressures in the elastic range
4 P dxdy
2
p ( x, y ) (1)
B x 2 y 2
1 1
B / 2 B / 2
From Eq. (1), it results that if x=y=B/2, pressures tend to infinitive and have a minimum value in
the middle (x=y=0). Therefore, Eq. (1) can reproduce the variation of contact pressures shown in
Fig. 2 a).
It has to be stressed that in deep compressed footing the upper part of the foundation is low stressed.
For this reason, the best shape to adopt should be the shaped. By contrast, the prismatic shape is the
According to the theory of Prandtl–Buisman [8], for a rigid foundation in soft soil, the effective
stress distribution is convex (Fig. 3), and the soil reaction underneath the punching cone is greater
Studies present in the literature (Viggiani [2], Toniolo [3] and Migliacci and Mola [4]) show that
distribution of contact pressures indicated in Fig. 2 can be idealized as in Fig. 4 a) for rigid soil or
Fig. 4 b) per soft soil. Reference values, for the definition of loading profiles, are those indicated in
Fig. 4. They are defined through the maximum and the minimum values correlated to the medium
value pm. Near failure conditions of footing, the hypothesis of uniform contact pressure (Winkler
model) is acceptable for safety conditions and the distribution of pressures that can be adopted is
Finally, when column footing failure occurs with soil in the elastic range, the better shape of contact
pressures is that of Fig. 4 a) and b), depending on the type of soil (rigid or soft); otherwise, the best
contact pressures distribution is that of Fig. 4 c). In this case, bending moments are higher than
Despite the complexity of the problem and with the aim of utilizing simplified models for the
preliminary design of column footing the contact pressures distribution are assumed linear or
Strut and tie and cantilever beam model with punching shear were here considered. In the strut and
tie model concrete crushing of compressed struts and yielding of steel bars are considered; while in
the cantilever or the slab models for flexible footing the shear failure due to punching failure is also
As shown below for 45° beam model with punching shear is more suitalbe, while for >45° and
Based on the previously mentioned models the arrangement of longitudinal steel reinforcement is
that of Fig. 5 a) for deep footings and that of Fig. 5 b) for flexible footing with specific punching or
For column footing, the minimum percentage of reinforcement of crack control according to
k c k Act f ct ,eff
As min (2)
s
Where Act is the concrete area in the tensile zone, kc=0.4 for simple bending and 1 for pure traction,
fct.eff the tensile strength of concrete in the presence of cracking (suggested value for Eurocode 2 [9]
is equal to 3 MPa), s< 0.9 fyk and ka between 0.5 and 1. It has to be stressed that this value is
referred, according to the Eurocode 2 [9], to the Serviceability Limit State, the case here not
examined.
If, in Eq. (2), it is assumed fy=fyk, fct,eff=0.1 fc, (with fc the compressive strength of concrete), kc=0.4
Af fy
and k=0.5, the mechanical ratio of reinforcement results 0.022 .
BH fc
Where Af is the area of longitudinal reinforcement equal in x and y direction, fy the yielding
reinforcements equal in x and y direction, are placed on the bottom of the column footing with
adequate cover and bent on the upper side with length calculated as the maximum between 200 mm,
10 times the diameter of the reinforcement and 1/3 of the footing side according to Eurocode 2 [9].
6
The calculus model here adopted is the strut and tie model shown in Fig. 6. This model is
constituted by two single trusses each loaded by P/2. Each truss is constituted by two struts and one
The force T in the tie and the force C in the strut are related to the external load P as follows:
P P
T tan C sin (3)
4 4
z
tan (4)
y
z 0.9 H (5)
is the cover thickness and y is the distance of P/4 from the face of the column.
If the column footing is prismatic and the contact pressures are supposed uniform y=B/4; while for
rigid or soft soil it results y= 1.22 B/4 and y = 0.77 B/4. Analogously, for shaped column footing it
results y= B/3, a y= 1.62 B/3 and y= 1.02 B/3, respectively, for Winker, rigid and soft soil.
In the case of deep footings, to take into account the nonlinear strain distribution within the cross
section, lower values of z has to be assumed. According to Leonhardt [11], its value depends on the
ratio a/d, where a is the shear span and d the effective depth of the beam.
The tensile force in the tie when concrete cracks in tension is calculated as:
f ct
Tc 0.2 f ct B H A f E s (6)
E ct
7
fct being the tensile strength of concrete, Ect the modulus of elasticity in tension and Es the modulus
The value 0.2 H in Eq. (6) defines the depth of tensile zone of concrete around steel members. The
value assumed is in agreement with values suggested in Leonhardt [11] and with results of
f
Pu ,c 4 0.2 f ct B H A f E s ct tan (7)
E ct
Ty A f f y (8)
Pu , y 4 A f f y tan (9)
From equilibrium of internal forces (see Fig. 6), it is possible to derive force Ccc on strut that, at
f 1
C cc 0.2 f ct B H A f E s ct (10)
E ct cos
Af f y
C cy (11)
cos
C cu b 2 f c (12)
8
Where b is the depth of column assumed equal to the depth of the strut, ξ is softening coefficient
that takes into account of the biaxial state of stresses (tensile-compression) on concrete struts. The
softening coefficient was here calculated through the expression proposed by Campione [12] in the
form:
1
2
(13)
a
1 0.66
z
Eq. (13) agrees with expression given in AASHTO [13]. Fig 7 shows the variation of the softening
coefficient calculated with Eq. (13) and with that given in AASHTO [13] expression with the
variation of a
Pu ,c 4 b 2 f c sin (14)
Eq. (14) is valid for B>3b. For B<3b it is necessary to introduce in Eq. (14) a corrective coefficient
B
that considers the limit case b=B for which it must results Pu ,c b 2 f c .
4b
If Eq. (9) and Eq. (14) are solved with respect to P it results:
0.9 H
Pu , y 16 A f f y (15)
Bb
1
Pu ,c 4 b 2 f c sin (16)
Bb
1 0.37
H
Similar expressions to Eq. (15) and Eq. (16) are given in the literature (Viggiani [2], Toniolo [3]
H
Pu , y 16 A f f y Viggiani [2] (17)
Bb
9
H 1
Pu , y 4 Af f y Toniolo [3] (18)
B b 0.2 H 1 b
4 B
1 1
Pu ,c 4 0.4 H b f c 2
2
Toniolo [3] (19)
b
1
H B
B b 0.2 H
4
0.875 H
Pu , y 16 A f f y Migliacci and Mola [4] (20)
Bb
0.4 H 2 0.50 H
Pu ,c 16 b fc Migliacci and Mola [4] (21)
Bb b
Comparison among expressions given by Eqn. (15, 17, 18, 20) gives results that differs for
between 45° and 60° in the range of 10%. Comparison among expressions given in Eqn. (16, 19,
21) gives results that differs for between 45° and 60° in the range of 20%.
minimum and a maximum value. These values are obtained placing Eq. (7) equal to Eq. (9)
f ct 1
min 0.2 (22)
fc
1 ct
y
b2
max cos (23)
BH
In Fig. 8 is showed the variation of the minimum and the maximum mechanical ratio of
reinforcement, with variation of θ for Winkler, rigid or flexible soil modelled assuming y=1.22 B/4
10
f ct
and y=0.72 B/4 respectively. In the application of Eq. (22) it was assumed that 0.1 and it was
fc
ct
supposed that was negligible.
y
From graphs of Fig. (8) it can be observed that the minimum mechanical ratio is equal to 0.02
which is very close to the value suggested in Eurocode 2 [9] equal to 0.022. Maximum value
When < max the ultimate load of deep footing is given by Eq. (15)
For flexible column footing having a square base, it was used a cantilever beam model derived
decomposing the footing in four independent cantilever beams subjected to uniform pressures, as
showed in Fig. 9.
In the case of uniform contact pressures, the maximum bending moment in the fixed section is
equal to:
M max
P
24 B 2
B 2 2 B 3b b 3 (24)
In the case of rigid soil (type a) of Fig. 4) the moment is equal to:
M max
P
56 B 3
B 3 5 B 7b b 3 B b (25)
While for soft soils, (type b of Fig. 4b) the moment is equal to:
M max
P
40 B 3
B 3 3B 5b b 3 3B b (26)
The maximum shear force, in the case of uniform pressures, is equal to:
11
P B b
Vmax (27)
2 B
The moment at cracking in the more stressed section (calculated not considering, for brevity sake,
H2
Bb
M cr B f ct with B For shaped footing (28-a)
6 2
H2
M cr B f ct For prismatic footing (28-b)
6
M rd A f f y z (29)
Puc 24 B 2 B H 2 f ct
2 For shaped footing (30)
A f f y z B 2 B 3b b 3 6 A f f y z
Puy 24 B 2
(31)
A f f y z B 2 (2 B 3b) b 3
Studies present in literature (Viggiani [2], Toniolo [3] and Menditto [14]) give similar expressions
to Eq. (31):
Pu 24 B 2
Viggiani [2] (32)
A f f y z B b 2 (2 B b)
Pu 8 B
Toniolo [3] (33)
A f f y z B b 2
12
Pu 1
Menditto [14] (34)
A f f y z B b
12 8
Comparing results obtained with Eqn. (31, 32, 33, 34) it is possible to see that values obtained are
different at least 10%. In the case of beam model, limits of maximum and minimum reinforcements
In this case, the ultimate load can be predicted utilizing one of the expressions given in the literature
[9-10] for punching shear failure for two-way slabs. In these expressions, it was taken into account
that under punching shear failure the plate of area A is punched when a vertical fracture cross-
section is established along the entire perimeter of the penetration body that is formed near the
column. It occurs in the so-called critical or control section of the perimeter Okp. Slab punching of
the effective depth d occurs when the shear stress in the critical cross-section reaches the shear
strength of the concrete. Therefore, punching calculation is reduced to checking the shear stress in
the critical section. Regulations for calculating ultimate shear stress at punching in the control
section greatly differ for the position and shape of the control section of area A0 and calculation of
the shear stress . In the case of punching shear taking into account the part of the soil reaction
1
Pu , p Okp d (35)
Ao
1
A
From Eq. (35) it emerges that the ultimate load increases with the following: - thickness of the slab;
Utilising Eq. (35) it is possible to obtain the ultimate load, that with expression of and Okp given
4 b d d
0.332 f c
Pu , p
1
B 2 b d
2
[9] (36)
B2
200 A
0.33
4 b d d
0.18 1 100 f f c 0.33
Pu , p
d B H
1 B b d
2 2
[10] (37)
B2
For flexible foundation as stressed in Bonić and Folić [20], the pressure exerted on the entire
contact surface of the footing on the granular soil is not uniformd and a corrective factro has to be
introduced. This factor is defined as the ratio of the average pressure under the punching body to the
average pressure under the entire footing and in Bonić and Folić [20] and a constant value of 1.4
For rigid foundations, in agreement with the soil pressure distribution assumed in Fig. 8, we have
Fc= 3/7 (0.428) and 9/5 (1.8) for rigid and flexible soil, respectively. While for elastic foundations
under elastic soil modelled with the Winkler model, it is possible to determine Fc analytically.
Hartog (1952) gives the expression for the centre deflection of an elastic beam subjected to a
concentrated load P:
P W 2 cos W B cosh W B
y center (38)
2k sin W B sinh W B
6k
where k is the Winkler soil constant and W 4 with Ec the elastic modulus of concrete
Ec H 3
4200 f c
(e.g. in the cracked state according to [10].
2
If the beam is rigid and a Winkler model for the soil is assumed, we have:
14
2P
y center (39)
k B B b
The Fc factor is obtained as the ratio between Eq. (38), and Eq. (39):
W B B d 2 cos W B cosh W B
Fc (40)
4 sin W B sinh W B
1 d 4 b d 0.33
200 Af
Pus 0.18 1 100 f c 0.33
Fc
B 2 b d
2
d BH
(41)
1 B2
The ultimate load for a slender footing is the minimum between Eq. (31) moreover, Eq. (41).
To validate the proposed model a comparison between experimental data, referred to real column
footings is shown in Fig. 11. Data of Siburg and Hegger [18] and Simões et al. [19] are used. All
speciemens were characterized by angle lower than 50° for which punching shear was observed
experimentally. Geometry and mechanical characteristics of specimens tested are shown in Table 1.
In the graph of Fig. 11 data were predicted with strut and tie model, with [9-10], and beam model
with punching shear and including the flexibility of foundation with Fc assumed 0.9 for simulate
rigid soil. From the comparison it emerges that the strut and tie model is the unsafe model, while the
more accurate is the flexural model with punching shear. The flexural model without punching
shear is unsafe. Eurocode2 [9] gives safe prediction, while ACI 318 [10] is unsafe.
If, in the strut and tie model, the equilibrium equations are utilized coupled with the
compatibility equation of the loaded joint and with the constitutive laws of strut and tie members it
is possible to derive the load-displacement (P-) curves of RC column footing. The load P is the
15
external load applied, while the displacement is the vertical displacement of the loaded joint. The
compatibility equations relate the elongation of the tie member δs and with the shortening of strut
c
s (42)
sin tan
If the Winkler soil model is adopted and the soil settlements have to be enclosed in the vertical
P 1
displacement, the effective displacement is obtained adding the term to with k being the
B2 k
Winkler coefficient, as described in Hetenyi [7]. It has to be stressed that the parameter δ represents
the shortening of footing without soil settlement. At this stage of the study, soil settlement is not
considered in the calculus of the displacement when refer to numerical simulation in which this
Ty a
sy (43)
Es A f
With Af the area of longitudinal reinforcement in x (or y) direction, Es the modulus of elasticity and
B B
a the shear span expressed as a for prismatic column footing and as a for shaped column
4 3
footing.
C z2 a2
c (44)
Ec b 2
Substituting Eq. (36) and Eq. (37) into Eq. (35) it results:
16
z2 a2 1 a 1
y Af f y (45)
4 E c b sin E s A f tan
2
And at strut crushing with tie to be yielded and in the hardening phase it results:
Puc
fy
a f y 4 A f tan Puc z a
2 2
1
u (46)
tan E s Eh 4 E sec b 2
sin
With:
fc
Esec (47)
2o
Eh Es (48)
h being the hardening modulus of steel and being the hardening coefficient of steel bars.
It has to be underlined that the ratio between δu and δy can be assumed as a measure of ductility
Puc z2 a2 1 a 1
cu 2
(49)
4 E c b sin E s A f tan
2
4
The load- shortening curve of the column footing (P-δ) is constituted by three linear branches.
If a cantilever model is adopted the deflection of a single beam subjected to the distributed load q is:
17
4
Bb
q
2
(50)
8 Ec J
bH3
J being the moment of inertia of the cross-section expressed as J cr before cracking or
12
1 bH3
Jy at yielding.
3 12
4
B b
2 2 q 2 q1 q 2
(51)
2 E J 4 15
With q1,q2 the soil reactions for rigid, flexible and Winkler soil, that are equal to:
q1 9 7 p m
For rigid soil (52-a)
q 2 3 7 p m
q1 9 5 p m
For soft soil (52-b)
q 2 3 5 p m
With:
P
pm 2 B (53)
B
The application of beam model gives three linear curves in which the first branch stops to cracking
load, the second one stops up to yielding load, and the third one (with strain-hardening effect) stops
to the ultimate displacement corresponding to the curvature in which steel failure in tension failure
occurs.
18
EXPERIMENTAL RESULTS
To better understand the problem here presented, and to have a confirm of the proposed model
finite elements analyses were carried out through the software ATENA [17]. The soil is here
modelled with independent springs having stiffness proportional to the type of the considered soil
or as fixed supports to simulate the effect of rigid soil. Shaped and single column footings were
modelled through isoparametric finite elements having nine joints. The concrete material utilized
was the so-called “SBeta” [17]. It has assumed a constitutive tensile law for concrete having linear
behaviour before cracking. When maximum principal stress exceeds tensile strength, cracks
develop. In this case, it was utilized a “smeared crack model” where cracks were modelled reducing
principal tensile stress according to a constitutive law of tensile concrete. Instead of represent a
single crack, distributed cracks having direction normal to tensile principal stress direction are
represented. This hypothesis is quite realistic for concrete, where the formation of cracks is
preceded by a microcracking of material. It has supposed perfect bond between concrete and steel.
Compressive softening coefficient has been assumed 0.7. Steel is assumed with an elastoplastic
For the numerical analyses, two types of column footings were considered. The first one is
constituted by shaped column footing having a square base of side 1000 mm, variable heights of
400, 500, 650, 1000 mm, with a reinforcing grid in x and y direction constituted by 8 bars of 14
diameter with yield strength fy=450 MPa. Concrete strength was assumed 45 MPa. Area of steel
Af
bars was adopted in the range of geometrical ratio of 0.00308, 0.00246, 0.00189,
bH
0.00123, respectively.
The second series of column footing was constituted by single prismatic members having square
base 800 mm. Reinforcement was constituted by 8 bars having 14 mm diameter in x and y
19
directions. Different heights of 800, 700, 600, 500, 400 mm were considered. Area of steel bars was
Af
adopted in the range of geometrical ratio of 0.0019, 0.0022, 0.00256, 0.00304, 0.0038,
bH
respectively.
For all column footing having side 1000 and 800 mm fixed supports at the base were considered.
For the footing having side 800 mm were also considered supports constituted by elastic springs
having stiffness that reproduces typical values of cohesive and non-cohesive soils. Particularly,
were examined sand and clay for which were assumed Winkler constant k for the considered base
equal to 118 N/cm3 and 12.5 N/cm3. Each column footing has been loaded through a growing
Fig. 12 shows shaped column footings analysed in numerical simulations. For each column footing
Fig. 13 shows the variation of the maximum load corresponding to yielding of longitudinal steel
and failure of compressed concrete struts with the height of the footing obtained numerically and
analytically. From the comparison, it results in good agreement between numerical and analytical
results. It is possible to observe that, with increasing the height of column footing, the load
increases. Exceeding the height of 800 mm, corresponding to θ=60°, failure of strut happens after
In Fig. 14 and in Fig. 15 are shown the curves for the column footing having base 800 mm. Fig. 14
gives failure load values with the variation of the height of the footing and Fig. 15 shows the
variation of the load with the variation of steel reinforcement. Good correspondence between
From Fig. 15 it is possible to observe that the variation of the peak load in the case of concrete
crushing, is lightly influenced by the variation of reinforcement area, whereas in the analytical
20
model it is not influenced. From the graph, it is possible to see that results obtained with strut and
tie model agree with numerical results and failure of deep beams (θ>60°) is due to strut crushing.
Fig. 16 shows the variation of failure load with variation of θ: a) for 3 different classes of strength
of concrete (Fig. 16 a); b) for different strength grades for steel (Fig. 16b); c) for rigid, flexible or
Winkler soils (Fig. 16 c). From the graphs, it is possible to see clearly the dependence of the failure
curves it was used the finite elements code ATENA. Fig. 17 shows load-displacement curves
deduced from the analytical model (beam model and strut and tie model) and with the numerical
model about single column footings having different heights. The comparison shows that, in the
considered field of investigation (θ between 45° and 60°), the strut and tie model better interprets
the numerical behaviour than the cantilever beam model. Yield load calculated with beam or strut
and tie models are much close to numerical values; strut failure happens after yielding of
reinforcement for the mechanical ratio of reinforcement higher than the value given by Eq. (24).
Strut and tie model overestimates in the range of 10% displacements obtained numerically, whereas
For all cases examined, calculus of ductility factor was carried out. Results are presented in Table 2.
From data of Table 2, it is possible to see that values of ductility calculated using beam model are
comparable with those obtained from the numerical model. All values of ductility are close to one
A comparison between theoretical and experimental results was made with data of Vacev et al. [16]
and Bonic and Folic [20]. Column footing had a square base of side 850 mm, height 125 mm with a
cover 25 mm and θ=19°. The same reinforcement in x and y direction were adopted and constituted
by a grid in the lower part of the foundation of seven bars having 8 mm of diameter and yielding
stress equal to 570 MPa and stress failure 637 MPa. Concrete had a compressed strength of 25 MPa.
21
Experimental analyses were carried out: - arrange the foundation on the ground surface; - applying
a centred load through a hydraulic jack of 1000 kN; - controlling each second the vertical
displacement of the loaded section and in other characteristics point as, for example, the corner of
footing.
In Fig. 18 are showed the load-displacement experimental curves and those obtained theoretically
using beam and strut and tie analytical [20] models. From results, it is possible to see that the beam
model is in good agreement with the experimental results, both regarding stiffness and
displacements during the different phases. Strut and tie model is stiffer than numerical one and after
yielding of steel considers concrete crushing. In Fig. 18 are also given the ductility factors
calculated.
Another comparison is carried out with data of Hallegren and Bjerke [21], where some column
footing having side 960 mm and height 275 mm lied on the ground through fixed supports with a
distance of 268 mm from the axis of the specimen were tested in compression. The same
reinforcement in x and y direction were adopted and constituted by a grid in the lower zone of the
footing having a mechanical geometrical ratio 0.0042, and yielding stress of 621 MPa. Concrete has
a compressive strength of 24.7 and 34.1 MPa for specimens 1 and 2, respectively. Experimental
analyses are carried out: - arrange the footing on fixed supports and applying a centred load through
a hydraulic jack on the column having side 250 mm. In Fig. 19 are showed load-displacements
experimental curve and those obtained theoretically with beam, strut and tie models. From results, it
has to highlight that strut and tie model is in good agreement with experimental results in the first
branch of response up to punching shear failure observed experimentally and predicted theoretically
in Bonic and Folic [19]; - but not considered in the current model.
22
4. CONCLUSIONS
In the present paper, a simplified calculus model to determine the structural behaviour of reinforced
concrete column footings subjected to centred compression is presented. Two different models are
used: strut and tie model for deep footing and cantilever beam model for flexible column footings.
Different loading profiles are considered, corresponding to different soil types (rigid or soft soil).
The role of the different position of soil pressures resultant, both regarding load both regarding
displacement capacity of column footing are considered. Results obtained stressed that angle
towards horizontal lower than 45° then beam model with punching shear is more suitalbe, while for
More in details results obtained for deep RC column footing showed that for the prediction of
ultimate load:
- the angle expressed as the ratio between the depth and the shear span of the footing) to define
the range of using strut and tie model from the range of using beam model is an important
- strut and tie model accurately predicts the ultimate load of deep column footing with yielding of
- strut and tie model allows one to define the minimum and maximum mechanical ratios depending
on the geometry of the footing, on the material’s characteristics and the type of soil corresponding
to the simultaneous occurrence of concrete cracking in tension and yielding of steel bars or yielding
- beam model including flexural and punching shear failure mode predicts the ultimate load of
flexible foundation;
23
- for mechanical ratio lower than the maximum mechanical ratio, a combined flexural-shear
mechanism governs the column footing behaviour, and yielding of steel bars occurs before concrete
crushing;
- for mechanical ratio higher than the maximum mechanical ratio, brittle failure governs the column
footing behaviour with concrete crushing of compressed struts and steel bars in elastic range;
- the maximum mechanical ratios decrease with the increasing of the height of column footing and
The mechanical model proposed for the determination of load-displacement curves of columns
footing allows one to predict in a satisfactory manner both numerical and experimental results
available in the literature stressing the importance in choosing the most appropriate simplified
model (strut and tie model or beam model) to predict the structural behaviour of column footing.
Variation of the main parameters included in the model and governing the structural behaviour of
column footing (geometry, size, concrete strength, the mechanical ratio of longitudinal bars) reflect
REFERENCES
[1] Albert L. Interazione fra struttura e fondazioni, Riunioni e Congressi 1964 (only available in
Italian).
[2] Viggiani C. Fondazioni. Helvelius Editore 1999 (only avalaibale in Italian); pp. 584.
[3] Toniolo G. Il cemento armato. 2° Volume, Zanichelli (only avalaible in Italian).
[4] Migliacci A, Mola F. Progetto agli stati limite delle strutture in c.a. Parte I e parte II. Ed.
Masson Italia, Milano 1984 (only available in italian).
[5] Simões JT, Faria DMV, Ruiz MF, Muttoni A. Strength of reinforced concrete footings without
transverse reinforcement according to limit analysis. Eng Struct 2016; 112; 146–161.
[6] Terzaghi K. Theoretical soil mechanics. John Wiley & Sons , Inc. - New York, London, Sidney
1964.
[7] Hetenyi M. Beams on Elastic Foundation. University of Michigan Press. Ann Arbor. MI. U.S.A
1964.
[8] Prandtl L. Uber die harte plasticher korper. Nachr. Ge.s Wiss. Goettingen, Math. Phys. Kl. 1920;
74-85.
[9] Eurocode 2: Design of Concrete Structures. Part 1-1: General Rules and Rules for Building
(ENV1992-1-1), 2004.
[10] ACI Committee 318. Building Code Requirements for Structural Concrete (ACI 318-08) and
Commentary, American Concrete Institute. Framington Hills MI 2005; 430 pp.
[11] Leonhardt F. C. A. & C. A. P. Calcolo di progetto e tecniche costruttive”, Edizioni Studio M
& B, 1980 (only available in Italian).
[12] Campione G. Flexural behavior of steel fibrous reinforced concrete deep beams. J Struct Eng
2012; 138(2); 235-246.
[13] AASHTO Guide Specifications for LRFD Seismic Bridge Design, 2nd Edition, with 2012,
2014, and 2015 Interim Revisions.
[14] Menditto G. Esercitazioni di tecnica delle costruzioni – volumi I e II. Liguori Editore (only
available in italian).
[15] Hegger J, Ricker M, Ulke B, Ziegler M. Investigation on the punching behavior of reinforced
concrete footing. Eng Struct 2007; 29; 2233-2241.
[16] Vacev T, Bonic´ Z, Prolovic´ V, Davidovic´ N, Lukic D. Testing and finite element analysis of
reinforced concrete column footings failing by punching shear. Eng Struct 2015, 92; 1–14.
[17] Cervenka, V. Simulating a response. Concrete Engineering International 2000; 4; 45–49.
[18] Siburg C and Hegger J. Experimental investigations on the punching behaviour of reinforced
concrete footings with structural dimensions. Struct Conc 2014; 15(3); 331-339.
25
[19] Simões JT, Bujnak J, Ruiz MF, Muttoni A. Punching shear tests on compact footings with
uniform soil pressure. Struct Conc 2016; 4; 603-617.
[20] Bonic Z, Folic R. Punching of column footings-comparison of experimental and calculus
results, Grandevinar 2013; 65(10); 887-899.
[21] Hallgren M, Bjerke M. Non-linear finite element analyses of punching shear failure of column
footings, Cem Concr Comp 2002; 24; 491–496.
26
List of tables
Tab. 1 – Characteristics of column footings tested in Siburg and Hegger [18] and Simões et al. [19].
Tab. 1 – Characteristics of column footings tested in Siburg and Hegger [18] and Simões et al. [19].
28
List of figures
Fig. 3 - Plastic soil-foundation pressures distribution for rigid footing as in Toniolo [3].
Fig. 4 - Simplified soil-foundation pressure distributions for rigid foundation of: a) rigid soil; b)
flexible soil; c) Winkler’s model of soil.
Fig. 7 - Softening coefficients for compressed struts of deep footing with variation.
Fig. 8 - Variation of minimum and maximum mechanical ratio of longitudinal reinforcement with
Fig. 11 - Analytical versus experimental load response of footings testes in Siburg and Hegger [18]
and Simões et al. [19].
Fig. 12 - Truncated pyramidal footing with H (mm): a) 1000; b) 650; c) 500 ; d) 400.
Fig. 16 - Load carrying capacity of RC footing with variation of for : a) concrete strength; b)
yielding stress of longitudinal steel; c) type of soil.
P
pm 2 B
B
a) b)
c) d)
Fig. 2 - RC column footing: a) rigid footing on rigid soil (clay); b) rigid footing on soft soil (sand);
c) flexible footing on rigid soil; d) flexible footing on soft soil.
32
Fig. 3 - Plastic soil-foundation pressures distribution for rigid column footing as in Toniolo [3].
33
a) b) c)
Fig. 4 - Simplified soil-foundation pressure distributions for rigid column footing on:
P P
b b
H
s s
H
s
s/2
B
s
B B
H
H
T
35 y
P/4
P b/4
b
b
B
B
C
H
z
y
T
y
P/4 b a
Fig. 6 - Strut and tie calculus model for RC deep column footing.
36
Fig. 7 - Softening coefficients for compressed struts of deep column footing with variation.
37
H
H
T 45°
y
38 P/4 a
b+
pt
P P P
b/4
b
b H/2 b failure surface
b
B
B
H
H
T y 45°
y
y pt
P/4 a
b+H
b a pt b a
b
B
B
y
y
b a b a
P P P
b
39 b H/2 b failure surface
C
H
H
H
T 45°
y pt
P/4 a
b+H
P P pt
b
B
B
45°
y
a y
pt
b+H
pt
b a b a
Fig. 10 - Punching shear calculus model for RC column slab footing.
b
B
b a
40
8
ACI model
UNSAFE Beam model without punching
7 Beam model with punching shear
EC2 model
6 Strut and tie model
Serie1
5
Pu,calulated/Pu,exp
1
SAFE
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Pu,exp [kN]
Fig. 11 - Analytical versus experimental load response of footings testes in Siburg and Hegger [18]
and Simões et al. [19].
41
a) b)
c) d)
Fig. 12 - Truncated pyramidal column footing with H (mm): a) 1000; b) 650; c) 500; d) 400.
42
15
Anal. concrete
crusching
12.5
Nc - numerical
10
P (MN)
7.5
Ny - numerical
5
Anal. yielding
2.5
0
0 200 400 600 800 1000 1200
H (mm)
Fig. 13 - Load carrying capacity of RC truncated pyramidal column footing with variation of H.
43
15
Anal. concrete
crusching
12.5
Nc - numerical
10
P (MN)
7.5
Analytical yielding
Ny - numerical
2.5
0
0 200 400 600 800 1000 1200
H (mm)
a)
b)
c)
Fig. 16 - Load carrying capacity of RC column footing with variation of for different: a) concrete
strength; b) yielding stress of longitudinal steel; c) type of soil (rigid, soft, Winkler).
46
16
16
H=800 mm
H=700 mm
14
14
beam model truss model
beam model truss model
12 12
10 10
numerical
P (MN)
P (MN)
8 8 numerical
6 6
4 4
2 2
0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
(mm) (mm)
16
16 H=500 mm
H=600 mm
14
14
truss model 12 numerical
12 truss model
10
10 numerical
P (MN)
P (MN)
8
8
beam model
beam model
6 6
4 4
2 2
0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
(mm) (mm)
tested in [19].
Model for compressive behavior of RC column footing with square cross-section
Strut and tie model for concrete footing with variation of geometry and soil type
Numerical model and experimental results do RC single and shaped footing
for mechanical ratio higher than the maximum mechanical ratio, brittle failure governs the
column footing behavior
concrete cracking in tension, yielding of steel bars and yielding of steel bars and concrete
crushing strut and tie model allows one to define minimum and maximum mechanical ratios