Allouzi 2018
Allouzi 2018
Allouzi 2018
10.1680/jstbu.18.00011
Accepted manuscript
As a service to our authors and readers, we are putting peer-reviewed accepted manuscripts
(AM) online, in the Ahead of Print section of each journal web page, shortly after acceptance.
Disclaimer
The AM is yet to be copyedited and formatted in journal house style but can still be read and
referenced by quoting its unique reference number, the digital object identifier (DOI). Once
the AM has been typeset, an ‘uncorrected proof’ PDF will replace the ‘accepted manuscript’
PDF. These formatted articles may still be corrected by the authors. During the Production
process, errors may be discovered which could affect the content, and all legal disclaimers
that apply to the journal relate to these versions also.
Version of record
The final edited article will be published in PDF and HTML and will contain all author
corrections and is considered the version of record. Authors wishing to reference an article
published Ahead of Print should quote its DOI. When an issue becomes available, queuing
Ahead of Print articles will move to that issue’s Table of Contents. When the article is
published in a journal issue, the full reference should be cited in addition to the DOI.
This paper investigates the non-linear behavior of slender reinforced concrete (RC) columns with inclination.
The available studies on the inclined columns (not strut) are not sufficient in comparison to those on the
conventional columns and struts. The critical buckling load of inclined slender columns is derived first using
second-order analysis then 3D finite element models are created to propose a relationship to estimate the
effective flexural rigidity of straight columns pushed up to buckling threshold. Finally, the buckling loads
estimated using FE models of inclined columns and proposed effective flexural rigidity are further used to
evaluate the effective length factor of inclined columns at various slenderness ratios, inclination angles, and
boundary conditions. It is concluded that the effective flexural rigidity depends on the slenderness ratio at
buckling threshold. The response of slender leaning RC columns is extensively studied and recommendations
Keywords: Buildings, structures & design; Columns, Concrete structures; Design methods & aids
As the complex shaped structures become a popular trend for architects, columns
tilted at an angle become widespread in many building, towers, and bridges. For example, in
Moscow city center the oblique columns are part of the design of many buildings such as
evolution Tower and Capital city Moscow Tower. Also, the substructure of many bridges is
composed from tilted columns as Mega Bridge in Thailand. These oblique or tilted columns
can be either struts or inclined columns. The former are structural members designed to
mainly resist longitudinal compression (Fig. 1 (b)). The latter are the columns which are
subjected to loads in the vertical direction compared to the tilted direction of the column and
its main reinforcements (Fig. 1(c)). The transverse reinforcements are kept horizontal
Inclined columns are subjected to larger moments compared to vertical columns due
addition to any exerted external lateral force. Taking into consideration these factors, inclined
slender columns can be analyzed by using elastic or inelastic second-order structural analysis.
Alternatively, the moment magnification method based on ACI 318-14 can be used to
estimate the amplified moment due to slenderness of straight columns (Chang, 1974). This
method is mainly depends of effective length calculation which is still not clear for inclined
slender columns. Chang (1974) studied slanted column which was represented as an inclined
member composed of an axisymmetrically deformable conical shell with rigid edges (Chang,
1974). The buckling of elastic slanted columns was derived by the tool of variational method
accounts for the boundary conditions to develop a set of nonlinear equations. The exact
solutions for these equations are gained for inclined columns with various end conditions.
The results are compared with eigenvalue method and other two conventional methods
(Chang, 2006).
BS8110 and applied it to a case study of Taman Ekologi Hutan Bandar, Johor Bahru that
involved three storey building composed of inclined column, ring beams and ring slabs.
Analysis and design of the inclined column resulted in a larger moment in the inclined
Another method of analysis permitted by ACI 318-14 is the finite element method.
Lam et al. (2012) used ABAQUS to model the axial behavior of inclined, tapered, and
straight–tapered–straight of steel tubular stub filled with concrete. The model was calibrated
based on experimental test conducted on RHS filled with concrete with limited range of
inclined angle and tapered angle. The results of compressive behavior of FE simulations were
in good agreement with experimental outcomes. Then, a parametric study was performed on
3 steel hollow sections with wider range of inclination angle and taperness angle.
Watan (2011) studied the shipside collision against inclined column. Two types of
boundary conditions of the column has been modeled; fixed-fixed boundary conditions and
boundary conditions with axial flexibility. Two different impact locations were simulated,
namely; middle span impact and quarter span impact. The ship model had been built using
columns with inclination is required and, from the results, the development of design
procedures for incorporation into building codes can be attained. This paper presents the
development of hybrid analysis composed from second-order analysis and finite element
analysis. A second-order analysis is developed first to estimate the critical buckling load of
inclined columns then a calibrated finite element models are developed to be capable to
simulate the response of RC slender columns. These calibrated finite element models are
further used to identify the effect of column inclination on the effective flexural rigidity of
these columns. These results are used in the developed critical buckling load to investigate
the effective length factor of these columns. Of particular interest for this research are
2. SECOND-ORDER ANALYSIS
magnitude of the applied vertical load (P) multiplied by tanФ (Fig. 2(a)). The components of
vertical and lateral forces in the direction along the column and the perpendicular direction
result on only one force along the column of the magnitude of P/cosФ as shown in Fig. 2(b).
(1)
(2)
(4)
( ) ( ) (5)
respectively.
The “A” represent the amplitude of deformed shape and shall not be zero. Accordingly,
(6)
Correspondingly,
(7)
This represents the elastic load at buckling threshold. The first mode controls when
( )
(8)
( )
This equation requires the estimation of buckling load and effective flexural rigidity to
evaluate the effective length factor of inclined RC slender columns. The simply supported
ends are generalized to include other possible end conditions based on the effective length of
buckling.
1- Finite element models are prepared based on Jenkines specimens for calibration.
2- Then, FE models are extended to estimate the effective flexural rigidity by using
equation (8) for straight pinned-pinned, pinned-fixed, and fixed-fixed columns under
3- Then, finite element models are developed for columns under various inclination
angles and slenderness ratio. The peak load just before buckling is obtained from FE
4- Rearrange equation (8) to estimate the effective length factor for each inclined
column. The effective flexural rigidity is obtained from step 1 and buckling load is
the behavior of structural components and members, calibrated finite element model is a
beneficial tool to further investigate the effect of other parameters on the response of these
structural members. In this paper, finite element models are developed using ABAQUS 6.13
and then calibrated based on four RC slender column specimens tested by Jenkines (2015).
columns under short-term and long-term loading. The main goal was to better understand the
behavior of these columns and develop improved design procedures since provisions for the
design of slender columns have not changed considerably since the first provisions
Four RC slender column specimens tested by Jenkines (2015) are considered in this
study to calibrate FE models. The cross setion is 155.6mmX 155.6 mm with main
reinforcement of 4ϕ16 at the corners for all columns. The transverse reinforcements are ϕ6
ratio of 40 and tested under 10 % and 25% eccentricities. The other two specimens are having
a slenderness ratio of 70 and tested under 10 % and 25% eccentricities. Then, these calibrated
models are extended to model 48 columns under various ranges of inclination angles and
slenderness ratios.
The models of columns studied in this paper are shown in Fig. (3) and Fig. (4). The
horizontal section (not the cross-section which is perpendicular to the column direction) is
kept the same for all columns. The columns are grouped based on the vertical length as 1.8 m,
2.2 m, 2.7 m, and 3.2 m. Each group consists of columns inclined by 0o, 10o, 20o, and 30o.
Also, all of these columns are modeled to be subjected to three different boundary conditions,
namely; pinned-pinned, pinned-fixed, and fixed-fixed. The results of these 48 columns are
studied in this paper. The vertical length of these columns and the inclination angles are
summarized in Table 1.
3D finite element models are developed for RC inclined slender columns from truss
concrete. Steel plates and rollers are modeled as 8-noded hexahedral 3D solid elements. Mesh
The Concrete Damaged Plasticity model is used in this paper to simulate the behavior of
concrete. This model is a continuum model that is based on the damaged elasticity, and
tension and compression plasticity for brittle materials. For compression plasticity
identification, the modified Popovics’ relation (Thorenfeldt et al., 1987) is used to define the
(9)
[ ]
(10)
where and are the concrete characteristic strength and corresponding strain without
considering any confinement, respectively. and are the concrete peak strength and
ascending side of the relationship and is adjusted for the descending side such that the final
( ( ))( ) (11)
( )
Richart et al. (1928) defined the peak strength and strain of concrete under hydrostatic
fluid pressure as
(12)
( ) (13)
( ̅
) (14)
(15)
̅ (16)
( ) (17)
(18)
( )
The concrete response under tension is modeled elastic until tensile strength ( ) is
reached. After this, the stress ( ) - strain ( ) relationship that identifies the plastic domain of
(19)
[ ]
( )
where is the strain at tension strength. and are parameters the influence the shape of
(20)
The definition of concrete behavior after crack (plastic range) is essential since it
simulates the load transfer across cracks to steel bars and is based on bond characteristics
Given that the main and transverse reinforcement bars are modeled as truss elements,
the elastic-plastic constitutive law is used to simulate the uniaxial behavior. The stress-strain
data are taken from test results provided by Jenkines (2015) (Fig. 6).
Elastic constitutive law is used to simulate the response of steel plates and rollers.
4 RESULTS OF FE ANALYSIS
The simulation results from proposed FE models are compared with experimental
outcomes of the four specimens used for calibration. The applied force versus lateral
are shown in Fig. 7 through 10. The measured applied force versus lateral displacement at
mid-height of columns curves based experimental results are shown in the same figures. The
simulation results compare very well with experimental outcomes for all considered
specimens. The initial stiffness based on FE results overestimates the measured initial
stiffness due to the assumed full bond between concrete and steel reinforcement at this stage
of loading since no cracks occurred yet. When cracks started to occur, the load transfer across
cracks to steel bars can be modeled with the implemented tool in FE models of tension
stiffening mentioned previously to represent the bond characteristics between concrete and
steel.
The damage pattern of tested specimen is best represented by plotting the equivalent
plastic strain of simulated specimens. The results of columns with slenderness ratio of 40 are
shown in Fig. 11. The simulated models could capture the crack of concrete and buckling of
the reference.
5 EFFECTIVE EI
The critical buckling load ( ) is calculated based on the effective flexural rigidity
and effective length. For straight pinned-pinned, pinned-fixed, and fixed-fixed columns
subjected to concentric load, the effective length factors (k) are 1, 0.7, and 0.5, respectively.
The calibrated finite element is used future to investigate the maximum load the column can
carry before buckling ( ) under wide range of slenderness ratios and the aforementioned
boundary conditions. Accordingly, the effective flexural rigidity can be estimated as follows
( )
( ) (21)
Based on ACI 318-14, the effective EI can be estimated by one of the following equations
( ) (22)
( ) (23)
( ) (24)
where is the modulus of elasticity of concrete and is the gross moment of inertia of the
moment of inertia of reinforcement about centroidal axis of the column cross-section. is the
is the ratio of sustained load to total load. Since the finite element model is subjected to
This reduces the number of variables in the equation for better estimation of the effective
( ) (25)
where is called the effective flexural rigidity factor and is evaluated based on the
( )
(26)
( )
The results are shown in Fig. (12). The linear regression of the outcomes is presented
as
(27)
where r is the radius of gyration taken as 0.3 times the cross-section depth.
Pinned-pinned straight columns are all slender. All straight columns under the fixed-
fixed boundary condition resulted in short column except the column with vertical length of
3.2 m. For pinned-fixed boundary condition, the straight columns with vertical length of 1.8
m and 2.2 m are short. All short columns are not considered in the development of equation
(27).
columns are 1, 0.7, and 0.5, respectively. For inclined columns, is the effective length factor
still the same? Based on the proposed effective flexural rigidity factor, the effective length
√ (28)
Since α depends on slenderness ratio , the equation is solved iteratively for all
column models adopted in this paper. The outcomes are summarized in Table 2 through
it short column (based on FE results) and they are addressed in these tables. Accordingly,
their k results are excluded especially for all fixed-fixed inclined columns considered herein.
The effective length factor is changed for column with an inclination at different
vertical length and boundary conditions. For specific vertical length and boundary conditions
of a column, the inflection point’s location changes as the column inclination increases
resulting in changing the mode from single curvature to double curvature or vice versa. The
effective length factor versus the vertical length of columns studied herein under various
inclination angles are shown in Fig. 13 and Fig. 14. Fig. 13 presents the result of pinned-
pinned columns. It is clear that as vertical length increases, the effective length factor
decreases if inclined at 10o or 20o. For inclination angle of 30o, the results of effective length
factor scatter between 0.9-1.17. This is also observed by the deflected shape of these modeled
columns in ABAQUS.
Fig. 14 presents the result of pinned-fixed columns. The trend here is more clear and
different than the pinned-pinned columns. As vertical length increases, the effective length
factor decreases if inclined at 20o or 30o. This can be attributed to the control the fixed
support enforces on these columns. The trend of columns inclined at 20o and 30o is the same.
At small inclination angle of 10o, the column of the highest vertical length is the only one that
stayed slender.
these columns are currently part of many structural systems to transfer loads. In this paper, a
hybrid analysis of analytical second-order analysis and nonlinear finite element methods are
developed to represent the response of RC inclined slender columns. The existing data from
experimental work assist calibrating these models and further used in this paper to study the
behavior of RC slender inclined columns under various ranges of slenderness and the
1. The developed numerical non-linear finite element models are capable to simulate
the performance of RC slender columns. The calibrated finite element models can
in the deformed shape. The buckling load of inclined slender columns is derived.
4. The effective length factors of inclined columns are not 1 and the inflection points
are moved due to the inclination of column direction. The effective length factor is
their values decrease as the vertical length increases at inclination at 20o or 30o. At
10o inclination, the columns are considered short except at vertical length of 3.2
m.
Notation list
ACI committee 318: Building Code Requirements for Structural Concrete and Commentary.
Chang, G.A. and Mander, J.B. (1994). “Seismic Energy Based Fatigue Damage Analysis of
Chang C. (2006). Buckling of Inclined Columns. In: Mechanics of Elastic Structures with
Inclined Members. Lecture Notes in Applied and Computational Mechanics, vol 22.
Dassault Systèmes Simulia Corp. (2011), ABAQUS Analysis User’s Manual Online
Dolhakim, S., 2008. ANALYSIS AND DESIGN OF INCLINED COLUMN. Master’s thesis.
Lam, D., Dai, X.H., Han, L.H., Ren, Q.X. and Li, W., 2012. Behaviour of inclined, tapered
and STS square CFST stub columns subjected to axial load. Thin-Walled Structures, 54,
pp.94-105.
Richart, F. E., Brandtzaeg, A., and Brown, R. L. (1928). “A study of the failure of concrete
Watan, R., 2011. Analysis and Design of Columns in Offshore Structures Subjected to
0 1.78 155.6
10o 1.81 153.2
1.8
20o 1.89 146.2
30o 2.05 134.7
0 2.24 155.6
o
10 2.27 153.2
2.2 o
20 2.38 146.2
30o 2.58 134.7
0 2.69 155.6
10o 2.73 153.2
2.7
20o 2.87 146.2
30o 3.11 134.7
0 3.20 155.6
10o 3.25 153.2
3.2
20o 3.41 146.2
30o 3.70 134.7
Table 4: Results of maximum load of fixed-fixed columns based on FE models and outcomes
of effective length factor
lv (m) ϕ Pmax(kN) k
0 1060 short
o
10 1237 short
2.2 o
20 1169 short
o
30 1025 short
0 1092 short
o
10 1237 short
2.7 o
20 1162 short
30o 1031 short
0 1019 0.18 0.5
o
10 1268 short
3.2 o
20 1195 short
30o 1029 short
Figure 1: (a) column, (b) strut, and (c) inclined column considered in this paper.
Figure 2: Inclined column subjected to vertical load (P); (a) reactions and (b) equilibrium
based on deformed shape.
Figure 3: Geometry of the models of columns considered in this paper
Figure 4: Finite element model of inclined columns of (a) pinned-pinned boundary, (b)
pinned-fixed boundary, and (c) fixed-fixed boundary.
Figure 5: Concrete stress-strain curves for 50 mm confinement at edges and 150 mm in the
middle.
Figure 6: Stress-strain used in ABAQUS and taken from Jenkines tests [4].
Figure 7: Force-lateral displacement at the middle for the column of slenderness ratio of 40
and eccentricity of 10%
Figure 8: Force-lateral displacement at the middle for the column of slenderness ratio of 40
and eccentricity of 25%
Figure 9: Force-lateral displacement at the middle for the column of slenderness ratio of 70
and eccentricity of 10%
Figure 10: Force-lateral displacement at the middle for the column of slenderness ratio of 70
and eccentricity of 25%
Figure 11: Damage pattern of specimens considered for calibration; (a) tested column with
slenderness of 40 and 10% eccentricity (Jenkins,2015), (b) tested column with
slenderness of 40 and 25% eccentricity (Jenkins,2015), (c) simulated column with
slenderness of 40 and 10% eccentricity, (d) simulated column with slenderness of 40
and 25% eccentricity
Figure 12: Effective flexural rigidity factor of straight columns of different slenderness ratio
and various boundary conditions
Figure 13: Effective length factor calculation for pinned-pinned columns under various
inclination angles
Figure 14: Effective length factor calculation for pinned-fixed columns under various
inclination angles