Sie sind auf Seite 1von 11

AIN SHAMS UNIVERSITY

FACULTY OF ENGINEERING
STRUCTURAL ENGINEERING DEPARTMENT

FIFTEENTH INTERNATIONAL CONFERENCE ON STRUCTURAL AND GEOTECHNICAL ENGINEERING


Advances In Construction Techniques

PLASTIC HINGING OF STRUCTURAL WALL ELEMENTS UNDER


LATERAL LOADING

MARWAN SHEDID
Department of Structural Engineering, Ain Shams University
Abbassia, Cairo, Egypt
E-mail: marwan.shedid@eng.asu.edu.eg
HASSAN BEDEIR
Department of Structural Engineering, Ain Shams University
Abbassia, Cairo, Egypt
E-mail: hassan.bedeir@eng.asu.edu.eg

ABSTRACT
Under seismic loading, inelastic deformations in cantilever walls are usually concentrated at
the base within a plastic hinge zone in which high inelastic curvatures and plastic
deformation of the vertical reinforcement occur as well as partial crushing of concrete or
masonry in compression. Top wall displacement, which is of prime importance in both force-
based and displacement-based seismic design, can be determined by integration of the
curvature profile over the wall height at different loading stages to capture the load-
displacement relationship of the component level in a building. Once such displacement is
determined, either at maximum load or at 20% strength degradation, displacement ductility
and consequently response modification factor used in design can be determine. Prediction of
realistic displacements beyond yield relies on attained curvatures at the base of the walls and
on a mathematically quantified equivalent plastic hinge length, lp, where lumped inelastic
curvatures are concentrated. This paper presents a review of equivalent plastic hinge lengths,
lp, models used in displacement prediction during seismic loading. The models reviewed
account for the variation in curvature profile after yielding, strain penetration inside concrete
foundation as well as effect of inclined flexure-shear cracking to determine wall top
displacements.

KEYWORDS
Plastic Hinge, Plastic Hinge Length, Displacement Ductility, Plastic Hinge Models,
Statistical Analysis
1 INTRODUCTION
During moderate to high seismic events, inelastic deformations are expected to occur in shear
wall buildings which result in a significant reduction in the building’s seismic demand. For
cantilever reinforced concrete and reinforced masonry shear wall, a ductile response can be
achieved through the development of a flexural plastic hinge at the base of the wall resulting
in significant energy dissipation and high ductility capacity (Park and Paulay 1975; Paulay
and Priestley 1992; Drysdale and Hamid 2008).
Although, the moment variation over the height of a cantilever wall with a lateral top load is
linear, the curvature profile over the wall height is nonlinear beyond the elastic limit as a
result of rigidity variation along the wall height. At ultimate load, high levels of inelastic
curvatures are attained at the base of cantilever walls within a plastic hinge zone. This zone is
characterized by the concentration of high inelastic curvatures and plastic deformation of the
vertical reinforcement.
For moderate to large earthquakes, it is uneconomical to design the seismic force resisting
system to behave elastically. Therefore, inelastic behavior through ductile response is
accepted where damage is allowed in pre-specified zones but without causing collapse. The
term ductility in seismic design is used to quantify the ability of a structure to undergo large
deformations in the inelastic range without a substantial reduction in strength. The
displacement ductility can be defined as the ratio between the ultimate displacement and the
yield displacement where the ultimate displacement is the summation of the yield (elastic)
and the plastic (inelastic) displacements. When force-based seismic design approach is
followed, displacement ductility capacity of walls, which is directly related to force
modification factor, is needed for design purposes. Whereas, when performance-based
seismic design approach is followed, ultimate displacements of walls are key input
parameters in the design process.
The quantification of the extent of the plastic hinge zone is important for predicting the
inelastic response of shear walls. Plastic hinging influences the displacement at maximum
load, Δu, and, consequently, affects the curvature ductility level, μφ, required to attain a target
displacement ductility level, μΔ.

2 PLASTIC HINGING CONCEPT


In order to idealize the actual curvature profile, Paulay and Priestley (1992) suggested
representing the inelastic curvature region by an equivalent plastic hinge length, lp, with the
remaining wall height deforming in a linearly elastic manner as shown in Fig. 1.
Fig. 1: Elastic and Inelastic Deformation of Cantilever Structural Walls
Plastic displacements and displacement ductility are significantly influenced by the levels of
strain attained in the concrete or masonry in compression and by the curvatures generated at
the wall bases within plastic hinge zones. An illustration of a typical strain profile for a wall
cross section is shown in Fig. 2 (a). The average curvature, φi, over a certain gauge length,
hgauge(i), along the wall height in experimental testing (see Fig. 2 (b)) can be calculated using
Eq. 1. The idealized curvature profile to be used in analysis and seismic design of walls as
well as calculation of plastic hinge length is shown in Fig. 2 (c).

εT + εC Δ Ti hgaugei + Δ Ci hgaugei
ϕi = = (1)
lw lw

where:
φi = Average curvature over a given segment height, hgauge(i), along the wall height;
εT= Average tensile strain over a given segment height, hgauge(i);
εC= Average compressive strain over a given segment height, hgauge(i);
∆Ci = Net contraction at the compression end of the wall over a segment height, hgauge(i)
∆Ti = Net elongation at the tension end of the wall over a segment height, hgauge(i)
hgauge(i) = Segment height corresponding to the measured ∆Ci and ∆Ti values.

(a)
(b) (c)

Fig. 2: Experimental and Idealized Curvature Profile over Wall Height


To estimate/predict plastic displacement at the top of cantilever walls and consequently
ultimate displacements and corresponding displacement ductility, equivalent plastic hinge
lengths over which plastic curvatures and rotations are concentrated/lumped should be
quantified. Therefore equivalent plastic hinge length could be considered as key parameter
when either performance-based design or forced-based design approaches are adopted in
design of cantilever walls.
When analysing a cantilever wall subjected to top lateral load using plastic hinge analysis
method, top displacement is obtained as the summation of yield displacement Δy and plastic
displacement ΔP. For the loading and support conditions of the cantilever wall presented in
Fig. 1, top wall displacement corresponding to first yield of the outermost vertical
reinforcement defined as yield displacement can be calculated by integration of the curvature
profile distributed along the height of the cantilever wall. On the other hand, displacement
beyond yielding defined as plastic displacement can be calculated by multiplying the plastic
rotation (θp) at the base by the height of the cantilever above the centre of the assumed
equivalent plastic hinge over which uniform curvature value is considered. This top
displacement can be expressed using the following equation:

H 2φ y
Δi = Δ y + Δ p = + (φu − φ y ) L p ( H − 0.5 L p ) (2)
3

where the plastic rotation is presented by the term (φ u − φ y ) L p . This rotation is based on the
assumption that the plastic curvature is uniform over the height of the equivalent plastic
hinge, as shown in Fig. (1). The relationships between the theoretical flexural displacement
ductility, μΔ, and curvature ductility, μφ, for different values of plastic hinge lengths, for
walls with aspect ratio of 2.0, are presented in Fig. 3 using the equations presented in Model
II in the following section.
Fig. 3: Relationship between Displacement and Curvature Ductility

3 CODE QUANTIFICATION OF PLASTIC HINGE LENGTH

Flexural plastic hinging of masonry shear walls is not well quantified (Paulay and Priestley
(1992) and Drysdale and Hamid (2008)) where the scarcity of data related to this subject is
evident from the widely differing, arbitrary, and changing expectations regarding the plastic
hinge length. For this critical feature of seismic response, Paulay and Priestley (1992)
reported equivalent plastic hinge lengths, lp, between 30% to 80% of the wall length, lw,
whereas the USA Masonry Standard Joint Committee code (MSJC 2008 - ACI 530/ASCE 5
/TMS 402) specifies that lp is to be taken equal to lw. The Canadian Standards Association,
CSA Standards S304.1 (CSA 2004) “Design of Masonry Structures” recommends the smaller
of lp = lw/2 or hw/6 for walls with limited ductility but up to lw for shear walls with moderate
ductility, where hw is the wall height. The Federal Emergency Management Agency
“Systematic Rehabilitation” (FEMA-273, (1997)) sets lp equal to half the flexural depth of the
wall cross section but not greater than one story high while the Egyptian Code of Practice
(ECP 2008) specifies the height of the critical zone at wall bases to be greater than hw/6 and
lw but less than 2lw. These ambiguities in estimating a parameter as fundamental as the
equivalent plastic hinge length would lead to inaccurate prediction of the wall displacement
ductility, which is a major factor in determining the wall seismic demand (design force).

4 EQUIVALENT PLASTIC HINGE MODELS


In this review, three different models to represent the equivalent plastic hinge length and/or
the ultimate displacement are presented and compared. Each model will be presented along
with its assumptions and the required parameters in the upcoming sections.

4.1 Model I
The first model investigated, proposed by Kazaz [Kazaz 2012], focuses on evaluating the
length of the equivalent plastic hinge. This analysis is performed under the assumption that
plastic hinge length is sensitive to several variables such as wall length, effective height M/V
axial load ratio (P/Aw fc’) and the amount of horizontal web reinforcement (ρsh). The equation
for the plastic hinge length proposed by Kazaz [2012] is:

0.45
 P  f ρ  M V 
L p = 0.27 Lw 1 − 1 − y sh   (3)
 Lw f c '  fc '  Lw 
Where Lw and M/V are in meters. This model is developed based on the regression analysis of
the aforementioned variables and is also calibrated based on a finite element shear wall model
developed according to the Turkish seismic code [TCS 2007]. It is noteworthy that two
different methods were adopted to evaluate the length of the plastic hinge zone in equation 3.
The first method is based on the ultimate curvature profile calculated from analysing section
strain profile when the outer tension steel reaches strain of 0.06. Such strain value was
selected as the limiting tensile strain of reinforcement for quantification of the ultimate
curvatures at the base of the walls based on strain profiles when using the calibrated finite
element shear wall models. For the second method used, the curvature profile computed,
using the shell element strains introduced in the finite element model, from the element
strains at the two wall ends at the same height in the same row from the analytical model. As
such, vertical strains at the compression and tension boundary ends of the walls were
evaluated from the finite element shear wall model without stating limits to calculate the
curvature based on this limit. These strains are used to calculate the curvature along the
height of the wall (∅=(ɛs- ɛc)/Lw).

4.2 Model II
Paulay and Priestley (1992) introduced an equation to calculate the ultimate displacement
form the displacement ductility for different reinforced concrete walls based on the
equivalent plastic hinge length Lp.
Lp  Lp 
μ Δ = 1 + 3 ( μφ − 1)  1 − 0.5  (4)
Lw  Lw 

where flexural displacement ductility and the theoretical flexural curvature ductility are
μ Δ = ( Δu Δ y ) and μφ = (φu φ y ) respectively. Then from experimental data, the equivalent
plastic hinge length can be determined knowing the theoretical curvature and experimental
displacement ductility. Based on analytical research conducted on reinforced concrete shear
walls, another model is proposed and stated that the equivalent plastic hinge length equals to
20% of wall length plus 7% of the moment shear ratio [Paulay 1992].
M  (5)
L p = 0.2Lw + 0.07  
V 
Priestley et al. (2007) proposed a plastic hinge model that meets specific serviceability and
damage control performance levels. The model includes some factors such as the tension shift
due to shear cracks, strain penetration depth into foundation, as a function of bar diameter,
and the ratio of the ultimate, fu, to yield, fy, strength of flexural reinforcement. The equation
proposed to evaluate Lp is as follows:
L p = k .H + 0.1Lw + L sp (6)

In which the tension shift presents 10% of the wall length; and the ratio of ultimate to yield
f 
strength of bar being represented by k = 0.2  u − 1 ≤ 0.08 . The strain penetration depth,
f 
 y 
Lsp, being affected by the yield strength flexural reinforced and the bar diameter can be
expressed as:
L sp = 0.022f y d b (7)
Furthermore, Priestley et al. (2007) introduced an empirical form for calculating the yield
curvature along the wall height in terms of tensile strain in the reinforcement and the length
of the wall with an associate factor describing the wall material.

4.3 Model III


A mechanics-based analytical model was developed by Shedid and El-Dakhakhni (2014) to
quantify wall top displacement for reinforced masonry walls. In order to facilitate analytical
predictions of wall displacement, curvature ductility, displacement ductility, and other
parameters, the wall experimentally-determined nonlinear load-displacement relationship
should be idealized into a simplified bilinear relationship. Such idealization will also enable
designers and researchers to easily determine critical displacement values used in
characterizing wall response as shown in Fig. 4. Relying on simple section analysis, the
initial yield strength, Q*y, corresponding displacement, Δ*y, as well as the ultimate capacity,
Qu, of a the wall can be easily quantified, and consequently, the yield displacement of the
bilinear system, Δy, can be calculated based on:

Δy=Δ*y(Qu/ Q*y) (8)


However, displacements corresponding to the ultimate and post-peak stages are not as easily
quantifiable due to several factors including the significant variation of curvatures past
yielding, tensile strain penetration in the foundation, and spread of plasticity resulting from
inclined flexure-shear cracking. For a typical wall having a height, Hw, a length, Lw, and
constructed on a reinforced concrete foundation with depth, tRC, while being subjected to a
top load, Qu, as shown in Fig. 4, the bending moment diagram over the height of the wall
corresponding to Qu is linear with a maximum value of Mu at the base and a corresponding
M*y value at a height equal to Hw(1- M*y/Mu) from the base (wall-foundation interface
level). The actual curvature profile over the wall height can also be assumed linear up to first
yield as shown by the dashed line. However, at the ultimate stage, the curvature significantly
deviates from the linear profile. Such deviation, shown in the actual curvature profile in Fig.
4, is primarily due to yielding of reinforcement resulting in an approximately parabolic
curvature profile in the region between M*y and Mu corresponding to curvature levels of Φ*y
and Φu, respectively (Paulay and Priestley, 1992). The extent of reinforcement yield
penetration in the foundation, (represented by L1 in the figure) and the spread of plasticity
due to inclined cracks leading to increase in curvatures (represented by L2 in the figure) also
affect the curvature variation over the wall height. Through quantification of L1 and L2, the
top wall displacement can be determined by double integration of the actual curvature
diagram over wall height. Idealization of the load-displacement should also be conducted to
represent the nonlinear response by a bilinear idealization, where a new value of yield
displacement, Δy, is to be determined. The corresponding yield curvature of the idealized
system which can be determined as shown in the modified curvature diagram of Fig. 4 can be
given by:
Φy,= Φ*y(Qu/ Q*y) (9)
Fig. 4: Model III Plastic Hinge Estimation
Finally, the idealized curvature diagram in Fig. 4 can be used to determine the wall
displacement at yield and at ultimate relying on a linear variation (up to Φy over a height
Hw), a parabolic variation (of [Φu –Φ3] over a height L3) a uniform (constant) distribution
above foundation level (of [Φu -Φy] over a height L2) and another uniform distribution
below the wall/foundation interface level (of Φu over a height L1). For simplification, if one
considers the parabolic distribution to be of the second degree, the top wall displacement at
yield, Δy, and at ultimate load, Δu, can be calculated as follows:
Mu
Δ y = Δ* y
M *y
(10)
 L   L 1  L
( )  2  2 4
( )
Δ u = Δ y + Φu − Φ y L2  H w − 2  + Φu L1  H w + 1  + Φu − Φ3 L3  H w − L2 − 3 
 3
(11)
Mu
Φ y = Φ*y
M *y
(12)
H w − L2
Φ3 = Φ y (13)
Hw

 M *y 
L3 = 1 −  H w − L2 (14)
 Mu 
 
Based on the steel strains presented in Fig. 5, it can be observed that strains of the M10 (db=
11.3 mm) vertical reinforcement up to 170 mm inside the reinforced concrete foundation
were higher than the yield strain regardless of the strain level at the interface between the
foundation and the wall. Whereas, strains slightly below, recorded at 200 mm inside the
foundation, were less than one third of those at 170 mm. It can be inferred that strain
penetration inside the foundation (L1) can be calculated as:

L1=0.03 fy db (15)

Fig. 5: Strain Penetration inside Foundation


Values of L1 (strain penetration), L2 (related to spread of plasticity due to inclined cracks)
need to be quantified. Due to the presence of planes of weaknesses in masonry construction at
the mortar joints, a close observation of the cracking pattern of test results from Shedid et al.
(2008 and 2010) and Bennett and El-Dakhakhni (2011) showed, that shear cracks connect
with the flexural cracks over the lower three to five courses and extend to the entire wall
length. This would result in spread of strains over a height varying from 0.15Lw, as a lower
bound, to 0.25Lw, as an upper bound, as a result of the inclined cracks. In other words, L2
can be given by:
0.15Lw ≤ L2 ≤ 0.25Lw (16)

5 ANALYSIS OF PLASTIC HINGE MODELS


Statistical analysis was performed by Siyam et al. (2015) to evaluate the reliability of the
three models to replicate the experimental data. For this purpose, experimental results of
fifty-six walls extracted from the work reported by Shing et al (1990), Thomson & Wallace
(2004), Shedid et al (2008), Kapoi (2012), Bachmann et al. (2009), and Sherman (2011) were
used.
The walls investigated were all flexure-dominated in order to ensure the formation of the
plastic hinge. The walls had a range of aspect ratios varying from 0.8 to 2.0 and were
subjected to axial compressive stress varying, from 0 (no axial load applied) to 12.8% of the
concrete or masonry compressive strength. Vertical reinforcement ratios for the chosen walls
varied from 0.3% to 1.3%. Yield strength of vertical reinforcement used in the walls varied
from 414 MPa to 675 MPa while the compressive strength ranged from 14.8 MPa to 45.6
MPa. This large difference in the compressive strength could be explained by the fact that the
studied samples included both concrete and masonry walls. Rigorous analysis is conducted
separately to the three models and is assumed to provide a good assessment for the reliability
of these mathematical models based on the selected walls failing in flexure.
The ratio between the predicted ultimate displacements and the experimentally determined
displacements based on Models I, II, III are 0.553, 0.591, and 0.79 respectively. Based on
these values, it can be concluded that Model I and Model II have almost the same ratio but
with a shift from the true value (being equal to one) while Model III present much less shift
from the true value but with larger deviation compared with the other two models as shown in
Fig. 6. It is important to report for all models error measures of central tendency (mean error)
and measures of spread (root mean square error) as these two types of statistics provide
different information. The root means square error is a measure of average deviation,
somewhat similar to standard deviation, but the root mean square error concerned with the
deviation from the true value where the standard deviation is concerned with the deviation
from the mean. The mean error of these models to predict the ultimate displacement
compared with the ultimate extracted from the experimental tests within the investigated
walls are -2.5, -1.8, and 1.3mm for Model I, Model II, and Model III respectively where
model III is the closest to the zero error line but at the same time it is the only model among
the three models that overestimate the ultimate displacement. Also model III has the biggest
root means square error for the data (36%) compared with the other two models (32% Model
I and 30% for Model II).
Fig. 6: Linear regression Model I, Model II, and Model III.

6 SUMMARY
Based on the review, it is apparent that there is no consensus on the quantification of the
equivalent plastic hinge length of reinforced concrete and reinforced masonry structural walls
which may be the most important parameter in inelastic displacement prediction under lateral
loading.
The mechanics-based model reviewed relying on curvature profile after yielding, strain
penetration inside concrete foundation as well as effect of inclined flexure-shear cracking to
determine wall top displacements was found to well predict past experimental results
following the regression analysis reviewed.

7 REFERENCES
[1] Banting, B. and El-Dakhakhni, W. (2012). ”Force- and Displacement-Based Seismic
Performance Parameters for Reinforced Masonry Structural Walls with Boundary
Elements.” J. Struct. Eng., 138(12), 1477–1491.
[2] Chopra, A. (2000). “Dynamics of structures: Theory and application to earthquake
engineering (2nd edition).” Prentice Hall, Inc., Englewood Cliffs, N.J., USA.
[3] Drysdale, R., and Hamid, A. (2008). “Masonry structures- behavior and design.” 3rd
Ed., The Masonry Society, Boulder, Colorado, USA.

[4] Kazaz, İ. (2012). “Analytical Study on Plastic Hinge Length of Structural Walls.”
Journal of Structural Engineering, 139(11), 1938-1950.
[5] Priestley, N., Calvi, G., and Kowalsky, M., (2007). “Displacement-Based Seismic
Design of Structures.” IUSS Press, Pavia, Italy.
[6] Shedid, M., El-Dakhakhni, W., and Drysdale, R. (2010) “Alternative Strategies to
Enhance the Seismic Performance of Reinforced Concrete-Block Shear Wall
Systems” ASCE. J. Struct. Engrg.

Das könnte Ihnen auch gefallen