13 views

Uploaded by senthilcivil

SR Hybrid

save

You are on page 1of 11

**semi-rigid connections
**

Mohammad Razavi, Ali Abolmaali ⁎

Center for Structural Engineering Research and Simulation (CSER), Department of Civil Engineering, The University of Texas at Arlington, Arlington, TX 76019, United States

a b s t r a c t a r t i c l e i n f o

Article history:

Received 14 November 2013

Accepted 22 February 2014

Available online 18 March 2014

Keywords:

Hybrid steel frames

Semi-rigid connections

Earthquake engineering

SAC frames

Reliability analysis

Performance based design

The concept of the hybrid steel frame system, as it pertains to mixtures of fully-rigid and semi-rigid steel connec-

tions used in 20-story SAC frames, is presented herein. Several different patterns and locations of semi-rigid

connection replacements within the frame were examined in order to identify hybrid frames with the best

seismic performance. The effective connection stiffness was identiﬁed by performing a parametric study on the

initial stiffness of the semi-rigid connections. Then, the cyclic behavior of the connections with the most effective

stiffness was obtained, using nonlinear ﬁnite element analysis. Inelastic dynamic analyses were conductedonthe

proposed selected frames by subjecting them to Los Angeles earthquake records characterized as those with 2%

and 10%probabilities of exceedance inﬁfty years. The maximumstory drift for the hybrid frames was determined

and comparedwiththe “life safety” and “collapse prevention” performance limits, as recommended by FEMA356

[12]. The story drift andmember forces for the proposedhybrid frames were comparedwiththose of convention-

al SAC frames without semi-rigid connections. Finally, a reliability analysis, utilizing the collapse margin ratio

presented in FEMA P695, was performed to quantify and compare the collapse performance of the proposed

hybrid frames and conventional rigid frames.

Published by Elsevier Ltd.

1. Introduction

Seismic performance of structural systems has been at the

forefront of research for many years. Occurrences of more than 18

severe earthquakes, with magnitudes of more than 5.8 Richter in

the state of California in the 15-year period between 1979 and

1994, have been reported. The premature brittle failures of welded

connections after the Northridge earthquake of 1994 particularly

motivated researchers to look beyond the conventional design

philosophies

In the past fewdecades, several researchers have introduced new

design concepts and approaches to improve the seismic perfor-

mance of steel structures. These include, but are not limited to, the

introduction of more ductile connections and new lateral resistant

systems, including energy dissipating technologies such as base

isolators, frictional or visco-elastic dampers, and active control

elements.

An innovative seismic design method is the Performance Based

Plastic Design (PBPD) that was introduced and developed by Professor

Goel and his associates at the University of Michigan [14,17]. The

development of this method was in response to shortages in the current

seismic design codes on satisfying a performance objective in a direct

manner. In PBPD, the base shear is calculated by equating the work

done by pushing the structure to a predeﬁned target drift monotonically

to the work done by an equivalent elastic-perfectly plastic single degree

of freedom system. The idealized inelastic response spectra by [19] are

used in this study.

Development of the seismic eccentric braced frames (EBF), intro-

duced by Popov and Englehardt in 1988, was another attempt to

enhance the seismic behavior of steel frames. Well-designed EBFs,

constructed with shear links, provide high elastic stiffness and

strength under low to moderate ground motions, combined with

high ductility and energy dissipation capabilities in severe ground

motions.

Recently, Abolmaali et al. [4] introduced a newlateral resistant steel

moment frame referredto as a “HybridSteel Frame.” This system, which

was the foundation of the researchleading to the work that is being pre-

sented in this manuscript, is based on the concept of introducing energy

dissipating mechanisms in the structural frame systems by targeting

and replacing selected rigid connections withmore ductile energy dissi-

pating semi-rigid connections.

The proposed structural system enhances the seismic behavior of

the steel frames by taking advantage of the inelastic energy dissipative

semi-rigid connections, along with stiffer and higher strength rigid

connections. In the case of low intensity earthquakes, while rigid

Journal of Constructional Steel Research 98 (2014) 1–11

⁎ Corresponding author at: 425 NeddermanHall, 416 Yates St., Box 19308, Arlington, TX

76019, USA.

http://dx.doi.org/10.1016/j.jcsr.2014.02.006

0143-974X/Published by Elsevier Ltd.

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

connections remain elastic, the semi-rigid connections absorb seismic en-

ergy and assist the systemin damping out the base motionaccelerationat

a faster rate. On the other hand, in the case of higher intensity earth-

quakes, this study shows that hybrid frames reduce the risk of collapse

when compared with its corresponding SMFs.

Current design codes have almost eliminated the partially restrained

connections in high seismic zones. However, there are several studies

that show high energy dissipating characteristics of the semi-rigid con-

nections with high stiffness and strength, as reported by Ackroyd and

Gerstle [5], Bjorhovde et al. [10], Astaneh et al. [6], Astaneh-Asl et al.

[8], Astaneh-Asl [7], Kukreti and Abolmaali [16], Abolmaali et al. [1],

Abolmaali et al. [2], and Abolmaali et al. [3].

Seismic behavior of SMFs with semi-rigid connections has been

investigated in several studies, both theoretically and experimentally

[18]. Excessive inter-story drift was a major concern for using semi-

rigid connections in steel frames. These studies showed that when

connection stiffness increases, base shear increases; however, inter-

story drift does not decrease proportionally.

Astaneh et al. [6] and Abolmaali et al. [3] studied the energy dissipa-

tion characteristics of different types of semi-rigid connections and

showed that they are capable of undergoing large inelastic rotation

(in excess of 0.05 rad), given that the connection is designed so that

the angle or plate yielding governs the behavior. In other words, if the

plate or angle thickness is relatively small compared to the bolt diame-

ter, bolt yielding and fractures are prevented, and plate yielding results

in a ductile connection behavior by undergoing large inelastic rotation.

In this research, The Los Angeles 20-story SAC frame is used as a

case study. SAC frame dimensions are presented in [15]. Selected

rigid connections were replaced by ductile semi-rigid connections.

A suit of 20 accelerograms [22], which was developed as a part of

the SAC project for the Los Angeles site, with different frequencies

was applied during inelastic dynamic analyses, and the results

were compared with the corresponding responses of the fully rigid

SAC frames.

A reliability-based analysis utilizing the Collapse Margin Ratio

(CMR), proposed in FEMAp695, was performed to quantify the collapse

performance of the Los Angeles SAC 20-story rigid frames and their

corresponding hybrid frames.

1.1. Nonlinear modeling

The concentrated plastic hinge model was adopted to introduce

nonlinear behavior in beams with rigid connections, beams with semi

rigid connections, and columns of structures. Beams with rigid connec-

tions and columns, as shown in Fig. 1, were modeled as compound

elements consisting of an elastic Bernoulli beam element at the middle

and conﬁned by two plastic hinges and two end-zones that connect the

member to the rigid connections. The assumption of the formation of

plastic hinges at the two ends of beams and columns was adopted

based on the hypothesis that the failure mechanism is governed by

the seismic loading.

The semi-rigidbeamcompounds, as shown inFig. 2, were deﬁned by

replacing the two stiff end zones with two non-linear moment-rotation

semi-rigid hinges in the rigid beam compounds. In this conﬁguration,

plastic hinges and semi-rigid connections are both sources of nonlinear-

ity. However, since the plastic moment of semi-rigid connections is

usually much smaller than the plastic moment of beam sections, the

behavior of the beam compound is governed by the behavior of the

semi rigid connections. In fact, the moment demand in beams cannot

exceed the plastic moment of the semi-rigid connections; therefore, it

will not reach the plastic moment of the beam section.

The nonlinear behavior of plastic hinges is commonly expressed by

presenting their moment-rotation backbone-curves. In this study, the

backbone curve for different beam and column members were con-

structed based on the beam deterioration modeling guideline provided

in ATC-72 [9] as shown in Fig. 3.

The plastic hinge parameters for the beams used in this study are

summarized in Table 1.

Parameters used for modeling of semi-rigid connections will be

presented after the effective connection parameters are identiﬁed.

2. Selection of hybrid frame patterns

For a reasonable placement of semi-rigid connections in a hybrid

frame, it is necessary to investigate the local and global effects of adding

semi-rigid connections to the moment resistance frames' performance.

These effects were investigated as follows.

2.1. Moment redistribution/act as a fuse

Semi-rigid connections act as rotational springs; consequently, they

change the distribution of moments between beams and columns.

Moreover, the plastic moment of semi-rigid connections is generally

Fig. 1. Beam compound with stiff end zone and plastic hinges.

Fig. 2. Beam compound with semi-rigid connections and plastic hinges.

Pre-capping plastic rotation (θ

p

)

Post-capping rotation range (θ

pc

)

Ultimate rotation (θ

u

)

Effective yield strength and rotation (M

y

and θ

y

)

Capping strength and rotation, (M

c

and θ

c

)

Residual strength, M

r

Effective elastic stiffness, K

e

Post yield tangent stiffness, K

p

Effective post capping tangent stiffness, K

pc

Deterioration parameter (Λ)

Fig. 3. Parameters of the monotonic backbone curve of the modiﬁed Ibarra–Krawinkler model.

After ATC72-10.

2 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

less than the plastic moment in their adjacent frame members (beams/

columns); thus, the moment cannot exceed the plastic moment of the

connection, and formation of plastic hinges in the adjacent structural

members is avoided.

A pushover analysis was performed on SAC 3-story rigid and hybrid

frames to investigate the effects of semi-rigid connections on moment

redistribution. Moment demands of the fully rigid 3-story frame and

its corresponding hybrid frame at roof drift of 3% are illustrated in Fig.

4-a and -b, respectively. In this study, a hybrid frame with a diagonal

pattern of semi-rigid connections was selected. Semi-rigid connections

are shown as rotational springs in Fig. 4-b. Moment demands of the

beams andcolumns adjacent tothe semi-rigid connections inthe hybrid

frame are noticeably less thanthe moment demands onthe correspond-

ing rigid frame members, as shown in Fig. 4.

2.2. Base shear reduction

In general, the stiffness of semi-rigid connections is smaller than the

rigid connections. Thus, implementing semi-rigid connections in a steel

frame reduces the overall stiffness of the frame and increases the

frame's fundamental period of the structure. Considering the shape of

the design spectra, as shown in Fig. 5, the softer frame experienced a

a) Moment in Fully rigid Frame (kip-in)

b) Moment in Hybrid Frame (kip-in)

-18364

25358

6994

-30083

25069

20028

-25042

-19342

7729

-25468

-5522

-3840.7

-1681

-12333

10882

-23215

10817

-10817

-12503

23441

12598

-8374

16737

-23450

-7733

23239

15505

1669

10826

12431

10724

21456

-10732

-30138 -24063 -29983 -23901

1602

-1659 1662

1669 -1605

Fig. 4. Bending moment demands at global drift of 3%.

Table 1

Parameters used for modeling plastic hinges in beams.

Section name θy θp θpc My Mu Mr Λ

W21X50 0.001272 0.031515 0.144247 6050 6655 2420 0.911012

W24X62 0.001123 0.027519 0.141778 8415 9256.5 3366 0.925926

W27X84 0.000974 0.022837 0.106695 13,420 14,762 5368 0.725497

W27X94 0.000967 0.022909 0.124601 15,290 16,819 6116 0.864289

W30X108 0.000881 0.021128 0.117753 19,030 20,933 7612 0.840001

W30X99 0.00089 0.020455 0.103769 17,160 18,876 6864 0.731785

3 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

reduced amount of acceleration. Fig. 5 shows that increasing the period

of the structures from t

1

to t

2

decreases the acceleration from Sa (t

1

) to

Sa (t

2

). However, since using semi-rigid connections will reduce the

system's stiffness, the stiffness of semi-rigid connections should be

determined to ﬁnd the minimum base shear, which still satisﬁes the

inter-story drift limit criteria. On the other hand, semi-rigid connections

may be used to shift the natural frequency of structure and avoid

resonance.

2.3. Shifting structures' period and changing their mode shapes

As explained in the previous section, implementing semi-rigid con-

nections in a SMRF reduces the stiffness of the frame; thus, the natural

periods of hybrid frames are higher than the natural period of corre-

sponding rigid frames. The natural mode-shapes of hybrid frames will

consequently differ from the fundamental mode-shapes of the corre-

sponding SMRF. The idea of the hybrid frame began with a simple as-

sumption that if a high-rise building is seen as a single cantilever

beam, theoretically, an indeﬁnite number of mode shapes is considered

for this system. The idealized simple model of rigid frame and its ﬁrst,

second, and third mode shapes are shown in Fig. 6.

Replacing rigid connections with ﬂexible semi-rigid connections at

certain story levels simulates spring development within the beam, as

shown in Fig. 7. These simpliﬁed models explain why the two cases of

rigidandhybridframes have a signiﬁcant difference inseismic behavior,

where the newly formed springs may help to reduce the frame's mode

shapes into two mode shapes, as shown in Fig. 7.

S

p

e

c

t

r

a

l

R

e

s

p

o

n

s

e

A

c

c

e

l

e

r

a

t

i

o

n

,

S

a

(

g

)

Period, T(sec)

SDS

t1 t2

Sa(t1)

Sa(t2)

Fig. 5. Schematic presentation of the Design Response Spectrum (ASCE 7–10).

a) Rigid Frame b) Simplified Model c) First Mode d) Second Mode e) Third Mode

Fig. 6. The simpliﬁed model of a rigid frame.

a) Hybrid Frame b) Simplified Model c) First Mode d) Second Mode

Fig. 7. The simpliﬁed model of a hybrid frame.

4 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

2.4. Increase energy dissipation

Introducing semi-rigid connections in SMRFs helps to reduce the

seismic demands in structures by dissipating seismic energy. In hybrid

frames, seismic energy dissipates in plastic hinges formed in frame

members (Beams/Columns) and in the semi-rigid connections. The

amount of the energy dissipated in each plastic hinge or semi-rigid

connection is found by calculating the area conﬁned by the outer loop

of the moment-rotation hysteresis loop. Fig. 8 shows the locations of

plastic hinge formations, in SAC 20 story frame's members subjected

to LA35 ground motion, by ﬁlled circles.

2.5. Hybrid frame pattern selection

Based on the aforementioned effects of the semi-rigid connections

on the response of steel moment frames, three different approaches

were adopted for placement of semi-rigid connections in the 20-story

hybrid frame. Frames are designated as HSAC20, where the letter “H”

stands for Hybrid and “SAC20” represents the 20 story SAC frame

designed for Los Angeles site.

The ﬁrst approach was based on introducing a rotational spring at

ﬁve middle stories of the frame. This approach was implemented by re-

placing rigid connections with ﬂexible semi-rigid connections in stories

9 to 13 of the SAC 20 story rigid frame, as shown in Fig. 9(a). The newly

formed springs decoupled the earthquake acceleration into two mode

shapes, as is shown in Fig. 7. Although the performance of this system

was tested with initial linear models in Abolmaali et al. [4], in this

study, the pattern was evaluated using a comprehensive nonlinear

model analysis.

The second approach was based on utilizing the semi-rigid connec-

tions as an energy dissipative tool. As shown in Fig. 8, for a SAC 20-

story frame subjected to LA35 ground motion, plastic hinges formed in

beams and columns located in stories 1 through 5. On the other hand,

structural members of stories 6 through 20 remained elastic for the

most part and did not contribute to the inelastic energy dissipation.

The analysis result of the 20-story frame, under other maximum credi-

ble earthquake (MCE) ground motions scaledfor the Los Angles site, has

commonly followed the same pattern of plastic hinge propagation over

the height of frame. These results are presented in Razavi [20]. In the

second approach, semi-rigid connections were placed in the beams,

Fig. 8. Energy dissipated in plastic hinges of the SAC20 frame under LA35.

Rotational Spring

Approach

Energy Dissipation

Approach

Stability

Approach

a) HSAC20-1 b) HSAC20-2 c) HSAC20-3 d) HSAC20-4 e) HSAC20-5

Fig. 9. Hybrid models based on three proposed approaches.

5 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

which remained elastic. Thus, these connections contributed to the en-

ergy dissipation of the frame. This approach resulted in two patterns of

HSAC20-2 and HSAC20-3, as shown in Fig. 9(b) and (c), respectively.

The former pattern had more energy dissipative members; however,

the system was softer. Nevertheless, the HSAC20-2 frame experienced

less acceleration since it was softer and consequently had a larger

period.

The last approachwas based onmaintaining stability of the structure

under strong ground motions. Collapse occurs when plastic hinges form

in all columns of two different stories. Semi-rigid connections, as men-

tioned before, might be used as a fuse to protect their adjacent columns.

This approach aims to protect at least one column in each story level.

Using this approach, two patterns of HSAC20-4 and HSAC20-5, as

shown in Fig. 9(d) and (e), were proposed.

2.6. Evaluation of proposed patterns based on inter-story drift angle

Inter-story drift angle, which is expressed as inter-story drift (δ), di-

vided by the height of the story (h) is known as one of the best measures

of seismic performance at the story level of steel moment resistant frames

(SMF). The story drift is a global parameter since it is related to the global

drift angle (roof drift angle), which is deﬁned as the roof displacement

divided by the height of the roof and consequently to the spectral

displacement demand. It is also a local parameter since it provides a

good estimation of member forces and deformation demands [15].

A major concern with using semi-rigid connections in steel frames is

that it may cause the inter-story drifts to increase beyond acceptable

limits. Although the use of hybrid frames causes a decrease in the initial

stiffness, the ground motions do not act similarly to static lateral loads

on frames. Ground motions exert forces to frames by introducing accel-

eration to stories' mass in story levels. Since semi-rigid connections shift

the period of structures, the amount of acceleration will reduce. On the

other hand, although the initial stiffness of a SMRF is more than the ini-

tial stiffness of its corresponding hybrid frame, the system stiffness of

the frame changes during an earthquake due to yielding in structural

members and nonlinear moment-rotation behavior of semi-rigid con-

nections. Consequently, the stiffness of the hybrid frame can be higher

than the stiffness of the corresponding moment frame during a ground

motion excitation. Thus, all frames are evaluated using an inelastic time

history analysis.

In order to determine the most effective pattern, the ﬁve proposed

hybrid frames were modeled using a comprehensive nonlinear model

and were subjected to the SAC ground motions at the Los Angeles site.

SAC ground motions are categorized into two levels: Design Based

Earthquakes (DBE) and Maximum Credible Earthquakes (MCE), based

on their return period. The models were ﬁrst subjected to the set of 20

DBE records, LA01 to LA21, to be evaluated for the ﬁrst performance

objective, which was to satisfy the Life Safety (LS) performance under

the DBE hazard level. The acceptance criterion for this performance

objective is to maintain an average inter-story drift of less than 2.5%.

The models were then subjected to the set of 20 MCE records, LA21 to

LA40, to be evaluated for the second performance objective, which is

to satisfy the Collapse Prevention (CP) performance under MCE hazard

level. The acceptance criterion for this performance objective is to main-

tain an average inter-story drift of less than 5%.

0

2

4

6

8

10

12

14

16

18

20

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0%

S

t

o

r

y

Drift

HSAC20-1

HSAC20-2

HSAC20-3

HSAC20-4

HSAC20-5

RIGID

C

o

l

l

a

p

s

e

P

r

e

v

e

n

t

i

o

n

L

i

m

i

t

Fig. 11. Average of story drift diagrams for various models of 20-story structure subjected

to L.A. MCE records.

0

5

10

15

20

1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%

S

t

o

r

y

Maximum interstory drift ratio, θ

max

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

C

o

l

l

a

p

s

e

P

r

e

v

e

n

t

i

o

n

L

i

m

i

t

The percentages refer to the

stiffness of the beam with semi

rigid connection with respect to

the stiffness of the beam with

rigid connection

Fig. 12. Average story drift diagrams for HSAC20-4 with various connection stiffness

subjected to LA MCE records.

0

5

10

15

20

1.0% 2.0% 3.0% 4.0% 5.0% 6.0%

S

t

o

r

y

Maximum interstory drift ratio, θ

max

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

C

o

l

l

a

p

s

e

P

r

e

v

e

n

t

i

o

n

L

i

m

i

t

The percentages refer to the

stiffness of the beam with semi

rigid connection with respect to

the stiffness of the beam with

rigid connection

Fig. 13. Average story drift diagrams for HSAC20-5 with various connection stiffness

subjected to LA MCE records.

0

2

4

6

8

10

12

14

16

18

20

0.0% 1.0% 2.0% 3.0% 4.0% 5.0%

S

t

o

r

y

Drift

HSAC20-1

HSAC20-2

HSAC20-3

HSAC20-4

HSAC20-5

RIGID

L

i

f

e

S

a

f

e

t

y

L

i

m

i

t

Fig. 10. Average of story drift diagrams for various models of 20-story structure subjected

to L.A. DBE records.

6 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

Figs. 10 and 11 show the average of all drift results for the ﬁve pro-

posed hybrid frames and the original SAC rigid frame subjected to DBE

and MCE records, respectively. In these graphs, each line shows an aver-

age of story drift ratios, which resulted from the dynamic inelastic time

history analysis of a frame subjected to 20 earthquake records. Thus,

each graph summarizes 120 nonlinear analysis drift results. Fig. 10

shows that all proposed models met the LS drift criteria, except the

HSAC20-1 model. However, HSAC20-4 and HSAC20-5 models have

the minimum amount of drift demand.

Similarly, drift results of the proposed hybrid frames subjected to the

MCE records conﬁrmed the effectiveness of the semi-rigid connection

location pattern of HSAC20-4 and HSAC20-5 frames.

These observations led to selecting patterns used in HSAC20-4 and

HSAC20-5 models as the most effective patterns.

3. Sensitivity study on semi-rigid connections initial stiffness

To determine the effective connectionproperties, a parametric study

was performed on the initial stiffness of semi-rigid connections used in

the selected hybrid frames. These frames were subjected to the Los

Angeles MCE level records, and the average of results is presented for

each frame in Figs. 12 and 13.

As shown in Fig. 3, the beam model consisted of an elastic beam–

column model, two plastic hinges, and two semi-rigid connections.

The ﬂexural rigidity of a beam is a function of the ﬂexural rigidity of

its components. The elastic beam–column member and the semi-rigid

connection can be considered as two rotational springs. The

compound's stiffness for this system of series springs is obtained by:

K

compound

¼

1

1

K

Beam

þ

1

K

Connection

¼

K

Beam

Ã K

Connection

K

Beam

þ K

Connection

: ð1Þ

Therefore, for connection stiffness equal to inﬁnity, the stiffness of

the compound is equal to the stiffness of the beam. On the other limit,

for a connection stiffness equal to zero, the compound's ﬂexural stiff-

ness equals to zero. Based on Eq. (1), the connections' initial stiffness

is back calculated for a total component stiffness of 10% to 100% of the

Column Beam

2

Top Angle Bolted

Seat Angle Bolted

15.6

3/4

d

4

Column

2

1/4

4

4

Weld

4

8.3

1/4

3/4

4

3/4

2

1/4

15

3

Web Angle

Fig. 14. Typical sketch of top- and seat-angle with double web angle connections.

Cyclic Load

Connection

Rigid End-Cap

Fig. 15. Top- and seat-angle with double web angle ﬁnite element model, and mesh properties.

7 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

stiffness of the beam. For an elastic beam–column member subjected

to double curvature bending, the initial stiffness of the beam equals to

6EI/L, where “E” is the modulus of elasticity of steel material, “I” is the

moment of inertia of the beam's cross section, and “L” is the length of

the beam. For the purpose of parametric study, the compound stiffness

is varied between 10 and 100% of the beams' stiffness. Then, the frames

are subjected to the MCE ground motions. Finally, the drift demands are

compared for selection of the best connections' initial stiffness.

Average drift diagrams for HSAC20-4 and HSAC20-5 frames with

various connection stiffness values subjected to LA MCE records are

illustrated in Figs. 12 and 13, respectively. These two graphs show that

in both semi-rigid patterns used for the 20-story frame, there is an

inverse relationship between connection stiffness and the maximum

drift. Further study on the frame drift demands subjected to single

earthquake records indicated that for ground motions with smaller ac-

celeration intensity, such as LA21, the greater connection stiffness

corresponded to the less drift demands. However, when frames were

subjected to a high intensity earthquake records, such as LA30, the

greater connection stiffness caused more drift demands. Story drift

ratio plots for individual ground motions are presented in Razavi [20].

Based on the presented study, a connection with stiffness of 30% to

50% of the stiffness of the beam was selected for implementation in

hybrid frames. Average story drift results support the hypothesis that

using connections with stiffness in this range reduces the probability

of collapse in MCE excitations; on the other hand, in the case of DBE

excitations, the system is stiff enough to meet the life safety criterion.

4. Semi-rigid connections

In general, bolted–bolted or welded–bolted connections with slip

critical bolts, which are pre-tensioned to 70% of their minimum tensile

strength, are known as semi-rigidconnections. Static and dynamic char-

acteristics of semi-rigid connections are categorized by their moment

rotation (M-θ) curves and hysteresis loops, respectively.

For the purpose of selection of the type of semi-rigid connection, re-

sults of several experiments conducted by Abolmaali et al. [3] and Chen

and Kishi [11] were studied. Based on the effective initial stiffness and

ductility of the semi-rigid connections, top-and seat-angle with double

web angle connections were found to be the best match for the purpose

of this study. The hysteresis behavior of beam–columnsteel connections

was predicted by means of a 3D non-linear ﬁnite element analysis

(FEA). Incremental nonlinear analysis takes into account material,

geometry, and contact nonlinearities in predicting moment-rotation

hysteresis loops. The results obtained from the ﬁnite element analyses

were validated by a series of full-scale structural tests performed by

Abolmaali et al. [3]. Modeling details are discussed in detail in Razavi

et al. [21] and Razavi [20]. Finally, semi-rigid connections used in this

study for different beam sizes were simulated. A bilinear curve was

ﬁtted to the moment-rotation hysteresis loop of each connection.

These curves were then used in global modeling of the hybrid frames.

A typical sketch of the top-and seat-angle with double web angle

connections selected for this study is shown in Fig. 14. A 3-D ﬁnite

element model of the top-and seat-angle with double web angle is

presented in Fig. 15. Fig. 16 shows the hysteresis loops for top-and

seat-angle with double web angle with W21X50 and the bilinear ﬁtted

curve.

The connections properties are also summarized in Table 2. Nota-

tions are explained in Fig. 3.

5. Force demands

To compare the shear and bending moments of the HSAC versus the

original rigid SAC frames, the ratios of the absolute value of maximum

forces for the HSAC were normalized with respect to the forces of the

SAC frame. Thus, a ratio of unity indicates that the HSAC forces are

equal to the rigid SAC forces, and the ratio of less than unity implies a

reduction in a given force for the hybrid frame (lower demand).

As an example, the beam's and column's moment ratios for the H-

SAC20-5 to the 20-story rigid frames subjected to two sets of ground

motions are plotted in Figs. 17 to 18. Each point in these graphs corre-

sponds to a normalized demand value, and the solid line shows the

-226

-169.5

-113

-56.5

0

56.5

113

169.5

226

-2000

-1500

-1000

-500

0

500

1000

1500

2000

-6%-5%-4%-3%-2%-1% 0% 1% 2% 3% 4% 5% 6%

M

o

m

e

n

t

(

k

i

p

.

i

n

)

Drift (%)

(

k

N

.

m

)

Fig. 16. FEMhysteresis loops for top- andseat-angle with double web angle with W21X50.

Table 2

Semi-rigid connections properties.

Beam Size Θ

y

M

y

(kip-in) Θ

u

M

u

(kip-in) K

e

K

p

W21X50 0.003 1032 0.05 1750 344,000 15,277

W24X62 0.003 1203 0.05 2000 401,000 16,957

W27X84 0.003 1411 0.05 2250 470,333 17,851

W30X99 0.003 1584 0.05 2450 528,000 18,426

W33X141 0.003 1920 0.05 2900 640,000 20,851

W36X150 0.003 2080 0.05 3300 693,333 25,957

0

5

10

15

20

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Ratio

LA21 LA22

LA23 LA24

LA25 LA26

LA27 LA28

LA29 LA30

LA31 LA32

LA33 LA34

LA35 LA36

LA37 LA38

LA39 LA40

Fig. 17. The ratios of the moment in beams for the 20-Story HSAC20-5/SAC subjected to MCE.

8 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

average in each story level. In addition, the average values of the mem-

bers' moment and shear ratios are tabulated in Table 3. On average, the

member forces are reduced between 6 and 16%.

6. Reliability analysis

This section focuses on evaluating and quantifying the collapse

performance of the Los Angeles SAC 20-story rigid frame and the pro-

posed hybrid frames by application of the FEMAP695 reliability analysis

methodology. An Incremental Dynamic Analysis (IDA) was performed

on the comprehensive nonlinear models of the rigid frame and its two

corresponding hybrid frames. The IDA was performed by applying the

20 records of the set of MCE, 2% probability of exceedance in 50 years,

ground motions. Finally, the collapse performance of the frames was

quantiﬁed and compared by computing the Collapse Margin Ratio

(CMR) value.

Incremental Dynamic Analysis (IDA), also known as Dynamic Push

Over (DPO), is an analysis method developed by Vamvatsikos and

Cornell [23] which aims at determining the global capacity of structures.

In this method, structures are subjected to one or more ground mo-

tion(s), each scaled to multiple levels of intensity Measure (IM). An IM

is a non-negative parameter, such as Peak Ground Acceleration (PGA)

or Spectral Acceleration (Sa), which represents the ground motion's

intensity. Then, the structure is analyzed under each scaled ground mo-

tion, utilizing a nonlinear dynamic analysis, and the Damage Measure

(DM) of interest is recorded. A DM is a measurable response of a struc-

ture such as ductility, global drift, or inter-story drift that is an output of

the nonlinear dynamic analysis under the prescribed seismic loading.

The smallest scale factor is selected to ensure an elastic response of

the structure, then the scale factor increases until the collapse limit

state is reached. The scale factor increment should be small enough to

capture the collapse point. Finally, a graph of DMs versus IMs is plotted

which also called IDA curve. In this study, the 5% damped ﬁrst mode

spectral acceleration (Sa (T1, 5%)) and the inter-story drift angle

(θ

max

) were selected as the IM and DM parameters, respectively.

A fragility function for collapse limit-state expresses the probability

of exceeding the limit state under a given ground motion with a certain

level of intensity. The fragility function is produced by using collapse

data from IDA results through a cumulative distribution function

(CDF) that expresses the probability of collapse as a function of ground

motion intensity (Ibarra et al., 2002). For this purpose, a lognormal dis-

tribution is ﬁtted to the collapse data obtained from IDA results. While

for a structure, the IDAis performed and the fragility curve is developed,

the median collapse intensity, SCT, should be identiﬁed. The lowest

intensity at which one-half of the records cause collapse is the median

collapse intensity, S

CT

[13]. In turn, this value is the spectral acceleration

corresponding to the 50% probability of collapse in the fragility curve.

The S

CT

value is a representative of the capacity of the structure. On

the other hand, the MCE intensity, S

MT

, which is deﬁned as the median

5%-damped spectral acceleration of the MCE ground motions at the

fundamental period of the structure, is a representative of the plausible

demands applied to a structure.

To quantify the collapse performance of steel frames, FEMA P695

introduced the collapse margin ratio, CMR, which is the ratio of the

S

CT

to the S

MT

, as shown in Eq. (2).

Collapse Margin Ratio CMR ¼

S

CT

S

MT

ð2Þ

Indeed a bigger CMR corresponds to a less probability of collapse.

The collapse point of a frame for a particular ground motion is

deﬁned as the lowest value of the following criteria: a) the point

where the slope of the IDA curve falls below 20% of the initial slope of

the curve, and b) the upper-bond inter-story drifts capacity of 10%.

The IDA curves and their corresponding fragility curves for different

20-story frames are shown in Fig. 19. Hollow circles in the IDA curves

correspond to the collapse points in structure. Moreover, each black

dot in the fragility curve corresponds to a cumulative probability of

collapse obtained from the IDA curves. The collapse margin ratio for

different SAC rigid frames and the proposed hybrid frames are shown

in Table 4. A larger CMR corresponds to a better performance.

Fig. 20(a) and (b) shows the ﬁnal deformed shape of SAC-20 story

rigid and the H-SAC20-5 frames subjected to LA36 ground motion, re-

spectively. The zigzag line in Fig. 20(b) was obtained by connecting

the locations of semi-rigid connections. As previously shown in Fig. 8,

formation of plastic hinges in stories 1 through 5 of the rigid frame

caused collapse in the original SAC frame. Although, the H-SAC20-5

has also shown excessive deformation, this structure did not collapse

when subjected to the identical earthquake record.

7. Conclusion

In this study, a new lateral resistant system called hybrid frame,

which is a combination of semi-rigid and fully rigid connections in

Table 3

Average hybrid/SAC demands response ratio.

Number of stories Model name Under DBE records Under MCE records

Beams' moment HSAC20-4 0.88 0.88

HSAC20-5 0.84 0.85

Beams' shear HSAC20-4 0.87 0.88

HSAC20-5 0.84 0.85

Columns' moment HSAC20-4 0.92 0.94

HSAC20-5 0.86 0.91

Columns' shear HSAC20-4 0.90 0.92

HSAC20-5 0.85 0.89

0

5

10

15

20

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

S

t

o

r

y

Ratio

LA21 LA22

LA23 LA24

LA25 LA26

LA27 LA28

LA29 LA30

LA31 LA32

LA33 LA34

LA35 LA36

LA37 LA38

LA39 LA40

Fig. 18. The ratios of the moment in columns for the 20-Story HSAC20-5/SAC subjected to MCE.

9 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

steel frames, was introduced to improve the performance of moment

resistant steel frames subjected to seismic excitations. The 20-story

SAC frame for the Los Angeles site was used as benchmark in this

study. Hybrid frames were designed by replacing selected fully-

rigid connections with more ﬂexible semi-rigid connections in the

SAC frame.

The effective semi-rigid connection patterns and the effective

connection stiffness were identiﬁed. Performance of the 20-story SAC

a) IDA Curves for Model SAC Rigid 20-Story b) Fragility Curve for Model SAC Rigid 20-Story

c) IDA Curves for Model HSAC20-4 d) Fragility Curve for Model HSAC20-4

e) IDA Curves for Model HSAC20-5 f) Fragility Curve for Model HSAC20-5

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0% 5% 10% 15%

S

a

(

T

1

,

5

%

)

/

g

Maximum interstory drift ratio, θ

max

Maximum interstory drift ratio, θ

max

Maximum interstory drift ratio, θ

max

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.2 0.4 0.6 0.8 1 1.2 1.4

P

r

o

b

a

b

i

l

i

t

y

o

f

C

o

l

l

a

p

s

e

Sa(T1, 5%) / g

0

0.2

0.4

0.6

0.8

1

1.2

0% 5% 10% 15%

S

a

(

T

1

,

5

%

)

/

g

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.2 0.4 0.6 0.8 1 1.2 1.4

P

r

o

b

a

b

i

l

i

t

y

o

f

C

o

l

l

a

p

s

e

Sa(T1, 5%) / g

0

0.2

0.4

0.6

0.8

1

1.2

0% 5% 10% 15%

S

a

(

T

1

,

5

%

)

/

g

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.2 0.4 0.6 0.8 1 1.2 1.4

P

r

o

b

a

b

i

l

i

t

y

o

f

C

o

l

l

a

p

s

e

Sa(T1, 5%) / g

Fig. 19. IDA and fragility curves for different 20-story models.

10 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

rigid and its corresponding hybrid frames were evaluated using the

reliability analysis presented in FEMA P695.

Comparison of performance of the 20-story SAC rigid frames with its

corresponding hybrid frames showed a superior performance for the

hybrid frame. Collapse Margin Ratio of the investigated hybrid frames

was between 20 and 33% better than the performance of the corre-

sponding rigid frame. This, in turn, reduced the probability of collapse

when the frame was subjected to severe ground motions.

The force demands in the beams and columns of the proposed 20-

story hybrid frames were reduced up to 16%. Considering the fact that

the structural members' (beams and columns) sections in the hybrid

frames were the same as those of the corresponding rigid frames, the

proposed frame system performed superior to conventional SMFs

when subjected to earthquake excitation.

References

[1] Abolmaali A, Kukreti AR, Razavi SH. Hysteresis behavior of semi-rigid double web

angle steel connections. J Constr Steel Res 2003;59:1057–82.

[2] Abolmaali A, Matthys JH, Farooqi M, Choi Y. Development of moment-rotation

model equations for ﬂush end-plate connections. J Constr Steel Res 2005;61:

1595–612.

[3] Abolmaali A, Kukreti AR, Motahari A, Ghassemieh M. Energy dissipation characteris-

tics of semi-rigid connection. J Constr Steel Res 2009;65:19–27.

[4] Abolmaali A, Razavi M, Radulova D. On the concept of earthquake resistant hybrid

steel frames. J Constr Steel Res 2012;68:34–42.

[5] Ackroyd MH, Gerstle KH. Behavior of type 2 steel frames. J Struct Div ASCE

1982;108(ST7):1541–58.

[6] Astaneh A, Nader MN, Malik L. Cyclic behavior of double angle connections. J Struct

Eng ASCE 1989;115(5):1101–18.

[7] Astaneh-Asl A. Seismic design of bolted steel moment-resisting frames. Moraga, CA

Steel Tips, Structural Steel Education Council; 1995 [82 pp.].

[8] Astaneh-Asl A, Nader MN, Harriott JD. Seismic behavior and design considerations in

semi-rigid frames. Proceedings of the 1991 national steel construction conference.

AISC; 1991.

[9] ATC-72. Guidelines on modeling and acceptance criteria for seismic design and anal-

ysis of tall buildings. Redwood City, CA Applied Technology Council; 2010 [ATC-72-

1].

[10] Bjorhovde R, Colson A, Brozzetti J. Classiﬁcation systemfor beam-to-column connec-

tions. J Struct Eng ASCE 1990;116(11).

[11] Chen Wai-Fah, Kishi N. Semi-rigid steel beam-to-column connections: data base and

modeling. J Struct Eng 1989;115(1):105–19.

[12] FEMA. Pre-standard and commentary for the seismic rehabilitation of buildings, re-

port FEMA-356. Washington, DC The SAC Joint Venture for the Federal Emergency

Management Agency; 2000.

[13] FEMA P695. Quantiﬁcation of building seismic performance factors. Washington,

D.C. Federal Emergency Management Agency; 2009

[14] Goel Subhash C, et al. Performance‐based plastic design (PBPD) method for

earthquake‐resistant structures: an overview. The, structural design of tall and

special buildings, 19.1‐2. , ; 2010. p. 115–37.

[15] Gupta A, Krawinkler H. Seismic demands for performance evaluation of steel

moment resisting frame structures (SAC Task 5.4.3). Report no. 132. CA John A.

Blume Earthquake Engineering Center, Stanford University; 1999.

[16] Kukreti AR, Abolmaali A. Moment-rotation hysteresis behavior of top and seat angle

steel frame connections. J Struct Eng ASCE 1999;125(8):810–20.

[17] Leelataviwat S, Goel SC, Stojadinovic' B. Toward performance-based seismic design

of structures. Earthq Spectra 1999;15(3):435–61.

[18] Maison Bruce F, et al. Performance of PR moment frame buildings in UBC seismic

zones 3 and 4. J Struct Eng 2000;126(1):108–16.

[19] Newmark NM, Hall WJ. Earthquake spectra and design. El Cerrito, California

Earthquake Engineering Research Institute; 1982.

[20] Razavi M. On the concept of earthquake resistant hybrid steel frames. A Dissertation

submitted as partial fulﬁllment of the requirement for degree of Ph.D. Arlington,

Texas Department of Civil Engineering, University of Texas at Arlington; 2013.

[21] Razavi M, Kiamanesh R, Abolmaali A. “Computational benchmarks in simulation of

cyclic performance of steel connections using three-dimensional nonlinear ﬁnite el-

ement method”, i-manager's. J Struct Eng September —November 2012;1(3):15–25.

[22] Somerville Paul G, et al. Modiﬁcation of empirical strong ground motion attenuation

relations to include the amplitude and duration effects of rupture directivity.

Seismol Res Lett 1997;68(1):199–222.

[23] Vamvatsikos Dimitrios, Allin Cornell C. Incremental dynamic analysis. Earthq Eng

Struct Dyn 2002;31.3:491–514.

Table 4

Collapse margin ratio for different frame models.

Structure type CMR

Rigid 1.48

HSAC20-4 1.80

HSAC20-5 1.99

a) Rigid SAC 20- Story b) H-SAC20-5

Fig. 20. The ﬁnal deformed shapes of the SAC20-story and HSAC20-5 frames subjected to

LA36 ground motion.

11 M. Razavi, A. Abolmaali / Journal of Constructional Steel Research 98 (2014) 1–11

- Structure B FORMUploaded bySujan Singh
- IRJET-Seismic Risk Assessment of RC BridgeUploaded byIRJET Journal
- Building 7 TesUploaded byodjan21_691446
- p751_ch1Uploaded byKy Visoth Sambath
- WCEE2012_3417Uploaded byMarcrft123
- C Earthquake EngineeringUploaded byMuhammad Shafiq
- Thermal Eng LabUploaded byPau Miralles Ferras
- Beam Design for Moment, Shear & TorsionUploaded bysidiq7
- ASCE 7 Seismic Loads SRRUploaded byJose Rojas-Guzman
- FrameUploaded bynatasha
- Coupling BeamUploaded by''Jeff Leong Manzanillo
- Sattar,Liel Seismic Performance of Infilled RC Frames FINALUploaded byChristian Alcedo Santi
- 2Uploaded byalaluf
- Effect of Structural Period and Ground Motion Parameters on the Earthquake Response of Asymmetric Buildings 1992Uploaded bycisco
- EN(450)Uploaded byreacharunk
- Mom III Module 18Uploaded byWolf Lord
- 2 AbstractUploaded bydhanaraj
- Bgscm13 SampleUploaded byGustavoYippeekayyayGus
- _3ff660b50abf55cd3d0a5983f6f398cb_MOM-III---Module-19Uploaded byHatnusen
- Rigid Head PilesUploaded byPrerak Agrawal
- RPP BFC 32102 Sem II 2014-2015Uploaded byTikar Mengkuang
- Concrete Example - SIUploaded bysanusi69
- LECTURE 7 - Design for ShearUploaded byHowell Tungol
- SAMBUNGAN APX-WF150Uploaded byPuji Kurniawan
- 1-s2.0-S1687404814000625-main.pdfUploaded byshailesh goral
- Bending StressUploaded byShaktivell Letchumanan
- asadsdfUploaded byJOSE
- second moment of inertiaUploaded byapi-219509070
- 4) 10CV3 - June 2013Uploaded byRadhaAnanthalekshmi
- RMI_Verif_NAFEMS_3_2.pdfUploaded byFernando Martinez

- Hand book for nuclear physicsUploaded byMd. Enamul Hoque
- ATC-40Uploaded byAZ314
- Structural SectionsUploaded bysenthilcivil
- 13 (a) Public Service CommissionUploaded byhadhai
- Steel Seismic DesignUploaded bysenthilcivil
- 164.THUploaded bysenthilcivil
- SeminarUploaded bysenthilcivil
- Guidebook on Non-Destructive TestingUploaded byMaurice Balkissoon
- SemiRigid Portal FrameUploaded bysenthilcivil
- Shortcut VirusUploaded bysenthilcivil
- TutorialsUploaded bysenthilcivil
- Performance of Steel BuildingsUploaded bysenthilcivil
- Beam ElementUploaded bysenthilcivil
- FEBeamNotesUploaded byAymaneng
- SAP ManualUploaded by76027074
- ObcUploaded bysenthilcivil
- Camera Work Shot LanguageUploaded byuks
- Integration FormulasUploaded byJoyen Sanjana
- SemiRigidUploaded bysenthilcivil
- Theory PTUploaded byimzee25
- Gyana PadalgalUploaded byapi-3728784
- Corrections IS800 2007Uploaded byVarun Kumar Munjal
- EC8 Seismic Design of Buildings - Worked examplesUploaded byMarco Fasan
- EC8 Seismic Design of Buildings - Worked examplesUploaded byMarco Fasan

- HIS Implementation GuideUploaded byJohn stevenson
- WWI Combatant Models by IntroductionUploaded byHarrierPotter
- ArturoToxtlePresentacionUploaded byDias Académicos Sharing Sessions
- KB3425L_4425LFINALUploaded byAppliance799
- (Cromatografia de Troca Iônica)Uploaded byJéssica Borges
- RefrigerationUploaded byZariv Torres
- Advanced Water Distribn Mod & MgtUploaded byCADsync
- 2. Fire Fighting Equipments_systemsUploaded bySam
- Ergonomia OrganizacionalUploaded bykarina
- CET111 - Lista de Exercícios - Primeira AvaliaçãoUploaded byIuri Carvalho
- Bombeo eolico1Uploaded byHugo Petit
- Tablas Para El Anclaje de Barras CorrugadasUploaded byDavid Fernandez
- rapport-de-stage.pdfUploaded byOussama Moudaoui
- FileUploaded bymbare80
- netionicequations-110220150937-phpapp01Uploaded byjkcfern
- BS-598-104-1989.pdfUploaded byEnd End
- Parcial Final GerenciaUploaded byfacundo cabral
- CLASE 2Uploaded byedwin franklin
- LosasUploaded byedgar
- Rubber Band OrnithopterUploaded byMohammad Salman
- 07 Fdn Repairs - UnderpinningUploaded byNasir Halimi
- Engenharia ClinicaUploaded byLucas de Melo
- RadiodifusãoUploaded byFabiano Meneghin
- DBA Question From OrafaqUploaded bymail2prb
- Trabajo 738-748 FinalUploaded byUnidad de Logística de la Evaluación CLMetropolitano
- Bgs Mx 003isb1Uploaded byrobertotoccaceliblas
- AUTOSAR_SWS_IO_HWAbstraction.pdfUploaded byhoaln
- BS STANDARDSUploaded bytakias
- Cronología de La Historia Concreto y CementoUploaded byEdward Yucra
- Day1 Scala Crash CourseUploaded byLokesh Dikshi