seismic retrofit of typical iranian steel buildings...

2008 FRF’s International Student Paper Competition in Honor of Professor Mehdi Ghalibafian, PhD, SE B. Shafei, M. Motavalli, S.R. Mirghaderi and P. Lestuzzi

1 Ph.D. Candidate, Department of Civil and Environmental Engineering, The Henry Samueli School of Engineering, University of California, Irvine (UCI), CA 92697, United States, e-mail: [email protected] 2 Assistant Professor, Department of Civil Engineering, Faculty of Engineering, University of Tehran, Tehran, Iran, and Head of Structural Engineering Research Lab, Swiss Federal Laboratories for Materials Testing and Research EMPA, Dubendorf, Switzerland, e-mail: [email protected] 3 Assistant Professor, Department of Civil Engineering, Faculty of Engineering, University of Tehran, Tehran, Iran, e-mail: [email protected] 4 Lecturer, IS-Institute of Structural Engineering, Ecole Polytechnique Fédérale de Lausanne EPFL, Lausanne, Switzerland, e-mail: [email protected]

SEISMIC RETROFIT OF TYPICAL IRANIAN STEEL BUILDINGS USING A NEW APPROACH FOR UPGRADING THE EXISTING CONNECTIONS

BEHROUZ SHAFEI1, MASOUD MOTAVALLI2, S.RASOUL MIRGHADERI3, AND PIERINO LESTUZZI4

ABSTRACT The construction of a wide range of typical Iranian steel buildings a large number of which were built before 1990 was based on no seismic code. According to the experiences of past earthquakes, especially Manjil Earthquake (1990) and Bam Earthquake (2003), it has been confirmed that many of these steel buildings may suffer extensive damage while earthquakes. Considering that some of these buildings are critical facilities that must remain operational following earthquakes, serious concerns are raised about their performance in future earthquakes. Observation of damaged Iranian steel buildings shows that most of them have failed due to their joint failure. The beam-to-column connections of these buildings follow the practice of Iranian typical connection named as “Khorjini connection”. In this study, following the evaluation of existing Khorjini connections, a connection upgrading method which significantly improves the seismic performance of these connections is proposed. In order to demonstrate the effectiveness of the proposed method, three Iranian common steel buildings that can be representative of low-, medium-, and high-rise buildings have been analyzed by linear static, nonlinear static (pushover), and nonlinear dynamic (time history) methods. Results obtained from these analyses also approve the success of suggested method in upgrading the seismic behavior of Iranian common steel buildings. Key words: Seismic Evaluation, Steel Building, Seismic Retrofit, Connection Upgrading

1. INTRODUCTION Iran is located in a high-seismic area, and experiences many earthquakes every year. Unfortunately, the construction of a wide range of Iranian steel buildings a large number of which were built before 1990 was based on no seismic code. These buildings can be seen from faraway rural areas to large cities of Iran, and are used widely as residential, commercial, and also administrative buildings. Based on the experiences of past earthquakes like Manjil Earthquake (1990) and Bam Earthquake (2003), it has been confirmed that many of these steel buildings may suffer extensive damage while earthquakes. Considering that some of these buildings are critical facilities like hospitals and schools that must remain operational following earthquakes, serious concerns are raised about their performance in future earthquakes.

2 out of 12

This category of Iranian steel buildings has obvious similarities in construction practices, and most of these common buildings suffer from the same deficiencies in their lateral load bearing systems. These buildings usually have the framing system in only one direction, and this framing system can just carry gravity loads. The beam-to-column connections of these frames often follow the practice of typical Iranian connection named as “Khorjini connection” which can be categorized as non-seismic semi-rigid connections. In most of subject buildings, the unidirectional frames have been joined together in perpendicular direction by beams using hinged connections. Because of non-rigidity of connections as well as non-existence of bracing systems in existing frames, typical Iranian steel buildings which belong to the study category show limited strength against lateral loads, and they are not capable of resisting moderate or large earthquakes. 2. SEISMIC PERFORMANCE OF TYPICAL IRANIAN STEEL CONNECTIONS Based on diverse research projects which have been conducted by Iranian authorities and academics (Tahooni 1990, Ghafory 1995, Moghaddam and Koohian 1994, Arbabi 1998, and Dehghani 2001), the main weak point of typical Iranian steel structures is located in their connections. These connections usually suffer from execution problems plus glaring mistakes in design and calculation. If the connection deficiencies are found out, and suitable strengthening methods are proposed, the seismic performance of these common buildings will be upgraded significantly. The typical Iranian steel connection named as “Khorjini Connection” is considered carefully in this section, and its seismic deficiencies are discussed comprehensively. Then, the authors’ proposed method of improving the seismic performance of this category of connections will be discussed. 2.1. An Introduction to Khorjini Connections Beam-to-column connections in majority of typical Iranian steel buildings built before 1990 are Khorjini connections. Figure 1 depicts two common details of typical Iranian Khorjini connections.

Figure 1: Two common details of typical Iranian Khorjini connections [Dehghani 2001, Photo: B.Shafei]

Steel frames with Khorjini connections are one of the most common structural systems in Iran. In this system, a pair of continuous beams passes along two sides of several columns, and each beam is placed on a seat angle welded to the column. Another angle connects the top flange of beam to the side of column. In some cases, vertical brackets may also be used inside the top and bottom angle sections in order to increase the load bearing capacity. The columns often consist of double rolled profiles joined together by tie plates. At the beam-to-column connection, a long plate is welded to the face of column, and both top and bottom angles are placed on this plate. The top angle is usually smaller than the bottom angle in size, and is also shorter. Figure 2 shows two real structures in which Khorjini connections have been used. All these connections are almost the same in configuration, similar to Figure 2. The only difference which can be observed in different frames is the length and the size of angle sections that vary according to connection design loads.

3 out of 12

Khorjini connections which have been developed in the past 50 years by practicing engineers of Iran are used extensively because of their simplicity and economic advantages. Construction using this type of connections saves on erection time and labor cost. Furthermore, because of limitations on the availability and the cost of deep rolled sections, Khorjini connections provide the possibility of using two parallel beams instead of one deeper beam (Tehranizadeh 2000 and 2001).

Figure 2: Two instances of existing structures having Khorjini connections [Photo: B.Shafei] 2.2. Seismic Deficiencies of Khorjini Connections In the absence of the bracing system in subject buildings, the seismic performance of existing frames could be satisfying only if the beam-to-column connections were able to provide required rigidity, but common Khorjini connections show low rigidity. The configuration of this type of connections is such that only two angle sections are located on the top and bottom flanges of crossing beams. Theoretical and experimental researches (Moghaddam and Koohian 1994, Dehghani 2001, and Pirouzanfar 2005) all demonstrate that these two angles are not capable of providing the expected strength and stiffness requirements of a rigid connection. As a result, the behavior of connection under cyclic loads is poor and vulnerable. The poor behavior of connections is emphatically contrary to the seismic code regulations which accept no failure in connections during the seismic behavior (BHRC 2005). Furthermore, connections should be detailed in a manner that they provide a safe load path for transferring the loads from beams to columns. An essential point regarding the reliable load path is that all connection components should participate properly in the load transfer with no interruption (Bruneau et al. 1998). But in Khorjini connections, just two angle sections welded to beams and columns (with poor quality welds in most cases) have the role of load transferring. Poor behavior of connection components which are not able to provide a reliable load path causes the failure of connection in first few cycles of loading. In order to obviate this important structural fault, the connection should be upgraded such that it can provide the reliable load path as well as desirable strength and stiffness. 3. PROPOSED METHOD FOR UPGRADING THE CONNECTIONS By the method proposed in the current research study, the Khorjini connection is upgraded in a manner that the load path changes from existing angle sections to new additional connection elements further to provide a rigid connection. Subsequently, the stress level in angle sections decreases significantly, and this results in preventing the brittle failure of the connection. In order to provide the expected seismic behavior, the connection is retrofitted by adding two vertical plates hereafter called “R-plates” perpendicular to the frame direction. These plates are placed between two beam profiles, and their edges are welded to beam and column flanges. Figure 3 shows the general shape of an R-plate plus its installation on an upgraded Khorjini connection. Both horizontal and vertical loads as

4 out of 12

well as flexural moments are transferred from beams to columns through these added plates instead of existing angle sections. R-plates lock the rotation of connection, and satisfy required strength and stiffness by connecting to beams by fully penetration welds, and to columns by fillet welds.

Figure 3: Typical existing Khorjini connection before upgrading (Left), proposed vertical plate (R-plate) to be added to both sides of the connection (Middle), and the upgraded Khorjini connection (Right)

According to a comprehensive experimental and analytical research conducted on structural behavior of both upgraded Khorjini connections and new seismic-rigid connections of continuous beams to the column, the proposed connections have demonstrated the acceptable behavior while cyclic loading. It can be inferred from experimental results that all experimented specimens sustained the interstory drift angle greater than 0.08rad without any significant loss of strength, which is far in excess of the requirements for a beam-to-column connection given by the latest seismic codes. The experimental and analytical results show that moment capacity of these connections is more than bending resistance of connected beams. Therefore, the structural ductility is controlled by the flexural behavior of beam ends; and the total behavior of connections and also of column might remain reasonably elastic. Hence, the upgraded frame can be categorized as a ductile system which has high energy dissipating capacity as well as stable hysteretic loops. Showing expected seismic performance, the frame can safely be used as a special or ordinary moment resisting frame. For the purpose of getting detailed information regarding the structural behavior of upgraded connection, authors would recommend that the reader refer to a related paper published by the second co-author of this paper in the Journal of Constructional Steel Research (Mirghaderi and Dehghani 2008). 4. SEISMIC CONFIGURATION OF PROPOSED UPGRADED CONNECTIONS Existing Khorjini connections should be upgraded such that they provide required rigidity as well as reliable load path. In order to achieve such desirable seismic behavior, the connection should be detailed and configured carefully. Hence, this section reviews briefly the seismic configuration of proposed upgraded connections. Figure 4 shows the complete configuration of the upgraded connection, and is considered as the reference figure during this section.

4.1. Providing the Rigidity in Khorjini Frame Direction One of main goals of the proposed method is to improve the non-rigid existing Khorjini connections to rigid connections which can resist earthquake loads properly. In the upgraded connection, the R-

5 out of 12

plates have the main role of providing rigidity in the Khorjini frame direction. As the Figure 4 shows, there are two R-plates which are connected to columns by vertical fillet welds perpendicular to Khorjini frame direction. In this case, the continuous beams are supported by R-plates, and are also welded to R-plates by fully penetration welds at their top and bottom flanges. Such configuration keeps the rotation of beams and columns equal, and causes to satisfy the rigidity requirements. 4.2. Providing the Rigidity in Perpendicular Frame Direction Beams perpendicular to Khorjini frames can not be connected to columns directly, and they are connected to Khorjini beams instead. In typical Iranian steel buildings, these perpendicular beams often connect to Khorjini beams only by hinged connections, and therefore, the building has almost no lateral load bearing system in perpendicular direction. In the proposed method, this structural fault can be eliminated by using two horizontal plates. According to Figure 4, these two horizontal plates named as By and Ty are placed on the top and bottom flanges of perpendicular beam, and their performance is similar to welded flange plates used in regular rigid connections. These two plates are not connected directly to columns, but are welded to R-plates by horizontal fully penetration welds instead. In fact, these cover plates provide the expected rigidity in perpendicular frames.

Figure 4: Complete configuration of the proposed upgraded Khorjini connection

4.3. Possibility of Adding Bracing Systems In cases that modifying the existing Khorjini frames to moment resisting frames can not provide the required lateral load bearing capacity, the proposed upgrading method foresights appropriate details for adding the braces. These details help the braces to be added to Khorjini frames by suitable seismic configuration, and as a result, the lateral load bearing capacity of these frames can be increased significantly in essential cases. According to Figure 4, braces can be welded to vertical gusset plates named as Gx, and transfer their loads to the connection through the reliable load path. The Gx-plates

6 out of 12

have the main role of transferring the brace loads to columns as well as balancing the horizontal and vertical load components. Tx- and Bx-plates which are placed on top and bottom flanges of Khorjini beams transfer the horizontal floor shear to connections, and they are also effective in ensuring the stable behavior of Gx-plates by contribution of S- and L-plates. These plates help significantly in preventing Gx-plates from buckling. 4.4. Formation of Plastic Hinges The proposed upgraded connection facilitates the formation of plastic hinges in connected beams. Since this connection has adequate strength and stiffness, it never fails prior to the adjacent beams. This fact results in the possibility of formation of plastic hinges in connected beams, and therefore, the upgraded frame can be categorized as a moment resisting frame which has a ductile behavior while earthquakes. According to Figure 5, the place of formation of plastic hinges in Khorjini beams is immediately after the Z-plate, and in perpendicular beams is beyond the end of By- and Ty-plates. Using the mentioned plates has caused the plastic hinges to be formed far enough from the connection, and this can ensure the reliable performance of the connection.

Figure 5: Expected location of the plastic hinge in

Khorjini beams (Left) and the perpendicular beam (Right)

4.5. Panel Zone One of the advantageous characteristics of existing Khorjini connections is that the panel zone area can be detected on them. Although the place of panel zone in Khorjini connections has been located on the web of Khorjini beams instead of columns, researches (Dehghani 2001 and Pirouzanfar 2005) have confirmed that if the upgraded connection is designed and configured properly, this panel zone can act similar to the panel zone of regular connections, and has a great role in providing suitable seismic behavior of connections. This area changes the direction of loads from horizontal to vertical for transferring to R-plates and then columns by its shear action. Figure 6 shows the panel zone area in Khorjini beams. As it can be inferred from the Figure 6, the panel zone area can also be strengthened easily by adding Z-plates in cases that the web thickness is not sufficient for providing required shear capacity.

7 out of 12

Figure 6: Panel zone area and its role in Khorjini beams (Mirghaderi and Dehghani 2008)

4.6. Continuity Plates Continuity plates have an undeniable role in ensuring reliable load path as well as increasing the load bearing capacity of the connection. Adding continuity plates can also prevent undesired local events like web crippling and buckling in panel zone area (Bruneau et al. 1998). Hence, because the panel zone area of Khorjini connections is located in the beam web, a pair of continuity plates shown by hatched plates in Figure 6 should be placed on both sides of beam web in direction of R-plates. The beveled continuity plates are connected to the inside of top and bottom beam flanges by fully penetration welds and to the beam web by fillet welds. Applying this detail can also help significantly in providing consistent load path between top and bottom parts of R-plates. 5. CASE STUDY BUILDINGS For the purpose of seismic evaluation of typical Iranian steel buildings which have Khorjini connections, three categories of them have been studied before and after upgrading their connections. These buildings include 3-, 6-, and 9-Floor existing buildings that can be representative of low-, medium-, and high-rise buildings respectively. These selected buildings have been modeled and analyzed by static and dynamic methods comprehensively in order to understand their seismic performance. Each of aforementioned buildings has been studied in three different stages: At the first stage, the existing building with its typical Khorjini connections has been considered. Then at the second stage, connections of the study building have been upgraded to rigid connections using the proposed method. Finally at the third stage, the frame elements including beams and columns have been strengthened in addition to the upgraded connections. Comparison of these three stages can give a clear view toward the seismic behavior of study buildings. The structural system of all study buildings follows the common practice of construction of steel buildings in Iran. These buildings have no bracing system, and their connections are Khorjini connections. According to DIN Standard, columns consist of double IPE-profiles joined together by tie plates, and beams consist of a pair of continuous IPE-profiles. It should be mentioned that the connections of columns to the foundation have been assumed to be hinged connections based on typical details of study buildings.

8 out of 12

6. MODELING, ANALYSIS, AND RESULTS 6.1. Assumptions and Modeling For the purpose of modeling and seismic analyses of study buildings, a 2-D frame has been extracted from each of aforementioned buildings. Figure 7 shows the general dimensions of selected 3-, 6-, and 9-Floor frames. It should be noted that the comparisons between 2-D and 3-D models have been confirmed that the 2-D models adequately capture the major static and dynamic behavior characteristics of study buildings.

Figure 7: Dimensions of selected 2-D frames including 3-, 6-, and 9-Floor frames subjected to seismic evaluation

All of case study buildings have been chosen from buildings which are capable of resisting gravity loads with no problem. This fact has been confirmed by analyzing the selected buildings under gravity loads. The gravity load values used in structural analyses of selected buildings are defined based on design load criteria suggested by Iranian national building code. In all selected buildings, the dead load values are 450, 600, and 450 kg/m2, while the live load values are 150, 200, and 500 kg/m2 for the roof (last floor), middle floors, and the parking (first floor) respectively. The study buildings have been modeled in the first stage assuming to have Khorjini connections, and in the second stage assuming to have upgraded connections. It should be emphasized that the beam and column section properties are the same in first two stages. But in the third stage, the frame elements have been strengthened by adding plates. These plates are added to increase the frame load bearing capacity as well as to provide desirable proportioning for satisfying the strong column-weak beam regulation. After some iteration, the suggested method for strengthening of beams and columns

9 out of 12

includes adding two plates to the top and bottom flanges of Khorjini beams located at the first floor, as well as adding two plates to sides of all frame columns. The added column plates can be considered as the continuation of R-plates upward and downward. In order to ease the comparison of seismic performance of buildings, the proposed strengthening method at the third stage has been assumed to be the same for all study frames. It is obvious that in real retrofitting projects, the best and optimum frame sections can be found by more iteration based on existing conditions of the building. One of important cases in modeling of typical Iranian buildings is the proper definition of existing Khorjini connections. Khorjini connections in contrast to the regular connections should be modeled in a manner that the continuous behavior of beams is preserved in connections. In other words, the rotation angle of beams should be essentially equal at both sides of connections. In order to satisfy the mentioned behavior, an element called “Panel Zone” has been assigned to connections while modeling. Assigning this element provides the possibility of different or even independent rotation of beams and columns. This element adds two rotational springs around the major and minor axes of connections, and specific stiffness can be defined for each of them. In the case of assuming zero stiffness, the rotation of beams and columns will be totally independent, but on the other hand, in the case of assuming considerable stiffness, the rotation of beams and columns will be the same, similar to rigid connections. It should be noted that the displacement of beams and columns at the joints will be essentially the same in all cases. 6.2. Modal Analysis Modal analysis is used to determine the fundamental vibration periods as well as mode shapes of structural systems. This analysis can show the general behavior of study structures, and is considered as the basis of response spectrum and time history analyses. Table 1 shows the fundamental vibration periods, and Table 2 shows first mode participating mass ratio of 3-, 6-, and 9-Floor frames in three aforementioned stages.

Table 1: Fundamental vibration periods (seconds) Model Type 3-Floor 6-Floor 9-Floor

Khorjini Connections 2.367 4.438 6.565

Connections Upgraded 0.584 1.047 1.518

Fully Upgraded 0.421 0.844 1.312

Table 2: Participating mass ratio - First mode (unitless)

Model Type 3-Floor 6-Floor 9-Floor

Khorjini Connections 0.871 0.836 0.827

Connections Upgraded 0.916 0.877 0.851

Fully Upgraded 0.866 0.810 0.790

As it can be inferred from Table 1, typical Iranian steel frames with Khorjini connections have the large fundamental vibration periods, and consequently low stiffness. Upgrading the existing Khorjini connections decreases the fundamental vibration periods to the regular range. According to Table 1, strengthening of the frame elements in addition to upgrade of connections reduces the fundamental vibration periods by about 20 percent in average. This improvement is achieved because of providing more stiffness in frames. Furthermore, Table 2 demonstrates that the first mode has a participating mass ratio of approximately 0.85 which is acceptable for regular buildings. 6.3. Nonlinear Static Analyses The objective of the nonlinear static analysis is to determine the lateral capacity of the structural frame, the failure mechanism, and the sequence of inelastic response events leading to near collapse. In the current study, both material and geometric nonlinearities have been applied to structural models. Material nonlinearity is added by defining appropriate plastic hinges on frame elements based

10 out of 12

on criteria mentioned in FEMA-273 (1997). Moreover, geometric nonlinearity is applied to the models by activating both P-Delta and large displacement effects. It should be declared that because the modeling of premature brittle failure of Khorjini connection prior to upgrading was not possible in the used software, SAP2000, related pushover curves (Figure 8) could continue infinitely. Hence, in the first stage of analyses, the pushover curves have been cut based on related experimental and analytical researches on the seismic behavior of Khorjini connections (Dehghani 2001).

0

1000

2000

3000

4000

5000

6000

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

Target Displacement (m)

Bas

e R

eact

ion

(kN

)

3: Fully Upgraded3: Connections Upgraded3: Khorjini Connections

0

1000

2000

3000

4000

5000

6000

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00


Bas

e R

eact

ion

(kN

)


0

1000

2000

3000

4000

5000

6000

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00


Bas

e R

eact

ion

(kN

)


Figure 8: Pushover curves of 3-, 6-, and 9-Floor frames Figure 8 shows the base shear-roof displacement curves obtained from nonlinear static analyses of 3-, 6-, and 9-Floor frames. Results of study frames approve that existing frames with Khorjini connections have low load bearing capacity, and they experience large displacements while earthquakes. On the other hand, by upgrading the connections using the proposed method, the load bearing capacity of frames increases significantly, and they can be expected to resist design earthquake loads with acceptable lateral displacements. Furthermore, if more increase in load bearing capacity is required, strengthening of the frame elements plus upgrading the connections results in an improvement of about 70 percent in lateral load bearing capacity of frames. 6.4. Nonlinear Time History Analyses In order to evaluate the response of study buildings, seven different earthquakes including Bam (2003), Manjil (1990), Tabas (1978), Naghan (1977), El-Centro (1940), Northridge (1994), and Kobe (1995) have been selected. These earthquakes time histories are widely used in the profession to represent typical earthquakes, and have different peak ground accelerations, strong motion durations, and frequency contents. While the selected earthquake time histories should be representative of the real ground motion at the building site, all chosen time histories have been scaled to 0.5g based on a method suggested by Iranian Code of Practice for Seismic Resistant Design of Buildings (BHRC 2005). This method is almost similar to the proposed method of FEMA-356 (2000). Ultimate responses of study frames obtained from nonlinear time history analyses have been assumed equal to the average of results calculated by each of seven time histories. The maximum values of average

11 out of 12

responses have been shown in term of maximum lateral displacement of floors (Figure 9). It should be noted again that results for the first stage of analyses are based on assuming proper connection behavior during all earthquakes, while in real cases existing connections fail after just few cycles because of premature brittle failure (Mirghaderi and Dehghani 2008). Considering the maximum lateral displacements of floors shown in Figure 9 demonstrates that upgrading the connections decreases lateral displacements significantly. Additionally, by strengthening of the frame elements plus upgrading the connections, the lateral displacements can be reduced by about 45, 25, and 15 percent in 3-, 6-, and 9-Floor frames respectively.

0

1

2

3

0 10 20 30 40 50 60 70 80 90

Story Displacement (cm)

Stor

y

Fully UpgradedConnections UpgradedKhorjini Connections

0

1

2

3

4

5

6

0 10 20 30 40 50 60 70 80 90


Stor

y


0

1

2

3

4

5

6

7

8

9

0 10 20 30 40 50 60 70 80 90


Stor

y


Figure 9: Maximum lateral displacement of floors for 3-, 6-, and 9-Floor frames 7. CONCLUSIONS This paper studies the seismic performance of typical Iranian steel buildings. Since a large number of these buildings have been constructed based on no seismic code, there are serious concerns about their performance in future earthquakes. The experiences of past earthquakes in Iran shows that many of these steel buildings may suffer considerable damage while earthquakes. Observation of damaged buildings approves that most of them have failed due to their joint failure. The beam-to-column connections of these buildings follow the practice of Iranian typical connection named as “Khorjini connection”. This connection should be categorized as a non-seismic semi-rigid connection, and its moment capacity is far less than the connected frame elements. The major reason of weak performance of this type of connections is the premature failure of welds. Hence, frames with Khorjini connections have extremely low ductility, and they often show undesirable brittle behavior during earthquakes. In order to improve the seismic performance of this common Iranian connection, the authors’ proposed method uses two vertical plates called R-plates for the purpose of satisfying strength and stiffness requirements as well as providing a reliable load path. Theoretical and experimental researches confirm that this upgraded connection coincides well with the seismic regulations established for connections of buildings in high seismic areas. Upgrading the existing Khorjini connection can provide the rigidity in both Khorjini frames and perpendicular frame. Furthermore, this method provides the possibility of adding braces if extra capacity is required.

12 out of 12

Comparison of results obtained from structural analyses confirms that the load bearing capacity of frames is increased significantly by upgrading the connections. Frames with upgraded connections are capable of resisting moderate or high earthquakes by their ductile behavior in contrast to the existing Khorjini frames that can bear just limited lateral forces by their brittle behavior. This pattern of strengthening also prevents from undesirable weak column-strong beam mechanism by formation of plastic hinges in beams instead of columns. In short, it can be concluded that the proposed upgrading method is an applicable cost-saving method which considers all possible deficiencies, and improves the seismic behavior of typical Iranian steel buildings significantly. REFERENCES [1] Arbabi F. (1998) Nonlinear deformation of Satchel (Khorjini) connections. Journal of Seismology and Earthquake Engineering, IIEES; 1:51-57. [2] Astaneh-Asl A., Modjtahedi D., McMullin C., Shen J.H., Amor E.D. (1998) Stability of damaged steel moment frames in Los Angeles. Journal of Engineering Structures; 20:433-446. [3] Bruneau M., Bhagwagar T. (2002) Seismic retrofit of flexible steel frames using thin infilled panels. Journal of Engineering Structures; 24:443-453. [4] Bruneau M., Uang C.M., Whittaker A. (1998) Ductile Design of Steel Structures, McGraw-Hill. [5] BHRC. (2005) Iranian Code of Practice for Seismic Resistant Design of Buildings, Standard No.2800, 3rd Edition. Building and Housing Research Center of Iran, Tehran, Iran. [6] Dehghani M. (2001) Analytical and experimental studies of Khorjini beam-to-column connections and new details of a rigid connection, MSc Thesis. University of Tehran, Tehran, Iran. [7] FEMA. (1997) Guidelines for the seismic rehabilitation of buildings. FEMA 273, National Earthquake Hazard Reduction Program, Federal Emergency Management Agency, Washington. [8] FEMA. (2000) Prestandard and commentary for the seismic rehabilitation of buildings. FEMA 356. National Earthquake Hazard Reduction Program, Federal Emergency Management Agency, Washington. [9] Ghafory-Ashtiany M., Moghaddam H., Tiv M., Ghane A. (1995) Dynamic modeling of Khorjinee connection, IIEES report no.74-95-6, IIEES, Tehran, Iran. [10] Hajjar J.F., Sullivan D.P. (1998) Analysis of mid-rise steel frame damaged in Northridge earthquake. Journal of Performance of Constructed Facilities; 12: 221-231. [11] Mirghaderi S.R., Dehghani M. (2006) The new details of rigid connection. In: Proceedings of the 13th European conference on earthquake engineering, Geneva, Switzerland; Paper no.1344. [12] Mirghaderi S.R., Dehghani M. (2008) The rigid seismic connection of continuous beams to column. Journal of Constructional Steel Research; doi:10.1016/j.jcsr.2008.01.015. [13] Moghaddam H., Hajirasouliha I. (2006) Investigation on accuracy of pushover analysis for estimating the seismic deformation of braced steel frames. Journal of Constructional Steel Research; 62:343-351. [14] Moghaddam H., Koohian R. (1994) Strength of steel buildings with semi-rigid connections (Khorjini connections) against earthquake loads. IIEES report no.10-96-74, IIEES, Tehran, Iran. [15] Pirouzanfar M. (2005) Analytical study and suggestion of new moment resisting Khorjini connections, MSc Thesis in Civil-Structural Engineering, University of Tehran, Tehran, Iran. [16] Reyes-Salazar A, Haldar A. (1999) Nonlinear seismic response of steel structures with semi-rigid and composite connections. Journal of Constructional Steel Research; 51:37-59. [17] Shafei B. (2007) Seismic evaluation and retrofit of existing steel buildings. MSc Thesis in Civil-Structural Engineering, University of Tehran, Tehran, Iran. [18] Shafei B., Lestuzzi P., Motavalli M., Mirghaderi S.R. (2006) Iranian Existing Steel Buildings: State of the Practice; Part 1: Non-Engineered Practices. In: Proceedings of the 13th European conference on Earthquake Engineering, Geneva, Switzerland; Paper no.1779. [19] Tahooni S. (1990) Stiffness of Khorjini connections. Report submitted to the Housing Foundation of Iran, Tehran, Iran. [20] Taranath B.S. (2005) Wind and Earthquake Resistant Buildings, Marcel Dekker, New York. [21] Tehranizadeh M. (2000) Approximate parameter for semi-rigid Khorjinee connections in dynamic torsional response of steel structures. Journal of Engineering Structures; 22:335-342. [22] Tehranizadeh M. (2001) Passive energy dissipation device for typical steel frame building in Iran. Journal of Engineering Structures; 23:643-655.

seismic retrofit of typical iranian steel buildings...

Documents