- Email: [email protected]
An alternative detail for continuity plates in steel beam to box-column moment-connections
Journal of Constructional Steel Research 167 (2020) 105952
Contents lists available at ScienceDirect
Journal of Constructional Steel Research
An alternative detail for continuity plates in steel beam to box-column moment-connections M.A. Najafgholipour a,⁎, Kianoush Peykari b, S.M. Dehghan a a b
Faculty of Civil and Environmental Engineering, Shiraz University of Technology, Shiraz, Iran Graduate Student of Civil and Environmental Engineering, Shiraz University of Technology, Shiraz, Iran
a r t i c l e
i n f o
Article history: Received 25 June 2019 Received in revised form 25 November 2019 Accepted 25 January 2020 Available online xxxx Keywords: Steel moment-connection Continuity plates Vertical stiffener Box-column Experimental study Finite element model
a b s t r a c t In this study, the use of an internal vertical stiffener in built-up box-columns parallel to the beam web, which turns the column section into a boxed I-section in the connection region to eliminate the common continuity plates, is assessed through an experimental and numerical study. For this purpose, two full-scale exterior Ibeam to box-column moment connections (one with common continuity plates and the other one with a vertical stiffener as an alternative detail) are tested under reversed cyclic loading. The test results in terms of hysteresis moment-rotation curves, strain in critical stations, and observations during the tests indicate that the connection with the vertical stiffener could survive story drifts of up to 4% without any strength loss and brittle failure such as weld fracture. For further investigation of the seismic performance of the proposed detail, Finite Element Modeling (FEM) of the test specimens is also conducted in the FE software ABAQUS and validated using the test results. The results of the numerical study also confirm the satisfactory behavior of the connections with the proposed detail. © 2020 Elsevier Ltd. All rights reserved.
1. Introduction Steel Moment Resisting Frames (MRFs) are well known as ductile earthquake resistant structural systems. They are commonly employed in buildings with different heights and occupancies mainly due to their minimum architectural interference. After Northridge (1994) and Kobe (1995) earthquakes whereby steel MRFs experienced severe damages especially in their connections, the crucial role of beam-column moment connections in the global seismic performance of steel MRFs was confirmed. In this regard, extensive studies were conducted to find the weak points of the existing connections and to make improvements to enhance their seismic performance. These studies led to specific guidelines for retrofitting existing joints and prequalified connections for new buildings [1,2]. Regarding the importance of steel moment connections and their effects on the seismic performance of steel MRFs, numerous experimental and numerical studies have been done to evaluate their behavior under cyclic loads. A significant number of studies have been devoted to the seismic performance of the common types of connections in recent decades [3–7]. In a number of experimental studies, some modifications have been proposed to improve the seismic performance of the connections by enhancing their strength as well as their plastic rotation capacity under reversed cyclic loads [8,9]. Innovative types of connections ⁎ Corresponding author. E-mail address: [email protected] (M.A. Najafgholipour).
https://doi.org/10.1016/j.jcsr.2020.105952 0143-974X/© 2020 Elsevier Ltd. All rights reserved.
with special features have been presented in a number of studies [10,11]. The performance of Reduced Beam Sections (RBS) as an improvement technique in moment resisting connections has also been evaluated in numerical and experimental studies [12–16]. To assess the effects of welding type, welding pattern, and welding damage on the seismic performance of steel moment connections, several studies have been performed [17,18]. A number of analytical and numerical studies have been conducted to explore the force-transferring mechanism in different types of connections leading to the development of design procedures for steel moment connections [19,20]. Due to the widespread use of built-up columns with double I-sections in some regions such as Iran, some experimental and numerical investigations have been conducted to evaluate the different aspects of these connections [21,22]. Wide-flange, box, boxed wide-flange, and flange cruciform columns are permitted to be used in steel MRFs according to AISC 341-16 [23]. Built-up steel box columns are widely used in regions where heavy hot-rolled steel sections are not available. Moreover, due to the inherent geometrical properties of box sections, they are more efficient compared to I-sections especially under bidirectional moments. In addition, beams can be joined to all sides of box columns simultaneously by means of appropriate types of moment connections. Furthermore, box sections are composed of stiffened elements reducing the widthto-thickness ratio requirements for seismic compact sections. A welded steel I-beam to box-column moment resisting connection is schematically illustrated in Fig. 1-a. The mechanism of transferring
2
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 1. Steel welded I-beam to box-column moment connection, a- Schematic view of the connection, b- Moment transfer mechanism in the joint.
internal actions in the joint region is shown in Fig. 1-b. The continuity plates across the column sections have a key role in transferring the bending moment from the adjacent beams to the column and vice versa as an internal diaphragm. A pair of tensile and compressive forces induced in the beam flanges (Puf), equivalent to the bending moment in the beam (Mu), are transferred to the joint panel through the in-plane action of the continuity plates. In the absence of continuity plates, the tensile and compressive forces cause a local bending moment in the column flanges which in turn leads to excessive flexural deformations. These deformations may be accompanied by the local yielding or local buckling of the column flanges. Various studies have been done to investigate the role of different types of stiffeners and details in the seismic behavior of steel moment connections with wide-flange and box columns. In a comprehensive research project, the nonlinear behavior of moment resisting beamcolumn joints with various detailing of stiffeners (doubler plates and continuity plates) was assessed [24–26]. As part of this research, Lee et al. [24] conducted experimental studies on six interior connections with different detailing of wide-flange column stiffeners. Afterward, Prochnow et al. [25], by focusing on the role of continuity plates in steel beam to column moment connections, conducted pull tests on beam flange to column flange specimens. In their report, a complete history of the investigations on the behavior of continuity plates and related code provisions can be found. Subsequently, in a numerical study, Ye et al. [26] assessed the nonlinear behavior of steel beam-column moment connections focusing on the role of doubler plates and continuity plates on their seismic behavior. For this purpose, they numerically investigated both previously conducted pull tests on column flange subassemblies and tests on beam-column connections. They also evaluated the behavior of continuity plates and doubler plates in wide-flange and box columns through a parametric numerical study. The analysis results showed that using continuity plates reduced the bending stress and strains in column flanges considerably. In another study, Ahmady Jazany and Hosseini Hashemi [27] conducted cyclic tests on six interior connections to assess the influence of the continuity plate configuration on the seismic behavior of the panel zone with unequal beam depths. The test results indicated that the configuration of the continuity plates not only affected the stiffness of the sub-assemblages but also influenced their plastic rotation capacity. Recently, through an analytical and numerical study, Amani et al. [28] investigated the local bending of column flange and the requirements of the continuity plates in double-web H-shaped columns. Despite the various advantages of box sections over other section types and their widespread use in some regions of the world, relatively few studies have been specifically conducted on the seismic performance of the steel beam to box-column connections. As an example,
Chen et al. [29] evaluated the performance of welded I-beam to boxcolumn moment connections. The results of their study demonstrated the role of internal diaphragms in the seismic performance of I-beams to box-columns moment resisting connections. As another early study on the beam to box-column connections, Shanmugam and Ting [30] performed tests on interior joints. Kim et al. [31] conducted tests on pre-Northridge steel beam to box-column connections with internal diaphragms. The test results as well as the numerical model of the connections led to some welding details of the continuity plates. Flange-plate connections of I-beams to box-columns were assessed by Gholami et al. [32] through an experimental and numerical study. The results of this study indicated that the connections were able to satisfy the AISC criteria for special moment connections. Chen and Shi [33] performed an experimental investigation on the seismic performance of a new end-plate joint with box-columns. The test results revealed that the proposed prefabrication technique had a satisfactory performance under cyclic loads. Mirghaderi et al. [34] proposed a through-plate connection for steel I-beam to box-column moment connections. They evaluated the behavior of the connection in an experimental study and investigated the dimensions and shape of the plate by means of a numerical model. A novel short-stub-beam connection was proposed by Erfani et al. [35] for I-beam to box-column connections. Finite Element (FE) method was employed to evaluate the seismic performance of this connection. Recently, Jahanbakhti et al. [36] evaluated the behavior of the panel zones in beam to box-column connections through an experimental study on three specimens. They examined the elimination of the continuity plates from the connection by providing a minimum column flange thickness. Considering the complicated welding of continuity plates in box-columns due to lack of accessibility, a number of researchers have proposed alternatives for continuity plates in box columns. For instance, Rezaifar et al. [37] assessed the seismic behavior of interior connections with external stiffeners as an alternative detail to eliminate the internal diaphragms. In a numerical study, Saffari et al. [38] examined a new detail as an alternative for common continuity plates. 1.1. AISC provisions for continuity plates in boxed wide-flange columns According to AISC 341-16 [23], in steel I-beam to boxed wide-flange column moment resisting connections and under some conditions, continuity plates are not required and the joint panel is able to transfer the internal induced actions properly. According to the code provisions, the column flange should necessarily satisfy the following limiting condition: t lim ¼
bbf 12
ð1Þ
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
3
Fig. 2. Fabrication sequence of the box-column section with a common continuity plates.
where, tlim is the minimum permitted thickness of the column flange and bbf is the width of the beam flange. In addition, the column flange and web should be designed under the concentrated force in the beam flange caused by the probable flexural moment capacity of the beam. For this purpose, the limit states of the local yielding, local crippling, sidesway buckling, and compression buckling must be checked for the column web. Moreover, the column flange should be designed for local bending. In the former edition of AISC 341-10 [39], the continuity plates could be eliminated in boxed wide-flange columns if the column flange thickness (tcf) satisfied the following inequalities:
t cf ≥
bbf 12
ð2Þ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi v" u # u bbf bcf F yb Ryb t 1:8bbf t bf 1− 2 bbf − t cf ≥0:4 4 F yc Ryc bcf
ð3Þ
where bbf and bcf are the widths of the beam and column flanges, respectively, and tbf is the thickness of the beam flange. Also, Fyb and Fyc are the specified minimum yield strengths of the beam and column, respectively, and Ry is Ratio of the expected yield stress to the specified minimum yield stress. 1.2. Research significance Although the horizontal continuity plates can be easily welded in wide-flange columns, proper welding of all the edges of the continuity plate to a box-column is difficult due to lack of accessibility. The fabrication sequence of a steel beam to box-column connection with common continuity plates is illustrated in Fig. 2 Accordingly, the welding of the fourth edge of the continuity plates to the column cannot be done properly using common welding procedures. To eliminate the commonly used continuity plates in steel I-beam to box-column moment connections, an alternative detail inspired from the aforementioned points about the necessity of continuity plates in
Fig. 3. Welded steel I-beam to box-column connection, a- Common detail with horizontal continuity plates, b- The proposed detail with a vertical stiffener.
4
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 4. Moment transfer mechanism in the connections with vertical stiffener.
Fig. 5. Fabrication sequence of the box-column section with a vertical stiffener.
boxed wide-flange columns is proposed in this study. For this purpose, a vertical stiffener is used in the column parallel to the beam web (see Fig. 3). Consequently, the column section becomes similar to a
wide-flange boxed section in the joint region. By satisfying the geometrical requirements and controlling the limit states, the elimination of the horizontal continuity plates may be possible by using a vertical stiffener.
Fig. 6. Dimensions of the test specimens.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 7. Fabrication details of the RBS-H connection, a- Side view, b- Top view, c- Dimensions of the continuity plates, d- Shear tab dimensions, e- Beam section.
5
6
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 8. Fabrication details of the RBS-V connection, a- Side view, b- Top view, c- Dimensions of the Vertical stiffener, d- Shear tab dimensions, e- Beam section.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
7
Fig. 9. Fabrication of the test specimens in shop.
Following the fabrication procedure of the proposed connection as illustrated in Fig. 5, its applicability is confirmed for practical cases. To evaluate the efficiency of this idea, the seismic performance of the I-beam to box-column moment resisting connections with the proposed detail is assessed through an experimental program. For this purpose, two full-scale exterior beam-column moment connections (one with common horizontal continuity plates and the other one with a vertical stiffener in the column) are tested under cyclic loadings. For further evaluation of the proposed technique and detailed investigation on the behavior of the test specimens, a Finite Element (FE) numerical study is implemented in the software ABAQUS [40] and validated with the test results. Fig. 10. Tensile test of the steel plates according to ASTM A370 [41].
Table 1 Mechanical properties of the steel plates and sections. Test specimen
Yield stress (MPa)
Ultimate strength (MPa)
Elongation (%)
HEA280 (Web) HEA280 (Flange) Plate 20 mm Plate 15 mm
368.8 298.1 300.7 292.2
521.1 474.9 475.1 476.7
24.8 28.9 24.2 20.1
Furthermore, by using the RBS detail in the beam, it is ensured that the plastic hinge in the beam occurs at a certain distance from the column face. A schematic view of the force transmission in the connections with vertical stiffness is illustrated in Fig. 4. Accordingly, the tensile and compressive forces induced in the beam flanges (Puf) result in internal forces in the column webs (Vcw) and the vertical stiffener (Vs).
2. Experimental program Two full-scale exterior joints were fabricated and tested in this study. The details of the test program including the fabrication process of the specimens and the test set-up as well as the results of the experimental study are presented here. 2.1. Test specimens The test specimens were exterior beam-column connections isolated from a steel MRF with the story height of 2.6 m and span length of 6 m (see Fig. 6). The connections were made up of a 300 mm × 300 mm built-up steel box column with the flange thickness of 20 mm. HEA280 was employed as the beam with the same height and flange width of 280 mm. The flange and web thicknesses of the beam section were 18 mm and 10.5 mm, respectively. Continuous groove weld was used to assemble the plates of the built-up box-column. The only difference between the specimens was that in the reference connection (RBS-H), a pair of square
Fig. 11. Test setup for the cyclic test of the exterior moment connections.
8
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 12. Loading protocol of AISC 341-16 in the reversed cyclic test of the steel moment connections [23].
Fig. 13. Configuration of the mounted strain gauges on the connections, a- RBS-H connection, b- RBS-V connections.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
9
Fig. 14. Hysteresis moment-rotation curve of the RBS-H connection.
continuity plates within the column (260 mm × 260 mm × 20 mm) were utilized in the plane of the beam flanges as the common detail for the moment resisting connections. In the other connection (RBS-V), a vertical stiffener in the column axis with the dimensions of 260 mm × 500 mm × 20 mm in the plane of the beam web was employed instead of the continuity plates. The vertical stiffener in the latter connection was extended 110 mm beyond the top and bottom flanges of the beam which makes the column section in the connection region similar to a boxed wide-flange section. The continuity plates and the vertical stiffener were jointed to the columns using groove welds. The fabrication and welding details of the test specimens including the dimensions of the different parts are illustrated in Figs. 7 and 8 for RBS-H and RBS-V, respectively.
It should be noted that the beam and column dimensions were determined based on a preliminary design according to AISC341–16 seismic provisions [23] to satisfy the strong column-weak beam criterion. The joint panel zone was designed for the expected shear stress due to the expected plastic moment of the beam to the column. Moreover, the shear tab was designed for the expected shear force developed in the beam corresponding to the plastic flexural strength of the beam section. In addition, to avoid extensive flexural plastic strain in the beam just adjacent to the column face, a Reduced Beam Section (RBS) was used with a 5% reduction of the flange width of the beam (see Figs. 7-b and 6-b). As is shown in Fig. 9, the connections were shop-welded by qualified welders. During fabrication, Ultrasonic Testing (UT) was performed on
Fig. 15. Measured strain at different levels of beam rotation.
10
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
all the groove welds to assure their quality and penetration. Finally, the specimens were painted with whitewash to monitor the inelastic deformations during the test. 2.2. Material properties To determine the mechanical properties of the steel plates and sections, tensile test based on ASTM A370 [41] was performed on standard coupons cut from the different parts of the connections (see Fig. 10). Using the obtained stress-strain curves from each tensile test, the key mechanical properties were determined as listed in Table 1. These properties were used in the numerical study and the verification phase. 2.3. Test setup and procedure To evaluate the seismic performance of the connections with common and alternative continuity plate details, reversed cyclic tests were conducted on the test specimens (Fig. 11). Under lateral load, it is assumed that contra-flexure points in the beams and columns of a MRF are located almost at the mid-stories and mid-spans of the frame. Thus, pin supports were utilized at the top and bottom of the column and at the beam tip as well (see Fig. 11). The cyclic displacement control load with the ascending amplitude following the standard loading protocol prescribed in AISC 341-16 [23] (Fig. 12) was applied to the end of the beam (3000 mm from the
column axis) by means of a 250 kN semi-automatic hydraulic actuator. However, in this study, the tests were terminated at the beam rotation of 0.04 rad. The displacement of the loading point and the applied load during the test were measured by means of a displacement transducer (LVDT) and a load cell, respectively. To avoid any lateral displacement and instability of the beam during the test, a lateral support was provided at a distance of 1500 mm from the column face. Unlike Reinforced Concrete (RC) beam to column moment connections, the axial load does not affect the behavior of the steel beam to column moment connections, especially in cases that the panel zone is designed to transfer the moment from the adjacent beams. Therefore, in this study, the connections were tested in the absence of column axial load. To measure the strain in regions where the yielding of the section was anticipated or stress concentration was probable, some strain gauges were mounted on the test specimens with the configuration presented in Fig. 13. It should be noted that due to some technical problems, a number of the strain gauges were disrupted and their data were not available. 2.4. Results of the experimental study The results of the cyclic tests on the connections including hysteresis load-displacement curves (moment-rotation curves) as well as observations during the tests are presented in the subsequent sections. It should be noted that the rotations reported in the subsequent sections, are the
Fig. 16. Observed behavior of the specimen RBS-H, a- 0.01 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
11
Fig. 17. The beam to column groove weld at the end of the test.
chord rotation of the beam which is the ratio of the vertical deflection of the beam end to the distance between the loading point and the column centerline. Moreover, the plastic moment strength of the beam (Mp) specified on the moment-rotation curves was determined as follows: L 2850 Mp ¼ Z RBS F yb b ¼ ð1065 298:1Þ ¼ 350:7 kN−m 2580 LRBS
ð4Þ
where ZRBS is the plastic section modulus of the beam at reduced section, and LRBS and Lb are the distances from the loading point to the center of the RBS and column centerline, respectively, 2.4.1. Reference connection with horizontal continuity plates (RBS-H) The experimental hysteresis moment-rotation curve of the RBS-H connection is plotted in Fig. 14. As was expected, the reference connection (RBS-H) with common continuity plates in the joint panel was able to survive the rotation of 0.04 rad without any strength loss. The initial stiffness of the connection was 27,618 kN-m/rad. The peak moment capacity of the connection corresponding to the rotation of 0.04 rad was 540 kN-m which was more than the expected plastic flexural strength of the beam (Mp = 350.7 kN-m). Hence, the connection could meet the AISC 341–16 requirements as a special moment resisting connection. The measured strains at different levels of the story drift are plotted in Fig. 15. Moreover, the test observations including the local
deformations of the beam and column sections are illustrated in Fig. 16. The joint behavior during the test and the measured strains at different stations indicate that the test specimen remained elastic up to the rotation of 0.01 rad and no obvious local excessive deformation was observed in this rotation (Fig. 16-a). Strain gauge number 3 which measured the strain on the top flange of the beam in the RBS region showed the first yielding in the test specimen at the rotation of 0.015 rad. Excessive local deformations with the flaking of whitewash (which may be a sign of yielding of the beam flange in the RBS region) were also observed in the test specimen (Fig. 16-b). According to the measured strain in the strain gauges 1 and 2, the top of the beam next to the column face yielded from the drift ratio of 1% to 2%. By increasing the rotation amplitude up to 0.04 rad, the flaking of whitewash increased on the beam flanges and developed in the depth of the beam at the RBS zone (Fig. 16-c and -d) and the measured strain on the beam flanges increased to around 5 times as much as the steel yield strain. The concentration of the plastic hinge in the beam and especially at the RBS region without any fracture in the beam to column groove welds (Fig. 17) confirmed the acceptable performance of the joint as a prequalified special moment resisting connection.
2.4.2. Proposed connection with a vertical stiffener (RBS-V) The hysteresis moment-rotation curve of the connection RBS-V with a vertical stiffener in the column is presented in Fig. 18. Similar to the
Fig. 18. Hysteresis moment-rotation curve of the RBS-V connection.
12
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
reference connection, the moment-rotation curve indicates that this connection could reach the rotation of 0.04 rad with no obvious strength loss. In addition, its maximum moment strength corresponding to the rotation of 0.04 rad was 549 kN-m which was more than the plastic moment of the beam. This means that the beam was able to reach its plastic flexural moment capacity. Consequently, the connection with the alternative detail was able to meet the requirements of the special moment resisting connections according to AISC341-16 [23]. The initial stiffness of the connection was 27,126 kN-m/rad. The comparison of the hysteresis moment-rotation curves of the RBS-H and RBS-V joints confirms that replacing the horizontal diaphragms in the box-columns with a vertical stiffener in the column axis had a slight influence on the global performance of the beam-column connection (see Fig. 19). The test observations and the measured strains in the specified stations at different levels of beam rotation are illustrated in Figs. 20 and 21, respectively. Considering the moment-rotation curve as well as the measured strains and performance of the test specimen, the connection remained elastic up to the rotation of around 0.01 rad (see Fig. 20-a). Further rotation of the beam started the plastic strain and led to the development of whitewash flaking on the beam flanges in the RBS zone. The strain gauge 3 mounted in this region confirmed the yielding of the top flange of the beam. The plastic strain in the beam was enhanced by increasing the displacement amplitude up to the beam rotation of 0.04 rad. As is shown in Fig. 20-c and -d, excessive whitewash flaking was confirmed with the measured strain in the RBS region which showed strains up to around 4 times as much as the steel yield strain. Despite some relatively minor yielding of the beam flange next to the column face as recorded with strain gauges 2 and 3, the measured strain was much less than that of the RBS-H connection. However, comparing the data from the strain gauges on the specimen and the test observations confirm that the plastic hinge was concentrated in the RBS region. It should be noted that although some minor whitewash flaking in the column web in the level of the beam flanges was observed, the strain gauges did not show yielding at these stations. Finally, the inspection of the groove welds in the test specimen indicated that no weld fracture occurred in these regions (Fig. 22).
3. Numerical study To further evaluate the efficiency of the proposed detail in connections with various geometrical properties, a numerical FE study was done. For this purpose, first, the model was validated using the results of the experimental study. Then, a parametric study was conducted to assess the behavior of the proposed alternative detail in connections with different beam and column dimensions. 3.1. Validation of the FE model To validate the FE model and to ensure the accuracy of its results, both tested connections (RBS-H and RBS-V) were simulated in the FE software ABAQUS [40] and analyzed under cyclic loading. The details of the conducted numerical modeling procedure of the connections are explained subsequently. • Constitutive material model for steel According to the material tests conducted in the experimental study and presented in Section 2.2, the stress-strain curves corresponding to the different parts of the specimens along with the key mechanical properties were available. Consequently, to define the uniaxial stressstrain curve for different parts of the models, the measured mechanical properties (i.e. the yield stress and ultimate strength) as well as the corresponding strains were employed. In addition, von Mises yield criterion were utilized to simulate the nonlinear behavior of the models under multi-axial stress states. Furthermore, although no local buckling was observed in the test specimens up to the last cycle of the loading protocol, an initial imperfection of 0.001 L was defined for the models to find out any local buckling in the numerical model. • Geometry of the models and element types The geometrical dimensions and fabrication details of the connections (Figs. 7 and 8) were followed for the simulation of the joints. The FE models of the connections are illustrated in Fig. 23. A quadratic
Fig. 19. Comparing the hysteresis moment-rotation curves of the RBS-H and RBS-V specimens.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 20. Observed behavior of the specimen RBS-V, a- 0.01 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
Fig. 21. The measured strains corresponding to different loading cycles in connection RBS-V.
13
14
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 22. The beam to column groove welds at the end of the test.
20-node solid element (denoted as C3D20R) with three translational Degrees of Freedom (DOF) in each node was employed for the analysis of the connections. Following a mesh sensitivity analysis, the element sizes ranging from 10 mm to 20 mm were used. In total, around 39,000 elements and 155,000 nodes were generated for the analysis of each connection (see Fig. 23-a). It should be noted that since no defect and failure was observed in the welds during the tests and the quality of welding was verified through frequent inspection and quality control tests, the welds were not modeled as a separate part in the FE models. Instead, the tie constraint in ABAQUS software was utilized. • Loading and boundary conditions The restrained DOFs in the FE model to simulate the boundary conditions of the test specimens are illustrated in Fig. 23-b. As described in the test program, a pair of pins were put at the top and bottom faces of the column. Hence, all of the translational DOFs of the two lines perpendicular to the plane of the connection at the top and bottom faces of the
column were restrained. To model the lateral support, the corresponding nodes were restrained against out-of-plane translation. The displacement cyclic loading was also applied to the beam tip following the loading protocol employed in the experimental study. • Analysis results
The results of the nonlinear analysis of both the RBS-H and RBS-V connections are presented in terms of hysteresis moment-rotation curves as well as distribution of the plastic strain in the models. As shown in Fig. 24, the comparison of the hysteresis moment-rotation curves obtained from the numerical and experimental studies shows the acceptable agreement of the results. The numerical model not only was able to predict the ultimate strength but also simulated the connection behavior reasonably in linear and nonlinear phases. Comparing the distribution and development of the plastic strains in different levels of the lateral drift with the test observations confirms
Fig. 23. The FE model of the specimens, a- The generated elements, b- The boundary conditions and loading defined in the FE model
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 24. Comparing the hysteresis moment-rotation curves obtained from the numerical study and the tests, a- RBS-H connection, b- RBS-V connection.
Fig. 25. Plastic strain distribution and development in the RBS-H connection at different levels of beam rotation, a- 0.015 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
15
16
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 26. Plastic strain distribution and development in the RBS-V connection at different levels of beam rotation, a- 0.015 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
Table 2 The geometrical details of the connections in the numerical study. Connection
Column size, bcf (mmxmm)
Column flange thickness, tcf (mm)
Beam flange width, bbf (mm)
Detail of column stiffener
RBS-B-300-280 RBS-H-300-280 RBS-V-300-280 RBS-B-400-200 RBS-H-400-200 RBS-V-400-200 RBS-B-400-280 RBS-H-400-280 RBS-V-400-280 RBS-B-400-340 RBS-H-400-340 RBS-V-400-340 RBS-B500-280 RBS-H-500-280 RBS-V-500-280
300 × 300
20
280
400 × 400
25
200
400 × 400
25
280
400 × 400
25
340
500 × 500
30
280
None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
17
Fig. 27. Hysteresis moment-rotation curves of the connections.
the ability of the FE model in simulating the behavior of the connections in the nonlinear phase. As illustrated in Figs. 25 and 26 and as observed in the tests, the beginning of yielding in the RBS-H specimen occurred on the beam flange at the rotation of around 0.015 rad. By increasing the lateral drift, plastic strains developed in the RBS region and in the depth of the beam. Finally, at the rotation of 0.04 rad, the beam section yielded completely and the plastic hinge formed. No local buckling was observed in the numerical model. A similar trend occurred in the RBS-V connection
where no plastic strain was observed in the test specimen up to the beam rotation of 0.015 rad. By further increasing the beam's rotation, the plastic strains which were concentrated in the RBS zone developed. 3.2. Parametric study To assess the influence of effective geometrical variables on the seismic performance of beam to box-column moment connections with the
18
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
proposed stiffening detail, a parametric study was conducted by means of a verified FE model. The details of the investigated connections and analysis results are presented in the subsequent sections. 3.2.1. Defining the models in the parametric study The column flange dimensions (width (bcf) and thickness (tcf)) as well as the beam flange width (bbf) were considered as geometrical variables which can affect the efficiency of the proposed detail. For this purpose, square box columns with side dimensions of 300 mm, 400 mm, and 500 mm were considered. According to the seismic provisions of ASCE 341-16 [23], the beams and columns of special MRFs must be seismic compact to minimize the probability of local buckling in the section components. Therefore, the thicknesses of the box-columns were considered equal to 20 mm, 25 mm, and 30 mm for the columns with the side dimensions of 300 mm, 400 mm, and 500 mm, respectively. To investigate the influence of the beam flange width on the seismic behavior of the connections, beams with three flange widths of 200 mm, 280 mm, and 350 mm but with the same moment capacity obtained by tuning the RBS dimensions were used. In each case, a connection with an unstiffened panel zone was analyzed to show the crucial role of the continuity plates and to confirm the necessity of an appropriate alternative detail for continuity plates. A summary of the properties of the investigated connections and their designations is listed in Table 2. 3.2.2. Analysis results The hysteresis moment-rotation curves of the investigated connections are plotted in Fig. 27. In each diagram, the curves corresponding to stiffened connections (with common continuity plates and a vertical stiffener) and an unstiffened one are presented. The plastic moment strengths of the beams are also specified on the diagrams.
In almost all cases, the beams of the bare connections did not reach their plastic flexural strength. According to the strain distribution in the bare joints at the story drift angle of 0.04 rad, the column flanges experienced yielding accompanied by excessive flexural outof-plane deformations. This was more significant in the connections with larger bcf /bbf. For instance, the developed plastic strain in the column flange is well demonstrated in Fig. 27-a for the RBS-B-400200 connection with the highest bcf /bbf (2.0) among the studied models. This phenomenon in unreinforced connections confirms the necessity of an appropriate stiffening detail for box columns in moment resisting connections. Moreover, the analysis results of the stiffened connections with both of the abovementioned techniques in terms of the momentrotation curves and plastic strain distribution confirm their satisfactory cyclic behavior as special moment resisting connections. According to Fig. 27, the stiffened connections were able to reach the plastic flexural strength of the beam with no strength degradation up to the drift of 0.04 rad. Additionally, although in using the two strategies there was a minor difference in strain distribution in the stiffened connections, their global performance was almost the same. Considering the strain distribution in the stiffened connections that are presented in Fig. 28-b and -c for RBS-H-400-200 and RBS-V400-200, respectively, the plastic strain was mostly concentrated in the RBS region which is expected in special moment connections. Accordingly, unlike the unreinforced models, in the stiffened connections the peak flexural strain in the beam flange occurred in the RBS region. Additionally, the analysis results showed that no plastic strain developed in the panel zone of the stiffened connections. Moreover, comparing the distribution of the shear stress in the column webs and the vertical stiffener, showed that the shear force induced in the vertical stiffener was around 50% to 65% of the total
Fig. 28. Plastic strain distribution in the connections at the rotation of 0.04 rad, a- RBS-B-400-200, b- RBS-H-400-200, c- RBS-V-400-200.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
19
(a)
(b)
(c)
(d)
(e)
(f)
Fig. 29. Plastic strain distribution over the beam flange at the rotation of 0.04 rad, a- RBS-H-400-200, b- RBS-V-400-200, c- RBS-H-400-280, d- RBS-V-400-280, e- RBS-H-400-340, f- RBS-V400340.
shear force that was transferred in the panel zone. The proportions of the shear force in the vertical stiffener and the column web depended on some geometrical ratios such as bcf/bbf and bcf/tcf. The plastic strain distributions over the beam flange are plotted for stiffened (with vertical stiffener and horizontal continuity plates) connections with column size of 400 mm, see Fig. 29 Accordingly, in the vertically stiffened connections, the plastic strain in the RBS region was more than that developed near the column face. Furthermore, the average plastic strain over the beam flange has been plotted in Fig. 30.
Accordingly, in almost all connections, the peak plastic strain was in the RBS region. 4. Conclusions An alternative detail for common continuity plates in steel beam to box-column moment resisting connections, inspired from continuity plate provisions of boxed wide-flange columns, was proposed in this study. In this regard, instead of horizontal continuity plates, a vertical
20
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 30. Plastic strain distribution along the beam axis, a- RBS-H-400-200, b- RBS-V-400-200, c- RBS-H-400-280, d- RBS-V-400-280, e- RBS-H-400-340, f- RBS-V-400-340
column stiffener was employed in the panel zone. In addition, to ensure the occurrence of plastic strain concentration at a certain distance from the column face, the RBS detail was used. To assess the efficiency of the proposed technique, cyclic tests were conducted on two full-scale exterior isolated joints (one with common horizontal continuity plates known as a code-based detail and the other one with a vertical stiffener as the proposed detail). In addition, a numerical study was carried out to investigate the performance of the proposed stiffening method in connections
with various geometrical properties. The major findings of this study are presented as follows: 1- The test results in terms of the moment-rotation curves, measured strains at key stations, and test observations indicated that the alternative stiffener was able to transfer properly the induced internal actions in the panel zone and the connection was able to survive the story drift angle of 0.04 rad with no strength degradation and undesirable brittle fracture in the groove welds. In other words, the
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
connection with a vertical column stiffener meets the requirements of special steel moment resisting connections. 2- The test observations and the measured strains in both connections indicated that the plastic strain was mostly concentrated in the RBS zone where the strain on the beam flange reached up to at least 4 times as much as the yield strain of steel. 3- Considering the welding procedure of the two stiffening strategies shows that the alternative vertical stiffener is more practical than common continuity plates in built-up box columns. 4- The results of the numerical study on unstiffened joint by using a validated FE model confirmed the need of an appropriate stiffening detail for the panel zone of the box columns. In unreinforced connections, a significant yielding of the column flange accompanied by excessive flexural out-of-plane deformations was observed preventing the beams from reaching their plastic flexural strength. This phenomenon was more prominent in connections with higher bcf/bbf. 5- The results of the numerical study on the connections designed with different values of bcf/bbf and stiffened using the proposed detail confirmed their acceptable performance up to at least the rotation of 0.04 rad. 6- The proposed detail is applicable to one-way moment resisting frames in its present form. Therefore, further investigations are required for utilizing it in two-way steel moment-frames. Declaration of Competing Interest None. Acknowledgment The authors would like to express their sincere gratitude to Tir & Setoon Pars Industrial Company (T.S.P) for fabricating the full scale test specimens. The assistance of Mr. Etemadi is highly appreciated. References [1] FEMA, Recommended Seismic Design Criteria for New Steel Moment Frame Buildings, FEMA 350, Federal Emergency Management Agency (FEMA), Washington, DC, 2000. [2] FEMA, Recommended Seismic Design Criteria for New Steel Moment Frame Buildings, FEMA 351, Federal Emergency Management Agency (FEMA), Washington, DC, 2000. [3] E.P. Popov, Seismic moment connections for MRFs, J. Constr. Steel Res. 10 (1988 Jan 1) 163–198. [4] K.C. Tsai, S. Wu, E.P. Popov, Cyclic performance of steel beam-column moment joints, Eng. Struct. 17 (8) (1995 Sep 1) 596–602. [5] B. Stojadinović, S.C. Goel, K.H. Lee, A.G. Margarian, J.H. Choi, Parametric tests on unreinforced steel moment connections, J. Struct. Eng. 126 (1) (2000 Jan) 40–49. [6] T. Kim, A.S. Whittaker, A.S. Gilani, V.V. Bertero, S.M. Takhirov, Cover-plate and flange-plate steel moment-resisting connections, J. Struct. Eng. 128 (4) (2002 Apr) 474–482. [7] G. Abdollahzadeh, S.M. Shabanian, Experimental and numerical analysis of beam to column joints in steel structures, Front. Struct. Civ. Eng. 12 (4) (2018 Dec 1) 642–661. [8] W. Wang, Y. Zhang, Y. Chen, Z. Lu, Enhancement of ductility of steel moment connections with noncompact beam web, J. Constr. Steel Res. 81 (2013 Feb 1) 114–123. [9] C.H. Lee, J.H. Jung, M.H. Oh, E.S. Koo, Cyclic seismic testing of steel moment connections reinforced with welded straight haunch, Eng. Struct. 25 (14) (2003 Dec 1) 1743–1753. [10] M. Morrison, D. Schweizer, T. Hassan, An innovative seismic performance enhancement technique for steel building moment resisting connections, J. Constr. Steel Res. 109 (2015 Jun 1) 34–46. [11] S. Momenzadeh, M.T. Kazemi, M.H. Asl, Seismic performance of reduced web section moment connections, Int. J. Steel Struct. 17 (2) (2017 Jun 1) 413–425. [12] C.P. Pantelides, Y. Okahashi, L.D. Reaveley, Experimental investigation of reduced beam section moment connections without continuity plates, Earthquake Spectra 20 (4) (2004 Nov) 1185–1209.
21
[13] C.H. Lee, S.W. Jeon, J.H. Kim, C.M. Uang, Effects of panel zone strength and beam web connection method on seismic performance of reduced beam section steel moment connections, J. Struct. Eng. 131 (12) (2005 Dec) 1854–1865. [14] D.S. Sophianopoulos, A.E. Deri, Parameters affecting response and design of steel moment frame reduced beam section connections: an overview, Int. J. Steel Struct. 11 (2) (2011 Jun 1) 133–144. [15] A.K. Swati, V. Gaurang, Study of steel moment connection with and without reduced beam section, Case Stud. Struct. Eng. 1 (2014 Jun 1) 26–31. [16] M.T. Roudsari, S.H. Moradi, Experimental and numerical assessment of reduced IPE beam sections connections with box-stiffener, Int. J. Steel Struct. 18 (1) (2018 Mar 1) 255–263. [17] D. Dubina, A. Stratan, Behaviour of welded connections of moment resisting frames beam-to-column joints, Eng. Struct. 24 (11) (2002 Nov 1) 1431–1440. [18] X.Y. Liu, Y.Q. Wang, J. Xiong, Y.J. Shi, Investigation on the weld damage behavior of steel beam-to-column connection, Int. J. Steel Struct. 17 (1) (2017 Mar 1) 273–289. [19] M. D'Aniello, R. Tartaglia, S. Costanzo, R. Landolfo, Seismic design of extended stiffened end-plate joints in the framework of Eurocodes, J. Constr. Steel Res. 128 (2017 Jan 1) 512–527. [20] A.B. Francavilla, M. Latour, V. Piluso, G. Rizzano, Design of full-strength full-ductility extended end-plate beam-to-column joints, J. Constr. Steel Res. 148 (2018 Sep 30) 77–96. [21] A. Deylami, M. Salami, Retrofitting of moment connection of double-I built-up columns using trapezoidal stiffeners, Procedia Eng. 14 (2011 Jan 1) 2544–2551. [22] S.M. Dehghan, M.A. Najafgholipour, S.M. Ziarati, M.R. Mehrpour, Experimental and numerical assessment of beam-column connection in steel moment-resisting frames with built-up double-I column, Steel Compos. Struct. 26 (2018 Feb 10) 315–328. [23] AISC, Seismic Provisions for Structural Steel Buildings, AISC/ANSI 341–16, American Institute of Steel Construction (AISC), Chicago, IL, 2016. [24] D. Lee, S.C. Cotton, R.J. Dexter, J.F. Hajjar, Y. Ye, S.D. Ojard, Column Stiffener Detailing and Panel Zone Behavior of Steel Moment Frame Connections. Structural Engineering Report No. ST-01-3.2, Department of Civil Engineering, University of Minnesota, 2002 Jun. [25] S.D. Prochnow, R.J. Dexter, J.F. Hajjar, Y. Ye, S.C. Cotton, Local Flange Bending and Local Web Yielding Limit States in Steel Moment-Resisting Connections. Structural Engineering Report No. ST-00-4, Department of Civil Engineering, University of Minnesota, 2000 Nov. [26] Y. Ye, J.F. Hajjar, R.J. Dexter, S.D. Prochnow, S.C. Cotton, Nonlinear Analysis of Continuity Plate and Doubler Plate Details in Steel Moment Frame Connections. Structural Engineering Report No. ST-00-3, Department of Civil Engineering, University of Minnesota, 2000 Sep. [27] B.H. Hashemi, R.A. Jazany, The continuity plate effect on panel zone ductility with unequal beams height on both sides of column in SMRF, Proc. 14th World Conference Earthquake Engineering, 2008. [28] R. Amani, H. Safari, A. Fakhraddini, Local flange bending and continuity plate requirements in double-Web H-shaped columns, Int. J. Steel Struct. 18 (2018) 199–209. [29] C.C. Chen, C.C. Lin, C.L. Tsai, Evaluation of reinforced connections between steel beams and box columns, Eng. Struct. 26 (13) (2004 Nov 1) 1889–1904. [30] N.E. Shanmugam, L.C. Ting, Welded interior box-column to I-beam connections, J. Struct. Eng. 121 (5) (1995 May) 824–830. [31] T. Kim, B. Stojadinović, A.S. Whittaker, Seismic performance of pre-Northridge welded steel moment connections to built-up box columns, J. Struct. Eng. 134 (2) (2008 Feb) 289–299. [32] M. Gholami, A. Deylami, M. Tehranizadeh, Seismic performance of flange plate connections between steel beams and box columns, J. Constr. Steel Res. 84 (2013 May 1) 36–48. [33] X. Chen, G. Shi, Experimental study of end-plate joints with box columns, J. Constr. Steel Res. 143 (2018 Apr 1) 307–319. [34] S.R. Mirghaderi, S. Torabian, F. Keshavarzi, I-beam to box–column connection by a vertical plate passing through the column, Eng. Struct. 32 (8) (2010 Aug 1) 2034–2048. [35] S. Erfani, A.A. Asnafi, A. Goudarzi, Connection of I-beam to box-column by a short stub beam, J. Constr. Steel Res. 127 (2016 Dec 1) 136–150. [36] E. Jahanbakhti, N. Fanaie, A. Rezaeian, Experimental investigation of panel zone in rigid beam to box column connection, J. Constr. Steel Res. 137 (2017 Oct 1) 180–191. [37] O. Rezaifar, M. Nazari, M. Gholhaki, Experimental study of rigid beam-to-box column connections with types of internal/external stiffeners. Steel and, Compos. Struct. 25 (5) (2017 Dec 10) 535–544. [38] H. Saffari, A.A. Hedayat, G.N. Soltani, New alternatives for continuity plate in Ibeam to box columns connections, Asian J. Civil Eng. (BHRC) 16 (2) (2015) 219–233. [39] AISC, Seismic Provisions for Structural Steel Buildings, AISC/ANSI 341–10, American Institute of Steel Construction (AISC), Chicago, IL, 2010. [40] ABAQUS, Theory Manual, Version 6.14, 2014. [41] ASTM, A370–18, Standard Test Methods and Definitions for Mechanical Testing of Steel Products, ASTM International, West Conshohocken, PA, 2018.
Contents lists available at ScienceDirect
Journal of Constructional Steel Research
An alternative detail for continuity plates in steel beam to box-column moment-connections M.A. Najafgholipour a,⁎, Kianoush Peykari b, S.M. Dehghan a a b
Faculty of Civil and Environmental Engineering, Shiraz University of Technology, Shiraz, Iran Graduate Student of Civil and Environmental Engineering, Shiraz University of Technology, Shiraz, Iran
a r t i c l e
i n f o
Article history: Received 25 June 2019 Received in revised form 25 November 2019 Accepted 25 January 2020 Available online xxxx Keywords: Steel moment-connection Continuity plates Vertical stiffener Box-column Experimental study Finite element model
a b s t r a c t In this study, the use of an internal vertical stiffener in built-up box-columns parallel to the beam web, which turns the column section into a boxed I-section in the connection region to eliminate the common continuity plates, is assessed through an experimental and numerical study. For this purpose, two full-scale exterior Ibeam to box-column moment connections (one with common continuity plates and the other one with a vertical stiffener as an alternative detail) are tested under reversed cyclic loading. The test results in terms of hysteresis moment-rotation curves, strain in critical stations, and observations during the tests indicate that the connection with the vertical stiffener could survive story drifts of up to 4% without any strength loss and brittle failure such as weld fracture. For further investigation of the seismic performance of the proposed detail, Finite Element Modeling (FEM) of the test specimens is also conducted in the FE software ABAQUS and validated using the test results. The results of the numerical study also confirm the satisfactory behavior of the connections with the proposed detail. © 2020 Elsevier Ltd. All rights reserved.
1. Introduction Steel Moment Resisting Frames (MRFs) are well known as ductile earthquake resistant structural systems. They are commonly employed in buildings with different heights and occupancies mainly due to their minimum architectural interference. After Northridge (1994) and Kobe (1995) earthquakes whereby steel MRFs experienced severe damages especially in their connections, the crucial role of beam-column moment connections in the global seismic performance of steel MRFs was confirmed. In this regard, extensive studies were conducted to find the weak points of the existing connections and to make improvements to enhance their seismic performance. These studies led to specific guidelines for retrofitting existing joints and prequalified connections for new buildings [1,2]. Regarding the importance of steel moment connections and their effects on the seismic performance of steel MRFs, numerous experimental and numerical studies have been done to evaluate their behavior under cyclic loads. A significant number of studies have been devoted to the seismic performance of the common types of connections in recent decades [3–7]. In a number of experimental studies, some modifications have been proposed to improve the seismic performance of the connections by enhancing their strength as well as their plastic rotation capacity under reversed cyclic loads [8,9]. Innovative types of connections ⁎ Corresponding author. E-mail address: [email protected] (M.A. Najafgholipour).
https://doi.org/10.1016/j.jcsr.2020.105952 0143-974X/© 2020 Elsevier Ltd. All rights reserved.
with special features have been presented in a number of studies [10,11]. The performance of Reduced Beam Sections (RBS) as an improvement technique in moment resisting connections has also been evaluated in numerical and experimental studies [12–16]. To assess the effects of welding type, welding pattern, and welding damage on the seismic performance of steel moment connections, several studies have been performed [17,18]. A number of analytical and numerical studies have been conducted to explore the force-transferring mechanism in different types of connections leading to the development of design procedures for steel moment connections [19,20]. Due to the widespread use of built-up columns with double I-sections in some regions such as Iran, some experimental and numerical investigations have been conducted to evaluate the different aspects of these connections [21,22]. Wide-flange, box, boxed wide-flange, and flange cruciform columns are permitted to be used in steel MRFs according to AISC 341-16 [23]. Built-up steel box columns are widely used in regions where heavy hot-rolled steel sections are not available. Moreover, due to the inherent geometrical properties of box sections, they are more efficient compared to I-sections especially under bidirectional moments. In addition, beams can be joined to all sides of box columns simultaneously by means of appropriate types of moment connections. Furthermore, box sections are composed of stiffened elements reducing the widthto-thickness ratio requirements for seismic compact sections. A welded steel I-beam to box-column moment resisting connection is schematically illustrated in Fig. 1-a. The mechanism of transferring
2
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 1. Steel welded I-beam to box-column moment connection, a- Schematic view of the connection, b- Moment transfer mechanism in the joint.
internal actions in the joint region is shown in Fig. 1-b. The continuity plates across the column sections have a key role in transferring the bending moment from the adjacent beams to the column and vice versa as an internal diaphragm. A pair of tensile and compressive forces induced in the beam flanges (Puf), equivalent to the bending moment in the beam (Mu), are transferred to the joint panel through the in-plane action of the continuity plates. In the absence of continuity plates, the tensile and compressive forces cause a local bending moment in the column flanges which in turn leads to excessive flexural deformations. These deformations may be accompanied by the local yielding or local buckling of the column flanges. Various studies have been done to investigate the role of different types of stiffeners and details in the seismic behavior of steel moment connections with wide-flange and box columns. In a comprehensive research project, the nonlinear behavior of moment resisting beamcolumn joints with various detailing of stiffeners (doubler plates and continuity plates) was assessed [24–26]. As part of this research, Lee et al. [24] conducted experimental studies on six interior connections with different detailing of wide-flange column stiffeners. Afterward, Prochnow et al. [25], by focusing on the role of continuity plates in steel beam to column moment connections, conducted pull tests on beam flange to column flange specimens. In their report, a complete history of the investigations on the behavior of continuity plates and related code provisions can be found. Subsequently, in a numerical study, Ye et al. [26] assessed the nonlinear behavior of steel beam-column moment connections focusing on the role of doubler plates and continuity plates on their seismic behavior. For this purpose, they numerically investigated both previously conducted pull tests on column flange subassemblies and tests on beam-column connections. They also evaluated the behavior of continuity plates and doubler plates in wide-flange and box columns through a parametric numerical study. The analysis results showed that using continuity plates reduced the bending stress and strains in column flanges considerably. In another study, Ahmady Jazany and Hosseini Hashemi [27] conducted cyclic tests on six interior connections to assess the influence of the continuity plate configuration on the seismic behavior of the panel zone with unequal beam depths. The test results indicated that the configuration of the continuity plates not only affected the stiffness of the sub-assemblages but also influenced their plastic rotation capacity. Recently, through an analytical and numerical study, Amani et al. [28] investigated the local bending of column flange and the requirements of the continuity plates in double-web H-shaped columns. Despite the various advantages of box sections over other section types and their widespread use in some regions of the world, relatively few studies have been specifically conducted on the seismic performance of the steel beam to box-column connections. As an example,
Chen et al. [29] evaluated the performance of welded I-beam to boxcolumn moment connections. The results of their study demonstrated the role of internal diaphragms in the seismic performance of I-beams to box-columns moment resisting connections. As another early study on the beam to box-column connections, Shanmugam and Ting [30] performed tests on interior joints. Kim et al. [31] conducted tests on pre-Northridge steel beam to box-column connections with internal diaphragms. The test results as well as the numerical model of the connections led to some welding details of the continuity plates. Flange-plate connections of I-beams to box-columns were assessed by Gholami et al. [32] through an experimental and numerical study. The results of this study indicated that the connections were able to satisfy the AISC criteria for special moment connections. Chen and Shi [33] performed an experimental investigation on the seismic performance of a new end-plate joint with box-columns. The test results revealed that the proposed prefabrication technique had a satisfactory performance under cyclic loads. Mirghaderi et al. [34] proposed a through-plate connection for steel I-beam to box-column moment connections. They evaluated the behavior of the connection in an experimental study and investigated the dimensions and shape of the plate by means of a numerical model. A novel short-stub-beam connection was proposed by Erfani et al. [35] for I-beam to box-column connections. Finite Element (FE) method was employed to evaluate the seismic performance of this connection. Recently, Jahanbakhti et al. [36] evaluated the behavior of the panel zones in beam to box-column connections through an experimental study on three specimens. They examined the elimination of the continuity plates from the connection by providing a minimum column flange thickness. Considering the complicated welding of continuity plates in box-columns due to lack of accessibility, a number of researchers have proposed alternatives for continuity plates in box columns. For instance, Rezaifar et al. [37] assessed the seismic behavior of interior connections with external stiffeners as an alternative detail to eliminate the internal diaphragms. In a numerical study, Saffari et al. [38] examined a new detail as an alternative for common continuity plates. 1.1. AISC provisions for continuity plates in boxed wide-flange columns According to AISC 341-16 [23], in steel I-beam to boxed wide-flange column moment resisting connections and under some conditions, continuity plates are not required and the joint panel is able to transfer the internal induced actions properly. According to the code provisions, the column flange should necessarily satisfy the following limiting condition: t lim ¼
bbf 12
ð1Þ
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
3
Fig. 2. Fabrication sequence of the box-column section with a common continuity plates.
where, tlim is the minimum permitted thickness of the column flange and bbf is the width of the beam flange. In addition, the column flange and web should be designed under the concentrated force in the beam flange caused by the probable flexural moment capacity of the beam. For this purpose, the limit states of the local yielding, local crippling, sidesway buckling, and compression buckling must be checked for the column web. Moreover, the column flange should be designed for local bending. In the former edition of AISC 341-10 [39], the continuity plates could be eliminated in boxed wide-flange columns if the column flange thickness (tcf) satisfied the following inequalities:
t cf ≥
bbf 12
ð2Þ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi v" u # u bbf bcf F yb Ryb t 1:8bbf t bf 1− 2 bbf − t cf ≥0:4 4 F yc Ryc bcf
ð3Þ
where bbf and bcf are the widths of the beam and column flanges, respectively, and tbf is the thickness of the beam flange. Also, Fyb and Fyc are the specified minimum yield strengths of the beam and column, respectively, and Ry is Ratio of the expected yield stress to the specified minimum yield stress. 1.2. Research significance Although the horizontal continuity plates can be easily welded in wide-flange columns, proper welding of all the edges of the continuity plate to a box-column is difficult due to lack of accessibility. The fabrication sequence of a steel beam to box-column connection with common continuity plates is illustrated in Fig. 2 Accordingly, the welding of the fourth edge of the continuity plates to the column cannot be done properly using common welding procedures. To eliminate the commonly used continuity plates in steel I-beam to box-column moment connections, an alternative detail inspired from the aforementioned points about the necessity of continuity plates in
Fig. 3. Welded steel I-beam to box-column connection, a- Common detail with horizontal continuity plates, b- The proposed detail with a vertical stiffener.
4
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 4. Moment transfer mechanism in the connections with vertical stiffener.
Fig. 5. Fabrication sequence of the box-column section with a vertical stiffener.
boxed wide-flange columns is proposed in this study. For this purpose, a vertical stiffener is used in the column parallel to the beam web (see Fig. 3). Consequently, the column section becomes similar to a
wide-flange boxed section in the joint region. By satisfying the geometrical requirements and controlling the limit states, the elimination of the horizontal continuity plates may be possible by using a vertical stiffener.
Fig. 6. Dimensions of the test specimens.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 7. Fabrication details of the RBS-H connection, a- Side view, b- Top view, c- Dimensions of the continuity plates, d- Shear tab dimensions, e- Beam section.
5
6
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 8. Fabrication details of the RBS-V connection, a- Side view, b- Top view, c- Dimensions of the Vertical stiffener, d- Shear tab dimensions, e- Beam section.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
7
Fig. 9. Fabrication of the test specimens in shop.
Following the fabrication procedure of the proposed connection as illustrated in Fig. 5, its applicability is confirmed for practical cases. To evaluate the efficiency of this idea, the seismic performance of the I-beam to box-column moment resisting connections with the proposed detail is assessed through an experimental program. For this purpose, two full-scale exterior beam-column moment connections (one with common horizontal continuity plates and the other one with a vertical stiffener in the column) are tested under cyclic loadings. For further evaluation of the proposed technique and detailed investigation on the behavior of the test specimens, a Finite Element (FE) numerical study is implemented in the software ABAQUS [40] and validated with the test results. Fig. 10. Tensile test of the steel plates according to ASTM A370 [41].
Table 1 Mechanical properties of the steel plates and sections. Test specimen
Yield stress (MPa)
Ultimate strength (MPa)
Elongation (%)
HEA280 (Web) HEA280 (Flange) Plate 20 mm Plate 15 mm
368.8 298.1 300.7 292.2
521.1 474.9 475.1 476.7
24.8 28.9 24.2 20.1
Furthermore, by using the RBS detail in the beam, it is ensured that the plastic hinge in the beam occurs at a certain distance from the column face. A schematic view of the force transmission in the connections with vertical stiffness is illustrated in Fig. 4. Accordingly, the tensile and compressive forces induced in the beam flanges (Puf) result in internal forces in the column webs (Vcw) and the vertical stiffener (Vs).
2. Experimental program Two full-scale exterior joints were fabricated and tested in this study. The details of the test program including the fabrication process of the specimens and the test set-up as well as the results of the experimental study are presented here. 2.1. Test specimens The test specimens were exterior beam-column connections isolated from a steel MRF with the story height of 2.6 m and span length of 6 m (see Fig. 6). The connections were made up of a 300 mm × 300 mm built-up steel box column with the flange thickness of 20 mm. HEA280 was employed as the beam with the same height and flange width of 280 mm. The flange and web thicknesses of the beam section were 18 mm and 10.5 mm, respectively. Continuous groove weld was used to assemble the plates of the built-up box-column. The only difference between the specimens was that in the reference connection (RBS-H), a pair of square
Fig. 11. Test setup for the cyclic test of the exterior moment connections.
8
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 12. Loading protocol of AISC 341-16 in the reversed cyclic test of the steel moment connections [23].
Fig. 13. Configuration of the mounted strain gauges on the connections, a- RBS-H connection, b- RBS-V connections.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
9
Fig. 14. Hysteresis moment-rotation curve of the RBS-H connection.
continuity plates within the column (260 mm × 260 mm × 20 mm) were utilized in the plane of the beam flanges as the common detail for the moment resisting connections. In the other connection (RBS-V), a vertical stiffener in the column axis with the dimensions of 260 mm × 500 mm × 20 mm in the plane of the beam web was employed instead of the continuity plates. The vertical stiffener in the latter connection was extended 110 mm beyond the top and bottom flanges of the beam which makes the column section in the connection region similar to a boxed wide-flange section. The continuity plates and the vertical stiffener were jointed to the columns using groove welds. The fabrication and welding details of the test specimens including the dimensions of the different parts are illustrated in Figs. 7 and 8 for RBS-H and RBS-V, respectively.
It should be noted that the beam and column dimensions were determined based on a preliminary design according to AISC341–16 seismic provisions [23] to satisfy the strong column-weak beam criterion. The joint panel zone was designed for the expected shear stress due to the expected plastic moment of the beam to the column. Moreover, the shear tab was designed for the expected shear force developed in the beam corresponding to the plastic flexural strength of the beam section. In addition, to avoid extensive flexural plastic strain in the beam just adjacent to the column face, a Reduced Beam Section (RBS) was used with a 5% reduction of the flange width of the beam (see Figs. 7-b and 6-b). As is shown in Fig. 9, the connections were shop-welded by qualified welders. During fabrication, Ultrasonic Testing (UT) was performed on
Fig. 15. Measured strain at different levels of beam rotation.
10
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
all the groove welds to assure their quality and penetration. Finally, the specimens were painted with whitewash to monitor the inelastic deformations during the test. 2.2. Material properties To determine the mechanical properties of the steel plates and sections, tensile test based on ASTM A370 [41] was performed on standard coupons cut from the different parts of the connections (see Fig. 10). Using the obtained stress-strain curves from each tensile test, the key mechanical properties were determined as listed in Table 1. These properties were used in the numerical study and the verification phase. 2.3. Test setup and procedure To evaluate the seismic performance of the connections with common and alternative continuity plate details, reversed cyclic tests were conducted on the test specimens (Fig. 11). Under lateral load, it is assumed that contra-flexure points in the beams and columns of a MRF are located almost at the mid-stories and mid-spans of the frame. Thus, pin supports were utilized at the top and bottom of the column and at the beam tip as well (see Fig. 11). The cyclic displacement control load with the ascending amplitude following the standard loading protocol prescribed in AISC 341-16 [23] (Fig. 12) was applied to the end of the beam (3000 mm from the
column axis) by means of a 250 kN semi-automatic hydraulic actuator. However, in this study, the tests were terminated at the beam rotation of 0.04 rad. The displacement of the loading point and the applied load during the test were measured by means of a displacement transducer (LVDT) and a load cell, respectively. To avoid any lateral displacement and instability of the beam during the test, a lateral support was provided at a distance of 1500 mm from the column face. Unlike Reinforced Concrete (RC) beam to column moment connections, the axial load does not affect the behavior of the steel beam to column moment connections, especially in cases that the panel zone is designed to transfer the moment from the adjacent beams. Therefore, in this study, the connections were tested in the absence of column axial load. To measure the strain in regions where the yielding of the section was anticipated or stress concentration was probable, some strain gauges were mounted on the test specimens with the configuration presented in Fig. 13. It should be noted that due to some technical problems, a number of the strain gauges were disrupted and their data were not available. 2.4. Results of the experimental study The results of the cyclic tests on the connections including hysteresis load-displacement curves (moment-rotation curves) as well as observations during the tests are presented in the subsequent sections. It should be noted that the rotations reported in the subsequent sections, are the
Fig. 16. Observed behavior of the specimen RBS-H, a- 0.01 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
11
Fig. 17. The beam to column groove weld at the end of the test.
chord rotation of the beam which is the ratio of the vertical deflection of the beam end to the distance between the loading point and the column centerline. Moreover, the plastic moment strength of the beam (Mp) specified on the moment-rotation curves was determined as follows: L 2850 Mp ¼ Z RBS F yb b ¼ ð1065 298:1Þ ¼ 350:7 kN−m 2580 LRBS
ð4Þ
where ZRBS is the plastic section modulus of the beam at reduced section, and LRBS and Lb are the distances from the loading point to the center of the RBS and column centerline, respectively, 2.4.1. Reference connection with horizontal continuity plates (RBS-H) The experimental hysteresis moment-rotation curve of the RBS-H connection is plotted in Fig. 14. As was expected, the reference connection (RBS-H) with common continuity plates in the joint panel was able to survive the rotation of 0.04 rad without any strength loss. The initial stiffness of the connection was 27,618 kN-m/rad. The peak moment capacity of the connection corresponding to the rotation of 0.04 rad was 540 kN-m which was more than the expected plastic flexural strength of the beam (Mp = 350.7 kN-m). Hence, the connection could meet the AISC 341–16 requirements as a special moment resisting connection. The measured strains at different levels of the story drift are plotted in Fig. 15. Moreover, the test observations including the local
deformations of the beam and column sections are illustrated in Fig. 16. The joint behavior during the test and the measured strains at different stations indicate that the test specimen remained elastic up to the rotation of 0.01 rad and no obvious local excessive deformation was observed in this rotation (Fig. 16-a). Strain gauge number 3 which measured the strain on the top flange of the beam in the RBS region showed the first yielding in the test specimen at the rotation of 0.015 rad. Excessive local deformations with the flaking of whitewash (which may be a sign of yielding of the beam flange in the RBS region) were also observed in the test specimen (Fig. 16-b). According to the measured strain in the strain gauges 1 and 2, the top of the beam next to the column face yielded from the drift ratio of 1% to 2%. By increasing the rotation amplitude up to 0.04 rad, the flaking of whitewash increased on the beam flanges and developed in the depth of the beam at the RBS zone (Fig. 16-c and -d) and the measured strain on the beam flanges increased to around 5 times as much as the steel yield strain. The concentration of the plastic hinge in the beam and especially at the RBS region without any fracture in the beam to column groove welds (Fig. 17) confirmed the acceptable performance of the joint as a prequalified special moment resisting connection.
2.4.2. Proposed connection with a vertical stiffener (RBS-V) The hysteresis moment-rotation curve of the connection RBS-V with a vertical stiffener in the column is presented in Fig. 18. Similar to the
Fig. 18. Hysteresis moment-rotation curve of the RBS-V connection.
12
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
reference connection, the moment-rotation curve indicates that this connection could reach the rotation of 0.04 rad with no obvious strength loss. In addition, its maximum moment strength corresponding to the rotation of 0.04 rad was 549 kN-m which was more than the plastic moment of the beam. This means that the beam was able to reach its plastic flexural moment capacity. Consequently, the connection with the alternative detail was able to meet the requirements of the special moment resisting connections according to AISC341-16 [23]. The initial stiffness of the connection was 27,126 kN-m/rad. The comparison of the hysteresis moment-rotation curves of the RBS-H and RBS-V joints confirms that replacing the horizontal diaphragms in the box-columns with a vertical stiffener in the column axis had a slight influence on the global performance of the beam-column connection (see Fig. 19). The test observations and the measured strains in the specified stations at different levels of beam rotation are illustrated in Figs. 20 and 21, respectively. Considering the moment-rotation curve as well as the measured strains and performance of the test specimen, the connection remained elastic up to the rotation of around 0.01 rad (see Fig. 20-a). Further rotation of the beam started the plastic strain and led to the development of whitewash flaking on the beam flanges in the RBS zone. The strain gauge 3 mounted in this region confirmed the yielding of the top flange of the beam. The plastic strain in the beam was enhanced by increasing the displacement amplitude up to the beam rotation of 0.04 rad. As is shown in Fig. 20-c and -d, excessive whitewash flaking was confirmed with the measured strain in the RBS region which showed strains up to around 4 times as much as the steel yield strain. Despite some relatively minor yielding of the beam flange next to the column face as recorded with strain gauges 2 and 3, the measured strain was much less than that of the RBS-H connection. However, comparing the data from the strain gauges on the specimen and the test observations confirm that the plastic hinge was concentrated in the RBS region. It should be noted that although some minor whitewash flaking in the column web in the level of the beam flanges was observed, the strain gauges did not show yielding at these stations. Finally, the inspection of the groove welds in the test specimen indicated that no weld fracture occurred in these regions (Fig. 22).
3. Numerical study To further evaluate the efficiency of the proposed detail in connections with various geometrical properties, a numerical FE study was done. For this purpose, first, the model was validated using the results of the experimental study. Then, a parametric study was conducted to assess the behavior of the proposed alternative detail in connections with different beam and column dimensions. 3.1. Validation of the FE model To validate the FE model and to ensure the accuracy of its results, both tested connections (RBS-H and RBS-V) were simulated in the FE software ABAQUS [40] and analyzed under cyclic loading. The details of the conducted numerical modeling procedure of the connections are explained subsequently. • Constitutive material model for steel According to the material tests conducted in the experimental study and presented in Section 2.2, the stress-strain curves corresponding to the different parts of the specimens along with the key mechanical properties were available. Consequently, to define the uniaxial stressstrain curve for different parts of the models, the measured mechanical properties (i.e. the yield stress and ultimate strength) as well as the corresponding strains were employed. In addition, von Mises yield criterion were utilized to simulate the nonlinear behavior of the models under multi-axial stress states. Furthermore, although no local buckling was observed in the test specimens up to the last cycle of the loading protocol, an initial imperfection of 0.001 L was defined for the models to find out any local buckling in the numerical model. • Geometry of the models and element types The geometrical dimensions and fabrication details of the connections (Figs. 7 and 8) were followed for the simulation of the joints. The FE models of the connections are illustrated in Fig. 23. A quadratic
Fig. 19. Comparing the hysteresis moment-rotation curves of the RBS-H and RBS-V specimens.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 20. Observed behavior of the specimen RBS-V, a- 0.01 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
Fig. 21. The measured strains corresponding to different loading cycles in connection RBS-V.
13
14
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 22. The beam to column groove welds at the end of the test.
20-node solid element (denoted as C3D20R) with three translational Degrees of Freedom (DOF) in each node was employed for the analysis of the connections. Following a mesh sensitivity analysis, the element sizes ranging from 10 mm to 20 mm were used. In total, around 39,000 elements and 155,000 nodes were generated for the analysis of each connection (see Fig. 23-a). It should be noted that since no defect and failure was observed in the welds during the tests and the quality of welding was verified through frequent inspection and quality control tests, the welds were not modeled as a separate part in the FE models. Instead, the tie constraint in ABAQUS software was utilized. • Loading and boundary conditions The restrained DOFs in the FE model to simulate the boundary conditions of the test specimens are illustrated in Fig. 23-b. As described in the test program, a pair of pins were put at the top and bottom faces of the column. Hence, all of the translational DOFs of the two lines perpendicular to the plane of the connection at the top and bottom faces of the
column were restrained. To model the lateral support, the corresponding nodes were restrained against out-of-plane translation. The displacement cyclic loading was also applied to the beam tip following the loading protocol employed in the experimental study. • Analysis results
The results of the nonlinear analysis of both the RBS-H and RBS-V connections are presented in terms of hysteresis moment-rotation curves as well as distribution of the plastic strain in the models. As shown in Fig. 24, the comparison of the hysteresis moment-rotation curves obtained from the numerical and experimental studies shows the acceptable agreement of the results. The numerical model not only was able to predict the ultimate strength but also simulated the connection behavior reasonably in linear and nonlinear phases. Comparing the distribution and development of the plastic strains in different levels of the lateral drift with the test observations confirms
Fig. 23. The FE model of the specimens, a- The generated elements, b- The boundary conditions and loading defined in the FE model
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 24. Comparing the hysteresis moment-rotation curves obtained from the numerical study and the tests, a- RBS-H connection, b- RBS-V connection.
Fig. 25. Plastic strain distribution and development in the RBS-H connection at different levels of beam rotation, a- 0.015 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
15
16
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 26. Plastic strain distribution and development in the RBS-V connection at different levels of beam rotation, a- 0.015 rad, b- 0.02 rad, c- 0.03 rad, d- 0.04 rad.
Table 2 The geometrical details of the connections in the numerical study. Connection
Column size, bcf (mmxmm)
Column flange thickness, tcf (mm)
Beam flange width, bbf (mm)
Detail of column stiffener
RBS-B-300-280 RBS-H-300-280 RBS-V-300-280 RBS-B-400-200 RBS-H-400-200 RBS-V-400-200 RBS-B-400-280 RBS-H-400-280 RBS-V-400-280 RBS-B-400-340 RBS-H-400-340 RBS-V-400-340 RBS-B500-280 RBS-H-500-280 RBS-V-500-280
300 × 300
20
280
400 × 400
25
200
400 × 400
25
280
400 × 400
25
340
500 × 500
30
280
None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener None Continuity plate Vertical stiffener
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
17
Fig. 27. Hysteresis moment-rotation curves of the connections.
the ability of the FE model in simulating the behavior of the connections in the nonlinear phase. As illustrated in Figs. 25 and 26 and as observed in the tests, the beginning of yielding in the RBS-H specimen occurred on the beam flange at the rotation of around 0.015 rad. By increasing the lateral drift, plastic strains developed in the RBS region and in the depth of the beam. Finally, at the rotation of 0.04 rad, the beam section yielded completely and the plastic hinge formed. No local buckling was observed in the numerical model. A similar trend occurred in the RBS-V connection
where no plastic strain was observed in the test specimen up to the beam rotation of 0.015 rad. By further increasing the beam's rotation, the plastic strains which were concentrated in the RBS zone developed. 3.2. Parametric study To assess the influence of effective geometrical variables on the seismic performance of beam to box-column moment connections with the
18
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
proposed stiffening detail, a parametric study was conducted by means of a verified FE model. The details of the investigated connections and analysis results are presented in the subsequent sections. 3.2.1. Defining the models in the parametric study The column flange dimensions (width (bcf) and thickness (tcf)) as well as the beam flange width (bbf) were considered as geometrical variables which can affect the efficiency of the proposed detail. For this purpose, square box columns with side dimensions of 300 mm, 400 mm, and 500 mm were considered. According to the seismic provisions of ASCE 341-16 [23], the beams and columns of special MRFs must be seismic compact to minimize the probability of local buckling in the section components. Therefore, the thicknesses of the box-columns were considered equal to 20 mm, 25 mm, and 30 mm for the columns with the side dimensions of 300 mm, 400 mm, and 500 mm, respectively. To investigate the influence of the beam flange width on the seismic behavior of the connections, beams with three flange widths of 200 mm, 280 mm, and 350 mm but with the same moment capacity obtained by tuning the RBS dimensions were used. In each case, a connection with an unstiffened panel zone was analyzed to show the crucial role of the continuity plates and to confirm the necessity of an appropriate alternative detail for continuity plates. A summary of the properties of the investigated connections and their designations is listed in Table 2. 3.2.2. Analysis results The hysteresis moment-rotation curves of the investigated connections are plotted in Fig. 27. In each diagram, the curves corresponding to stiffened connections (with common continuity plates and a vertical stiffener) and an unstiffened one are presented. The plastic moment strengths of the beams are also specified on the diagrams.
In almost all cases, the beams of the bare connections did not reach their plastic flexural strength. According to the strain distribution in the bare joints at the story drift angle of 0.04 rad, the column flanges experienced yielding accompanied by excessive flexural outof-plane deformations. This was more significant in the connections with larger bcf /bbf. For instance, the developed plastic strain in the column flange is well demonstrated in Fig. 27-a for the RBS-B-400200 connection with the highest bcf /bbf (2.0) among the studied models. This phenomenon in unreinforced connections confirms the necessity of an appropriate stiffening detail for box columns in moment resisting connections. Moreover, the analysis results of the stiffened connections with both of the abovementioned techniques in terms of the momentrotation curves and plastic strain distribution confirm their satisfactory cyclic behavior as special moment resisting connections. According to Fig. 27, the stiffened connections were able to reach the plastic flexural strength of the beam with no strength degradation up to the drift of 0.04 rad. Additionally, although in using the two strategies there was a minor difference in strain distribution in the stiffened connections, their global performance was almost the same. Considering the strain distribution in the stiffened connections that are presented in Fig. 28-b and -c for RBS-H-400-200 and RBS-V400-200, respectively, the plastic strain was mostly concentrated in the RBS region which is expected in special moment connections. Accordingly, unlike the unreinforced models, in the stiffened connections the peak flexural strain in the beam flange occurred in the RBS region. Additionally, the analysis results showed that no plastic strain developed in the panel zone of the stiffened connections. Moreover, comparing the distribution of the shear stress in the column webs and the vertical stiffener, showed that the shear force induced in the vertical stiffener was around 50% to 65% of the total
Fig. 28. Plastic strain distribution in the connections at the rotation of 0.04 rad, a- RBS-B-400-200, b- RBS-H-400-200, c- RBS-V-400-200.
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
19
(a)
(b)
(c)
(d)
(e)
(f)
Fig. 29. Plastic strain distribution over the beam flange at the rotation of 0.04 rad, a- RBS-H-400-200, b- RBS-V-400-200, c- RBS-H-400-280, d- RBS-V-400-280, e- RBS-H-400-340, f- RBS-V400340.
shear force that was transferred in the panel zone. The proportions of the shear force in the vertical stiffener and the column web depended on some geometrical ratios such as bcf/bbf and bcf/tcf. The plastic strain distributions over the beam flange are plotted for stiffened (with vertical stiffener and horizontal continuity plates) connections with column size of 400 mm, see Fig. 29 Accordingly, in the vertically stiffened connections, the plastic strain in the RBS region was more than that developed near the column face. Furthermore, the average plastic strain over the beam flange has been plotted in Fig. 30.
Accordingly, in almost all connections, the peak plastic strain was in the RBS region. 4. Conclusions An alternative detail for common continuity plates in steel beam to box-column moment resisting connections, inspired from continuity plate provisions of boxed wide-flange columns, was proposed in this study. In this regard, instead of horizontal continuity plates, a vertical
20
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
Fig. 30. Plastic strain distribution along the beam axis, a- RBS-H-400-200, b- RBS-V-400-200, c- RBS-H-400-280, d- RBS-V-400-280, e- RBS-H-400-340, f- RBS-V-400-340
column stiffener was employed in the panel zone. In addition, to ensure the occurrence of plastic strain concentration at a certain distance from the column face, the RBS detail was used. To assess the efficiency of the proposed technique, cyclic tests were conducted on two full-scale exterior isolated joints (one with common horizontal continuity plates known as a code-based detail and the other one with a vertical stiffener as the proposed detail). In addition, a numerical study was carried out to investigate the performance of the proposed stiffening method in connections
with various geometrical properties. The major findings of this study are presented as follows: 1- The test results in terms of the moment-rotation curves, measured strains at key stations, and test observations indicated that the alternative stiffener was able to transfer properly the induced internal actions in the panel zone and the connection was able to survive the story drift angle of 0.04 rad with no strength degradation and undesirable brittle fracture in the groove welds. In other words, the
M.A. Najafgholipour et al. / Journal of Constructional Steel Research 167 (2020) 105952
connection with a vertical column stiffener meets the requirements of special steel moment resisting connections. 2- The test observations and the measured strains in both connections indicated that the plastic strain was mostly concentrated in the RBS zone where the strain on the beam flange reached up to at least 4 times as much as the yield strain of steel. 3- Considering the welding procedure of the two stiffening strategies shows that the alternative vertical stiffener is more practical than common continuity plates in built-up box columns. 4- The results of the numerical study on unstiffened joint by using a validated FE model confirmed the need of an appropriate stiffening detail for the panel zone of the box columns. In unreinforced connections, a significant yielding of the column flange accompanied by excessive flexural out-of-plane deformations was observed preventing the beams from reaching their plastic flexural strength. This phenomenon was more prominent in connections with higher bcf/bbf. 5- The results of the numerical study on the connections designed with different values of bcf/bbf and stiffened using the proposed detail confirmed their acceptable performance up to at least the rotation of 0.04 rad. 6- The proposed detail is applicable to one-way moment resisting frames in its present form. Therefore, further investigations are required for utilizing it in two-way steel moment-frames. Declaration of Competing Interest None. Acknowledgment The authors would like to express their sincere gratitude to Tir & Setoon Pars Industrial Company (T.S.P) for fabricating the full scale test specimens. The assistance of Mr. Etemadi is highly appreciated. References [1] FEMA, Recommended Seismic Design Criteria for New Steel Moment Frame Buildings, FEMA 350, Federal Emergency Management Agency (FEMA), Washington, DC, 2000. [2] FEMA, Recommended Seismic Design Criteria for New Steel Moment Frame Buildings, FEMA 351, Federal Emergency Management Agency (FEMA), Washington, DC, 2000. [3] E.P. Popov, Seismic moment connections for MRFs, J. Constr. Steel Res. 10 (1988 Jan 1) 163–198. [4] K.C. Tsai, S. Wu, E.P. Popov, Cyclic performance of steel beam-column moment joints, Eng. Struct. 17 (8) (1995 Sep 1) 596–602. [5] B. Stojadinović, S.C. Goel, K.H. Lee, A.G. Margarian, J.H. Choi, Parametric tests on unreinforced steel moment connections, J. Struct. Eng. 126 (1) (2000 Jan) 40–49. [6] T. Kim, A.S. Whittaker, A.S. Gilani, V.V. Bertero, S.M. Takhirov, Cover-plate and flange-plate steel moment-resisting connections, J. Struct. Eng. 128 (4) (2002 Apr) 474–482. [7] G. Abdollahzadeh, S.M. Shabanian, Experimental and numerical analysis of beam to column joints in steel structures, Front. Struct. Civ. Eng. 12 (4) (2018 Dec 1) 642–661. [8] W. Wang, Y. Zhang, Y. Chen, Z. Lu, Enhancement of ductility of steel moment connections with noncompact beam web, J. Constr. Steel Res. 81 (2013 Feb 1) 114–123. [9] C.H. Lee, J.H. Jung, M.H. Oh, E.S. Koo, Cyclic seismic testing of steel moment connections reinforced with welded straight haunch, Eng. Struct. 25 (14) (2003 Dec 1) 1743–1753. [10] M. Morrison, D. Schweizer, T. Hassan, An innovative seismic performance enhancement technique for steel building moment resisting connections, J. Constr. Steel Res. 109 (2015 Jun 1) 34–46. [11] S. Momenzadeh, M.T. Kazemi, M.H. Asl, Seismic performance of reduced web section moment connections, Int. J. Steel Struct. 17 (2) (2017 Jun 1) 413–425. [12] C.P. Pantelides, Y. Okahashi, L.D. Reaveley, Experimental investigation of reduced beam section moment connections without continuity plates, Earthquake Spectra 20 (4) (2004 Nov) 1185–1209.
21
[13] C.H. Lee, S.W. Jeon, J.H. Kim, C.M. Uang, Effects of panel zone strength and beam web connection method on seismic performance of reduced beam section steel moment connections, J. Struct. Eng. 131 (12) (2005 Dec) 1854–1865. [14] D.S. Sophianopoulos, A.E. Deri, Parameters affecting response and design of steel moment frame reduced beam section connections: an overview, Int. J. Steel Struct. 11 (2) (2011 Jun 1) 133–144. [15] A.K. Swati, V. Gaurang, Study of steel moment connection with and without reduced beam section, Case Stud. Struct. Eng. 1 (2014 Jun 1) 26–31. [16] M.T. Roudsari, S.H. Moradi, Experimental and numerical assessment of reduced IPE beam sections connections with box-stiffener, Int. J. Steel Struct. 18 (1) (2018 Mar 1) 255–263. [17] D. Dubina, A. Stratan, Behaviour of welded connections of moment resisting frames beam-to-column joints, Eng. Struct. 24 (11) (2002 Nov 1) 1431–1440. [18] X.Y. Liu, Y.Q. Wang, J. Xiong, Y.J. Shi, Investigation on the weld damage behavior of steel beam-to-column connection, Int. J. Steel Struct. 17 (1) (2017 Mar 1) 273–289. [19] M. D'Aniello, R. Tartaglia, S. Costanzo, R. Landolfo, Seismic design of extended stiffened end-plate joints in the framework of Eurocodes, J. Constr. Steel Res. 128 (2017 Jan 1) 512–527. [20] A.B. Francavilla, M. Latour, V. Piluso, G. Rizzano, Design of full-strength full-ductility extended end-plate beam-to-column joints, J. Constr. Steel Res. 148 (2018 Sep 30) 77–96. [21] A. Deylami, M. Salami, Retrofitting of moment connection of double-I built-up columns using trapezoidal stiffeners, Procedia Eng. 14 (2011 Jan 1) 2544–2551. [22] S.M. Dehghan, M.A. Najafgholipour, S.M. Ziarati, M.R. Mehrpour, Experimental and numerical assessment of beam-column connection in steel moment-resisting frames with built-up double-I column, Steel Compos. Struct. 26 (2018 Feb 10) 315–328. [23] AISC, Seismic Provisions for Structural Steel Buildings, AISC/ANSI 341–16, American Institute of Steel Construction (AISC), Chicago, IL, 2016. [24] D. Lee, S.C. Cotton, R.J. Dexter, J.F. Hajjar, Y. Ye, S.D. Ojard, Column Stiffener Detailing and Panel Zone Behavior of Steel Moment Frame Connections. Structural Engineering Report No. ST-01-3.2, Department of Civil Engineering, University of Minnesota, 2002 Jun. [25] S.D. Prochnow, R.J. Dexter, J.F. Hajjar, Y. Ye, S.C. Cotton, Local Flange Bending and Local Web Yielding Limit States in Steel Moment-Resisting Connections. Structural Engineering Report No. ST-00-4, Department of Civil Engineering, University of Minnesota, 2000 Nov. [26] Y. Ye, J.F. Hajjar, R.J. Dexter, S.D. Prochnow, S.C. Cotton, Nonlinear Analysis of Continuity Plate and Doubler Plate Details in Steel Moment Frame Connections. Structural Engineering Report No. ST-00-3, Department of Civil Engineering, University of Minnesota, 2000 Sep. [27] B.H. Hashemi, R.A. Jazany, The continuity plate effect on panel zone ductility with unequal beams height on both sides of column in SMRF, Proc. 14th World Conference Earthquake Engineering, 2008. [28] R. Amani, H. Safari, A. Fakhraddini, Local flange bending and continuity plate requirements in double-Web H-shaped columns, Int. J. Steel Struct. 18 (2018) 199–209. [29] C.C. Chen, C.C. Lin, C.L. Tsai, Evaluation of reinforced connections between steel beams and box columns, Eng. Struct. 26 (13) (2004 Nov 1) 1889–1904. [30] N.E. Shanmugam, L.C. Ting, Welded interior box-column to I-beam connections, J. Struct. Eng. 121 (5) (1995 May) 824–830. [31] T. Kim, B. Stojadinović, A.S. Whittaker, Seismic performance of pre-Northridge welded steel moment connections to built-up box columns, J. Struct. Eng. 134 (2) (2008 Feb) 289–299. [32] M. Gholami, A. Deylami, M. Tehranizadeh, Seismic performance of flange plate connections between steel beams and box columns, J. Constr. Steel Res. 84 (2013 May 1) 36–48. [33] X. Chen, G. Shi, Experimental study of end-plate joints with box columns, J. Constr. Steel Res. 143 (2018 Apr 1) 307–319. [34] S.R. Mirghaderi, S. Torabian, F. Keshavarzi, I-beam to box–column connection by a vertical plate passing through the column, Eng. Struct. 32 (8) (2010 Aug 1) 2034–2048. [35] S. Erfani, A.A. Asnafi, A. Goudarzi, Connection of I-beam to box-column by a short stub beam, J. Constr. Steel Res. 127 (2016 Dec 1) 136–150. [36] E. Jahanbakhti, N. Fanaie, A. Rezaeian, Experimental investigation of panel zone in rigid beam to box column connection, J. Constr. Steel Res. 137 (2017 Oct 1) 180–191. [37] O. Rezaifar, M. Nazari, M. Gholhaki, Experimental study of rigid beam-to-box column connections with types of internal/external stiffeners. Steel and, Compos. Struct. 25 (5) (2017 Dec 10) 535–544. [38] H. Saffari, A.A. Hedayat, G.N. Soltani, New alternatives for continuity plate in Ibeam to box columns connections, Asian J. Civil Eng. (BHRC) 16 (2) (2015) 219–233. [39] AISC, Seismic Provisions for Structural Steel Buildings, AISC/ANSI 341–10, American Institute of Steel Construction (AISC), Chicago, IL, 2010. [40] ABAQUS, Theory Manual, Version 6.14, 2014. [41] ASTM, A370–18, Standard Test Methods and Definitions for Mechanical Testing of Steel Products, ASTM International, West Conshohocken, PA, 2018.