Comparison of Component Method with Experimental

www.ijoss.org

Steel Structures 9 (2009) 00-00

Comparison of Component Method with Experimental Tests for Flush End-Plate Connections using Hot-Rolled Perwaja Steel Sections Mahmood Md. Tahir *, Mohd Azman Hussein , Arizu Sulaiman , and Shahrin Mohamed 1,

2

3

1

1

Professor, Steel Technology Centre, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 UTM, Skudai, Johor, Malaysia

2

Master Research Student, Steel Technology Centre, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 UTM, Skudai, Johor, Malaysia

3

Senior Lecturer, Steel Technology Centre, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 UTM, Skudai, Johor, Malaysia

Abstract A component method has been introduced by Steel Construction Institute to predict the moment resistance of partial strength connection. The design philosophy is taken directly from Eurocode 3 with strength checks on bolts, welds, and steel which have been modified to comply with BS 5950-1:2000. The accuracy of the method however needs to be validated with the experimental tests especially for hot-rolled sections other than typical British Section (BS). Six experimental tests on beamto-column connections have been carried out for Flush End-Plate (FEP) connections consisting of variable parameters such as thickness of end-plate, size and number of bolts, size of columns, and beams. The tests were set-up using local hot-rolled steel sections known as Perwaja Section (PS) for beams and columns instead of typical British Section (BS). The strength of materials for end-plate, column and beam sections were tested for tensile strength and used in predicting the moment resistance for component method. The moment versus rotation of the test results were plotted and compared with the moment resistance derived from component method. The study concluded that the moment resistance of the tested flush end-plate connections was higher than the predicted moment resistance from component method which showed good agreement between the two moments. The study also concluded that the tested FEP connections met the requirements and criteria of partial strength connections.

Key words: 1. Introduction

Connections in steel frames are usually designed either as pinned joint or rigidly joint. The rigid connection and pinned connection are the idealized assumptions that indicate full moment transfer and zero moment transfer; the semi-rigid connection in actual condition stands in between these two limits. The beams are assumed as simple supported with pin jointed connections and the columns are assumed to sustain axial and nominal moment (moment from the eccentricities of beam’s end reactions) only. The connection is simple but the sizes of beams obtained from this approach have resulted in heavy and deep beam. This type of connection is usually Note.-Discussion open until October 1, 2009. This manuscript for this paper was submitted for review and possible publication on 0000 00, 0000; approved on 0000 00, 0000 *Corresponding author Tel: ; Fax: E-mail: [email protected]

associated with simple construction. On the other extreme, rigidly jointed frame results in heavy columns due to the end moments transmitted through the connection. Hence, a more complicated and more expensive fabrication of the connection could not be avoided. This type of connection is usually associated with continuous construction. One approach, which creates a balance between the two extreme approaches mentioned above, has been introduced. This approach, termed as partial strength connection is usually associated with a connection having a moment capacity less than the moment resistance of the connected beam, Allen . (1995). Two typical types of connections usually associated with partial strength connections are Flush End-Plate (FEP) and Extended End-Plate (EEP) connections as shown in Fig. 1(a) and (b) respectively. They are commonly used in the design of both braced and un-braced frames due to their higher moment resistance. These types of connections are related to the design of steel frame as semi-continuous construction. However, in this paper only flush end-plate connections will be presented as extended end-plate connections are discussed elsewhere Hussein (2001). et al

2

Mahmood Md. Tahir

et al.

Figure 1. Flush and Extended End-Plate connections as partial strength connections.

Eurocode 3 (2005) allows building frames to be designed using semi-rigid or partial strength connection, a type of design that utilized a condition between the simple and rigid design; provided that the moment resistance of the connection can be quantified. However, to quantify the moment resistance of the connection for partial strength connections, experimental tests should be carried out so as to understand explicitly the behavior of the connection. This is very costly and time consuming. Steel Construction Institute Allen . (1995) has established a component method to predict the moment resistance of the connections by adopting EC 3 (EC 3, 2005) and BS 5950-1:2000 (BSI, 2000). The method however needs to be validated by experimental tests. A series of test have been carried out by Bose (1993) at Dundee Institute of Technology for typical hot-rolled British Sections (BS) to validate the prediction of moment resistance of the connections by the component method. The thickness of the end-plate used was up to 60% of the diameter of the bolt. The ductility of the connections was performed up to 686 × 254 universal beams section. Details of the tests results submitted to Steel Construction Institute however were classified as confidential. Nevertheless, in this paper, a series of tests on flush end-plate connections with Perwaja steel sections will be presented and compared with the predicted moment resistance from the component method. The objective of this paper is to compare the moment resistance of the connection predicted by component method with the experimental tests and to show that the component method is applicable to other type of sections, not just for BS. Some of the significant characteristics of FEP connections such as the moment resistance, the rotational stiffness (rigidity), and the ductility (rotation capacity) will be discussed in this paper. The proposed isolated tests were set-up as beam-to-column connections comprised of six specimens of FEP connection where the columns and the beams were of hot-rolled Perwaja steel sections (local sections produced in Malaysia). Studies on partial strength or semi-rigid connections et

al

have been carried out by many researchers. Among them were Chen & Kishi (1989), who developed a computerised databank system at Purdue University Computer Centre. The database was aimed at providing moment-rotation characteristics and corresponding parameters of semirigid steel beam-to-column connections used frequently in steel construction. Abdalla & Chen (1995) then expanded the database by adding another 46 experimental tests data. Recent tests on extended end-plate connections were conducted by Abdalla (2007) to investigate the force distribution in high-strength bolt. Six full scale beam-to-column joints with extended end-plate connections were tested with eight high strength bolt of M20 grade 8.8 were used. From the study, it was found that the tension forces in the upper bolts above the beam flange could be reasonably determined using an equal distribution method. However, the use of equal distribution method was not suitable for upper bolts below the beam flange. Coelho . (2004) carried out experimental tests on eight statically loaded end-plate moment connections. The specimens were designed to cause failure on the endplate or bolts. The study concluded that an increase in end-plate thickness had resulted in an increase in the connection’s flexural strength and stiffness. Mahmood (2008) had carried a series of experimental test on extended end-plate connections with variable parameters. The sections used for beams and columns in the tests were from the Perwaja Sections produced locally in Malaysia. The tests concluded that the theoretical values calculated by component method showed good agreement with the experimental values in most cases. However, these tests were carried out for hot-rolled steel section and the connection was a non-composite connection. Shi (2007) tested composite joints with flush end plate connection as partial strength connection under cyclic loading. The composite joints with flush end plate connection showed large strength resistance and good ductility, and the slippage between the concrete slab and steel beam was very small, which showed that between the concrete slab and steel beam, full interaction could be et al

et

at.

et al.

running title

Figure 2.

3

Critical zones to be checked for failure.

obtained by proper design of shear connectors. Studies on the partial strength connection were also carried out by Tahir (2008) & Sulaiman (2007). A series of non-composite connections comprising of flush and extended end-plate connections was tested and compared with composite connection. The beam sections were a built-up section known as trapezoid web profiled steel sections whereas British Sections were used as columns. It was concluded that the moment resistance and the initial stiffness for composite connection were higher than the non-composite connection. Further study on the partial strength connection with TWP steel section was carried out by (Tahir, Sulaiman and Saggaff, 2008). Two full scales testing with beam set-up as sub-assemblage and beam-to-column connections set-up as flush and extended end-plate connections were carried out. It was concluded that the use of extended end-plate connection had contributed to significant reduction in the deflection and significant increase in the moment resistance of the beam than flush end-plate connection.

paper only S275 steel sections were used for the experimental tests. The PSS sections are produced locally in Malaysia.

2. Perwaja Steel Section

3.1. Distribution of bolt forces

Perwaja Steel Section (PSS) is a structural hot-rolled Ishaped steel section referred in this paper as a section produced in accordance with the specifications of JIS G 3192:1994 (1994). These sections are used for structural steel that can be welded based on BS EN 10113 (1993). The section designation used for PSS is not the same as the BS; the dimension of the section is round up to a more fixed number which is more user-friendly. For example, in BS, the column is designated as UC 203 × 203 × 46, however for PSS the section is designated as HB 200 × 200 × 49.9. The designation of HB is used to represent both the beams and columns sections. The thickness of the PSS section for HB200 × 200 × 49.9 is 8.0mm thick whereas the thickness of the BS section for HB 203 × 203 × 46 is 7.3 mm thick, lesser than the PSS. This also applies to all other sections of PSS where the thickness of the web is thicker than the BS. The width of PSS (200 mm) is smaller than the width of BS (203 mm) steel. The steel design strengths for PSS however are the same as the typical BS, namely S275 and S355. In this

3. Component Method The design model of component method presented in this study is adopted from Steel Construction Institute (SCI) and Eurocode 3 (2005). For checking on the details of strength of bolts, welds, and steel section, modification was made to suit BS 5950-1:2000 (BSI, 2000). Checking on the capacity of the connections is classified into four zones namely tension zone, horizontal shear zone, compression zone, and vertical shear zone as shown in Figure 2 Allen . (1995). Details of the checking on each of the zone are listed in Table 1. The basic principles on the distribution of bolt forces need to be addressed first before details of the checking on all possible modes of failures can be discussed. et al

The moment resistance of a connection transmitted by an end-plate connection is through the coupling action between the tension forces in bolts and compression force Component design checks Zone Reference Component to be checked 1 Bolt tension 2 End-plate bending 3 Column flange bending Tension 4 Beam web tension 5 Column web tension 6 Flange to end-plate weld 7 Web to end-plate weld Horizontal Shear 8 Column web panel shear 9 Beam flange compression 10 Beam flange weld Compression 11 Column web crushing 12 Column web buckling 13 Web to end-plate weld Vertical Shear 14 Bolt shear 15 Bolt bearing (plate or flange) Table 1.

4

Mahmood Md. Tahir

et al.

Figure 3. Elastic and plastic analysis of bolt forces distribution in FEP connection.

Figure 4. Three typical types of failures in tension zone

at the centre of the bottom flange. Each bolt above the neutral axis of the beam produced tension force whereas the bolts below the neutral axis are dedicated to shear resistance only. Eurocode 3 (2005) suggests that the bolt forces distribution should be based on the plastic distribution instead of the traditional triangular distribution. Figure 3 shows the forces in the connection and the corresponding distributions. The forces of the bolts are based on the plastic distribution which represents the actual value calculated from the critical zones in Fig. 2. The force from the top bolt row transmits to the end-plate connection as tension force which is balanced up by the compression force at the bottom flange of the beam to the column. The end-plate is welded to the web and both flanges of the beam. The formation of tension at the top and compression at the bottom contribute to the development of moment resistance of the connection. Tests on the connections have shown that the centre of compression flange which bears against the column was found to be the centre of rotation of the connection (Bose, 1993). The force permitted in any bolt row is based on its potential resistance and not just on the length of the lever arm. 3.1.1. Tension zone

The resistance at each bolt row in the tension zone may be limited due to bending of column flange, end-plate,

column web, beam web, and bolt strength. Column flange or end-plate bending is checked using Eurocode 3 (2005) which converts the complex pattern of yield lines around the bolts into a simple ‘equivalent tee-stub’ as shown in Fig. 4. Details of the procedures are illustrated in SCI publication (SCI & BCSA, 1995). This approach has resulted in the classification of three types of mode of failures as follows: Type 1: complete flange yielding In this failure mode, the strength of the column flange or end-plate is weaker than the strength of the bolts. Upon failure, the flange or end-plate will yield but the bolts are still intact as shown in Fig. 4(a). As a result, a ductile failure can be achieved. This type of failure is the most preferred failure mode in semi-continuous construction as suggested by SCI (SCI & BCSA, 1995) as abrupt failure of the connection can be avoided. To achieve this type of failure, SCI has suggested that the size of 12 mm thick end-plate is used together with M20 bolt and the size of 15mm thick end-plate is used together with M24 bolt. Type 2: bolt failure with flange yielding In this failure mode, the strength of the column flange or the end-plate and the bolts are about the same. As a result, the column flange or the end-plate and the bolts yield together upon failure. This mode of failure is shown in Fig. 4(b). This type of failure can be used in the design

running title

of semi-continuous construction provided that the moment resistance of the connection can be quantified and the connection can still be classified as ductile connection. Type 3: bolt failure In this failure mode, the strength of the bolts is weaker than the strength of the flange. Upon failure, the bolts will yield (or even break) but the flange or the end-plate is still intact as shown in Fig. 4(c). This type of failure is not suitable for semi-continuous connection and should be avoided as the connections possess an abrupt type of failure which is not allowed by BS 5950:2000 and EC 3 code of practice.

3.1.2. Compression zone

The checking on the compression zone follows the same procedures as described in BS 5950-1:2000 (2000) which requires checks on web bearing and web buckling. The compression failure modes can be on the column side or on the beam side. The column side should be checked for web buckling and web bearing due to the compression force applied to the column. The use of stiffener or the effect of having other beam connected to the web of the column is not included so as to reduce the cost of fabrication and simplified the calculation. The compression on the beam side can usually be regarded as being carried entirely by the beam flange, however when large moments combined with axial load, the compression zone will spread to the web of the beam which will affect the centre of compression Allen . (1995). Therefore, the stiffening of the web of the beam needs to be done. However, in this study the stiffening of the web was not considered so as to reduce the cost of fabrication and in line with the geometrical configuration of the connections suggested by SCI. Allen . (1995). et al

et al

3.1.3. Shear zone

Column web can fail due to shearing effect as a result of tension and compression force applied to the web of the column. The failure to the shearing of the web is most likely to happen before it fails due to bearing or buckling. This is possible because the thickness of the flange is more than the thickness of the web. Again in this shear zone, stiffener is not provided so as to reduce the cost of fabrication. For one-sided connection with no axial force, the shear in the column web can be taken as the compressive force. However, for two-sided connection with balanced moments, the shear is taken as zero and for two-sided connection with unbalanced moments; the shear is taken as the addition of the compression force and the tension force.

3.1.4. Welding

Fillet weld is preferred to be used than the butt welds as the welding of beam to the end-plate is positioned at 90 degrees. The end-plate is connected to the web of the beam by an 8 mm fillet weld, whereas a 10 mm fillet

5

weld is suggested for connecting the end-plate to the flange. The weld is designed in such a way that the failure mode of the connection is not on the welding. This is to ensure the ductility of the connection which is necessary for partial strength connection.

4. Test Arrangement and Procedures For a full-scale testing, a test rig was designed and erected to accommodate a column height of 3 m and a cantilever beam span of 1.3 m. The rig consists of channel sections pre-drilled with 22 mm holes for bolting purposes. The sections were fastened and bolted to form loading frames, which were subsequently anchored to the laboratory strong floor as shown in Fig. 5. The height of the column was kept at 3 m to represent the height of a sub-frame column of multi-storey steel frame. The column was restrained from rotation at both ends. The beam was also restrained from lateral movement. The load was applied at a distance of 1.3m from the face of the column using a hydraulic jack. This distance was taken to represent the approximate length of hogging moment occurred on the sub-assemblage steel frame. After the instrumentation system had been set-up and the specimen had been securely located in the rig, data collection software in the computer was used to check the reading of all connected channels to the instruments on the specimen. Correction factors from calibration and gauge factors from manufacturer were set into the software prior to each test. The specimen was then loaded about one-third of the predicted value. The reading of load was taken as point load applied for the ease of monitoring. After reaching one-third of the predicted load capacity, the specimen then unloaded back and reinitialized. This procedure was carried out so as to enable the specimen to be in the state of equilibrium prior to the actual test. The specimen was loaded again after re-initializing the instrumentation system; however, the applied load was not restricted to one-third capacity. An increment of about 5kN was adopted so that a uniform data and gradual failure of the specimen can be monitored. The specimen was further loaded until substantial deflection of the beam can be observed. At this point, the loading sequence was controlled by the increment of the deflection as a small increment of load has resulted to substantial increase in the deflection. Therefore, the load was continuously applied but each increment of the load was limited to the deflection of 2 mm of the beam instead. This procedure was continued until the specimen had reached its failure condition. The failure condition was considered to have reached when an abrupt or significantly large reduction in the applied load or when a large rotation of the connection due to deformation of the tested specimen. For each loading, a set of reading was taken for deflections, rotations, and applied load.

6

Mahmood Md. Tahir

Figure 5.

et al.

Test rig for full scale testing

4.1. Description of specimens

Six specimens were arranged for the testing of flush end-plate (FEP) connections. The typical geometrical configuration of the connection was shown in Fig. 1(a). The size of beam, column and end plate, and the diameter of bolts were shown in Table 2. The specimens are designated as FEP 1-FEP 7. Specimen FEP 5 is not given in Table 2 as the specimen has the same parameter as FEP 6. The size of column used was HB300 × 300 × 83.5 designated as heavy column and HB200 × 200 × 56.2 designated as light column. This is purposely done so as to understand further the effect of changing the size of the column to the failure mode of the connection. The size of beam varied from 250 mm to500 mm deep. This increment is necessary as the moment resistance of the connection varies significantly with the increment of the lever arm of the connection which is related to the depth of the beam.

The moment resistance of the connection is also related to the size of bolts and end-plate. To optimize the strength of the bolts and the thickness of the end-plate, bolts of size M20 of grade 8.8 were used together with 12 mm thick end-plate and bolts of size M24 of grade 8.8 were used with 15mm thick end-plate. The width of the endplate varied from 200 mm to 250 mm so that the effect of changing the width of the end-plate can be studied. The number of tension bolt row was also varies from one to two bolts rows. The vertical distance and the gauge between bolts remained the same for all specimens. Both of the beam flanges were welded by a 10 mm fillet weld to the end-plate whereas the web was welded to the endplate by a 8mm fillet weld for all specimens. The weld was expected not to fail so that other variable parameters of the connection could be studied without any interference of weld failure.

Dimensions of the thickness of flange and web of (PSS) and the FEP connections Size of end-plate Size of bolt No. of bolt Column sections Beam sections Width Thickness Depth (in mm) row in tension (in mm)

Table 2.

Test No. FEP 1 FEP 2 FEP 3 FEP 4 FEP 6 FEP 7

HB 200 × 200 × 56.2 Flange=12 mm thick Web=12 mm thick HB 200 × 200 × 56.2 Flange=12 mm thick Web=12 mm thick HB 250 × 250 × 63.8 Flange=11 mm thick Web=11 mm thick HB 250 × 250 × 63.8 Flange=11 mm thick Web=11 mm thick HB 300 × 300 × 83.5 Flange=12 mm thick Web=12 mm thick HB 300 × 300 × 83.5 Flange=12 mm thick Web=12 mm thick

HB 250 × 125 × 25.1 Flange=8 mm thick Web=5 mm thick HB 250 × 125 × 25.1 Flange=8 mm thick Web=5 mm thick HB 400 × 200 × 65.4 Flange=13 mm thick Web=8 mm thick HB 400 × 200 × 65.4 Flange=13 mm thick Web=8 mm thick HB 500 × 200 × 102 Flange=19 mm thick Web=11 mm thick HB 500 × 200 × 102 Flange=19 mm thick Web=11 mm thick

M20

1

200 12 300

M24

1

200 15 300

M20

2

200 12 500

M20

2

250 12 500

M24

2

200 15 600

M24

2

250 15 600

running title

Table 3.

No.

Beams and columns.

1

200 × 200 × 56.2 (flange) 200 × 200 × 56.2 (web)

2

250 × 250 × 63.8 (flange) 250 × 250 × 63.8 (web)

3

300 × 300 × 83.5 (flange) 300 × 300 × 83.5 (web) 250 × 125 × 25.1 (flange) 250 × 125 × 25.1 (web) 400 × 200 × 65.4 (flange) 400 × 200 × 65.4 (web) 500 × 200 × 102 (flange) 500 × 200 × 102 (web) End-plate (12 mm) P1 P2 P3

4 5 6 7

Material properties of beams, columns, and end-plates Ultimate Strength, fu Yield Strength, fy (N/mm2) (N/mm2) 367 528 385 547 (avg. 376) (avg. 538) 351 510 351 540 (avg. 351) (avg. 525) 370 516 359 510 (avg. 364.5) (avg. 513) 388 521 356 506 335 405 312 509 299 471 357 499

End-plate (15 mm) P4 P5 P6

8

467 491 470 (avg. 476)

203 205 204 (avg. 204)

310 311 308 (avg. 309.7)

515 524 507 (avg. 515.3)

204 205 203 (avg. 204)

f

E

f

f

f

Modulus of Elasticity, E (kN/mm2) 194 198 (avg. 196.0) 193 192 (avg. 192.5) 202 194 (avg. 198) 208 193 201 198 193 195

305 308 309 (avg. 307.3)

Coupon tests were carried out on the flange and the web of the tested beams, columns, and end-plates of the specimens to check the material properties. The mean values of yield strength y, ultimate strength u and modulus of elasticity were recorded as shown in Table 3. The expected value for yield strength y is 275 N/mm2. However, the results showed that the experimental values were higher than the specified values for both y and u. The tested values of y were used in the component method to predict the moment resistance of the connections. This predicted moment resistance was then compared with the maximum moment resistance derived from the experimental tests. f

7

f

5. Test Results

Full-scale tests were conducted by setting up a 1.3 m length beam connected to a column. The arrangement of the cantilever beam was designed to study the interaction or the effect of using partial strength connection to the moment resistance of connection on the actual beam. The test results were presented based on the moment resistance, rotation stiffness, and ductility of the connection which were derived from graph plotted for moment versus rotation curve. The expected modes of failure were predicted from the calculated moment resistance value proposed by component method and compared with the experimental results.

5.1. Modes of failure

There were definitely no apparent visual deformations which can be observed in all of the tests during initial stage of loading. This was expected since the application of loads was intended for all components of the joint to be stabilized or to be in equilibrium. At this stage all instruments used to collect the data were checked to be in good working order prior to the actual commencement of the tests. After re-initializing, each specimen was then loaded until there was an indication that ‘failure’ had been attained. This indication was observed from the declination of the load after reaching a period of constant ultimate load. The test was then brought to a stop as any increment of load would result in to further deformation of the specimen. During the tests, there was no occurrence of any vertical slip at the interface between the end-plate and the column. This was mainly due to the adequate tightness of the bolts carried out during the installation. The first visible deformation was observed around the vicinity of the connection; and this deformation was localized to the tension region (the critical zone) of the joint due to the tension forces exerted through the top bolt rows and the compression zone where the bottom flange of the beam exerted a compression force to the column flange. The mode of failures of the connection however was dependent on the geometrical configuration of the connection. The typical types of failure can be divided

8

Mahmood Md. Tahir

et al.

Failure mode predicted from component method and compared with experimental tests Specimens Component method Experimental tests FEP 1 Deformation of end-plate in tension zone End plate deformed in tension zone. flange deformed first followed by the deformation of FEP 2 Deformation of column flange in tension zone Column end-plate in tension zone. of column flange in tension zone. FEP 3 Deformation of column flange in tension zone Deformation No slippage of bolt. of column flange in tension zone. FEP 4 Deformation of column flange in tension zone Deformation No slippage of bolt. of column flange in tension and compression zone. FEP 6 Deformation of column flange in tension zone Deformation No deformation of end-plate. of column flange in tension and compression zone. FEP 7 Deformation of column flange in tension zone Deformation No deformation of end-plate Table 4.

Deformation of end-plate in tension zone for FEP 1 specimen. Figure

6.

into two areas, namely the beam and the column areas. For the beam area, the most likely modes of failure are the deformation of the end-plate, the deformation of the beam flange, the crushing of the beam web, and the slippage of the bolts’ tread. For the column area, the most likely modes of failure are the deformation of the column flange, the crushing of the column web and the shearing of the column web. These modes of failure however can be predicted using a component method proposed by Steel Construction Institute, Allen . (1995). The predicted mode of failure was then compared with the mode of failure from the experimental tests. For the experimental tests, the increment of applied load was stopped until the mode of failure of the connection was clearly recognized. Some typical modes of failure from the experimental tests are shown in Fig. 6 to Fig. 11. The comparison between mode of failure predicted from component method and mode of failure from the experimental tests is shown in Table 4. The results in Table 4 show that the modes of failure predicted from the component method are in consistent with the experimental tests. Modes of failure of the connections were mostly in the tension zone. Only specimen FEP 1 showed the deformation et

Deformation of end-plate and column flange in tension zone for FEP 2 specimen.

Figure 7.

al

Deformation of the column flange in tension zone for FEP 3 specimen. Figure

8.

of the end-plate which was also considered as Type 1 failure. This expected type of failure was due to the same thickness (12 mm) for both the end-plate and the column flange of HB200 × 200 × 56.2. Moreover, the size of bolt used for FEP 1 was one bolt row of M20 which has not developed enough tension force to deform the column flange. As the thickness of the end-plate increases from

running title

Deformation of column flange in tension for FEP 4 specimen.

Figure

9.

mode of failure from component method as shown in Figures 7-11 (FEP 2-FEP 7). The end-plate however did not show any significant deformation for most of the specimens. This is because the thickness of end-plate used was thicker than the thickness of column flange. For FEP 2, FEP 6 and FEP 7, the thickness of end-plate (15 mm) was thicker than the thickness of column flange (12 mm) of HB 300 × 300 × 83.5 and HB 200 × 200 × 56.2. The failure mode where the column flange yield first was more explicit when 15mm thick end-plate was used with two number of M24 bolt rows. All failure modes were in consistent with the failure modes predicted from the component method as shown in Table 4. In compression zone where the bottom flange of the beam exert a compression force to the column flange, only two columns namely, FEP 6 and FEP 7 experienced a crushing type of failure to the flange of the column. However, this mode of failure only occurred after a large deformation of column flange in tension. There was no shear deformation on the columns web throughout the experimental programme of the tests as the thickness of the column web, 12 mm for H300 × 300 × 83.5 and 11 mm for H250 × 250 × 63.8 was thick enough to resist the shear failure.

5.2. Moment-rotation curves

Deformation of column flange in tension and compression zone for FEP 6 specimen.

Figure 10.

Deformation of column flange in tension and compression zone for FEP 7 specimen.

Figure 11.

12mm to 15mm and the size of bolts increases from M20 to M24, the mode of failure of the connection started to shift from the deformation of the end-plate in the first stage to the deformation of the column flange. No slippage of tread occurred to any bolt for all tested specimens. This deformation corresponded to the mode of failure for specimens other than FEP 1. This type of failure mode known as type 1 was in consistent with the predicted

9

The prediction of moment resistance and the stiffness of the connection are related to the size of the connected members, types of joints, and orientation of the column axis, Tahir, (1997). Beam-to-column connections generally developed linear and non-linear moment-rotation curves. Initially, the connections developed a stiff initial response which was then followed by a second phase of much reduced stiffness. This second phase was due to an inelastic deformation of the connections’ components or those of members of the frame in the immediate vicinity of the joint. These deformations need to be accounted for because they contribute substantially to the frame displacements and may affect significantly the internal force distribution. The structural analysis needs to consider this non-linearity of joint response to predict accurately both stiffness and resistance for a semicontinuous frame in case the joint behavior exhibits a form of material non-linearity. The curves of the experimental results for the M-Φ curve are shown in Figs 12-17 for FEP specimens. The maximum moment resistance (Mmax) listed in Table 5 was determined from the M-Φ curves plotted in Figure 12 to 17. The graphs of the M-Φ curve show that the connections behaved linearly in the first stage followed by non-linear behavior and gradually losing the stiffness with the increase in rotation. From this plot, the behavioral characteristics of a particular joint can be determined based on the three significant parameters; the moment resistance (strength), the rotational stiffness (rigidity) and the rotational capacity (ductility). Table 4 also presents the theoretical values of the moment resistance calculated from the

10

Mahmood Md. Tahir

Figure 12.

Moment vs rotation for FEP 1

Figure 13.


Figure 14.

et al.

Figure 15.


Figure 16.


Figure 17.



component method proposed by SCI as mentioned earlier. The theoretical values for component method were calculated using the average value of the actual design strength of steel (fy) from the coupon tests results presented in Table 2. The average values in Table 2 were calculated for sections with the same web and flange thickness and for the end-plate with the same thickness.

The theoretical moment resistance is designated as Mcm (Table 5). Details of the component method are presented in SCI publication (SCI & BCSA, 1995). The overall results show that the experimental values of maximum moment resistance were greater than the theoretical

running title

Comparison between the experimental and theoretical values of maximum moment resistance Theoretical Ratio (component Specimens Experimental Mmax/Mcm Test, Mmax method) M Table 5.

FEP 1 FEP 2 FEP 3 FEP 4 FEP 6 FEP 7

67.1 80.3 121.8 168.5 292.2 312.3

39.0 54.0 116.0 118.0 225.0 226.0

cm

1.72 1.49 1.05 1.43 1.30 1.38

values with the ratio in the range of 1.05 to 1.72 as shown in Table 5. The results indicate that the theoretical values calculated by component method showed good agreement with the experimental values for most of the tested specimens. 5.3. Rotation stiffness

The rotation stiffness of the connection depends on the geometrical configuration of the connection. Factors such as number of bolts, thickness of end-plate, and depth of the beam play an important role to determine the stiffness of the connection. Therefore, it is best to present the rotation stiffness of the connection by comparing the moment resistance and relate to other connections’ parameters. The results of the moment resistance, initial

11

stiffness, and maximum rotation at maximum load are tabulated in Table 6. The results generally show that the higher the moment resistance, the higher the initial stiffness of the connection. These results also show that the deeper the beam, the higher the initial stiffness of the connection. This was because the use of deep beam had resulted in longer distance between tension zone and compression zone which reduced the rotation capacity of the connection. The thickness of the end-plate and the size of bolts also affects the stiffness of the connection. For 12 mm thick end-plate in conjunction with M20 bolts, the stiffness of the connection was lesser than the 15 mm thick end-plate in conjunction with M24 bolt. The increase in initial stiffness from FEP 1 (4.17 kNm/mrad) to FEP 2 (8.47 kN/mrad) showed that the stiffness of the connection increased significantly. This was because the 15mm thick end-plate together with one bolt row of M24 has contributed significantly the tension capacity of the connection. The ductility of the connection is measured by the capacity of the connection to rotate to act as a plastic hinge Allen , (1995). SCI has suggested that for the connection to be classified as partial strength connection, the rotation achieved of the connection should be at least 20mrad and the moment resistance of the connection should be at least 25% of the moment resistance of the connected beam. In these tests the maximum rotation of the connections was in the range of 39.8 to 104.9 mrad as et al

Test results based on the moment versus rotation curves Stiffness, Max. rotation at and no. of End-plate thick- Moment Rotation, Φ Initial Size of beam Sizebolt S row =M ness max. load., Φ Resistance, M R j,ini r/Φ (in mRad) (kNm/mRad) HB (kNm) (in mm) (mm) (in mRad) 20 12 250 × 125 × 25.1 (1 bolt row) 47.1 11.3 4.17 104.9 (W=200) 24 15 250 × 125 × 25.1 (1 bolt 70.3 8.3 8.47 96.5 row) (W=200) 12 400 × 200 × 65.4 (2 bolt20rows) (W=200) 103.7 3.2 32.41 39.8 12 400 × 200 × 65.4 (2 bolt20rows) (W=250) 105.8 2.6 40.69 45.4 15 500 × 200 × 102 (2 bolt24rows) (W=200) 214.6 5.9 36.37 79.2 15 500 × 200 × 102 (2 bolt24rows) (W=250) 204.0 4.5 45.33 42.90 Table 6.

Specimen FEP 1 FEP 2 FEP 3 FEP 4 FEP 6 FEP 7

Comparison of moment resistance of connection versus moment resistance of connected beam Max. moment resistance from Moment capacity of the beam. Ratio of moment Size of beam Mcx (kNm) Mmax/Mcx experimental tests, Mmax (kNm) HB 250 × 125 × 25.1 67.1 86.0 0.78 HB 250 × 125 × 25.1 80.3 86.0 0.93 HB 400 × 200 × 65.4 121.8 361.0 0.34 HB 400 × 200 × 65.4 168.5 361.0 0.47 HB 500 × 200 × 102 292.2 661.0 0.44 HB 500 × 200 × 102 312.3 661.0 0.47

Table 7.

Specimen FEP 1 FEP 2 FEP 3 FEP 4 FEP 6 FEP 7

12

Mahmood Md. Tahir

shown in Table 6 and the moment resistance of the connection was in the range of 34% to 93% of the moment resistance of the connected beam (Mcx) as shown in Table 7. The Mcx values shown in Table 7 were calculated based on BS 5950-1:2000 with the design strength of steel fy, taken from the actual test results in Table 3. The test results show that the moment resistance of the connection Mr was more than 25% of the moment resistance of the connection beam as suggested by SCI. As a result, all of the tested connections could be classified as partial strength and possessed a ductile connection behavior with the ability to form a plastic hinge. 6. Discussion of Results

The interaction of the component elements that formed the connection represents the behavior of the connections. However, significant effect to the behavior of the connections can only be achieved by considering the combination of proper connections’ parameters. For example the endplate thickness of 12 mm should be used together with M20 bolts and 15 mm thick end-plate should be used together with M24 bolts. This is to avoid premature deformation of end-plate or the bolts if the combined connections parameters are not properly selected. The size of the beam, the number, size and distance of the bolt Table 8.

Specimen FEP 1 (1 row M20 bolts) vs FEP 2 (2 rows M24 bolts) FEP 3 (2 rows M20 bolts) vs FEP 4 (2 rows M20 bolts) FEP 6 (2 rows M24 bolts) vs FEP 7 (2 rows M24 bolts) FEP 1 (1 row M20 bolts) vs FEP 3 (2 rows M20 bolts) FEP 2 (1 row M24 bolts) vs FEP 6 (2 rows M24 bolts)

et al.

and the thickness of the end-plate may significantly affect the moment resistance and the rotation stiffness of the connection. To understand further the effects of these geometrical configurations the moment resistance, rotational stiffness, and the ductility of the connection should be compared and discussed by referring to the M-Φ curves. These effects can be well understood by comparing the behavior of the tested specimens based on the moment resistance and the initial stiffness of the connections.

6.1. Effect on varying the geometrical parameters of the connections

The use of 12 mm thick end-plate with M20 bolts of Grade 8.8 and 15 mm thick end-plate with M24 bolts of Grade 8.8 thick was suggested by SCI. This practice was suggested so as to ensure the ductility of the connection and the deformation capacity of the end-plate and the tension capacity of the bolt can be balanced up or in equilibrium. The effect of increasing the size of bolts from M20 in conjunction with 12mm thick end-plate to M24 in conjunction with 15mm thick end-plate could be seen by comparing specimen FEP 1 with specimen FEP 2. The results presented in Table 8 show that the moment resistance increased by 19.7% and the initial stiffness increased by 103.1%. The increase in moment resistance was not that significant as the thickness of the end-plate and the thickness of the column flange was equal to

Effect of increasing the size of bolts and the thickness of the end-plate Max. moment Connection’s parameters Percentage Initial Stiffness, Percentage resistance from End-plate difference Sj,ini =MR/Φ experimental tests, difference (kNm/mRad) Thickness Width % % Mmax (kNm) 12 200 67.1 4.17 HB250 × 125 × 25.1 19.7% 103.1% HB250 × 125 × 25.1 80.3 8.47 15 200 12 200 121.8 32.41 HB400 × 200 × 65.4 5.6% 25.5% HB300 × 300 × 65.4 168.5 40.69 12 250 15 200 292.2 36.37 HB500 × 200 × 102 6.9% 24.6% HB500 × 200 × 102 312.3 45.33 15 250 12 200 67.1 4.17 HB250 × 125 × 25.1 81.5% 677.2% HB400 × 200 × 65.4 121.8 32.41 12 200 15 200 80.3 8.47 HB250 × 125 × 25.1 289.0% 329.4% HB500 × 200 × 102 312.3 36.37 15 200

running title

12 mm. An increment of the end-plate to 15 mm thick did not contribute significantly to the moment resistance of the connection as the 12 mm thick column flange started to deform. However, as the failure mode shifted from deformation of end-plate to the deformation of the column flange due to this increment, the increase in the initial stiffness was significant as the percentage difference showed a 103.1% increase. The effect of increasing the width of the end-plate from 200 mm to 250 mm could be seen by comparing specimen FEP 3 with FEP 4, and FEP 6 with FEP 7. The results of the comparison between these specimens are tabulated in Table 8. The results show that the percentage of increment in moment resistance was in the range of 5.6% to 6.9% and percentage of increment in initial stiffness was in the range of 24.6% to 25.5%. The results show that the increase in the width of the end-plate with the same thickness did not result to a significant increase in both the moment resistance and the initial stiffness of the connection. The effect of increasing the depth of the beam from 250 mm with one bolt row of M20 to 400 mm with two bolt rows of M20 could be seen by comparing specimen FEP 1 with specimen FEP 3. The increase in the depth of the beam could also be studied by looking at the effect of increasing the depth of the beam from 250 mm with one bolt row of M24 bolts to 500 mm with two bolt rows of M24 bolts. This could be seen by comparing specimen FEP 2 with specimen FEP 6. The percentage of increment in moment resistance was in the range of 81.5% to 289.0% and initial stiffness was in the range of 329.4% to 677.2% started to improve significantly as the depth of the beam changed from 250 mm to 500 mm together with the change in the number of bolt from one bolt row to two bolt rows. This increment was due to the combination of both the increment of the size of beam that contributed to the increase in the lever arm of moment resistance and also the contribution to the number and size of bolt that contributed to higher tensile force which was used to determine the moment resistance of the connection. However, as the number of bolt increased from one bolt row to two bolt rows and the size of bolt increased from M20 to M24, the stiffness of the connection started to reduce from 677.2% to 329.4% as higher tension force from M24 has resulted to the deformation of the 12mm thick column flange. This showed that the increased in the end-plate thickness from 12 mm to 15 mm did not contribute to the stiffening of the connection. However, the moment resistance has improved from 81.5% to 289.0%. The overall results showed that the increase in the thickness of the end-plate from 12 mm to 15 mm has not contributed significantly to the increase in moment resistance and stiffness of the connection. The overall results also show that the moment resistance and the stiffness of the connection increased significantly as the size and number of bolt and the depth of the beam

13

increased. Again the thickness of the column flange which was 12mm thick had limited the increment as the column flange deformed first before other connection’s components started to develop any deformation.

7. Conclusions From the test results, the following conclusions can be drawn: 1. Tests on the flush end-plate connections revealed that the behavior of the connections satisfied the requirements or criteria set by SCI for partial strength connections with moment resistance of the connections recorded more than 25% of the moment resistance of the connected beam and the rotation capacity of the connections recorded more than 20 mrad. 2. The increase in the moment resistance and initial stiffness of the FEP connections was significantly improved as the size, the number of bolt and the depth of the beam increased. However, the increase in the thickness of the end-plate should also take into consideration the thickness of the column flange. 3. The use of 12 mm thick end-plate together with M20 bolts resulted in the deformation of the end-plate first followed by the column flange. However, for 15 mm thick end-plate used together with M24 bolts, the failure mode of the connection was shifted to the column flange. 4. The increase in moment resistance and stiffness of the connection depended on the thickness of the column flange not the size of the column. 5. No visible deformation or shearing effect on the column web could be seen for all tested specimens. 6. All tested specimens showed a ductile behaviour as there was no abrupt failure and the formation of plastic hinge occurred at the connection. 7. The theoretical values calculated by component method showed good agreement with the experimental results for Perwaja Steel Sections.

Acknowledgment This study was part of a research work carried out towards M.Phil by one of the authors. The overall research was funded by Steel Technology Centre, Universiti Teknologi Malaysia, Skudai, Malaysia. Special thanks to all technicians in Structural Engineering Laboratory who were involved in this project.

References

Abdalla, K.M., Abu-Farsakh, G.A.R., Barakat, S.A., “Experimental investigation of force-distribution in highstrength bolts in extended end-plate connections”, Steel and Composite Structures, SCS, Vol. 7, No. 2, 2007, pp 87-103. Abdalla, K.M., and Chen, W.F., “Expanded Database of Semi-Rigid Steel Connections”, Computer and Structures, Vol. 56, No. 4, 1995, pp 553-564.

14

Mahmood Md. Tahir

Al-Jabri, K.S., Burgess, I. W., Lennon, T. and Plank, R.J. “Moment Rotation-Temperature Curves for Semi-Rigid Joints” , Vol. 61, 2005, pp 281-303. Allen , Steel Construction Institute and British Constructional Steelwork Association Limited. 1995. Joints in Steel Construction. Volume 1: Moment Connections. Ascot, Berks: Steel Construction Institute. Bose, B. “Tests to verify the performance of standard ductile connections”, Dundee Institute of Technology, 1993. British Standards Institute BS 5950-1. 2000. Structural Use of Steelwork in Building Part 1: Code of Practice for Design-Rolled and Welded Sections. London: British Standards Institution. British Standards Institute BS EN 10113. 1993. Hot rolled products in weldable fine grain structural steels. London: British Standards Institution. Eurocode 3: EN 1993-1-1. 2005, Design of Steel Structures: General Rules and Rules for Buildings. Brussels: British Standards Institution. Chen, W.F. (ed.) 1993. Semi-rigid Connections in Steel Frames-Council on tall Buildings and Urban Habitat. New York: Mc Graw-Hill. Chen, W.F. and Kishi, N. “Semi-rigid Steel Beam-toColumn Connections: Database and Modelling”. Vol. 115, No. 1, 1989, pp 105119. Coelho, A.M.G., Bijlaard F.S.K. and Silva, L.S.D. “Experiemental assessment of the ductility of extended end-plate connections”. , Vol. 26, No. 9, 2004, pp 1185-1206. Couchman, G.H. 1997. Design of Semi-Continuous Braced Frames. Ascot, Berks: Steel Construction Institute. Hussein, M.A. 2001. “Performance of connections on Major Axis using Local Sections”. , Universiti Teknologi Malaysia, Malaysia. Jones, S.W., Kirby, P.A. and Nethercot, D.A. “The Analysis of Frames with Semi-Rigid Connections-A State of the Journal of Constructional Steel Research

et

al

Journal

of Structural Engineering,

Journal of Enginering Structures

M.Phil. Thesis

et al.

Art Report”. , Vol. 3, No. 2, 1983, pp 2-13. Mahmood, Md T, Hussein, A M; 2008 (Dec), “Experimental Tests on Extended End-Plate Connections with Variable Parameters”. , Vol. 8, No. 4, pp 369-381. Nethercot, D. and Zandonini, R., “Methods of Prediction of Joint Behaviour”. In: Narayanan, R. (ed). . Essex: Elsevier Applied Science, 1989, pp 23-62. Sayed-Ahmed, E.Y., “Design aspects of steel I-girders with corrugated steel webs” , EJSE, Vol. 7, 2007, pp 27-40. Shi, W.L., Li, G.Q., Ye, Z.M. and Xiao, R.Y., “Cyclic Loading Tests on Composite Joints with Flush End Plate Connections”, , Vol. 7, 2007, pp 119-128. Sulaiman, A. 2007. “Behaviour of partial-strength connection in semi-continuous construction for multistorey braced steel frame using TWP sections”. Universiti Teknologi Malaysia, Malaysia. Tahir, M.M. 1997. “Structural and Economic Aspects of the Use of Semi-Rigid Joints in Steel Frames”. . University of Warwick, UK. Tahir, M.Md., Sulaiman A, Anis S, 2008 (March) “Experimental Tests on Composite and Non-Composite Connections Using Trapezoid Web Profiled Steel Sections”. , Vol. 8, No. 1, pp 43-58. Tahir, Sulaiman, and Saggaff, “Structural Behaviour of Trapezoidal Web Profiled Steel Beam Section Using Partial Strength Connection”, , Vol. 8 pp 55-66. Weynand, K., Jaspart, J.P., Steenhuis, M., “Economy Studies of Steel Frames with Semi-Rigid Joints”. , Vol. 46, No. 1-3, 1998, Paper No. 63. Journal of Constructional Steel Research

International

Journal

of

Steel

Structures

Structural

Connections-Stability

and

Strength

Electronic Journal of Structural

Engineering

International Journal of Steel Structures

PhD

Thesis.

PhD Thesis

International Journal of Steel Structures

Electronic

Journal

of

Structural Engineering

Journal

Constructional Steel Research

of