A DESIGN PROCEDURE FOR BOLTED TOP-AND-SEAT ANGLE CONNECTIONS FOR USE IN SEISMIC APPLICATIONS

Transcription

1 A DESIGN PROCEDURE FOR BOLTED TOP-AND-SEAT ANGLE CONNECTIONS FOR USE IN SEISMIC APPLICATIONS Jared D. Schippers, Daniel J. Ruffley, Dr. Gian A. Rassati, and Dr. James A. Swanson School of Advanced Structures, University of Cincinnati, Cincinnati, OH ABSTRACT Since the 1994 Northridge and 1995 Kobe earthquakes, bolted moment connections have garnered considerable interest for their application in Seismic Lateral Resisting Systems (SLRS). However, the considerable amount of research conducted over the last two decades has not produced many design procedures that would allow the applications of bolted connections either as fully-restrained or partially-restrained. This paper outlines a step-by-step design procedure for the design of bolted top-andseat angle moment connections for seismic applications. The proposed procedure is used to design three practical examples of top-and-seat angle connections: two fullstrength and one partial-strength. The connections are then are modeled in ABAQUS following a validated modeling approach that has been verified against multiple experimental tests, both quantitatively and mechanistically. The analysis results of these models are subsequently compared to the expected outcomes from the design procedure, as a proof-of-concept. The results of this comparison are presented and commented, and it is concluded that the proposed procedure is suitable for the design of top-and-seat angle connections for seismic applications. 1. INTRODUCTION In the wake of the 1994 Northridge and 1995 Kobe earthquakes, numerous moment connections were investigated and studied. The earthquakes demonstrated that welded moment connections were far more brittle than previously thought, and as a result there arose an increased interest in bolted moment connections. Moment connections can be classified in terms of strength, stiffness, and ductility. For strength, a connection is considered full-strength (FS) if the connection has enough capacity so the beam can develop a full plastic hinge. If the capacity of the connection is not enough for this to occur, it is considered to be partial-strength (PS). Concerning stiffness, a connection is considered fully-restrained (FR), partiallyrestrained (PR), or simple, depending on the relative rotational stiffness of the connection with respect to the connected beam. When the initial stiffness is greater than 0EI/L, the connection is FR. If the initial stiffness is less than EI/L, the connection is simple. Anything between these two limits classifies the connection as PR. Finally, a connection is considered ductile if it has at least 80% of its nominal strength at a plastic rotation of 0.03 radians. Figure 1 shows a full-strength, partiallyrestrained, ductile connection (Swanson, 1999). Currently, only full-strength, fully-restrained moment connections are allowed for use in seismic lateral resisting systems per ANSI/AISC (010) in intermediate

2 moment frames (IMF) and special moment frames (SMF), and all other connections must be considered simple gravity connections. Previous research has shown that accounting for the moment contribution of these gravity connections in moment frames adds a considerab le amount of lateral resistance during a seismic event (Barber, 011, and Zhang, 01). With more and more research going into PR connections and frames consisting of PR connections s, it is anticipated that the contribution of lateral resistance from PS, PR connections will eventually be allowed to be incorporated in seismic design per ANSI/AISC 341. Additionally, it is envisioned in the future that the primary Figure 1: Moment-Rotation Curve lateral resisting (Swanson and Leon, 000) system in SMFs and IMFs will be permitted to also consist of FS, PR connections. Given these assumptions, thiss paper presents a general design procedure for bolted top-and-seat angle connections for use in seismic design. The design procedure has been verified through finite element modeling, both quantitatively and mechanistically, using the software ABAQUS. Three example connections, two FS and one PS, have been designed using the proposed procedure and modeled in ABAQUS. The results are then presented and commented.. BACKGROUND In 1995, after the Northridge and Kobe earthquakes, the SAC Joint Venturee and FEMA entered into an agreement to further research in seismic design pertaining to steel moment frames and connections (FEMA, 000). SAC Subtask 7.03 was performed at the Georgia Institute of Technology and was concerned with bolted T-stub and TSA connections. As part of this research, 8 full-scale T-stub connections, fullangle scale TSA connections, 48 bolted T-stub components, and 10 bolted clip components weree experimentally tested (Smallidge, 1999; Swanson, 1999; and Schrauben, 000). All full-scale tests were performed cyclically and the component tests led directly to Swanson (1999) developing thee Modified Kulak Model for predicting prying forces in T-stub connections. Swanson and Gao (000) later developed a similar prying model for predicting prying forces in heavy clip angle components, using previously compiled data from SAC subtask 7.03.

3 Schrader (010) compiled the documentation to prequalify bolted T-stub connections as FR connections for use in IMFss and SMFs per the provisions of ANSI/AISC (010). He used the moment-rotation to using existing data to meet the criterion for prequalifying a connection, a design procedure was created. This design procedure implemented the Modified Kulak Model and is currently being reviewed by the AISC and other experimental data gathered from SAC Subtask In addition Connection Prequalification Review Panel (CPRP). The design procedure outlined in this paper is molded after the design procedure in Schrader (010). 3. TOP-AND-SEAT ANGLE DESIGN PROCEDURE 3.1 Methodolog gy for Design Procedure In order for AISC-CPRP to prequalify a connection, the connection must be qualified as FR to be considered for use in SMFs and IMFs. Previous TSA experiments have shown insufficient stiffness to be classified ass FR, so this paper outlines a design procedure under the assumption that future provisions will allow the use of PR connections in seismic design. Under this assumption, this procedure is based on mechanistic principles and mostly follows provisions in ANSI/AISC (010) and ANSI/AISC (010). The portion of the procedure considering prying usess the Modifiedd Kulak Model developed by Swanson and Gao (000)) and Gao (001). Figure : Typical TSA Connection Figure 3: System Detail showing plastic hinge location 3. General Top and seat angle (TSA) connections use a top angle and seat angle to provide the moment resistance in the connection. The angles are connected to the column and beam flanges by high-strengt th bolts as shown in Figure. The top and seat angles must be identical so the connection has equal resistance for a negative or positive moment. The shear connector is designed to carry all the shear resistance in the

4 connection. The shear connector shown in Figures and 3 is a shear plate bolted to the beam flange and welded to the column flange. Due to the length limitations for this paper, a detailed list of all system limitations, provisions, and requirements could not be included. Most off these items strictly follow current standards in ANSI/ /AISC (010), ANSI/AISC (010), and ANSI/AISC (010). For a detailed design procedure listing all these items, see Schippers (01). Figure 4: Column and Shear Tab Details Figure 5: Beam Details Figure 6: Angle Details 3.3 Design Procedure For commentary on design procedure, see Schippers (01). Step 1: Compute the maximumm expected moment (occurs at the beam hinge) % 1,

5 For a FS design, PS % equals 100%, or 1. R y =R t =1.1 per ANSI/AISC (010). Step : Compute the maximum shear bolt diameter To ensure a ductile failure in the beam, the following must be met: 1 " Step 3: Determine the preliminary shear strength per bolt. " 7 Bolt Shear.4 Beam Bearing.4 Angle Bearing 8 Step 4: Estimate the number of shear bolts needed for each beam flange Step 5: Estimate the location of the plastic hinge in the beam. 10, 11 Step 6: Calculate the shear force at the plastic hinge in the beam Step 7: Find the expected moment and corresponding force at the column face

6 Step 8: Approximate the thickness of the angles and size of the tension bolts. 17 " Step 9: Determine a preliminary configuration for the angles., ,, ,, 5 5 Step 10: Find the required thickness of the angle when considering prying. 7 Three limit states can control the tensile capacity of the connection. For more information on these limit states, see Swanson (1999) Step 11: Compute the actual force in the horizontal angle leg. (34)

7 Step 1: Confirm that the shear bolts provide adequate resistance. (35) Step 13: Back-check the capacity of the horizontal angle leg. Check that (in the order shown): gross section yielding, net section fracture, and compressive yielding or buckling. (36) 1 16 (37) 0.75, , (38) If / 5, compressive yielding governs, is the exact same as in gross section yielding. If / 5, flexural buckling governs and the provisions of Section E3 of the ANSI/AISC (010) apply. Step 14: Back-check all three limit states for tensile failure defined in step 10 (ϕt 1,,3 ). Step 15: Finalize Design. Lastly, bearing and tear-out and block shear in the beam flange and horizontal angle leg should be checked in accordance with Sections J3.10 and J4.3 in the ANSI/AISC (010). Also, the shear connection needs to be detailed accounting for eccentricity. All applicable shear limit states should be checked per Chapter J in ANSI/AISC (010). Panel zone strength shall be in accordance with Section.4.4 and 6 in ANSI/AISC (010). Finally, lateral bracing requirements shall meet the lesser length found in either ANSI/AISC 360/10 (010) or ANSI/AISC (010). 4. FINITE ELEMENT MODELING (FEM) Figure 8: FS-01 Comparative Plots Figure 9: FS-0 Comparative Plots

8 4.1 Modeling Existing Experimental Data As previously mentioned, Schrauben (000) tested two full-scale TSA connections and both of these experiments have been modeled in ABAQUS. For a detailed summary of the modeling procedure, see Ruffley (011). Figures 8 and 9 show both force-displacement results of Schrauben s (000) experimental data and the curve obtained by modeling the same connections in ABAQUS. It can be observed that the modeling procedure produced highly accurate results. Ruffley (011) also modeled component tests that Swanson (1999) tested, in order to verify the procedure s capability of predicting various failure modes, obtaining very satisfactory results. 4. Modeling New Connections In an attempt to verify the accuracy of the design procedure, three new connections have been modeled using the procedure outlined in Ruffley (011). Two were FS, and one was PS (60%). Table 1 shows the summary of calculations of the three connections designed using the proposed procedure. Table shows comparative results between the predicted forces computed in the design procedure and actual forces from analyzing the models. It should be noted that the analyses of all three models showed no signs of block shear in the beam flange or horizontal angle leg, which verifies the expected over-conservative nature of the block shear resistance calculation for this connection. Schrader (010) had similar conclusions when analyzing T-stub connections concerning block shear. For this reason the design procedures allows a 10% reduction in F f when designing block shear. Prying forces were calculated by taking each element stress multiplied by its corresponding area, and then summing the forces for all elements in a cross section of a tension bolt. The plastic mechanism in the angle, block shear, gross section yielding, and net section fracture were all analyzed by visually inspecting the equivalent plastic strain contours. Shear bolt forces were analyzed calculating the actual force transmitted by the horizontal leg of the angle. This force was calculated by summing the stress in each element of the horizontal leg of the angle and multiplying it by the element s crosssectional area. It was assumed for the sake of simplicity that all shear bolts carried an equal load. All values in Table correspond to the instant in which the beam in the model develops M pr as calculated in step 1 at the expected hinge location calculated in Step 5 of the design procedure. 4.3 Modeling Results The FS W16x31 and PS W4x6 connections were both anticipated to be controlled by tension bolt capacity, which is precisely what the models verified. The quantitative errors in these two models were 11% and 9%, respectively. For the FS W18x35 model, the limiting state was expected to be formation of a plastic hinge in the top angle, and the model showed correspondingly signs of widespread inelastic deformation. The fact that the angle had yielded indicates that the capacity is being approached, although model is not capable of quantifying it explicitly. From a visual inspection, it is concluded that the prediction of plastic hinges forming in the angles is accurate, so the model is deemed to reproduce the predicted outcome. It should also be noted that the analyzed prying forces in the tension bolts were within 1% of the expected forces from the design procedure, once again validating its accuracy.

9 Table 1: Design Procedure Results DESIGN PROCEDURE RESULTS Beam Size W16x31 W18x35 W4x6 Column Size W14x11 W14x11 W14x11 FS or PS Design FS FS PS (60%) Shear Bolt Size, d vb (in) 3/4 7/8 7/8 Shear Bolt Grade A490X A490X A490X No. Shear Bolts, n vb Tension Bolt Size, d tb (in) 1 1/4 1 1/ 1 1/ Tension Bolt Grade A490X A490X A490X No. Tension Bolts, n tb Tension Bolt Gauge, g t (in) 7/8 3 1/4 3 1/4 Angle Width, W a (in) Angle thickness, t a (in) /8 Set-back, SB (in) 3/8 3/8 3/8 M pr (k-in) M f (k-in) F f (kips) φt 1 resistance (kips) φt resistance (kips) φt 3 resistance (kips) Prying effect, φt 3 /φt (%) 10.% 7.% 5.% Bearing resistance (kips) Block shear resistance (kips) Angle GSY resistance (kips) Angle NSF resistance (kips) Shear bolt resistance (kips) Table : Table : Analysis Results DESIGN PROCEDURE AND FEM COMPARITIVE RESULTS W16x31 FS Connection Expected prying force in tension bolts 1.10(40) = 65 kips Actual bolt tensile force at M pr 93 kips Percent error Plastic Hinge Devolpment (φ T 1 ) at M pr Actual force in shear bolts at M pr Block Shear GSY, NSF W18x35 FS Connection 11% Minimal Hinge Development 194 kips No inelastic deformation (I.E.) No I.E., No I.E. Analysis shows hinge development Expected Limiting State, φt 1 in angle; limit state has not been reached but is being approached Expected pyring forces in bolts, φt 1.7(70) = 343 kips Bolt Tensile Force at M pr in F.E.M. 384 kips Percent error 11.9% Actual force in shear bolts at M pr 46 kips Block Shear No inelastic deformation (I.E.) GSY, NSF No I.E., No I.E. W4x6 PS (60%) Connection Expected prying force in tension bolts 1.5(85) = 357 kips Actual bolt tensile force at M pr 389 kips Percent error 9.0% Plastic Hinge in Angles ( φt 1 ) minimal Actual force in shear bolts at M pr 6 kips Block Shear No inelastic deformation (I.E.) GSY, NSF No I.E., No I.E. 5. CONCLUSIONS A design procedure has been presented for full-strength and partial-strength bolted top-and-seat angle beam-to-column connections for use in seismic design. It has been discussed that the moment contribution of these connections currently cannot be incorporated in IMFs and SMFs, but it is envisioned that this could change in the future. With this assumption, the design procedure closely follows the connection prequalification standards in ANSI/AISC (010). The accuracy of the procedure has been demonstrated through finite element modeling in ABAQUS by a modeling procedure that had been verified through actual experimental data. Using this modeling procedure, two full-strength and one partial-strength connections were designed using the proposed procedure and modeled in ABAQUS. The results clearly show the quantitative accuracy of the design procedure, especially concerning prying forces in the tension bolts. In addition, all three models show that the predicted limit state was the likely cause of failure, verifying the mechanistic accuracy of the design procedure. Future research should include physical experiments of the three connections discussed in this paper, which would provide the ultimate verification of the design procedure s accuracy.

10 REFERENCES ANSI/AISC (010). Specification for Structural Steel Buildings, American Institute of Steel Construction, Chicago, IL. ANSI/AISC (010). Prequalified Connections for Special and Intermediate Steel Moment Frames for Seismic Applications, American Institute of Steel Construction), Chicago, IL. ANSI/AISC (010). Seismic Provisions for Structural Steel Buildings, American Institute of Steel Construction, Chicago, IL. Barber, (011). Contribution of Shear Connections to the Lateral Stiffness and Strength of Steel Frames, MS Thesis, School of Advanced Structures, University of Cincinnati, Cincinnati, OH. FEMA-350 (000). Recommended Seismic Design Criteria for New Steel Moment- Frame Buildings, Federal Emergency Management Agency, Washington D.C. Gao, X. (001). Strength Determination of Heavy Clip-Angle Connection Components. MS thesis, Dept. of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, Ohio. Ruffley, D.J. (010). A Finite Element Approach for Modeling Bolted Top-and-Seat Angle Components and Moment Connections. MS Thesis, School of Advanced Structures, University of Cincinnati, Cincinnati, Ohio. Schippers, J.D. (01). A Design Procedure for Bolted Top-and-seat Angle Connections for Use in Seismic Applications. MS Thesis, School of Advanced Structures, University of Cincinnati, Cincinnati, Ohio. Schrader, C.A. (010). Prequalification and Design of Rolled Bolted T-stub Connections in Moment Resisting Frames. MS Thesis, School of Advanced Structures, University of Cincinnati, Cincinnati, Ohio. Schrauben, C.S. (1999). Behavior of full-scale bolted beam-to-column T-stub and clip angle connections under cyclic loading. MS Thesis, School of Civil and Environmental Engineering. Georgia Institute of Technology, Atlanta, Georgia. Smallidge, J.M. (1999). Behavior of Bolted Beam-to-Column T-stub Connections under Cyclic Loading. MS Thesis, School of Civil and Env. Engrg, Georgia Institute of Technology, Atlanta, Georgia. Swanson, J.A. (000). Characterization of the strength, stiffness, and ductility behavior of Tstub connections. PhD Dissertation, School of Civil and Env. Engrg. Georgia Institute of Technology, Atlanta, Georgia. Swanson, J.A. and Gao, X. (000). Strength Determination of Heavy Clip-Angle Connection Components, in Connection in Steel Structures IV: Steel Connections in the New Millennium, Roanoke, VA. Oct, 000. Swanson, J.A., and Leon, R.T. (000). Bolted Steel Connections: Tests on T-stub Components. Journal of Structural Engineering, 16(1): Zhang, W. (01). Study of the Influence of Gravity Connections on the Lateral Response of Steel-Concrete Composite Moment Frames, MS Thesis, School of Advanced Structures, University of Cincinnati, Cincinnati, OH.