Dcb 68 Sliding Hinge t 1p1p5w

17-19 Gladding Place P O Box 76 134 Manukau City, New Zealand Phone: Fax: Email:

+64-9-262 2885 +64-9-262 2856 [email protected]

No. 68

June/July 2002

The author(s) of each article in this publication are noted at the beginning of the article.

The procedure detailed herein has been the subject of review by a number of people. The effort and input of these reviewers is greatly appreciated.

Introduction

In This Issue

As readers will be aware, HERA and the University of Auckland are engaged in a long-term research project aimed at developing new forms of semirigid ts for moment-resisting, steel framed seismic-resisting systems (MRSFs). Two t types have been developed from this programme as the preferred options for the beam to column connections of MRSFs. These are the Flange Bolted t (FBJ) and the Sliding Hinge t (SHJ). The experimental and analytical phases of this project are now completed and the final phase (writing up and presenting of results) has begun. The FBJ was the first t to be developed. Design and detailing procedures for it have already been published – in DCB No. 58, principally, and with a minor corrigenda in DCB No. 62 and an extension to its original scope of application in DCB No. 64. This t has been used in at least two building developments (one in Auckland and one in Napier), which was the intention behind the design and detailing requirements being published prior to the release of the full research report [1]. With the completion of the analytical work on the SHJ, the research has now reached the stage where final design and detailing provisions for the SHJ can be made. This issue presents these recommendations, covering the design and detailing of the t itself and the design of moment-resisting steel framed system incorporating the t. It also presents a detailed design example on a particular SHJ.

The Sliding Hinge t

1

Member Compression Capacity of a Solid Section

33

References

33

The Sliding Hinge t: Design and Detailing Provisions and Design Example This article has been written by G Charles Clifton, HERA Structural Engineer, John Butterworth, Senior Lecturer at the University of Auckland Department of Civil and Resource Engineering and Tanja Miller, Undergraduate Student from the Fachhochschule Weingarten on Study Leave (Industrial Practice) at HERA.

1.

Introduction and Scope of Article

1.1

Brief history of the overall project

HERA and the University of Auckland are in the final stages of a long-term research project aimed at developing innovative new forms of semi-rigid ts for moment-resisting steel framed seismicresisting systems (MRSFs). These ts are designed and detailed to achieve the following performance characteristics: •

Remain fully rigid up to the design level serviceability limit state earthquake moment

•

Remain reasonably rigid above the serviceability limit state level and up to the design level ultimate limit state earthquake moment

•

Allow inelastic rotation between beam and column to occur when the design ultimate limit state earthquake moment is exceeded

Also covered is a short article on a specific design issue that has arisen in recent times.

HERA Steel Design & Construction Bulletin

Page 1

Page

No. 68, June/July 2002

•

Be able to withstand the inelastic rotation demand associated with the design level earthquake with negligible damage, such that the post-design earthquake building response under serviceability conditions is not significantly affected

•

Withstand greater levels of rotation demand with increased damage but not failure.

Of the five t types that have been researched for this project, two t details have emerged as preferred options for the beam to column connections of MRSFs. These are the Flange Bolted t (FBJ) and the Sliding Hinge t (SHJ). These two ts are designed and detailed to meet the performance criteria in different ways. Very briefly: •

•

The FBJ is designed for higher strength, low design ductility demand applications. It is very simple to fabricate and erect. It has a low inelastic rotation damage threshold, but is capable of withstanding high levels of inelastic rotation demand if necessary. The SHJ is designed for lower strength, high design ductility demand applications. It is slightly less simple than the FBJ to fabricate and more complex to erect and is designed to withstand fully ductile levels of design inelastic rotation with minimum damage.

The FBJ development was completed in 2001. Guidance on design and detailing of the FBJ and MRSFs incorporating the FBJ has been given in DCB No. 58, with a minor corrigenda in DCB No. 62. In the latter half of 2001, it became apparent from the numerical integration time history (NITH) analyses that the originally proposed scope of application of the FBJ, which was for low ductility demand applications only, could be widened, and work on this was undertaken, with the results published in DCB No. 64. Up to the end of 2001, all NITH studies were undertaken in accordance with NZS 4203:1992 [2]. However, the March 2002 version of the draft replacement to that standard, which has been under development for several years, contained detailed guidance on the selection and scaling of earthquake records for NITH. The selection and scaling of earthquake records used up to the end of 2001 was rather ad-hoc (see details in section 6.4 of HERA Report R4-88 [3] and summary details in section 3.4, pp. 16-17 of DCB No. 64) and so, in 2002, the opportunity has been taken to use the provisions of DR1170.4 [4] to produce a revised suite of earthquake records and scale factors. Details of these will be summarised in DCB No. 69. The FBJ designs were then reanalysed under this new suite of earthquake HERA Steel Design & Construction Bulletin

Page 2

records. Given that the records were selected and scaled in accordance with the new draft, [4], while the FBJ frames had been designed to the existing standard [2], some comparative 5 and 10 storey frames were redesigned to the new standard to see the differences in seismic design actions, P - ∆ effects and subsequent member sizes. The member sizes for a given application turned out to be the same from both standards for each case studied. For the SHJ NITH studies, the frames were all designed and analysed to the draft provisions. Because the suite of earthquake records cover three soil/fault conditions, the designs were undertaken for these conditions. The three conditions covered were: (1) (2) (3)

Class C – shallow soil [4] – with near fault action Class C – shallow soil [4] – without near fault action Class D – soft soil [4] – without near fault action

Designs were undertaken for two seismic zones (Auckland, Wellington). The near fault action option is only applicable to Wellington. The SHJ NITH studies were completed in June 2002. With their completion, the design and detailing provisions for the SHJ have been finalised and are presented herein. Summary details of the NITH studies and the frame options will be given in DCB No. 69. Writing up of the entire project is also progressing concurrently and is due for completion in the first quarter of 2003 [1]. A summary paper [7] of the research into both ts and systems was presented at the 2001 NZSEE Technical Conference. 1.2

Scope of This Article

This article presents the design and detailing provisions for the SHJ and for MRSFs using the SHJ. The former is presented in section 3 and the latter in section 4. This is followed with a SHJ design example, in section 5. However, prior to presenting these provisions, this article looks briefly at the performance of SHJs in severe earthquakes, in of the design philosophy, target performance requirements and behaviour from experimental tests. These issues have already been mentioned in DCB No. 59 and that article will be cross-referenced as appropriate. They will also be covered in detail in the thesis report [1] on the whole project.

No. 68, June/July 2002

Fig. 68.1 Sliding Hinge t: Isometric and Exploded View

HERA Steel Design & Construction Bulletin

Page 3

No. 67, June/July 2002

Fig. 68.2 Layout and Notation for the Sliding Hinge t

Fig. 68.3 Lever Aims for Moment Capacity Determination

HERA Steel Design & Construction Bulletin

Page 4

No. 68, June/July 2002

moment from the large-scale test 4 and in Fig. 68.4 herein, which shows the t rotation versus moment from the large-scale test 3.

Before commencing with the performance in earthquakes, a description of the SHJ is in order. Fig. 68.1 shows an isometric and exploded view of the t. Fig. 68.2 shows an elevation with the layout and notation, while Fig. 68.3 shows the lever arms for determination of the t moment capacity. 2.

Performance of the Sliding Hinge t in Severe Earthquakes

2.1

Design philosophy operation

and

modes

On rotation reversal, the t unloads abruptly, then the moment capacity builds up in the reverse direction, as shown in Fig. 59.28 or 68.4. The increase in moment with increasing reverse rotation occurs in two stages; one as sliding occurs along the first interface (beam to plate) and then with a further increase in shear capacity as the second interface (plate to cap plate) is activated.

of

The slotted hole is designed to accommodate a t rotation of ± 30 mrad (radians x 10 -3) multiplied by an over rotation factor of 1.25; if the inelastic rotation demand exceeds this, the t undergoes further inelastic behaviour through flange plate yielding, in the same manner as for the FBJ (see DCB Issue No. 58). The first largescale SHJ specimen, tested to destruction in test 2, still developed its design moment capacity at over 120 mrad rotation!

The design philosophy behind this t has been to establish dependable behavioural characteristics for the SHJ and for the MRSF system for the serviceability limit state condition and for two levels of ultimate limit state conditions. These are described in section 2.3. The first level of ULS condition is the design level ultimate limit state earthquake, as stipulated by NZS 4203 [2] or DR 1170.4 [4] and the second is the more severe maximum considered event. All the experimental and analytical work undertaken on the SHJ has been planned and executed with this philosophy in mind.

Under the design level ULS earthquake, inelastic rotation demand is expected to be not greater than the 37.5 mrad accommodated within the slotted holes. At this level of rotation demand, minimum t degradation will occur and only minor slab cracking, such that no post-earthquake repair is required.

The mode of operation of the SHJ is relatively simple. The beam is pinned laterally at the top flange level, using nominal sized bolt holes and FBJ details. This keeps lateral movement in the floor slab to 2-3 mm, thus minimising undesirable floor slab participation and slab damage. t rotation is achieved through sliding at the bottom flange and the web bottom bolt level (see Fig. 68.1 for the location of these components and Fig. 64.10, DCB No. 64, for an illustration of this mechanism).

Under the maximum considered event (MCE), the MRSF with SHJs will retain its integrity, to allow evacuation and post-earthquake assessment, but will suffer controlled t damage, which may necessitate replacement of components. However, the results from the NITH studies show that, in most instances, little or no reinstatement would be needed after most maximum considered events, especially for buildings not subject to near fault action.

The sliding details are shown in the isometric view of Fig. 68.1. The sliding layers are between the brass shims and plate (web plate, bottom bolts for bottom flange plate). The holes for the web bottom bolts in the web plate and for the bottom flange bolts in the bottom flange plate are slotted to allow this sliding to occur. The beam flange or web and the associated cap plates all have nominal sized holes.

In of the force based seismic design philosophy of [2, 4], the design procedures developed for these semi-rigid systems utilise either the equivalent static or modal response spectrum methods, in conjunction with NZS 3404 [5] and, where appropriate, HERA Report R4-76 [6]. The preliminary sizing / design method, in particular, is easy and rapid to use. These procedures are given in sections 3 and 4 below.

When the moment demand on the SHJ from earthquake generates internal beam axial forces which exceed the sliding resistance available through the bottom flange bolts and web bottom bolts, the t will slide, allowing beam rotation to occur. As sliding occurs, the cap plate is anchored in position relative to the beam flange or web by the bolts, allowing the cap plate to also slide relative to these surfaces. Once the imposed moment reduces, there comes a point where the sliding stops and the t becomes rigid again. This is illustrated in Fig. 59.28 of DCB No. 59, which shows the t rotation versus HERA Steel Design & Construction Bulletin

2.2

Design role of t components

This is described in section 3.3.2 herein. 2.3

Performance characteristics

The MRSFs with SHJs have been developed to deliver the following performance characteristics for the three levels of earthquake described in section 2.1. Page 5

No. 68, June/July 2002

(1)

(2)

which is the only location likely to be subjected to appreciable inelastic demand)

For the serviceability limit state earthquake (ie. as represented by DR 1170.4 [4] Section 2.1.1, involving a return period of 20 years for normal structures (as defined by Table 3.1 of AS/NZS 1170.0 [8])): (i)

The t and system shall remain effectively rigid, with negligible inelastic action from any component

(ii)

This condition shall apply even when the system has been subjected to a prior ultimate limit state design level event.

For the design level ultimate limit state earthquake (ie. as represented by [4, 8] involving a return period of 500 years for normal structures): (i)

Negligible inelastic demand in the beams

(ii)

Minimal inelastic demand in the columns at base level (such that fixed column bases will be readily repairable) and none at higher levels

(iii)

The rotation demand on the ts is not to cause the bottom flange bolts to the ends of the slotted holes

(iv)

Column zone rotation demand to be ≤ 1%

(v)

Lateral drift not to exceed 2%

(vii)

The positioner replacement

bolt

may

need

For the maximum considered earthquake (ie. based on a 2000 year return period event or higher): (i)

Negligible inelastic demand in the beams, except in the vicinity of bolts to the flange and web plates

(ii)

Inelastic demand in the columns to be able to be dependably resisted (this applies especially at the base,

HERA Steel Design & Construction Bulletin

(iv)

zones may rotate in excess of 1% strain demand

(v)

Lateral drift to be within sustainable limits, including the influence of P - ∆ effects

(vi)

The positioner replacement

(vii)

Minor cracking only to the concrete floor slab surrounding the frame.

bolt

will

need

(a)

Analysing the frame for the design level earthquake using the Equivalent Static Method or the Modal Response Spectrum Method from [2,4], and sizing the and connection components to meet the required strength and stiffness criteria for this event

(b)

Following the t design and detailing provisions given herein (section 3) such that the t can sustain the MCE rotational demands while delivering the performance characteristics of (3) above.

2.4

SHJ behaviour from experimental tests

There has been extensive experimental testing undertaken on the SHJ, involving both small-scale component and large-scale assemblage tests. Some details of the large-scale tests are given on pages 26-30 of DCB No. 59 and a very brief overview of these large-scale tests is given in section 3.3 of [7]. Details of the small-scale component tests are given on pages 28, 29 of DCB No. 64.

(viii) Minor cracking only to the concrete floor slab surrounding the frame. (3)

In the extreme case, t rotation demand may cause the bolts to impact the ends of the slotted holes, requiring replacement of the sliding bolts and possibly bottom flange plate replacement

Application of the design procedures for the forcebased method of design involves:

P - ∆ effects to be ed for either through provision of suitable frame stiffness (ie. satisfying Equation 6.1 (1) of [4]) or through increased strength (ie. satisfying Clause 6.5.4 of (4)).

(vi)

(iii)

There has also been extensive finite element analysis (FEA) modelling of the sliding hinge t sliding assemblage. This work was undertaken in two stages; that from 2001 is summarised on pages 24-33 of DCB No. 64 and presented in full detail in HERA Report R4-110 [9]. That from 2002 will be summarised in DCB No. 70 and presented in [1].

Page 6

No. 68, June/July 2002

SHJ Test 3, 04/08/2000, Plastic Rotation vs Moment and Simultest3 from Hysteresis Model 800 600

Moment [kNm]

400 200 0 -30

-25

-20

-15

-10

-5

0

5

10

15

20

25

30

-200 -400 -600 -800 Rotation [mrad] Test 3 experimental data

Simultest3

Fig. 68.4 Experimental and Simulated Moment-Rotation Behaviour for Large-Scale Tests Without Belleville Springs • • • •

Fig. 68.4 shows the moment-rotation characteristics from the large-scale test 3, which involved the final proposed t configuration without Belleville Springs to the bottom flange bolts. The moment-rotation characteristics of the SHJ are markedly different to those of any other semi-rigid t, because of the two stage sliding from the sliding components. In order to accurately represent the t behaviour in the NITH analyses, a mathematical model of the moment-rotation characteristics has had to be developed and implemented into the computer program, RUAUMOKO [10] used for the NITH analyses. This has been done; see details in [11]. The simulated moment from that model generated by the experimental rotations from test 3 is also shown in Fig. 68.4.

bolt layout and orientation flange plate and cap plate thickness presence/absence of Belleville Springs effect of loading rate: seismic-dynamic and pseudo-static • effect of repeated loading on assemblage, including after a delay time of 4 weeks One of the component experimental test results is shown in Fig. 64.16, DCB No. 64. On the basis of these component tests, a bolt design model has been developed to give the bolt sliding shear capacity. Details of that model are given on pages 29, 30 of DCB No. 59. The basic mechanisms assumed for that model were confirmed by FEA modelling, as described in [9] and more briefly on pages 24-33 of DCB No. 64.

As the large-scale experimental tests could only investigate one size and layout of bolt, plate and cap plate and only at a pseudo-static rate of loading, a series of small-scale tests on the bottom flange sliding assemblage were undertaken during 2000/2001 on representative connections to determine the influence of the following parameters.

The completion of the experimental testing programme, FEA modelling and NITH studies has allowed the design and detailing provisions for the SHJ and the MRSF systems incorporating the SHJ to be finalised. These are given in the next two sections, starting with the design and detailing of the t itself, in section 3.

• bolt size – range from M24 to M30 HERA Steel Design & Construction Bulletin

Page 7

No. 68, June/July 2002

3.

Design and Detailing of the Sliding Hinge t

3.1

General

Section 3.2 presents these requirements. It starts with limitations on the flange and web plate grade and thickness, followed by the material selection for the brass shims.

The SHJ is intended for high ductility demand; µdesign = 4 is used. In theory it is possible to use µ = 6, the maximum allowed from [2 or 4]. The t’s ductility capacity is more than adequate for this. However, as noted in section 2.3 (1), one of the performance criteria set for the t is to remain effectively rigid, even after the t has been subjected to a design level ultimate state earthquake.

This is followed by edge distances, bolt pitches and gauges, then by the very important provision of clearance between the beam face and the column flange.

Such an event is associated with some permanent softening of the t, hence the decision to use µdesign = 4 as the ULS design ductility.

Section 3.2 ends with surface treatment requirements for the ply surfaces.

The dimensioning of all components is then covered. This is followed by aspects of bolt selection and installation and forming of the slotted holes.

3.2.1

Designers should be aware of the very great advantage that the SHJ and the FBJ offer over conventional rigid-ted MRSF systems. This advantage is the ability to de-couple seismic and gravity requirements in the frame and connection design. The approach used involves a variation on the procedure for design of multi-storey windresisting MRSFs in non-seismically active countries, such that: (i)

The beams are designed to resist the maximum applied gravity loads (dead, live loads) in a simply ed condition

(ii)

The t is sized to resist only the moment generated by the earthquake action, ie. Mcode, µdesign. This moment is calculated and applied independently of the beam’s section moment capacity.

(iii)

The brass shim material must be specified as UNS C2600 – ½ Hard Temper, eg. to AS 1566 [15]. It is very important that the ½ Hard Temper is included in the specification, as that defines the hardness, yield stress and tensile strength required and on which all the research has been based. 3.2.2

material

Limit on flange and web plate thickness as a function of bolt diameter

The same relationship as is used for the Flange Bolted ts should be used for the bottom flange plate and web plate. This is given by equation 68.1 and has been determined from the component testing;

As with all structural components designed to deliver dependable performance under severe seismic action, the detailing requirements and selection of appropriate materials are as important to the final behaviour as the design itself. HERA Steel Design & Construction Bulletin

t

Grade of steel for the flange, web plates and cap plates is to be 250, 300 or 350. It is important, when sizing the plates, that the use of grades 300 or 350 in order to reduce the plate thickness for a given width is clearly specified in the contract documents so that the grade used in design is supplied in practice. Designers can always opt for use of grade 250 material; this is also consistent with the approach used in R4-100 [12].

Details of the MRSF design are given in section 4. Coverage of the t design itself now commences, first with the all-important detailing provisions and material selection. These should be read in conjunction with Fig. 68.1 for general details and Fig. 68.2 for specific layout and notation. The notation used herein is consistent with that of the Structural Steelwork Connections Guide, HERA Report R4-100 [12]. and

the

The positioner bolt is a Property Class 4.6 black bolt to AS 1111.1 [14]. Only one positioner bolt per t is used and it has the same diameter as the bottom flange bolts. It must be supplied black finish, to make it visibly different from the HSFG bolts. Black finish is the default surface treatment for this property class of bolt.

The columns are designed to resist the overstrength action developed by the t, not that from the beam.

Detailing requirements selection

for

The bolts used, except for the positioner bolt, are Property Class 8.8 Structural Bolts (HSFG bolts) to AS/NZS 1252 [13]. For calculation of bolt shear capacity, threads are assumed to be in the shear plane. These bolts are to be supplied galvanized (this is the default surface treatment specified by [13]).

Thus the beam depth can be chosen for gravity strength and lateral stiffness control without impacting on the column design.

3.2

Material selection components

Page 8

No. 68, June/July 2002

ti,max = 0.9df

parallel to the line of principal applied force, shall be ≥ df. This dimension is a'ep in Fig. 68.2.

(68.1)

where: ti,max = maximum thickness of bottom flange, web plate df = diameter of bolt

The distance between the centreline of the last pair of sliding bottom flange bolts and the centreline of the positioner bolt is given by;

This translates to: • • • •

' Sp,bfb→pb = Max(2aep; 0.5Lsh + aep + aep)

16 mm for M20 bolts 20 mm for M24 bolts 25 mm for M30 bolts 32 mm for M36 bolts

(68.2) where: Lsh = length of slotted hole; see equation 68.6 in section 3.2.6.

For the top flange plate, which is sized on the basis of the actions generated by the sliding bolts (bottom flange and web bottom bolts), this limit can be relaxed slightly in the larger bolt diameters, up to: • • • •

3.2.4

Spf = Sgf = Sgw = 70 mm for M20 bolts Spf = Sgf = Sgw =

16 mm for M20 bolts 20 mm for M24 bolts 32 mm for M30 bolts 40 mm for M36 bolts

3.2.3

aep

90 mm for M24, M30 bolts

Spf = Sgw = 140 mm for M36 bolts Sgf = 140 mm (preferred) for M30, M36 bolts = 90 mm (alternative) for M30, M36 bolts, where the beam flange width is inadequate to accommodate the sum of 140 mm plus at least 4df.

Edge distances required

For the edge distances to all the nominal sized holes, these are ≥ 2df. This applies to the web top bolts, and the top flange bolts. The relevant distances are shown in Fig. 68.2, namely: aet

Pitches and gauges

Note that the minimum beam flange width required from (Sgf + 2aet,f,b) will not allow the SHJ to be used for beams with bf < 170 mm.

= edge distance transverse to the line of principal applied force = edge distance parallel to the line of principal applied force

Table 68.1 gives the relevant values for each dimension that have been used for th range of practical bolt diameters for the SHJ, along with the design sliding shear capacities, determined in accordance with equations 59.4 to 59.10 of DCB No. 59.

For the slotted holes, the minimum distance from the end of a slotted hole to an adjacent free edge,

Table 68.1 Bolt Sliding Shear Design Capacities and Detailing Properties BOLT SLIDING SHEAR DESIGN CAPACITIES AND OTHER PROPERTIES Bolt Designation

Plate Thickness

φVfss

kN

φVfss, bs

kN

φVfn

kN

df

mm

df '

mm

aep

mm

aet

mm

Sgw

mm

Sgf

mm

Sp

mm

mm

Plate thickness limit, bottom flange & web plates

M20 M20

12 16

42 38

51 47

93 93

20 20

22 22

50 50

50 50

70 70

70 70

70 70

16 16

M20 M24

20 12

36 65

44 78

93 133

20 24

22 26

50 50

50 50

70 90

70 90

70 90

16 20

M24 M24

16 20

60 56

73 69

133 133

24 24

26 26

50 50

50 50

90 90

90 90

90 90

20 20

M24 M30 M30

25 16 20

52 104 98

64 124 118

133 214 214

24 30 30

26 33 33

50 65 65

50 65 65

90 90 90

90 90 90

90 90 90

20 25 25

M30 M30

25 32

91 83

111 102

214 214

30 30

33 33

65 65

65 65

90 90

90 90

90 90

25 25

M36 M36

16 20

162 153

190 182

313 313

36 36

39 39

75 75

75 75

140 140

90 90

140 140

32 32

M36 M36

25 32

144 132

173 162

313 313

36 36

39 39

75 75

75 75

140 140

90 90

140 140

32 32

HERA Steel Design & Construction Bulletin

Page 9

No. 68, June/July 2002

3.2.5

(3)

Clearance between beam face and column flange

Length of bottom flange plate See Fig. 68.2 for these .

This is the dimension fSHJ shown in Fig. 68.2. ' Lbfb = fSHJ + Lsh(0.5nbfb – 0.5) + 0.5nbfp aep

The dimension is calculated on the basis that, when the sliding hinge t is subject to maximum design negative rotation, thus causing the beam bottom flange to move its closest in towards the column flange (see the right hand figure, Fig. 64.10 of DCB No. 64), there is still a clear length of flange plate of 2.5tbfp available. This gives the following requirements for fSHJ in mm; fSHJ

≥ 10 + 1.25 θp,desdb + 2.5 tbfp

+ Sp,bfb→pb + aep (68.5) where: Lsh = 2.5 θpdb + d f' (68.6) Sp,bfb→pb = as given by equation 68.2 θp = 30 x 10 -3 radians d'f = diameter of nominally sized bolthole to NZS 3404 Clause 14.3.5.2.1 (mm) db = depth of beam (mm)

(68.3)

where: 10 = gap to clear weld between column and bottom flange plate (mm) θp,des = 30 x 10 -3 radians db = depth of beam (mm) tbfp = thickness of bottom flange plate (mm)

3.2.7 (1)

The value from equation 68.3 should be rounded up to the nearest 5 mm.

(2)

Thickness of both brass shims = 3 mm

(3)

Length of upper brass shim Lubfbs = Lbfp - fSHJ

(4)

3.2.8

Width of bottom flange plate:

(1)

(2)

Width, bb

(3)

3.2.9 (1)

The initial estimate of thickness is determined from section 3.5 and confirmed from section 3.7. The limiting thickness as a function of bolt size from section 3.2.2 must also be met.

(68.10)

= Min (tbfp ; 20 mm)

(68.11)

Length Lb

Thickness of bottom flange plate; tbfp

= bbfp

Thickness, tb

Where possible, use a flat bar to minimise fabrication cost.

HERA Steel Design & Construction Bulletin

(68.9)

Dimensions of bottom flange cap plate

bb

(68.4.1) (68.4.2)

where: df,bfb = diameter of bottom flange bolts Sgf = bolt gauge (Fig. 68.2 and Table 68.1)

(2)

Length of lower brass shim

where: Lb = length of flange cap plate, from section 3.2.8

This depends on the number of bottom flange bolts, which are determined from sections 3.6 and 3.7. Once this is determined, the dimensions of the bottom flange plate are determined as follows:

≥ 4df,bfb + Sgf ≤ 1.05bfc

(68.8)

Llbfbs = Lb

Dimensions of the bottom flange plate

bbfp,min bbfp,max

(68.7)

where: bbfp = width of bottom flange plate

The 10 mm gap for the weld applies, irrespective of the type of weld used between the bottom flange plate and the face of the column.

(1)

For both brass shims (upper and lower): Width = bbfp + 40 mm

As specified in section 3.2.6 of DCB No. 58, the FBJ has a constant clearance gap, fFBJ, of 20 mm. In contrast, the SHJ has a variable clearance gap that, in practice, varies from 50 mm to 100 mm or more.

3.2.6

Dimensions of bottom flange plate brass shims

' = 2aep + 0.5(nbfp – 2) (LSH + aep ) (68.12)

Dimensions of web plate Depth This is the dimension dwp in Fig. 68.3. The web plate should be as deep as is practicable for the given depth of beam,

Page 10

No. 68, June/July 2002

(3)

leading to the following recommendations for hot – rolled beams. dwp, minimum = db – 2tfb – 58

(68.12.1)

dwp, maximum = db – 2tfb – 48

(68.12.2)

dwp, average = db – 2tfb – 53

(68.12.3)

This is 3 mm. (4)

(5)

Thickness

3.2.11 (1)

Length (2)

(3)

Dimensions of web cap plate Depth (68.16)

Thickness (68.16)

Length Lw = Lwp – fSHJ

3.2.12

Lwp ≥ Max [(fSHJ + 2aep + (nwtb – 1) Sg,w); ' ' ( aep + nwbb (Lsh + aep )] (68.13)

(68.18)

Dimensions of top flange plate

This depends on the number of top flange bolts, which are determined from section 3.11. Once this is determined, the dimensions of the top flange plate are determined as follows:

where: nwtb = number of web top bolts, from section 3.8 nwbb = number of web bottom bolts, from section 3.6

(1)

Dimensions of web brass shims Depth of web inner brass shim

Width of top flange plate btfp, min ≥ 4df,tfb + Sgf

(68.19.1)

btfp, max ≤ 1.05bbfc

(68.19.2)

where: df,tfb = diameter of top flange bolts

As shown in Fig. 68.1 and, to a lesser extent, in Fig. 68.2, the inner brass shim extends the full depth of the web plate, with a return at the top of 15 mm. This return is to allow the brass shim to hook over the web plate during erection, thus making it self-ing while the beam is being put into position.

Where possible, use a flat bar to minimise fabrication cost. (2)

Thickness of top flange plate This is determined from section 3.11; the limit of section 3.2.2 as a function of bolt size must also be met.

The inner web brass shim is therefore dwp clear depth with a 15 mm return to either the left or right as appropriate.

(3)

Length of top flange plate Ltfp = fSHJ + 2aep + (0.5ntfb – 1) Spf (68.20)

Depth of web outer brass shim

where: ntfb = number of top flange bolts, from section 3.12.

This is equal to the web cap plate depth.

HERA Steel Design & Construction Bulletin

(68.15)

tw = Min (twp ; 20mm)

Thus the length of the web plate is controlled by:

(2)

Length of outer web brass shim

dw = 2aet

As can be seen from Fig. 68.2, the spacing of the web top bolts and the web bottom bolts is controlled by different criteria. The web top bolts align with the top flange bolts, while the web bottom bolts align with the bottom flange bolts.

(1)

(68.14)

The additional length of the outer brass shim is to allow it to be held for positioning during erection, once the web cap plate is in place.

twp = tbfp is initially used and is increased only if required from section 3.9. This has not been required in any of the designs undertaken for the NITH studies.

3.2.10

= Lwp - fSHJ

Lowbs = Liwbs + 30 mm

The limits for dwp are given in mm.

(3)

Length of inner web brass shim Liwbs

where: db, tfb are the beam depth, flange thickness (mm).

(2)

Thickness of both brass shims

Page 11

No. 68, June/July 2002

3.2.13

Dimensions of shim

optional

decking

• • •

This is shown in Fig. 68.1. Its use facilitates laying of decking around the connection, reducing cost and enhancing constructability. It is formed from 3 mm thick steel plate. (1)

•

3.2.15

(68.21)

where: bbf = width of beam flange (mm) This allows 50 mm overlap each side of the wider of the beam flange or the top flange plate. (2)

Thickness = 3 mm

(3)

Length

• no Belleville Springs • Belleville Springs to the bottom flange bolts Fig. 68.1 shows the former option, while Fig. 68.2 shows the latter.

Ldss = Ltfp – 20 mm

(68.22)

If Belleville Springs are to be used, then they must be of sufficient number and strength to develop close to the bolt proof load, from NZS 3404 Table 15.2.5.1, when fully compressed.

The outer edge of the decking shim and top flange plate coincide; the inner edge extends past the face of the beam towards the column face, as shown in Fig. 68.2, with a gap of 20 mm adjacent to the column. 3.2.14

Preferred bolt groupings

Use of Belleville Springs

These are optional for the bottom flange bolts. They increase the bolt sliding shear capacity, as described in DCB No. 59 pages 29, 30, through reducing the loss of installed bolt tension due to the interaction of moment and axial force in the bolt shank when the t is sliding. This benefit is of principal importance for the bottom flange bolts and, throughout this project, the research has concentrated on the following options:

Width bdss ≥ Max (btfp ; bbf) + 100

nwtb ≥ nwbb df,bfb = df,wbb = df,wtb df,tfb = df,bfb is preferred df,positioner bolt = df,bfb

sizes

and

From the manufacturer’s load charts [16] for alloy/carbon steel springs, the following designation and number of springs are required to achieve this:

bolt

• • • •

For an initial guesstimation of bolt sizes, use M24 for beams up to 600 mm deep and M30 for beams above 600 mm deep.

When Belleville Springs are installed, they are to be placed under the nut end of the bolt, between the hardened washer and the face of the cap plate, as shown in Fig. 68.2.

The minimum sliding bolt group layout is: • 4 bottom flange bolts (2 rows of 2 bolts) • 3 web bottom bolts (3 rows)

When determining the nut rotation from the snugtight position to apply, for the given bolt length, from Table 15.2.5.2 of [5], an extra ½ turn must be added to allow for compression of the Belleville Springs. This extra ½ turn applies for all bolt diameters used (M20 to M36). A background to this will be given in [1].

This is the layout shown in Fig. 68.2. When increasing the number of sliding bolts to develop the design moment, do this as follows: • Add one row of bottom flange bolts to give 6 bottom flange bolts (3 rows) and 3 web bottom bolts (3 rows); then

3.2.16

• Add one row to each bolt group (ie. increase the sliding bolt numbers in groups of 3 at a time).

Allowance for manufacturing tolerances in the ed beam and inclusion of a decking shim

As described on pages 23, 24 of DCB No. 56 and in section 3.2.7 of DCB No. 58 for the FBJ, allowance must be made for manufacturing tolerances in the beams by offsetting the positions of the top and bottom flange plates.

This keeps sliding bolt group proportions in line with those experimentally tested. Other constraints on bolt sizes and groupings are:

HERA Steel Design & Construction Bulletin

For a M20 bolt, 2 No. 12-EH-168 springs For a M24 bolt, 3 No. 16-H-168 springs For a M30 bolt, 3 No. 20-H-225 springs For a M36 bolt, 3 No. 24-H-262 springs

Page 12

No. 68, June/July 2002

Finally, note mention of the decking shim in Fig. 68.1. This is made from 3 mm thick Grade 250 or 300 plate. It extends 50 mm beyond the top flange plate on whichever side(s) of the beam (s) steel decking and provides a to the decking during construction. It is also detailed in item 35 of HERA Report R4-58 [17]; see especially item 35c therein in this regard. It is an extra component to consider in fabrication and erection but one which greatly facilitates placing of the decking around the connection. Note also the 3 mm thick plate extensions welded onto the underside of the top tension/compression stiffeners in Fig. 68.1 and Fig. 68.5 for the same purpose.

In the case of the FBJ, the magnitude of the offset was not important to the operation of the t, thus the final recommendations were driven only by constructability considerations. However, in the case of the SHJ, it is important to minimise the offset between the bottom face of the beam and the bottom flange plate. In the first large-scale test specimen, this offset was 3 mm, whereas in the second test specimen, it was only 1 mm. The greater offset from the first test resulted in an appreciable loss of bolt tension and hence sliding shear capacity of the t. However the effect of the offset in test 1 was exacerbated by allowing for a minimum gap between the bottom corner of the beam and the column face of only 15 mm under maximum negative rotation. In the second test specimen, this minimum gap was increased to 40 mm using equation 68.3, thus reducing the pull-down effect on the bolts by a factor of 18. While this increase in clearance has a significant effect, it is also desirable to limit the maximum net extent of mismatch likely between the top surface of the bottom flange plate and the bottom surface of the beam to 2 mm. This results in the following recommendations for manufacturing tolerance allowances in the SHJ: (1)

(2)

3.2.17

The bolts are to be positioned in the directions shown in Fig. 68.1 and tightened from the nut end. This is particularly important to avoid clashes between the web and flange bolts during installation. The positioner bolt is used during erection to stabilise the bottom of the t and to prevent undue rotation.

The allowances are provided as an offset of each flange plate away from the specified centreline position of the beam (see Fig. 68.2)

Once the frame is aligned, the bolts should all be snug tightened, starting with the bottom flange bolts and working up. The tightening pattern should be to NZS 3404 Clause 15.2.4.1. For each group of bolts (eg. the bottom flange bolts) this means starting with the bolts closest to the column face and working along the row away from the column face. For the flange bolts, this may require two or more rounds of snug tightening to get all bolts snug tight, pulling the flange plate in hard against the flange upper brass shim.

The up offset for the top flange plate is as follows: • 3 mm for beam depths up to 610 mm • 4 mm for beam depths above 610 mm • 3 mm is added to all the above to accommodate a decking shim, where used.

(3)

The down offset for the bottom flange plate is as follows:

The bolts are then fully tensioned, starting again with the bottom flange bolts and working up. Tensioning is to the part turn method of NZS 3404 Clause 15.2.5.2. For bottom flange bolts where Belleville Springs are installed, tighten by an extra ½ turn from snug tight over that specified in Table 15.2.5.2 of [5].

• 2 mm for all beam depths • 3 mm is added to all the above to accommodate the flange upper brass shim, which is always required. In practice, these tolerance allowances will lead to a gap existing between the beam flange and top flange plate in most instances; this gap is readily closed by the bolt tightening, for which the moment developed is at least an order of magnitude greater than the weak axis plastic moment capacity of the plate.

3.2.18

Tightening of large diameter HSFG bolts

The SHJ connections will routinely require the use of fully tensioned M30 high strength structural bolts and occasionally the use of M36 bolts. It is important to ensure that, when this size is specified, they are fully tensioned.

The web plate must also be offset from the column flange centreline by an amount equal to half the beam web thickness plus 3.5 mm. 3 mm of this is to accommodate the web inner brass shim. HERA Steel Design & Construction Bulletin

Bolt tightening sequence and method of tightening

This task is beyond the scope of a standard impact wrench. Suitable equipment is readily

Page 13

No. 68, June/July 2002

available; details are given on page 24 of DCB issue No. 56. 3.2.19

(1) Calculate the design sliding shear capacity, ΣφVfss, of the bottom flange plate bolt group and the web bottom bolt group.

Forming of the slotted holes (2) Take moments of each of the sliding shear capacities from (1) about the top of steel beam. The lever arms are shown in Fig. 68.3. The sum of these moments = φMSHJ.

The slotted holes in the bottom flange plate and web plate (see Figs. 68.1 – 68.3 for their location) can be formed by machine flame cutting or water jet cutting to the required dimensions.

The design vertical shear (seismic plus gravity) is carried by the web top bolts; thus the design shear capacity, φVSHJ, of the SHJ = the design shear capacity of the web top bolt group. For most applications, only one row of web top bolts will be required to carry the design vertical shear, however, if this is large, two rows of web top bolts may be needed.

However, they can also be formed by drilling a nominally sized hole at each end of the slotted hole, then gas cutting across the top and bottom of this pair of drilled holes to form the slotted hole. This gas cut surface need be no smoother than that from good practice hand gas cutting, provided that the rounded ends of the slotted hole are of drilled surface smoothness. If this method is adopted, then the width of slotted hole, as measured between the adjacent gas cut surfaces,

In all the representative designs undertaken in this project, only one row has been needed.

must lie between d'f and (d'f + 2) mm. This has been the approach used in all the SHJ experimental tests undertaken.

The t is sized to develop the following design moment and shear capacities:

3.2.20

* φMSHJ ≥ Mdesign

(68.23)

φVSHJ ≥ VGQU + VEµdesign

(68.24)

φVSHJ ≥ VGQmax

(68.25)

Surface treatment of the ply surfaces

The sliding surfaces are between steel and brass. The steel surfaces must be clean and free of any surface coatings, loose scale, loose rust, visible grease or oil marks.

where: M*design

The brass surfaces must be clean and free of surface coatings, visible grease or oil marks. Because of these restrictions on surface condition of the sliding surfaces, the SHJ is principally intended for application in corrosion category C1 to ISO 9223 [18] (very low rate of corrosion, typically found inside heated or air conditioned buildings with clean atmospheres). The surfaces for the bottom flange bolts and web bottom bolts must be as specified above for SHJs in corrosion categories C2 to C5 of [18]. Non- surfaces can be protected with an appropriate surface treatment; the edges of the surfaces should be sealed against water ingress. The positioner bolt will need to be painted in these applications. 3.3

3.3.1

VGQU

= design shear force from load combination G + Qu (dead and live load for use in conjunction with earthquake).

VEµdesign

= design shear force derived from out-of-balance design seismic moments acting on the clear beam length.

VGQmax

= design shear force for full factored loading, eg. 1.2G + 1.6Q from [2].

3.3.2

Design role of t components

Refer to Fig. 68.1 in conjunction with this section. The design roles of the SHJ components are as follows:

Design concepts for the sliding hinge t Development of moment and shear capacity

The design moment capacity of the SHJ, φMSHJ, is determined as follows:

HERA Steel Design & Construction Bulletin

= design moment for the SHJ from the most critical of earthquake or wind; see section 3.4.

Page 14

•

The top flange bolts act as the anchor point for t rotation, pinning the beam top corner in place relative to the column

•

The web top bolts resist the applied vertical shear force. They are subject to only small movement in the longitudinal direction due to No. 68, June/July 2002

Step 3 : Determine initial bottom flange plate width and thickness and initial web plate thickness Step 4 : Determine bolt size and numbers for moment adequacy, then finalise bottom flange plate width and thickness

their proximity to the pinning action of the top flange bolts. •

The web bottom bolts and bottom flange bolts develop the sliding shear resistance

•

The cap plates provide the to the bolt end remote from the beam of the sliding bolt groups

•

The brass shims facilitate smooth sliding between the steel surfaces at a near constant level of shear friction, which is essential to the maintenance of stable and sufficient bolt tension when the t is sliding

•

•

Step 5 : Design web top bolts for vertical shear resistance Step 6 : Design web plate Step 7 : Design top flange bolts and plate Step 8 : Check on reduced tension capacity of the beam at the bolted connection

The Belleville Springs, which are optional additions to the bottom flange bolts, assist these bolts to retain bolt tension under sliding. This sustains the bolt sliding shear capacity, Vss, at a higher level than is the case without the springs and retains t stiffness in the post-sliding regime of behaviour.

Step 9 : Design column

It functions as a locater bolt for the sliding bolts, ensuring that they are located in the middle of the slotted holes in the erected t

(iii)

It provides a rapid visual indicator as to whether the t has gone into the sliding mode following a severe earthquake; if this happens and the t inelastic rotation exceeds around 10 mrad, the positioner bolt shears through and the lower half drops out.

3.3.3

plates

and

Step 11 : Design, detail positioner bolt and shims Step 12 : Design tension/compression stiffeners Step 13 : Calculate t overstrength capacity Step 14 : Design t zone The full SHJ design procedure, starting with determination of t design moment and design shear, is given in sections 3.4 to 3.21.

It acts as a stability bolt for erection purposes, making the t rigid for erection by developing moment resistance in conjunction with the top flange bolts

(ii)

between

Step 10 : Dimension flange and web plates

The positioner bolt is a black finish class 4.6 bolt that connects between the beam flange and bottom flange plate only, through nominal sized holes in each ply. It has the same diameter as the rest of the bolts (which are all galvanised finish property class 8.8 structural bolts). The positioner bolt has three very important roles, namely: (i)

welds

3.4 3.4.1

Calculation of the design moment and design shear Design earthquake moment

As has been mentioned in section 3.1, the t itself is sized to resist the code-derived earthquake moment alone, ignoring t moments induced by gravity only, with the beam designed to resist the full factored gravity load (ie. 1.2G + 1.6Q to NZS 4203 [2]) as a simply ed beam. For the SHJ, the design earthquake moment, ME* µdesign , is determined from [2 or 4, 5 and 6] for low-rise and medium-rise MRSFs. The t design earthquake moment is given in sections 4.2 and 4.3 herein.

Sequence of design actions

3.4.2

Design shear force

The full SHJ design procedure involves the following 14 steps:

This is given by the largest of equations 68.24 and 68.25.

Step 1 : Determine design moments and shears

The seismic component of shear, VE*µdesign , is given by:

Step 2 : Determine sliding bolt group layouts

VE*µdesign = HERA Steel Design & Construction Bulletin

Page 15

3 ME* µdesign (Lb - d c )

(68.26)

No. 68, June/July 2002

0.7

= strength reduction factor for tension friction action 0.8 = strength reduction factor for bolt sliding shear capacity determination 0.85 = kh for short slotted holes, from NZS 3404 Clause 9.3.3.1.

where: 3 ≈ 1.4 x 1.1 x 2 1.4 = overstrength factor on t 1.1 = allowance for φMSHJ / M* 2 = moment pattern factor (equal and opposite end moments) (Lb – dc) = clear length of beam 3.4.3

In practice, the length of slotted hole will typically be such as to classify it as a long slotted hole. However the cap plate provides much more robust confinement than an oversized washer, thus the value of kh for short slotted holes rather than for long slotted holes is used.

Design wind moment

The SHJ has been developed as a semi-rigid t for seismic-resisting systems. However, it must also perform satisfactorally under wind loading.

Step 3: Check if equation 68.28 is satisfied

In designs for New Zealand application, in accordance with NZS 4203 [2] or its proposed replacement [4, 19], it is possible that ultimate limit state wind design may govern some ts in buildings over around 10 storeys high. This will be especially the case for designs to the draft Loadings Standards [4] which are located in the lowest seismic regions.

φMSHJ,WSLS

In practice, for designs to either the current Standard [2] or the new draft [4 and 19], it is likely that, where wind action governs, it will be the ULS rather than SLS that is critical. This is because the ratio of (M*WSLS / M*WULS ) will typically be less than 0.75.

Wind ultimate limit state

Having determined the design moment and shear, the t design proceeds as follows. For this procedure, the ULS design moment is designated M*design , which covers the critical ULS moment being from either earthquake or from wind, as appropriate.

The t design for the wind ultimate limit state moment, M*WULS , uses the principles and procedures as given in sections 3.5 to 3.16. In saying this, it is conservative to apply the relevant overstrength factors as ductility demand is not anticipated under M*WULS .

3.5 It follows, in checking for the wind ultimate limit state, that if M*Eì > M*WULS , then the earthquake condition governs design for the ultimate limit state.

3.5.1

3.5.2

The SHJ must remain rigid at the wind serviceability limit state. This is easily checked as follows:

Step 2: Determine the moment associated with rigid action of the t from equation 68.27. (68.27)

First estimate of bottom flange plate thickness

N t,*design =

where: 0.75 = (0.7/0.8) x 0.85 HERA Steel Design & Construction Bulletin

Bottom flange plate width

The bolt sliding shear capacity is a function of the plate thickness, hence the t moment capacity is also a function of the plate thickness. This means it is desirable to obtain a rapid estimate of bottom flange plate thickness as soon as the t design moment is known. This is determined from the following two equations.

Step 1: Calculate the design wind serviceability limit state moment, M*WSLS .

= 0.75φMSHJ,Eµ

Determine bottom flange plate width and initial thickness

See section 3.2.6 (1) for the limits. Select a plate width, bbfp, within these limits.

Wind serviceability limit state

φMSHJ,WSLS

(68.28)

If it isn’t, then add an extra set of sliding bolts in accordance with section 3.2.14 and recheck. This will affect the overall t design and overstrength action and require reconsideration of the t and system design for earthquake.

For this reason, brief guidance on SHJ design for each wind limit state is given below.

3.4.3.2

* MWSLS

If it is, the t design is satisfactory.

Also, because the levels of wind loading associated with the serviceability and ultimate limit states are closer (see eg. Table 5.4.2 of [2]) than for earthquake, it is possible that either wind limit state may govern some SHJs in buildings as low as 10 storeys high.

3.4.3.1

≥

Page 16

* 1.2 Mdesign

db

(68.29)

No. 68, June/July 2002

tbfp ≥

N t,*design 0.9 (bbfp − 2df' ) fy,bfp

φVfss,bs (68.30)

where: 0.9 = strength reduction factor d'f = bolt hole diameter for nominal sized hole, from NZS 3404 Clause 14.3.5.2.1. fy,bfp = bottom flange plate yield stress 3.5.3

= design sliding shear capacity, with BS = get from Table 68.1 for bolt size and plate thickness

Fig. 68.3 shows the lever arms for the moment capacity determination. The value of 26.5 used in equation 68.33 comes from the average web plate depth, from equation 68.12.3. 3.6.3

Check plate thickness limit in relation to bolt size

Check moment adequacy

This is given by:

Check that the limit of section 3.2.2 is satisfied; if it isn’t, then a larger bolt diameter is needed for the given plate thickness.

φMSHJ ≥

3.5.4

where: M*design is from section 3.4; typically section 3.4.1.

3.6

Apply this estimate of thickness to the web plate

3.6.4

Determine sliding bolt size and numbers for moment adequacy

3.6.1

• Bolt size, numbers and layout from section 3.2.14.

(1)

• Either increase the bolt numbers in accordance with section 3.2.14 and recalculate; or • Increase the bolt size and recalculate; or • Increase the bolt numbers and bolt size and recalculate.

Calculate moment capacity of t ts with no Belleville Springs

(68.31)

3.7

ts with Belleville Springs in bottom flange

where: nbfb = no. of bottom flange bolts = 4 for initial trial, from section 3.2.14 nwbb = no. of web bottom bolts = 3 for initial trial, from section 3.2.14 φVfss = design sliding shear capacity, no BS = get from Table 68.1 for bolt size and plate thickness • Bolt size for initial trial from section 3.2.14 • Plate thickness from section 3.5.2 = db – tfb – 26.5 – aet (mm) (68.33)

HERA Steel Design & Construction Bulletin

Design of bottom flange plate

There are four cases to consider, three of which require calculation and the fourth of which is dealt with by detailing. These are:

φMSHJ = nbfb φV fss,bs db + nwbb φV fss ewb (68.32)

ewb

Review bolt numbers and size

If equation 68.34 is not satisfied;

φMSHJ = nbfb φV fss db + nwbb φV fss ewb (2)

(68.34)

If equation 68.34 is easily satisfied, reduce bolt size to M20 and recalculate; this gives φMSHJ, minimum for the given beam size.

Start with the following

3.6.2

* Mdesign

Page 17

(i)

Suppression of net tension yield prior to the bolts developing their sliding shear capacity; see section 3.7.1

(ii)

Suppression of net tension fracture while t is in active sliding mode; see section 3.7.2

(iii)

Suppression of compression yielding while t is in active sliding mode; see section 3.7.3

(iv)

Suppression of premature bolt shear fracture when end of slotted hole is reached; this is covered by compliance with the bottom flange plate thickness to bolt diameter ratio given by equation 68.1.

3.7.1

Net tension yield

* N ty, bfp

= 1.15 nbfb φVfss,bfp

(68.35)

No. 68, June/July 2002

φNty,bfp = 0.9 (bbfp - 2df' ) f y,bfp tbfp

φNcu,bfp = 0.85bbfp tbfp fy,bfp

(68.36)

(68.43)

where: nbfp = no. of bottom flange bolts from section 3.6.4 1.15 = 0.9 / 0.8 = difference in φ between bolt and plate (b,t,fy)bfp = from section 3.6 d'f = function of df,bfb from NZS 3404 Clause 14.3.5.2.1 (see also Table 68.1) φVfss,bfp = φVfss or φVfss,bs as appropriate, from Table 68.1.

where: 0.85 = 0.9 x 0.942 0.942 = α v from Table 6.3.3 of NZS 3404 for α n = 25 and α b = 0.5

φNty,bfp

* ≥ N ty, bfp

Use the resulting tbfp for the web plate.

3.7.2

Check for net tension failure

is required

φNcu,bfp ≥ Nu,* bfp is required where: * Nu, bfp is given by equation 68.38

(68.37)

3.8

φVfss,bfp φ

Csp 0.9

3.8.1

where: VE*µdesign

(68.38)

* VGQu

≥ Nu,* bfp

3.7.3

Compression capacity

is required

(68.39) nwtb ≥ where: φVfn,wtb

(68.40)

 Le,bfp  λn,bfp =    0.29 tbfp 

 fy,bfp     250 

(68.41)

(68.42)

3.9

=

as given by equation 68.26

=

as given by the tributary area vertical loading for the appropriate factored maximum (dead + live) loads

Determine the number of web top bolts required * Vwv φVfn,wtb

(68.46)

= design capacity, threads included, same bolt diameter as for web bottom bolts. (See eg. [20] for this information).

Design of web plate

The web plate thickness, twp, has been set equal to be bottom flange plate thickness, from step 3.7.3. The web plate’s capability to resist the vertical shear and horizontal tension actions now needs to be determined.

Check if λ n,bfp ≤ 25. If it is, proceed to the next equation. If it isn’t, then α v for input into equation 68.43 needs to be re-evaluated from Table 6.3.3 of NZS 3404 [5] for the value of λ n,bfp from equation 68.42. HERA Steel Design & Construction Bulletin

(68.45)

If nwtb < nwbb, where nwbb has been determined from section 3.6, then add additional web top bolts such that nwtb = nwbb. The additional web top bolts are then used to resist the forces developed by the sliding groups of bolts, in conjunction with the flange top bolts, in section 3.11.

First the slenderness ratio of the bottom flange plate must be checked Le,bfp = 0.7 (fSHJ + 1.25θpdb)

* , VGQmax

3.8.2

where: fu,bfp = ultimate tensile strength of bottom flange plate φNtu,bfp

Vertical design shear force

* * * Vwv = Max(VE*µdesign + VGQu ; VGQmax )

where: nbfb = no of bottom flange bolts, from section 3.6 Csp = 1.45 when no springs are used = 1.55 when Belleville Springs are used 0.9 = ideal capacity factor φNtu,bfp = 0.77 (bbfp - 2df' ) fu,bfp tbfp

Design of web top bolts

These are designed to resist the applied vertical shear, in bearing, with threads included in the shear plane.

This is determined from the design action developed under the design level of rotation. The ideal capacity of the plate is used to resist this action, therefore the ideal capacity factor is incorporated into the design action determination, thus: Nu,* bfp = nbfb

(68.44)

Page 18

No. 68, June/July 2002

Vertical shear will be resisted over the full depth of plate less the width involved in resisting horizontal actions from the web bottom bolts. Horizontal tension/compression is developed by the sliding resistance of the web bottom bolts. This is resisted by the strip of web plate under the web cap plate for commencement of yield and by 1.5 x bw for tension fracture under overstrength action. 3.9.1

Calculate design capacity of plate

vertical

φNtu,wp ≥ 0.77(1.5d w - d f' )t wp fu,wp

(68.54)

* φNtu,wp ≥ N tu, wp is required

(68.55)

3.9.5

This can now be done; see section 3.2.9. 3.10 Sizing of cap plates and brass shims

shear

3.10.1

φVvn,wp = 0.9 x 0.6 x 0.6 x 0.83(d wp - d w) fy,wp t wp α v = 0.27(d wp - d w) f y,wp t wp α v

αv

(d wp - d w) t wp

≤ 82

fy,wp 250

Bottom flange cap plate

See section 3.2.8 for determining the width, thickness and length of bottom flange cap plate, using the values determined above.

(68.47)

3.10.2

(68.48)

See section 3.2.7.

where: α v = 1.0 if

Sizing of web plate

3.10.3

= 1.0 otherwise; see Clause 5.11.5.1 of NZS 3404.

Bottom flange upper and lower brass shims

Web cap plate

See section 3.2.11. 3.10.4

Web inner and outer brass shims

The second 0.6 is to for moment / shear interaction.

See section 3.2.10.

dw

3.11 Design of top flange bolts and plate

3.9.2

= as given by section 3.2.11(1) Check vertical shear adequacy of plate

* φVvn,wp ≥ Vwv

The top flange plate anchors the beam laterally and operates as a hinge about which the beam can slide. It is designed to resist the combined shear developed by the web bottom bolts and bottom flange bolts, using bolts of the same diameter. The shear from these is the greater of the overstrength sliding shear or the threads excluded design shear capacity. The latter will always govern and is therefore the only check needed. Because of this it is simply a matter of matching bolt numbers, incorporating any web top bolt unused capacity to resist the lateral force.

(68.49)

where: * Vwv = design vertical shear force from equation 68.45. 3.9.3

Check for net tension yield

This is checked under the design sliding shear, for the width of web plate under the cap plate only. * N ty, wp = 1.15nwbb φVfss

(68.50)

φNty,wp = 0.9(d w - d f' )t wp fy,wp

(68.51)

φNty,wp ≥

(68.52)

3.9.4

* N ty, wp is

required

3.11.1

Number of bolts required

Using the same bolt diameter as for the web bottom bolts and the bottom flange bolts. nftb required = where: kr

Check for net tension failure

This is checked for the overstrength sliding action associated with reaching the end of the slotted hole, with this action being resisted by a depth of web plate = 1.5 x depth of cap plate.

nwtb,calc

(

(

))

1 nwbb + nbfb - nwtb,used - nwtb,calc kr (68.56)

= reduction factor for bolts in a line from NZS 3404 Table 9.3.2.1. = no. of web top bolts required from equation 68.46, section 3.8.2

nwtb,used * N tu, wp = nwbb

φVfss,wb φ

C sp 0.9

HERA Steel Design & Construction Bulletin

= no. of web top bolts used from section 3.8.2 If the length of the t, as measured from the first to the last bolt, exceeds 15df, then kr < 1.0.

(68.53)

Page 19

No. 68, June/July 2002

3.11.2

Determine the top flange plate width required

φNt,tfp and φNc,tfp ≤ 0.85ntfp φVfn,tfb

(68.62)

See section 3.2.12(1) for the limits on btfp. Select a plate width within these limits.

If equation 68.62 is not satisfied, add an extra pair of top flange bolts and recheck.

3.11.3

3.12 Check on the reduced tension capacity of the beam at the bolted connection region

Determine required thickness suppress tension yielding

to

This is sized so that the plate can develop the sliding shear capacity of the bottom flange and web bottom bolts, without tension yielding. t tfp,tension ≥

1.15 (nbfb φVfs + nwbb φVfs ) 0.9(btfp - 2d f' )fy,tfp

The purpose of this check is to suppress yielding of the beam cross-section through the loaded end of the beam under moment-induced tension during the sliding phase of the t. Such yielding would cause unwanted loss of bolt tension and hence sliding shear moment capacity.

(68.57)

3.12.1

where: φVfs = φVfss or φVfss,bs as required fy,tfp = yield stress of top flange plate d'f = diameter of bolt hole to NZS 3404 Clause 14.3.5.2.1

* N tb = 0.5 x

Where possible, use a flat bar to minimise fabrication cost. 3.11.4

Undertake a slenderness ratio check on the top flange plate, if no concrete slab is present

If there isn’t a concrete slab, then: = 0.7 (fSHJ + aep,tf,b)

(68.58)

 fy,bfp     250 

(68.59)

 Le,tfp  λn,tfp =    0.29 t tfp 

Check if λ n,tfp ≤ 25, when no concrete slab is present. If it isn’t and no slab is present, then α v for input into equation 68.61 needs to be reevaluated from Table 6.3.3 of NZS 3404 [5].

φNc,tfp = 0.85btfp t tfp f y,tfp

(68.61)

φNtb = 0.45Agfyfb

(68.64.2)

Check that the following is satisfied (68.65)

If this equation is not satisfied, use a larger beam size so that it is satisfied. Do not use beam reinforcing plates with the SHJ.  φMSHJ   ≤ 0.76, the beam end In practice, if    φMsx,b  capacity is likely to be adequate. For preliminary design, one can use (M*design / φMsx,b ) ≤ 0.76/1.15 ≈ 0.66 as a target value for beam selection.

Check HERA Steel Design & Construction Bulletin

(68.64.1)

* φN tb > N tb

Calculate (68.60)

φNtb = 0.39Anb fub

3.12.3

Check top flange plate and bolt adequacy for the ULS condition

φNt,tfp = 0.77(btfp - 2d f' )t tfp fu,tfp

(68.63)

where: Anb = net area of the beam cross section, calculated in accordance with Clause 9.1.10 of [5] fub = tensile strength of the beam fyfb = yield stress of the beam flange

where: fSHJ is determined from equation 68.3 aep is given in Table 68.1 for the given bolt size.

3.11.5

1.15φMSHJ Nt φMsx,b

where: φMSHJ = the t design moment capacity from section 3.6 φMsx,b = the design section moment capacity for the beam size chosen; eg. from [20] Nt = the nominal section gross yielding capacity, determined from NZS 3404 Equation 7.2.1 3.12.2 Calculate the design tension capacity of the beam from the lesser of

If there is a concrete slab in with the top surface of this plate, which will be the typical case, no slenderness check is needed.

Le,tfp

Calculate the design tension action, N*tb , on the tension half of the beam, from equation 68.63

Page 20

No. 68, June/July 2002

3.13 Welds required between column flange and bottom flange plate

where: * vtw = φNt,tfp from equation 68.60 bmin = lesser of (btfp ; bfc) btfp = width of top flange plate

The bottom flange plate has been sized to dependably resist the maximum force expected under the maximum design rotation, in accordance with section 3.7.2. This carries the t-overstrength factor, which means that the weld need only be designed to develop the design tension capacity of the flange plate, not the overstrength tension capacity. 3.13.1

* vw, bfp =

3.14.2

Determine fillet weld size required as for the bottom flange plate, see section 3.13.2. If tw > 15 mm use a BW. 3.15 Welds required between column flange and web plate

Design action on bottom flange plate weld * Ntw, bfp

2bmin

= φNtu,bfp from equation 68.39 lesser of (bbfp; bfc) width of bottom flange plate width of column flange

3.13.2

Select fillet weld size such that:

* φv w ≥ v w, bfp

These welds are subject to two very different sets of conditions. The first is combined moment and vertical shear generated by the web top bolts and resisted by the clear depth of web plate for shear and the full depth for moment. The second is moment-induced axial tension generated by the web bottom bolts at the end of their sliding regime, taken over a thickness of web plate equal to 1.5 x the thickness of the web cap plate. The two cases are considered separately and the design action is the maximum from the two cases, but not required to be greater than the design tension capacity of the plate. All this involves:

(68.66)

where: N*tw,bfp bmin = bbfp = bfc =

Design of welds

3.15.1

(68.67)

where: φvw = design capacity of category SP fillet weld from [5]

Calculate actions vertical shear

* vwv, wp,v =

Values of φvw are listed in [20]

* vwv, wp,h =

This is the fillet weld size required on each side of the flange plate to column flange. 3.13.3

* vwv, wp

From consideration of welding economics and clearance requirements, determine if the fillet weld size from 3.13.2 will be used or if a complete penetration butt weld is required.

* Vwv

dwp dw ey

3.14 Welds required between column flange and top flange plate

HERA Steel Design & Construction Bulletin

(68.70)

2 d wp

) + (v )  2

2 * wv,h

0.5

(68.71)

= design vertical shear force, from equation 68.45 = average depth of web plate, from equation 68.12.3 = depth of web cap plate, from section 3.10.3 = fSHJ + aep,wt,b + ((nwtb – 1)/2)Sg,wt (68.72)

Design action on top flange plate weld

2bmin

from

(68.69)

* 3Vwv ey

(

3.15.2

* Ntw, tfp

* Vwv 2(dwp - d w)

=  v *wv,v 

A similar situation applies to that for the bottom flange plate, namely:

* vtw, tfp =

weld

where

If tw > 12 - 15 mm, use a complete penetration butt weld (BW). For most fabricators engaged in multi-storey construction, the changeover point to a BW will be tw > 15 mm.

3.14.1

on

* vwh, wp =

Calculate actions on weld from axial tension generated by web bottom bolts φN tu,wp 3d w

(68.73)

where: φNtu,wp = capacity given by equation 68.54

(68.68)

Page 21

No. 68, June/July 2002

3.15.3

• Replace all related to the beam flange with the same term for the bottom flange plate, ie: Abfp replaces Afb; fy,bfp replaces f yb; tbfp replaces tfb; bbfp replaces bfb; twf relates to the weld calculated from 3.13.3 above.

Calculate design actions on weld between column flange and web plate

 0.9t wp fy, wp  * * *  vw, wp = Min Max v wv,wp ; v wh,wp ; 2 x 103   (68.74)

(

3.15.4

)

• More simply, use equation 50.2 from section 3.2(2) of DCB Issue No. 50 with the same substitutions as stated above.

Design weld

Select fillet weld size such that * φvw,wp ≥ v w, wp

(3) (68.75)

where: φvw = design capacity of a category SP fillet weld, eg. from [20]

Design and detail the tension/compression stiffeners to section 3.2 of DCB issue No. 50 (with the above modification to section 3.2(2) of that issue)

3.18 t overstrength moment, MoSHJ

The size is used on each side full length of the web plate to column flange.

This is determined as follows:

If tw,required ≥ 15 mm use a BW.

 φMSHJ  o MSHJ = φoms   φ 

3.16 Selection and location of the positioner bolt

(68.76)

where: φMSHJ = t design moment from section 3.6 φ = 0.8 φoms = 1.4 for the SHJ with or without Belleville Springs

The role of the positioner bolt is described in section 3.3.2; its grade and appearance in section 3.2.1. This bolt connects between the bottom flange plate and beam flange only. It is placed as shown in Fig. 68.2; the distance from the centreline of this bolt to the centreline of the adjacent row of bottom flange bolts is given by equation 68.2.

This overstrength factor has been derived from the experimental testing, using the methodology as will be described in [1].

This bolt is intended to be snug tightened only, but can be fully tensioned to hold the bottom of the t in place during erection, if desired.

3.19.1

3.19 t zone requirements

The zone design moment for input into NZS 3404 Equation 12.9.5.2(1) is the t overstrength moment given by equation 68.76. However, compared to the layout of a rigid welded t, the top and bottom flange plates are more widely spaced apart (see Fig. 68.2) which reduces the unbalanced shear force on the connection.

3.17 Tension/compression stiffener requirements These are determined using NZS 3404 Clause 12.9.5.3.1, modified as described below, in conjunction with section 3.2, page 13, DCB Issue No. 50. (1)

(2)

These two aspects are incorporated into equation 68.77, which gives the design shear force on the zone of a SHJ.

Provide tension/compression stiffeners positioned opposite the flange plates, so that top of steel is the same for each element.

o  MSHJ VP,* SHJ =   db + tbfp 

(

Use NZS 3404 Equations 12.9.5.3(3) and 12.9.5.3(4) to determine the area of stiffener required for each design action, with the following modification:

)

o    +  MSHJ   db + tbfp L 

(

)

  - VCOL  R (68.77)

where: The subscripts L and R refer to the left and right hand beams at the connection. MoSHJ = as given by equation 68.76.

• The tension/compression stiffener design is based on the bottom flange plate dimensions for both the top and the bottom pair of stiffeners. This may mean that the top pair of stiffeners are slightly thinner than the top flange plate.

HERA Steel Design & Construction Bulletin

Design shear force on zone

For preliminary design and for most final designs, VCOL can be ed for as described in NZS 3404 Commentary Equation C12.9.5(1).

Page 22

No. 68, June/July 2002

3.19.2

The design shear capacity of the zone, φVc , is calculated to NZS 3404 Eq 12.9.5.3.(5).

3.19.3

The zone capacity when

has

φVc ≥ Vpz∗

The effects of the slight foundation flexibility should be ed for; in lieu of a more detailed analysis, use the rotational stiffness given by NZS 3404 Clause 4.8.3.4.1(b).

adequate Design and detailing concepts for momentresisting column baseplate connections are given on pages 11-20 of DCB Issue No. 56.

(68.78)

The advice in both articles is written to utilise, as much as possible, the standard details and provisions in HERA Report R4-100 [12].

Doubler plates, if needed, should be designed in accordance with sections 4 and 6 of DCB Issue No. 57, pages 23-25 therein, which, although written for FBJ connections, actually covers both FBJ and SHJ connections.

3.20.3

Design actions and detailing requirements are given in section 4.2.2, page 23 of DCB Issue No. 50. The advice therein is also written for use in conjunction with [12]. Note that, for analysis, a “pinned” connection should be assigned a realistic rotational stiffness. This can be obtained from Clause 4.8.3.4.1 (a) of [5].

With the SHJ, doubler plates are not typically going to be necessary when only one beam frames into the column, but will often be required when two beams frame into the column. 3.20 Connections at column bases 3.20.1

Options available and impact on building performance

3.20.4

Ring spring bases

Fig. 15 of [21] shows a ring spring test setup which would also be applicable to a column base application.

The most commonly used column base connection type for a MRSF is a fixed base connection. This has the advantage of reducing lateral deflection in the superstructure. As mentioned in section 2.3, with fixed base columns the inelastic demand on the ts under the design severe seismic event is within the performance criteria specified for the columns in sections (2) and (3) therein. With pinned base columns, these limits are slightly exceeded in some types of earthquake record, principally those exhibiting positive near fault directional motion.

The ring spring t is well suited to application at the column base of a MRSF with SHJs or FBJs. This is because it combines the benefit of the pinned base, in protecting the column from inelastic action at its base, with the ability to generate a rapid increase in moment capacity with increasing rotation demand. The t also has good self-centering capability, which will assist in returning the building to its pre-earthquake position at the end of the strong ground motion shaking.

A third option is a ring-spring type detail at the column base. This is mentioned in [21], with a picture of such a t shown as Fig. 15 of [21].

Design of the ring spring t for this system is relatively straightforward. It is referenced from section 6.2 of [21] and will be described in [1]; further details are not given herein. the HERA Structural Engineer for more information.

When subjected to a design level severe seismic event, it is anticipated that minor damage to the yielding regions of columns adjacent to the column bases would occur in columns with fixed base connections. For columns with pinned base connections, minor damage would be expected within the baseplate detail. In each case, minimal or no repair would be anticipated to be necessary from this level of event.

3.21 Guidance on practical aspects of sliding hinge t design • The flange plates should be made as wide as possible, within the limits of sections 3.2.6 (1) and 3.2.12(1)

The ring-spring base would be dependably undamaged by this level of event.

• The top flange plate will typically be the same thickness or the next thickness up from the bottom flange plate and the web plate

Brief guidance on each type of column base connection is now given.

• The maximum number of bottom flange bolts should be 8, in order to keep the required bottom flange plate thickness within the maximum thickness allowed for the given bolt diameter. If the design from section 3.6 indicates that nbfb = 10 is required, look at

13.20.2 Fixed bases The design actions for fixed bases are given in section 4.2.1, pages 22,23 of DCB Issue No. 50 and are directly applicable to these semi-rigid systems. The actions are based on µdesign = 4. HERA Steel Design & Construction Bulletin

Pinned bases

Page 23

No. 68, June/July 2002

increasing the bolt size. With nbfp = 8, nwbb = nwtb = 4 will result, with ntfb = 12, typically.

the number of bays (more than one scheme may be required).

• Refer to Table 68.2, section 4.2, step 7 for typical values of φMSHJ/φMsx,b that have resulted from the many representative frames designed as part of this project.

The SHJ has been developed for perimeter frame application and the guidance given in step 2.2 and step 6 herein for the member sizes to meet frame stiffness requirements is formulated on that basis. (Perimeter frames and internal frames are as defined in NZS 3404[5]).

4.

Design of Moment-Resisting Steel Frames Incorporating Sliding Hinge t Connections

4.1

Step 2

This estimate should be made at the first level above the seismic base level, at the level of uppermost principal seismic mass level and at selected intermediate levels.

General and scope of guidance given

Section 4 presents guidance on the design of the MRSF system that incorporates the SHJ. This guidance is very similar to that for MRSFs with FBJs and follows the same format as that given in DCB No. 58 for the FBJ systems.

Guidance on the number of intermediate levels to consider is given on page 5.3 of [6]. For buildings up to 4 storeys in height, do the check at every level. For buildings up to 8 storeys in height, check levels 1, 3, 5 and 8. For buildings between 8 and 12 storeys in height, check levels 1, 3, 5, 8, 11 and 12. For buildings above 12 storeys in height, check 1, 3, 6 then every 4th level. However, the SHJ is probably not the most costeffective system to use on buildings above this height, because of the limited ductility demand required compared with what the SHJ can deliver. For such high-rise buildings, the FBJ offers a potentially more cost-effective solution, especially in low to medium seismic zones.

Section 4.2 covers preliminary design, while section 4.3 covers final design. The design procedures presented herein are based, in format and content (wherever possible), around the procedures incorporating capacity design presented in sections 5 and 6 of HERA Report R4-76 [6] for preliminary and final design, respectively, of category 1 or 2 MRSFs with rigid beam to column connections. For such systems, strength and stiffness cannot be de-coupled, so the columns must be designed to resist the beam section overstrength actions (or the upper limit seismic actions Emax).

Step 2.1 To carry gravity loads Use the approach given in step 2.1, section 5.2 of [6], except use the denominator value of 8 in equation 5.1 of [6] instead of 10. This corresponds to a simply ed condition, which is required for design in accordance with section 3.1(i) herein.

In contrast, for the semi-rigid systems incorporating SHJs, strength and stiffness are considered separately and the columns are designed to develop only the overstrength moment from the t. This requires some modifications to the R4-76 [7] procedures, but is a considerable simplification from the designer’s view point.

Use the lightest category 3 section from NZS 3404 [6] within a particular designation to resist the design moment, such that M*≤ φMs. (In the 1992 edition of NZS 3404, this category was designated 3A, which is still used in HERA report R4-76 [6]. This point is picked up in the summary notes Tips on Seismic Design of Steel Structures which are included in all post-July 2000 copies of [6]).

Given that this guidance is being written at the time of transition from NZS 4203:1992 [2] to the new Loadings Standard [4], wherever practicable the requirements of both documents are referenced. 4.2

Procedure for MRSF preliminary design

Step 2.2 To provide suitable frame lateral stiffness

The preliminary design procedure presented below is based around that given in R4-76 section 5.2 for preliminary design of category 1 and 2 MRSFs. It is presented in the same step by step format as section 5.2 of [7] and with the same headings. Step 1

Establish preliminary frame layouts

Formulate the preliminary frame layout or layouts in of the beam and column spacings and

HERA Steel Design & Construction Bulletin

Estimate beam sizes required

(1)

For perimeter frame MRSFs, select beam depths from the target span to depth ratios given by equations 68.79.1 to 68.79.3

(1.1)

For the lower half of the structure (up to 0.5H) (L/d*) = (9 or 11)10 > (11 or 13)5 (68.79.1)

Page 24

No. 68, June/July 2002

(1.2)

Step 4

For the three-quarter height (0.75H) of the structure

Determine the design bending moments at the ts (ie. M∗Eì and M∗Emax ) for load cases Eµ and Emax

(L/d*) = (11)10 , (13 or 15)5 (68.79.2) (1.3)

(1)

For the uppermost level (1.0H) of the structure (L/d*) = (15)10 , (14 or 15)5

(68.79.3)

In equations 68.79;

Use the procedure given in step 3.2, section 5.2 of [6], except that, instead of calculating M∗Eì as given on page 5.7 of [6], use equation 68.81 below.

ME∗ =

• The figures in ( )10 are the target L/d* ratios for a 10 storey building; those in ( )5 are for a 5 storey building. Interpolate for building levels between these.

L = span of beam (centre to centre) d* = target depth of beam

∑ M t, i

= sum of the out-of-balance moments at each t on level i

nbeam, i

= number of beams in the semi-rigid system on level i

(2)

In applying step 3.2, section 5.2 of [6], assume that the columns above and below that level are all the same section.

(3)

Calculate both M∗Eì and M∗Emax .

Step 5 At each level, select the lightest weight category 3 section, from NZS 3404 [6], for which db ≥ d* (db = depth of beam).

from section 3.4.3.1 and substitute it for M∗Eì if required.

Calculate the design seismic loads

Step 6

Use the procedure given in step 3.1, section 5.2 of HERA Report R4-76 [7], except that the estimate of fundamental period should be given by : (i)

(ii)

Step 7

(68.80.1)

Select the lightest designation I-section cross section that complies with the following: (68.80.2) (1)

There are two levels of seismic load to determine, namely: (2) (1)

(2)

That associated with determining the design t moment. This is load case Eµ (see section 3.4.1 herein), determined for µdesign = 4.

Flange slenderness complies with NZS 3404 Table 12.5 [5] for a category 2 member Web slenderness complies with Equation 8.4.3.3(2) of [5] for ( N∗g / Ns ) ≤ 0.3. This gives  d1  44   ≤ t fy /250  w  storey 1 where: d1,tw,fy are as defined in NZS3404

That associated with determining the upper limit design seismic actions on the secondary . This is load case Emax and is determined for µmax = 1.25.

HERA Steel Design & Construction Bulletin

Estimate the column sizes required

Step 7.1 For the first seismic storey

For a perimeter MRSF in which dcol < 0.8dbeam T1 = 0.15h0.75 n

Reassess the beam size in order that beam web reinforcing plates are not required.

This involves checking that M∗design / φMsx,b ≤ 0.66, as described in section 3.12.3.

For a perimeter MRSF in which dcol ≥ 0.8dbeam T1 = 0.12h0.75 n

Check magnitude of wind moment and use if this exceeds the earthquake moment

Calculate the ULS design wind moment, M∗WULS ,

Step 2.3 Select the largest beam size from steps 2.1 and 2.2 Step 3

(68.81)

where:

• Where two figures are given, the first is for a frame with sufficient stiffness to meet the P - ∆ OK deflection limit of NZS 4203 Equation 4.7.1 [2] or of DR 1170.4 Equation 6.5(1). The second is for a frame that does not meet this limit and for which the strength is increased to compensate, in accordance with DR1170.4 Clause 6.5.4.

(2)

(∑ M t,i )/ 2nbeam,i

Page 25

(68.82)

No. 68, June/July 2002

(3)

moment from section 3.18, using φMSHJ estimated for the given beam size from Table 68.2.

Check the column moment and axial load capacity at the first level (and then at all other levels where the beam size is checked) by using step 2.2, section 5.2 of [6], in conjunction with the overstrength moment for the t from section 3.18. To do this requires determination of φMSHJ. For preliminary design, the following guidance can be given for expected ratios of φMSHJ / φMsx,b, based on the range of representative frames designed for this project. Having determined the beam size and hence strength from step 2.3, a preliminary estimate of φMSHJ can be made from the following:

Step 8

Use the procedure given in step 4, section 5.2 of HERA Report R4-76 [6]. When applying equations 5.14 and 5.15 of [6], there are two general lateral deflection regimes to aim for. These are:

Table 68.2

(1)

Meeting the P - ∆ OK drift limit, such that no increase in strength to for P - ∆ actions is required

(2)

Meeting the maximum drift limit, as given by Clause 2.5.4.5 of [2] or Clause 8.5.2.1 of [4], and applying the P - ∆ enhancement given by Clause 6.5.4 of [4] if the P - ∆ drift limit is exceeded.

Indicative Values of φMSHJ / φMsx,beam Location in MRSF

Lowest seismic zone, P - ∆ OK

Other

≤ 0.5 H 0.75 H 1.0 H

0.25 0.35 0.30

0.45 – 0.6 0.45 – 0.55 0.40 – 0.45

For buildings in the highest seismic zones, the preferred option is (1). For buildings in the lowest seismic zones, the t strength required in meeting (1) may be considerably lower than that required to resist wind loading, thus making (2) the preferred option. Buildings in intermediate seismic zones may benefit from having both options checked.

Notes to Table 68.2: (1) H = height of structure (2) Lowest seismic zone, P - ∆ OK means design for Zmin and for the stiffness limits of NZS 4203 Equation 4.7.1 or DR 1170.4 Equation 6.5(1) to be met (3) Other means design for increased strength to cater for P - ∆ actions in conjunction with meeting the maximum drift limit for the lowest seismic zone, or any design for the highest seismic zone.

If the member sizes need increasing, increase the beam sizes to a greater extent than the column sizes, as the beam sizes have a greater effect on the MRSF stiffness. Step 9

Step 7.2 For the levels above the first seismic level

(2)

4.3

Procedure for MRSF final design

The procedure for final design is based on the established capacity design procedure from [5,6] for conventional category 1 and 2 MRSFs. It uses the detailed procedure given in section 6.2 of [6], wherever possible. It is presented in step by step format, using the step numbers and headings corresponding to those of section 6.2 of [6].

The column cross section (flange, web slenderness) complies with NZS 3404 Table 12.5 and Clause 12.8 for a category 3 member. (Category 3 rather than category 4 is used, because the ts develop an appreciably higher overstrength moment than conventional MRSFs, hence the columns above the base are marginally more likely to be subject to slight inelastic action than those of a conventional MRSF). For this reason, the cross sections are made category 3 to give them a small dependable inelastic rotation capacity without loss of performance.

Step 1

Analyse the frame for the required load cases and load combinations

Follow step 1 in section 6.2 of [6]; load case Emax is based on µ = 1.25. Step 2

Assess P – delta effects and check the seismic lateral deflections

The frame elastic stiffness (ie. with the ts in their closed condition) should be such as to comply with the appropriate lateral deflection regime (see section 4.2 step 8 above).

Column moment and axial load capacity complies with step 2.2, section 5.2 of [6], in conjunction with the estimated overstrength

HERA Steel Design & Construction Bulletin

Design the connections

This involves applying the detailed procedure given in section 3 herein.

At each level where the beam size has been determined, choose the column from the lightest designation of I-section type cross section which complies with all the following: (1)

Review member sizes to control lateral deflection

Page 26

No. 68, June/July 2002

Step 3

movement and inelastic rotation demand in the SHJs, however at the expense of column base damage. Pinned bases are very applicable for SHJ frames.

Derive the beam bending moments using moment redistribution

This step is not required as the gravity and seismic moments are applied as separate cases for beam design. This is a major simplification from conventional MRSF design for earthquake loading. Step 4 (1)

(2)

Step 8

The columns at the lowest level of fixed based frames are designed as category 2 in accordance with NZS 3404 Clause 12.8.3, using the design actions from step 7. Note especially the axial load/web slenderness requirements of Clause 12.8.3.1.

Determine the required beam sizes

The positive moment capacity must be able to resist the maximum moment from applied vertical loading (eg. from 1.2G & 1.6Q for [2]) in a simply ed manner. This is a more severe requirement than equation 6.2 of [6].

The columns at the higher levels of fixed based frames are category 3 . The columns at all levels of pinned based frames may be category 3 , however the pinned based detail itself must be detailed for a dependable inelastic rotation of 30 milliradians, in accordance with section 3.20.3 herein, which onreferences to section 4.2.2, page 23 of DCB Issue No. 50.

The second criteria on selecting beam size is to provide adequate frame stiffness. The preliminary design beam size selection is likely to dependably cover this requirement. If the beam size is required to be increased, the moment input through the t into the column does not have to increase accordingly.

Step 5

Step 9

Determine the beam overstrength moment capacities and design shear forces

4.4 4.4.1

Guidance on practical aspects of the MRSF design Estimation of fundamental period

The range of representative frame designs undertaken for the NITH over 2001/2002 have given an indication of the accuracy of equations 68.80.1 and 68.80.2. These frames have covered the following:

o capacity of the t, MSHJ . This is given by equation 68.76, section 3.18 herein.

In step 5.2, Ccol = 1.0 is used in all instances. Step 5.4 from [6] does not need to be applied, due to the nature of vertical shear force transfer from the beam to the column via. the t. This avoids the concentration of shear force in the beam web immediately adjacent to the t that occurs with rigid ted MRSFs.

• 5 and 10 storey • Auckland (low seismic zone) and Wellington (high seismic zone) • Intermediate soil conditions, with and without positive near fault action and soft soil conditions.

Evaluate the overstrength factors at each beam-column t

These designs have shown that the equations slightly over-predict the period determined by RUAUMOKO [10] for the frames in the most severe applications and underpredict it to as much as 15% in the least severe applications. The extent of overprediction (which is potentially unconservative) is not more than 5%.

Follow step 6, from [6], for ts in the superstructure, using the overstrength t capacity calculated from step 5 above. Follow section 3.20 on pages 23 herein for ts at the column bases.

4.4.2 Step 7

Design and detail the connections

This involves applying the detailed procedure given in section 3 herein.

The general details of steps 5.1 to 5.4 of section 6.2 of [6] are applicable, except that the overstrength moment capacity of the beam, o Mbeam , is replaced by the overstrength moment

Step 6

Design the columns

Determine the design actions for the typical levels and for the base

Choice of to use for the beams and columns

Follow the guidance given previously herein in conjunction with that given on pages 21 - 22 of DCB Issue No. 49.

Follow step 7, from [6], for the superstructure. Follow section 3.20 on page 23 herein for the column bases; fixed bases limit frame lateral

HERA Steel Design & Construction Bulletin

Page 27

No. 68, June/July 2002

to a 610 x 229 x 171W column, as shown in Fig. 68.5.

When using generically one-sided welded beams, these beams must be double sided welded through the connection region and for a reasonable distance beyond the bolt(s) furthest in along the beam flanges or web from the beam end adjacent to the column. A distance of 300 mm is recommended.

The design example is taken from level 1 of a 5 storey MRSF for Auckland, soft soil (soil class D from [4]). In this case, the P - ∆ provisions for stiffness are not met, so the force multiplier provisions of Clause 6.5.4 are used. These increase the design seismic base shear by 1.73 in this instance.

Because this semi-rigid system decouples strength and stiffness, the same beam size can be used over many levels, with the moment capacity of the t reduced at successively higher levels by reducing the number and diameter of the bolts. This offers considerable scope for matching moment capacity to moment demand more closely over each level of the MRSF than is possible with a rigid framed system, while using the same beam size. 5. 5.1

Design of the t shown in Fig. 68.5 is covered in section 5.2. 5.2

Design of the t

The detailing requirements from section 3.2 are all met by the SHJ detail shown in Fig. 68.5 and are not elaborated on further in this example. The allowances for manufacturing tolerances given in section 3.2.16 are incorporated into the flange plate offsets from the beam centreline shown in Fig. 68.5.

Sliding Hinge t Design Example Scope and introduction

Section 5 presents a design example for the SHJ. It relates to a 530UB82 beam connected

∗

∗

Fig. 68.5 Sliding Hinge t Design Example Notes: (1) The beam s a 120 mm slab on trapezoidal steel decking, deck rib height 54 mm, which is not shown. Concrete strength = 25 MPa. (2) The design moment and shear is: * M∗E = ME µdesign = 377 kNm (both directions ) - wind moment is not considered here ∗ VG, Qu =

(3)

∗ 122.4 kN (due to G & Qu) ; VGQmax = 185 kN The beam span, between column centrelines, is 7 m.

HERA Steel Design & Construction Bulletin

Page 28

No. 68, June/July 2002

This t design is not using Belleville Springs which would be the typical case.

ewb = 528 – 13.2 – 26.5 – 65 = 423 mm φVfss for M30 and 20 mm plate, = 98 kN

5.2.1

Design moment (section 3.4.1)

M ∗design

5.2.2

= M *Eµdesign

φMSHJ,new = 435 > M∗design = 377

= 377 kNm

Thus the design solution involves:

Design shear force (section 3.4.2)

VE∗µ design

=

3M E* µ design (Lb - d c )

M30 bolts; nbfb = 6; nwbb = 3 3 x 377 = = 177 kN (7.0 - 0.61)

5.2.5

Design of (section 3.7)

5.2.5.1

Net tension yield

∗ Vdesign = VGQu + VE∗µdesign = 122 + 177 = 299 kN ∗ = 185 kN - not critical VGQmax

5.2.3 (1)

plate

bbfp = 240; d'f = 33; fy,bfp = 250; tbfp = 20

bbfp,min (equation 68.4.1) = 4 x 24 + 90 = 186 mm bbfp,max (equation 68.4.2) = 1.05 x 229 = 240 mm from section 3.2.14, M24 bolts are used as first estimate for the 530UB82 beam size Try bbfp = 240 mm.

φNbfp > N*ty, bfp 5.2.5.2

√ O.K.

Net tension failure 98 x 1.45 x 0.9 = 959 kN 0.8 (equation 68.39) = 1098 kN

∗ Nu, bfp (equation 68.38) = 6 x

(2)

First estimate of bottom flange plate thickness 1.2 x 377 N∗tdesign (equation 68.29) = = 857 kN 0.528

φN∗tu,bfp

fu,bfp

= 410 MPa

φN tu,bfp ≥ N*u,bfp

857 x 103 (equation 68.30) ≥ 0.9(240 − 2x26) x 250 = 20.3 mm

5.2.5.3

√ O.K.

Compression capacity

fSHJ (equation 68.3) ≥ 10 + 1.25 x 30 x 10-3 x 528 + 2.5 x 20 = 80 mm

fy,bfp = 250 MPa is used From section 3.2.2, tbfp,max = 20 mm for M24 bolts. Determine sliding bolt numbers for moment (section 3.6)

flange

φN ty,bfp (equation 68.36) = 783 kN

Bottom flange plate width

5.2.4

bottom

N∗ty,bfp (equation 68.35) = 1.15 x 6 x 98 = 676 kN

Determine bottom flange plate width and initial thickness (section 3.5)

t tfp

√ OK accept

The answer for fSHJ is rounded up to the nearest 5mm

size and adequacy

Le,bfp (equation 68.41) = 0.7 (80 + 1.25 x 30 x 10-3 x 528) = 70 mm

φMSHJ, initial estimate (equation 68.31) = (4 x 56 x 528 + 3 x 56 x 438) x 10-3 = 192 kNm ewb (equation 68.33) = 528 – 13.2 – 26.5 – 50 = 438 mm

λ n,bfp (equation 68.42) 70  250  = 12.1 =    0.29 x 20  250

φVfss for M24 and 20 mm plate, Table 68.1 = 56 kN

As λ n,bfp < 25, equation 68.43 does not need modification

φMSHJ, initial estimate = 192 < M∗design = 377 NG

φNcu,bfp (equation 68.43) = 1020 kN

As this is considerably below requirements, increase both the bolt diameter and add one set of bolts to the bottom flange.

φN cu,bfp ≥ N *u,bfp 5.2.6

φMSHJ, new = (6 x 98 x 528 + 3 x 98 x 423) x 10-3 = 435 kNm HERA Steel Design & Construction Bulletin

√ O.K.

Design of the web top bolts (section 8)

* Vwv (equation 68.45) = Max (299; 185) = 299 kN

Page 29

No. 68, June/July 2002

5.2.9.3

* V1.2G + 1.6Q ≈ 1.85 (see Note (2) in Fig. 68.5)

n wtb ≥

299 214

= 1.40 ⇒ require 2 bolts

Determine plate thickness suppress tension yielding

t tfp,tension (equation 68.57)≥

As nwtb,required = 2 < nwbb,required = 3, increase nwtb to 3. The additional bolt is used as part of the top bolt group to anchor the sliding bolts. Design of web plate (section 3.9)

5.2.7.1

Check for vertical shear adequacy

1.15 x 882,000 = 25.8 mm 0.9 (240 - 66) 250

√ O.K.

Select ttfp = 25 mm – 4% under 5.2.9.4

5.2.7

Check top flange plate and bolt adequacy for the ULS condition

φNt,tfp (equation 68.60) = 1373 kN

φVvn, wp (equation 68.47) = 429 kN

φNc,tfp (equation 68.61) = 1275 kN

α v (equation 68.48) = 1.0 dwp = dwp,average (equation 68.12.3) = 448 mm dw = (equation 68.16) = 2 x 65 = 130 mm

0.85 ntfp φVfn,tfp = 0.85 x 8 x 214 = 1455 kN

* φVvn, wp = 429 kN > Vwv = 299 kN

5.2.7.2

Max (φNt, φNc) ≤ 0.85 ntfp φVfn,tfp √ OK

5.2.10

√ O.K.

Calculate beam tension adequacy in the connection region (section 3.12)

Check for net tension yield

N∗ty,wp (equation 68.50) = 1.15 x 3 x 98 = 338 kN

N*tb (equation 68.63) = 0.5 x 1.15 x

φN ty, wp (equation 68.51) = 437 kN

φMsx,b = 558 kNm, from [20]

d'f

= 33 mm

φN ty,wp > 5.2.7.3

Nt =

√ O.K.

N*ty,wp

3 x 98 x 1.45 X 0.9 = 479 kN 0.98

2840 , from [20] 0.9

fub Anb

φNtu,wp (equation 68.54) = 1023 kN φN tu,wp > N *tu,wp

Sizing of cap plates and brass shims (section 3.10)

5.2.9

Design of top flange bolts and plate (section 3.11)

5.2.9.1

Number of bolts required

= 440 MPa for the grade 300 beam = Ag – 4 x 33 x 13.2 – 2 x 33 x 9.6 = 8124 mm2

φNtb,2 (equation 68.64.2) = 0.45 x 10,500 x 0.300 = 1417 kN

√ O.K.

Min (φNtb,1 ; φNtb,2) = 1395 kN < N*tb = 1414 kN Beam is 1% overstressed for this check – accept.

This is done in accordance with section 3.10. The resulting sizes are shown in Fig. 68.5.

5.2.11

Design of welds between column flange and bottom flange plate (section 3.13) 1098 = 2.39 kN/mm run 2 x 230 = 1098 kN (section5.2.5.2)

v *w,bfp (equation 68.66) =

 1 ntfb,required (equation 68.56) = Even  (3 + 6 - (3 - 2 )) = 8 1.0 

kr = 1.0 as Lj = 3 x 90 = 270 mm < 15df = 450 mm 5.2.9.2

435 2840 x = 1414 kN 558 0.9

φNtb,1 (equation 68.64.1) = 0.39 x 8124 x 0.440 = 1395 kN

Check for net tension failure

N*tu,wp (equation 68.53) =

5.2.8

to

N*tw,bfp bmin

= Min (240 ; 230) = 230 mm bfc = 230 mm

Determine top flange plate width

φvw for a 15 mm leg length, category SP, fillet weld made with E48 weld metal = 2.44 kN/mm

Use btfp = bbfp = 240 mm

As φvw > v *w,bfp , use a 15 mm FW both sides.

HERA Steel Design & Construction Bulletin

Page 30

No. 68, June/July 2002

(This is changed to BW in section 5.2.12 for consistency with the top flange plate requirements).

5.2.15

5.2.12

This uses the procedure in section 3.2 of DCB No. 50, with modifications as described in section 3.17 herein.  fyp  A s,pair,required ≥ (bbfptbfp - t wc tbfp )    fys    (DCB No. 50, eq 50.2)

Design of welds between column flange and top flange plate (section 3.14) 1373 = 2.98 kN/mm 2 x 230

v *w, tfp (equation 68.68) =

N*tw = 1373 kN (section 5.2.8.4)

≥ (240 x 20 − 15.5 x 20)

As φvw for 15 mm FW < v *w, tfp , use a BW to the top flange plate. Same weld details are then also used to bottom flange plate for consistency. 5.2.13

Determine area of tension/compression stiffeners required (section 3.17)

250 = 4490 mm 2 250

fys = 250 MPa, as grade 250 plate is used. As starting point, select same thickness as bottom flange plate, ie. 20 mm.

Welds between column flange and web plate (section 3.15)

bs,min ≥ (0.9 bfp – twc) / 2 (DCB No. 50, eq 50.1)

5.2.13.1 Weld actions from vertical shear v *wv, wp, v (equation 68.69) =

v *wv, wp,h (equation 68.70) =

299 = 0.47 kN/mm 2 (448 - 130) 3 x 299 x 235 = 1.05 kN/mm (448)2

where: C1 = 15 (based on the incoming beam category of 3) Try 110 x 20 FL for stiffeners – Grade 250 used

ey (equation 68.72) = 80 + 65 +90 = 235 mm

(

v *wv, wp = (0.47)2 + (1.05)2

)

0.5

= 1.15 kN/mm

As,supplied = 2 x 110 x 20 = 4400 mm2

5.2.13.2 Actions on weld from momentinduced axial tension, web bottom bolts * Vwh, wp (equation 68.73) =

t s,min

= (0.9 x 240 – 15.5) / 2 = 100 mm b  f  ≥ ( s )  ys  = 6 mm C1  250    (DCB No. 50, eq 50.3)

Use 2 110 x 20 plate stiffeners for each pair of tension / compression stiffeners.

1023 = 2.62 kN/mm 3 x 130

Check 2bs + tw = 220 + 15.5 = 235.5 ≈ bfc = 230 mm 5.2.16

2 x 10

* v w, s,cf =

= 2.25 kN/mm

√ O.K.

Welds between stiffeners and column flange adjacent to incoming beam

v *w, wp (equation 68.74) = Min (2.62 ; 2.25) = 2.25 kN/mm

3

√ accept

As,supplied = 98% of As,required

5.2.13.3 Final design action

0.9 t wp fy,wp

√ O.K.

0.9bst sfys 2bs

(DCB No. 50, eq 50.4)

= 2.25 kN/mm

5.2.13.4 Sizing of weld

Use 14 mm leg length, category SP welds, E 48 filler metal

φvw for 14 mm leg length category SP FW, E48 weld metal = 2.28 kN/mm > v *w, wp

* φvw = 2.28 kN/mm > v w, s,cf

√ O.K.

Use a 14 mm leg length FW each side. 5.2.17 5.2.14

Welds between stiffeners and column web

Selection and location of positioner bolt * v w, s,cw =

See details in Fig. 68.5.

0.9bs ts fys C2d1c

(DCB No. 50, eq 50.5)

= 0.43 kN/mm HERA Steel Design & Construction Bulletin

Page 31

No. 68, June/July 2002

where: C2 = 2.0; 1 beam frames into connection (see Fig. 68.5) d1c

The spreadsheets cover the connection design, including the procedure given above. There are worksheets for the ts without Belleville Springs to the bottom flange plate and for the ts with Belleville Springs.

= 573 from [22] for the 610 x 229 x 171 W column

The spreadsheets are written for 5 and for 10 storey buildings, however this can be altered.

(Assume no doubler plate required at this stage; with a SHJ in a perimeter frame, this will always be the case for 1 beam framing into the column).

The spreadsheets do allow rapid t design to be made.

Use 5 mm leg length, category SP welds φvw

√ O.K.

= 0.82 kN/mm

5.2.18

They have been produced on Microsoft ® Excel for Office 97 and a copy of each is available free-ofcharge on a “use at your own risk” basis. They have been checked against the design example in section 5 and had quite thorough informal checking, but have not been through a formal quality assurance checking programme.

t overstrength moment (section 3.18) 1.4 x 435 = 761kNm 0.8 φMSHJ = 435 kNm (section 5.2.4).

MoSHJ (equation 68.76) =

5.2.19

Design shear force on zone (section 3.19)

Vp,* SHJ (equation 68.77) = Vcol hc

Associated with the spreadsheets is a data set of section properties that are called up by a macro routine and used in the frame design. 7.

761 - 217 = 1171kN (0.528 + 0.020)

= 761 / 3.5 = 217 kN = 3.5 m

5.2.20

5.2.21 φVc

= = = =

2. Dr John Butterworth Hank Mooy and Jos Geurts, University of Auckland, for assistance with planning and undertaking the extensive pseudo-static and seismic-dynamic experimental testing involved in this project.

27.9 mm; 15.5 mm; 629 mm 275 MPa zone adequacy (section 3.19)

= 1599 kN >

Vp,* SHJ

3. Dr John Butterworth, for his guidance and input as principal PhD supervisor to Charles Clifton.

= 1171 Kn

No web doubler plates are needed.

4. The Foundation for Research, Science and Technology, for providing the principal funding for this project.

That ends the design example. See Fig. 68.5 for the t details. 6.

The HERA Structural Engineer, principal author of this article, would like to acknowledge the contribution of all persons/organisations involved in this research, with special mention of: 1. The undergraduate students from who have undertaken and continue to undertake the setting up of experimental tests, the processing and presentation of data from this testing, the development of analytical modelling data and other essential work.

Design shear capacity of zone

φVc = 0.9 x 0.6 x 275 x 629 x 15.5 x 1.0 x [1.10] x 10 -3 = 1599 kN tfc twc dc fywc

Acknowledgments

Spreadsheet Programs are Available

Detailed spreadsheets have been developed for the design of the representative 5 and 10 storey frames. The design of the frames is to the seismic provisions of the new draft loadings standard [4], however the same approach is used with the current standard [2]. HERA Steel Design & Construction Bulletin

Page 32

No. 68, June/July 2002

Member Compression Capacity of a Solid Section

the points of end restraint. However the cross section at any point along the member length will not undergo distortion, it will simply move as an entity in the manner shown in Fig. 9.4 (a) of the HERA Limit State Design Guides Volume 1 [23]. However, if the member is restrained also along its sides, its behaviour changes markedly, as described in section 9.2 of [23]. With both end and side , kf ≠ 1.0 in all instances and must be calculated to Clause 6.2.4.

This article has been written by G Charles Clifton, HERA Structural Engineer.

Recently a design query has been received regarding calculating the member compression capacity of a solid section ed at its ends and loaded in compression through those ends. Two questions were raised, namely:

2. The compression member section constant, α b, is given in Table 6.3.3 (1) of NZS 3404 [5] for most types of cross section, but not for solid cross sections. For such sections, α b = -0.5 is used.

1. What value of form factor is applicable? 2. What value of member section constant is applicable? The answers to those are as follows:

Another point with regard to solid rectangular

Calculation of design member compression capacity, φNc, is undertaken to NZS 3404 [5] Clause 6.3. It is a two stage operation, the first stage being the calculation of section compression capacity, φNs, to Clause 6.2.

cross sections is that r = I / A = 0.29t, where t is the thickness in the direction of buckling. This means that rectangular cross sections ed only at their ends typically have low values of r, resulting in high slenderness ratios and a member compression capacity much lower than their section compression capacity.

When calculating the compression capacity of a steel member, the issue of buckling under the compression load is of paramount importance. This subject is comprehensively dealt with in Section 6 of the Standard [5]. Put simply: •

•

References 1. Clifton, GC; Thesis Report on the Development of New Semi-Rigid ts for Moment-Resisting Steel Frames. In preparation, due for publication first quarter of 2003.

Local buckling of elements of a cross-section under compression load is an issue when determining the section compression capacity and is addressed in Clause 6.2. This form of buckling involves one part of a cross section undergoing buckling relative to another part.

2. NZS 4203:1992, General Structural Design and Design Loadings for Buildings; Standards New Zealand, Wellington, New Zealand.

Member buckling involves the member moving out-of-plane between points of restraint. In member buckling, the whole cross section moves from its at-rest position, with this movement being effectively zero at the points of restraint and reaching a maximum within the midspan regions furthest away from the points of restraint.

3. Clifton, GC et. al.; Development of MomentResisting Steel Frames Incorporating SemiRigid Elastic ts 1995/96 Research Report; HERA Manukau City, 1996, HERA Report R488. 4. DR 1170.4 PPC 5 DR4/V Draft t Earthquake Loadings Standard, July 2002 Version; Standards New Zealand, Wellington.

With this background and through reference to the NZS 3404 provisions, the above two questions can be readily answered.

5. NZS 3404: 1997, plus Amendment No. 1: 2001, Steel Structures Standard; Standards New Zealand, Wellington, New Zealand.

1. The form factor, kf, is associated with the effectiveness of the cross section against local buckling. kf = 1.0 means that the cross section will not undergo any local buckling. A solid cross-section is in this category, so kf = 1.0 is the appropriate value to use. This applies even for a thin plate member restrained only at its ends and loaded through those ends in compression. Such a member will have low compression capacity, limited by member buckling between HERA Steel Design & Construction Bulletin

6. Feeney MJ and Clifton G C; Seismic Design Procedures for Steel Structures; HERA, Manukau City, 1995, HERA Report R4-76 ; to be read with Clifton, GC; Tips on Seismic Design of Steel Structures; Notes from Presentations to Structural Groups mid-2000; HERA, Manukau City, 2000.

Page 33

No. 68, June/July 2002

21. Clifton, GC et.al.; Moment-Resisting Steel Framed Seismic-Resisting Systems With Semi-Rigid Connections; SESOC Journal, Vol. 11, No. 2, 1998, pp. 21-41 and 43-52.

7. Clifton, GC et.al.; Two New Semi-Rigid ts for Moment-Resisting Steel Frames; NZSEE 2001 Conference, Wairakei; New Zealand Society for Earthquake Engineering, Wellington, 2001. 8. AS/NZS 1170.0:2002 Structural Design Actions Part 0: General Principles; Standards New Zealand; Wellington.

22. Structural Sections to BS4: Part 1 and BS 4848: Part 4; Corus Sections, Plates and Commercial Steels, Redcar, Teeside, UK, 1999.

9. Mago, N and Clifton, GC; Sliding Hinge t FEA Study; HERA, Manukau City, 2001, HERA Report R4-110.

23. Clifton, GC; Steelwork Limit State Design Guides Volume 1; HERA, Manukau City, 1994, HERA Report R4-80.

10. Carr, AJ; RUAUMOKO – the Maori God of Volcanoes and Earthquakes; University of Canterbury, Civil Engineering Department, Christchurch, 1998. 11. Pantke, M; Development of analytical Models for SHJ and SHJs; Report Produced for Second Industrial Internship, HERA; Manukau City, 2001. 12. Hyland C; Structural Steelwork Connections Guide; HERA, Manukau City, 1999, HERA Report R4-100. 13. AS/NZS 1252:1996, High Strength Bolts With Associated Nuts and Washers for Structural Engineering; Standards New Zealand, Wellington. 14. AS 1111.1; 2000 ISO Metric Hexagon Commercial Bolts; Standards Australia, Sydney, Australia. 15. AS 1566:1997, Copper and Copper Alloys – Rolled Flat Products; Standards Australia, Sydney, Australia. 16. Belleville Springs (Product Manual): Solon Manufacturing Company, Chardon, Ohio, USA. 17. Manual of Standard Connection Details for Structural Steelwork, Second Edition; HERA, Manukau City, 1990, HERA Report R4-58. 18. ISO 9223:1992, Corrosion of Metals and Alloys – Corrosively of Atmospheres – Classification; ISO, Geneva, Switzerland. 19. AS/NZS 1170.2: 2002, Structural Design Actions Part 2; Wind Actions; Standards New Zealand, Wellington. 20. Design Capacity Tables for Structural Steel, Third Edition, Volume 1: Open Sections; Australian Institute of Steel Construction, Sydney, Australia, 2000.

HERA Steel Design & Construction Bulletin

Page 34

No. 68, June/July 2002