Design of Portal Frame Buildings ‘Third Edition S.T. Woolcock Director, Bonacei Winward Consulting Engineers S. Kitipornchai Professor of Civil Engineering The University of Queensland M.A. Bradford Professor of Civil Engineering The University of New South Wales Published by Australian Institute of Steel Construction Level 13, 99 Mount StreetAUSTRALIAN INSTITUTE OF STEEL CONSTRUCTION A.CNN, 000 973 839 DESIGN OF PORTAL FRAME BUILDINGS Published by: AUSTRALIAN INSTITUTE OF STEEL CONSTRUCTION Enquiries should he sessed othe peblisher: Busines address ~ Level 13,99 Mount Stet, Noch Sydney, NSW. 2060, Austin, Postal address - P.O, Box 6366, North Sydney. NSW, 2059, Australia, E-mail adress ~ enguiries@aise.com.au Website - wwe ise.comau ‘© Copsrighe 1999 Australian Insite of tel Construction All igh ceserved. This book or any part thereof must not be reprodued in any form ‘witha tke writen permission of the Austin Institut of Stet Consrution. Published as Design of Parl Frame Buildings ~ It ekon to AS 1250)~ 1987 Limit Sate Design of Porta Frame Buildings Ist eition fo AS 4100)— 1991 Limit Sate Design of Pra Frame Buildings 2nd edition (10 AS 4100) ~1995 Desien of Portal Frame Buildings ~ 3d edition (to AS 4100) ~ 1999 (his edition) [ational Library of Australia Cataloguing-inePublication entry: Weolcock.S.T. Design of poral tame buildings. edad Bibliogaphy. Includes ines ISBN. 909945 84 5 | industrial bullngs~ Design and constuction. 2. Building, on anc sie ~ Design and construction. f Kikpomehs, Sl, Bradford, Mark A. (Marke Andrew) I Ausiralian Instvte of Ste! Construction 1 Tite 6931 Production by Robert Burton Printers Py Ld 653 Cartingtord Steet, Seton, NSW 2162, Australia DISCLAIMER Every effort has boen made an all easoable eae taken to ensure the accuracy of the ‘material contained inthis Publeation. Howeve. to the exent pemnied by law the ‘Authors, Edo and Publishers ofthis Publication: {wil rot be held lable oe espesible in any ways and (expen disclaim any Hibiiy or esponsibility, for any lost, demoge, cost or expenses insured in connection with this Publication by any person, whether that person te purchaser ofthis Publcaon or not, Without lation, this inches fst, damage, ons end expenses neared If any person wholly or partially relies on any pat ofthis Publication, and fos damage, costs and expenses incurred as a resul ofthe negligence a the Authors, Editors er Publishes WARNING “This Pbtiuion should not be weed without the services of a competent professional persan wih expen knowledge in the relevant fed, and under no ecastances should this Pubtication be relied upon fo replace any or all of the kaowledge and expertise of sich a person,Contents PREFACE NOTATION INTRODUCTION LA Portal Framed Buildings 1.2 Limit State Design 1.2.1 Background 1.2.2 Design for the Strength Limit State 1.2.3 Design for the Serviceability Limit State 13 Design Example 14 References Loaps 2.1 Background 2.2 Dead Loads 2.3 Live Loads 2.4 Wind Loads 2.4.1 General 2.4.2 Regional Wind Speeds 2.4.3 Wind Directions 2.44 Terrain Category 2.4.5 Basic Wind Speeds 2.4.6 Calculation of Pressures 2.4.7 Extemal Pressures 2.48. Internal Pressures 2.49 Area Reduction Factor 2.4.10 Local Pressure Factors 2.5 Load Combinations 2.5.1 Strength Limit State 2.5.2 Serviceability Limit State 26 Desien Example Londs Dead Loads Live Loads Wind Loads 2.63.1 Basic Wind Data 2 Extemal Wind Pressures 3 Internal Wind Pressures 4 Peak Local Pressures 2.6.4 Load Cases for Portal Frames 2.6.5 Load Combinations 2.7 References PurLins & Girts 3.1 General 3.2 Roof and Wall Sheeting 33 Frame Spacing.iv Contents 3.4 Purlin Strengths 3.4.1 Manufacturers" Brochures 3.4.2 _R-Factor Method .5 Deflections 5, Axial Loads 66 | Purlin Cleats 7 Purlin Bolts 8 Equivalent UDL’s for Peak Pressure ‘9 Design Example - Purlins 3.10.1 Member Capacity Brochures 3.10.2 Outward Loading - Cross Wind 3.10.3 Outward Loading - Longitudinal Wind 3.10.4 Purlin Selection for Outward Loading, 3.10.5 Check Inward Loading 3.10.6 Purlin Deflections 3.10.7 Purlin Summary 3.10.8 R-Factor Method 3.11 Design Example - Girts 3.11.1 Side Wall Girts 3.11.2 End Wall Girts with Span of 6250 mm 3.12 3.12 References 3, 3, 3, 3 3 3 FRAME DESIGN 4.1 Brame Design by Elastic Analysis 4.2 Computer Analysis 42.1 Load Cases 4.2.2. Methods of Analysis 423 Moment Amplification for First Order Elastic Analysis, 43 Rafters 43.1 Nominal Bending Capacity Mu, in Rafters 4.3.1.1 Simplified Procedure 4.3.1.2 Alternative Procedure 43.2. Effective Length and Moment Modification Factors for Bending Capacity 4.3.2.1 General 43.2.2 Top Flange in Compression 4.3.2.3 Bottom Flange in Compression 43.3. Major Axis Compression Capacity Nex 43.4 Minor Axis Compression Capacity Noy 43.5 Combined Actions for Rafters 4.3.6 Haunches for Rafters 4.4 Portal Columns 4.4.1 General 44.2 Major Axis Compression Capacity Nex 443 Minor Axis Compression Capacity Noy 44.4 Nominal Bending Capacity My, in Columns 44.4.1 General 4.4.4.2. Inside Flange in Compression 4.4.43. Ouiside Flange in Compression 4.5. Combined Actions 45.1 General 45.2 In-Plane Capacity 45.2.1 In-Plane Section Capacity‘tse DPFBAS Portal Framed Buildings 4.5.2.2 In-Plane Member Capacity 4.5.3 Out-of-Plane Capacity 45.3.1 Compression Members 45.3.2 Tension Members 46 Central Columns 4.6.1 General 4.6.2. Effective Lengths for Axial Compression 4.6.2.1 Top Connection Pinned Combined Actions with First Order Elastic Analysis ‘Combined Actions with Second Order Elastic Analysis 4.7 End Wall Frames 47.1 General 472 End Wall Columns 4.73. End Wall Columns to Rafter Connection 473.1 General 4.73.2 Continuous Rafter 4.733 Discontinuous Rafter 48 Braces 48.1 Fly Braces 48.2 Purlins as Braces 49. Deflections 4.9.1 General 4.9.2 Problems of Excessive Deflection 4.9.3 Recommended Deflections 4.10 Design Example - Frame Design 4.10.1 Frame Analysis 4,10,1.1 Preliminary Design 4.10.1.2 Haunch Properties 4.10.13 Methods of Analysis 4.10.2 Frame Deflections 4.10.3 Columns (460UB74) 4,10.3.1 Column Section Capacities 4.10.3.2 Column Member Capacities 4.10.33 Column Combined Actions 4.10.4 Rafters (360045) 4,10.4.1 Rafter Section Capacities 4,10.4.2 Rafter Member Capacities 4.10.4.3 Rafter Combined Actions LIMSTEEL Results End Wall Frames End Wall Columns 4.10.7.1 Inside Flange in Tension (Inward Loading) 4.10.7.2 Inside Flange in Compression (Outward Loading) 4.10.73 Axial Compression Under Gravity Loads 4.11 References ao ABS Bes ese FRAME CONNECTIONS 5.1 General 5.2 Bolted Knee and Ridge Joints Base Plates Design Example - Frame Connections 5.4.1 GeneralContents 55 5.4.2 Knee Joint 5.4.2.1 General 4.2.2. Calculate Design Actions for Bolts, End Plate and Stiffeners 5.4.2.3 Bottom Flange Connection 5.4.2.4 Top Flange Connection 5.4.2.5 Column Web Shear Stiffeners Ridge Connection Base Plates End Wall Column Connections 5.4.5.1 General 5.4.5.2 Centre Column - Top Connection 5.4.5.3 Quarter Point Columns - Top Connection References gee Bee ERG Roor & WALL BRACING 6.1 6.2 63 64 65 6.6 67 68 69 General Erection Procedure Forces 63.1 Longitudinal Wind Forces 6.3.2 Rafter Bracing Forces Bracing Plane Bracing Layout Tension Rods Tubes and Angles in Tension Tubes in Compression End Connections for Struts and Ties 6.9.1 Tubes 6.9.1.1 ‘Tubes in Tension 6.9.1.2. Tubes in Compression 6.9.2 Angles 6.10 Eccentricity 6.11 Design Example - Roof and Wall Bracing 6.11.1 Longitudinal Forces 6.11,1.1 General 11.1.2 Forces dite to Longitudinal Wind 11.1.3 Forces due to Rafter Bracing 11.4 orees in Root Bracing Members es r Tension Diagonals ‘onnections 11.4.1 End Connections for Struts 6.11.42 Bolts 6.11.5 Side Wall Bracing egygetes 6.12 References Footings & SLABS 1 12 73 Th General Design Uplift Forces Pad Footings Bored Piers 74.1 General Isc DPFH03 109 109 109 12 128 131 134 137 ity 14h Mi 142 143 145 145 146 146 146 146 147 147 152 155 159 161 161 161 163 163 163 163 163 163 164 166 166 166 168 m im 172 173 1B 191 191 192 192 193 193Asc DPFneS Portal Framed Buildings vii 7.4.2. Resistance to Vertical Loads 195 7.4.3. Resistance to Lateral Loads 195 7.5 Holding Down Bolts 196 75.1 General 196 } 752 Design Criteria 198 : 753 Grouting or Bedding 199 754 Bolts in Tension 199 7.5.4.1 Anchorage of Straight or Cogged Bars 199 7342 Cone Failure 199 73.43 Bmbedment Lengths 201 7.5.44. Minimum Edge Distance for Tensile Loads 201 75.5 Bolts in Shear 204 75.6 Corrosion 205 7.6. Slab Design 205 7.6.1 Design Principles 205 7.6.2. Slab Thickness 205 76.3 Joints 206 7.6.3.1 General 206 7.632 Sawa Joints 206 6.3.3 Cast-In Crack Initiators 207 7634 Keyed Joints 207 7.6.3.5 Dowelled Joints 209 7.63.6. Joint Spacing and Reinforcement 209 7.7 Design Example - Footings 210 7.1.1 Typical Portal Footings 210 TIAA Bored Piers 210 7.7.1.2 Compare Pad Footings 212 7.1.2. Bnd Wall Column Footings 2212 7.73 Main Portal Footings in Braced Bays 213 7.13.1 Comer Columns 213 7.13.2 Column on Grid B2 214 7.733. Columns on Grids A2, A8 and BS 214 7.73.4 Holding Down Bolts for Portal Columns 214 -73.5_ Holding Down Bolts for End Wall Columns 215 78 Design Example - Slab 215 78.1 Design Criteria 215 7.82. Slab Thickness Design 216 783° Joints 216 7.84 Reinforcement 217 7.9 References 217 8 PLASTIC FRAME DESIGN ag 8.1 General 219 8.2 Plastic Analysis 219 8.2.1 General 219 8.2.2 Direct Mechanism Method 220 3 Iterative Mechanism Method 223 2.4 Statical Method 25 8.2.5 Second Order Effects 225 83 Basis of Plastic Design in AS4100 25 8.4 Member Capacities 226 85 Design Example - Plastic Frame Design 207 8.5.1. Preliminary Design . 227viii Contents Isc DPFHNS 85.1.1 Gravity Load Case 27 8.5.1.2 Cross Wind Load Case 229 85.1.3 Deflections 231 8.5.2 Detailed Design 21 85.3 Columns 238 8.5.3.1 Section Capacities 238 85.3.2 Member Capacities 239 85.4 Rafters 242 8.5.5 Serviceabitity 244 8.5.6 Comparison of Plastic and Elastic Solutions 244 8.6 References 244 9 (GANTRY CRANES & MONORAILS 247 9.1 General 247 9.2 Design Procedure for Gantry Cranes 248 9.3 Crane Runway Beams 249 93.1 General 249 9.3.2 Design Loads and Moments 250 9.3.3 Member Capacity in Major Axis Bending Mj, 251 9.3.3.1 A84100 Beam Design Rules 251 9.3.3.2 Proposed Monosymmetric Beam Design Rules, 252 9.4 Portal Columns Supporting Crane Runway Beams 254 9.5 Monorail Beams 234 9.5.1 General 234 9.5.2 Member Capacity Tables 234 9.5.3 Local Bottom Flange Bending 255 9.6 Design Example - Gantry Crane 255 9.6.1 Load Cases 255 9.62 Crane Runway Beam 258 9.6.2.1 Major Axis Bending Moments 258 9.6.2.2. Minor Axis Bending Moments 259 9.6.2.3 Combined Actions 262 9.6.2.4 Check Major Axis Compound Section Capacity ¢Mpx 263 9.6.2.5 Deflections 263 9.6.2.6 Vertical Shear Capacity 263 9.6.2.7 Shear Buckling Capacity 264 9.6.2.8 Shear and Bending Interaction 264 9.6.2.9 Bearing Capacity of Crane Runway Beam 264 9.6.2.10 Check Bffect of Eccentric Corbel Loading on Column 266 9.6.2.1 Check Effect of Vertical Loads on Web 267 9.6.2.12 Check Effect of Eccentric Rail Loading on Crane Runway Beam Web 268 9.6.2.13 Check Effect of Web Buckling Under Vertical Loads 2m 9.6.2.14 Fatigue 2 9.6.3 Check Portal Frame 22 9.6.3.1 Loads 222 9.6.3.2 Load Combinations 223 9.6.3.3 Columns 273 9.7 References 271 Appendix 9.1 Design Capacity Tables 279 Appendix 9.2 Background to Design Capacity Tables 287 A9.2.1 General 287 ‘A9.2.2 Section Moment Capacity Mex 287AISC DPFBIOS Portal Framed Buildings A9.2.3 Member Moment Capacity Mex ‘A9.2.4 Member Capacity to AS4100 APPENDIX]: DRAWINGS APPENDIX I: COMPUTER OUTPUT Geometry; Load Cases; Deflections Second Order Analysis; Load Combinations; Member Forces; Reactions Joint and Member Numbering; Displaced Shapes; Bending Moment Diagrams Elastic Critical Load Analysis APPENDIXID: LIMSTEEL OUTPUT SUBJECT INDEX 289 290 293 303 305 3 317 323 327 333Preface In October 1985, Scott Wooleock and Sritawat Kitipornchai presented a non-technical paper entitled Some Aspects of the Design of Industrial Buildings to a conference of! the Australian Institute of Construction Supervisors at the Gold Coast, ‘The paper outlined some of the grey areas in the design of portal framed buildings. AISC were very interested in the paper and invited these two authors to write the eatlier working stress version of this book. It was entitled Design of Portal Frame Buildings and was published in 1987, The working stress version was then completely rewritten for the change to limit states design. ‘The first limit state edition was published in 1991 and was entitled Limit State Design of Portal Frame Buildings. Further changes were made for the second limit state edition in 1993 to incorporate amendments to AS4100 and AS 1170.2, to reflect changes in the AISC structural connections manual and to generally refine the limit state design process. This third limit state edition has been almost completely rewritten to cater for the change in basic steel grade from 250MPa to 300MPa and the change in roof wind loads in Amendment No, 2 of AS1170. The release of the limit stale cold formed structures code A$4600 in 1996 and the publication of the Lysaght and Stramit limit states purlin and girt brochures in 1999 have also been fully accounted for. Because limit state design is now well established, the title has reverted to the simpler, original litle - Design of Portal Frame Buildings. ‘A new chapter dealing mainly with the design of portal frame buildings for overhead travelling cranes has been added. It covers the design of crane ranway beams and addresses some ambiguities and inadequacies in AS4100's treatment of monosymmetric beams. The chapter includes design capacity tables for top flange (and above top flange) loading of some standard combinations of UB's and WB's with PFC top flange channels. In addition, the effect of crane loads and crane deflection limits on the design of the portal frames is addressed. Some typical details are provided. The theory is extended to bottom flange (and below bottom flange) loading of UB and WB monorails, and design capacity tables are presented. The design capacity tables for crane runway beams and monorails should prove to be of great assistance to designers because there has been little if anything published since the sixth edition of AISC’s Safe Load Tables for Structural Steel in 1987. The 1987 tables were working stress design tables based on a steel grade of 250 MPa and did not account for above top flange or below bottom flange loading, ‘The design capacity tables for CHS and SHS roof and wall bracing struts, which are unique to this book, have been expanded to cater for the Duragal range of sections, These tables account for the effect of self-weight bending in combination with axial compression. Tension capacities and maximum spans for span/150 deflection are now given for each CHS and SHS section. ‘The previous chapters on footings and slab-on-the-ground have been merged. The design of bored piers is now generally in accordance with the limit state design approach of AS2159-1995 although different geotechnical capacity reduction factorsASC DPFRAS Preface xi are proposed depending on whether the piers are classified as short or long, A quadratic expression derived from Broms’ work by the authors for use in determining the lateral load capacity of short bored piers is now presented in the text of the footings chapter rather than being somewhat hidden in the design example. This expression facilitates the preparation of spreadsheet programs for determining the lateral capacity of bored piers. ‘The authors’ association started at The University of Sydney where all three obtained doctorates conducting research into steel structures under the supervision of Professor NS Trahair. This association has continued over the years. Overall, this addition should prove to be of great assistance to practicing engineers and students. The authors gratefully acknowledge the positive feedback from many users, Firstly they would like to thank consulting engineers Bonacci ‘Winwatd and the Departments of Civil Engineering at The University of Queensland and The University of New South Wales for their support in preparing this book. Although Bonacci Winward’s Brisbane office prepared most of the diagrams, Brice Engineers of Townsville prepared the three dimensional view in Chapter 1 using Strucad, and this is much appreciated. Simon Pikusa’s idea for and contribution towards the plastic design chapter in the first edition is acknowledged. In particular, the authors would like to thank Arun Syam, National Manager - Technology at the ‘Australian Institute of Steel Construction for his continued encouragement. Finally, the authors would like to express their appreciation for the continued support of their wives and families during the preparation of this edition. Scott Wooleock Sritawat Kitipomehai ‘Mark Bradford September 1999Notation ‘The following notation is used in this book. Where there is more than one meaning to a symbol, the correct one will be evident from the context in which it is used. Generally, the notation has been chosen to conform where possible to that in the relevant design standard, bn cross-sectional area, or tributary area which transmits wind forces to elements. core cross-sectional area of bolt shank area of rod flange area at critical section flange area at minimum cross-section ‘gross area of cross-section net area of cross-section cross-sectional area of tension reinforcement, or tensile stress area of bolt or bracing rod, or effective area of stiffeners gross sectional area of web area of column web dimension used in defining extent of application of local wind pressure factors eight of application of load below shear centre of a monosymmetric beam = Wye- spl? = (6g bo = 22 edge distance from bolt centreline to top or bottom edge of end plate distance from bolt centreline to face of rafter flange effective value of ar for bolted moment end plate cexige distance from bolt centreline to side edge of end plate = (b- s,)/2 "distance between crane wheel loads overall dimension of square hollow section frame spacing, or length of building normal to wind stream web bearing width used in AS4100 at the neutral axis of the member ‘web bearing width used in AS4100 at the junction of the web and inside face of flange effective width of plate clement stiffener outstand from face of web flange width of beam flange width of column width of end plate = tye + 2re width of railhead xiialse DPFROS ds Ey Ey Sat & Notation xiii average breadth of shielding buildings normal to wind stream, or stiff bearing length wind pressure coefficient extemal wind pressure coefficient | intemal wind pressure coefficient cross wind undrained cohesion dead load, or beam depth, or rod diameter, or tube diameter, or hold down bolt diameter, or building spacing parameter in determining shielding dead load minimum roof plan dimension, or depth ofa building parallel to windstream, or bored pier diameter clear depth between flanges ignoring fillets or welds twice the clear distance from the neutral axis to the inside face of the compression flange ‘beam depth column section depth distance between flange centroids, or nominal bolt diameter volt hole diameter in bolted moment end plate minimum depth ofhaunch (equal to rafter depth) depth of web plate, or clear distance in Appendix I of AS4100 column section depth between fillets = de = 2ke ‘Young's modulus of elasticity - minimum edge distances for hold down bolts subjected to tensile load and shear, respectively eccentricity above ground line of applied load to bored pier, or eccentricity of crane loading allowable working stress in AS1250 clastic buckling stress in AS1250 vertical distance from knee to ridge for plastic analysis axial stress axial stress in cable or rod bonding stresses in stiffeners at end wall column to rafter connection characteristic strength of concrete design valuc of shaft adhesionNotation AISC OFFS tensile stress, or tensile strength of conerete ultimate tensile stress ultimate tensile strength of bolt normal tensile strength of weld material average design shear stress in web maximum design shear stress in web yield stress column yield stress of flange or web, respectively yield stress of CHS yield stress of doubler plate yield stress of bolted moment end plate yield stress of stiffener equivalent design stress on web panel factors in elastic monosymmetric beam buckling formula nominal dead load, or shear modulus of elasticity end restraint parameters for a compression member in AS1250 part of dead load which resists instability column height for plastic analysis, or column height for effective length calculation design lateral force on bored pier height of rail design lateral bored pier capacity caves height, or height of structure above ground monorail load height eaves height average height of shielding building ridge height, or rail height value of f, for column internal pressure second moment of area of flange polar moment of area value of fy for rafter intemal suction second moment of area of web, or ‘warping section constant second moments of area about major (x) and minor (y) principal axes, respectivelyAISC DPF BMS Ie J de Li, LaLa Len bey Notation xv ‘minor axis second moment of area of compression flange Saint Venant torsion constant torsion constant for rail beam parameter for monosymmetric beam distance on column from outer face of flange to inner termination of root radius tye * Pe member effective length factor form factor for a member subjected to axial compression load height effective length factor spring stiffness proportion of design moment transmitted by web coefficient to allow for additional bolt force due to prying, effective length factor for restraint against in-plane lateral rotation twist restraint effective length factor, or Joad eccentricity reduction factor for tension members ratio of area of web {0 total cross sectional area span, ot member length, or rafler span, ot embedded length of bored pier embedment lengths of hold down bolts for singe cone, two intersecting cones and four intersecting cones respectively length of column of compression member or laterally unsupported beam value of Z- about major (x) and minor (y) principal axes, respectively holding down bolt cog length {ength of web along which rail load is uniformly distributed live load length of rafter measured between centre of column and apex average spacing of shielding buildings bending moment “design bending moment ‘gust wind speed multiplier for terrain category ‘cat’ at height z nontinal bending moment capacity value of Mj, about major principal x axis, wind direction reduction factor nominal in-plane member moment capacity, oF structure importance multiplier for design wind speed maximum calculated design bending moment along length of a member or in a segmentxvi Notation isc DPFBa3 clastic critical uniform bending moment for a beam with ends fully restrained against lateral translation and twist rotation but unrestrained against minor axis rotation amended elastic buckling moment for a member subject to bending reference elastic buckling moment obtained using Le elastic ertical bending moment calculated by elastic buckling analysis and incorporating moment gradient, height of loading and restraint conditions ‘Mg for a scament, fully restrained at bolt ends, but unrestrained against lateral rotation and loaded at the shear centre . nominal out-of-plane member moment capacity about major (x) axis required design moment by plastic frame analysis nominal plastic moment capacity reduced by axial force value of Mpr about major principal x axis value of May reduced by axial force nominal section moment capacity in bending, or shielding multiplier for design wind speed value of M, about major principal x axis topographic multiplier for design wind speed torsional moment in rail design first yield moment ultimate moment capacity of bored pier design bending moment on web panel moment modifying factor in monosymmetric beam buckling formula design axial force, tensile or compressive nominal member capacity in compression design axial force on column nominal strength of stiffener in compression value of N; for buckling about the major (x) and minor (y) principal axes, respectively total compression design force in flange total tension design force in flange PEP clastic flexural buckling load of member value of Nom for braced member nominal capacity of bolted end plate in bending valuc of Now for sway member design axial force in rafter reduced nominal axial capacity of horizontal tubular strut due to self weight bending nominal section capacity for compressive axial forceAISCDPFBAS ™ Me Ny Ne Ne Now Nr Ne Nw 1 Rs P Pa Ps Q % R Notation xvii nominal section capacity for tensile axial force nominal capacity of bolts at tension flange nominal tension capacity of a bolt or bracing rod. design bolt tensile force nominal strength of stiffener in tension capacity of tube wall near cap plate nominal strength of stiffener designed to resist excess shear in column, design force on stiffener due to shear nominal capacity of fillet or butt weld for flange subjected to axial force number of bolts in bolt group number of upwind shielding buildings within 45° sector or radius 20h, applied load, or magnitude of anchor head of holding down bolt applied load crane dynamic wheel load design wind pressure at height z nominal live load free stream gust dynamic wind pressure resulting from Vz reduction factor in cold-formed code, or redundant force in plastic analysis, or support reaction nominal bearing buckling capacity nominal bearing yield capacity nominal capacities of column adjacent to beam compression flange nominal capacity of stiffened column adjacent to beam compression flange rafler length along slope from column centreline to apex in plastic design nominal buckling capacity of stiffened web nominal capacities of column adjacent to beam tension flange nominal capacity of column flange with doubler plates adjacent to beam tension flange nominal capacity of stiffened column flange adjacent to beam tension flange design bearing force or reaction on web panel used in Appendix I of AS4100 nominal capacity radius of gyration root radius of column section ratios used for tapered member in AS4100 radius of gyration about the major (x) and minor (y) axes, respectively plastic section modulus design action effect distance between fly bracesxviii Notation aise prs distance between purlins or girts ratio of plastic section modulus of column to unhaunched rafter safe working load purlin spacing, bolt gatige bolt pitch flange thickness, or force in tension diagonal, or thickness of anchor head of holding down bolt thickness, or web thickness, or thickness of tube wall thickness of doubler plate end plate thickness flange thickness ‘beam flange thickness column flange thickness root radius in rail design thickness of stiffener fillet weld throat thickness ‘web thickness, or fillet weld leg length beam web thickness column web thickness thickness of web doubler plate. regional basic gust design wind speed design shear force nominal shear buckling capacity design shear force in column nominal capacity of single bolt in shear used in AISC’s connections manval nominal shear capacity of bolt used in AS4100 nominal shear capacity of bolt group used in AISC’s connections manual design shear force in bolt used in AS4100 basic wind speed for permissible stress method nominal capacity of plate in shear basic wind speed for serviceability limit state basic wind speed for ultimate limit state, or nominal shear capacity of web in uniform shear ‘nominal shear capacity of web vertical design shear force at interface of end plate and column nominal web shear capacity in the presence of bending momenta HPS OE RPRRE Notation xix nominal shear yield capacity of web basic design gust wind speed at height z nominal capacity of fillet weld per unit length weld force component in y direction weld force component in z direction nominal wind load, or ‘weld size used in anchor head of holding down bolt external work internal work serviceability wind load ultimate wind load uniformly distributed load design uniformly distributed toad distributed dead load equivalent uniformly distributed load distributed live load nominal load nominal loads in plastic frame analysis sag in cable or rod effective section modulus effective section modulus of web used in Appendix I of AS4100 distance or height above ground level angle of slope of roof, or reduction coefficient for adhesion on bored pier, or load position parameter for monosymmetric beams compression member section constant value of @, about major (x) and minor (y) principal axes, respectively moment modification factor for flexural-torsional buckling, slenderness reduction factor slenderness reduction factor for monosymmetric beams reduction factor for tapered member shear buckling coefficient ratio of smaller to larger bending moment at ends of member monosymmetry parameter ratios of compression member stiffness to end restraint stiffness frame or member deflection sway deflection ‘moment amplification factor for braced member moment amplification factor, taken as the greater of d and J moment amplification factor for sway memberSUPRER OBR ON y 4 Seas = Notation AISC DPEBAS, load height parameter angle of deviation of wind stream from axis of structural system, or virtual angle of rotation in plastic analysis clastic buckling load factor : factors for calculating lateral capacities for bored piers depending on whether piers behave as long or short modified compression member slendemess value of 4, about major (2) and minor ()) principal axes respectively web plate element slendemess ‘web plate element yield slendemess limit degree of monosymmetry short term load factor capacity reduction factor capacity reduction factor for bending in cold-formed structures code ‘geotechnical reduction factor for bored piers value of g, for long or short pile, respectively.1 Introduction i 1.1 PORTAL FRAMED BUILDINGS Portal-framed steel clad structures are the most common type of industrial buildings. They find extensive use as industrial factory and warehouse structures, and as indoor sporting venues. The major components of a portal frame building are a series of parallel portal shaped frames as the major framing elements. Each frame is rigid, and resists horizontal wind forces and gravity loads in the plane of the frame by flexural action. A typical portal frame is, shown in Figure 1.1. Longitudinal wind forces that are perpendicular to the frames are resisted by triangulated bracing systems in the roof and walls which prevent the frames from falling over. Am illustrative isometric view of the steel skeleton of a bracéd bay of a portal frame building is shown in Figure 1.2. This book presents limit state design procedures for the design of portal framed buildings based on Australian standards, as described in Section 1.2, Large clear spans of about 40 metres can be achieved economically using Universal Beam (UB) or Welded Bearh (WB) rafters such as those manufactured by BHP [1]. The columns are generally larger than the rafters because the rafters are haunched near the columns to cater for the peak bending moments at the columns. For larger spans, some form. of roof truss, as shown in Figure 1.3, is often used in lieu of UB or WB rafters. As the span increases, the weight saving offered by trusses becomes more pronounced, until the higher cost per tonne for truss fabrication is eventually offset. ‘The crossover point is difficult to rominate because of the many variables. One of the difficulties of the comparison is that a building with roof trusses is higher than a building with portal frames, assuming that the same internal height clearances are maintained. ‘The main drawback of a trussed roof is the need for Span Figure 1.1. Typical Portal Frame2 Introduction Ise DPEwOS bracing the bottom chord, Nevertheless, it is recommended that the cost of using portalised trusses in preference to portal frames for a particular project be investigated where the span exceeds 30 metres or 50. Bolted moment fend plete connection. at riege bracket Doubler plotes \ copenn flange stiffener Double diagonal wall bracing Figure 1.2: Structural Components in a Braced Bay l Roof truss with diagonal UB or WB column web members orientated to ‘suit dominant uplift loan Figure 1.3 Portalised Trussassc PFDs Portal Framed Buildings 3 Although portal framed buildings are very common, the number of manuals and handbooks dealing with their design is comparatively small. This book considers the design of portal framed buildings in accordance with the Australian limit states stec! structures code AS4100 [2], which was first introduced in 1990 in response to an international trend towards limit state design. Prior to the mid-eighties, the design of structural steelwork in most westem, countries was undertaken using permissible or working stress methods. Very little mention of these methods will be made in this book, since they have now been superseded. Apart from the 1978 Canadian code [3], limit state design standards for steel structures were released after 1985: in 1985 and 1990 in the United Kingdom [4], in 1986 in the United States [5], in 1990 and 1998 in Australia [2] and in 1992 in New Zealand [6]. Background information on the development of the Australian limit state code is given in Section 1.2. It may be thought that the design of portal-framed buildings is simple and straightforward, However, some aspects of AS4100 and the wind loading code [7] are ambiguous, and the behaviour of many aspects of the structure is not well understood. For example: . Methods of Analysis ‘There are now three main methods of analysis which could be used in the design office as follows. «+ Blastic analysis. This requires separate manual amplification of the moments which in tum requires the determination of the frame buckling load factor. This is achieved by using appropriate formulae such as those developed by Davies [8] or by utilising an elastic critical load analysis using commercially available programs such as Microstran [9] or ‘Spacegass [10]. + Nonlinear or second order elastic analysis. This is readily available in proprictary programs, and does not require the amplification of moments. © Plastic analysis, This is described in Chapter 8. Note that more advanced analysis programs are starting to become available, but to date these are generally only research tools Loads + External pressures ace generally prescribed clearly in AS1170.2 but two values of roof pressure cocfficients are given, ie -0.9 or -0.4; -0.5 or 0; -0.3 or +0.2; -0.2 or +0.3. Some designers use the first coefficients mixed with the second to produce the worst effect, whereas the intention of the code is that the first and second coefficients be used as alternative sets. + The choice of internal pressure coefficients is largely a matter of judgement for the designer. This means that different designers can arrive at different solutions for a given project. Effective Lengths of Compression Members (Flexural Buckling) Effective lengths of compression members in portal frames need to be determined as shown in Chapter 4 for: ‘+ _In-plane or major axis buckling under axial load alone (Lexis generally # L). © Inplane or major axis buckling for assessing in-plane member capacity under combined actions (Ler = LOL).4 Introduction Ais DPFBAS © Outoftplane or minor axis buckling for assessing out-of-plane member capacity under combined actions (Ley is generally < 1.0L because of restraint by purlins and girts). Effective Lengths of Beams (Flexural-Torsional Buckling) The rules in AS4100 for determining effective lengths of beam segments are relatively complex, and depend on: End lateral restraints. End torsional restraints. End minor axis bending (lateral rotational) restraints, Height of loading with respect to the shear centre. Tension Members under Self Weight ‘The tensile capacity of horizontal tension members such as double diagonal roof bracing members under self weight bending is not widely understood. For example: + Some designers consider the combined actions of tension and self weight bending ‘moments in tubes and angles whereas tension only is an adequate consideration, ‘+ Some designers are uncertain about appropriate limits on the deflection of roof bracing. ‘members under self-weight alone. Guidance is given in Chapter 6. ‘+ The level of prestress needed for zods, its control on site and the effect, if any, on the limit state tensile capacity of the rods are also issues not well understood and are addressed in this book. «The design of welded T-end connections is not well covered in the literature, and guidance is given in Chapter 6, Roof Bracing Struts under Self Weight Under AS4100, the capacity of a strut under transverse loads is effectively determined by comparing moments rather than axial forces. These moments are relatively small and sensitive to the level of axial load, Designers therefore cannot readily develop a feel for the axial capacity. Design compression capacities of CHS and SHS members under self-weight, unique to this book, are presented in Chapter 6 Holding Down Bolt Embedment ‘The design of holding down bolts is not covered by either the stee! or concrete standards, and there is wide variation in practice. ‘The earlier working stress version of this book [11] presented information on embedded bolts drawn from research by the American Concrete Institute on nuclear safety-related structures, and this information has now been incorporated into the AISC’s Structural Connections book {12}. ‘The essential details in a slightly revised format are presented in this edition. Geotechnical Limit State Design There are no Australian standards for the working stress or limit state design of pad footings for buildings, The piling code [13,14] covers the limit state design of bored piers and the Bridge Design Code addresses the limit state design of pad footings in a comprehensive but overly complex way. This book presents some useful information on the limit state design ofAISC PFE Portal Framed Buildings 5 bored piers including 2 unique formula (see Equation 7.4 in Chapter 7) for assessing the lateral capacity of bored piers in cohesive soils. Design for Gantry Cranes and Monorails ‘The limit state design of erane runway beams and the portal frames which support them is not covered comprehensively by Australian standards or handbooks. Chapter 9 addresses these issues and ineludes a design example. Comprehensive design tables are presented for the designer to help choose the correct composite runway beam for a given crane loading based on a rational buckling analysis of the monosymmetric runway beam. Tables for the bending, capacity of monorails with central concentrated loads at bottom flange level and 200 mm below bottom flange are also presented, This book has two essential aims, It attempts firstly to provide an interpretation and explanation of the limit state approach to the design of portal frame structures using AS4100. Secondly, it attempts to throw some light on many of the problems encountered in portal frame design. It tries to deal with the problems not normally covered by textbooks, and to provide a state-of-the-art book on the limit state design of portal frame buildings from the roof sheeting down to the slab-on-ground and footings. Although not intended to be a complete step by step design manual, the book presents a comprehensive worked design example which is followed through each chapter. The brief is given in Section 1.3. Material readily available in other publications such as industrial pavement brochures, geotechnical standards and standard connection manuals is not reproduced here, but ‘comments are provided, The Australian loading standards AS1170.1-1989 Part 1: Dead and live loads and load combinations {15} and AS1170.2-1989 Part 2: Wind loads (7} are used ‘throughout this book, as are the design standards AS4100-1998 Stee! structures [2], AS/NZS4600-1996 Cold-formed steel structures [16] and AS3600-1994 Concrete structures [17]. Other material is referenced as used in the text. 1.2. LIMIT STATE DESIGN 1.2.1 Background ‘The rational technique of treating loads and strengths as random variables has led to the development intemationally of limit state design procedures, and these design procedures, have been adopted for use in Australia, Until 1990 when AS4100 was first released, portal frame buildings had to be designed predominantly in accordance with working stress or permissible stress philosophies [18]. Since 1996, following the release of AS/NZS4600-1996, the cold formed steel structures code, it has become possible to design all components of a portal frame building using limit state design procedures. Although the superstructure of a pottal frame building can be designed totally in accordance with limit state principles, some of the geotechnical aspects of the foundations must still be designed to working stress principles. ‘The limit state approach for the design of structures arose because it was recognised that different types of load (dead, live, wind, earthquake and even snow) have different probabilities of occurrence and different degrees of variability. Furthermore, the probabilities associated with these loads change in different ways as the degree of overload increases.6 Introduction AIScDPFBIS Limit state design thus differs from working stress design in that not only are load factors used, but different load factors are also used for different load types and different limit states, and different capacity reduction factors are used for different materials. ‘The advantage of limit state design over working stress design is that it is more logical and provides a more consistent margin of safety [19,20]. It can serve better to evaluate existing structures, and should result in more economical portal frame buildings. One of the major advantages of limit state design is that it leads to more rational load combinations. This climinates the problem encountered in working stress design of combining wind upli with dead loads, which was discussed in Reference [11]. Jn the limit state approach, the structure must satisfy simultaneously a number of different limit states or design requirements. It must possess adequate strength, be stable against overtuming or uplift, and perform satisfactorily under service loads. The structure must also be durable, possess adequate fire protection, resist fatigue loading and satisfy any special requirements which are related (o its intended use, Codes of practice specify design criteria which provide a suitable margin of safety against a structure becoming unfit for service in any of these ways. When a particular limit state is satisfied, the probability of exccedance (eg. the probability that a column or rafter will buckle or that a deflection will be excessive) is very small. The limit state design criteria adopted for use in AS4100 were calibrated [21] so that this probability is comparable with historical exceedance probabilities implied in the superseded working stress design code AS1250 [18]. The limit states of strength (including stability against overtuming) and serviceability must be considered separately, and satisfaction of one does not ensure satisfaction of the other. For each limit state, the designer must compare the capacity of the structure with the appropriate extemal loads. The latter are obtained from the loading codes ASI170.1 and AS1170.2, while the capacities are obtained from the relevant steel or concrete standard, The loads and load combinations for industrial portal frame buildings are discussed in the next chapter, while the remaining chapters are devoted to examining the capacities of these structures. 1.2.2. Design for the Strength Limit State The design action effect $* is calculated by the methods of structural analysis from the most severe load combination for the strength limit state (see Section 2.5.1). At a particular cross- section, the design action effect may be the axial force IV", the shear force ¥”*, the bending moment M*, or combinations of these. Computer programs such as Microstran [9] and Spacegass [10] are almost invariably deployed to calculate these design action effects. ‘The design strength of a member is taken as the product of its ultimate strength or nominal capacity R,, and an appropriate reduction factor 1 The capacity reduction factor is introduced to account for the variability of the steel (or concrete or soil), the degree with which the structural model approximates real behaviour, and the likelihood of uunderperformance. For the steel frame, a value of g of 0.9 is used for the column and rafter members, while gtakes lower values in the design of connections.AISCDPFBAS Limit State Designs = 7 ‘The design requirement for the strength limit state is that the design strength or capacity is greater than or equal to the design action effect, that is S'S, aay ‘This requirement must be satisfied at each cross-section and at each connection throughout the frame. Of course, in satisfying Equation 1.1, several different load combinations must be considered. 1.2.3 Design for the Serviceability Limit State In design for serviceability, the designer must ensure that the structure behaves satisfactorily, ‘and can perform its intended function at service loads, The most important serviceability limit states to consider for a portal frame building are those of limiting excessive deflection and in some cases preventing excessive vibration. ‘The load combinations employed in design for the serviceability limit state are discussed in Section 2.5.2. Deflections are calculated by the usual methods of structural analysis, and guidance on these is given in Section 4.9, Vibrations of portal frame buildings, particularly in response to dynamic crane loadings, are not considered in this book, although crane loadings are considered in Chapter 9. While most of the design standards are devoted to calculating the capacities R, for the strength limit state, this does not indicate that the strength limit state is always more important than the serviceability limit state. Some portal frame designs may be governed by the limiting of deflections, and it is important fo check that a structure which possesses sufficient strength will perform satisfactorily at service loads. In some cases, it may be desirable to proportion the members to satisfy serviceability criteria first, and then to check that the structure possesses an adequate reserve at the strength limit state. 1.3. DESIGN EXAMPLE ‘The material presented in the chapters of this book is illustrated with a worked design example, Where appropriate, reference is made to code clauses, tables, figures and other information on the right hand side of the design calculations. The design brief is for a factory in.a wind Region B industrial estate with the following constraints: Building Size (Figure 1.4) Length = 72 m (frame centres) Width 25 m (column centres) Height 7.5 m (floor to centreline at knee) Frame (Figure 1.5) Steel portal = single span across 25 m width Spacing om Pitch e8 Introduction AISC DPFBAS Personnel doors Reller gates Door (850) 0.9m x 2.2m dem x 3, S e\g ° I Figure 1.4 Design Brief: Plan and Elevation (ee Elevation Typical Section Figure 1.5. Design Brief: Cross-Section‘Asc DPFRIOS Design Example 9 Shielding buidings 42 x 25 x Sen high nga hus 87 hs = 90 0 Is = 87( 545) = 87 Typicol upwind sector ‘Allotments in industrial estote Subject Building 725m, overalt Figure 1.6 Shielding Buildings in Design Example Floor Reinforeed concrete to carry 4.5 tonne forklift with unlimited passes ‘Subgrade CBR 5 Roof and Walls ‘Trimdek 0.42 BMT (Base Metal Thickness) sheeting Ventilator Full jength ventilator with 600 mm throat Doors 4xroller shutter doors each 4 m x 3.6m high 4xpersonnel doors each 0.9 m x 2.2 m high10 10. u. 13. 4 16. 17. 18, 20. nv Introduction AISC DPFBAS Soit Conditions Stiff clay with ¢, = 50 kPa Footings Bored piers or pad footings Shielding Buildings Refer to Figure 1.6 REFERENCES Broken Hill Proprietary (1998). Hot Rolled Structural Steel Products, BHP, Melbourne. Standards Australia (1998). AS#/00 Steel Structures, SA, Sydney. Canadian Standards Association (1978). CAN3-SI6.1-M78 Steel Structures for Buildings — Limit States Design, CSA, Rextale, Ontario. British Standards Institution (1990). 855950, Structural Use of Steel in Buildings, Part 1, Code of Practice for Design in Simple and Continuous Construction: Hot Rolled Sections, BSI, London. American Institute of Steel Construction (1986). Load and Resistance Factor Design ‘Specification for Structural Steel Buildings, AISC, Chicago. Standards New Zealand (1992). NZS3404 Steel Structures Standard, SNZ, Wellington, NZ. Standards Association of Australia (1989). 4S1170.2-1989 Part 2 Wind Loading Code (with amendments), SAA, Sydney. Davies, J.M, (1990). Inplane stability in portal frames, The Structural Engineer, 68(4), 141- 147, Engineering Systems Pty Ltd (1996). Micrastran Users Manual, Engineering Systems, Sydney. Integrated Technical Software Ply Ltd (1995). Spacegass Reference Manual, ITS Pty Ltd, Werribee, Victoria Woolcock, 8.7. end Kitipomchai, 8. (1987). Design of Portal Frame Buildings, AISC, Sydney. Australian Institute of Steel Construction (1985). Standardized Structural Connections, 3" edn, AISC, Sydney. Standards Association of Australia (1978). AS2/59-1978 SA4 Piling Code, SAA, Sydney. Standards Australia (1995). 452/59-1995 Piling — Design and Installation, SA, Sydney. Standards Association of Australia (1989). AS/170.1-1989 Part | Dead and Live Loads and Load Combinations, SAA, Syéney. Standards Australia/Standards New Zealand (1996). AS/VZS4600 Cold Formed Steel Structures, SA, Sydney, SNZ, Auckland. Standards Australia (1994). Concrete Structures, SA, Sydney. Standards Association of Australia (1981). ASI250-1981 SAA Steel Structures Code, SAA, Sydney. Kennedy, DJ.L. (1974). Limit states design - an innovation in design standards for steel structures, Canadian Journal of Civil Engineering, \(1), 1-13. Leicester, R.H., Pham, L. and Kleeman, P.W. (1983). Conversion to limit states design codes, Metal Structures Conference, Brisbane, May, 29-33. Pham, L., Bridge, R.Q. and Bradford, M.A. (1985). Calibration of the proposed limit states design rules for steel beams and columns, Civil Engineering Transactions, Institution of Engineers, Australia, CE27(3), 268-274.2 Loads 2.1 BACKGROUND ‘As part of the development of the limit state design approach for structures, the loading codes were drafted using a rational probabilistic basis. The relevant loading codes for limit state design appeared some time ago, being AS1170.1-1989 Part 1: Dead and Live Loads and Load Combinations [1] and AS1170.2-1989 Part 2: Wind Loads [2]. The wind code has had two amendments. Both loading standards will be used extensively throughout this book. The loads to be considered in the design of portal frame buildings are dead, live, wind and occasionally snow loads, and combinations of these. Live loads generally represent peak loads which have a 95% probability of not being exceeded over a 50 year return period, while for wind loads, different return periods are used for the strength and serviceability limit states. Snow loads are not considered in this book. Dead loads G, live loads Q and wind loads Ware discussed in Sections 2.2, 2.3 and 2.4 respectively. ‘The load combinations used to obtain the factored design loads for the strength and serviceability limit states have been determined on a probabilistic basis, and these are discussed in Section 2.5. Crane loads are treated in Chapter 9. 2.2 DEAD LOADS ‘The dead loads acting on a portal-framed industrial building arise from its weight including finishes, and from any other permanent construction or equipment. The dead load will vary during construction, but will remain constant thereafter, unless significant modifications are made to the strueture or its permanent equipment. As a guide for preliminary analysis, a dead load of 0.1 kPa can be allowed for the roof sheeting and purlins. ‘The weight of the rafter should be included, but the weight of roof bracing, cleats and connections is not usually considered as being significant. 2.3 Live LOADS ‘The live loads acting on the roof of a portal frame building arise mainly from maintenance loads where new or old roof sheeting may be stacked in concentrated areas. The roof live loads for cladding, purlins and rafters are specified in the loading code ASIL70.1, the roofs of industrial buildings being of the non-trafficable category. Roof cladding must be designed to support a concentrated load of 1.1 KN in any position, but this is usually taken account of by the sheeting manufacturer which nominates the maximum spans that will sustain this load. uR Loads AISC DPFBOS For purlins and rafters, the code provides for a distributed load of 0.25 kPa where the supported area A is less than or equal to 14 m®, the area A being the plan projection of the inclined roof surface area. For areas A less than 14 mé, the code specifies the distributed load wo to be Wo (2+o.2) KPa @1) ‘This formula is equivalent to a distributed load of 0.12 kPa plus a load of 1.8 KN distributed over a span of the member, and ensures that the minirmm load to be supported by short members such as purlin cantilevers and end wall fascia members is 1.8 kN. Presumably, such a load would cater for the case of a heavy worker standing on the edge of the roof or at the ‘edge of an opening, and lifting materials on to the roof. In addition to the distributed live load, the loading code also specifies that portal frame rafters be designed for a concentrated load of 4.5 KN at any point. Such a load is not critical for large roofs in high wind areas. It should be noted that the distributed live load given in Equation 2.1 need not be considered acting simultaneously with any wind load (see Section 2.5). AS1170.1 requires that the structure be designed to support either the distributed live load or the wind load, whichever produces the worse effect. Note that the distributed live load of 0.25 kPa is significantly less than the live load in the UX, Europe and North America where snow loads must be catered for. 2.4 WIND Loaps . 2.4.1 General ‘The wind loading specified in AS1170.2 is generally the major loading influence in the design of industrial buildings, even in low wind areas. It is therefore important to evaluate the wind loads carefully. Table 2.1 compares the wind speeds and the possible uplift pressures in various regions of Australia for Terrain Category 3. It can be seen that the wind pressures in Region C (cyclonic areas) are almost twice the Region A pressures. After some deduction has been made for the factored dead weight of the roof sheeting, purlins and rafters, the net uplift ‘on @ portal frame rafter in coastal northem Australia could be more than twice that in southem and inland Australia, ‘The wind code provides a simplified procedure for calculating wind loads. The simplified method is applicable to reasonably small rectangular buildings located on flat or generally undulating ground. A more detailed procedure covers almost all contingencies in the design of industrial buildings. Because of this, and because the simplified procedure can be overly conservative in many cases, the detailed procedure is recommended for the design of industrial buildings. The simplified procedure is not considered in this book.Isc DPEBICS Wind Loads Table 2.1 Comparison of Terrain Category 3 Wind Loads Region A | Region B | Region C Perth Brisbane | (cyclonic areas ‘Adelaide except Region D) Wind Loading Melbourne Darwin Parameter Canberra Sydney Basic Wind Speed 7, , m/s 50 60 70 Design Gust Wind Speed for Mca = 0-80, M, = 0.85, M,= 1.0 34.0 40.8 47.6 My=1.0 Gust Dynamic Wind Pressure q. (KPa) | 0.69 1.00 136 Typical Uplift Pressure p, 083 120 163 = (0.740.5)q, “ ” ° ‘Typical Dead Load of Sheeting, ?Parlins & Rafters p, (kPa) 015 O47 0.20 Design Uplift =p, -0.8p, (kPa) O71 1.06 1.47 Design Uplift STA Ee Po 5 2 Design Uplift for Region A 100 we o 13 ‘In the wind code, the basic wind speeds ¥, and V, are given for the strength (ultimate) and serviceability limit states respectively. These speeds are then converted into wind pressures for design’. The wind code also gives permissible stress design velocities /, which were intended for use in the design of purlin and girt systems to working stress procedures. However, purlin and girt designs are now undertaken in ‘accordance with limit state procedures, and permissible stress velocities /, will nt be used inthis book.4 Loads aiscprraas 2.4.2 Regional Wind Speeds The basic wind speeds V, and ¥, for the strength and serviceability limit states are clearly specified in the wind code for the four different wind speed regions throughout the country. ‘These are standardised for a building of height 10 metres in Terrain Category 2. The basic wind speeds are factored to culculate the design gust wind speeds as discussed in Section 24.5. 2.4.3 Wind Direction The basic wind speeds for the strength and serviceability limit states for some major population centres are given in the code for specific wind directions. The code allows for the basic wind speed to be adjusted for specific wind directions in areas where sufficient ‘meteorological information is available. Where sufficient information is not available, the code allows a reduction factor of 0.95 on the design wind speed for major framing elements in Regions B, C and D". Because the factor applies to wind speed, the reduction in pressures is about 10% which is significant. The reduction factor is used in the design example for determining not only the loads on portal frames, but also overall wind bracing forces. It should be emphasised that the reduction factor does nor apply to the wind loads on purlins and girts. 2.4.4 Terrain Category Most wind speed data have been recorded at airports at a height of 10 metres. The terrain near most airports is basically very similar, and is designated as Terrain Category 2. Because so much of this information is available, wind speeds at a height of 10 metres in Terrain Category 2 are taken as the basic or reference wind speeds V with height multipliers equal to unity. The terrain category factors given in the wind code lead to a wide variation of wind pressures as shown in Table 2.2, It is therefore important to select the appropriate terrain category carefully. The code uses four terrain. categories defined specifically in terms of roughness length. This allows for interpolation between the categories on a logarithmic basis. In selecting the terrain categories, due allowance for any future changes in terrain should be made, such as the development of neighbouring areas. For example, a factory in a new industrial estate may be more exposed in its first few years than in the remainder of its life. "Amendment No. 1 of ASI170.2 2] reintroduced a wind direction eduction factor onthe design wind speed for ‘major faming elements in Regions B, C and D (except for ¥, in Region B). It is worth noting that such a factor was first introduced in the 1983 edition of AS11702 with a value of 0.9. ‘The factor was changed to 0.95 when the 1989 edition wes published, but it applied to overall buildings and not to major framing elements. With Amendment No, 1, major framing elements are again included. ‘This appears to apply to the portal frames of industrial buildings. Some designers take advantage of this, while others are not aware of itor choose not to use itAisc Pras Wind Loads 15 Ifso, it would be reasonable to assume Terrain Category 3 for design purposes rather than Terrain Category 2 of 2'/, _ Table 2.2 Relative Wind Pressures for Different i Terrain Categories (h = 7.5m) Terrain Height Multiplier | Relative Category Meow Pressures 1 1.09 1.90 2 0.96 148 2h 0.88 124 3 0.79 1.00 2.4.5 Basic Wind Speeds ‘The design gust wind speed V, is obtained from the regional wind speed 7 (whether for the strength or serviceability limit states) using ¥, =VM ceo MM, (2.2) where Mzeqy is the terrain and height multiplier for a particular terrain category, M, is a shielding factor, M, is a topographic multiplier and M, is an importance multiplier. The values Of My cq ate specified clearly in AS1170.2 as functions of the terrain category (or roughness) and height z, The code permits interpolation for intermediate values of z and roughness. ‘The shielding multiplier M, accounts for the shielding effect of surrounding buildings of equal or greater height than the portal frame building under consideration. When the building spacing parameter D for the surrounding buildings is less than 1.5, the shielding factor M, drops down to 0.7, whereas M, is unity when D is greater than 12. Shielding cannot be disregarded if the most economical structure js to be achieved. “The topographic multiplier M, applies if the building is located in a local topographic zone, and may under exceptional circumstances result in a 50% increase in the design gust speed. The importance factor M, should be taken as 1.0 for an industrial building, unless the building has a post-disaster function or some other special purpose. 2.4.6 Calculation of Pressures ‘The free stream gust dynamic wind pressure q_ (KPa) is calculated from the design gust wind speed ¥, (m/s) by16 Loads AIsc DPF 9, = 0.6¥2 «10% 23) The wind pressure p, at height z for the relevant limit state is then calculated from the pressure coefficient C, for the surface by the expression Ps Onde 4) The extemal wind pressure coefficients are set out clearly in the code, and their determination is straightforward. However, Amendment No. 2 introduced some additional complexity with alternative external pressure coefficients for the roofs of industrial buildings, as mentioned in Section 2.4.7. The determination of internal pressure coefficients has traditionally caused some confusion amongst designers, and these are discussed in Section 248. 2.4.7 External Pressures Although more complex than coefficients in British and US wind codes, extemal pressure coefficients in AS1170.2 were relatively simple for rectangular industrial buildings until Amendment No.2 was issued in 1993. ‘This amendment introduced altemative sets of roof coefficients C, for cross winds on buildings with roof pitches leSs than 10° and for ongitudinat winds, such that designers must use -0.9 or -0.4 for a distance h from the windward edge; ~0.5 or 0 for the zone from to 2h; ~0.3 or +0.2 for the zone from 2h to 3s and ~0.2 or +0.3 beyond 3h. The first coefficient in each pair should be combined to form one set (-0.9, ~0.5, -0.3 and ~0,2), and the second coefficient to form the other set (-0.5, 0, +¥0.2 and +0.3). The set which gives the worst effect should be used. ‘The coefficients from ‘one sct stiould not be mixed with the other. For typical industrial buildings, this amendment results in two main cross wind options whereas there was one previously. These options are: ‘* Maximum uplift using coefficients: ~0.9, -0.5, -0.3, -0.2 © Minimum uplift using coefficients: 0.4, 0, +0.2, +0.3 For longitudinal winds, the altemative coeflicient approach introduces the option of a downwind frame having downwind external pressure on the roof. If this downwind pressure ‘combines with intemal suction, then the resulting combination can be more severe than the gravity load combination of 1.25G + 1.5Q. This outcome is surprising when it is considered that portal-framed buildings have been designed and built for decades without accounting for ‘such load combinations. If the maximum internal suction coefficient —0.65 is combined with downward roof pressures, then the comparison is even more severe. This situation could theoretically arise if there are roller doors open in the side walls at the windward end of the building in the -0.65 wall suction zone, and the rest of the building is closed. Previously, external suctions were counteracted by intemal suctions to some extent, and so these load combinations were not considered.Ise DPFBVG3 Wind Loads 7 In summary, while wind tunnel testing has undoubtedly revealed that downwind pressures can be exerted on the roofs of some buildings, these pressures are at odds with previous practice and intemational wind loading codes: Perhaps this is because the probability of a load combination comprising downward extemal pressure and internal suction is low enough compared with other load combinations not to warrant serious consideration of such a combination. 2.4.8 Internal Pressures ‘The internal pressure coefficients in AS1170.2 range from a positive coefficient of +0.7 to a suction coefficient of -0.65, as shown in Figure 2.1 Vy —{—n Wind > | 40.7 \ Wind > -0.65 D\ Kk Pressure Suction Figure 2.1 Maximum Internal Pressure Coefficients ‘The code permits calculation of the permeability ratio to determine intemal pressure coefficients C,,. The permeability ratio is the ratio of the opening area in the windward wall to the sum of the opening aress in the roof and other three walls, provided any opening in the roof (such as a ventilator) is in an extemal suction zone. However, this calculation becomes a matter of judgement because it is up to the designer to choose which of the doors and windows may be relied upon to remain closed under design winds. Jtcan be argued that the worst winds occur without warning, eg, during thunderstorms, and that the windows and doors may not be closed when the design winds occur. However, unless buildings have permanent openings, most are only open, on average, 10 hours per day and 5 days per week, which is only 30 percent of the time. Assuming that the worst winds are likely to occur with equal probability at any hour of the day, then the average recutrence interval should be 0,350 years which is 15 years. Although thunderstorms can occur at any time of the day, the probability of occurrence during a 24 hour period may not be uniform. Hence it would be prudent to assume an average recurrence interval of, say, 25 years instead of 15 years for the intemal pressures when designing to the strength limit state. Moreover, there is the statistical probability that the building will not have the worst combination of ‘windows and doors open and shut. The foregoing probabilistic approach to internal pressure18 Loads Aisc Pru does not have any basis in the wind code, and is provided here as background information only to assist designers in justifying internal coefficients which are less than the maximum in some cases. : AS1170 (E3.4.7) states that industrial and farm buildings can have permeabilities up to 0.5% of the wall area but the actual percentage can be difficult to quantify. A realistic assessment of leakage could be made by calculating the area of ribs and gaps at the wall/floor and wail/roof junctions, The uncertainty with this approach is in the width of the gap between the wall and roof sheeting and between the floor edge and wall sheeting, and whether the ribs have been sealed for bird proofing or other reasons. if one considers only the area of the ribs for say Trimdek roof sheeting, the area of ribs for a 50 m x 20 m building would be as tollows: For leeward and side walls: _ 0.05% 0.025 . + wall/floor: 599 (50+ 20420) = 0.56 mi wall/eaves: (as for wall/floot) = 0.56 m* For windward wall: wall/floor: 299% 2925, 59-031 m* wall/eaves: = Permeability ratio assuming no other openings 0314031 0564036 ~ °° Hence internal pressure coefficient C,; = #0.1 If one roller door is added on the windward face, say 4 m x 3.6 m= 14.4 mv area, then: 14440314031 056+056 permeability ratio 3.4 Hence intemal pressure coefficient C, ;=+0.7 It may thus be concluded that the effect of ribs will not be significant if there are major wall openings such as vehicle doors. Some designers prefer to use roof ventilators to reduce internal pressures. However, roof ventilators are quite expensive and their cost can outweigh the savings in structural steelwork and footings resulting fom reduced intemal pressures, Part of the problem is that the equivalent free area of a ventilator is only about 30% of the throat arca, As a result, ifa 50 m long industrial building has a ridge ventilator with a 600 mm throat for the full length of the ridge, the equivalent free area would be 50x0.6x30/100 = 9.0 m’. In this case, the permeability ratio would be (14.4 + 0.31 + 0.31)(0.56 + 0.56 + 9.0) = 148. The internal pressure coefficient C,, would then drop from +0.7 to +0.3. Another problem which arises, particularly in cyclonic areas, is the effect of flying debris on windows, and the failure of roller shutter doors because they bow under pressurealscDPFHS Wind Loads 19 and pull out of their guides. These problems can be overcome by providing cyclone shutters or security grilles over glass windows and by fitting wind locks to roller doors. There is some uncertainty, however, regarding the effectiveness of wind locks on rolfer shutters. Consideration should also be given in non-cyelonic areas to the ability of roller shutter guides to withstand wind forces, and to the possibility that the doors will blow out of their guides. In particular, it appears that roller shutter doors are often attached inadequately to their supports. 2.4.9 Area Reduction Factor The area reduction factor for extemal pressures allows basically for the fluctuating nature of these pressures, and the fact that the average pressure when the area is large is less than the coefficients indicate. ‘The area reduction factor applies to roof and side wail loads. It does not apply to internal pressures, or to windward and leeward wall loads. This means that for a portal frame under cross wind, only the rafter loads due to external pressures may be reduced. ‘Under longitudinal wind, both rafter and column loads due to external pressures may be reduced. If the area supported by the rafter or a column is greater than LOO mr’, the area reduction factor is 0.8. This factor is significant and cannot be ignored in the design if an ‘economical structure is to be achieved. 2.4.10 Local Pressure Factors ‘The code requires all wall and roof claddings, together with their ishmediate supporting members and fixings, to be designed for peak local pressures as shown in Figure 2.2. The local pressure factors of 1.5 and 2.0 apply to negative external pressures (suetions) whereas the factor of 1.25 applies to positive external pressures anywhere on the windward wall. Note that the local pressure factors do not apply to intemal pressures (positive or negative). 2.5 LOAD COMBINATIONS 2.5.1 Strength Limit State ‘The loading code AS1170.1 stipulates that to produce the most adverse effects’, the design loads for the strength limit states shall be the following combinations of dead load (G), live Toad (Q) and ultimate wind load (7): ‘Early working stress versions of the steel structures code did not specify load combinations, but they di permit 8 25% overstress when wind loads were present. ‘The limit on overstress was increased to 33% in the 1972 dition of the code, which was consistent with American practice at that time. However, the permissible stress proach to steel design had an inherent danger that if wind load and dead load act in opposite directions and are of similar magnitude, then the difference between the loads-is @ small value which is very sensitive 10 inaccoracies. This was illustrated in Reference (3).2» Loads AIsc PFO () 1.25G+159 . ©) 1256+, © 08g+1259 @) 08G+¥, i A separate load combination is also given if earthquake forces are to be considered. ‘The above load combinations are used for the instability of uplift limit state, except that the part of the dead load which resists the instability (G*) is separated from the total dead load. i Locol pressure foctors are rant Pe not opplicable at ridge where o=0 roof pitch < 10° a< 60" Height hy for © = 90° ‘ond for O = 0 when a 2 60° SS Wind © = 90° hy $ 25.0m Areaaxa: Local pressure foctor 1.5 on negotive pressures i Area a/2 x 0/2: Locol pressure factor 2.0 on negative pressures G Area 0/2 x 0/2: Locol pressure factor 1.25 on positive pressures on windward wall @ = he, 0.2b or 0.24, whichever is least Figure 2.2 Peak Local Pressures In an attempt to remedy this situation, the 1975 edition of the working stress code AS1250 [7] removed the 33% ‘overstress (or the 0.75 load factor) for cases where wind and dead load act in opposite directions. Unfortunately, this did litle to improve the potentially dangerous load combination because the resulting 33% increase in design load stil did not adequetely cater for small errors in the dead load of for underestimates of the wind load. ‘The problems of load combinations for permissible stress design 2s outlined above were overcome in the limit state loading code AS1170.2 [2] which appeared in 1989,AISC PFD Load Combinations 21 2.5.2. Serviceability Limit State ‘The loading code AS1170.1 includes load combinations for the serviceability limit state. The following combinations of dead load (G), live load (Q) and serviceability wind load (W,) are to be considered: @ % ®) 42 © Gr, @ G+ HQ where y, is the short-term load factor given in the code and taken as 0,7 for the roofs of industrial buildings. Strictly speaking, this means that in checking rafter deflections, only 0.7 times the live load need be considered. However, the deflection limits suggested in this book are only guidelines based on a survey of practising engincers [4]. In any case, the limit suggested for live load deflections applies to the full live load. ‘Therefore, there does not seem to be any point in considering a reduced live load for the serviceability limit state of a portal frame. 2.6 DESIGN EXAMPLE ~ LOADS 2.6.1 Dead Loads ASII70.4 Sheeting: Trimdek 4.3 ke/m? = 0.043 kPa Purlins: 220019 at 1200mm centres with 15% laps Aysaght [5] _ LAS x 5.68% 9.82107 12 Total we =0.043 + 0.053 =0.096 kPa say 0.1 kPa Hence sheeting and purlin load on rafter = 0.10x9 = 0.90 kN/m (along slope) Frame self-weight will be included under the gravity option (GRAV) in the computer analysis = 0.053 kPa In some buildings, an allowance for miscellancous dead loads such as bracing, roof exhaust systems, lighting and soffit linings or ceilings will be appropriate. 2.6.2. Live Loads AsiI70.1 Wo AG& +012) 0.13 kPa but not less than 0.25 KPa ASI/70.) C148.1.1 * Hence wo = 0.25 kPa Live load on rafter = 0.25x9 = 2.25 Nim (on plan projection)2 Loads aise DPFBIS As the computer program Microstran [6] does not have a load type with vertical load distributed on the plan projection of the rafter, it would be more accurate for steep-pitched roofs to convert the live load to a distributed load along the slope, In this case, the pitch is not steep and so the effect of pitch on live load is insignificant, ie, live load on rafter along slope = 2.25xcos3° = 2.25 KN/m. In addition, a concentrated load of 4.5 KN will be applied at the ridge. 2.6.3 Wind Loads ASII70.2 2.6.3.1 Basic WIND DATA Region B: Basic wind speeds: Ultimate ¥,,= 60 m/s ASII70.2 Table 3.2.3, Serviceability V,= 38 m/s AS1170.2 Table 3.2.3 Terrain Category 3 (industrial area) Column height: 7.5 m at intersection of rafter centreline Portal span: 25 m between column centres Roofpitch: 3° (see Figure 1.5) Eaves height assuming 310 UB rafter, 200 purlins 0310 = 154+ +0200= 7.85 m say 8.0m Ridge height = 8 04 ctan3* .655 m say 8.7m Average spacing of shielding buil Average height of shielding buil ings = 87 m s = 9m ‘Average breadth of shielding buildings = 42 m 87 Buildit i eter: D= =45 AS11702 C1 3.2.7 lng spacing parameter: D= ES a Shielding multiplier: Af, = 0.85 ‘AS1170.2 Table 3.2.5.4 © Cross Wind . h=80m Termain and height multiplier: Mg.) = 0.80 ASII70.2 Table 3.2.5.1 Shielding multiplier: M4, = 0.85 (AS1170.2 Table 3.2.7 Ultimate: V, =0.80x0.85x60 = 40.8 m/s ‘AS1170.2 C13.2.2 Gz =0.60x40.8%10" = 1.00 kPa AS1170.2 13.3atsc pPFwas Serviceability: ¥, = 080x0.85x38 =25.8 m/s G_ = 0.60%25.8°x10° = 0.40 kPa © Longitudinal Wind h=8.7m Terrain and height multiplier: Mig ,3) = 0.81 Shielding multiplier: Mf, = 0.85 Ultimate: FV, =081%0.85x60 = 41.3 m/s Gz = 0.60%A1.3%10° = 1.02 kPa Serviceability: V, =0.81%0.85x38 =26.2 m/s q, = 0.60%26.2*10" = 0.41 KPa 2.6,3,2. EXTERNAL WIND PRESSURES © Cross Wind (= ©) ‘Leeward wall: 8 gays 7082 < 05 Design Examples -Loads 23 ASLI70.2 C13.2.2 ASIII0.2 C13.3 AS1170.2 Table 3.2.5.1 AS1170.2 Table 3.2.7 ASHI70.2 C13.2.2 AS1170.2 13.3. ASHI70.2 C13.2.2 AS1170.2 C1 3.3 ASIITO.2 Fig. 3.3 AS1170.2 Table 3.4.3.1(A) ASI170.2 Table 3.4.3.1 (B) Two sets of C,. values for the roof are given in Amendment 2 of ASI170.2 Table 3.4.3.2(A), Therefore, adopt pressures shown in Figures 2.3(a) and (b). + Longitudinal Wind (0= 90) (see Figure 2.4) Tributary area for rafter under cross wind = 25% Hence reduction factor for rafters = 0.8 AS1170.2 Table 3.4.3.2(A) 225 mi?4 Loads aise DPBS 8000 ro 7|_4500_ 3500 -09 FO 4 8000 =03 |-o2 +07 -0.5 (0) Moximum Roof Uplift Coefficients go00 4500, 3500 Tol | 8000 x4 402 Iso 40.7 ~0.5 4 (b) Minimum Roof Uplift Coefficients Figure 2.3 External Pressure Coefficients under Cross Wind ‘Tributary area for rafter and columns under longitudinal wind = (2x75 + 25)x9 = 360 m* Hence reduction factor for columns under longitudinal wind = 0.8 ASIJ70.2 Table 3.4.4 2.6.3.3 INTERNAL WIND PRESSURES * Cross Wind To calculate the intemal pressure coefficients C,,, it is necessary to determine the equivalent free arca of the ventilator. Manufacturers give coefficients in theit brochures for converting the throat width into an equivalent free throat width. In this case, take the coefficient as 0.35, so that the equivalent free area is 0.35x0.6x72 = 15.1 m*.AISC DPFBIOS Design Examples-Loads 28 Permeability ratio for worst intemal pressure under cross wind = 243642409422 9 151 2 Henee C,, = 05+|— > (0.605) =+0.52 ASI170,2 Table 3.4.7 For the worst internal suction under cross wind when dominant openings are on the leeward wall, use the value of C,,; for leeward external wall surface Ga = -0.50 AS1170.2 Table 3.4.3.1(B) ‘Note that roof ventilators can be expensive and the saving in cost due to reduced internal pressures will be offset to some extent by the cost of the ventilators. ASI170.2 Table 3.4.7 # Longitudinal Wind Permeability ratio for worst intemal pressure (end wall door open, others closed) 4x36 =", 7088 ASIIT0.2 Table 3.4.7 For internal suction under longitudinal wind, the worst case would be with the side doors ‘open and the end doors closed. Hence should strictly speaking adopt the worst side wall pressure coefficient C,, = ~0.65 but this will mean that the combination of extemal downward pressure and maximum internal suction will now govern the portal frame di ‘whereas this was not so prior to Amendment 2 of AS1170.2. For the purpose of this design example, adopt C= the code, and C,,; 2. 4 PEAK LOCAL PRESSURES ‘The peak local pressure roof plan is show in Figure 2.5. 7m or @=0.2b=0.2x72.5= 14.5 m or @=0.2d=0.2x26= 5.2m whichever is least. Hence a= 5.2m (see Figure 2.5) 2.6.4 Load Cases For Portal Frames. Primary Load Cases: LCi: DL of 0.90 kN/m + frame self weight 0.3 for portal frame design although not strictly in accordance with 0.65 for purlin and girt design.26 Loads _.,8700 8700 8700 0.7 =0.25 Windword Leoword wall woll 8 @ 2000 = 7200 72500 opprox overall (a) Roof and End Wall Pressure Coefficients Maximum Upii 8700 8700 8700 PT yo [02 . 5 Roof ik * IF Windward Leeward wall wall 8 @ 9000 = 7200 8 @ 9000 = 7200 72500 opprox overall (b) Roof and End Wall Pressure Coefficients Minimum Upit 8700, 8700, 8700 im os! -o2 Cy ~ | qe Th jf selon | | || ~02 6: 8700 “8700 “8700 (c) Woll Pressure Coefficients Figure 24 External Pressure Coefficients under Longitudinal Wind AISCDPFDASAISCDPFRAS Design Examples -Loads 27 ‘ _ jl} gis ‘ | General area Ws S als | General oreo I) gs L Jt} <1 AL 5.2m x 5.2m zone 2.6m x 2.6m zone Local pressure foctor Local pressure factor 15 2.0 Approximately 72.5 overall length Figure 2.5 Peak Local Pressure Roof Plan LC2: LL of 2.25 kNin + 4.5 KN at ridge (Note that both the dead and live loads are usually input as negative loads in Microstran because they usually act in the negative direction of the local or global member axes.) LC3: Cross Wind Maximum Uplift (CW1 — see Figure 2.6): g, = 1.00 kPa Wind direction reduction factor for major framing elements such as portal frames = 0.95? ASH70.2 C1 3.2.3 Area reduetion factor for roof only = 0.8 ASHI70.2 C13.4.4 UDL (windward column) = 0.95°x0.7x1.00%9.0 UDL (leeward column) 0.95*«0.5x1.00x9.0= 4. UDL (roof, 0t08m,C,,=-0.9) = 0.95*0.8x0.9%1.00x9.0 = 5.85 kNim UDL (roof, 8 mto 16m, C,,= 0.5) = 0,95%«0.8x0.5%1.00%9.0 UDL (roof, 16 m to 24 m, C,,,= -0.3) = 0.95%«0.8x0.3%1.00x9.0 UDL (roof, 24 m to 25 m; C,, = —0.2), adopt 1.95 kN/m 8000 4500. 3500 3.25 Il3.25, 5.85 1.95 5.69 4.06 Figure 2.6 Cross Wind Frame Loads - Maximum Uplift (kN/m)28 Loads Isc DrFwios Note that the 8 m roof wind loading zones strictly speaking should be measured from the caves edge of the roof which is approximately 0.5 m upwind of the intersection point between the rafter and column. For simplicity, the extent of frame loading has been taken between the frame intersection points. LC4: Cross Wind Minimum Uplift (CW2 - see Figure 2.7): , = 1.00 kPa Wind direction reduction factor = 0.95 Area reduction factor = 0.80 g000___4500_ 3509 Teo TTT ao oO 5.69 4,06 Figure 2.7 Gross Wind Frame Loads - Minimum Uplift CW2 (kN/m) UDL (columns) as for LC3 UDL (roof, 0 t0 8 m, C,.=—0.4 UDL (roof, 8 m to 16m, C,,=0)=0 UDL (roof, 16 m to 24 m, C,,¢ = 40.2) = -1.30 kNim UDL (oof, 24 m to 25 m, C,.™= +0.3) adopt ~1.30 Nim (The negative sign indicates the loads acts downwards in the negative direction of the local member y axis according to Microstran Toad input conventions.) 60 kN/m LCS: Longitudinal Wind First Internal Frame (LW1) Area reduction factor for roof and walls = 0.8 As h is approximately equal to the frame spacing of 9 m in this case, take / = 9 m to simplify calculation of the panel loads. (Jn other cases where h is quite different from the bay spacing, the frame UDL’s should be determined by simple statics, ic. assuming the purlins are simply supported beams with the frames as supports.) UDL (rafters) = 0.95? x0. an(22e08 }a.02%9.0 = 4,64 Nin 1.02x9.0 =3.81 kN/m UDL (columns) = 0.95 x0.8x| (eS + 3) LC6: Longitudinal Wind with 0.39, Exteral Roof Pressure and 0.2g, Wall Suction (LW2)scorns Design Examples - Loads 29 ‘Area reduction factor for roof and walls = 0.8 UDL (rafters) = 0,95x0.8x0.3x1.02x9.0 = 1.99 KN/An UDL (columns) = 0.957x0.8x0,2%1.02x9.0 = 1.32 kN/m, LCT: intemal Pressure Under Cross Wind (IPCW): C, ,= +0.52 ‘Area reduction factor does not apply to internal pressures UDL (rafters and columns) = 0.95"x0.52«1.00x9.0= 4.21 KN/m LC8: _ Intemal Pressure Under Longitudinal Wind (IPLW): C,,;= +0.1 UDL (rafters and columns) = 0.95°x0. 1x1.02x9.0 = 0.83 KN/m. LC9: Internal Suction Under Cross Wind (ISCW): C,,;= - 0.5 = -0.96 xLC7 LC10: Note that load cases LC9 and LC10 are not included in the computer analysis and combinations containing LC9 and LC10 are obtained by factoring LC7 and LC8. Table 2.3 Factors Used in Load Combinations Load Factors cases | cr | nc2 | 103 | uca | xcs | 208 | ter | xcs LC2 } 1.25] 15 Lea | 08 10 1.0 Lc22 | 08 1.0 1.0 1023 | 1.25 1.0 -0.96 ‘ Lez | 08 10 10 Leas | 1.25 1.0 3.0 The following load combinations LC20 to LC25 may be obtained by factoring the primary load cases (see Table 2.3). These load combinations have been analysed by both30 Loads ‘ASC DPHIOS clastic and plastic methods. In each case, use was made of the computer software Microstran- 3D Structural Analysis Program [6]. The computer output is listed in Appendix II for the second order method of elastic analysis. LC20: 1.25DL + LSLL = 1.25LCI + 1.5LC2 LC21: O.8DL + CWI (maximum uplift) + IPCW = 0.8LCI + LC3 + LC7 LC22: 0.8DL + CW2 (minimum uplif) + IPCW = 0.BLCI + LC4 + LCT LC23: 1.25DL + CW2 (minimum uplift) + ISCW = 1.25LC1 + LC4- 0.96LC7 LC24: O:8DL + LP! (maximum uplift) + IPLW = 0.8LCI + LCS + LC8 LC2S: L.25DL + LW2 (maximum downward) + ISLW = 1.25LCI + LC6- 3.0LC8 Note that LC25 combines the dead load and downward roof pressures which can occur on downwind frames under longitudinal wind with internal suction, As discussed in Section 2.4.7, such combinations should be viewed sceptically if they govem the design. In this design example, the internal suction coefficient has been arbitrarily reduced for this combination from -0.65 to ~0.3. This is not in strict accordance with the current wind loading code. Designers need to make their own judgements on the validity of such combinations. 2.7 REFERENCES 1, Standards Association of Australia (1989). AS/170.1-1989 Part 1 Dead and Live Loading Code, SAA, Sydney. 2. Standards Association of Australia (1989). AS//70,2-1989 Part 2 Wind Loading Code, SAA, Sydney. 3. Weotcock, $:T. and Kitipornchai, S. (1987). Design of Portal Frame Buildings, ASC, Sydney. 4, Weoleock, 8.1. and Kitipomehai, S. (1986). Deflection limits for portal frames, Steel Consiruction, AISC, 20(3), 2-10. 5. Lysaght (1999). Zeds and Cees Purlin and Girt Systems, BHP Building Products. 6. Engineering Systems Pty Ltd (1996). Microstran Users Manual, Engineering Systems, Sydney. 7. Standards Association of Australia (1981). AS1250-1981 SAA Steel Structures Code, SAA, Sydney.3 Purlins & Girts 3.1 GENERAL Purlins and girts are the immediate supporting members for roof and wall sheeting respectively. They act principally as beams, but also perform as struts and as compression braces in restraining rafters and columns laterally against bucking. In some buildings, purlins and girts also act as axial members to transfer end wall wind loads to the braced bays, while in smaller buildings they may even act as the struts of the triangulated roof bracing system. Purlins and girts are now almost universally zed (Z) and channel (C) section members, cold-formed from zinc coated 450 MPa steel of 1.5, 1.9, 24 and 3.0 mm thickness. Stee! sections cold formed from-1.2 mm 500 MPa steel are also available, as well as those of 1.0 am thickness cold-formed from 550 MPa steel. Timber purlins arc still used occasionally, especially in certain corrosive environments. Figure 3.1 shows a typical Z section purlin and girt arrangement, Strength is not the only consideration when designing purlins. Purlin spacing must be chosen to suit the type of roof sheeting and ceiling system if any. The use of translucent fibreglass roof sheeting will also restrict the purlin spacing. Some suspended ceiling systems require a maximum purlin spacing of 1200 mm, and some riggers and roofers object to purlin spacings in excess of 1200 mm, Purlin deflections must also be controlled. Because of the thin walls of the cold-formed sections, their design and analysis are more complex and the limit state cold-formed steel structures code AS/NZS4600 [1) must be used in lieu of AS4100 [2]. Until 1996, this presented added difficulty as the prevailing cold- formed stee! structures code AS1538 [3] was written in a working stress format, Fortunately, puslin and girt manufacturers provide comprehensive design capacity tables [4,5] and it is not Typical assembly using Zed sections and Hook-Lok bridging. Figure 3.1 Dypical Lysaght Purlin and Girt Details 3132 Purlins & Girts AIS DPFBIOS usually necessary to refer to AS/NZS4600 unless the designer wishes to take advantage of the ‘Refactor method’ described briefly in Section 3.4.2. 3.2 ROOF AND WALL SHEETING ‘The first step in purlin design is to consider the spanning capacity of the proposed roof sheeting. Sheeting manufacturers provide data on minimum roof pitch, and on allowable intemal and end sheeting spans. Their brochures give maximum spans for average conditions in non-cyclonic areas, as well as allowable wind pressures for various spans in cyclonic and non-cyelonic areas. It is important to remember that the maximum spans for roof sheeting arc determined not only from wind load considerations, but also from live load requirements, including the 1.1 KN concentrated load of ASI170.1 {6]. Therefore, these maximum spans should not be exceeded for roof sheeting, even if the allowable wind pressure table for the sheeting profile indicates that the sheeting has the capacity to do so. The peak local pressure zones around the perimeter of the roof govem the purlin spacing in these areas, and the purlin spacing chosen in the end bays is usually adopted for the rest of the toof. A typical purlin and girt layout is given in Appendix I. In some cases, extra purlins are used in the end bays to halve the purlin spacing used in other bays that are only partially subjected to peak local pressures. Because the extra purlins are simply supported, however, it may be necessary to use the heaviest purlin thickness for strength, and even then the deflections may be excessive. As a result, the use of extra single span purlins in end bays and fewer purlins in interior bays is not generally worthwhile, In larger buildings, it can be advantageous to extend intermediate purlins in the end zones over two or three bays, thereby providing the continuity needed. In cyclonic areas, special design criteria are required because of the cyclical loading and the possibility of fatigue failure. Cyclone or load spreading washers may be necessary. 3.3. FRAME SPACING There are many variables to be considered in optimising the frame spacing for a particular building. Apart from the portal frames and purlins, it is necessary to consider the length of roof bracing struts and tension ties, and of course the footings. Although the size of the portal frames obviously increases with frame spacing, the weight per unit area of the portal frame building decreases. Theoretically, the price of the steelwork per tonne should also decrease because the sections are heavier, and there will be less labour per tonne, By contrast, the cost per square metre of purlins and girts will increase with frame spacing, but in steps comesponding to the depth or thickness increases. Because of the limited range of purlin sizes and the consequent sudden jumps in purlin capacity and cost as sizes increase, itis difficult to optimise frame spacing. It will obviously be cheaper, for a given purlin depth, to incresse the frame spacing to ensure that the purlin system is working to its capacity. However, this is rarely possible as site or other tequirements usually fix the overall length of the building, and so the choice of frame spacing, is limited.AISc DPFBS Frame Spacing 33 ‘An extensive study in the United Kingdom [7] examined hundreds of different options {for an industrial building 90 m long. Portal frame spacings of 4.5, 6.0, 7.5 and 9.0 m were considered. The results indicated that the 7.5 m spacing was the most economical for portal frame spans in excess of 20 m. For spans less than 20 m, the 4.5 m spacing was most ‘economical. However, basic wind speeds, purlin types afd unit costs are different in the United Kingdom, and it is uncertain whether the same conclusions apply in Australi 3.4 PURLIN STRENGTHS 3.4.1 Manufacturers’ Brochures ‘The limit state design capacity tables produced by manufacturers such as Stramit [4] and Lysaght [5] cover simply supported spans, double spans, double lapped spans, lapped continuous spans and increased thickness end spans in lapped continuous systems. The Stramit tables also include reduced end spans in continuous lapped systems. “The Lysaght tables for lapped systems are based on standard laps of 600, 900, 1200 and 1800 mm, the laps being a minimum of 10 percent of the span. This is conservative in most cases because the standard laps are usually greater than 10 percent. The Stramit tables are based on laps of 15, percent of the span. Both the Stramit and Lysaght load capacities take account of section yielding, flexural- torsional buckling, distortional buckling, combined bending and web shear, and bolt eapacity. ‘The method used to calculate flexural-torsional buckling capacities is the so-called rational clastic buckling analysis in Clause 3.3.3.2(b) of AS/NZS4600. This clause contains the design ‘equations from AS1538 converted to limit states format [8,9]. Z-section putlins have their principal axes inclined to the plane of the roof sheeting, and therefore rely on the lateral and twist-rotational restraint from the roof sheeting to prevent Jateral displacement of the purlin under inward loads. Under outward loads, the bottom flange of the Z-section purlin is in compression over most of each span, and it is necessary to provide bridging between purlins as shown in Figure 3.1 to reduce the effective lengths to control flexural-torsional buckling, During erection, Z-purlins are obviously unrestrained, To limit excessive deflections prior to fixing roof sheeting, it is an advantage to orient the top flange pointing up the slope as this results in a smaller inclination of the weak principal axis to the vertical. Indeed, experience has shown that a single row of bridging is advantageous during erection, even if not required for strength in service. Accordingly, Lysaght rocommends at least one row of bridging in every span, and that unbridged lengths be restricted to less than 20 times the section depth, Stramit also limits the unbridged length to 20 times the section depth, or 4000 mm, whichever is less. It should be noted that these bridging requirements are those recommended by the manufacturers, and although bracing is treated in detail in AS/NZS4600, there is no specific limit on the unbridged lengths. End spans are usually the critical area for purlin design. This is because end spans not only have higher bending moments and higher deflections for a given uniform load, but also higher loads because external suctions including local pressure effects are highest at the windward end under longitudinal winds.34 Purlins & Girts Asc DPFBS Because of these problems, it is common to use a section with a heavier wall thickness for end bay purlins. For example, if Z15015 sections are used in internal bays, then Z15019 and Z15024 sections would be used in the end bays. Significant reductions in external suction coefficients on downwind surfaces in the more recent versions of ASI170.2 [10] have accentuated the differgnce in loads and bending moments between end span and internal span purlins. It is therefore advantageous for economical design to consider: ‘+ increased wall thicknesses in end span purlins, o reduced end spans by closer portal frame spacing, or + extra bridging in end spans, provided this increases the design strength of the purlins. ‘The Stramit brochure provides design capacity tables for lapped systems with end spans reduced by 20%, and this shows that significant economies can be achieved. The Stramit and Lysaght load capacity tables do not cater for the non-uniformity of the base load ‘on the purlin system, although the Stramit brochure has provided methods to cater for peak local pressure zones to allow the designer to convert the non-uniform pressures to equivalent my distributed loads. Such methods are also considered in Section 3.9 of this book, With the extemal suction coefficients reducing under longitudinal wind from -0.9 to ~0.5 to ~0.3 to -0.2, the end span will generally have a higher base load than the next span. The moments in the end span would then be higher than if the load on the putiin system were uniform. 3.4.2, R-Factor Method AS/NZS4600 [1] provides a method in Clause 3.3.3.4 known as the ‘R-factor method’ for the flexural design of members having one flange through-fastened to sheeting, This method is relatively simple, and can be used by designers in lieu of using manufacturers’ design capacity tables. However, it requires a lapped continuous beam analysis for purlins in various roof zones to determine the design bending moments and shears Mf” and 7”. The reactions also need 10 be calculated to check bolt capacities. The laps can be easily modelled by inserting members with double the second moment of area into the beam system. Peak local pressure loading and reductions in the downwind extemal pressure coefficients under longitudinal wind can be easily accounted for. ‘The member bending capacity is taken in this method as toMy = byRZ ef, Ga) where Z, is the effective clastic section modulus (calculated using the effective widths of the compression and bending elements of the purlin section as given in AS/NZS4600), f, is the yield strength of the purlin, ¢, is the capacity reduction taken as 0.90 and R is the reduction factor. The cold-formed stcel structures code AS/NZS4600 presents values of reduction factor to be used under both uplift loading and downward loading, Its use is restricted to roof and wall systems which comply with a number of limitations which can be met in standard designs. Depending on the arrangement of lapped and unlapped spans and bridging, the R factor vaties from 1.0 to 0.60.Isc DPFB«3 R-Factor Method 35 3.5 DEFLECTIONS In its current brochure, Lysaght [5] does not give recommended deflection limits for purlins and girts. However, Lysaght did provide recommendations in previous brochures on the basis of extensive practical experience. These recommendations were as follows: (a) Under maximum of total design load: span/120; (b) Under combined dead and live load: span/150; and (c) Under live load alone: span/180. These limits applied in @ working stress design environment and as such, the appropriate regional basic design wind speed for calculating deflections was the same as the strength design wind speed. This wind speed corresponds to ¥, in the current wind loading code AS1I70.2 [10] and is greater than the current serviceability wind speed F,. In view of the lack of current recommendations from Lysaght, and as Stramit recommends a maximum deflection limit of span/150, the following deflection limits are now proposed. © Under dead load alone: Span/360 ‘© Under live load alone: Span/180 ‘© Under serviceability wind load alone: Span/I50 ‘The limit of span/150 for serviceability wind load alone may be more stringent than before but some account has been taken of the reduction in wind speed from , to V,. Both the Stramit and Lysaght tables present’ distributed loads corresponding to a span/i50 deflection for the serviceability limit state, These tables can be factored readily to give a span/360 or a span/180 deflection. 3.6 AXIAL LoaDS ‘As mentioned in Section 3.1, purlins may be required to act as compression members to transfer end wall wind loads to the nodes of the triangulated roof bracing system with the assistance of roof sheeting acting as a diaphragm or a deep beam. Under this condition, the purlins are therefore subjected to combined actions (bending and compression). Lysaght presents formulae in their design brochure for the axial capacity of purlins * based partly on.any feserve of flexural strength. If there is no reserve of flexural strength, the axial capacity is taken to be zero. ‘The Stramit brochure does not present specific formulae, but directs the user to the provisions of the combined axial load and bending Clause 3.5 of ASINZS4600. 3.7 PURLIN CLEATS Standard purlin and girt cleats have ably stood the test of time and are generally used without analysis or design. The standard sizes for lapped purlins which require only two bolis are 75x8 flats for up to 250 purlins and girts, and 75x12 flats for 300 and 350 purlins and girts. ‘When purlins are unlapped, four holes in the cleat are required and the cleat width increases from 75 mm to 130 mm.36 Purlins & Girts AISC DPF BOS Purlin cleats are subjected not only to axial loads, but also to bending moments. The bending moments result from the component of the weight of the roof sheeting in its own plane and from the restraint provided by the sheeting to prevent lateral buckling. In the case of Z profiles, there are also bending moments from lateral forces due to the inclination of the principal axes to the plane of the roof, i ‘When the gap between the purlin and rafter (or girt and column) is much greater than the nominal 10 mm gap, thicker cleats or angle cleats such as 75x75%5 equal angles can be used for strength. This situation can occur when rafters are horizontal and the purlin cleats are graded in height to provide the roof pitch. Angle cleats also provide greater robustness during. transport and erection. The maximum overall height of an 8mm thick cleat should be 250 mm, while a 12 mm thick cleat should be no more than 450 mm high. The height at which designers specify an angle cleat in preference to a rolled steel flat is fairly arbitrary, but a practical requirement is that cleats higher than 450 mm should be angles. One yardstick for robustness is that girt cleats should not yield when stood on by a heavy worker. This would equate to a 1.1 KN load applied to the tip of the cleat with a 1.5 load factor to allow for dynamic effects as the worker climbs the steelwork, 3.8 PURLIN BOLTS The standard bolt is an M12-4.6/8 which comes either with loose washers or with a flanged hhead and nut (je. with washers integral with the bolt head and nut), Although the latter bolts are about 2.5 times as expensive as the standard M12 bolts, it appears that the extra expense is considered worthwhile by riggers because of the speed of handling only two components rather than four. Occasionally under high shear, M12-8.8/S bolts should be used. Surprisingly, 8.8/S bolts with flanged heads and nuts are only about 20% to 25% more expensive than their 4.6/8 counterparts. M16 bolts are required for Z/C300 and Z/C350 sections. It should be remembered that washers under both the head and nut are essential if the bolts with flanged heads and nuts are not used. This is because the standard punched holes in purlins are 18mm high by 22mm long and the standard hole diameter in cleats is 18 mm. ‘These hole sizes are too big for M12 bolt heads end nuts even though the height of the hole through lapped purlins is less than 18 mm diameter because of the lapping. By comparison, the width across the flats of an M12 bolt is only 18 mm and the washer diameter is 24 mm. Z. and C-sections with depths of 300 mm and 350 mm require M16 bolts, for which the holes in the cleats can be 22 mm diameter. 3.9 EQUIVALENT UDL’s FOR PEAK PRESSURE In order to use the purlin capacity tables, it is necessary to convert the loads due to peak pressures over part of the span to uniformly distributed loads (UDL’s) over the whole span. Lysaght gives guidance on converting point loads to equivalent UDL’s but not peak pressures as noted earlier. Stramit presents a table of factors to cater for peak loads as partial load blocks both at the end of the span and in the middle of the span (even though Stramit'sAISC DPBS Equivalent UDL's for Peak Pressure 37 explanatory diagram indicates a partial load block at the end of the span). A. simple alternative approach is to take a weighted average of the exira peak load block as shown in Figure 3.2 and add this to the base uniformly distributed load. ‘The multipliers of 1.3 and 2.0 for the end span in Figure 3.2 were determined from computer analysis of a four-span continuous lapped beam with different lengths and locations of load blocks on the end span, ‘They are the maximum factors by whith the average load over the full end span needs to be increased to give equivalent maximum moments (in the mid-span region of the end span) to those for the actual load block. ‘The moments at the first internal support are not critical. ‘The multipliers are conservative in achieving equivalent UDL’s over the end span only. However thé Lysaght and Stramit tables arc derived for uniform loads over all spans and not just the end span. For an equivalent UDL over all spans, the corresponding maximum multipliers are approximately 1.6 and 2.5. This is not surprising when it is realised that a iniform load over a full end span would need to be factored by 1.24 to achieve the same maximum end span moment in the mid-span region as that for a uniform load over all four spans. ‘As the Stramit approach does not differentiate between mid-span and end-of-span load blocks, it tends to be conservative for end-of-span load blocks. ‘The end result is that the simple approach of using multipliers of 1.3 and 2.0 as proposed in this book is not as ‘unconservative compared with the Stramit approach as it might appear. For example, for a Kyy value of 1.5 and a g value of 0.5, the Stramit factor on the nominal base load is 1.47. (K,yis, the ratio of the total UDL in the peak local pressure zone to the nominal base UDL in the absence of local pressures, and g is the ratio of the length of the load block to the span length.) By comparison, the simple approach in this book gives corresponding factors on the nominal pone ef 1/2 L/2 wes 3M sw wes 2 sy Figure 3.2. Equivalent Uniformly Distributed Loads for Peak Pressure Load Blocks38 Purlins & Girts AIC DPFILOS base load of 1.33 (= 1.0+0.5x0.5x1.3) for an end-of-span load block and 1.50 (= 1.0 +0.5x0.5%2.0) for a mid-span load block. As the corresponding accurate factors are 1.40 and 1.47, the approach in this book is unconservative by 5% (= 1.40/1.33) for the end-of-span load block, and conservative by 2% (= 1.50/1.47) for the mid-span load block. If K,, equals 2, the comresponding percentages are 9% and 3%. - i 3.10 DESIGN EXAMPLE — PURLINS 3.10.1 Member Capacity Brochures Brochures produced by Stramit [4] and Lysaght [5] present design capacity tables for purlins and girts subjected to outward and inward loading. ‘The case of outward loading tends to ‘govern the design in the majority of cases, since it produces predominantly compression in the unrestrained flange. Because of this, itis logical to select a purlin section from the outward design capacity table, and then to check this section using the inward design capacity tables. Deflections also need to be checked Outward loading is produced by internal wind pressure combined with external suctions, reduced by 0.8 times the dead toad in accordance with the load combination in AS1170.1 [6]. On the other hand, inward loading is produced by wind load and dead load combinations, and by dead and live load combinations, again as in AS1170.1. The inward wind plus dead load case consists of intemal suctions and extemal pressures plus 1.25 times the dead load, while the dead and live load combination is 1.25 times the dead load plus 1.5 times the gravity live load. 3.10.2 Outward Loading — Cross Wind As noted in Chapter 2, ctoss wind loading produces extemal suctions over a number of different zones. The maximum iitemal pressure coefficient is +0.52 (see Section 2.6.3.3). © Edge Zone 0 to 2600 mm from Eaves The cross wind coefficients for this region are shown in Figure 3.3. They include @ peak local pressure zone at midspan which produces the worst effect. These loads apply to both internal and end spans. The equivalent UDL for this local pressure is w, = 2wx/l, where wa is the additional peak load distributed over a length x. ‘The wind pressures are calculated using g, = 1.00 kPa for cross wind (see Section 2.6.3.1). Using Load Combination (4) in Section 2.5.1, total equivalent UDL for spacing » (in metres) is we {(00s BES28 +032) 100-08%03] x5 = 1.865 kNimAsc DPBS Design Example -Purlins 39 2600. Local pressure load block on end or = -09 internal spans NI] % Cp = -0.9 Cp = +052 9000 Figure 3.3 Cross Wind Coefficients 0 to 2600 mm from Eaves «Edge Zone 2600 mu to 5200 mm from Eaves ‘The cross wind coefficients for this region are shown in Figure 3.4. From Figure 3.2, x= 5200 mm, 0 2x/L = 2x5200/9000 = 1.16 > 1, and so w, = w. Hence the total equivalent UDL for spacing s is {(0.9+0.45 +0.52)x1.00~0.8%0.1}xs=1.79s kN/m 5200 Local pressure load block on end or internal spans Neo] Cp = 0.45 +0.52 9000 Figure 3.4 Cross Wind Coefficients 2600 mm to 5200 mn from Eaves + Zone 5200 mun to 8000 mm from Eaves ‘The peak pressure zone starts at the end wall and is therefore at the end of the purlin. Hence take the equivalent UDL w, as 1.3wx/L. There are two cross wind peak pressure coefficients for this case, viz. 0.9 over a length of x= 2600 mm or —0.45 over a length40 Purlins & Girts aise DPF ofx=5200 mm, Both produce the same effect according to the equivalent UDL formula in Figure 3.2, and the cross wind coefficients for this zone are shown in Figure 3.5. Hence the total equivalent UDL for spacing s is wee {(oo-t2x0ases2, os2)s 100-0201} * Zone 8000 mm from Eaves to Ridge ‘The cross wind coefficients assumed for this zone are as shown in Figure 3.5 except that the external pressure coefficient is —0.5, and so —0.5 should replace ~0.9, and ~0.25 should replace ~ 0.45. Hence total equivalent UDL for spacing s is {(osets22#52 059), 100-0203} =L13s KN/m 8700 {Cp = -0.9 I Llc, = +052 9000 Figure 3.5. Cross Wind Coefficients 5200 mm to 8000 mm from Eaves 3.10.3 Outward Loading - Longitudinal Wind ‘+ Edge Zone 0 to 2600 mm from Eaves ‘The external pressure coefficient is -0.9 over 8.7 m of the span and —0.5 over the remaining 0,3 m of the span while the internal pressure coefficient is +0. instead of +0.52, ‘These coefficients are shown in Figure 3.6. Although the longitudinal wind pressures are calculated using g, = 1.02, the cross wind case in Figure 3.3 is clearly critical.ASC DPFBINS Design Example Purlins 4 8700 2600 9000 Figure 3.6 Longitudinal Wind Coefficients 0 to 2600 mm from Eaves «Edge Zone 2600 mm to 5200 mm from Eaves Cross wind is again clearly critical, as the intemal pressure coefficient under longitudinal wind is +0.1 and under cross wind is +0.52. © Zone 5200 mm from Eaves to Ridge ‘The effect of the local peak pressure coefficient of ~0.9 acting over 2600 mm is identical lo that of -0.45 acting over $200 mm according fo the equivalent UDL formula in Figure 3.2. The longitudinal wind coefficients for this zone are shown in Figure 3.7. Conservatively adopting the ~0.9 extemal pressure coefficient over the whole span, the {otal equivalent UDL for spacing s is wea {{o9s 122258882 soa} 02-08% 04}xe=1 295 kKN/m 3.10.4 Purlin Selection for Outward Loading Maximum peak pressure on sheeting = (0.9x2 + 0.52)x1.00 = 2.32 kPa For Trimdek 0.42 BMT using Lysaght’s limit state sheeting brochure: ‘Maximum end span = 1300 mm Maximum internal span = 1900 mm © Maximum unstiffened overhang = 150 mm42 Purlins & Girts aise DrFBIS 8700 +~ Alternative peok pressure zone 9000 Figure 3.7 Longitudinal Wind Coefficients 5200 mm from Eaves to Ridge In order to reduce purlin twists and deformations, it is recommended by Stramit [4] and Lysaght (5) that the maximum bridging spacing be 20 times the purlin depth. For a purlin depth of 200 mm, the maximum spacing is 4000 mm and so two rows of bridging are recommended in all 9 m spans. Both Stramit and Lysaght present capacities for thicker purlins in the end spans. ‘The Stramit capacities for various purlin configurations including thicker end span purlins are reproduced in Table 3.1 while the corresponding Lysaght capacities are given in Table 3.2. It can be seen from these tables that the Stramit and Lysaght design capacities are somewhat different, with the differences presumably being attributable to the different lap lengths, For the purposes of the remainder of this design example, the Stramit system with its longer lap lengths (15% laps) will be adopted, + Edge Zone 0 10 2600 mm from Eaves Spacing required for the Stramit 2200-19 purlin system for flexure alone: 110m where 2,05 kN/m is the tabulated Stramit load shown in Table 3.1. Spacing required for the Stramit 2200-24 purlin system for flexure alone: 2.13 47m where 2.73 kN/m is the tabulated Stramit load shown in Table 3.1.tse DPFIVES Design Example—Purlins 43, Hence ADOPT the Stramit Z200-19 purlin system at 1100 mnt maximum centres © Edge Zone 2600 mm to 5200 mm from Eaves Spacing required for the Stramit 2200-19 purlin system for flexure alone: I Spacing required for the Stramit 2200-24 purlin system for flexure alone: 2.73 1D Hence ADOPT the Stramit 2200-19 purlin system at 1100 mm maximum centres 153m © Zone 5200 mm to 8000 mm from Eaves ‘Spacing required for the Stramit 2200-19 purlin system for flexure alone: Spacing required for the Stramit 2200-24 purlin system for flexure alone: 273 =28 m1 satay 763m Hence ADOPT the Stramit 2200-19 purlin system at 1200 mm maximum centres Table 3.1 Stramit Capacities for 5 or More Lapped Continuous 9 m Spans (2 Rows of Bridging) Sesion Tm | OM | tn | Spas 7200-15 450 1.36 1.36 0.94 2200-19115 | 5.74/4.50 | 1.37 137 1.19 2200-19 5.74 2.05 2.06 1.27 7200-24/19 | 7.215.74 | 2.07 2.07 1.56 2200-24 721 2.73 2.91 1.68 2250-19 6.50 2.62 2.62 2.06 2250-24/19 | 8.17/6.50 | 2.63 2.63 287 2250-24 8.17 3.65 3.73 2.76 Bold capacities require Grade 8.8 purlin bolts.44 Purlins & Girts AIsC DPFBAS Table 3.2 Lysaght Capacities for Four Lapped Continuous 9 m Spans (2 Rows of Bridging) (220015 4.44 1.16 1.16 0.93 Z20015/24 | 4.44/7.15 1.22 1.22 1,60 220019 5.68 1.77 1.77 1.26 220024 715 2.44 2.57 1.68 (225019 643 2.24 2.24 2.04 Z25019/24 | 6.43/8.10 231 2.31 2.70 225024 8.10 | 3.26 3.29 277 Bold capacities require Grade 8.8 purl bolts, *¢ Zone 8000 mm from Eaves to Ridge Spacing required for the Stramit 2200-19 purlin system for flexure alone: 113 Could adopt the Stramit 2200-19 purlin system at 1800 mm maximum centres but first check deflections and the possible use of translucent sheeting, and consider the ease of erection given that some riggers and roofers prefer 1200 mm maximum centres. + Preliminary Arrangement Based on the outward loading design capacities, try the Stramit Z200-19 purlin system with two rows of bridging for all spans and 1350 mm laps with the following maximum spacings: 1100 mm: 0 to 5000 mm from eaves 1200 mm: 5000 mm from eaves to 8000 mm from eaves 1800 mm: 8000 mm from eaves to ridge 1300 mm: maximum end sheeting spans at eaves and ridge for foot traffic 3.10.5 Check Inward Loading + Zone 0 t0 5200 mm from Eaves (1100 mm spacing) ‘Combining the external pressure coefficient of 40.3 with the intemal suction coefficient of -0.5 under cross wind:AISC DPFIOS Design Example -Purlins 45 (0.3 + 0.5)x1.00 +1.250.1}x1-1=1.02 KNim < 2.05kN/m OK and under longitudinal wind, combining the worst extemal pressure coefficient of +0.3 with the worst internal suction coefficient of —0.65: we = {(0.3-+0.65)1.0241.25x0.1}x1.1=1.20 kNim < 2.05kNim OK © Zone 5200 mm to 8000 mm from Eaves (1200 mm spacing) ‘The same assumptions in previous section will be adopted for both cross wind and longitudinal wind except that the external pressure coofficient under cross wind will be +0.2 rather than +0.3: w* = {(0.5+0.2)x1.0041.25x0.1}x1.2=0.99 KNim < 2.05KN/im = OK and under longitudinal wind: ((0.3+-0.65)x1.02 +1.25%0.1}x1.2=131 kNim < 2.05kNim OK + Zone 8600 mm from Eaves to Ridge (1800 mm spacing) ‘Adopt the same assumptions as in previous section Hence under cross wind: ww = {(0.5-+-0.2)x1.0041.25x0.1}x 1.8 =149 kNim < 2.05kNim OK and under longitudinal wind: = {(0.3+0.65)x 1.024 1.25%0.1}x1.8=1.97 kN/m < 2.05KNim OK * Check Dead Plus Live Load For worst case of 1800 mm purlin spacing: wt = (5.74x9,82x107 +4,3x9,82 x10” x1.8)x1.25+0.25x1.8x1.5 = 0,132x1,25 +0.45x1.5 = 0.84 kN/m < 2.05kNin OK where the self weight a Z200-19 is 5.74 kg/m and the self weight of 0.42 BMT Triméek is 4.3 kg/m?. As the tributary area of 9x1.8 = 16.2 m? is greater than 14 m’, the live load. according to AS1170.1 [6] is 0.25 kPa. 3.10.6 Purlin Deflections ‘The dead load deflection of a Stramit 2200-19 purlin system spaced at 1800 mm centres using the dead load calculated above:46 Purlins & Girts ATSC DPFB/OS =m CM og 1475 300 where 1.27 KN/m is the tabulated Stramit load shown in Table 3.1 to give a span/150 deflection and 0.13 kN/m is the dead load calculated in the previous section. The live load deflection of Stramit Z200-19 purlins spaced at 1800 mm centres: 9000 — =213 127° 150 me span, PRO 423 180 To check the maximum deflection under wind load alone, a designer could adjust the maximum wind load combination for the strength limit state by eliminating the dead load ‘component. However the wind load is quite dominant in this case and so the combined wind and dead load UDL will be adopted as the wind load alone. The maximum wind uplift plus ead load UDL is approximately equal to the strength capacity of 2.05 kN/m, Converting this from an ultimate to a serviceability wind load by applying a factor of (38/60), the serviceability wind load alone is 20se( 22) = 0.82 Nim < 1.27kKNim OK where 1.27 KN/m is the tabulated Stramit load shown in Table 3.1 to give a span/150 deflection. 3.10.7 Purlin Summary The purlin system and spacing arrangement suggested in Section 3.10.4 is also satisfactory under inward loading and its deflections are acceptable. However, as translucent sheeting will be used and the recommended maximum purlin spacing is 1500 mm for Alsynite 3050, the maximum purlin spacing needs to be restricted to 1500 mm. ‘The final system and spacings adopted for this design example match the spacings used in the previous edition of this book which limited the spacing to 1200 mm. The adopted purlin system is as follows: Stramit 2200-19 system with 15% laps Two rows of bridging M12 4.6/8 bolts Purlin arrangement from eaves: 5 at 1000 mm centres Sat 1200 mm centres 2 at 800 mm centres (with top purlin 300 mm from ridge to suit ridge ventilator with 600 mm throat) This compares with 2200-20 in the end spans and 2200-16 in the intemal spans adopted in the previous edition [11] in accordance with AS1538 [3] and the working stress design purlin and girt capacities.AISCDPFBIOD Design Example~Purlins 47 3.10.8 R-Factor Method ‘AS/NZS4600 [1] permits a simple alternative method for calculating the member capacities of purlins and girts under certain conditions. The so-called R-factor (or reduction factor) method js potentially applicable in this case with the use of 0.42 mm BMT roof sheeting, purlin laps greater than 13% and other conditions satisfied. : In this edition, it is not proposed to undertake a purlin and girt design for the whole building by the R-factor method but to investigate the capacity of one purlin run. For the zone located 5200 mm to 8000 mm from the eaves and under cross wind Joading, the purlin spacing is 1200'mm and the base UDL is derived from a -0.9 external pressure coefficient and a +0.52 internal pressure coefficient. The peak local pressure zone has a coefficient of 0.5x(-0.9)=-0.45 and is 5.2 m long from the end support as shown in Figute 3.5. These loads are applied in combination with 0.8 times the dead load to an eight- span continuous beam with the 1350 mm laps which straddle each intemal support simulated by doubling the second moment of area J,. The maximum bending moments f° in the end are 12.1 kNm in the mid-span region and 18.0 kNm at the first internal support as shown. in Figure 3.8a, The maximum moment in unlapped Z20015 members in the intemal spans is 8.5kNm. 124 Figure 3.81 Bending Moments for R-factor. Method for LW 5200 mm from Eaves to Ridge 101 Figure 3.8 Shears for Refactor Method for LW 5200 mm from Eaves to Ridge48 Purlins & Girts AISCDPFBOS © Member Moment Capacity Using an R-factor of 0.95 comesponding to two rows of bridging, the member capacities gM, for Z200-19 and Z200-15 sections are obtained from Equation 3.1 using the minimum ‘élues tabulated by Stramit [4] and g equal to 0,9 as follows: For 2200-15: gM, = 0.9x 0.95% 23.0x10? x 450=8.85 kNm For 2200-19: gM, = 0.9x0.95x33.8%10? x 450 = 13.0 kNm For moment capacity alone, 2200-19 in the end span and Z200-15 in the intemal spans are both adequate as follows: For Z200-15: M* For Z200-19; M* .5kNm < 8.85 kNm 2.1 KNm < 13.0kNm_ * Combined Bending and Shear A check must be made for combined bending and shear. Note that strictly speaking it would be necessary to re-analyse the purlin system to account for the thinner Z se the intemal spans as this would result in slightly different bending moments. For combined bending and shear, the relevant member actions are as follows. At the end of the lap in the ond span, the moment is 5.9 KNm and the shear is 7.9 KN. In the second span at the end of the lap near the first intemal support, the moment is 8.5 kNm and the coincident shear is 6.0 kN. For 220019: Shear capacity: d, = 203-2x(5+1.9)=189.2 mm 2x10* 5.34 1.6 > 1.41Sx J J2x10" *5:34 _ 68, 99.6 > 1.415; T LAS x, 750 68.9 Hence BV, = 0.9 x{ 220% EEE) gg (0.905% 210° x 5.34 mo 4, ” 189.2 a, 1892 1 19 =31.5kN Combined bending and shear: (iz) i) 9 GM, =0.95%33.8%10° x 450 =14.4 kNm Note that ¢, equals 0.95 for section capacity, not 0.90, Table 1.6 AS/NZS#600ASC DPBS Design Example Purlins 49 [At the end of the lap in the end span: s9¥ (19V 22) (2) =0.17+0.06=0.23 < 1.0 (2) (8) +006 0.23 < 1.0 OK oR £2001 Shear capacity: d, = 203-2x(5+1.5)=190 mm fy — 199 _ 167>68.9 as before te 15 Hence = 0.9x{ 2.905% Eke 0.905 x 2x 10% x 5.34% 1.5" BV, = 0.9: (oe } ogy See ) =15.5KN Combined bending and shear: OM, =0.95%23.0%10° x450= 9.83 kNm Table 1.6 AS/INZS4600 At the end of lap in first internal span near first internal support: 85) ,( £2) .0564015=071 < 10 OK 9.83) “155. ‘The Stramit Z20019/Z20015 system is therefore adequate for combined bending and shear as well as for maximum moment alone. There is ample reserve of combined bending and shear strength and sufficient reserve of bending strength to preclude the need for re- analysis of the continuous beam for the Z20019/Z20015 combination. The 720019/Z20015 system is lighter than the Z20019 system obtained by using the Stramit tables. 3.11 DESIGN EXAMPLE - GIRTS 3.11.1 Side Wall Girts Inward Pressure Coefficients: +0.7 extemal pressure (CW) 05 0.5 40,52 0.65 40.1 intemal suction (CW) external suction (CW) internal pressure (CW) ‘external suction (LW) internal pressure (LW) Clearly the cross wind case is more critical than the longitudinal wind case because of the much higher internal pressure.50 Purlins & Girts AIsC DPEHOS © Outward Loading ‘The assumed pressure coefficients for cross wind loading including the local pressure zone are shown in Figure 3.9. Equivalent UDL for cross wind loading with spacing, s, is 21s kim (sx t2xogsesz., . s2}+1.00%8 = 8700 Alternative peck local pressure zones 0.25 Cp = -05 Cp = +01 9000 Figure 3.9 Cross Wind Coefficients for Outward Loading on Side Wall Girts © Girt Selection For a Z200-24/19 system, spacing required for flexure alone: 2.05 s =1.69 m 121 Hence try the Stramit Z200-19 system at 1700 mm maximum centres © Inward Loading with 1700 mm Spacing UDL= (0.7 + 0.5)x1.00x1.7= 2.04 kN/m < 2.05kNim OK © Summary Adopt the Stramit Z200-19 girt system at 1700 inm maximum centres with 1350 laps and two rows of bridging on all spans.ASC DPBS Design Example Girts 51 3.11.2 End Wall Girts with Span of 6250 mm Inward Pressure Coefficients: +0.7_ extemal pressure (LW) 0,65. internal suction (LW) Outward Pressure Coefficients; -0.65 ektemal suction (CW) 40,52. internal pressure (CW) 0.25. extemal suction (LW) [a/b = 72/25 = 2.88] 40,1 intemal pressure (LW) © Outward Loading Clearly cross wind will govern the design and the relevant coefficients arc shown in Figure 3.10, Referring to Figure 3.2, the total equivalent UDL with peak pressure zone under cross wind for spacing, s, is = (0051123052852 .052)x1.00%5 =151siNin Alternative peok local pressure Figure 3.10 Cross Wind Coefficients for Outward Loading on End Wall Girts + Girt Selection ‘To match the side wall girt spacing, try s= 1.7m Outward loading = 1.51x1.7 = 2.57 KN/m For Z150 girts (whose Stramit capacities are listed in Table 3.3 and Lysaght capacities are listed in Table 3.4 using linear interpolation), the recommended maximum bridging52 Purlins & Girts Isc oP FBS spacing of 20D = 3040 mm. This length is quite close to half of the span, so one row of bridging may be justified. ‘Try Stramit Z200-15 systom with one row of bridging at 1700 mm centres Capacity = 2.77 kN/m > 2.57 kN/m OK i © Inward Loading with 1700 mn Spacing UDL = (0.7 + 0.5)x1.02x1.7 = 2.34 KNim < 2.771kNim = OK Table 3.3 Stramit Capacities for 3 o 4 Lapped Continuous 6.25 m Spans Outwards Inwards Section | Mass N/m kN/m_| Deflection tem | teow | 2Rows | tow | Spav/is0 zis0-10 | 243 | 099 | 118 | 1s | o78 z150-12 | 290} 126 | 149 | 149 | 098 zso-1s | 359] 163 | 196 | 196 | 126 z15019 | 451 | 227 | 277 | 274 | 167 z1so24 | 567 | 314 | 389 | 376 ] 215, z200-15 | 450 | 277 | 277 | 297 | 247 Table 3.4 Lysaght Capacities for Lapped Continuous 6.25 m Spans Outwards Inwards Section | Mass Nim kNin_| Deflection 4s | rrow | 2Rows | 1Row | Spav150 zisolz | 284] 133 | 149 | 133 1.05 zisos | 3.54] 173 | 196 | 1.96 1.34 zis019 | 4.46 | 235 | 276 | 271 1.79 Z18024 | 5.62 | 3.26 | 388 | 3.77 234 Z20015 | 4.44 | 265 | 265°] 265 2.58aise DPFEM3 Design Example — Girts 33 + Summary For end wall girts, adopt Stramit 2200-15 girt system at 1700 mm centres with 1000 mm laps. Use one row of bridging in all spans. 3.12 REFERENCES 1, Standards Australia/Standards New Zealand (1996). AS/NZS4600-1996 Cold Formed Steel Structures Code, SA, Sydney, SNZ, Auckland. 2, Standards Australia (1998). AS4/00-1998 Steel Structures, SA, Sydney 5, Standards Association of Australia (1988). AS/538-1988 SAA Cold-Formed Steel Structures Code, SAA, Syéney. 4. Stramit (1999). Siramit Purtins and Girts, Stramit Metal Building Products. 5. Lysaght (1999). Zeds and Cees Purlin and Girt Systems, BHP Building Products. 6 Standards Association of Australia (1989). AS/170.1-1989 Part 1 Dead and Live Loading Code, SAA, Sydney. 7. Horridge, JF. and Morris, LJ. (1986). Single-storey buildings cost considerations, Proceedings, Pacific Structural Steel Conference, New Zealand Heavy Engineering Research Association, August, 265-285. Standards Australia/Standards New Zealand (1998). AS/NZS4600-1996 Supplement 1: 1998 Cold-Formed Structures - Commentary, SA, Sydney, SNZ, Auckland, 1. Hancock, G.J. (1998). Design of Cold-Formed Steel Structures. 3" edn, AISC, Sydney. 10, Standards Association of Australia (1989). AS/170,2-1989 Part 2 Wind Loading Code, SAA, Sydney. Li. Wooleock, 8.7, Kitipomchai, 8. and Bradford, M.A. (1993). Limit State Design of Portal Frame Buildings, 2 edn, AISC, Sydney.54 AISC DPFBAS4 Frame Design 4.1 FRAME DESIGN BY ELASTIC ANALYSIS Traditionally, portal frame analysis and design in Australia has been elastic rather than plastic because of the non-uniform, asymmetric nature of the wind load. Although AS4100 {1] is a Jimit state code with section and member capacities based on the plastic moment of resistance, the main method in the code for determining the forces and bending moments in a frame is still elastic analysis, However, plastic analysis may in some cases lead to more economical sructures, and this is considered in Chapter 8 In the Australian wind code AS1170.2 (2], coefficients for external suction decrease in steps starting from -0.9 at the windward edge to -0.5 to -0.3 ‘to -0.2, or altematively from ~04 to 0, +0.2 and +0.3. This non-uniform pressure can be handled easily by an elastic analysis using a plane frame computer program. In fact, it would be extremely difficult to take advantage of the reduction in pressure and achieve an economical structure without recourse to a plane frame computer program. In the design of rafters and columns in portal frames, the selection of the member sizes may be governed by the ultimate or strength limit state, or by limiting deflections in the serviceability limit state, For the strength limit state, the design axial and bending capacities @N- and gMyy respectively are obtained through a consideration of flexural and flexural- torsional buckling respectively. To obtain an economical rafter design, it is important to ensure that the design bending sirength is as close as possible to the section capacity ¢Vf=,, which for many sections will be the plastic moment capacity Sf. This capacity is usually achieved by the use of adequate restraints such as fly braces to restrain the inside rafter and column flanges laterally when in compression. Of course, there are some cases where deflections govern the design, and these are discussed in Section 4.9 of this chapter. 4.2 " COMPUTER ANALYSIS 4.2.1 Load Cases For the computer analysis, itis best to use load cases which are complete in themselves. For example, intemal pressure should be a load case by itself, and not combined with an external pressure case. The loads on columns and rafters should not be separated. Recommended load ‘eases for a computer analysis are as follows: Dead Load (DL) Live Load (LL) Cross Wind Maximum Uplift (CW1) (external only) Cross Wind Minimum Uplift (CW2) (external only) 3556 Frame Design AISC DPBS Longitudinal Wind on First Internal Frame (LW1) (external only) Longitudinal Wind on Downwind Frame (LW2) (external only) Intermal Pressure under Cross Wind (IPCW) Intemal Pressure under Longitudinal Wind (IPLW) Extra load cases may be necessary for non-symmetrical buildings, for buildings where the cross wind terrain category is different on one side from the other, and for buildings where it may be an advantage to consider different wind speeds in different directions. Cross wind Toad combinations with internal suction are not often critical, but designers should check such combinations nevertheless. It is possible that the hogging moment at the downwind knee joint will be worse under dead load, cross wind and intemal suction (1.25DL + CW + IS) than under dead load plus live load (1.25DL + 1.5LL). This particularly affects the downwind column as its unrestrained inside flange will be in compression. ‘The internal suction case (IS) can be obtained simply by factoring the internal pressure load case by an appropriate negative number. ‘The recommended load combinations for a computer analysis are: L2SDL + 1.SLL O.8DL + CW! (maximum uplift) + IPCW ODL + CH2 (minimum uplif) + 1PCW L25DL + CW2 (minimum uplift) + ISCW O.8DL + LW (maximum uplifi) + IPLW L.2SDL + LW2 (minimum uplif) + ISLW Note that the loading code AS1170.1 [3] states that it is not necessary to consider live Joad and wind load acting simultaneously. There is some doubt about the validity of LC25 as discussed in Sections 2.4.7 and 2.6.5. The trial section properties used in the first computer run will not affect “the distribution of bending moments, provided that the column and rafter second moments of area are in the same proportion as those finally adopted. Some computer programs allow for shear deformations, although the effect is not significant. To account for shear deformations in Microstran, the web area, which can be taken as the overall depth D times the web thickness 44, must be input. 4.2.2. Methods of Analysis AS4100 permits a number of types of analysis consisting of first and second order elastic analysis, first and second order plastic analysis and advanced structural analysis. First order plastic analysis is considered in Chapter 8, while first and second order elastic analysis is treated in this chapter. First order elastic analysis assumes the frame remains elastic and that its deflections are so small that secondary effects resulting from the deflections (second order effects) are negligible, First order analysis is generally carried out using plane frame analysis computer programs. Despite the basic assumption of first order analysis, second order effects are not negligible. Second order effects are essentially P-4 effects which arise from the sway A of the frame, or P-Seffects which arise from the deflections 5 of individual members from the‘aise DPFBANS Computer Analysis 7 straight lines joining the members’ ends. AS4100 requires that the bending moments calculated by first order analysis be modified for second order effects using moment amplification factors. ‘The use of moment amplification factors can be avoided by using second order elastic analysis. Second order analysis is now widely adopted by designers as suitable programs are commercially available, and it is easier and more accurate to obtain elastic second order moments directly than to amplify first order moments. Second order elastic analysis is used as the first preference in this book with variations for first order analysis also being given ‘where appropriate, It should be noted that second order analysis should only be performed for Joad combinations and not for individual load cases. 4.2.3 Moment Amplification for First Order Elastic Analysis ‘AS4100 requires a rational analysis of non-rectangular sway frames to determine the frame clastic buckling load factor 2... ‘The first order bending moments in the columns and rafters are then amplified’ using the amplification factor 4; given by (4.1) ‘The factor 4, can be determined by commercially available clastic critical load computer packages. However, as these seem to go hand in hand with second order elastic analysis programs, there is little point in determining 2, in this way when direct second order analysis which avoids the use of A, is available, It should be noted that for pinned base portals, the approach used by these packages does not take advantage of the nominal base restraint” allowed in AS1250 [4] and therefore should be conservative. For designers without access to such computer packages, simple approximate expressions for determining 4, for pinned and fixed base portal frames may be found in Reference [5]. These expressions ignore the stiffening effect of any haunches and the nominal base restraint’’ allowed ini AS1250 and therefore should be conservative. For pinned base frames: 3EL, A 42 £(NTK, +0.3N7E,. a2) “in AS1250, moment amplification was effectively applied in the combined stress rules where the amplification factor U(I-f/0.6Fqq) was used to increase the in-plane bending stresses. To determine F, and Fug in the ‘combined stresses equation, the designer was requied to calculate the in-plane effective length of the columns. In the absence of any betiet technique, it was customary to regard the portal frames es rectangular frames with zero axial loads in the beams or rafters and use the Gy and. Gy factor approach in Appendix E of AS1250. However, such an approach was of doubiful validity because rafters are inclined and exrry axial loads. Nominal base restraint was represented by a G value of 10 fo pinned base in AS1250 wien using the Gy and Gy factor approach for determining effective lengths58 Frame Design AIC OPFRS «For fixed base frames: 5£(10+R) (43) SNe + 2RN he i z in which Tel, Root 4.4 Thc 4) and Eis Young’s modulus, NZ isthe axial force in the column, ;,_ is the axial force in the rafter, Jz isthe second moment of area of the columa, T, _ isthe second moment of area of the rafter, hie is the height to the eaves, and £, is the length of rafter between the centre of the colunm and apex Once the first order moments are amplified, the combined actions section (Section 8 of AS4100) applics. Member moment capacities are calculated using actual lengths of rafters and columns when determining the axial capacity N, as required by Clause 8.4.2.2 of AS4100 taking an effective length factor k, of 1.0. Clause 8.4.2.2 of AS4100 also requires the rafters and columns to be checked under axial load alone using the effective lengths Z, determined from the frame elastic buckling load factor 4, as discussed in Sections 4.3.3 and 4.4.2. The effective length of a rafter or column can be determined from ET, L,=#% (45) AN where N" is the design axial force in the rafter or column and I, is the respective second moment of area about the x axis. . 4.3.1.1 SIMPLIFIED PROCEDURE AS4100 uses @ semi-empirical equation to relate the nominal bending capacity Mg, to the elastic buckling moment M, and the section strength Mz, which for Universal and Welded Beams and Columns can be taken as Zeyh. This philosophy uses a set of semi-empirical equations to relate the member strength to the plastic moment and the clastic flexural- torsional buckling moment,Asc Pros Rafters 59 Clause 5.6.1.1 of AS4100 expresses the nominal member bending capacity Mi a Moy = On isM ox (4.6) where @q is a moment modification factor to account for the non-uniform distribution of major axis bending moment, and a, is a slendemess reduction factor which depends on Max and the elastic buckling moment of a simply supported beam under uniform moment Mp. The code gives comprehensive values of 2 which would be met in practice. ‘The conservative option of taking oy, equal to unity is also permitted. The slendemess reduction factor is expressed in Clause 5.6.1.1 of the code as vo (“) tel <10 47 Me Me where Mog may be taken as either (i) My which is the elastic buckling moment for @ beam with 2 uniform bending distribution and with ends fully restrained against lateral translation and twist rotation but unrestrained against minor axis rotation; or (ii) a value determined from an accurate elastic buckling analysis. The elastic buckling moment M, may be determined from the accurate expression [1] given in Clause 5.6.1.1 as wel, M, == JEI,GI, (4.8) Lv” P GI 48) where Ly is the effective length, and £J,, GJ and Ely are the flexural bending rigidity, the torsional rigidity and the warping rigidity respectively. Values of the section properties J,, J and Ip are given in the BHP Section Properties Handbook (6) and in AISC’s Design Capacity ‘Tables for Structural Steel [7]. The use of Equation 4.8 requires the effective length Lp, and the determination of this is discussed in subsequent sections. 43.1.2 ALTERNATIVE PROCEDURE Clause 5.6.4 of AS4100 allows the designer to use the results of an elastic buckling analysis, although in most cases this is not practical for design offices and is really a research tool. If an elastic buckling analysis is to be used, then the clastic buckling momnent Mog, which allows for the moment gradient, restraint conditions and height of loading, is determined either from ‘a computer program or from solutions given in the literature [8,9]. Having obtained Mop, the value of Mga to be used in Equation 4.7 is caloulated from a9) where values of ay are obtained either from the code or from an elastic buckling analysis such thatAlsc Dera (4.10) The moment Me; is the elastic buckling moment corresponding to Myp for the same ‘beam segment with the same bendijg moment distribution, but with + shear centre loading, ends fully restrained against lateral translation and twist rotation, and + ends unrestrained against minor axis rotation. ‘The moment Moy is the critical uniform bending moment M, given by Equation 4.8 with Ze taken as the laterally unsupported length L. In the event that the whole rafter is designed as a tapered member fabricated by diagonally cutting, rotating and welding the web, an accurate elastic buckling analysis must be used. This also applies to the haunched segment of a conventional rafter. The values of ‘Moy-and Mos for tapered raiters may be found in Reference {10]. 4.3.2 Effective Length and Moment Modification Factors for Bending Capacity 4.3.2.1 GENERAL If the simplified design procedure in Clause 5.6.1.1 of AS4100 (incorporating Equations 4.6, 4.7 and 4.8 above) is used, then the effective length Ze of the rafter must be determined in accordance with Clause 5.6.3, The effective length depends on the spacing and stiffness of the purlins and fly braces, and the degree of twist and fateral rotational restraint as follows: © Whether the connection between the purlins and rafter is rigid, semi: pinned. * The flexural rigidity of the purlins, in that AS4100 classifies purlins qualitatively as flexible or stiff. No numerical yardstick is given. + The load height in that AS4100 allows, for example, for the destabilising effect of loads applied at or above the shear centre in a beam subjected to downward loads, + Whether the inside or outside flange is the critical flange. For a portal frame, the compression flange is the critical flange as explained in Clause C5.5 of the AS4100, Commentary [11]. © The degree of lateral rotational restraint provided at the ends of a segment by adjoining segments. gid oF 4.3.2.2 ToP FLANGE IN COMPRESSION Under gravity loads, the top flange is mostly in compression, except near the knees. Purlins provide lateral restraint to the top flange, but full twist restraint to the rafter from the purlins cannot be relied upon because standard oversized 22 mmx18 mm holes are generally used in purlins with only M12 bolts. Although this means the holes in the purlins are effectivelyAISC DPFBIS Rafters 6 slotted, the bolts are tightened and so the purlin to rafter connection using a standard purlin Cleat and two bolts can be regarded as a partial twist restraint connection in terms of Figure $.4.2.1(b) in AS4100. Fortunately, the code permits partial twist restraint at the critical flange {Gn association with lateral restraint) to be classified as full restraint of the cross-section. ‘Therefore for each segment between purlins when the top flange is in compression, both ends are fully restrained (FF) and the fwist restraint factor ky is 1.0. ‘Although gravity loads are applied through the purlins at the top flange, the load height factor ky of 1.4 in Table 5.6.3(2) in AS4100 does not apply because the load is not free to move sideways as the member buckles. In other words, the load is applied at a point of lateral restraint and ky should be taken as 1.0. ‘The degree of lateral rotational restraint provided at each end of the segment by adjoining segments depends on whether the adjoining segments arc fully restrained laterally or not, as described in Clause 5.4.3.4 of AS4100. (A fully restrained segment in accordance with Clause 5.3.2 is essentially one with M, not less than M, which means its @cr, value is greater than unity.) The code permits full lateral rotational end restraint or none. No intermediate option is provided. While segments between purlins under downward loading are short and are likely to be fully restrained laterally, full restraint in accordance with Clause ‘5.3.2 cannot be guaranteed, It follows that lateral rotational restraint should strictly speaking be disregarded. There is, however, a high degree of lateral rotational restraint which would allow & to be taken safely as 0.85. In summary, the effective length L- is given by kkk, L as 1, = 10x 1.0% 0858, = 0.855, @.ly) Because the spacing between purlins is short in comparison with the length of the rafter, the moment médification factor 0p, should usually be taken as 1.0. 4.3.2.3 BOTTOM FLANGE IN COMPRESSION © With Fly Bracing under Uplift Under uplift, most of the bottom flange of a portal frame rafter is in compression, In such cases, the rafter is attached to the purlins at the tension flange level, and the compression flange of the rafter is unrestrained, In order to achieve increased member capacity, it is customary to restrain the bottom flange of the rafter laterally by providing fly bracing using small angle section members joining the bottom flange to the purlins. With the bottom flange in compression, AS4100 classifies a fly brace restraint as a full ot partia! cross-sectional restraint depending on whether the purlins are flexible or stiff. No numerical criterion is given for assessing the flexibility or stiffness of purlins. ‘Therefore if partial cross-sectional restraint is assumed conservatively at each end of the segment (PP), the twist restraint factor &; will be greater than 1,0 in accordance with Table 5.6.3(1) of AS4100. However, unless fly braces are closely spaced or the rafter has an unusually high flange to web, thickness ratio, f, will normally be close to 1.0.62 Frame Design aise DPFBts Considering that the partial restraint assumption is probably conservative, a k, value of 1.0 is recommended for simplicity. It may appear that there should be a useful reduction in effective length because the wind loads act at the more favourable tension flange level. However, the benefit of this is not significant as most of the bending moment within a segment is due to end moments, and the segment should not be likened fo a simply supported beam under uniformly distributed load applied at the tension flange level. Moreover, the reduction in effective lengths of a simply supported beam under such loads is limited in some cases as discussed in the next subsection and AS4100 offers no concession for bottom flange loading. For this reason, k, should be taken as 1.0. For a segment between fly braces and with the bottom flange in compression, the lateral rotational restraint provided at the ends of the segment by adjoining segments should strictly speaking be disregarded because it is unlikely that the adjoining segments are fully restrained laterally in accordance with Clause 5.4.3.4 of AS4100. There is, however, a degree of lateral rotational restraint which would allow &; fo be taken as 0.85. Tn summary, the effective length Le for segments between fly braces is given by kik, Las L, -0%1.0 0.858, = 0.85S- (4.12) The moment modification factor o,, for segments between fly braces will usually be greater than 1.0. For segments which have a reversal of moment, part of the segment will have its compression flange restrained by purlins but this benefit should be ignored. © Without Fly Bracing under Uplift Although some fly bracing is recommended, it is interesting to consider the rafter behaviour vider uplift where there is no fly bracing at all. In this ease, the full portal span should be taken as the effective length, and @q should be based on the bending moment distribution across the rafter span. Even though the validity of this approach for a kinked member is doubtfil, the large effective length should equate to such a low capacity that some fly bracing will be necessary. Designers often feel that the lateral restraint offered by purlins to the tension flange under uplift conditions should also increase the lateral buckling capacity. | However, theoretical and experimental studies [9,12] of the bracing of beams have confirmed that translational restraint alone acting at the level of the tension flange, such as that provided by purlins, is virtually ineffective. ‘These studies show that if the lateral restraint is combined with some twist restraint, the buckling capacity is increased. It is possible to design the purlin-rafter connection for some rotational capacity by providing two or four friction bolts to the cleat, or by using wider cleat plates with more bolts. ‘There may be architectural advantages in avoiding fly bracing, such as when a ceiling is required above the bottom flange level. Investigations have been carried out [12] into the effectiveness of standard purlin connections in providing rotational restraint to the rafters. The results revealed in part that the requirement for rotational stiffness is a function of the initial geometricAISCDPFBAS Rafters 6 imperfections in the rafter. That is, for very crooked rafters, greater stiffness in the brace is required. ‘The theoretical and experimental studies have so far indicated that ordinary or standard purlin connections are effective to some degree, provided that the bolts“are properly tightened. Further tests and analyses are needed, but in the meantime tension flange bracing should be disregarded. One fly brace neor ridge Fly brace near column to stabilize knee 7 He Fly brace 4| A No girts this I~ side say St LOCATION EFFECTIVE LENGTH Outside flange in compression 0.85 Sp Inside flange in compression 0.85 Sf Column without girts or fly bracing] 0.85 Figure 4.1 Effective Length Factors for Bending in Rafters and Colunns © With Fly Bracing under Downward Load The effect of the bottom flange near the columns being in compression due to gravity loads or other loading should be considered even though most of the bottom flange of the rafter is in tension, A fly brace is recommended near each knee and near the ridge to restrain the inside comers of the frame at kinks. A stiffener between column flanges as indicated in Figure 4.1 effectively exténds the bottom flange of the haunch to the outside column flange which is restrained by girts. This effectively provides some restraint to the inside of the knee, However, a fly brace near the knee is still recommended. With fly braces at least at the knees and the ridge, the effective length will be 0.85 times the spacing between fly braces. The value of the moment modification factor dy, for the segment should be determined using one of the three methods in AS4100, but Method (iii) in Clause 5.6.1.1(a) is likely to be most appropriate if there is no intermediate fly brace between the knee and ridge. It is recommended that any haunch should be ignored in determining the design bending capacity ¢Mj, of the segment, but the applied bending moments should be reduced by4 Frame Design AISC DFFDS factoring the moment at any haunch section by the ratio of the elastic section modulus of the unhaunched section to the corresponding elastic modulus of the haunched section, Alternatively if each end of the haunch happens to be fly braced as in the design example, the haunch may be treated as a tapered segment in accordance with AS4100. | 4.3.3 Major Axis Compression Capacity Na In AS4100, the nominal member capacity Ne, is required in the combined actions rules for determining the in-plane member capacity in Clause 8.4.2.2, It is obtained from Clause 6.3.3 as Nex rch y An fy 4.13) where 4, is the net rafter cross-sectional area, which is generally the gross area for portal frame members (see Clause 6.2.1 of AS4100). The member slenderness reduction factor a¢ is given in tabular form in the code for values of the modified slendemess ratio Ane = (Le Fe )ofky ff, 7250 where Le is the effective length equal to kel based on the actual rafter length L from the centre of the column to the apex. ‘Two effective lengths need to be used under Clause 8.4.2.2 of AS4100. For combined actions, the effective length factor ke should be taken as 1.0. The rafter also needs to be checked under axial load alone using effective lengths determined from the fraine clastic buckling load factor 4... This factor can be obtained either by using Equation 4.5 with the Davies method [5] outlined in Section 4.2.3 of this book, or by using commercially available computer packages such as Microstran [13] or Spacegass [14]. The check under axial load alone is unlikely to be critical for portal frames without eranes because they are principally flexural frames with low axial loads in all members. The form factor kr which accounts for local plate buckling is given in the BHP section handbook [6]. 43.4 Minor Axis Compression Capacity Vy The nominal member capacity Ney for buckling about the y axis is required in the combined action rules of AS4100 for determining the out-of-plane capacity in Clause 8.4.4.1. Tt is obtained by taking the effective length Las the distance between purlins, since the purlins are restrained longitudinally by roof sheeting acting as a rigid diaphragin spanning between the roof bracing nodes. The theoretical effective length of an axially loaded member (rafter ot column) with discrete lateral but not twist-rotational restraints attached to one of the flanges may be greater than the distance between the restraints, Unfortunately, there is no simple method of determining the effective length of such a member, In the case of a rafter restrained by purlins, some degree of twist-rotational restraint would also exist. The combined full lateral and partial twist-rotational restraint provided by the purlins to the outside flange should be effective in enforcing tie rafter to buckle flexurally between the purlins, The capacity Ney is obtained from the minor axis modified slendemess ratio Any =(Le ly ify fy 1250