Tubular Steel Structures Theory and Design M S Troitsky 2nd Ed 1990 PDF

ow) ey oF 5 0 Oe La SMe Most a)a TUBULAR STEEL i STRUCTURES — Theory and Design SECOND EDITION : by MLS. TROITSKY, D.Se. Professor of Engineering Concordia University MontrealTo my wife Tania © Copyright April 1982 Second Edition, 1990 Library of Congress Number — 81-8093, ec OPTED ESE ICS publisher’s preface As authors of major works in highly specialized areas of technology know, finding a publisher for the manuscript may be a difficult task, When the subject matter is of narrow interest and highly circumscribed, it is outside the province of the commercial publisher, who must realize a sales volume more than enough to cover his costs. The more comprehensive and scholarly the work becomes — and, thus, the more useful to, the technology — the more difficult is the problem of finding a publisher. When publication requires hundreds of pages and a great mass of art work, itis usually beyond the scope of the journals of pertinent technical societies. In such eases, a patron may be required if the scholarly work is to be put in print, hard-covered, and made available to the public Fortunately, there are patrons in most areas of scienee, engineering, and technology — foundations, associations, endowments, and even individuals who serve to {ill the gap in publication possibility. The James F. Lincoln Are Welding Foundation is such a patron for comprehensive studies pertaining to are-welding technology and arc-welded stecl design. As a part of its broad function in providing educational serviees that enhance the progress of are welding, the Foundation subsidizes the preparation, publication, and distribution of scholarly works that benefit designers, industrial managements, and others concerned with the technology and engineering of arcewelded steel. Dr. Troitsky’s monumental work on “Tubular Sree! Structures, Theory and Design” fits admirably into this function of the James F. Lincoln Arc Welding Foundation. ‘The growing importance of tubular steel in the design of structures is well known to engineers. The properties of the circular cross section and the achievements of strength with light weight that ean be realized with it are intriguing tothe designer. But even though tubular steel is being used extensively in structures, the technology of its use is widely distributed through the literature. Dr. Troitsky in this book brings together the scattered information on tubular design and analysis in a well organized manner that should make it a valuable text and reference source. Dr. Troitsky is not unknown to beneficiaries of the James F. Lincoln Are Welding Foundation’s publishing ventures. He is also author ofthe work “Orshotropie Bridges, Theory and Design”, published under the aegis of the Foundation in 1967. This book, as the present one, also was considered as having too limited outlet possibilities for a commercial publisher — although the first printing of 5,000 copies sold out within a year. Such widespread distribution of volumes of highly specialized interests is facilitated by the Foundation’s subsidized distribution policy, which supplements its patronage of the preparation and printing costs. In effect, the Foundation offers the books it publishes at but a fraction of costs — just a nominal sum to guarantee that the recipient has a real need and interest in the work ‘Other well-known design books published by the Foundation inelude “Design of Weldments" and “Design of Welded Siructures”, both by Omer W. Blodgett, In addition, the Foundation has sponsored many other works of more widespread interest, such as “Metals and How 10 Weld Them”, “Design Ideas for Weldments", and “Modern Welded Structures”. 9902-aHI-£0 oT 62: SWdd ‘ld Woe (RAH OL OVeo'd$$ ‘The James F. Lincoln Are Welding Foundation was formed in 1936 by the Lincoln Electric Company to provide educational services in the burgeoning art of arc welding and arc-welded steel design. Although expending most of its efforts on publications, the Foundation has also sponsored various Award Programs to encourage architects, engineers, and designers to use arc-welded steel more efficiently and esthetically in weldment and structural designs, ‘We feel certain that “Tubular Stee! Structures, Theory and Design” will be a worthy addition to any engineering library. Richard 8, Sabo, Executive Director, The James F. Lincoln ‘Are Welding Foundation The serviceability of a product or structure utilizing this type of information is ‘and must be the sole responsibility of the builder/user. Many variables beyond the control of The James F. Lincoln Are Welding Foundation affect the results obtained in applying this type of information. These variables include, but are not limited to welding procedure, plate chemistry and temperature, weldment design, fabrication ‘methods and service requirements, preface In this second edition the contents have been thoroughly revised and every effort was ‘made to correct errors and misprints detected in the first edition. Although much of the data of the first edition has been retained, the following changes and additions were made in the contents of the second edition. In Chapter 7 “Bins and Bunkers” sections retained to the Ring Girders and Roofs for Circular Bins, were substituted by the new text for Circular Girders Supporting Bins and Roofs for Circular Bins. Also in Chapter 12 “Tubular Conveyor Galleries” new sections concerning the design of "Ring Girders at Supports, Method by Krupka” were added. 9902-aH-£0 oT 62: SWdd ‘ld Woe (RAH OL oLea'dresume Saree eee ee cere eee eee reas eer ee cere eee MS. Troitsky, D.Sc. is Professor in the Department of Civil Engineering at Concordia University, Montreal. He has taught at Ecole Polytechnique and McGill University, Mon- treal. He has varied industrial experience in a fulltime and consulting capacity, and was responsible for the design and supervision of the construction on several major bridges, He is the author of the books: Orthotropic Bridges; Cable-Stayed Bridges; Stiffened Plates: Bending, Stability and Vibrations; Design Guidelines for Rectangular Stee! Bins, Tubular ‘Steel Structures; and Prestressed Stee! Structures; as well as numerous papers on structural ‘and bridge enginecring. He is a member of @ number of technical societies. acknowledgements {In preparing a book in the field of engineering, an author must depend upon the publishing works of many engineers, scientists, mathematicians and engineering firms. The author and publisher of this book gratefully acknowledges the help and coopera tion received from many sources, Special acknowledgement is herewith made to the Canadian Steel Industries Construction Council and Imperial Oil Limited, who have financially supported the preparation of the research reports by the author on some topics treated in this book. ‘The author is also indebted to many org: ns and individuals for their help, assistance and permission to reproduce the designs. tables and graphs. The author also wishes to thank Mr. G. Korioth for the preparation of the draw- ings, and to Miss M. Stredder for typing the manuscript. Special acknowledgement is herewith given to the Canadian Institute of Steel Construction for whom the author has prepared a number of research reports and which work inspired the author to prepare this book 9902-aHI-£0 oT 62: SWdd ‘ld Woe (RAH OL Orecontents CHAPTER 1 — INTRODUCTION General Concept - 3.6 Edge Stresses and Deformations along. Intersections of Closed Cylindrical and 9902-aHI-£0 oT 62: : M4 12 Advantages of Tubular Structures ra Conical Shells under Internal Pressure : 1.21 Optimum Utilization of Material J... Ft 3.6.1 Design Loading 122 Pabrication i 3.6.2 Force According to the Membrane 1.3 Structural Behavior of Tubular Structures... 1-2 Theory oo 218 3 Lh1 Application of Thin-Walled Shells 3.6.3 Determination of Redundants X; and Xp ; Theory to Tubular Structures 2 and Values of My and Hl. 38 3 1.32. Buckling Stability of Tubular 3.64 Total Local Stresses of Cylindrical Shell 3 Structures... IB {in the Given System) at Level ava a 13. Junetion of Shells of Difierent Figure 3.9 (a) a8 Geometry | Sees 3.6.5 Summary of the Edge Siresses of the Conical Shell at Level a-a, Figure 3.9 (a) 3419 3.6.6 The Circumferential Edge Stresses of CHAPTER 2 — LOCAL AND OVERALL ete et 9 BUCKLING OF CVLINIMOAL SHELLS 3.64 Tangential Edge Stsss ofthe Conical s 24 Introduction 2 3.6.8 Displacements in the Plane a-a of the 22 Overall and Local Buckling +24 Points at the Intersection Line of the 2.3 A Paradox in the Buckling Analysis of Middle Surfaces of the Shells, Directed Cylindrical Shell under Axial Compression ‘Along the Radius r from the Axis 2.4 Importvetions of the Shell Shape and ofthe Shell. . 9 Edge Ftect : 5 3.6.9 Twisting of the Section at Level a-a a 25 Practical Application of Experimental Dat Considering Clockwise Direction as the 26 Allowable Design Stresses... : Positive, Figure 3.11 (6) x9 z References. Chapter 2 Edge Effect at the Intersection of Closed 2 Cylindrical and Spherical Shells under a CHAPTER 3 — EDGE EFFECT AT TUBULAR Internal Pressure, Figure 3.12 .... 319 Eerie 3.7.1 Design Scheme and Determination of | the Forces « 29 3 Physical Concept Sopra oe 3.7.2 Summary of the Local Stresses of the 3.2. Methods for the Solution of Edge fect Cylindrical Shell in a Given System at Problems... ae Level aa, Figure 3.12 (a) 320 3.2.1 Solution of the Problem by the 37.2.1 Meridional Edge Stresses of the [Force Method vee 2 Spherical Shell 3.20 2.2 Solution of the Problem by the 3.7.2.2 The Circumferential Bdge Stresses Method of Deformations. 33 ‘of the Spherical Shell ...2+2....2++ 321 e 3.3 Influence of Edge Effect on the Carrying 3.7.3 The Increase of the Radius ¢ at a 2 E Capacity of the Shell B13 Level a-a, Equal to AR, and Turn of the & i 34 Local Stresses and Deformations under Section at some Level Clockwise ¥. S ; Internal or External Pressure along the Figure 3.12 oss... 321 8 I Intersection of Closed Cylindical Shells 3 Bug Elles of Closed Cuindical Shei : 3.411 Designations having Stiffening Ring at Their Junction... 3-21 3.4.2 Rule of Signs : 3.9 Edge Effect Stifening Rings of Cylindrical 3.5 Determination of the Stresses under Internal’ Shells under Internal Pressure Po... 324 Pressure in the Thin-Walled Shells According 3.10 Edge Effect due to the Eccentrie Junction to the Membrane TReOry «s+... Bus of the Shells .esesseeeseeeeeseees 32sc |APTER 4 — THERMAL STRESSES IN TUBULAR STRUCTURES References — Chapter 3 3:26 4.1. Basic Assumptions and Designations 4 4.2 Thermal Stresses and Deformations in Walls of Cylindrical Shetls 42 4.2.1 Stresses and Deformations under Uniform Heating — Diagram “1 42 4.2.2 Stresses and Deformations under Uniform Heating ~ Diagram “" 2.2... 4-2 4.2.3 Suresses and Deformations under Non-Uniform Heating — Diagram “ 42 4.24 Stresses at Free Edge of Circular Shells under Non-Uinitorm Heating — Diagram "20" on 43 References — Chapter 4 e 5 52 5. 54 35 56 |APTER 5 ~ SELF-SUPPORTING STACKS Types of Stacks . St Design Loads .. Se 521 Dead Load i 52 5.22 Live Load + 52 5.23 teing .....) 52 5.24 Wind Forces oa 52 5.25 Stack Vibration 4 52 5.2.5.1 Transverse Ossillations 4 5.2.5.2 Effect of Vortex Shedding 55 5.2.5.3 Ovalling Oscillations 55 5.26 Earthquake Forces .... cn BS 5.2.6.1 Base Shear . fee 5S 5.2.6.2 Distribution of Lateral Forces... 5-8 5.2.6.3 Overturning Moment . 38 5.2.64 Supports sens SB 5.2.6.5 Erection Conditions «2.1... 58 5.2.7 Temperature Influence . Brees | Stress Analysis for Self-Supporting Stacks... 5-9 5.3.1 Dead Load .. Sader ey 39 5.3.2 Wind or Earthquake Loads 50 Design of Self-Supporting Stacks 10 5.4.1 Dead Load Stress : 5-10 5.4.2 Dead Loads and Wind Stresses 10 5.4.3 Dead Loads and Earthquake Force Stresses . 510 5.44 The Allowable Stresses 1... 5.4.5 Checking of the Wind Induced Oscillations : Base Design eciie 5.5.1 Design of Anchor Bolts... 5.5.2 Base Plate eee ceeere 5.5.2.1 Base Plate without Gussets 5.5.2.2 Base Plate with Gussets 5.5.2.3 Practical Consideration in Designing Basc-Plates -.-......... 5-13 5.5.2.4 Design of Gusset Plates for Compression Rings . os SAS Approximate Breech Opening Analysis ..... 5-16 5.6.1 Conception of the Method. 5-16 $.6.2 Stresses at the Edge of Breech Opening 5-17 5.6.3 Geometrical Properties of the Reduced Cross-Section of the Stack $19 5.6.3.1 Statical Moment of the Compressed Area ..... 19 5.6.3.2 Statical Moment of Tensile Arca 5-20 5.6.3.3 Moment of Inertia of the Compressed Area .......... $220 5.6.3.4 Moment of Inertia of the Tensile Area : $22 5.64 Stresses at Breceh Opening. Experimental Tests 5-23 5.6.4.1 The Effect of Square and Rectangular Cutouts. cesses $23 3.7 Design of Stiffeners at Breceh Opening... 5.25 5.7.1 Design of Vertical Stiffeners 5.25 5.7.1.1 The Load Imposed on each Stiffener by Stee! Shell and Lining . 5-26 5.7.1.2 The Load Imposed on each Stiffener due to Wind Bending Moment . ates 5.7.1.3 The Eccentric Bending Moment due to Eccentricity of Stiffener ..... 5-26 5.7.1.4 The Effect of the Stiffened Breech Opening on the Strength of the 5-26 Stack «2... sees 526 5.7.1.5 Buckling Stability of Verticat Stiffeners seveeee $27 5.7.2 Ring Stiffeners .... fees 5:28 5.8 Rigorous Breech Opening Analysis 21.1... 5-28 5.8.1 The Finite Element Program 1. .... $28 5.8.2 Method of Finite Element Analysis... 5.28 5.8.3 ANSYS Computer Program nee 528 Numerical Example No. I. Approximate Method : S31 Numerical Example No. 2. Rigorous Analysis, Application of Finite Element Method ae | 37 References — Chapter 5. 531 CHAPTER 6 — MULTILEVEL GUYED STACKS 6.1 Parameters used in the Analysis of a Guyed Stack 2.0.2... 6.1.1 Introduction... : 6.1.2 Guyed Stack Structural System 6.1.3 Support Flexibility ........... 6.2 Analysis of the Stack by the Method of Deformation aie : 6.2.1 Equilibrium Equations for Span c-f 6.2.2 Equilibrium Equations for Spans e-0 aNd FE eeeseeees 65 62.2.1 Span eo... 66 6.2.2.2 Span fy : vise 66 6.2.3 “Additional Formulas used in the Static ‘Analysis of the Guyed Stack 66 6.3 Buckling Analysis of a Guyed Stack 6-7 6.3.1 Introduction . : 67 6.3.2 Buckling Load for a Guyed Stack 1... 6-7 64 Dynamic Analysis of Guyed Stacks... 0... 6-11 64.1 Introduction 7 so Gl 6.4.2 Approximate Determination of the Natural Vibrations of a Guyed Stack by the Method of Deformation IL 64.2.1 Moments and Shears for Span fez 7 6-13 6.4.2.2 Moments and SPAN C0 sees eeeeee 613 6.4.2.3 Transverse Shear at Stack- ‘Supports from the Cables ots 6.5 Forced Vibration of a Guyed Stack . els 6.5.1 Introduction ....... LOIS 65.2 Forced Vibrations from a Harmonie Disturbing Force oe ots Appendix A. Numerical Example ols References — Chapter 6 oa? CHAPTER 7 — BINS AND BUNKERS. 7.1 Types of Bins « pocertts Seed TLL Methods for Classifiations of Bunkers and Bins .....- : Siete 7.1.2 Structural Systems of Square and Rectangular Bunkers and Bins 2 7.1.3 Square and Rectangular Bunkers 72 7.14 Square and Rectangular Bins... 14 1.2 Properties of Bulk Solids ..+-...cecccssce0 74 7.2.1 Bulk Density my 7.2.2 Compressibil : 1a 7.2.3 Particle Size 7.2.4 Angle of Repose =... 1 7.2.5 External and Internal Angles of Friction -...+.+ 7.26 Flow Properties 727 Flow... ce : 7.28 Conditions Affecting Flow Properties 7.2.8.1 Moisture Content : 728.2 Temperature 7.28.3 Gradation 7.2.84 Segregation 7.285 Degradation .. : 7.2.8.6 Corrosiveness sie 7.28.7 Abrasion 7.2.9 Flow in Bins 7.2.10 Pressure Field in Bins 7.2.11 Properties of Stored Materials 7.3 Functional Design of Square and Rectangular Bins ......e..200+ 7.3.1 Types of Bins 73.2 Mass-Flow Bins 73.3 Funnel-Flow Bins 734 Expanded-Flow Bins 3.5 Hoppers ..... 736 Outlets 73.7 Archings en 73.8 Outlets for Mass-Flow Bins 7.3.9 Hoppers for Mass-Flow Bins 7.3.10 Outlets for Funnel-Flow Bins . T3101 Live Storage 7 7.3.12 Rate of Discharge -..-...- 74 Loads 1... a 74.1 Design Considerations =... 7.4.2 Bin Loading from Stored Material 74.3 Pressure on Bin Walls. Janssens Formula +. .-.++0++4 : 7S 7.44 Computation of Static Pressure- Lateral and Vertical .. : 7s 744.1 Janssen’s Method ....cccccsc. 716 74.4.2 Reimbert’s Method ... + TAT 744.3 Static Pressure on Flat Bottoms 7-18, TAA Static Forces — Vertical Friction frees 74.5 Overpressure and Overpressure Cae a 7.4.6 Static Pressure in Bin Hoppers 7.4.7 Design Pressure in Bins Factors, 7A8 Pressures on Bin Walls, Eecentrie Outlets... testes 7.4.8.1 DIN Specifications Metho 7.4.8.2 Imaginary Bin Method 7.49 Literal Pressure Design Curve id 7.4.10 Limiting the Height of a Bin Wall 74.11 Bunker Loads and Forces 7.4.11. Top Surface of the Stored Material is Horizontal .....+ 74.11.2 Top Surface of the Stored Material Sloped 7.4.12 Bunker Design Pressures 7.8 Structural Analysis and Design of Square and Rectangular Bins 7.5.1 General Concept 7.5.2 Forces and Moments Acting on Structural System of the Bin 7.5.3 Vertical Walls, Lateral Pressure the 7.5.4 Vertical Walls, Tension and Bendin Moments 754.1 Tensile Forces 7.5.4.2 Bending Moments 7.58 Vertical Walls, Plating 7.5.5.1 Small Deflection Theory» 78.5.2 Large Deflection Theory 755.3 Determination of the Plating Thiekness 78.5.4 The Effective Widih of the Plating 7.36 Vertical Walls, Horizontal and Vertical Stiffeners, 7.5.7 Walls Analysis of Multicel! Bins 7.5.8 Hoppers. Determination of Dimensions . eee 7.5.9 Forces Acting on Hopper Walls 7.5.9.1 Design Pressure Normal to the Tnelined Walls 7.5.9.2 Horizontal Tensile Forees 7.5.9.3 Meridional Tensile Forces at ‘Symmetrical Hopper 7.5.9.4 Meridional Tensile Forces at Unsymmetrical Hopper 7.5.10 Hoppers. Design of Plating « 7.5.10.1 ‘Average Pressures on Hoppers Panels 7.8.10.2 Converting of Triangular and ‘Trapezoidal Panels into Rectangular ined Panels : 7.5.1 Hoppers. Horizontal and Incli Stiffeners 2.2.2. ree 7.S.11.1 Horizontal Stiffeners 75.112 Inclined Stiffeners 7.5.12 Procedure for the Design of Bi 7.6 Square and Rectangular Bunkers. Structural Analysis and Design « 7.6.1 General Concept 7.6.2 Structural Analysis 7.6.3 Symmetrical Bunkers 7 7.6.3.1 Forces Acting on Side W: alls 9902-aHI-£0 oT 62: SWdd ‘ld Woe (RAH OL O1-s0'd7.6.3.2 Forces Acting on Hopper Walls 7-38 7.64 Unsymmetrieal Bunkers... 738 7.64.1 Forces Acting on Walls s1...... 7-38 7.6.5 Side Wall Design. s 7-40, 7.6.5.1 Side Wall Plating + 780 7.6.5.2 Side Wall Stiffeners 7-40 7.66 Hopper Wall Design .. fall Tal 7.6.6.1 Hopper Wall Plating” 74t 7.6.1 Hopper Stiffeners 741 7.7 Supports for Square and Rectangular Bins and Bunkers ..... . 742 7.2.1 Types of Supports 7-42 7.72 Loads 7 742 7.73 Wind and Earthquake Loads 7-42 7.7.4 Axial Forces in Column under Vertical Loads ..... ocen 7.78 Bending Moments in Columns 7-42 der Wind Loading 743 1.2.5.1 Hinged Supports 743 7.7.8.2 Fixed Supports .. 18a 78 Circular Bins, . 184 1.8.1 Vertical and Horizontal Static | Pressures on Cylindrical Wall ces 744 7.8.2 Determination of Forces and Stresses ||. 7-45, 7.8.2.1 Cylindrical Shell... ......112 7-45 7.8.2.2 Conical Hopper Shell 2. 245 7.8.3 Circular Girder Supporting Bin... 7-46 7.8.3.1 Shear Forces and Bending Moments in Circular Girder oe TA6 7.8.3.2 Stiffening Ring . | 2 TAT 7.8.33 Checking of the Stiffening ing for Buckling... 747 7.8.3.4 Horizontal Bracing System |. 7-47 7.84 Columns... . cos 788 7.8.5 Roofs for Circular Bins ||| +749 7.8.5.1 Tension Ring 7.49 9 Circular Bumpers 754. 7.9.1 Forces in the Cylindrical Section of Bunker... = 154 7.9.2 Pressures in Cylindrical Section. 7.54 7.9.3 Meridional Tension in Cylindrical Section... <. 18S 7.9.4 Forces in the Conical Section” 155 7.9.8 Stresses in the Walls <2. 2.11.51. 755 7.9.6 Ring Beam and Supports»... 7-56 Appendix A. Example No.1... 0.0.0 sscc. 5. 7556. Appendix B. Example No, 2 Ce 268 References — Chapter 7.0.2.1... 72 CHAPTER 8 — HORIZONTAL STORAGE TANKS, 8.1 Introduction 8.2 Principal Design Codes 8.3 Loads and Design Assumptions | ‘8.4 Tank under Internal Pressure «0... 8.5. Heads of the Tanks under Internal Pressure 8.5.1 Hemispherical Heads . 8.5.2 Conical Dished Heads 8.5.3 Ellipsoidal Dished Heads 85.4 Flat-Plate Heads ....... 8.6 Stability under Axial Compression 8.7 Bending Stability tf 8.8 External Pressure Effeet se 86 8.8.1 The Relationship for the Conditions Beyond the Critical Length se 86 8.8.2 Instability of Cylinders Shorter than Critical Length a7 8.8.3 The Design-Allowabie Curve for External Pressure .... sees BT 8.9 Stability of Unstiffened Tanks. Combined Loadings Be 8.9.1 Internal Pressure any Compression . 88 8.9.2 Intemal Pressure and Bending 88 8.9.3 Axial Compression and Bending 89 8.9.4 Axial Compression and External Pressure a) 8.10 Approximate Stress Analysis of the Tanks |. 8-9 8.10.1 Stresses in the Tank under the Weight of Liqui 8.10.3 Tank having Rigid Stiffening End Rings, Either Freely Supported or Fixed under the Weight of the Liquid and Its Own Weight .. 8.11 Stability of Unstiffened Tanks. : 8.11.1 Critical Length between Stiffeners 8.11.2 Collapsing Pressure on Tank with Circumferential Stiffeners ss. ..c...00. 11.3 Design of Circumferential Stiffeners 8.11.4 Short, Ring-Stiffened Tanks ...... 8.11.41 Material Failure : 8.11.42 Buckling between Rin 8.1.4.3 General Instabilt 8.12 Stability of Heads under Uniform Pressure... 8.12.1 Spherical Head”? - 8.122 Hemispherical and Torispherical Heads 8.12.3 Ellipsoidal Head... 8.12.4 Conical Head 8.13 Stability of Horizontal Tank with Saddle Supports : 7 References — Chapter 8. CHAPTER 9 — ABOVE GROUND PIPELINES 9.1 Introduction . : 9.2 Design Loads and Forces... 9.2.1 Dead Load . 9.2.2 Live Load ... 9.2.3 Wind Forces 9.2.4 leing 9.25 Temperature Influence 9.2.6 Earthquake Forces 9.3 Pipe Stress Analysis ......! 9.3.1 Straight Pipe under Internal Pressure 9.3.1.1 Circumferential or Hoop Stresses... 9.3.1.2 Barlow’ Formulas... 9.3.1.3 The Allowable Unit Stress” 9.3.1.4 Longitudinal Stresses 9.3.2 Stresses in Circular Bends 9.3.2.1 Circumferential or Hoop Stresses - 9.33 Bending Stresses - 9.4 Buckling Stabil under Own Weight and Liquid Pressure ..... 9-3 9.5 Multispan Pipe under Different Loadings 9.6 Buckling of the Pipe due to Ovalling Bifect .. 9-6 9.7 Buckling of the Pipe as Beam-Column 9'8 Buckling Stability of Curved Pipes .....00. 97 9.8.1 Basic Parameters of the Curved Pipe. 9-7 918.2 Effect of Flatening of the Cross Section during Bending « 9.8.3 Longitudinal and Transverse Stresses due to Bending .... veces OB 9.8.4 Bending of the Pipe In-Plate of Curvature... 9-8 9.8.5 Bending of the ‘Curvature 9.9 Buckling Stability of Stffened Pipe 9.9.1 Method of Solution ..... 9.9.2 Normal Stress Components for Loading . sitsteseetenses OID 9.9.3 Determination of the Critical Buckling Stresses ctsceersne OB 9.10 Temperature Effect upon a Restrained Pipe ....+. 9.13 9.10.1. Free Expansion of Contraction of the Pipe --.-. + 9-13 9.10.2 Stresses Produced by Restraints at Uniform Change of Temperature ...... 9-13 9.10.3 Stresses Produced by Restraints Non-Uniform Change of Temperature .. 9-14 1 Stresses Produced by Bending and ‘Temperature under Restraint 9.111 Bending Stresses under Weight of i ine and Liquid ....- : ois 9.11.2 Bending Stresses due to the Eccentricity of the Axial Restrain Load .. 9-15 ois 9.11.3 Axial Compressive Stress... 9-16 9.12 Criteria of Safety for Complex Stress ‘System. = 916 9.13 Loads on Anchor and Buttress Blocks... 9-17 9.13.1 Anchor Blocks + 917 9.13.2 Buttress Blocks 9.17 9.14 Buckling of the Pipe at Supports ......... 9-18 9.14.1 Saddle Supports... 9.18 9.14.2 Stress Effect due to Saddle Support. 9-18 9.14.3 Ring Girder Supports 9-20 9.14.3.1 Stress in Pipe Shell 9.20 9.14.32 Rim Stresses, 9-2 9.14.3.3 Stress in Ring Girder 9-2 9.15 Vibration of Pipeline 9.22 9.16 Influence of Imperfections 9.23 9.16.1 Introduction . 9.23 5162 Stability ofthe Pipe having Local Eccentricity 9.23 9.163 Out-Of-Round Pipe .- 9.24 9.16.4 Out-Of-Round Curved Pine + - 9.26 References — Chapter 9 9.26 CHAPTER 10 — UNDERGROUND PIPELINES 10-1 10-1 10.1 Introduction ......++ 10.2 Types of Loads... 10.3 Determination of Load seeeeees 102 10.4 Thin-Walled Pipe Installed with Controlled Compaction . ceceee 104 10.5 Ring Stress. + 10-5 10.5.1 Critical or Collapse Stress ses 1065 10.5.2 Selection of Wall Thickness ........ 10-5 10.5 seen, 1046 10.8 Stability of the Pipeline at Longitudinal and ‘Transverse Flexure .... we 1066 10.9 The Determination of Critical Stresses 1... 10-8 10.10. Stability of the Pipeline having Initial Flexure : 09 10.11. Stability of Pipeline at Eecentrieal | Compression 7 10-10 10.12 Differential Equation of Stability of Straight Pipeline cee 10H 10.13 Stability of Underground Pipeline along Straight Section ......csscesesuseseeses WD 10.14 Stability Analysis of Deformed Shapes of ne 10-13 ps ieee 10,15 Assymptotic Forms of Bulging of Underground Pipelines .... seve 10d 10.16 Stability of Pipeline at the Bends of Greater Radius ....- 10-15 10.17 Stability of Pipeline at Single Span” Crossings 10-16 10.18 Stability of Underground Pipeline a Bends of Small Radius 10-18 10.19 Stability of Pipelines at Branching 10-19 10.20 Stability of Parts of a Pipeline on the Banks ofa Crossing -.-..s+++0+s0se4++0++ 10-20 10.21 Recommendations for Stability Design of Buried Pipeline .......2+e0++ 10-21 10.22 Soil Resistance Against Longitudinal Displacements of Pipelines sees 1021 10.23 The Resistance Against Transverse | Displacements of Pipeline . 10-24 10.24 Determination of Compressive Stresses in 1 Pipeline seseeee 1025 10.25 The Influence of Internal Pressure «1... 10-25 10.26 Design of Pipeline for Stability « 10-25 Appendix 10-26 Numerical Example No. | 10-26 10:27 Numerical Example No. 3 10-27 References — Chapter 10. 1929 CHAPTER 11 — UNDERWATER PIPELINES. U1.1 Introduction 11.2 Pipelaying Methods 11.2.1 Bottom Pull Method <2... 11.2.2 Flotation Method . 11.23 Reeled Pipe Method 11.24 Lay Barge Method 11.3 Pipeline Construction 11.4 Pipeline Coating -.. 11.5 Pipeline Anchoring . 11.6 Stability of Pipeline during Construction 11.6.1 Lay Barge Method of Construction 11.62 Method of Free Launehing ofthe Pipeline . 7 9902-aHI-£0 oT 62: SWdd ‘ld Woe (RAH OL O20"11.6.3 Stresses in the Pipe during the Lowering of Its End by Filling with Water .....-- 11-9 11.64 Critical Tension and Inclination of @ Deep-Water Pipeline Laid from the Barge . eee 112 1.6.4.1 Natural Catenary Method 112 11.6.5 Stresses due to the Lif Submerged Pipeline . 13 11.7 Structural Analysis and Design of Pipelin seveeteeee Held IL7.1 Stability Requirements 0000000007) tt 11.7.2 Stability During Construction =...) 1-14 11.73 Stability In-Place ..-......00000 Meld 118 Hydrodynamic Forces Acting on ‘Submarine Pipeline ......... Its 11.9 Preliminary Structural Design 1s 11.10 Pipe Bending Stress and Radius of CUIVALUTE vee eee eee eeeee tees ug 1.11 Concept of Structural Stability HAD 11.12 Pipe Strength Problem tee HID 11.12.1 Buckle Propagation _. 120 11.13 Buckling of Pipeline under Hydrostatic Pressure sees eceee eves seeeee 1120 1.13.1 Long Unstiffened Pipe’. fee E20 11132 Stiffened Pipe .... 2. 11-20 11.13.3 Critical Length between Stiffening Rings . 7 M2 1113.4 Collapsing Pressure of Pipeline with Ring Stiffeners os. .....ceeeeeeese HQL HIBS Design of Ring Stiffeners 2..70.7) 14.92 1135 Design of Ring Stiffeners 1122 11.13.66 Design of Long Pipe having Imperfections ...... + 1123 11.14 Vertical Stability of the Sagging Pipeline 11-23 1114.1 Criterion for Sinking of the Pipe .. 11-24 11.142 Symmetrical Sag of a Pipeline in Plastically Deformed Sediment 24 11.143 The Basic Equation of Combined Flexure and Elongation in a Pipe ..... 11-24 11,144 Longitudinal Slippage at the End of the Pipe ss... sereesssee 126 11,145 The Characteristic Dimensionless Parameters 0 2eeeeecveseseveee 11-26 1114.6 Practical Forms of the Equations for Moments, Slope and Relative Sag ..,, 11-27 11.147 Induced Stresses... fess 128 11.148 The Dimensiontess Stress Parameters ., : 11-29 References — Chapter ii 1129 CHAPTER 12 — TUBULAR CONVEYOR GALLERIES 12.1 totreduction See: ea 12.2 Internal Arrangement and Exterior Protection of the Tube . 122 12.2.1 Shuiceway 2.401.) 122 12.2.2 Thermal Insulation 122 12.3 Advantages of Tubular Conveyor Galiery 12-2 12.4 Supporting Structures for Conveyor Gallery setetetsecseeeeesseevene ID 124.1 The Inverted V-Bent ..., 123 12.42 The Single Post ‘ tee 123 12.4.3 Advantages of Inverted V-Bents and Single Posts over Conventional Supports eee eteceeeeee ee 123 12.44 Erection of Tubular Galleries. 4 12.5 Fabrication . cer [24 12.6 Loads and Forces... : 24 12.7 Structural Analysis and Design 4 12.7.1 Vertical Loads ......- 124 12.7.2 Horizontal Forces... 125 12.7.3 Temperature Effects. 25 12.7.4 Structural Behavior 11. 1 26 12.8 Stress Analysis... veeeses 128 128.1 General Data se 128 12.8.2 Vertical Loads 3002222222277 7 1! aa 12.8.3 Horizontal Loads 128 12.8.4 Combined Cases . | 129 12.8.5 Allowable Stress 2129 12.9 Stability Requirements 2129 129.1 General Data... 2.177) 129 12.9.2 Local Buckling or Wrinkling |... 1 "12:10 12.10 Overall Buckling 2 123 1210.1 Tube Under Combined Loading... 12.13 12.11 Intermediate Stiffening Rings... 2126 12.11.1 Design of the Stiffening Ring 1246 12.12 Ring Girders at Supports... 02. 12. 1219 1212.1 Analysis and Design of Ring Girder eee eee AD 12.12.2 Horizontal Loads 2222021201111 42.22 12.13 Ring Girders at Supports, Method by Krupka wee... ee... 12.23 12.13.1 Vertical Supports ‘| 12.23 12.13.2 Saddle Support. 12-24 12.14 Euiptical Tube for Conveyor Gallery |. 12.25 12.14.1 Loads and Forces tee 1225, 12.142 Stress Analysis + 1225 12.143 Vertical and Horizontal Loads ||! 12-26 12.144 The Principal Stresses...) 1... 12-26 12.145 Horizontal Loads... 212.111) 12.26 12.146 Buckling Stability 171 1!) 7 12.26 1214.61 Bending 0002272002221 12236 12147 Torsion 2.2.0 22) co. 1228 References — Chapter 12 + 1229 APPENDIX A [New AWS Code Specifies the Design of ‘Tubular Structures... eee AL CHAPTER 1 introduction 4.1 General Concept Tubular structures consist of stiffened and unstiffened shells. or eylinders constructed fram steel sheets with large diameter-to-thickness ration D/t. Such structures are used for the storage of granular solids, liquids and gases, transportation of liquids and gases, as supporting galleries for conveyors and for discharge of cool or heated gases Tubular stect structures in most cases are space type structures such as: stacks, bins and bunkers, liquid stor. age tanks, pipolines and conveyor galleries, Cylinders with large diameter-to-thickness ratios ‘most often are called shells. Typically shells are stiffened ‘When a single descriptive term is required, th ‘or “tubular member" is wilized, 1.2 Advantages of Tubular Structures Any advantages of tubular structures may besummarized as follows: 1.2.1 Optimum Utilization of Mat Structurally, the shell is very efficient, since it is a continuous medium and a high proportion of the material ‘can be used to capacity. The cylindrical shell provides the ‘most efficient cross-sectional shape under axial compression having equal lateral restraint in all directions normal to the oylinder axis, The material is convenientiy distrib- tuted along the perimeter and the radius of gyration is constant for all axis passing through the centroid of cross- section. To insure the stability of compressed section in all directions the circular eylinder is the best by a comparison of all other sections and the most economical, The particular advantages of some tubularstructures, Consists in their capacity to perform technological fune- tions and simultaneously to serve as a carrying structure, 4 for instance, in the case of pipelines. Generally, itis possible to design tubular structures having a minimum weight of metal, simplicity in fabrication and erection, and durability in operation. The second factor permitting the reduction of the Weight of tubular structures is substantially smaller aero- dynamic and hydrodynamic resistances under wind and Water actions, respectively. For instance, for relatively high tubular structures where the stress often depends by 80%5 0 90% of wind pressure, the amount of metais in a structure is possible to be reduced more than twice by comparison with conventionally designed structure. It should be mentioned also that the material of tubu- Jar structures by comparison with the equal cross-section. al areas has a minimal perimeter of contact under atmos- pheric influences. Therefore, tubular structures require relatively smaller quantities of paint or other protective Measures. Also, experience indicates that the effect of corrosion al the exposed area of tubular structuresis from 16%% to 30% smaller than at conventional structures, 1.2.2 Fabrication Although the advantages of the tubular structures have been known for a long time, it was the introduction of Welding which made its extensive use possible. Welding has facilitated the design and construction ofa great vari- oty of tubular structures due to simplification in the fabri cation of joints. Old types of joint connections of tubular ‘members by bolts and rivets were expensive and did not Provide simple and safe joints. Fusion welding is the most widely used method of fabrication for the construction of tubular structures, This method of construction is virtually unlimited with regard to size and is extensively used for the fabrication and erection of large-size tubular structures in the field. Often such structures are fabricated by the method of sub- assembly. In this process, sections of the unit are shop welded and then assembled in the field. Members of tubur- lar structures having a size sufficiently small to permit transportation by trucks, rail, or barge are usually completely shop welded because of the lower cost and greater control of the welding procedure in the shop. There are two types of fusion welding that are extensively used for the fabrication of tubular structures. These |. ‘The gas welding process, in which a combustible mix- ture of acetylene and oxygen supply the necessary heat for fusion, and, 2. The electric-are welding process, in which the heat of fusion is supplied by an electric current. 9902-aHI-£0 oT 62: SWdd ‘ld Woe (RAH OL O1e0'dAre welding is the preferred process because of the reduction of heat in the material being welded, the reduction of oxidation and better control af the deposited weld netal. A wide range of arc-welding equipment is av able, from the small portable welding units to the large automatic welding machines. Small arc-welding machines ate widely used in welding shops that fubricate small equipment whereas automatic machines are better suited for the welding of heavy sections involving the deposition of a large quantity of weld metal ‘One of the most recent and successful developments in the field of arc welding is the submerged-are welding process. The process involves submerging of the are be- neath a blanket of granulated mineral flux. In addition to completely protecting the weld metal from the atmos- phere, this process makes the weld metal virtually free of hydrogen, ‘The success of fabrication by welding is dependent upon the controt of the welding variables such as experience and training of the welder, the use of proper materials, and welding procedures, A number of codes and standards have been established for control of such welding variables ‘The American Welding Society (AWS) established the basic standards for qualifying operators and procedures. These standards of qualification form the basis for ‘most of the standards in the various codes. For practical purposes, therefore, the rules for qualifying welders and Welding procedures are essentially the same in the various codes and standards. Regardless of whether or not the welded tubular structure is intended to meet one of the codes or standards. itis advisable that the welding con- form to one of the minimum standards. A variety of types of welded joints are used in the fabrication of vessels. The selection of the type of joint ‘depends upon the service, the thickness of the metal, fabrication procedures, and code requirements, 1.3 Structural Behavior of Tubular Structures 1.3.1 Application of Thin-Walled Shells Theory to ‘Tubular Structures Tubular structures which have found a wide application in modern industry consists generally of the eylindrical, conical and spherical shells of their combinations. Forthe practical analysis and design of tubular structures, we ‘may apply the theory of thin-walled shells, however, with certain limitations. This is because the formulas for thin- walled shells were developed under the theoretical assumption that the ratio t/R is very small, However, in practice, in application to tubular structures, this ratio UR may have greater values and the results obtained always have an approximate character. Therefore, itis important to find the maximum value of the ratio t/R when the deviation obtained by the theory of thin shells is practically permissible, or to determine the limiting value of ratio t/R at which the tubular structure may be considered as “thin-walled.” We approach this problem under certain consideration which may permit us to obtain a practical solution of the problem. Tubular Steet Structures — Theory and Design Let us consider a closed cylinder under internal press sure, having internal and external radii, and R respectively, and {= R =1, isthe thickness of the wall. The maxi- ‘mum and minimum values of the circumferential stresses applying the theory of thick wall eylinders are Con ay ae But from the :heoryof thin-walled eylinders. it isthe snide rads of the Wall we have a= et a2) Because R=1+Land =e .aftersubstituting these values into (1.1), we obtain The deviation of the approximate value of a from the exact will be given in per cent by the expressions pp tte? 1+ (tse)? ag = 190 /t = /dte/e) 2 Ln t/e + 1/a(e/ey? 4) Assuming certain values of tin function or r, we may find corresponding values of a and ay shown in Table tl TABLE 1.1 — Values of Coeticlents a; and a, e oar | 045 _| 02 oar 025 | 086 0.90 38, 78 7 25 5 This Table indicates that even at = 0.2r the deviation from omax does not reach 1%. Regarding the deviation of 23.5% from amin, this deviation may be considered as the factor of greater safety. Therefore, for the design analysis of eylindrical shells, used in practice, formulas developed for thin-walled shells, may be applied up to the values of 2 For the analysis of cylindrical shells also of great in= terest is the determination of edge forces. These forces basically depend on the deformations at the shells edges under external forces, The change of ther= ] mM I t 3 — . : | 8 R | ¢ s, | | g NOT H g4 : 5 HH 3? P= 1 7 | | | Pe a a eT) 0 10’ 10 Rt FIGURE 2.5 — Correlation factors for unaitined circular cylinders subjected to axial compression. Local and Overall Buckling of Cylindrical Shells AT 2. sRECOMMENDED Lt ons Sara a A er Bb cmae os) BB conso0- navn Lee | wan A. —o,* 388. mcconmmercens yer BUCKLING STRESSES NN KSL FATT « Sareea eae ove FIGURE 2.8 — The recommended allowable buckling stresses. AC yield point for mild steel Fy = 36 ksi the limits are cbc 92 of the ‘of the forces XX, and NfeosB, considering these Frees as applied at the level a-a, or at the middle plane of the stiffening ring "st", and the loadings P, acting on the ring ‘long the small depth te then the equations(3.74) may be Simplified substantially as Follows: nea sebeeSt) + xa 0h) + aa(c88h) + xan + act. gst aa mit anfsing) of ct st se gagt each xy (2808) + x, GOSS) + aKa + LOFT = 85D ast oe 2 (8ie0s8) nical) tan. + matetSy + a 8S) - watt a ast a (sfsins) 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL Ore324 and + AE FBS) + HV CURE 6S) - st 4nfcoss) = 0g +0 3.76) Using Tables 3.1 to 3.12 and Equation (3.75) we may determine the unknown factors Xj. Xo. Xy and X,, and alter the forces M,, Hand Q. acting on the separated shells of the basic system, Figure 3.14, in Sections aa, nd a:-a,, apart from the membrane farces Nf and Nf. ‘The positive directions of the forces are shown in Figure 3S, "he Forces M,. H amd Q per unit length of the eircum- forential sections aj-a, and ay-a,, at the locations of the original edge effect, will be determined by the following expressions. In general eases, on a7 ici? Wg * Hysins In particular cases when the directions of X,, Xo Xyand X, ave coinciding with the direction used in Figure 3.14, wet = x, 4 Boe x 6.78) Ogg * “Hat Og © -Kvsind When the forces N. M, Hand Q are known, its possible to determine the local stresses in the shells using the formulas 3.36) to (3.43) and (3.44) to (3.54), The forces acting per unit length of thestiffening ring in the basic system at the positive directions, shown in Figure 3.16, have the following values. weg! oe ie FIGURE 3.16 — The postive clrections of the foreesinthe silten- Ing rings Tubular Steel Structures — Theory and Design Jn the general ease N= Xi + te + pty, + NS cose tee GE a + xa 4 CtarcsenScose) “BE x In the particular case with the coincidence of directions of Xi, Xs, Xa Nyand N with the diteetions shown in Figure 3.12 Mona + xy + pty = NT cosd : ty, O80 = Xa + Ute-KesnSeose) —3* It is necessary to remember that xo e wSsing and, therefore, the forge acting on the stiffening cing song the axis ofthe sell (fom the plane of the ving) eau to ver, The displacement in the direction ofthe rads from the axis ofthe shell, Figure 3.14 i seta, = setyst at = oa nf Bes fyiBes SM) > Ao Gs, 8S tt, WSoos8) - go uo - 0S, The turn of the section at level a-a (positive turn) is accepted clockwise for the left joint or unclockwise forthe right joint, sete pehsour ct akcereeecian tees geta Fa ete" eee Ap sot (382) 7 Sa ie 7 Se 3.9 Edge Effect at the Stiffening Rings of Cylindrical Shells Under Internal Pressure p This ease is shown in Figure 3.17 The increase of shells radius in zona located at some sistance from the original edge effect is oeBa-y as The increase ofthe radios a the stiffening ing deg, = (ecvam) Ee (384) where A is a cross-sectional area of the stiffening ring Force equations for the ease under consideration are yee Effect ot Tubular Structures ch ech 4 got mi 8ey + HES + aS = 0 G85) chy ygck 4 get mogy + noch + ach = 0 3.86) Pale i therefore According 10 Tables 3.1 10 2.12, A84 p = 0, theres fog * > Sin oan After subsitting into (3.87) the values of fom Tables Sito 312. we obtain *Sco | Boe (3.88) amt” lect Irom this, flows that 1s, - st 3.89) m= See 89) tgp ~~ (orrtesy) 8.90) After substituting into Equation (3,86) the value (3.89), the values from Tables 3.1 to 3.12. and the value 8.90), considering also Equations (3.1) and (3.84), we obtain us, " + (poem EE ~ BERG we ER Efoy 0 391) and, considering Equa n (3.12), we find et Pad x? |v. Ee eieie + wou - PE = 6 a9) From Equations (3.91) and (3.92) we obtain ye PEGA PScp [A-B) ~ Cty) = ES, ap) “hen Bhp (AD) Cen) Epp l0.2553,a + 2E*S at BSc 13.93) a cc ot yk ‘ At infinitely rigid tion (3.92) gy We may abiain from Eguae w= Psy Ge) 0.94 The stresses we may find from the formulas (3.36) to (3.43) using MeHg" X, and Hy = 0 3.95) 3.10 Edge Effect Due to the Eccentric Junction of the Shells The values of the bending moments originating in the shell at an eccentric junction of separate elements, may be determined by the joint solution of the equations. mt aloe ra) my wee where m= summary of the momentsacting in both parts of the shell per unit length which is equal naa (3.97) where mg = moment per unit length in the left part of the shell mp = moment per unit lengthin the right part of the shell ‘A. = value which depends on the configuration of those parts of the shells at the junetion and used according to Table 3.2, assuming that the twisting angles of both parts of the shell are equal 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL Ot-sa'd3:26 TABLE 3.14 Tubular Steel Structures — Theory and Design Eccentric Junctions of Shelie, ‘Anglos of Twit tthe Shes of Junction Let ight x ah Se ay > te ' ESE Hse i Son 3 V3 (1-08) 5 é x vr 0.3 8. = 0.78FE ed tein, tine fale Ey81g8N1 ta8) BW Lee z pees 3 Hav i Hi B ig 7 0. 7800E (2-20) eega (2-20) eta REFERENCES — CHAPTER 3 DAI] Timoshenko, S., Strength of Materials Part 11, [3.7] Lessig, E.N., Likev, A.F, and Sokolov, A.G., “Steel Advanced Theory and Probert, ade D. Von Not. Shoet Structures, Sate Ef of Laeratuie on Cone teand Company, Ine. New York, 1953, pp. 164168 rutin and Architecture, Meskow, 986, Gn Ros- Sian), pp. 372-02, [3.2] Roark R.., Formulas for Stress and Strain, Sedo MeGrawcHil Book Company. Ine. New York. 1984, [B48] Coates, W.M. "State of Stress in Full Heads of Pres op. 264268 sure Vesscis" Trans Am. Soe: Mech Engrs, Vol 32 Paper No. APM-S2:12 193), pp. 117-131 [3] Bievo, CB. and Gramm, R. Engineering Dynn- Se VOLE, Blacke€ Son Ltd, London, 1956,pp.246- [39] Wats, G.W., and Barrows, W.R. “The Base Elastic 38 Theory of Vessel Heads under Inieral Pressure" J. Appl. Mechanics, Vol 11, March 1989, pp. 38°73. [54] Timoshenko, Sand Woinowsky-Keleger,S. Theory of Plates and Shel, 2nd od, McGraw-Hill Hook Com- [3.10] Gil. 5. Editor, The Sess Analysis of Presre Ves pany Ine, New York, 1953, pp. 82-48, sclsand Pressure Vessel Components, Pergamon Pcs Oxford, 1970. pp 3630, (351 Fhugge W. Suesesin Shells 2nd. Springer-Verlag. Nest York, 1978. pp. M6381 (B11) Wat, OW. and Lang, A, “Stress ina Pressure Vessel with '@ Conical Heal, Trarsacons AM. Soe: Mech 56] Browns LE, and Young, EA, Process Equipment Design—Vessel Design, John Wiley & S Ine. 1968, pp. HES-119, 126-128 Engrs, Vol. 74, 1952, p. 315. CHAPTER 4 thermal stresses in tubular structures 4.1 Basic Assumptions and Designations ‘The change of steel temperature is assumed in limits ofthe constant modulus of elasticity and stresses inthe limits of Hooke’s Law: The diagrams of change in temperature are shown in Figure 4.1, where & initial temperature tmaxs'min = ae maximum and minimum tempera tures, respectively ‘The signs are shown according to the positive change of temperature (cating) In the case of reverse changes of temperature it is necessary to change the signs of the corresponding ordinates, The average temperature of the shell or ring is, toax + tas min (4) The values of edge ordinates + ( of the diagram of temperature change from tj(when tj = {,),are as follows, ¢ = max 5 Sin (42) FIGURE 4.1 — Diagram of the chango of temporature: (a) Arbitrary change: (b) Linear change, With a change of temperature the deformation may be expressed as an = atse a where a= coefficient of thermal expansion, For stel eta f= dimension ofthe member At = penral increase of temperature The relative deformation i ee hw ate aa) At free change of temperature deformations thermal stresses do not originate in the constructions. However, when the deformations are restricted, then the stresses originate, and according to Hooke’s law os ce = sa0ne 45) At the two-dimensional state of stress of the shell, we have the known relations INTERNAL SIDE tmax 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL O1-s0'd42 Tubular Steel Structures — Theory and Design (4.6) an substituting nto (4.6) the values from Equation 44) Pgrte se weobain ayo = 2 Batt an The linear relations between the changes of tempera ture, deformations and streses permit the superposition of the diagrams of temperature changes 4.2 Thermal Stresses and Deformations in the Walls of Cylindrical Shell Let us now consider the influence on the shell by the line early changing temperature, Figure 4.1 (b) ‘The diagram of changeable temperature we resolve as follows: the t,-diagram representing uniform distribution as defined by the expression (4.1) and the diagram indi- cating changeable t along the thickness of the wall, as defined by the expression (4.2). 4.2.1 Stresses and Deformations Under Uniform Heating — Diagram “t" Let us consider the sections located far from the ends. a. The case when one end of the shell is free In this ease the deformations are free and the stresses 0 ‘The values of the displacements and deformations we may determine using formulas (4.3) and (4.4) after substituting into At the value tj and instead of the corresponding dimension of the shell (radius, length of the cit- Eumference or the length of the generattix.) b.Thecase when theesternal connections do not perntit the increase of the lengths of the generatrix, but the radius mar freely increase ‘The value of the restrained deformation along the generatrix is ers aty (48) ‘The value of the restrained deformation along the rine e20 (49) After substituting these values ¢, and e; into the expression (4.6), we obtain eee (4.10) a) __vBaty ~ a8) The clongation of the generatrix is equal to zera, and elongation of the radius is ar = arty (4.10) The case when the external connections do not permit the elongation of the radi and in the direction of the generatrix there are no connections As inthe previous case, Ay = 0.and the elongation af the generatrix is Aa = atts (4.12) Restrain of the deformations is ere (4.13) en = aty The stresses are oe 2 (ev?) zor, av) (4.14) 6. Case of the total resteain of the expansion at heating er ter so aty Baty ira) (4.15) This is according to the expressions (4.6) and (4.1), 4.2.2 Stresses and Deformations Under Uniform Heating- Diagram “t7 Let us consider edges of the shells a. Case when the edge of the shell is free The stresses and deformations will be the same asin the case considered above, namely as at the sections remote from the edge at the similar edge conditions. 4.2.3 Stresses and Deformations Under Nonuni- form Heating — Dingram “st” Let us consider section remote from the edges ofthe shell. Figure 4.1 In this case there will be no deformations. because the shape of the circular shells and resistance of the adjoining sections do not permit their origin. ‘The diagrams of stresses a, and 0, will be the same and similar to the dia gram, The values ofthe stresses of the edge fibers according to Equation (4.7) are or = 02 = = Bat (4.16) mal Siresses in Tubular Structures on During the linear change of the temperature, Figure itis necessary 1oimpase on the bending moment, given by “a (h) the state oF stress is equivalent to the case of the the expression (4,17), the equal moment but having the = 2 2Blty) on fy These moments M, alty directions on the circumferenti id M, are weting in (wo mutu- and meridional 4.2.4 Stresses Non-U For the transfer from the case considered above to the ‘ease under consideration, Figure 4.1 (b) and to satisfy conditions at the free edge Free Edge of Ci orm Hea and ‘opposite sign. If the shell is heated from the inside, Figure 4.2, the it is necessary to impose on the bending moment —- M,. the edge moment + My. Then at the edge will he My == +m = 0 (4.18) The imposing of the moment + M, will initiate at the edge in the ring’s ditection. The appearance of the addi- ment is equal C0 My M hecause M, = M;, and also displacement of the edge out: side and appearance of the additional ring force * Ny due to the displacement a. Cylindrical shell, Figure 4.2 (a) According to Table, the displacement is si, sigh 4.19 ct Sa ea (4.19) migy = 2 ay = $8 at Zeg,08 7 “2D According to Hooke’s law the displacement is ot. Sgr. Aye x meek = Ht. Ge (4.20) FIGURE 4-2 — Designing schemes of the shells under the influence of vaiable temperature 1 ess of the wal. which changes linearly through the thiek= 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL O20"44 e equalizing (4.19) and (4.20), we obtain wr, = Set@iEte | 2em. _ drat (ey) _ = Br 2 2 ~ Soe Sooty = 4.21 _ Bt ytavT=0 en YTa-) The total ring moment atthe edge is eMa = = My + uMa = = Ma (Ley) = -2a82Gov?) % Btiat (4.22) ae The total ring stress at the free edge is wth ev (4.23) land the meridional stress is oso (4.24) In Equation (4.23) the sign plusis related to the exter nal fibres and sign minus — to the internal. At the re- versed thermal action signs should be changed Atv = 0.3 the maximum ring stress at the edge ac~ cording to formula (4.23) is approximately 25% greater than the stresses determined alter formula (4.16) for the sections located far from the edges. b. Conical shetts On the basis of similar consideration as that under (a), itis possibleto find the total stresses atthe free edge of Tubular Steel Structures — Theory and Design the conical shell under heating, using diagram *t, Figure 4.1 (D. For the conical shell, Figure 4.2 (b) + Silla , EeayT=pE * Bat Ho Gv (4.25) oso (4.26) Stresses at cylindrical shells having hinged support under nonuniform linear heating, according to diagtem Ht, Figure 4.1 1b). With reference tothe above case itis possible to con- clude that 2M, = Oand a, = 0 The displacement M,5!IM = 0 and from the Equa- tion (4.20), it follows that Ny = 0. The bending ring moment Etiat a amy = 2 42 at change of temperature according to the scheme in Fig- ure 42. The ring stress at the edge oa" # Bot (4.28) ‘The support reaction at the edge H will be determined from the equation Mabe * Hoa = 0 (4.29) Reactions H are directed towards the centre at the flow of temperature according to Figure 4.2, and from the centre at the reverse of the flow. The stresses at the points of zona of the edge effect, which are not at the edge itsell, we may find by thesuper~ position of stresses after the formula (4.16), siresses due to the forces H and M, applied atthe edge. The signs of H and M are necessary to be used accordingly in corres- pondence with the actual direction of the thermal flow. REFERENCES — CHAPTER 4 [4.1] Timoshenko, 5.P. and Goodier, J.N., Theory of Elas= ticity. Ard. ed.. McGraw-Hill Book Company, New York, 1970, pp. 433-439, [42] Lossip, E.N., Lileev. ALP, and Sokolov, A.G., Steet Sheet Structuces, State Edition of Literature on C structions and Architecture, Moscow, 1956, (in Rus- sian), pp. 417-419, [43] Kent, CHL, “Thermal Stresses in Thin-Walled Cylin= ders", Transactions of the Am. Soc. of Mech, Engrs. Vol. 75, 1953, pp. 167-180, CHAPTER 5 self-supporting stacks 5.1 Types of Stacks “There are two basic types of steel stacks: self-supporting stacks which together with the foundation should remain Hable uncer all working conditions without additional Support. and guyed stacks in which all external applied Toads are not totally carried by the stack shell and for which puys or stays are provided Lo ensure stability [5.1} Self-supporting stecl stacks are usually made with the upper part cylindrical and with the lower part flared or belled. The height ofthe flare usualy varies from 10 Ye inhejght. Where the height of the flare is / the height of the stack, the diameter of the base with a conical flare will be /spreater than the diameter of the upper part of the stack. The ratio ofthe diameter ofthe base of the flare to the diameter of the stack in well designed stacks varies from 4/3 to 5/4 [5.2] Figure 5.1 shows some of the more common types presently being employed Stee! stacks are either lined or unlined. Independent, self-supporting brick lining should have enclosing steel we BRACED STACK ® © joy LL lo, bands at regular intervals in its height and, in particular, immediately below each changein lining thickness, where the weight of the lining above is applied eccentrically to the section below the offset. Brick lining may be supported at intervals by shelf angles in the steel shell. Linings of high-alumina cement and expanded shale aggregate have also been used extensively. This is pneumatically applied directly to the inside of the shell over mesh reinforcement attached by welded studs to the steel shell. Independent, ‘continuous steel plate linings, using either carbon or corrosion-resistant steel, insulated on the outside of the lining may prove quite effective. 5.2 Design Loads “The stack shell should be designed to resist stresses resulting from the following loads and forces or their combinations when they exist [5.3, 5.4, 5.5, 5.6, 5.7, 5.8}. STIFFENING RING. ANCHOR, SOUS AND BASE RING FIGURE 5.1 — Types of stool stacks, 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL O1e0'd52 Tubular Steel Structures — Theory and Design ‘Self-Supporting Stacks 33 2 Deed toad G+ CVy in piste velocity wind pressure a eee be load where Lia > 109 FORCE COEFFICENTS 2 lee Wind or earthouake forces whichever are grater © = 0027 if Vw in miles per hour 8 & Tempsraureeteets otnvertzaly andcecomeren- Gg. = the gust eet for S vail Ce = theexposure factorfor the appropriate height of the surface or part of the surface A = de = the projected area of the cable in eleva tion Inaddition to these main loads and forces, secondary TOTAL FORCE Fs 6, 4-CyCyA effects should also be considered. These are stresses resulting from support and erection procedures, Aedel le2s |>25. SMOOTH WIRES, RODS,PIPES © | 12 |05 5.2.1 Dead Load 5.2.4 Wind Forces Dead loads are computed on the basis of the unit weights of the materials. Dead load consists of the particular weight of the shel, linings, ladders with safety rings and any permanent equipment. The structure should be divid- ‘ed into units or sections for weight tabulations, Divisions can be made at convenient changes in taper, cross-section, oF plate thickness, at lining changes, and s0 as to avoid any one segment being too long. 5.2.2 Live Load According to the German Specification DIN 4133 [5.5] ‘The stack should be designed forall loads and) or defleo- tions caused by wind on the stack calculated in accord ance with ACI307 “Specification for the Design and Con Struction of Reinforeed Concrete Chimneys" (5.6), the National Building Code of Canada[38]. or ather specified wind loads. The force of the wind on a stack of cireulareross-seo- tion shall be taken asthe resultant horizontal pressure on its projected area considered as bands of uniform loading applied to the stack within height zones. This force should not be less than the design pressure specified by the Figure 53 (58. Total wind force acting on the stack is MOD, SMOOTH WIRES, RODS 0 FINE WIRE CABLES @ 9902-aHI-£0 er THick wire cases = @® Be: FIGURE 5.2 — Force coellicient Cq for poles, rods and wires. live load, including snow load should be 50 psf, considering platforms and stairs. In addition, any part of the F = Gy.9.0y-C,h (52) stack, platform and stairs may be loaded withthe vertical concentrated load of 650 Ibs at locations producing crit- where cal effect. All parts of the structure which have projec- tions, such as canopies, which are at an angle of 30 degrees SWdd ‘ld Woe TOTAL FORCE F= Cyq + Cy CyrA WHERE Ashed Cy = force coefficient dV/q > 2.5, used according Cy» FORCE COEFFICIENT FOR d¥q>2.5 (RAH OL or less with the horizontal, with the exception of the lad- to Figure 5.2 der step or rings, should be designed using a vertical con- @ = the reference velocity pressure SLENDERNESS h/d= —e [25] 7 | 4 sented Toad of 220 hs and a dsrbuted loading of 6 Cg. = the gust effet factor Pst, acting simultaneously. For the design of horizontal : CROSS SECT, AND ROUGHNESS Cr | Gp | C, failings, the horizontal uniformly distributed load of 30 Ce = the exposure factor a {En Ibs/ft should be applied, A. = dh = the projected arca ofthe stack in eleva MODERATELY SMOOTH {it is suggested that for American conditions, the tion inetol timber,conerete) [07 |0-6 |0.5 applied snow load, s, bes = Csg, where Cs is equal 100.8 facie ioteeter blake aaok Ce g IROUGHT SURFACE rT ‘and gis the ground snow load as specified in American : a URFACE [es fos for National Building Code [5.9]. S should be used todeter- hh = height of the stack, ft Berea! {rounded ribs h=_2¥ed) mine the applied snow load on catwalks, roofs canopies, 21 VERY ROUGH SURFACE spiral access stars, sampling platforms and whatever 2.5 Stack Vibration (sharp ribs h= 8%) | 1.2/1.0 | 08 otter trutures ae attached to the stackand are capable °°? 8 Se oS - , SHOOTHEANDTROUGH t of transferring loads to the stack structure tack shells are prone to oscillations due to wind action 14 }12 | 10 : (5.10, 5.11, 5.12, 5.13, 5.14, 5.15, 5.16, 5.17, 5.18, 5.19, Surface sharp edges l 520 SU, 522, $23, 5.24, 525, 826 $27, 5.28, 828, GpeBXTERNAL PRESS COEFF. FOR eb/@S35 AND MODERATELY SMOOTH SURFACE 5.23 Icing ; In locations where the strongest winds and icing may occur simultaneously, structural memebers, cables and ropes must be calculated assuming an ice covering based fn climatic and local experience. The total force, Fp, for the guyed cable as specified in Supplement No. 4 of NBC [5.8] is given by the following equation In order to guard against these itis necessary to eal late the natural frequency of the shell and to compare it with the frequency of wind vortex shedding, inducted cither by the stack itself, or by adjacent structures, at steady wind speeds of the order of 10 t0 40 mph. Vibration is manifested in rhythmic swaying to large w]e] [| [oe [os] a0 10.8 |+0.1 |-0.9 25 |-28|-18 [os [07 Pel amplitudes or transverse oscillations and in rhythmic pe {*10 [0.8 | 40.1 |-0.8 |= 17 | -22)-2.2]-1.7 |-08 |-08 Oteo'd ovelling of the chimney, both of which occur at right i Fat yl ge Sg ® (5:1) gles to the direction of the wind. This phenomenon 1 | ape fet |roe feos |-o7)-12 |-1.9)-17 |-12 [07 |-05|-0.4)-0.2)-04 attributed to resonance of the chimney with respect to where eddies, called von Karman vortices, formed in the ait- pT Cpa Cg Ce STACK FULLY OPERATING C;; Cq_ = force coefficient, depending on the Reynolds’ stream at regular intervals on alternate sides of the chim . number Re which is expressed by dq, where ney. If the natural frequency of the stack as a cantilever, Pes Sea + Cg: Ce STACK THROTTLED _Cyi=-0.8 dis the diameter of eableinfeet and q isthe approximates the frequency of the eddy formation, reso- ‘elocty wind pressure in pst. Figure 3.2 pro- nance malead to major vibrations despite the small mage vides values of Cn for poles, rods and cables. nitude of the forces imposed by the vortex tral aud 6 == Wee coemetertt@t'd WLOL sa Tubular Steel Structures ~ Theory and Design 4 pSupporting Stacks 55 5.2.5.1 Transverse Oscillations ieee er DEFLECTED pep, 170 tat (5.12) British Standards of Steel Chimneys provides the follow- jr posmon “sn were p= p> (59) Asn i" ing data (5.3} So long as the self weight of the stack is small com= pared to the critical buckling load (ic. the critical buck~ Ting load for the stack considered as a vertical cantilever, without regard to the local buckling of the shell wall) the natural frequeney f for a cantilever stack of constant section is given by: a ty = 2/5 6.3) where {= natural frequency of stack, eyeles/sec. H-= height of stack, ft E = modulus of elasticity of shell material at oper~ ating temperature Ibs/ft? T= moment of inertia of stack section, ft W = weight per unit length of stack, Ib/ft 8 = gravitational constant, 32 fi/sec? isa constant depending on the mode of vibration, For the first mode (the natural frequency), c= 3515 In calculation for lined stacks the section properties of an equivalent steel section may be calculated using a suitable value for the modular ratio, e.g, Esteet Flining The natural frequeney is 54) where W = thedead load weight of the section applied as a lateral load, and Xi = the deflection at its centroid, which will be produced if the stack were considered to bea vertical cantilever. ‘The natural frequeney (ist mode) for a stack of vary- ing cross-sections can be calculated by dividing the stack into a number of sections, Figure 5.4 According to the British Standard (5.3), “The maximum deflection of steel chimney shall be limited to 1/200 of their unsupported length. This deflection shall be calculated on the design thickness alone without including the corrosion allowance.” ‘An approximate formula is given by Dickey [5.31] for unlined steel stacks of straight or flared shape at normal temperature £, = 2520 D(+LA.dae) A, (5.5) where the equivalent height is aS Eu'x's 4% #109 Xp nn AGS FIGURE 5.4 — Weights of sections and thelr deflections. , nays Bek, Hor Bet gee ft 66 D = diameter, ft ‘ hickness of shell wall at bottom of straight section, ft ‘The natural frequeney of a tapered stack, Figures 5.1 (b) and 5.1 (¢) cam be determined by the formula given by Dockstader [5.32] 3.52 D, Fee n 6 ae Fp = natural frequeney, eps sravity acceleration = 386 in/sec? E = modulus of elasticity, psi ws = unit weight of steel shell = 0.283 Ibyin? he equivalent height, fe De = the equivalent diameter, ft The natural frequency is determined by using & straight eylinder of equivalent stiffness. The diameter De and height He of the equivalent eylinder are given by the following equations From Figure 5.1 (b) 2D, 2 Hert Hise) where Dp= > G8) and from Figure 5.1 (c) 5.25.2 Effect of Vortex Shedding nder exposed structural elements such as tacks should en faust reo the effec of vortex shedding, Miueture may be considered slender inthis context i the fatio of height to diameter exceeds 5. The filial wind speed at which vortex shedding may oocur can be calculated from the following formula, fp v2 (5.10) where f= natural frequency of stack, eycles/see. = stack diameter, ft $= 0.20 Stroubal number V__ = eeritical wind speed, ft/sec. ‘The resonant wind velocity can be determined by v= Me 3,600 57280 ‘res ~ 022° 6 32D, 6.1) Where a steel stack is lined and the weight of the lin ing is carried by the shell, an added mass must be set in vibration by the vortex trail. Since there is no major in crease in the moment of inertia ofthe cantilever, the lined stack will have a lower natural frequency than an unlined stack of the same dimensions and will be relevant at lower wind velocities. The approximate natural frequency and resonant wind velocity of a stack with a uniform weight of lining throughout its height can be computed by Eqs. (5.10) and (5.1) provided the value of ws is properly determined. For this purpose, ws may be taken as the weight of the shell plus lining, divided by the thickness of the plate. The weight and thickness should be taken at about 1/4 the actual height above the base 5.25.3 Ovalling Oscillations In addition o swaying oscillations of sacks, vortices also tend to produce diametric oscillations. This phenome non, referred to as “ovalling” or “breathing”, has been observed with the largest amplitude atthe top of the stack 15.33, 5.34, 5.35, 5.36, 5.37]. This type of motion has generally been seen on stacks before lining was in place and consists of a rhythmic change in the circular cross-section ofthe stack where the diameter alternately increases and decreases by several inches. This phenomenon occurs separatcly or in combination with a bending oscillation, ‘The ovalling has cecurred at twice the wind velocity indicated by the Stroukal number and may have invoived “subharmonic excitation” in which the exciting fore acted at twice the Frequency of the oscillation, Itisevident that two cycles of ovaliing occurred during one vortex cycle. According tothe British Standard "Stecl Chimneys {5.3}, the frequency of ovalling ofa ciular stack of con- Saat section is given by frequency of ovalling (eycles/ sec) thickness of stell ({t) R= radius of stack (ft) The critical wind speed at which ovalling may oceur ‘can be calculated from the following formula ves, By 5.13) If the critical wind speed V is fess than or equal tothe steady wind speed, then it may be assumed that the shell will be prone to oscillations. ‘A simple practical cure against ovalting is to use stif= fening rings near the top of the stack. 5.2.6 Earthquake Forees Stacks in earthquake areis shall be designed and constructed to withstand a6 a minimum the lateral seismic forces, assuming that the forces can act in any direction (5.38, 5.39, 5.40, 5.41, 5.42, 5:43, $44, 5.43). The project specification should state applicable earthquake coefficient 2(5.46). With the availability 9 the high-speed digital eom- puter, stacks can now be analyzed for accelerograms of actual earthquake records, Although a rational analysis based on model response calculations is recommended for final design, certain procedures that are adaptable to hand computations are useful in preliminary design, The most thoroughly studied and generally applied earthquake design criteria in the United States are those developed by the Structural Engincers Association of California in 1967 [5.7}. The SEAOC Code is of broad scope and has included criteria for “structutes other than buildings”. Stacks fall into this eategory. The Code as applied to stacks provides the basic criteria that define forces, shears, and moments 5.2.6.1 Base Shear ‘The base shear of total lateral force, Figure 5.5, shall be computed according to the National Standard Building Code [5.46] by the following equation V=SUCW or V=ZUCW, (5.14) where Y= total shear at the base Z = zone coefficient for earthquake magnitude from the map of the United States showing the zones of approximately equal seismic probability in the latest edition of the Uni- form Building Code. This coefficient shall not be less than 0. Tor Zone 1,0,5 for Zone 2, and 1.0 for Zone U = use factor and varies from 1.3 to 20, C= numerical cosificient for base shear which is equivalent to KC in the Uniform Building Code 9902-aHI-£0 oT Be: SWdd ‘ld Woe (RAH OL eVet'd56 Tubular Steel Structures — Theory and Design —_— Vv FIGURE 5.5 — Base shear. W = total weight of stack without lining, Ib W, = total weight of stack including corbel sup- Porting lining, Ib, ‘The value of C shall be computed by Equation (5.15) c= Ons (5.15) The value ofthe fundamental period of vibration for 4 cantilevered structure of uniform cross-section may be expressed by the following formula a r= 1.79 For steel stack this can be transformed as p= 76s x 1078 es (ED (5.16) 17) fundamental period in seconds height, t iameter, ft weight per unit height, Ibs/ft shell thickness, inches modulus of elasticity, psf moment of inertia of the cross-section, fit acceleration of gravity = 32.2 ft/sec? If the base of the steel stack is flared, the period of such a stack can be calculated by Tsar (5.18) where 5 = the calculated deflection in feet at the top of thestack due to 100% of its weight applied as a lateral load Ifthe stack is lined, the weight per foot must include the weight of the lining, and the added stiffness due to the lining must be determined as an equivalent thickness of stecl. The modified thickness of steel then should be used in the appropriate formula given above. This has a good correlation with measured periods of lined stacks. Gunite lining, for example, ean reduce the period of steel stack in the order of 15 Mitchell [5.48] developed a form of solution which is ‘especially useful in calculating fundamental periods of cylindrical, tapered-cylindrical, and step-tapered-cylindrical structures. = cHLy (ESE t= Cygg]?/ Ba ae (5.19) where T = fundamental period in seconds H_ = height in feet Ww = weight per foot of height over a uniform, or assumed uniform section of the structure in pounds per foot P= concentrated loads that may be attached to the structure at any levels, which add mass but do not contribute to the stiffness E = modulus of elasticity in psi D__ = diameter of each section in fet thickness of cach section of shell in inches difference in thea values between the topand bottom of each section constant atthe seetion at which concentrated loads P are attached Ailerence in the 7. values between the top fand bottom of each section 2 = summation of the product of the quantities shown over the height of the structure Values of the a. § and 7 in the function ofthe ratio hhy/ Have given in Table 3.1 = 5 FIGURE 5.5 — Distrbution of base shear ‘TABLE 8.1 — Values of 9, and 7 In the function of hx/Hl 7 cy 7 @ 6 m 8 7 2108 3347 77000000 050 0.1088 0.9863 ‘95ers 2o21 B21 +1.000000, 0.49 0.0986 0.9210 ogstas 941 7.808 +1.000000, 048 0.0909 ose ossess 11.863 7878 “1090000, oa? 0.0826 o79e7 ossi89 4767 7.461 1.090000, a6 ora arse o.g9e6s ane 7.248 o.s96808 a5 0.0878) 876 ‘oes0e7 042 7037 o.spe808 oa 0612 0.5361 192095, 1573 6.830 0.980807 043 0.0551 05372 osres4 1508 5.626 o.g9ee64 042 0.0498 0.5409 o9n73 1.440 8.425 o.seeee oat oaae 0.4971 90843 4377 0227 o.g9ec82 00 0.0385; 04567 ‘o.s0e7e 1316 8.032 o.g00671 038 0.0361 oaie7 oases 1256 5.840 o.956886 038 ost 0.3801 9.88003 1199 5.652 o.s5esc4 037 0275 0.3486 87033 143 5.467 0.996805 0.38 o.caa2 oie o.set23 71.080 5.285 0.996867 038 ooze 0.2833 ord 11938 5.108 0.906817 034 018s, 0.2582 ozaos2 0.938 4.950 o.909754 033 ote 0.2291 82601 0.939 4758 o.s0eer4 032 ootao 0.2050 osiri0 0.802 4500 0.900576 031 0.0120 0.1826 0.80059 0.887 “aa24 0.900865 030 o.oroes | 016200 o7014 0.804 4261 2.900309, 028 o.oos7es | 014308 ore 0762 4102 0.06133 028 coors | ai2s76 0.7682 0722 3.966 0.86823 027 0.006229 | 0.10007 7480 0.683 3.794 0.996676 028 0.005222 | _o.08s64 0.7321 0.646 3646 ‘o.996385 025 ‘o.oo1s2 | 008267 07185 9810 3.499 o.s96047 024 o.o03s64 | 07101 0.6961 0576 3.386 0.907656 023 .002807 0.06056 0.6800, 0543 3217 0.997208 032 ooc2s19 | a0sr26 2.6610 ost 3.081 o.s06e89 021 coors | a04s08 05813 0481 209 0.986101 020 0.001485 0.6207 0.453 2020 0.995434 018 9.001159 0.5002 0.425 2694 o.s04681 018 .000863 osres 0.399 Ber .ossea4 on? 0.000677 08536 0374 2.482 o.902085 016 8.000604 0.8285 0.3407 2.3065 0.99169 015 0.000068 ‘0.5064 0.3269 22240 0.88065 014 .000263 oares 0.3082 218 o.sa0as 033 000183 asia 2246 2.0089 saree ore .ooot2a 4201 2650 49062 0.98630 ont 0.000081 0.3940 0.2866 1068 0.98455 0.10 0.000051 0.3699) 0.2288 17107 0.88262 908 .0c0089 | 0.00165 03327 02122 ve177 0.98052 0.08 0.000017 | aoot0s 0.3003, 0.1965 15278 eres 007 8.000009 | —a.00082 0.2669 or8%6 14813 097573 0.08 .oco004 | 0.00094 0.2325, 0.1676 13579 0.97301, 005 ‘o.000002 | 0.00016 0.1965, 11565, 125 o.s7007 9.04 9.000001 ‘.00007 1587 01421 42002 .96603 a3 ‘9.000000 | 9.00002 ox216 0.1906 1259 o.9eses one ‘.e00000 | 0.00000, ‘.oazs o.r196 10547 0.95673 oot .ccoco | 0.900000, o.o4ts 9902-aHI-£0 Teet SWdd ‘ld Woe (RAH OL ette'd58 Tubular Steel Structures — Theory and Design 5.2.6.2 Distribution of Lateral Forces The following Figure 5.6 indicates how the design base shear is distributed up through the structure as equivalent lateral static forces. First, if the height-to-width ratio of the structure is greater than 3, then part of the base shear, up to 0.15 V ‘maximum, is arbitrarily assigned to the top of the structure. The purpose of this is to increase the shear in the upper parts of the structure to account for the increasing igher modes of vibration in the re- nized that the fundamental mode deflection curve departs from the assumed straight line as the structure deflection ig due more to bending action rather than shcar deflee- tion. This tends to cause greater shears in the top of the structure than would result from the so-called “triangular distribution, 7 (520) Re (521) ., eT RSV for “BS 6.2 (5.22) ¢ ) Whe Pye WR) . Sb wh (5.23) der tt vent Ey (524) yer ret Ee, tet Em (5.25) Most stack structures have height-to-width ratios greater than 6.6; hence, generally, the distribution of the base shear will be 0.15 V as F; at the top, with the remaining 0.85 V distributed in accordance with the triangular distribution formula (5.23). In applying this formula, the structure arbitrarily can be divided into convenient incre- ments, depending somewhat upon the weight distribution but usually not more than 10 sections. The transverse design shear at any horizontal plane in the structure is the cumulative sum of the lateral forces above that plane, resulting from the distribution of the base shear by the criteria given by formulas (5.20) and (5.23). Formula (5.24) states tis for any plane x above the base, giving V. Formula (5.25) provides the corresponding value at the base V, which of course, should check the base shear that previously has been distributed and now is re-accumu- lated 5.2.6.2 Overturning Momenis Having determined the base shear and the distribution of that base shear, the remaining basic design criteria relate 015 M FIGURE 5.7 — Overturning moments. to the overturning moment on the structure as a whole and at any horizontal plane x, Figure 5.7, [5.49], fe * FyLD-1SV(h-D,) + E Py (hy-BI7 (5.26) in which the value of Jx is determined from wi ag o4 ang (29 where oe (8.28) and J shall be not less than 0.45 nor greater than 1.0. 5.2.6.4 Supports I it is expected that conditions at supports may change. this should be considered at the design stage [5.5] 5.2.6.5 Erection Conditions Loads and forees which may appear under the erection conditions, for instance, temporary supports or jacking, should be considered at design stage (5.5). 5.2.7 Temperature Influence The influence of the temperature variations on the design of steel chimneys is proposed following German Stand- ards DIN 4133 [5.5] a. Permissible remperature ‘The influence of the temperature should be considered in the analysis of shell if the temperature is high- cr than + 50°C. b. Thermal forces Provision shall be made for stresses resulting in temperature. The rise of + 35°C and fall of - 35°C in temperature shall be figures from an assumed temperature at the time of erection, In the case of changes of temperature higher than + 50°C or lower than ~ 30°C, the difference of both temperatures should be considered in the design ees aj-Supporting Stacks “5.3 Stress Analysis for Self-Supporting Stacks “The self-supporting stack may be treated as. vertical can- The seipeam subjected to loads caused by the horizontal ser rensure and the weight ofthe stack Tt may be ana- wind by beam theory forthe purpose of determining the ‘ythtng longitudinal stresses, The ony eror introduced {i the analysis of stack shell by the beam theory, as op- in oto shell analysis, isthe asumption that plane sur pon emain planer and the stresses are proportional to feet ptances ftom the neutral axis, However, the error In duced by this assumption is small and conservative ‘The maximum unit stress is usually compression and court dur to the wind load, An allowance for posible Sccipquake stress must be made or thestacks to becon- Structed in certain localities Indesigninga horizontal joint, the means of fastening roost be able to transmit the maximum compressive tresses ovcutringin any pat ofthe joint without exceed- {ng the allowable unit stresses in bearing and shear. The goss thickness ofthe plate must be able tosafely take the Eompressive stress, while the effective thickness of the plate must be able to safely take the maximum tensile Eres in any part of the joint "The stretes in the stack shell are (550, 5.51 a. The Compressive Stress resulting from dead loads including the weight of the shell itself, the weight of the lining and attached equipment. ‘Also, stresses caused by an eccentricity resulting from irregular load distribution. b._ Stresses resulting from bending moments caused by ‘wind or seismic loads acting on the stack. ‘Shearing siresses due to wind or earthquake. Thermal stresses. Stresses resulting from fabrication procedures such as cold forming and welding Stresses resulting from erection. & Allowance should be included for higher resulting stresses due to corrosion. 5.3.1 Dead Loads Cémpressive stresses caused by dead loads may be considered in two parts for convenience: a. Suress induced by shell and lining Weight of steel in chimney having a constant shell thickness in pounds W = Ey DytHtog, abs (529) ‘The stress due to the dead weight of shell My £7 apis Pst 630) ‘Ws. = weight of shell, Ibs inside diameter of shell, ft {= thickness of shell, ins 2 H = height of the stack, ft Ps = density of shell material, 490 lb/ft? for steel © = corrosion allowance 7 ‘Weight of lining, in pounds mytEemtty B G3n My (5.32) £,* myterey Pet where We = weight of lining, Ibs Dg = mean diameter of lining, ft tg = lining thickness, ins Hg = height of the lining, ft %e = density of lining material, Ib/f t= thickness of shell, ins b. Stress induced by eccentric loadings In stacks, externally attached equipment such as: platforms, ladders, pipes, etc., act eccentrically and ‘may be reduced to vertical loads acting along the shell axis and moments acting eccentrically with respect to the shell axis, If the total vertical equipment load is EG, then the combined moment is Me = DGe and the Compressive and tensile stresses due to the eccen- tic loads are an, psi (5.33) fe" pif t wyte-oy £ where fe = unit stress in extreme fiber or the shell, psi @ = eccentricity, the distance from the shell axi to the resulting vertical load 2G DE (e-e) 5.3.2 Wind or Earthquake Loads a. Stresses induced by wind loads In general, a selfsupported steel stack may be consi ered as a vertical cantilever beam subjected to the forizontal wind pressure. According to formula (.2), the total wind foree act ingonthe stackis F=Cy"q"Cq"Ce"A. Therefore, the maximum bending tnoment is nerd 634) and the stress due to wind is eee Fant kee) P88) b. Stresses induced by earthquake loads ‘According to formula (5.26) the total bending mo- ‘ment under earthquake force acting on the stack is We gf01svaah) + F ray-ho] 630 sx and the stress due to earthquake is eee (0.25v(h, ~A) fet o.asving-h) + BO wee) ne co +E wy tah od ah, OY 9902-aHI-£0 Teet SWdd ‘ld Woe (RAH OL OVeo'd5-10 Tubular Steel Structures — Theory and Design 5.4 Design of Self-Supporting Stacks 5.4.1 Dead Load Stress According to formulas (5.30), (5.32) and (5.33) fon" fe t fy t tee (5.38) a, tw, +26 at 1 BBE yee pitt sD? (t-e) 5.4.2 Dead Loads and Wind Stresses According to formulas (5.30), (5.32), (5.33) and (5.35) ttt ey 2rH HDF (eo) (5.39) 5.4.3 Dead Loads and Earthquake Force Stresses Accarding to formulas (5.30), (5.32), (5.33) and (5.37) fous * fs fot tee My tmyt to aM, 7 ) eae E Py ‘TD? (t-c) (5.40) 4s 2 +E foasweneh) + Try thyhy)] wD (tec) dex * 5.4.4 The Allowable Stresses The allowable buckling stresses for mild steel are used according to Equation (2.10). At the condition 2 361, then according to the formula 2.15) = 58 ket According to the German specification DIN for design of steel stacks [5.5], the allowable buckling stress can be estimated by using the following formula Fatiow = red) (41) where the coefficient K is xe, ate (5.42) and “e” js the eccentricity of the cross-section of the stack due to fabrication imperfections e = gety 20 + py > ey in (543) mt roo) > 5.4.5 Checking of the Wind Induced Oscillations The resonant wind velocity for the stack configuration of Figure 5.1 (a) and 5.1 (b) is determined according to the formula (5.11) Yees = 3-20F, where the first mode of vibration is after formula (5.7) 3.520, wea For the stack configuration of Figure 5.1 (c) D should be taken as the actual diameter at 1/8 the height below the top. The critical wind speed at which ovalling may occur is calculated following the formula (5.13) 22 Yer * fora I the critical wind speed V is less than or equal to the steady wind speed, than it may be assumed that the shell will be prone to oscillations, 5.5 Base Design To analyze the stresses between the base ofthe stack and its foundation, itis important to consider the degree of fixity of the base, which depends on the connection de- tails, eis factor which allows or this degree of fixity and is graphically explained in Figure 5.8 5.5.1 Design of Anchor Bolts In designing anchor bolts fora self-supporting steel stack, the weight of the lining is not considered as the steel work is usually built first and the masonry lining added after wards. Sometimes a considerable portion of the lining is removed and renewed during the life of the stack. This ‘means that the anchor bolts must be large enough to keep. the stack Irom overturning before the lining is removed and renewed during the life ofthe stack. ‘The load acting on the stack is transferred either as a compressive or tensile load to the concrete footing through the anchor bolts. The bending moment M and the weight of the stack W results in a loading condition in the concrete footingsimilarto that as shown in Figure 5.9. In the calculation it isassumed that the bolt ring isthe center of the bearing plate. The moment and weight of the stack result in a tensile load on the left-side anchor bolts, fand a compression load on the right-side [5.52]. Calling kD the distance between the neutral axis and the mean circumference on compression side, asshown in Figure 5.10, we have by similar triangles ng, cease (5.44) QOQOQ ky Sg SS PONT CONTACT rhe + - has wit Ww War Ago eet +075 €=05 E2025 Ex0 Ace AREA OF CONTACT AND Aya TOTAL POSSIBLE AREA OF CONTACT [COMPRESSION ZONE) FIGURE 5.8 — 35 2 lunction of base area contact of stack where n= <2= the ratio of moduli or elasticities of si€el to the concrete. From Figure 5.10, the location of the neutral axis n-n may be defined in terms of angle a eee fel (5.45) DR oe 1. 26 ‘The total force, T, on the tension side of the sections (r-a) (cose + cosa) etfs ““Uireesay 1? tren sina + (n-a)cos: caveaay (sine + ( 6.46) Since any given position of the neutral axis deter mines a, this equation may take the form 7 gates (547) in which Cy is a constant fora given position of the neu tral axis and shown in Table $ ‘The moment of the total tensile force, neutral axis is (rad Mp = 2 jf tye, ElS0S8 + cont) tay T2088) 2 qooay [t-adcosta+ = erty 7 3 eee + Feinacose + bina) 7 a Dividing M7 by T, we have as the distance of the center of tension from the neutral axis, 9902-aHI-£0 Teet SWdd ‘ld Woe (RAH OL oLea'dS13 WIND — ome STEELIN TENSION. CONCRETE IN ‘COMPRESSION aa FIGURE 5.10 — Load on anchor bolts and bearing plate. task reat (r-ayeos*a +} sinacosa + F(raa ay EO 3 Boa) (sina + (r-a)coss] (6.49) Similarly, we determine the total compression force C= Cpr (te + ney) 6.50) in which Cp is a constant for a given position of the neutral axis shown in Table 5.2 ‘The moment of the total compressive force C about the neutral axis is 12 Mom (Be + BES)ECK! Crmcoaay si) acoe’s ~ 2 2 (ucos*s - } sinacose + da] TABLE 5.2 — Value of & Cp, Cr Zand Funaton of k op or 2 1 as) [~b600 | Sans) paso} are oso | bese | 2357 | bas | 766 oso | to | 272 | dase | arr | s2o | tee | 256 | pase | O70 0250 | 1370 ass | 0779 o%00 | 0 Sass | Oren oa | jee daar | Ore a400_| 1368 Baie _| ee 0450 [84 | 40 [705 9300 | 2000 2300 | oze8 os | ire Sas | oes Sa00_| 2294 Sas | aes and the distance of the center of compression from the neutral axis is 3 1 Ny _ Cacos?a - 3 sinacosa + $a) (5 59) Ap go = ——etns = aeoaay ‘The system of forees, as shown in Figure 5.10, must be in equilibrium. Hence. taking moment ab out the foree PL we may write TD = MPD But Cysts Therefore CyfertgjD = M ~ PD Whence ‘The total area of steel requited is Ag = 2rty Therefore 2n (P20) "CED (5.53) From Table 5.1, it may be seen that the constant j ‘changes but slightly for a considerable variation in the position of the neutral axis, Taking = 8 for all cases, Equation (5.53) may be a gmep2D) oT (5.54) Applying now the condition that the summation of ail vertical forces must be zero, we have c-T=P Substituting the values of C and T as previously found, we Cpfgt (te + Mts) - Cyfgtt, 2 P Solving for te, we obtain P+ (Cpl, ~ Cpfomirt, te oe EE 555) ‘The number of anchor bolts for a self-supporting steel stack should never be less than 8 and should prefer ably be 10 or 12 or more depending on the size of the stack. Generally, for given values of M and P, and assumed. number and cross-sections of anchor bolts itis required to determine the maximum stresses in the anchor bolts fy and the concrete under compression fe. The problem is solved by a method of successive trials, since the position of the neutral axis is not known, The procedure is as follows: |. Assume a position of the neutral axis, select the constants accordingly, substitute into Equations (5.54) and (5.55) and solve them for fy and fe Check the position of the neutral axis as fixed by these values of fy and fe is the same as the position assumed at the start, If the two positions agree, then fg and fe as found are the actual stresses. If not, a new position of the neutral axis must be assumed, new constants selected, and new valucs of fy and fe computed. Thus.a series of trials must be made tuntil the location of the neutral axis as assumed is consistent with the computed values of fe and fo, 5.5.2 Base te 5.5.2.1 Base Plates Without Gussets formly loaded cantilever beam with f¢ the uniform load, at the junction of the stack shell and the base plate for unit £8? Mercer A 66 The maximum stress in an elemental strip of unit wiath is cw Max, Hott 659 ieee pete ee where t, is the base-plate thicknessin inches. Letting fmax. = fall and solving for t, gives to WT 5.5.2.2 Base Plates With Gussets If gussets are used to stiffen the base plates, the loading conditions on the section of the plate between two gussets may be considered to act similarly to that of a rectangular uniformly loaded plate with two opposite edges simply supported by the gussets, the third edge joined to the shell, and the fourth and outer edge free. For this particular case Timoshenko and Woinowsky-Kreiger {5.53} have tabulated the deflections and bending moments as shown in Table 5.3. To determine the base-plate thickness from the bending moment the following formula should be used 2 Je ‘all (5.58) (5.59) Note that in Table 5.3, for the case where L/b = 0 (no gussets or gusset spacing b = e) the bending moment is reduced to Equation (5.56), and the thickness of the base plate is determined by Equation (5.58). Also when Lb is equal to or less than %, the maximum bending moment occurs at the junction withthe shel! because of cantilever action. If jb is greater than %, the maximum bending moment occurs at the middie of the free edge. 5.5.2.3 Practical Considerations in Designing Base Plates Rolled-angle base plates may be used for stacks ifthe calculated thickness of the base plates is 4 in, or less. The steel angle is rolled to fit as shown in Figure 5.11 [5.54]. If the required base-plate thickness is '4 in. to in. a design using a single-ring base plate may be used as shown in Figure 5.12. ‘TABLE 5 — Maximum Bending Moments In 9 ‘Base Plate with Gusset, Tweet) | i) | ; > a 1 0.0972teb? 0.1198 e 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL Oresel Tubular Steel Structures ~ Theory and Design self-Supporting Stacks S15 ‘TABLE 5.4 — Number of Chae for Varios Size Stack Diameter, SKIRT FULL FILLET WELD Stack Diameter ft No.of Chak aie cussey pute 3 < ON EITHER SIDE OF ANCHOR BOLTS Z 7 6 2 1c FLLET z 8 ee 8 18 PLAN VIEW 88 tu3's3/8" OR LARGER 3 2 7 10 2 ca BOLT SIZE+ 1/8 FIGURE 5.11 — Rolled-angle base plate. Ifthe required base-plate thickness is ¥, in, or greater a bolting “chair” may be used as shown in Figures 5.13 and 5.14 Although the number and size of bolts required should be checked for each individual design, some typi cal values of maximum numbers of chairs can be obtained from Table 5.4 for a given stack base diameter. When checking the base-plate thickness for a centered chair, Figure 5.13, the plate inside the stiffeners is considered to act as a concentrated loaded beam with fixed ends. The concentrated load, P, produced by the Ddoltis equal to the maximum bolt stress multiplied by the bolt root thread area, ‘The maximum moment in the base plate occurs on the line of symmetry centered inside the chairand is given by Bb Mnax ” “F (5.60) The anchor bolt hole reduces the effective width of the plate. Taking this into consideration, the base-plate thickness, tis os fenaey =r (5.61) where by is the width of the base plate, bhd is the bolt-hole diameter and fall is the allowable stress in psi If the number of bolts required is greater than that given in Table 5.4, an external bolting chair may be used, as shown in Figure 5.14, aren Et sack oF ANCHOR Sous “WURLLET WELD BO SIZE + 1/8" La FIGURE 5.12 — Single ring base plate. With reference to Figure 5.14, plan view, with yin the radial direction and 2 in the circuraferential direction, the maximum bending moments Mz and Mx are given by my = i Catan Ey a3 - (5.62) +B tomePM 2 oa. 663) = (Q-p-ra) 2) where H_ = Poisson's ratio (0.30 for steel) fn = natural logarithm a= radial distance from outside of skirt to bolt sirce, inches {= radial distance from outside of skirt to outer edge of compression plate, inches b= gusset spacing, inches € = radius, of action of concentrated load, inches or one-half distance across flats of bolting nut, inches uy Ty = constants from Table 5.5 when b— unity, Mx = Mz, and when b> 1, Mz>Mxand therefore Mz controls, For the case in which a = (/2 and Mz is controlling, Equation (5.62) reduced to (= eta) bbe ey] (5.64) ‘The maximum stress in the compression ring of unit ‘ TABLE 5.5 — Constants for Moment Calculations. eg] wo [12 [ss [ie] 8] 20] = 7 | 086s | 0.950) 0217 [0.125 | 0073 | 0042 | 0 1 | 0138 | 0.115 | 0.085 | 0087 | 0.087 | 0023 | 0 Note: For a,b, © ess than 1.0 invert BC and rot WASHER THCVESS, LU BRAC Bor Bia * ELEVATION SECTION AA FIGURE 5.19 — Center anchor-bolt chair. BOLT Sze +s" a BO SIZE 1/0" BOT SZE + 8° GUSSET PLATES Noa se ® FIGURE 5.14 — External bolting chat. Where tis the thickness of the compression ring. ‘The moment of inertia of the gusset about the axis of gyration is given by 55.2.4 Design of Gusset Plates for Compression = teas? 666) Rings — the gussets are spaced evenly as shown in Figure 5.15, : they may be considered to behave as vertical columns - (3.67) 15.56), 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL Ot-sa'dTubular Steel Structures — Theory and Design’ where a area of cross-sction, in 1 = radius of gyration, in te = gusset-plate thickness, in € = wideh of gusset in and ge Be 28000 gg 60 « nye « 200 2 a+ (nt/ieooor’) (5.68) where b= height of gusset, in From Equations (5.66), (5.67) and (5.68) we may obtain 18,000 2} ~ (bolt 1oaaye? - Mibope toad) (5.69) Ith is small the third term in the equation may be dis- regarded, therefore simplifying Equation (5.69) to HEAR 620 P, 2/3 1/3 t= LB GE r 7D where 1 = radius of the stack at the point under consideration, inches P= maximum bolt load, pounds 4 = radial distance from outside of stack shell t0 the anchor bolt eile, inches = gusset height, inches ‘A (see Figure 5.14) or bolt spacing 5.6 Approximate Breech Opening Analysis From a functional point of view, steel stacks have to resist the effects of wind and of their own weight. In the case of steel stacks the effects of the wind predominate bya comparison of the influence of their own weight With respect to their structural behavior there are two types of steel stacks: the self-supporting type and uyed stacks. ‘The self-supporting type acts as 2 column resist own weight, acting vertically, and the overturning effect of the wind, the base only being fixed. In other words, i isa cantilever subjected to a relatively large overturning ‘moment and a relatively small vertical load. Guyed stacks resist the same forces, but itis assisted by lateral supports in the form of guyed cables equally spaced around the stack and anchored at the other end to concrete blocks, some distance from the stack footings. ‘The effect of those guy cables is to produce a somewhat smaller wind bending moment in the stack. Near the base of the stack a relatively wide breech ‘opening or flue inlet is cut. To facilitate the inspection and lining of stacks and to enable soot, etc., to be removed, it is usual to fit an access door near the base. AS in the case Of the breech opening, any opening of this nature should be adequately reinforced. ‘The breech opening weakens the wall of the stack considerably and some forms of compensation are required to take the loadingand provide safety against local buckling. In the following discussion the breech opening analy sis and reinforcement design are considered for self-supporting stacks. However, it should be noted that such an analysis and design may be applied to guyed stacks. This is because guyed stacks are affected by such similar toad- ings as self-supporting stacks. Only the vertical load may be increased on account of the vertical reactions of gay cables, In the following are given two methods for the analysis and design of breech openings: one approximate and another more rigorous method. ‘An approximate method may be useful for the preliminary analysis and design and a more rigorous method is recommended for the final design [5.55]. 5.6.1 Conception of the Method Experiments show that the applied load causing the general collapse of the thin-walled cylindrical shell axially. loaded is slightly greater thar load initiating local buckling. Prior to local buckling, a shell is quite sensitive to slight disturbances. When the applied load is just below the local buckling load any lateral disturbance would cause local buckling to occur in the hole region cut in the shell Once the local buckling had occurred, the shell did ‘not seem as sensitive to sligit disturbances because these disturbances seldom led to its general collapse. Buckling loads continued to decrease as the hole dimensions increased. During the experimental tests the loads were applied differently by the testing machines used in shell tests and FIGURE 5.15 — Stack shell wth external bolting chairs. s17 CROSS-SECTION A-A FIGURE 5.16 — Shel and hole geometry. the measured shell buckling loads were reduced to approximate equivalent applied stresses in order to provide 2 common basis for comparison. This was done by replac~ ing the applied load with a svaticalli-equivalent mem brane stress system acting on the plane which may be called she applied stress plane and which is perpendicular to the cylinder axis, Figure 5.16. Therefore, the statically-equivalent applied mem: brane stress system must equilibrate the applied axial load plus a static bending moment due to wind action and ‘eccentricity of axial load with respect to the neutral axis of cross-section To simplify the form of this applied membrane stress system the following assumptions were made: The membrane stress applied at y = *a, Figure5.16, con the hole edge can be used to represent the applied stress corresponding to local buckling of the shell, and “The hole is far enough from the ends of the shell, so that the assumed applied stresses are not influenced by the end support conditions. 5.6.2. Stresses at the Edge of Breech Opening Reduced cross-section of the stack shell is underinfluence of axial load P and wind bending moment Mw. Based on the geometry of the applied stress plane shown in Figure 5.16, the resulting stresses in the wall of the stack are given by pe By Met PHI 57) ios where A= reduced eros-sectional aren of the stack T= moment of inertia ofthe reduced crosses tional arca of the stack the resulting stresses remain negative (compression only), then there will be no anomlly,and the uswal laws of Met llli TT 4 1s 7 a | scscree \ \ \4y : KE AN | =| jeje / | Li FIGURE 5.17 — Cross-section of the stack. Stress distribution diagram 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL O1-s0'ds-I8 Tubular Steel Structures — Theory and Design Statics will be valid. If, however, the stress distribution diagram includes a tensile portion, this should be consid- fered in further analysis, assuming that the position of neutral axis is known, With reference to Figure 5.17, the following nota- tions are introduced, SeST = statical moments of wall under compression aand tension, respectively, referred to the neu tral axis, Ie,IT = moments of inertia of wall under compression and tension, respectively. referred to the neutral axis C.T = total compressive and tensile forees, respee= tively MoMT = moments capable of being resisted by the compressive and tensile sides, respectively. = 11 = the slope of the stress line and 15.73) Then, the compressive force is : ce fean=(2) fees me 1 By analogy, the value of tensile force is . 8p (5.75) Bending moment capable of being resisted by the compressive area is ny fx. ans (2) fers = or, (5.76) cin = (at < By analogy. the value ofthe bending moment capable of being resisted by the tensile are is My = oly (s.) From the condition of the equitibsium of vertical forces, we have P+t-c=0 Pec-T=N a) Total bending moment is MEMS + Mp My + Bx, 6.79) Alter substituting into formula (5.78) values (5.74) and (5.75), we obtain N=C-Tee(SS=S_) (5.80) sand formuta (5.79), using (5.76) and (5.77) may be repre- sented as ™ M+ Pxye 6 gt tg) (581) From (5.81), the value of @ is Ay + Px, oe at (5.82) ‘The value of N determined from the formula (5.80) should be equal to P, or Nee (80+ Sp) =P (5.83) The average compressive stress in the wall may be obtained ‘from the formula We 634) a te ie £ where L is the total length of the compressed wall, Figure 5.17 T= arc BF + arc GH BRO = a + 8) ne2x MOD ess) (585) Using midwall radius rey reo (5.86) we obtain ale + £490 - 2+ 6) Se (87) Therefore, the average compressive stress is 6s, 180 fev ETE ESOT aT aT (S88) The maximum value of the compressive stress occur- ring at the edge of the opening can be approximately determined by considering | inch wide strip at the edge, Figure 2.18 From the diagram of stresses we may find the following ratios a Gar x 1 (5.89) E 5 ge = ox, 2 ‘The maximum stress at the middle of the wall is itt - $ + x, fnax peaee zy riz 6.90) (5.91) : supporting Stacks 5.19 wy 4 5.6.3 Geometrical Properties of the Reduced Cross- Section of the Stack 5.6.3.1 Statical Moment of the Compressed Area ==> Figure 5.19 (y, SS Ordinate of centroid of the sector AOB is 2s sinyy_sinyy : Pao a a 7% ty at Area of the sector AOB Aer san) Staical moment of the sector AOB is ae siny, ein, avg 2 ety (ZEEE es 5.54) and for sector A,OB, 2, sinyy sim, Pt = = ts © sini 95) sc ty, (FEA 4s eins) 90 Total sttical moment i 4 pts 8,7 2(5;°8)) = 4 (ee) stay ainy, + sar(rt-rhy, sing FIGURE 5.18 — Diagram of stresses in the wall of stack. (eter, (5.96) ee y FIGURE 5.19 — Geometric data forthe determination ofstatical moment of compressed area. 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL O20"$20 Tubular Steel Siructures — Theory and Desigh S21 where 1597) (5.98) 5.6.3.2 Statical Moment of Tensile Area: Figure 5.20 Ordinate of centroid of segment DEF 4c con's. ee" STRORE) sin TET (5.99) Area of the segment DEF is aeF foro -sn2e] gon Statical moment of the segment DEF with respect to the neutral axis N-N is A (yg r sing) = aed P (90-8) = sin 28] x 4r_cos*8 ] Tew = sin zayT FP BL = $x conte ~ = sina[2(a0-8) - saad (5.101 By analogy, satial moment of the segment D,,P, rere nierernie rene 5,* 72 (4 2, costs =x sine [ais0-8) - sin 2g ata (5.102) 4 Total statical moment of the tensile area with respect to neutral axis N-N is 2 (ret) costs - y= 84-8, = 2 (x2}) cos" - Ecteepp ein px fargo -s)- sin 24 (5.103) 5.6.3.3. Moment of Inertia of the Compressed Area; Figure 5.21 The moment of inertia of the compressed area with respect to the neutral axis N-N is Wet ac-yP 6.108 FIGURE 5.20 — Geometric data or determination of statical moment of tenalle area. Sf) ne z 8gt8 2 2 [ee costae = wo =F lay, - sin (ass) cos (o-8)] = Wrate where " 2 (5.106) By analogy, the moments of inertia of two sectors, AjOB, with respect to neutral axis x-x are Bagge [21 ~ ain Cot8) co8 (or8)] (5.107 ‘Therefore, the moments of inertia of two rings ‘AA,BB, are Bet Tie Tax * et ep (5.108) b. Determination of the expression Ay (V3 ~ ¥4) Where Ay is the area of the sector AOB, aWhoyp © rin et yy, 7 yD = 4x siny, siny, 4 siny, sin a EE ot 3 = rty, sine (5.109) €. Determination of the expression A; (92, ~ ¥8) Where As is the area of the sector A,OB, Agi, 7 y}) * iy, er + ¥3) Wen - 3) * a ee A stay stay esr (S.110) ‘The resulting value of the expressions (5.109) and (5.110) is Y FIGURE 6.21 — Geometrical data ordetermination of momento! inertia of compressive area. 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL O1e0'd5.22 Tubular Steel Structures — Theory and Desig = Sz ginty = pode sin? (90-8) Yo" Ptaq sin By ~ TEGO 8) > sin Bay yy sing x (11a i (! shay, ainyy :) Secor eee Aydt yf) + ado ype eee eee Moment of inertia of segment EFG with respect to centroid axis xp Xe is (5.111) ae 4 sin? 4 cose aap (i+ Faget 20( SEER) uy Therefore, total moment of inertia of compressed area with respect to the axis N-N is Determination of the expression Aryi,. where A, is the arca of the segment EFG 2 2 Fe = sin 20) 2 ayy = PGeaginzel vy - 2 cose)? = a F(29 - sin 29) [ar int = Bengt) epi a) a From (5.115) and (5.116), we obtain the expres- (eS Fy) pa fen = sin(ate) cos al] + 4 siny, aiay, + retets ety, sing (22M gy seco epiy, sine (A 4) (5.112) 5.6.3.4 Moment of Inertia of the Tensile Area, Fi = + 7 (26 ~ sin 26) a5 sin Toy ¢ : y FIGURE 5.22 ~ Geometrical data or termination of moment of inertia of tensile area a pie a - 2204 ~ sin 29) (s sin’s z as cose] 6.19 From (5.118) and (5.119), we obtain the expression for moment of inertia of segment E,F\G, a + 4 sin’ ¢Cosg| + fey sintg PRs - cose]! (5.120) Therefore, total moment of inertia of tensile area with respect to the neutral axis N-N is obtained by substructing from formula (5.117) the value (5.120), or ign yet # F [las - stn 20 + + A sins cons] = Fen ~ ain 20 oe [29 = sin ap + 4 sing cose] . F] 24 sin 20 | ary sins z SaaS eae Tay z I tet ep =x cosqf =e SD [ae = ss2 20 . 8 sin’ +4 ein’ cons] + 8 tet eb eis - Fe (er) sin? sco + Supporting Stacks 523 women! of inertia of segment E,F(G, wih re ete 4 t0 centroidal axis Xe, ~ Xe, + RE coste (25 = sin 20) ry ae 4s 48in'e cong peta sy 7 Beate ( a seep |B [ae ain 20 + ous Hee 8 sints tbe nes othe epson EF - 45 £3) sin? + ie rFU(2@ - sin 20) aint D . 2 ow (yg. - © cosg)? = AM c. siete et ED corty ce = city Gran After substituting = 90 ~ 8, we obtain + 4 c0s%6 sings 8 sta | 3 (S08) = SEE | 4e (et rf) costs sing + ~ $e ce rp costs sing rtirt rh) + ATE [e009 -8) - sin 28] sinte (5.122) 5.6.4 Stresses at Breech Opening. Experimental Tests In this chapter the effect of rectangular cutouts on the buckling of circular cylinder is discussed based on experi mental investigations. Because of the nonsymmetric nature of a cylinder with a cutout, analytical solutions for the buekling loads of such cylinders are difficult to obtain. This is especially tue for large cutouts where nonlinearities become more pronounced. Consequently, studies of the effect of eut- ‘outs on the buckling of cylindrical shells have generally been limited to a few analytical solutions and to experi mental tests Brogan and Almroth (5.$6, 5.57] carried out a theoretical and experimental investigation of the effect of rectangular cutouts on the buckling loads of cylinders, The analysis was based on a two-dimensional finite-difference scheme, and this numerical solution entails treat- ment of a large system of nonlinear algebraic equations. 5.64 1 The Effect of Square and Rectangular Cut- outs ‘The effects of square and rectangular cutautson the buckling of cylinders loaded by central axial compression force are summarized by the results which ate bounded by the curves shown in Figure 5.23. Test cylinders had shell radius to thickness of the wall ratios r/t = 400, The dashed curve in Figure 5.23 repre- 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL Oteo'd@t'd WLOL 5-24 Tubular Steel Structures — Theory and Design a8, NASTRAN nee 02 FIGURE 5-23 — The effect of equare sents the results of a finite-element analysis. Most of the cylinders tested consisted of specimens with increasingly larger concentric square or rectangular cutouts in their sid2s. A limited number of cylinders had both square and rectangular cutouts with a common center. ‘The available prebuckling analyses aresimilar to that they all provide solutions that are dependent on a nondi- mensional geometric parameter Fe aor er) where T= a sharacteriie hole dimension R= the shell ais the wall thickness @ ® FIGURE 5.24 — Square and rectangular cut in cylinders REPRESENTATIVE SHELL © LOCAL BUCKLING A GENERAL COLLAPSE id rectangular cutouts on the buckling of circular eyinders loaded by central axial compression, SSOUARE COTOUTS BOUNDARIES OF EXPERIMENTAL The characteristic hole dimension in the circulation parameter T is taken with reference to Figures 5.24 & and 5.24 b as follows: ‘The characteristic hole dimension as shown in Figure 5.24 a equal to one-half of the side length for the squares, ror=2 5 Fest (5.120) and for rectangular hole shown in Figure $.24 b ate atb Teer (5.125) Even for such diversely different configurations as longitudinal and circumferential rectangles with aspect ratios equal to 2.0, all the experimental results for the cylinders with squave and rectangular cutouts fill within the relatively narrow scatterband of Figure 3.1 when plotted with respect to # For values of F less than zpproximately 1.2, the buc ling behavior of eylinders with square and rectangular cutouts was much the same as the behavior of eylinders with citcular holes, Figure 525 For values of T greater than about 1.2 the buckling loads continued to dcerease with increasing F, but ata ‘much smaller rate of decline than that for the smaller values of F. For rectangular and square cutouts there was no transitional range as there was for cireular cutouts, and the general collapse of the eylinders was always preceded by a stable local buckling made with aninward postbuckling deformation pate For F between approximately 1.2 and 1.6, the stable local buckling mode always occurred in a symmetrical pattern approximately in the form of an ellipse with its Semimajor axis tangent to the shell circurnference, For F larger than about 1.6, the symmetveal stable local buckling mode was usually preceded by a nonsymmetrical mode aligned with one of the cutout diagonals whieh either snapped into its postbuckling form or just began growing inward as « larger local deformation. For square and longitudinal ectangular cutouts with F greater than approximately 2.5, noticeable symmetric utward prebuckling deformations of the longitudinal For r= py-Supporting Stacks 5.25 ‘ 1 os \ 0.7 \ . \_NasTRAN \— fA RESULIS cutout sides preceded the above nonsymmetrical defor mation, For circumferential rectangles with T greater than about 3.2, noticeable symmetric outward prebuckling deformations of the circumferential cutout sides preceded the nonsymmetrical pattern. In all cases, the difference in the loads for the first noticeable local bucklingand general collapse was small, For F less than approximately 4.5, there was no de~ tectable difference (beyond normal experimental scatter) between the general collapse loads for longitudinally and circumferentially oriented rectangles for the limited num= ber of evlinders tested with each cutout configuration, For f greater than 4.5, the general collapse loads of the circumferentially oriented rectangles tended to oc- cupy the lower portion of the scatter bond in Figure 5.25. ‘This phenomenon appears to be a result of the large pre~ buckling deformations and stable local buckling mode for the eylinder with circumferentially oriented rectangles which cause intensive stress redistribution to occur far- ther around the cylinder circumference. Generally, for a given value off, the general collapse loads for eylinders with square and rectangular cutouts as as CIRCULAR CUTOUTS fu cL 02] = FIGURE 5.26 — The effect of circular and square cutouts on Sucking ofa rela ylinder onde by conta sla compre REPRESENTATIVE SHELL WITH RA 2 400 © LOCAL’ BUCKLING GENERAL COLLAPSE BOUNDARIES OF EXPERIMENTAL RESUS FOR 16 SHELLS were found to be slightly lower than those for eylinders with circular cutouts. Asan example, the difference in the general collapse loads for circular and square cutouts in the same cylinder are shown in Figure 5.26. 5.7 Design of Stiffeners at Breech Opening 5.7.1 Design of Vertical Stiffeners In order to avoid eccentric loading and a change in the direction of the stress, it isnecessary to keep the neutral axis throughout the Compensated cross-section on the centre line of the whole stack. It should be noted that itis not easy to obviate some of the eccentric loading on the compensating members. Such eccentric loading is offset in part by the fact that the line of action of the loading in the stack plates above and below the breech opening, ‘often as far from the centreline of the stack, asthe neutral axis of its compensating member. ‘The statical moment of the removed plate, where 2a is the angle in radians subtended at the stack center by the are removed, Figure 5.27, is tg fonda) seose = 2eR%sina (5126 Using two verte! stile. we my determine the reavired frossacctonal ae of ach se by eal ing smal amen othe cstesetionl wear a the sUtines 2Ay andthe removed pat ofthe such, or 2a,d = 2eR'sing (5.127) from which a, = Biging (5.128) where d is the distance between the centroid of eross-seetion of stiffener and axis «x of the stack 9902-aHI-£0 oT 2b: SWdd ‘ld Woe (RAH OL eVet'd5-26 5.7.1.1 The Load Imposed on Each Stiffener by ‘Steel Shell and Lining 6 = BOE cgtgog + Opt yrs) — (512) where G__ = total load on single stiffener Ds.De = inside diameter of the stack and lining, respectively tot = the thickness ofthe wall and lining, respectively 3,0). = specific weights of the steet and lining materials, respectively H_ = height of the stack above the breech opening, ‘The axial compressive stress in each stiffener, under the weight of the wall and lining above the breech opening is ao ols (os) 0s 5%) 5.7.1.2 The Load Imposed on Each Stiffener Due to Wind Bending Moment Compressive stress in the stack’s wall due to the wind load “ 6.13) woe where My = bending moment due to wind load Dm = mean diameter of the stack at cross-section under consideration Cross-sectional area of the wall of the stack is rat (8.132) “Therefore, the approximate load imposed on each stiff 4, 1.0 xm,ex 385 = woe eee eee a (5.133) and the axial compressive stress in each stiffener under the wind pressure on the stack is ee in TDs sine Hyd 22.5 eA sina 6.3 5.7.1.3 The Eccentric Bending Moment Due to Eccentricity of Stiffener ‘By designating the distance between wall and centroid of, theerossaceion ol thestffener "the ceentrc bending moment due to the loads (G + Py) is My = (6+ Pye (5.135) Tubular Steel Structures — Theory and Design (ane) tL , ELEVATION, FIGURE 5.27 — Breech opening geometry. and corresponding axial stress is wr Pe sa 6 . atite where Ssiff is section modulus of the stiffener cross- section. Total axial stress is f (5.137) Feoe 5.7.14 The Effect of the Stiffened Breech Opening on the Strength of the Stack To evaluate this effect itis necessary to determine the moment of inertia of the removed area of the are with respect to the stack center line, 28 follows, Figure 5.28 “The moment of inertia of the area of the arc removed about a stack center line is Self-Supporting Stacks Weortade = Roe BEBE) (5.138) The approximate moment of inertia of the stiffener cross-section with respect to the stack center line is, Teeiee = Aye (5.139) E where As = cross-sectional area of the stiffener d= distance between the center of gravity of the ‘eross-section of stiffener and axis x- Assuming Teese ? Trem 65.140) then Bese ~ Trem 7 2Agl? = ren = ST (5.141) 2 ti tim tn tae son a eee et ep iLtamrE ees Ttar Sseise 7 “S ; (6.143) Assuming that Sorig > Ssease 6.144) Therefore, the reduction in section modulus of stiff ened cross-section of the stack in % is reach opening and sitteners geometry. 5.27 Hi P 1 i iN uv SIDE view PLAN FIGURE 5.29 — Vertical stittener as beam-column. Seeice ‘orig It should be remembered, however, that the stress in the extreme edge of the stiffener is ay (6.145) £. EM fetige " Soc site (5.146) wiveresthe stesin the stack pats themselves atu lessened in the ratio - ually x a (5.147) or, the original moment of inertia divided by the new ‘moment of inertia. Thus, the original section modulus is hot seriously weakened, and the plate stress is reduced. 5.7.1.5 Buckling Stability of Vertical Stiffeners Vertical stiffeners under action of axial load and bending moments due to the eccentricity of axial loads at both ends, may be considered as beam-columns. To achieve the sreater factor of safety, the stiffeners will be considered as isolated beam-columns, neglecting the plating of the wall to which the stiffeners are connected. Both ends of the stiffeners are assumed as hinges, Figure 5.29. ‘The maximum stress at mid length of the column is given by the secant formula (5.58) 80 9002-40-90 SHdd ‘ld WO pe: OH OL ‘OI Ta'a5.28 Tubular Stee! Structures — Theory and Design FIGURE 6.90 — Ring girders. 5.7.2 Ring Stiffeners In addition to the compensation members which are placed as vertical stiffeners, horizontal reinforcement should be provided by using ring girders above and below the breech opening. These ring stiffeners should be designed as beams fixed to the wall of the stack to carry the "unsupported parts of the stack above and below breech opening, Figure 5,30. ‘The span in bending is the width between the vertical stiffeners, but the girder should encircle the stack to preserve circularity at the opening, ‘To form each ring girder, stel plates should be placed to act astop and bottom flanges. The shell of the stack can serve as a web, ‘Around the stack. each ring girder must be capable of carrying a uniformly distributed load Ib/in of are due to axial vertical load “ok (5.14) and due tothe wind bending moment, when the stress i £ he (5.150) ‘Therefore, the corresponding vertical load is fea, 6.151) The total distributed load in Ib) wewp tye (5.152) ‘The bending moment in the girder is wet (ao gM we Beaks (2+ B) orm 5.8 Rigorous Breech Opening Analysis Ever since the advent of plate steuctures, the structural arrangements in the vicinity of openings in them have been a matter of considerable concern. When analyzing and designing such structures, var~ ious factors have to be taken into consideration. One of these is the “stress concentration” around openings, recesses and cutouts in plane and three dimensional parts of structures, The initial theoretical work on this subject was per- formed by Inglis [5.59] who approximate the effect of a rectangular opening with rounded corners by a pair of cllipses intersecting obliquely. Savin [5.60] studied both the square and the rectangle openings with rounded corners. Greenspan [5.61] working independently and by another method, produced a solution for the square with rounded corners, None of these analytical approaches is applicable to a wide variety of rectangular openings. The first step to rectify this situation was taken by Brock [5.62] who presented a solution for the entire family of rectan- ales with rounded corners. ‘An outstanding theoretical study of stresses around rectangular openings in cylindrical shells appears by ‘Muskhelishvili [5,63]. He makes use of the complex-var able method in conjunction with conformal mapping technique in investigating such a problem. His solution's based on the assumptions of plane elasticity: homogene- ‘ous, isotropie material within the elastic limit, uniform stress across the thickness of the shell with no stress nor~ mal to it, an opening “small” relative to the shell and “small” displacements. ‘An accurate analysis of the stresses around openings in stacks until recently has been beyond the tate of the art in shell analysis. The large number of parameters involved makes in impossible to produce design charts for designing reinforcement around openings by use of a purely empirical approach and a theoretical analysis has been prohibited by the lengthy and complicated mathe- ‘matics involved. Consequently, design of opening reinforcement has been based on rules of thumb which are generally quite conservative due to the uncertainty involved. However, recent advances in the Finite Element Method of Analysis coupled with improvements in computer technology and numerical analysis methods have brought the state of the art toa level where it now appears feasible to establish design procedures with a more solid foundation [5.64, 5.65, 5.66, 5.67, 5.68]. 5.8.1 The Finite Element Program ‘The invention of digital computers has significantly im- proved the output of engineering profession. Manual ‘methods were the only means of performing engineering calculations up to the end of the Second World War. They ‘were not practical because of the enormous calculation ‘work involved in solving a large number of simlutaneous. equations encountered in analyzing any reasonably large structure, To overcome this difficulty, engineers resorted to comparatively easier but approximate methods, e.g. relaxation and successive approximation methods. These ‘methods, though still in use at present, have been virtually replaced by computerized stiffness solution methods — self Supporting Stacks 5.29 incorporating the finite element technique and matrix ‘methods for structural analysis [5.69]. Computers seem to be ideally suited for modern structural analysis problems because of their versatility and tremendous speed resulting in substantial savings in time and man-hours required for a particular project, In addition, computer programs using finite elements have the following distinct advantages: 1. Their ability to use elements of various types, sizes and shapes and to model a structure of arbitrary geometry. 2. Their ability to accommodate arbitrary support conditions and arbitrary loading, including thermal loading. 3, Their ability to model composite structures involving different structural components such as stiffening ‘members and a shell and combination of plates, bars and solids, et. 4, The finite clement structure closely resembles the aciual structure instead of being quite a different abstraction that is hard to visualize. Invariably, the finite element method of analysis entails afew disadvantages as well, as shown below: 1. A specific numerical result i obtained for a specife problem. A general close form solution, which Would permit one to examine system response to changes in various parameters, is not produced. 2. Experience and judgement are needed in order to construct a good finite element model. 3. A large computer and a reliable computer program are essential, 4. Input and output data may be large and te. prepare and interpret. 5.8.2 Method of Finite Element Analysis Finite element analysis has come about through the use of digital computers solving with matrix algebra methods the many equations of compatibility and equilibrium created by classical solution techniques such as slope deflection. Generally, most finite element analysis pro- _Brams utilize displacement methods of solution, resolving deflections first, followed by a stressing routine, which solves for member forces (stresses) from the deflections icc, a stiffness solution method (5.70, 5.71]. What must be appreciated, however, is the fact that all finite element models are at best approximations of the ‘actual structures they may represent. Firstly, the applica~ bility of finite element analysis must be determined with respect to a problem in hand. Ifit is possible to model an actual structure under consideration, it then becomes Recessary (0 carefully choose a proper element and grid pattern such that the errors, that incur when the actual structural continuum is replaced by the finite element model, may be minimized. Furthermore, though must be Used to establish the appropriate constraints that must be applied about the model so that the model and actual structure tend to behave similarly. Over the past years, as finite element technology Brew, its popularity also increased. Subsequently, at the ‘moment, good evidence exists verifying the viability of the finite element analysis method as quite reliable when applied properly. Generally speaking the steel stack analyzed in this reportis well ordered and noacomplex problem ta solve. is anticipated that the results of the analysis presented in Appendix B, are very reflective of the forces and deflections that would exist if the modelled structure and the applied loads were in actual existence. 5.8.3 ANSYS Computer Program ANSYS is a large seale general purpose computer program for the solution of several classes of engineering Analysis problems. Itsanalyzing capabilities include static and dynamic; plastic, creep and swelling: small and large deflections: steady state and transient heat transfer and steady state fluid flow. The matrix displacement method of analysis based upon finite element idealization is employed throughout the program. The library of finite elements available namn- bers more than forty for static dynamic analyses and ten for heat transfer analyses. This variety of elements gives the ANSYS program the capability of analyzing frame structures (two dimensional frames, grids and three dimensional frames), piping systems, two dimensional plane and axisymmetre solids, lat plates, three dimensional solids, axisymmetric and three dimensional shells and nonlinear problems including interfaces and cables. Loading on the structure may be forces, displace- iments, pressures, temperatures or response. spectra Loading may be arbitrary time functions for linear and nonlinear dynamic functions for linear and nonlinear dynamic analyses. Loadings for heat transfer analyses include internal heat generation, convection and radia~ tion boundaries, nd specified temperatures or heat flows 18.73, ‘The ANSYS program uses the wave front (or “fron- tal”) direct solution method for the system of simultane ous linear equations developed by the matrix displace- ‘ment method, and gives results of high accuracy ina minimum of computertime. The program bas the capability of solving large structures. There is no limit on the number of elements used ina problem. The number of nodes can be in exeess of 2500 for three dimensional problems, and 5000 for two dimensional problems. There is no “band width” limitation in the problem definition, however, there is a “wave front” restriction. The “wave front restrietion depends on the amount of core storage available for a given problem. Up to 376degrees of freedom on the wave front limitation tends to be restrictive only for analysis of arbitrary three dimensional solids or inthe use of ANSYS on a small computer. ANSYS has the capability of generating substrue- tures (or super-clements). These substructuces may be stored ina library file for use in other analyses. Substrue- turing portions ofa model can resutin considerable com puter time savings for nonlinear analyses, Geometry plotting is available for all elements in the ANSYS library, including isometric, perspective and section views of three dimensional structures. Plotting sub- routines are also available for the plotting of stresses and O05 Total force, Fn, NBC Supplement 4, p. 77 is Fa Syd Sy Se AL Ceand Ac are functions of Hy. i., the height of the stack at a point x from the top as shown in Figure 5.33, ‘Therefore, ng o rr aes v.s0.0 [ea] co (ise Se a. Wind pressure from 0 to 40 ft (exactly 0 10 418°) The exposure factor Ce = 0.50 su, a7 065 9.9 0.5 x 2.0 (iE = HE) = = 103 = 0.299 ay 8. Wind pressure from 40 to 200 ft 2)! su, #, + 0165 x 9.9 x 0.6 (24) 2.0 fie = ook = (25.98 - 0.03.) ff, FIGURE 5.33 — Distribution of wind load acting on the stack. For Hx Fy = 103 Hx Fn = 95.28 Hy Fh = 140.71 2. Part 2. Bending Moments von \ \ | : \I i \ +t fore sey fe ? FIGURE 5.34 — Determin ‘under wind loading. bonding moments In stack 80 9002-2du-90 wold be. SWdd ‘Ld (RAH OL a owe532 Tubular Steel Structures — Theory and Design ‘Self-Supporting Siacks 5-33 Mx for uniform loading if x < 160-0" Meat is erm y ols - = (10 = 0.1405%)x* Mx for triangular loading if x > 160-0" Mxt = 0,0938x! Total moment when x < 160-0" My = (70 ~ 0,1406x)x4+ 0.0938x? = = 70x? ~ 0.0468x* For the bottom part x > 160-0" the exact expression is na Walt + 63x = 126401 (x = 160) (0.2% + 253) ns soo. ae However, for the bottom 40 ft. the wind loading can be considered to be uniform and set equal to 95 1b/{t, or S(x = 150)* ia z Moment at the base = 47.5 (x = 160)? 2 FL } (95 + 140) 160 - 19800 Ibs 160 (2 x 140 + 95). 460 x 375 Sado + SST 3x 238 a5 fe My = 18800 x (85 + 40) = 2,250,000 ib/tt = } 6 28,200,000 ab/in 2 = 205 + 100 40 = 2960 we ay = 40 = $2 32 98) 973 ee Mp = 3960 x 19.73 = 78130.8 Ib/tt = 2 2 937,570 Ib/in Meoe = 28+200,000 + 937,570 = = 29,137,570 Ib/in 3, Part 3. Determination of Stack’s Wall Thickness In order to determine the required thickness of stack’ wall, the moment due to wind should be determined. The critical Duckling stress can be determined from equations: 3 Wytee ay DF (a0) Combining both of the above equations and substituting cor = fe+ fw yields ty 4, ee BT * oF (ered setting (” = te, the equation can be rearranged so that t can be optimized. Therefore, 4, 4 eget eri” ne5,0,* ors Assume t' = 0,50 inch at base 2 Mex merefore 3, = 157% since 2 < 400, the Wilson-Newmark equation cannot be used, therefore, the AISI formula applies. = 384 Since 13000 269 < 2 0 reo
You might also like
Shoe Dog: A Memoir by the Creator of Nike
From Everand
Shoe Dog: A Memoir by the Creator of Nike
Phil Knight
Rating: 4.5 out of 5 stars
4.5/5 (537)
Never Split the Difference: Negotiating As If Your Life Depended On It
From Everand
Never Split the Difference: Negotiating As If Your Life Depended On It
Chris Voss
4.5/5 (838)
Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future
From Everand
Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future
Ashlee Vance
4.5/5 (474)
Magazines
Podcasts
Sheet music
The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life
From Everand
The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life
Mark Manson
Rating: 4 out of 5 stars
4/5 (5783)
Grit: The Power of Passion and Perseverance
From Everand
Grit: The Power of Passion and Perseverance
Angela Duckworth
4/5 (587)
Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race
From Everand
Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race
Margot Lee Shetterly
4/5 (890)
The Yellow House: A Memoir (2019 National Book Award Winner)
From Everand
The Yellow House: A Memoir (2019 National Book Award Winner)
Sarah M. Broom
4/5 (98)
On Fire: The (Burning) Case for a Green New Deal
From Everand
On Fire: The (Burning) Case for a Green New Deal
Naomi Klein
4/5 (72)
Yes Please
From Everand
Yes Please
Amy Poehler
4/5 (1888)
The Little Book of Hygge: Danish Secrets to Happy Living
From Everand
The Little Book of Hygge: Danish Secrets to Happy Living
Meik Wiking
3.5/5 (399)
Principles: Life and Work
From Everand
Principles: Life and Work
Ray Dalio
4/5 (599)
A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story
From Everand
A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story
Dave Eggers
3.5/5 (231)
Team of Rivals: The Political Genius of Abraham Lincoln
From Everand
Team of Rivals: The Political Genius of Abraham Lincoln
Doris Kearns Goodwin
4.5/5 (234)
Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America
From Everand
Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America
Gilbert King
4.5/5 (265)
The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers
From Everand
The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers
Ben Horowitz
4.5/5 (344)
Fear: Trump in the White House
From Everand
Fear: Trump in the White House
Bob Woodward
3.5/5 (738)
The Emperor of All Maladies: A Biography of Cancer
From Everand
The Emperor of All Maladies: A Biography of Cancer
Siddhartha Mukherjee
4.5/5 (271)
Steve Jobs
From Everand
Steve Jobs
Walter Isaacson
4.5/5 (806)
John Adams
From Everand
John Adams
David McCullough
4.5/5 (2409)
The World Is Flat 3.0: A Brief History of the Twenty-first Century
From Everand
The World Is Flat 3.0: A Brief History of the Twenty-first Century
Thomas L. Friedman
3.5/5 (2219)
The Unwinding: An Inner History of the New America
From Everand
The Unwinding: An Inner History of the New America
George Packer
4/5 (45)
The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are
From Everand
The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are
Brené Brown
4/5 (1090)
Rise of ISIS: A Threat We Can't Ignore
From Everand
Rise of ISIS: A Threat We Can't Ignore
Jay Sekulow
3.5/5 (137)
Angela's Ashes: A Memoir
From Everand
Angela's Ashes: A Memoir
Frank McCourt
4.5/5 (440)
Bad Feminist: Essays
From Everand
Bad Feminist: Essays
Roxane Gay
4/5 (1015)
The Glass Castle: A Memoir
From Everand
The Glass Castle: A Memoir
Jeannette Walls
4.5/5 (1711)
The Outsider: A Novel
From Everand
The Outsider: A Novel
Stephen King
4/5 (1800)
The Woman in Cabin 10
From Everand
The Woman in Cabin 10
Ruth Ware
3.5/5 (2322)
Brooklyn: A Novel
From Everand
Brooklyn: A Novel
Colm Tóibín
3.5/5 (1937)
The Light Between Oceans: A Novel
From Everand
The Light Between Oceans: A Novel
M.L. Stedman
4.5/5 (789)
The Sympathizer: A Novel (Pulitzer Prize for Fiction)
From Everand
The Sympathizer: A Novel (Pulitzer Prize for Fiction)
Viet Thanh Nguyen
4.5/5 (119)
A Man Called Ove: A Novel
From Everand
A Man Called Ove: A Novel
Fredrik Backman
4.5/5 (4609)
Wolf Hall: A Novel
From Everand
Wolf Hall: A Novel
Hilary Mantel
4/5 (3811)
Manhattan Beach: A Novel
From Everand
Manhattan Beach: A Novel
Jennifer Egan
3.5/5 (791)
Little Women
From Everand
Little Women
Louisa May Alcott
4/5 (104)
The Perks of Being a Wallflower
From Everand
The Perks of Being a Wallflower
Stephen Chbosky
4.5/5 (2099)
The Art of Racing in the Rain: A Novel
From Everand
The Art of Racing in the Rain: A Novel
Garth Stein
4/5 (4193)
Her Body and Other Parties: Stories
From Everand
Her Body and Other Parties: Stories
Carmen Maria Machado
4/5 (821)
A Tree Grows in Brooklyn
From Everand
A Tree Grows in Brooklyn
Betty Smith
4.5/5 (1929)
Sing, Unburied, Sing: A Novel
From Everand
Sing, Unburied, Sing: A Novel
Jesmyn Ward
4/5 (1103)
The Constant Gardener: A Novel
From Everand
The Constant Gardener: A Novel
John le Carré
3.5/5 (104)

Tubular Steel Structures Theory and Design M S Troitsky 2nd Ed 1990 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tubular Steel Structures Theory and Design M S Troitsky 2nd Ed 1990 PDF

Uploaded by

Copyright:

Available Formats

You might also like