GUIDELINES SHV bES These guidelines are reported by the ASCE Gommittee on Cable-Stayed Bridges: Carl C. Ulstrup*, Chairman 1983-1989 Man-Chung Tang”, Chairman 1989— Bogusz Bienkewisz Guoha Li Niels J. Gimsing* Raymond J. McCabe Manabu Ito* Walter Podolny Gerhard Joehnk* R. L. Wardlaw* John M. Kulicki Wilhelm K. Zelner* “Contributing Members 1 | OFe00 95 3 _Na Z oles: ae Senco | Had iech . TH a - x Pubihsbectiy the) PR EOE g American Society of , 345 East 47th Street New York, New York 10017-2398 ngineersLe for design as wel a3 . neler testing, com See anne ore of anchorages is included. Library of Congress Cataleging-in-Publication Data Guidelines tor design of cable-stayed bridges / Carl C. Wlntrug.. [et al). “those qodeines are imported by the ASCE Committee on Cable-Stayed Bndges.” Inchurieess inch. IS8N 0-87 A62-800-7 | Groges, Gable-Stayrd — Design and construction. |. Uisiruo, Car G. fl. Amarican Somety of Civil Enginonr. Commttes on Cable-Stayed Bridges. Tans GBS 162 dro S247073 or > material presented in this publication hag beer pre- accordance wilh generally recognized angmesning Apgliratan. coments of this publication are not intended te be Id iol be construed to Daa stancardof the Amencan Slalules, or any ather legal document. reference made in this publication ta any specific praducl, process, of service constitutes or implies emem. recommendation, o wararnly thereo! by ASCE makes no represantahion or warranty of any kind, InrESS OF implied, concemmy Ine accuracy. com- suitability or utility of any inlonmalign apparatus, or process discussed mm this publication, and uma no liability therefor nore uliizing this information assumes al kability sing rom such war including but not limeted te infringe: 1 of any palent or palents. Autnongaban 10 pholOCOpY Material for intemal oF personal un mci ice te hag wn Mit Lae vans ot tha Gapynght Act is granted by ASCE tolibranes and ner Users regslered with ine Copyrignt Clearance Center Ch eens noc Fine panies set se 28 per ar lie 5 3.15 per pa irectly OOC, 2? Congress Street. Salem, MA OIE 70 5 ite sik cabon tor ASCE Books is 0-B7262:92. $1 + 15. Requesis tor special permission o¢ bulk copying should be aridiessed to Appants!Permessions Department Copyright © 1992 by the American Sochely af Cia Engineers, ALL FA a Gangret Calalag Card No: 9247073. abana ISBN 0-87262°900 7 : Manutlactured in the United States of America (Gower photo: Talmarige Memorial Drige, Savannah, Georgia INTRODUCTION These guidelines are intended for the design of cable- stayed bridges. They are intended as performance specifications to be uged with any national coda. However, national codes for live loads and other important provisions differ. Yet safety level depends en all loads and resistance paraneters. ‘Therefore, 4 mre complex treatmant of the design requirements may be required undar cartain conditions. The guidelines in these papera are the best estimates by the ASCE Committee on Cable Suepended Bridgea. They have been developed based on currant practice and the lateat thinking for this type of bridges. Pquations given in S$I-UNTTS with aquations in parenthesis in English mits.CHAPTER CHAPLER CHAPTER CHAPTER CHAPTER CHAPTER Safety Factors, Loada, and Impact Cables and Anchorages Stability of Cable-Stayed Bridges Dynamic Wind Leads Special Considerations for Railroads Seismic Design iy 21 Be 4 45TER 1 SAFETY FACTORS, LOADS, AND IMPACT 1.0 GENERAL The loads given in this chapter are intended for general guidance only {including wind, ice, straan flow, etc.). Maximum deflections and retationa may likewise vary with national codes. In sll cases, loads must be considered in conjunction with allowable stress and/or load factors. 1.1 DEAD LOADS The dead load shall conaiat of the waight of the entire structure, including the roadway, sidewalks, car tracks, pipes, cunduite, cables, and other public utilities. For roadway bridges it shall include parapets, medians, separate and future wearing surfaces, and railings. For railroad bridges it shall include tracks, ties, ballast, deck and walkways, and railings. For stay forces, the weight of steel tendon, sheathing and grout, when present, shall be included as dead load in the stay or cable. 1.2 LIVE LOADE it ig yecommendes that live load follows the loads given in the governing coda for that structure. Number of traffic lanes, railroad tracks, impact, longitudinal forces, centrifugal forces, reduction in lead intensity for multiple lane or tracks, railing loading, sidewalks loading, etc., according te codes. In addition to these loads, checkerpoard loading to produce maximum torque should be investigated. Impact ghall be computed in accordance with the national cade weed. Membars to be included are superatructura menbers and stays. Piers and towers are excluded from impact. 1.3 WIND LOADS All wind loadsa in this chapter are "static," for “dynanic" wind loads, see Chapter 4. The wind loadsa shall be as defined in the national code, however, the - 1applied horizontally at right angles to the nal axle of the structure ghall at least ba: ma! O.2 2 2 % Vag -% Cy Hh gee 25.6 0.2 2 z xv xc, “i. = i abs tt?) 600 # = Heignt in m (feet) of the top of the floor ayetem above ground level, or ion (30 feet), whichever is greater. (Vyo) — Fastest mile wind speed, in kpshr (mph), at 20m (30 feet) above ground, for 100-year maan recurrence interval (ece Figure 1 for use in the U.5.A.). For other countries reference is made to almilar statistical information. S, = Shape facter fer herixental wind load. yp, unless wind tunnel data confiran different values, shall be at least: for plate and boy girders cy for trusses (applied to c, one plane only) 1.5 2a ecommeanded Design Loads for Bridges," sT 7, 1961, ASCE Journal of the Structural om (30 feet), Vyq = i6akm/hr (V9 = 100 mph) 8 given for cp, Ve wind load is 2,400 Pa (50 and 3,600 Fa (75 lbsytt*) recpectively. For loads under different skew angles, the lateral and udinal load components may be found by force ponen=s of using the game ratios as given in AASHTO n Association of State Highway and eneportation Officials. Standard Specifications far Highway Bridges Lath ed. 1989). Whe wind loade on cables may be found by the sanc Squation given above using ¢, = .7. Proper danping of Gables must be provided, due beth to vertex shedding and to cacillations due to variations in cable tension. 1.4 WIED ON LIVE Loap This load shall be applied 1.am (6 feet) above the deck tor highway bridges. For railroad bridges see Chapter 5. The magnitude of thie load is: o.2 2 * Vag % Spy We See Pa 25.6 a.2 2 Z x Y. zt 30 h (HY ee sec Lba ft?) 600 Where Sa = 1-0 and Vp (Vg) * shall not exceed 93 kmjnr [55 mph) . The magnitude of the exposed area shall be the loaded length of superstructure multiplied by = height of in (10 feet). 2.5 WIND OW SUBSTRUCTURE The horizontal wind loads per square toot of exposed area of sunatructure is calculated from the expression under Section 1.3 with: Ch — +7 for circular €, = 1-4 for octagonal {, = 2.0 for rectangular sections aa Por wind directions assumed skewed te the substructure, thie force shall be resolved into components or may be applied according te AASHTO. 2.6 VERTICAL WIND - OVERTURNING FORCES An upward force shall be applied at the windward quarter point. This force per unit araa of deck is calculated for the axpreggien in 1.3 where cy is substituted for Cy. C, may be assumed to be Cy = .8.“| a | wt Se) NP } ead ao x i ae: te lea. | 2 ae | Sey oe Table 1 — Streuhal Number tor Various Groae Sections Crose—section MAISIAST (WHA LONGUS ate 40 THOR 3IRY “HL ATOT -83LNLS GALINA JIL HI SHIA BASH uo SHOAL GIO HOR STM WOU CALSVHLNa “IWAUTLNE USHTMYOSE MAW UA-DEL ‘OXNOME anGGY 14 OF osade aun SS315¥4 lWaNLea TnNy -f Yalues of C, and different pointes of action nay if justified py wind tunnel test. i)? AcChows WIND VIBRATION OF BTRUCTURAL MEMBERS @ider structural membere such a6 hangers, truss ra, andeel columns, and cables, shall be Agned to rasist resonant vibrations caused by vortex ding, The critical wind velocity is determined by: Vere 7 2.68 2b ka/he 0.62 fb mph) a Worit = Strouhal number (as given in Table 1) frequency in Herts ~ characteristic width of cross section in m (feet) mts 1 7 shajl, not be ese than 0.91 x 291 x wy ee ere Bee Be cae iether wah ee ag @ effect cf wind loads on cablas and ta the nead for Ohanises to dampen cable oscillations. Cable felllations also occur due te changea in cable tenealon Qaueed by live load. During rain, water rivulets (Whhing down the cable may, in combination with wind, oa disturbing cable oscillations. Additional wind @ijeiderstions are given in Chapter 4 - Dynamic Wind Lada. i. THERMAL FORCES Ih addition te the uniform thermal forces normally used fay Btructures in a particular climatic area, Mpersture differential between cables and deck or Qebies and pylons must be considered. A minimum of iorc (1 temperature differential should be applied for white teped or painted stays. Site specific (Webers for temperature differentials for othar colors Bhould developed and may be as high as 20°C (36°F) for black setaya. Thermal gradients for box girders Wuet be considered. A 210°C (118°F), gradient through feck is recommended. A © (f15°F) differential fetween the inside and outside of the box should be Investigated, but not aimultaneously with the deck Gradient. For open deck structures, a +5°C (+9°F) Gradient through the deck should be considered. Purthernore, reference is made to "National Cooperative Highway Research Program Report 276 - Thermal Effects in Concrete Bridge Superstructures." 1.9 6HOW AND ICE ON SUPERSTRUCTURES In aress where it can be expected that the bridge may be closed to traffic due to accumulated snow, tha bridge should be designed for such loading conditions, 4,20 FORCES FROM STREAM CURRENT AND FLOATING ICE Forces from stream flow may be assessed by the equation: Pe515 xk x v2 (Pek x vy pressure in Pa (pounds per square foot) water velocity m/sec (ft/sec) form factor; 1.4 for square ends 7 for circular piers -5 for angle ands 30° or lesa When currents may become of such magnitude that scour is possible, proper scour protection is required. Ice forces on piere shall be selected with regard tc site condition and mode of ice action such as dynamic or static ice pressure, ice jams, or static uplift. Dynamic ice forces may be calculated by: FeC,xpxtxw F = ice pressure in 4 (pound=) C, 7 nose coefficient (1.0 - vertical nose .5 = ao" to as" nose inclination to vertical) Fp = ice strength .7 te 2.78 uymm? fran to 400 pounds per square inch) t = ice thickness mm (inch) w = width of pier in contact with ice mm (inch) 1.11 FORCES FROM SHIF COLLISION Piers on navigational waters should be designed far ship impact unless otherwise protected by fendering systems, etc.1.12 BUOYANCY AND EARTH PRESSURE Buoyancy and earth pressure ahall be considered whera it affects the design of substructure or superstructure. 1.13 EARTHQUARES In regione where aarthquakes should be anticipated, structures shall be designed te resist earthquake notions considering the relationship of the sita to active faults, selamic response of soils, and the h of the structure. It may be done by an aquivalent static force method, responsa spectrum method, or multi-mode spectral procedures. tructures a time-history response analysis may be considered. Multimode spactyal procedures should be considered empecially if the acceleration coefficient for the bridge site is greater than 0.2. Unlike classical sspension bridges, vibrations of cable-stayed bridges cannot be categorized as solely vertical, lateral or torsional; instead, a three-dimensional motion is associated with almost every mode of vibration. For long structures, it is furthermore obvious that the structure is subjected to different motions at each of its foundations. additional consideration is given in chapter 6 - Seismic Degiqn of Cable-Stayed Bridgas, 1.14 CONBTRUCTION LOADS Leads applied during conatruction nust be investigated. Some recent construction methods Jlika halanced cantilever construction leave the structure more vulnerable during construction stages than in its inished etage. Wind oscillationa about a horizontal andjor vertical axis at the tower must be investigated, For structures which are constructed or designed for a balanced loading condition dua to dead load, an unbalanced losd, up on ona arm and down on the other mugt be added. The magnitude of the unbalanced load sall be stipulated on basis of construction tolerances and variation of material densities. superstructure towers and/or piers should be designad for thie unbalanced load which is s part of the dead lead. In areas with earthquake, the risk should be evaluated. 1@ maximum unbalanced moment due to sequential addition of segments or panels is the moment produced by one geqment or panel, plus erection equipment needed for construction (traveler, forms, craneg, etc.}. In addition te these leads, an additional 460 Pa (10 pst) For complex shall ba applied on the loaded cantilever and 240 Pa {5 Pst) on the second cantilever to account for unanticipated construction equipment. During construction it is customary to apply 100¢ wind oh one cantilever and 50 on the other for design purposes. Due to infrequent occurrence, seiamic loadings are usually strongly reduced for the Construction staga (see Chapter 6}. 1.15 LOADING COMBINATIONS Since live loads are following national codes it is only natural that loading combinations, allowable stresses and/or load factors fallow the consistency. same code for 1.16 FATIGUE LoADe Again it is recommended to follow the national for live load to be extendad for fatigue loads. codes Fatigue leade for cables are given in “Recommendations for Stay-Cable Design and Testing" by thea PTT Ad Hoo Committee on Cable-Stayed Bridges. Cable fatigue stresgee are discussed in chapter a - Cables. 1.17 MAXIMUM DEFLECTION 24D ROTATION Thera is mo direct set limitation for deflections, hewaver, Lt is part of the P-f analysis. Maximum slope changes due to live load are more of a service conaideration and dictated ss follows for the use of bridges: Maximum elope changa for truck loading: 2.0% (1.157) For railroad gee Chapter 5. 1.18 STRUCTURAL STABILITY AWD LINEARITY Cable etructures are ones in which loads are transmitted from tha deck to the towers by tensile forces with no compression allowed. changes in loads cause changes in cable geometry, The literature contains ample guidance with respect to suspension bridges. Fer ceble-stayad bridges the minor sag effects may be accommodated by the use of an equivalent nodulu= Of elasticity. Because the deck structures often are .{large span/depth ratio), second order effects by the deck thrust must be accounted for, The Boat common method ia to use a get of Uniferm and concentrated loada for linaar combinatian Of maximum live load cases. The second order effocts [Pedelta; will be determined separately for maximum loading cases along the girder. The tower structure should likewise be investigated for ®econd order efrecta. There are computer programs on the market whigh will Wive the second order effects, from which the nU® order Bay be predicted. Non-linear programs are available, Wut may becoma rather tedious for a full invest igqat ion. The deck structure must be properly investigated Against buckling, and it should be noticed that proper action equations for bending, thrust, and shear used. Reference is made to "Proposed Design poificetions for Steel Box Girder Bridgas" - Report Wo, Fiwa-vs-s0-205, Further discussion i mado in Chapter a — stability of @able-Stayed Bridges. PERPUSTARAAT tues | epee 46s 2 OOH eS F | Ra [oe cee ee asepey | i | ‘CHAPTER 2 CABLES AND ANCHORAGES 2.0 GENERAL The following types of cables are generally used 4h Gable stayed bridges: 1, ‘Parallel bars 2. Parallel seven wire strands 1. Parallel or gemi-parallel wire cablas 4. Locked coil stranda and ropes 5. Helical galvanized bridge strands The first two types are extensively used in prestressed concrete structures and when applieg as cables in stayed girders bridges, a very sindlar installation technology is used. The cable is installed in a polyethylene tube or steel pipa which ia then filled with grout. The grout add# shout 30% to the weight of the cable. This adds to the cable sag and decreases ite stiffness. @ three latter typas have been used for # long time in atructures supported primarily by tension nesbers. The guiding principle here has always been case of inspection and maintenance. 2.1 CAERLE STRewcTH Under normal design load (dead plus live plus impact) the static stresa in the cable shall not exceed 454 of the guaranteed ultimate tensile strength (GUTS) of the cable, The design fatigue stress range of the cable is determined by teata. The fatigue strength of the cable is assumed to be the stress range at which $% of the wires have failed when the cable is subject to an average tensile strength of 452 GUTS. the nuaber of cycles used in the test is normally 2 * 10° for highway bridges and 1 x 107 for railway bridges.fhe design fatigue stress range is defined as: : iorat = 498 geet Whore Fat = G max -G min > fatigue stress range as determined by tests 4 d Fat = factor of satety may vary from 1.25 to 1.50 WOrmeally two curves are used to determine the allowable | fatic range. , . \ 7 The Smith Diagram. The stress range is the , PPkamples res. 1. The Wohler curve. ordinate, the mumber of cycles the abscissa. The tensile force in the cable does not enter tha picture. whe stress range is the ordinate, the average strass is the abscissa The number of cycles are given. ef these curves are shown in the appendix in 1 through §. Figur is customary to specify both static and fatigue ts for the cables in the steel plant before @weepting dalivary. The easiest things to test are @isgie wires and bars and from thease results compute the « @ strength taking inta account thea “group effect, This means that the combined strenqtn of Several tensile members is less than the gua of the imdividual members. A cabie made up of several seven wire strands have two p affects." The strength of a strand is less than Seven tines the strength of one wire and the strength ef the cable ig less than the sum of the individual strands. The nanufacturer'a catalog will uaually suggest wnat the value of those "group effects" are for preliminary design purposes, but the final design strength le determined by tests. Most @ngineera will normally specify that tests ba mada On full size cables with anchoragea attached at each end. Tt is sufficient to use three cables for each type of test. 12 In fatigue tests the median caple atrege shall be at least 45% GUTS. The cable must break outside the anchor socket. The langth of the test cable is usually about 4 meters (13 feet). 2.2 DESCRIPTION OF CABLES E - Young's modulus in w/a? [kei = Kips per square nen). GUIS means guaranteasd ultimate tensile strength. Creep of steal can be neglected since the maximum stress under normal leading (dead plus live plus impact) must not exceed 45% curs. Whenever an applicable ASTM specification applies, it is mentioned, Normally, it applies only to plain steel tension @lemente, which means they are not galvanized. Since the strength of thease elements is constantly subject to change in metallurgy and protective corrosion comting, testing is necessary and the standard specifications are of little valua. In whet follows reference is made to some ASTM specifications and some typical design stresses that may be used in a preliminary design. But in the final design, actual values obtained fron tests must be used. B.2e1 Parallel Bara The bars aro covered by ASTM A722. They have been used extensively in prestressed concrete structures. When used in cablea, they have to be coupled and their fatigue strength is relatively low. In order to improve the fatigue strength, the anchorage is split into «= dynamically acting part and a statically acting part. The ultimate strength is normally 1,090 Nynm® (159 ksi) but may up as high as 1,470 M/mn* (213 ksi} E = 200,000 N/mm? (29,000 ksi} aaa.2.2 Parallel Seven Wires Strands strana have been uged for a leng time in stressed concrete, Those used in cables normally vo a@ diameter of 15 mm (0.59"), Like bars; the bundle of atrands are ae a mule Asserbled in the figld. The cable may have a GUTS az ' aa 1,870 Nfmm* (270 kei). If the wires in the atrands are galvanized, the CUTS has @ lower value. The plain seven wire strands are covered by ASTM A416 and Aree. BR- 1.8 - 1.3 x 109 wymm* (26,000 - 28,000 ksi) Bea. Barallel or dent-parsliei Wire cables (Typical Values) Poralie! bright wiro cables have = GUTS of 1,670 W/nm? (042 kai) and covered by ASTM aaz1. Be 200,000 wymm? (29,000 ksi) ‘Galvanized parallel wire cables have 2OUTS of 1,570 fan (228 ksi) and £ = 190,000 N/m {27,500 ksi) on crosa Area. re. lay Galvanized wire cables nave GUTS of 1,570 few? (220 kei) and E = 190,000 mymm* (27,500 kai). wwe three types of cables are shop assambled snd wight to the site coiled on druma. baa Locked Goi] strands and Ropes: Whe locked coil atrand has several layers of round Mfee whereas the locked coil rope has saveral layers f differcntly shaped wires. ply the outer layer may be galvanized or #11 wires may ga d. Thase cables should be Prestratched to abou a Mymp? (228 ksi) Nyame (25,000 kai) [_ PeERSIT AS SUMATe 2.2.5 Helical Galvanized Bridge stranas The qurs isjonly 670 N/m (97 kai) and & = 1.6 - 1.65 x 107 N/mm? (23,000 - 24,000 kei} depending on the strand diameter. The strand should ba Prestretched to about 55% of GUTS, Tts fatigue strength is low and it can only be used in cable stayed bridges where the live load is very small in comparigon with the dead load, 2.3 CORROSION PROTECTION Different types of corrosion Protection are used for the various cable systems. 1. Parsllel bars are inserted in polyethylene or ateel tubes which are then filled with cement grout or some other corrosion resistant Substance. The bare may algo have an epoxy coating. 2. Parallel seven wire strands are treated like Parallel bars, de Parsllal or seni-parallel wire cables will normally be placed in a polyethylene tube which is filled with cement grout or another Bubstance if the wirea are not galvanized, A polysthylene sheathing is used for galvanized wires, 4s Locked coil and strands and ropes. These cables are usually left without any cover added, 5. Helical galvanized bridge strands. The sebles nay consist of = single strand or several parallel atrands spread apart. Ne ether protection is previded, 2.4 ANCHORAGES The general requirements ia that the cables fail before the anchorage, the loading being static or dynamic. 45The anchorage for the bars is very simple: ute and bearing plates. fipecial anchorages are required for cablea made up of parnileL seven wire strands. These may be eimilar to those for prestressing cables or as described undar b. below. For cables made up of parallel wires, there are two basic systems: Steel sockets filled with zinc. These anchorages havea keen in use for a long time. Tt they are improperly designed the cable tends to break at the mouth ef the anchorage. This can be improved by providing a smooth rounding where the cable exists. A vibration damping device, such as @ neoprene washer rigidly supporting the cable a short distance avay from the anchorage alao improves the eable's fatiqua life. Having been out of favor for soma years, there seems to be a return te the zinc filled gockat of improved design, one reason being ite low cost. b. Steel gockata filled with epoxy into which small steel balls of various sizes are anbedded. These balls are supposed to provide # wedging action between the wires. The latter are provided with button hands at each ond which rest against a perforated hearing plate. They have a high fatigue resistance and have been widely used. The disadvantage ic high cost and the fact that epoxy apparently melts at about #0 degrees Celsius (180 degraas F). Cable connections at tower and superstructure require caraful design. It has to be assumed that water cannot be completely kept out of the cables or connections and thia water must be allowed to eacape. The transition betwaan cable and anchorage should have a certain flexibility to avoid high bending stresses caused by live load and wind. At the sane tine, a device to prevent cable vibration from taking place at the anchorage should be provided. 16 Sinca corrosion cannot be totally prevented, visual inspection of the cable-anchoraga connection is essential. Gable-stayed bridges should be designed so that the failure of one stay or anchorage aoe allows the bridge to carry reduced live 17Au MPa) sco Stress eee . too 4— 4 i A Ari Pea] ‘et 2 * af of 4 number of eycies a i: Tyekal Winier's carves deduced from satgue tant ot Copenhagen University Loeorstory wire ia. 7 men (2767) Soe one aes ¥ men 276") im 2 4 wh number of ccien Fig 2 Wiler'a curves tor a single strand dia. Te omm (517) ano stgy cba with Freyeeinet oncherages, D202 0405 aR 0708 8 10 Fig. 4: Mamimurm stress in termux ot xm Tee .ooo 4300 Te Nig 3) Meumum stress in teem of A= Cupar — Tiere The Figures 1 throws § howe bean reproduced by the permision of the Fraysinnet Company. The cures: de not inciuda o fector of aofaly Mia = 14S inl CHAPTER 3 STABILITY OF CABLE-STAYED BRIDGES 3.1 GENERAL Thia Chapter deals only with Stabllity under static loadings. For aerodynamic stabilities and stability of ekructure under seismic loading, sae the anproprista chapters, Both the towers and the girder of =» cable-ctayed bridge are compression members. But these alements alao carry signiticant bending stresses. ‘The P-delta effect, then is basically a nonlinear behavior which usually increases the stresses in the structural eclenenta, Tt is not directly a buckling problen. Que to thea nonlinear effects, atresses and applied lead= are no longer proportional to each other. The incressa_ in stresses and deformations are usually faster than the increase of applied loade. Principle ef suparposition is not valid. The monlinesr effect should be considered gicbal analysis and the local anslysia, both the 3.2 GLOBAL NONLINEAR ANALYSIa S.R2 Nenlinsar Behavior The forces and deformations in a global analysia shall consider the nonlinear behavior of tha a@tructural systen. There are three major causes ef nonlinaarity in the global analysis: 1. Nonlinear behavior of the cable, 2. The P-dalta cffact of the girder and tha towers. 3. Material nonlinearity. . 21Feeented by a chord mepber, the instantaneous jee of a stayed cable is =f G? costa EA BA , 7 eA fae [mod.} Gat ' G = Total weignt of cable A = Crogs-sectLonal Area E = Young's Modulus H Horizontal component of cable force 9 = Angle between cable chord and horizental fhe cable force changes and ita horizantal me varies from HW, to Hy under a specific , the effective cable “stiffness can be hed as ay 6? costs EA (HytHy) EAtmoa.) = FR’ [2+ “=e ion can be applied in the analysis The _B-Delta Effect ® effect (Geometric Wonlinearity) may be by waing the nomlinear equations for beam However, it is sufficiently accurate to do siation | iteratively using the deformed ral shape. per slope of the cable shall also be adjusted in ion dua to the displacement at the cable inte. nonlinearity can usually be neglected in is of cable-stayed bridges under normal However, for exceptional loadinga or if the state of the cable-ctayed bridge is to be lated, the nonlinear relationship between moment fotation under tha influence of axial forces due to [ding of the steel and/or cracking of concrete shell Gonsidered by applying the actual effective Pfvees cf the structure iteratively in tha analysis. az Slackening of cables especially for bridge strands vnen the cable force ia above the pre-stretching force shall be considered, 30265 Ease Structural System A base structural system is the shape ot the structure after all losdings are removed from the structure, including cable adjuatment forces. In a rigorous nonlinear analysie 411 loads must be applied to the base structural system simultanecusly. 3.2.6 Analysis Tt is recommended te increase all leade including cable adjustment forces to 120% of their intended valuea in # nonlinear analysia. This is to assure that inaccuracy in estimating the weight of the structural clements as well as overloading will not cause axcegsive stresses bacauae of the nonlinear behavior of the atructure. This procedure is to be applied to both allewable stress design and load factor design. For load factor design the value of 1.20 is used in licu of the load factors. 3.3 BUCKLING LoaD a.3.2 System Safety the safety factor (}) against elastic buckling of a cable—stay: fase may Be calculated for a specific loading te ted Vn [esas + DCF LIER pocay) 2 Pa! as where: Mode shapa ordinates 7 Cable forces = Cable length = Cross-sectional area = Young's Modulus = Moment of Inertia = Axial force in mamber (compression positive) (") = Denotes derivative with respect to a Jan Fe om The above Formula dosa net consider the nonlinear benavior of the material properties. . 23