Most structural damage to steel bridges is the result of repetitive loading from trucks or wind. Often, the damage is caused by secondary effects, for example, when live loads are distributed transversely through cross frames and induce large out-of-plane distortions that were not taken into account in design of the structure. Such strains may initiate small fatigue cracks. Under repetitive loads, the cracks grow. Unless the cracks are discovered early and remedial action taken, they may create instability under a combination of stress, loading rate, and temperature, and brittle fracture could occur. Proper detailing of steel bridges can prevent such fatigue crack initiation.
To reduce the probability of fracture, the structural steels included in the AASHTO specifications for M270 steels, and ASTM A709 steels when ‘‘supplemental requirements’’ are ordered,* are required to have minimum impact properties (Art. 1.1.5). The higher the impact resistance of the steel, the larger a crack has to be before it is susceptible to unstable growth.
With the minimum impact properties required for bridge steels, the crack should be large enough to allow discovery during the biannual bridge inspection before fracture occurs. The M270 specification requires average energy in a Charpy V-notch test of 15 ft-lb for grade 36 steels and ranging up to 35 ft-lb for grade 100 steels, at specified test temperatures. More conservative values are specified for FCM members (Art. 11.9). Toughness values depend on the lowest ambient service temperature (LAST) to which the structure may be subjected.
Test temperatures are 70F higher than the LAST to take into account the difference between the loading rate as applied by highway trucks and the Charpy V-notch impact tests.
Allowable Fatigue Stresses for ASD and LFD Design. Members, connections, welds, and fasteners should be designed so that maximum stresses do not exceed the basic allowable stresses (Art. 11.8) and the range in stress due to loads does not exceed the allowable fatigue stress range. Table 11.20A lists allowable fatigue stress ranges in accordance with the number of cycles to which a member or component will be subjected and several stress categories for structural details. The details described in Table 6.27 for structural steel for buildings are generally applicable also to highway bridges. The diagrams are provided as illustrative examples
and are not intended to exclude other similar construction. (See also Art. 6.26.) The allowable stresses apply to load combinations that include live loads and wind. For dead plus wind loads, use the stress range for 100,000 cycles. Table 11.20B lists the number of cycles to be used for design.
Stress range is the algebraic difference between the maximum stress and the minimum stress. Tension stress is considered to have the opposite algebraic sign from compression stress.
Table 11.20A (a) is applicable to redundant load-path structures. These provide multiple load paths so that a single fracture in a member or component cannot cause the bridge to collapse. The AASHTO standard specifications list as examples a simply supported, singlespan bridge with several longitudinal beams and a multi-element eye bar in a truss. Table 11.20A (b) is applicable to non-redundant load-path structures. The AASHTO specifications give as examples flange and web plates in bridges with only one or two longitudinal girders, one-element main members in trusses, hanger plates, and caps of single- or two-column bents.
Improved ASD and LFD Provisions for Fatigue Design. AASHTO has published ‘‘Guide Specifications for Fatigue Design of Steel Bridges.’’ These indicate that the fatigue provisions in the ‘‘Standard Specifications for Highway Bridges’’ do not accurately reflect the actual
fatigue conditions in such bridges; instead, they combine an artificially high stress range with an artificially low number of cycles to get a reasonable result. The actual effective stress ranges rarely exceed 5 ksi, whereas the number of truck passages in the design life of a bridge can exceed many million.
For this reason, these guide specifications give alternative fatigue-design procedures to those in the standard specifications. They are based on a more realistic loading, equal to 75% of a single HS20 (or HS15) truck with a fixed rear axle spacing of 30 ft. The procedures accurately reflect the actual conditions in bridges subjected to traffic loadings and provide the following additional advantages: (1) They permit more flexibility in accounting for differing traffic conditions at various sites. (2) They permit design for any desired design life.
(3) They provide reasonable and consistent levels of safety over a broad range of design conditions. (4) They are based on extensive research and can be conveniently modified in the future if needed to reflect new research results. (5) They are consistent with fatigueevaluation procedures for existing bridges.
The guide specifications use the same detail categories and corresponding fatigue strength data as the standard specifications. They also use methods of calculating stress ranges that are similar to those used with the standard specifications.
Thus, it is important that designers possess both the standard specifications and the guide specifications to design fatigue-resistant details properly. However, there is a prevailing misconception in the interpretation of the term ‘‘fatigue life.’’ For example, the guide specifications state, ‘‘The safe fatigue life of each detail shall exceed the desired design life of the bridge.’’ The implication is that the initiation of a fatigue crack is the end of the service life of the structure. In fact, the initiation of a fatigue crack does not mean the end of the life of an existing bridge, or even of the particular member, as documented by the many bridges that have experienced fatigue cracking and even full-depth fracture of main load-carrying members. These cracks and fractures have been successfully repaired by welding, drilling a hole at the crack tip, or placing bolted cover plates over a fracture. These bridges continue to function without reduction in load-carrying capacity or remaining service life.
Fatigue Provisions for LRFD. The AASHTO load-and-resistance factor design specifications can be best understood by considering a schematic log-log fatigue-resistance curve where stress range is plotted against number of cycles, Fig. 11.5. The curve represents the locus of points of equal fatigue damage. Along the sloping portion, for a given stress range, a corresponding finite life is anticipated. The constant-amplitude fatigue threshold represented by the dashed horizontal line defines the infinite-life fatigue resistance. If all of the stress ranges experienced by a detail are less than the stress range defined by the fatigue threshold, it is anticipated that the detail will not crack.
The LRFD Specifications attempt to combine the best attributes of the Guide Specification, including the special fatigue loading described previously, and those of the Standard Specifications, including the detail category concept. The LRFD Specifications define the nominal fatigue resistance for each fatigue category as
The term (A/N)^1/3 in Eq. 11.18 represents the sloping line in Fig. 11.5, and ( F)TH 1/3 the horizontal line. The multiplier of 1⁄2 represents the ratio of the factored fatigue load to the maximum load. In other words, if the stress range due to the factored fatigue truck is less than 1⁄2 of the constant-amplitude fatigue threshold, the detail should experience infinite life.
The load factor for fatigue is 0.75, Table 11.8. The truck loading for fatigue is shown in Fig. 11.6.
The fatigue resistance defined in LRFD is similar to that in earlier specifications, although the format is different. Complete LRFD design examples, including fatigue designs of typical girder details, have demonstrated that design in accord with the LRFD Specifications is basically equivalent to design in accordance with the provisions for redundant structures in
the Standard Specifications. In developing the LRFD provisions, it was determined that because of the greater fracture toughness specified for non-redundant structures, a reduction in allowable stress range for such structures was unnecessary.
An understanding of the fatigue susceptibility of various details is important for the design of reliable structures. Numerous references are available to maintain familiarity with the state of the art, including:
Fisher, J. W., Frank, K. H., Hirt, M. A., and McNamee, B. M. (1970). Effect of Weldments on the Fatigue Strength of Steel Beams, NCHRP Report 102. Highway Research Board, Washington, DC.
Fisher, J. W., Albrecht, P. A., Yen, B. T., Klingerman, D. J., and McNamee, B. M. (1974).
Fatigue Strength of Steel Beams with Transverse Stiffeners and Attachments. NCHRP Report 147. Highway Research Board, Washington, DC.
Fisher, J. W., Hausammann, H., Sullivan, M. D., and Pense, A. W. (1979). Detection and
Repair of Fatigue Damage in Welded Highway Bridges. NCHRP Report 206. Transportation Research Board, Washington, DC.
Fisher, J. W., Barthelemy, B. M., Mertz, D. R., and Edinger, J. A. (1980). Fatigue Behavior
of Full-Scale Welded Bridge Attachments. NCHRP Report 227. Transportation Research Board, Washington, DC.
Fisher, J. W. (1974). Guide to 1974 AASHTO Fatigue Specifications, American Institute of Steel Construction, Chicago, Ill.
Keating, P. B. and Fisher, J. W. (1986). Evaluation of Fatigue Test Data and Design Criteria. NCHRP Report 299, Transportation Research Board, Washington, DC.