Structural Analysis and Design of Steel Railway Bridges - Design of Modern Steel Railway Bridges - page 209

5.3.2.2.2 Fatigue Design Criteria
The fracture ∗ of steel by fatigue may be caused by modern high magnitude cyclical
railway live loads. Fatigue cracks may initiate and then propagate at nominal tensile †
cyclicalstressesbelowthetensileyieldstressatstressconcentrationsinthesuperstruc-
ture. The cyclical railway loading accumulates damage (which may be manifested as
plastic deformation, crack initiation, and crack extension) at the stress concentrations
that may precipitate fracture, leading to unserviceable deformations or failure at a
certain number of cycles, N f .
The high cycle ‡ fatigue life, N

N f , of a member or detail is determined by
constantamplitudecyclicalstresstestingofspecimenstypicalofsteelbridgemembers
and details. The testing of representative specimens makes the determination of stress
concentration factors and consideration of residual stresses unnecessary for ordinary
steel bridge design. § Therefore, fatigue analysis and design may be performed at
nominal stresses.

≤

5.3.2.2.2.1 Railway Fatigue Loading The variable amplitude cyclical railway
live load must be developed as an effective or equivalent constant amplitude cyclical
design load because fatigue strength is established by constant amplitude cycli-
cal stress testing of materials, components, members, and details. This effective
cyclical fatigue design load must accumulate the same damage as the variable
amplitude cyclical load over the total number of stress cycles to failure. The stress
ratio, R S remin , may be used to

describe constant amplitude loading (Figure 5.33) . Constant amplitude loads are
used for fatigue testing. They are often performed with R

= S remin /S remax , and stress range, Δ S re = S remax −

= 0 (cyclical tension

S min ) . The mean stress ∗∗ with S min = 0) or R =− 1 (fully reversed with S max =−

is S remean = S remin )/ 2.

The variable amplitude cyclical load corresponding to the AREMA (2008) design
load mid-span bending moment is shown in Figure 5.34. The uniformly distributed
load (8000 lb/ft for Cooper's E80 design live load) creates no change in stress and
is shown truncated in Figure 5.34. Actual variable amplitude freight rail traffic loads
measured on in-service bridges are more complex, often making the assessment (life
cycle analysis) of existing bridges difficult from a fatigue perspective. †† The number
of stress range cycles and their magnitudes can be determined directly from load

(S remax +

∗ The fracture limit state can be defined in various ways, such as crack propagation to some critical
length or number of cycles to appearance of a visible crack (generally considered to be on the order
of 1-5 mm in the stress-life approach to fatigue). It is also defined as when initiated fatigue cracks
propagate through the thickness of the component, member, or detail.
† At members and details with a net applied tensile stress, since there is no fatigue cracking in purely
compression regions that never experience tension stress.
‡ High cycle or long life fatigue analysis is appropriate for steel railway bridge design.
§ However, the use of nominal stresses without stress concentration factors should be carefully reviewed
in areas of high stress gradients.
∗∗ Since constant amplitude cyclical fatigue testing of members and details includes the effects of stress
concentrationsandresidualstresses(presentfromrolling,forming,fabricating,andweldingoperations),
the fatigue life is not influenced by mean stress effects and the range of stress is important.
†† Also, lack of records regarding historical rail traffic may hinder assessment.