Civil Engineering Reference
In-Depth Information
and little plastic deformation). Linear elastic fracture mechanics (LEFM) methods
are also applicable to steel railway bridges but are not often used in ordinary steel
bridge design because of lack of information regarding initial crack shapes and sizes
with which to conduct crack growth rate analyses.
Fatigue damage accumulation occurs at stress concentrations in tension zones
∗
making location and detail characteristics of prime importance. These characteristics
are compiled within various fatigue detail categories
†
in AREMA (2008) based on
the number of cycles to “failure,”
‡
N
, at various constant amplitude stress range tests,
Δ
S
re
. Since railway live load is applied as a high cycle (long life) load, testing must
alsobeconductedathighcycleconstantamplitudestressranges.Theallowablefatigue
stress for design of a particular detail is based on a probabilistic analysis (without FS)
of high cycle test data and, therefore, it is appropriate to perform fatigue design at
service load levels. Also, since stress concentration effects are accounted for within
the various fatigue detail categories, a nominal applied stress approach for fatigue
design is recommended in AREMA (2008).
The allowable fatigue stress range for design,
S
rall
, from Equation 5.51, depends
on the number of equivalent constant amplitude stress range cycles over the member
or detail life,
N
,as
Δ
A
N
1
/m
Δ
S
rall
=
(5.57)
or
Log
(N)
=
Log
(A)
−
m
Log
(
Δ
S
rall
)
,
(5.58)
which is plotted in
Figure 5.36
for
m
3 and various values of constant
A
(as shown
established from regression analysis of test results such that Equation 5.58 describes
S
-
N
behavior for details with 95% confidence limits for 97.5% survival (2.5% prob-
ability of failure). Testing has also indicated that there is a constant amplitude fatigue
limit (CAFL) stress range,
=
S
CAFL
, below which no fatigue damage accumulates.
§
The CAFL is also shown in Table 5.5 and by the horizontal lines in Figure 5.36.
2.0
Δ
10
6
cycles is considered an infinite life condition in terms of fatigue testing
(Taly, 1998). In
Table 5.4,
the number of applied equivalent constant amplitude stress
range cycles over the member or detail life,
N
, clearly exceeds 2.0
×
10
6
cycles for
loaded lengths less than 100 ft.Therefore, the allowable fatigue stress range for loaded
lengths or spans less than 100 ft is limited to the to CAFL stress range (Table 5.5),
which provides for infinite life.
×
∗
Therefore, the presence of residual tensile stresses from rolling or welding processes may be important.
†
These are designated as A, B, B
,C,D,E,andE
details according to the number of constant amplitude
stress cycles to “failure” at a given stress range.
‡
“Failure” in terms of fatigue design does not mean failure as defined by the strength limit state. Fatigue
“failure” is a criterion based on data at some standard deviation (generally, 2 or 2.5) from the mean of
test data for the member or detail (AREMA, 2008 uses a standard deviation of 2.5).
§
Even a small number of cycles exceeding the CAFL may effectively render it as nonexistent. Therefore,
fatigue design ensures that all design live load stress ranges are below the CAFL.
