TWI Frequently asked questions
by Tim Gurney
The answer to this question is strongly influenced by the effect of residual stresses. Consequently, welded and unwelded details are treated in different ways, as follows:
Welded details
In welded details it has to be assumed that high tensile residual stresses will probably exist at the point of crack initiation, and that such residual stresses will effectively convert nominal applied compressive stress ranges into tensile ranges. They will therefore be fully effective in propagating a fatigue crack. In such instances, therefore, the whole stress range, including any compressive component, should be taken into account.
Unwelded details
There are two possible situations. If the stress range, allowing both for dead load and any residual stresses, is entirely compressive, it can be assumed that fatigue failure will not occur. Alternatively, if the resulting stress range involves stress reversals through zero, the effective stress range to be used for fatigue calculations can be deduced by adding 60% of the compressive part of the range to the whole of the tensile part of the range.
Further information
More information on this subject is contained in Section 3 of BS 7608: 1993 Code of Practice for 'Fatigue design and assessment of steel structures'.
See also:
FAQ:
What is the effect of small stress cycles under fatigue loading?
