Effect of heat from weld on fatigue in steel - after the weld has gone

[kodak79]

I am looking at fatigue in a UB member that had a plate welded across the flange, but was subsequently taken away. In accordance with EN 1993-1-9 this would be treated as Detail Category 50 for calculating the fatigue damage with the plate in place.

However, as the plate has been removed and the welds grinded off there will not be a stress concentration due to the plate (Detail Cat. 50). I assume there would be some reduction in allowable stress in the member due to residual stresses from the localised heat from the welding process. My questions are:

1. Is this assumption correct?
2. How can any reduction be quantified?

Many thanks,

Chris

 

[maulidinbastian]

Depend on what type residual stress you have. If you have more compress stress, it will reduce the stress applied. But if you have high tensile stress on residual stress, it will break even on small stress you applied

 

[kodak79]

Do you mean it's dependant on the stress in the member when the weld was applied? Or the type of stress that will be applied in service?

The member is subject to a fully reversible load so each flange will see both compression and tension. The weld was applied when the frame was unloaded (except for dead weight).

 

[metengr]

What you can do, if not performed previously, is have a surface NDT (PT or MT) of the previous weld region to ensure you have no surface cracks. Once this is completed, portable hardness testing can be performed to evaluate the tensile strength of the remnant heat affected zone in the flange base material and decide on an appropriate knock down factor, if any.