Abstract
The residual safety assessment of steel structures, an important subject in practice, is given to much attention. Life prediction in the planning course of steel structures under fatigue loading is mainly based on fatigue design criteria resulting from S-N curves. But for any reason cracks have to be assumed due to fabrication failures or fatigue loading in service which can lead total fracture of structures. The life prediction can be carried out by means of fracture mechanics using Paris-Erdogan equation($da/dN=C {\cdot}{\Delta}K^m$). The paper presents results from charpy test to interpret transition behaviour of charpy energy($A_V$) in a wide temperature range and from constant-load-amplitude test to measure fatigue crack growth of various steels widely used in steel bridges since beginning of 20 centuries in Europe. In the normal service temperature range of steel bridges, the steel S355M shows higher maximum charpy energy($A_{Vmax}$) and lower transition temperature($T_{AVmax/2}$) than other steels considered. The C and m of Paris-Erdogan equation on the steels appear to be correlated, and to be affected by the R-ratios due to crack closure, especially at a low fatigue crack growth rate. Scanning electron microscopy analysis was carried out to interpret an influence of the crack closure effects on the correlation of C and m.
