We don't design a structure and then calculate its design life. Based on the importance of the structure, we design it for a targeted design life. This target design life takes care of many factors due to which design load may increase and load carrying capacity of the structure reduces. Hence in normal circumstances, the structure is expected to behave normally during this designed service life.
Yes, in most cases an operational life of a structure is programmed on the stage of its designing. But even in this cases some techniques are used for life estimation. Usually such techniques or calculation (FEA simulation) uses S-N curves and andurance limit of material, obtained from it. At non-monotonic cyclic loading so caled rainflow method is used. Also there are cases when it is needed to estimate a residual life of existing structure.
My question is about the accuracy of existing approaches to life-time estimation in all cases.
Can someone share such information from the personal experiance or give references to literature data?