This paper is based on the experience from investigating over 400 structural collapses, incidents and serious
structural damage cases with steel structures which have occurred over the past four centuries. The cause
of the failures is most often a gross human error rather than a combination of “normal” variations in
parameters affecting the load-carrying capacity, as considered in normal design procedures and structural
reliability analyses. Human errors in execution are more prevalent as cause for the failures than errors in
the design process, and the construction phase appears particularly prone to human errors. For normal
steel structures with quasi-static (non-fatigue) loading, various structural instability phenomena have been
observed to be the main collapse mode. An important observation is that welds are not as critical a cause of
structural steel failures for statically loaded steel structures as implicitly understood in current regulations
and rules for design and execution criteria.
This paper contains a compilation of results from some 50 000 steel specimen tests and close to 5 000
measurements of the cross-sectional properties of rolled steel members. Based on the statistical
distribution of these properties the statistical distribution of the sectional capacity of such steel members is
evaluated using a numerical integration procedure. For standard structural steel members the variations of
the strength properties are reasonably well-known and may be used in reliability assessment methods for
the design of structures. However, it has been observed in many actual failures with steel structures that
the cause of such failures normally is one gross human error, rather than a combination of “normal”
variations in parameters affecting the actions and response of the structures. Another observation from
failures experienced with steel structures is that gross human errors in execution are more critical than
gross errors in the design process.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.