Nondeterministic Aspects in Structural Design: Proposals for Classification of Errors

This paper examines the question of human error as a factor in structural failures. Even though increasingly sophisticated tools and methods are available, designs for structures must, necessarily, also be based on the analysis of errors or situations that have arisen in the past; although these past events might be considered “negative” experiences, they do have an important didactic function for future operations. In order to understand such past events, an interpretative reference model must be defined within which the fundamental distinctions and classifications can be organized. This poses many problems that may arise either from the researcher's ability to identify the primary and the secondary causes of a failure or from the type of approach adopted. This paper seeks to show how the interpretation of events may be influenced by different factors, i.e., the availability (or lack) of information, which constitutes a limitation, or the scientific validity of any analyses carried out, which may themselves be limited. Hence, interpretation will be affected by the deterministic level of research adopted by the researcher.