The human factor in structural engineering: A source of uncertainty and reduced structural safety

Sammanfattning: It is known that human errors are the cause of most structural failures. Extreme loads or material deficiencies are normally of secondary importance. This Project has studied the human factor in the design phase; how subjective decisions, individual knowledge and the use of advance tools and codes affect structural safety and structural design. 17 Swedish structural engineers from the house-building sector participated in a round-robin investigation. This investigation was divided into two separate tasks; both designed to resemble real design situations, and performed individually by the participants. The first task was the preliminary design of a five storey concrete building and the second was the conceptual design of a 68 m span roof structure. To better understand the results from the investigation, a qualitative interview was held afterwards. The results from the investigation reveal a large variation between engineers. Despite a gross error free result, the ratio between the lowest and highest value of the column design loads of the first task, is approximately three (for the majority of the individual columns). This variation is related to differences in total applied load but, more importantly, to the distribution of loads between columns. In order to describe the importance of subjective decisions performed in the transformation from architectural drawings to computational models, the term Engineering Modelling Uncertainty EMU is introduced. This uncertainty has a large impact on structural safety. The second task resulted in a geometrically uniform truss design. It was found that the majority of the engineers used the architectural sketch as input for the same type of structural analysis software. Yet, the estimated steel weight of the trusses varied between 20 and 50 tons. The most important finding from the interviews is that the majority of the engineers experience a lack of review of calculations from their practice. This may explain why faulty knowledge has developed into biased best practice. Altogether, the study indicates that the use of advance tools and complex design codes prevents young engineers from the development of knowledge and conceptual understanding, as these tools force their users to focus on details rather than the whole problem; in particular if they are used with limited supervision to compensate for lack of knowledge.