Uncertainty approaches in ship structural performance

Abstract

Uncertainty quantification in ship structural performance remains uneven. While the ship structure community has long developed probabilistic models for the ocean environment and structural responses, complete quantification frameworks remain the exception not the rule. Much of the current uncertainty focus is around design code development, and more piecemeal responses in areas where current design codes are not adequate. This chapter briefly reviews the current state of the art in ship structural performance, including examples of failures in service. Then, the current design code landscape is reviewed. In design codes today, partial safety factors make scattered appearances, but a wide adoption of such measures is still some time off. While more general concepts of risk are gaining traction in the marine industry for risk-based approval, direct structural performance simulation remains the exception rather than the rule. Different uncertainty types and underlying data are then presented and reviewed. Initially, basic design parameters and underlying operational data are discussed, and then strength and loading models are presented. The fragmented nature of uncertainty quantification is clear across all these topics - while many key responses are described in stochastic frameworks, an equal number of modeling, fabrication, and operational parameters are subject to a large amount of epistemic uncertainty today. This limits the number of integrated uncertainty quantification computations that can be completed today. Finally, an overview of published uncertainty frameworks to date is presented. While none of these frameworks are all-encompassing, they demonstrate both the practical application of uncertainty quantification to current problems and a foundation for future probabilistic modeling advances.