Concept exploration of bridge structures based on comprehensive topology optimization under minimum representation of design conditions

Abstract

This paper discusses computer-based comprehensive exploration of preferred design principles of bridge structures. A preferred design principle means here a system-level, conceptually optimal structure that differs by each system scale range. When recognizing that each principle is derived from a different system scale under the unique optimality, it could be assumed that classification of comprehensive optimal alternatives derives all principles at once in advance of individual design activities. Such a framework should be effective for conceptual design. This paper demonstrates such a possibility in the case of bridge structures, the preferred design principles of which are explicit historically. After the existing design principles are reviewed, the minimum representation of bridge structures is proposed as abstract but essential description. Following, topology optimization is performed comprehensively for different conditions. By classifying the obtained alternatives with a data-science approach, essential design principles hidden in the design problem is systematically derived. Finally, some future works toward the framework are discussed under such a demonstration.