Multi-parametric MINLP optimization study of a composite I beam floor system

Abstract

This paper deals with a comparative study of an optimal design regarding composite steel and concrete floor structures based on the performed multi-parametric mixed-integer non-linear programming (MINLP) approach, and Eurocode specifications. The optimization models COMBOPT (COMposite Beam OPTimization) were developed and cost optimizations of the structure alternatives were performed. A series of MINLP optimizations were executed over a wide range of various design parameters: different material unit prices, hourly labour costs, imposed loads, structure spans, steel and concrete grades, welded and standard steel sections, plastic and elastic cross-sectional resistances, different positions of the neutral axis and different locations of the centre of gravity axis of the transformed (all steel) section.The Modified Outer-Approximation/Equality-Relaxation (OA/ER) algorithm was applied. The minimal self-manufacturing costs of the structure, steel and concrete grades and standard sizes were obtained through each individual MINLP optimization. All the results were analysed and compared. Comparative diagrams and a recommended design for a composite floor system were determined. In addition, this study answered some questions regarding the influence of unit prices on the optimal design, the most suitable position of the neutral axis, the competitive spans of the structures with welded and standard steel sections, the competitiveness of the plastic cross-section resistance when compared to the elastic resistance, the spans and loads where the ultimate and the serviceability limit states are predominant and the adequate location of the centre of gravity axis of the transformed (all steel) section.