Analyze of colapse modes in the optimum design of alveolar steel beams through the Genetic Algorithm method

Resumo

The application of steel structures in Brazil has grown considerably in recent years, not only
in industrial sectors, such as large warehouses, but also in buildings with a high number of floors.
With this, new solutions in the field have become more studied, such as beams of sequential apertures
in the web (alveolar) that are formed by "I" or "H" laminated profiles previously cut in the web to the
established standard and later displaced in the longitudinal to weld them, thus increasing the capacity
resistant to bending stress without adding weight to the profile. Moreover, these profiles have a great
advantage to the installations, such as electrical and plumbing systems, because the journey by web is
not restricted. The objective of this work is realizing an analysis and comparison of the optimized
sizing of castellated and cellular steel beams by means of genetic algorithm (AG) in a system
implemented in the MATLAB. The optimization process is of greater relevance in the elaboration of
less expensive projects, whereas is found a solution with smaller steel consumption to comply with all
requirements established in standards and resit to the active efforts. To validate and show the
applicability of the proposed problem, we will analyze the sizing of 12 alveolar beams (roof and floor)
for each type of cut, cellular and castellated (Peiner, Litzka, Anglo-Saxão), observing the linear weight
of optimized profiles and the collapse modes that govern the sizing of the same.