Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi

Yazarlar: Ahmet ÖZBAYRAK

Konular:İnşaat Mühendisliği

DOI:10.28948/ngumuh.523939

Anahtar Kelimeler:Earthquake effect,Reinforced concrete structures,Flat slabs,Finite element method,Artificial neural networks

Özet:    In this study, the bending moments obtained from reinforced concrete flat slab structures under the influence of horizontal forces were analysed by multi-layer artificial neural networks (ANN). In the analysis method with ANN, the variables in the input data related to the structure are storey heights, slab thickness, earthquake forces, axles spacing and column dimensions. As the output data, the resultant bending moments of the slabs were used. In SAP2000 generate this data, ten structures with four storeys were analysed. The Levenberg-Marquardt algorithm in MATLAB software was used to process this data. Accordingly, a proximity of 98.2% was obtained and it was observed that there was no significant difference between ANN analysis results and actual design bending moments. As a result, the use of ANN in the detection of design moments of reinforced concrete flat slabs can be evaluated as advantageous in terms of obtaining fast and reliable results.

---