ANFIS-based prediction of moment capacity of reinforced concrete slabs exposed to fire. Issue 4 (May 2016)

• Record Type: Journal Article
• Authors: Bilgehan, Mahmut; Kurtoğlu, Ahmet
• Is Part Of: Neural computing & applications. Volume 27:Issue 4(2016)
• Page Start: 869
• Page End: 881
• Publication Date: 2016-05
• Subjects: High temperature -- Fire -- Reinforced concrete slabs -- Bearing capacity -- Adaptive neuro-fuzzy inference system; Neural networks (Computer science) -- Periodicals; Neural circuitry -- Periodicals; Artificial intelligence -- Periodicals; Neural Networks (Computer) -- Periodicals; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux nerveux -- Périodiques; Intelligence artificielle -- Périodiques; 006.32
• DOI: 10.1007/s00521-015-1902-3
• Languages: English
• ISSNs: 0941-0643

Abstract

Abstract Reinforced concrete slabs, just as the other structural elements, are highly affected by the high temperatures. Due to the decrease in strength of reinforced concrete members under high temperature, bearing moment capacity of reinforced concrete slabs also decreases. In this study, a prediction model is investigated in order to determine the bearing moment capacities of reinforced concrete slabs under high temperature. Pre-calculated moment capacities of slabs exposed to fire are predicted by the implementation of adaptive neuro-fuzzy inference system (ANFIS) and the prediction performance of ANFIS model is investigated. The bending capacities of slabs with different concrete characteristics and different times of exposure are calculated. High temperature resulting from the duration of fire exposure is calculated as a function of time in accordance with ISO 834. The temperature distribution inside the slab is determined by the adoption of a steady-state one-dimensional heat transfer. The slab was separated into slim slices and the heat in each slice is determined depending on the time of exposure. Forces and resistance of materials under fire exposure are calculated by applying the reduction coefficients in Eurocode 2. Results confirm the high prediction capability of ANFIS model.