Composite metal deck slabs are designed and reinforced at the bottom for positive moments with minimum thermal reinforcement at top. However, when the dynamic impact loading is applied to these slabs without sufficient upper part reinforcements, it may be failed under negative moments. The current study investigates the performance of composite metal deck slabs under free drop weight impact loading. The study is conducted in two main parts: generating a data set using numerical simulation analysis, and training several machines based on the generated data set using machine learning procedures. In the first part, 165 models were produced and analysed using LS-DYNA commercial software. In machine learning part, the FEM results were used as a data set to train the machines and to make prediction on performance of composite metal deck slabs with high accuracy. The main results of the conducted analyses are reported in terms of the maximum negative moment, maximum deflection, and elastic and plastic behaviour of the slab. It was seen that in the cases with high striker velocity, the specimens experienced an ultimate internal negative moment at the range of 60 to 80 kN.m.
Emami, F., & Kabir, M. Z. (2023). Strength prediction of composite metal deck slabs under free drop weight impact loading using numerical approach and data set machine learning. Scientia Iranica, (), -. doi: 10.24200/sci.2023.59582.6321
MLA
Fakhreddin Emami; Mohammad Z. Kabir. "Strength prediction of composite metal deck slabs under free drop weight impact loading using numerical approach and data set machine learning". Scientia Iranica, , , 2023, -. doi: 10.24200/sci.2023.59582.6321
HARVARD
Emami, F., Kabir, M. Z. (2023). 'Strength prediction of composite metal deck slabs under free drop weight impact loading using numerical approach and data set machine learning', Scientia Iranica, (), pp. -. doi: 10.24200/sci.2023.59582.6321
VANCOUVER
Emami, F., Kabir, M. Z. Strength prediction of composite metal deck slabs under free drop weight impact loading using numerical approach and data set machine learning. Scientia Iranica, 2023; (): -. doi: 10.24200/sci.2023.59582.6321
)