| Summary: | The use of castellated beam in construction are now applied widely,
because it has advantages, such as capacity of the larger bending moment because of adding to high beams without adding to the self weight, a beautiful artistic value is compared originial I-beam and the hole can be used for mechanicalelectrical installations. Besides having advantages castellated beam also have some disadvantages that the shear force and buckling because of some modifications. To reduce these disadvantages should be optimized dimensionalhexagonal hole to be able to carry the load optimally.
This study was conducted in two stages, the first stage to make numerical
models with hexagonal hole dimensional variations use finite element method
analysis. The second stage to perform simulation result of analysis of numerical
models using Artificial Neural Network to obtain a mathematical function used as a prediction of the most optimal hole dimensions. The input parameters used in the mathematical function is depth and width of the hexagonal holes while output parameters is stress and deformation that occur on the castellated beam.
The result of study shows that the higher and the wider hexagonal hole, so
the larger stress on the web and the smaller the ratio of hole, so the smaller shear force. The result of obtained generalization of data using Artificial Neural
Network to target output is close to finite element method and its average ratio is 1.01.
The result of obtained also using Artificial Neural Network simulation, the most optimal-hexagonal hole the case study castellated beam 225x75x7x5 mm dimension with span of 1 meter and 2 points load amount of 400 kN are : depth of hole = 100 mm, dept of stem = 62.5 mm, width of web post = 80 mm and width of hole = 80 mm.
|
|---|