Experimental study and modelling of deformation of rectangular plates subjected to hydrodynamic loading using neural network

In this paper,experimental responses of fully clamped aluminium and steel rectangular plates are presented subjected to hydrodynamic loading. the gmdhtype neural networks (group method of data handling) are then used for the modelling of the mid-point deflection thickness ratio of the rectangular plates using the experimental results. the purpose of this modelling is to display how the variations of the significant parameters changes with the mid-point deflection. in addition,the results indicate that dimensionless input variables provide simpler polynomial expressions in comparison with actual physical parameters. it should be mentioned that,for validation of presented model,the yielded results of modelling are contrasted with experimental results. this comparison demonstrates that the results of modelling have satisfying compatibility with experimental results. in general,regarding the presented model,80% of data are in the reliable range. hence,utilizing the mentioned equations for modelling of the mid-point deflection of rectangular plates is appropriate. © school of engineering,taylor’s university.