期刊名称:International Journal of Engineering Research
印刷版ISSN:2319-6890
出版年度:2014
卷号:3
期号:4
页码:230-234
出版社:IJER
摘要:Hydro forming pro cess is one of the new technolo gies in the manufact uring process in which ho llow pieces a nd sheet works a re fo rmed under fluid p ressure. This pap er d eals with the develop ment of a sheet hydro fo rming (SHF) set up and model to p redict the deformation caused by the hyd ro fo rmi ng usi ng Artificial Neural Networks(ANN). During the s heet hyd ro fo rmi ng (SHF) process, t he hydra uli c pressures were mea sured a nd simultaneousl y the req uired hemispherical products were ob tained with d ome hei ght a nd specific thickness. To va lidat e result of this set up wo rk, a hemispherical pa rt was formed i n a fabri cated hydro forming set up using sing le sheet. The da ta used t o train the Arti ficia l Neural Network (AN N) mod el was collected from sheet hydro forming (SHF) set up tha t wa s designed a nd b uilt in our labo rato ry. Singl e- and two -hidd en-layer feed fo rward neural network mod els were used to capt ure the nonl inear correlat ions between the input(hydraulic p ressure, thickness and dia met er of work piece) a nd o utput (dome hei ght o f deformed p roduct). Neural network mod el was ab le to predict the d eformed dome heig ht, t hick ness of ci rcul ar sheet wi th go od a ccuracy.
关键词:Sh eet hydro -fo rmin g; A NN; Hyd raulic pressu re; Dome ; heig ht.