welding pool
2.In allusion to welding pool of aluminum alloy, the compound filter system which is composed by a 1064nm central wave length narrowband filter and a 0. 1% neutral filters is adopted. Clear welding pool images like taper are collectted. The welding pool of aluminum alloy is mostly covered with blazing arc.
3.SiO2 and V2O5 were served as experimental surface activating fluxes. Vacuum EB welding and DCSP TIG welding were respectively carried out to study the effects of oxide on weld penetration only when flux had influence on arc behavior of TIG welding or welding pool behavior. By this way, mechanism of oxides improving weld penetration of A-TIG welding for aluminum alloy was investigated.
4.Based on the SISO fuzzy-neural network controller--FNNC for regulating welding current in our former work, the expert system controller for regulating welding travel velocity has been designed in this paper, which establishs a DIDO intelligent controller for welding pool dynamic process and realizes simultaneously manipulating the top and back weld widths and molding.
7.With the EN polarity time increasing, the liquid metal of welding rod end congregates and its diameter increases, which shows the arc effect of EN polarity for welding rod consumable, Arc creeps up and the brightness of arc column region decreases with EN polarity of AC PMIG welding, and the brightness of arc column region increases with background level of EP polarity. This kind of AC arc is propitious to reduce the arc plasma jet force and the arc force on molten pool.
8.Using the ideas of fuzzy control,an artificial neural network controller with error,error change and error acceleration as inputs is established for uncertain objects.A self-learning neural network control approach to the controlled objects with uncertainties is presented in this paper.The results of experiment on control of the dynamic process of weld pool in the pulse TIG welding show that the self-learning neural network control scheme presented in this paper is effective.
