Review : Friction Stir Welding Process and Process Parameters

Welding of low melting temperature alloy metals is very difficult through conventional welding techniques. It faces so many problems such as cracks, formation of oxide at weld region, reduce strength of joint, large HAZ area, incomplete penetration, solid inclusion, spatters, etc. Other joining processes like joining through nut & bolt, snap joint, mechanical riveting increase the weight of product, which is not acceptable. Friction stir welding is a stable method for welding low melting temperature alloy of aluminium, Copper etc. It uses a simple and specially designed tool that includes a shoulder and a pin (or probe) that enables thermomechanical operation to join two workpieces. The shoulder is connected to the machine by a vertical machine to transfer axial loads to the work area for friction and placement in the work area. The rotating pin (monitor) moves and mixes the plastic material along its length and moves in the direction of the weld seam. Shoulder size and pin geometry affect the characteristics of the welding equipment. Therefore, tools designed to improve integration are difficult. This connected technology is energy efficient, environmentally friendly and versatile. External filler metal will not be required to join them It can be used to join high-strength aerospace aluminium alloys and other metal alloys that are difficult to weld by conventional fusion welding. FSW is considered the most important development in metallurgy. FSW is also enhance grain size which improves the mechanical property like Strength, fatigue of weld zone than base metals. In order to examine the weld quality, different quality tests will be performed at room temperature. Tensile strength will be measured on UTM machine to check the overall strength of the weld joint. Joint should have sufficient hardness to overcome localized plastic deformation. Artificial Neural Network will be implemented to predict response without further physical experiment. For that Artificial Neural Network methods will be explored to find results with precision.