A Statistical Scrutiny of Three Prominent Machine-Learning Techniques to Forecast Machining Performance Parameters of Inconel 690

2. The Motivation Of The Present Study

Due to the wide usage of heat resistive alloys, the machining of Inconel materials has turn out to be a very crucial subject for investigation in the arena of manufacturing. In this chapter, a tentative inquiry was carried out to recognize the machinability behavior of Inconel 690. Generally, Inconel alloys are portrayed by excellent oxidation, high strength, creep resistance at elevated temperature. These properties are accountable for low machinability, high tool wear, high cutting temperature, and high cutting force. Thus, it is very demanding to machined Inconel alloys in the industrial ambiances.However, computational systems with its unique capability to concurrently get across an appreciably reduced surface roughness, resultant cutting force and cutting temperature footmark as paralleled to conventional milling have been correctly considered to be an imperative technical revolution to grace the milling machines of today.

Numerous alternatives crop up as a solution to these machinability problems; the use of novel tool materials with special cutting geometry, new lubrication strategies and use of refrigerant fluids. In fact, to advance the process efficiency, latest tool materials such as coated carbide tools, coated CBN, PCBN and whisker reinforced ceramics cutting tool are regularly used as cutting tools for machining of Inconel 718 (Li 2006, Muammer et al 2007; Costes, 2007).Where the usage of superior pressure Jet-assisted cooling technology during the machining of heat resistive super-alloys, delivers temperature diminution at the cutting zone (Colak 2012)..Nitrogen has also being worn as a coolant in the milling of diverse heat resistive alloys.

“Black box” predictive approach like ANN does have a confirmed reputation of sensation for certain specific problem fields. But the neural network is still viewed by many of investigators as being magical hammers which can explain any machine learning problems and consequently, tend to apply it indiscriminately to problems for which they are not appropriate.