share article

Share on facebook
Share on twitter
Share on linkedin

Training neural networks for machining applications


By Orhan Güngör,Burdur Mehmet Akif Ersoy University, and Abdülkadir Çakir, Associate Professor, Isparta Uygulamalı Bilimler University, both in Turkey

Machining is one of the most important techniques in modern-day manufacturing, and it continually evolves. One vital raw material in machining is the nickel-based super-alloy. More recently, nickel-based super-alloys have become more widespread, and especially in the aviation sector, industrial gas turbines, space applications, engines, nuclear reactors, submarines, steam production facilities, petrochemical devices, heat-resistant applications, and many more.

Each of these sectors uses machining tools, which are subject to many forces that leads to excessive vibrations that cause deterioration and damage as well as affect processing stability and quality. To develop a stable processing strategy, it is necessary to have a tool and a machine vibration model that will help analyse the forces at play and help identify the right solutions to mitigate them or avoid them entirely.

Relying on studies

In the machine industry, tool wear and surface roughness are important in determining cutting conditions. Studies have shown that determining the exact values for, say material and tool lubrication, for various cutting conditions allows a more economical and efficient cutting process.

To reduce speed and surface roughness on the cutting tool, the cutting process is performed under high pressure. One of the intentions behinds applying pressure is to reduce fractures on the cutting edge and the working pieces, and hence reduce vibrations. Increasing the pressure of the cooling water allows to increase the cutting speed.

Predicting dimensional changes in cutting tools during the cutting process is especially important for understanding the machine’s functionality and the likely lifespan of the tools. Variables such as cutting speed, depth and feed rate play an important role in predicting vibrations. Anova variance analysis testing can also be used to determine the effects of various cutting parameters.

The magnitude of the vibrations on the cutting tool is indicative of an increase in surface roughness. It is possible to say that during the cutting process of aluminium, light steel and PVC materials, cutting depth has minimum effect on tool vibration, whilst cutting force has maximum effect.

The system used for determining cutting force and vibration during cutting under high cooling-liquid pressure during turning is shown in Figure 1. The Inconel 718 super alloy material was subjected to the turning process under various water pressure values. Cutting forces (progress and tangential) were measured with a dynamometer, whilst displacement was measured with a vibration meter. Surface roughness and tool wear values were determined based on experiment parameters. Using experimental data, the cutting parameters (cutting speed, feed rate, cutting depth) and material properties – which together constitute the vibration-related factors – were collected in a database. Data was first taken into the Labview program, after which training was performed in Matlab on an artificial neural network (ANN). This enabled to prepare vibration graphs to show tool wear, surface roughness and cutting forces according to different input values.

Share this article

Share on facebook
Share on twitter
Share on linkedin

Related Posts

View Latest Magazine

Subscribe today

Member Login