Single Input Multi Output Adaptive Network Based Fuzzy Inference System for Machinability Data Selection in Turning Operations

The selection of machining parameters needs to be automated, according to its important role in machining process. This paper proposes a method for cutting parameters selection by fuzzy inference system generated using fuzzy subtractive clustering method (FSCM) and trained using an adaptive network based fuzzy inference system (ANFIS). The desired surface roughness (Ra) was entered into the first step as a reference value for three fuzzy inference system (FIS). Each system determine the corresponding cutting parameters such as (cutting speed, feed rate, and depth of cut). The interaction between these cutting parameters were examined using new sets of FIS models generated and trained for verification purpose. A new surface roughness value was determined using the cutting parameters resulted from the first steps and fed back to the comparison unit and was compared with the desired surface roughness and the optimal cutting parameters ( which give the minimum difference between the actual and predicted surface roughness were find out). In this way, single input multi output ANFIS architecture presented which can identify the cutting parameters accurately once the desired surface roughness is entered to the system. The test results showed that the proposed model can be used successfully for machinability data selection and surface roughness prediction as well.

Statistical Wear Model for Adhesively Bonded Tools

Computerized machinability data systems are essential for the selection of optimum conditions during process planning, and they form an important component in the implementation of computer integrated manufacturing (CIM) systems. Since statistical models for adhesively bonded tools are unavailable, the present paper presents a study of the development of a tool life, surface roughness and cutting force models for turning constructional steels, using adhesively bonded tools. These models are developed in terms of cutting speed, federate and depth of cut. These variables are investigated using design of experiments and utilization of response surface methodology (RMS).

An integrated approach to the high-speed machining of moulds and dies involving both a knowledge-based system and a chatter detection and control system

Over the past decade, high-speed machining (HSM) has provided a step change in productivity for the manufacture of moulds and dies, particularly those made from tool steels in the fully hardened state. Advantages include reduced machining times and costs and favourable workpiece surface integrity. The knowledge and expertise that have been generated in academia and industry are widespread and important for the efficient utilization of HSM. Following a survey of relevant publications, the present paper details the development of an integrated system for the HSM of moulds and dies. This included the design and implementation of a knowledge-based system (KBS) and a chatter detection and control system. The main modules of the KBS related to machine tools, cutting tools and machinability data, toolholders, cutter path strategies and troubleshooting, together with a database management centre. Ranking techniques were employed with combined weighing factors, together with tool rationalization and reduction strategies, fuzzy logic and machinability data optimization. The chatter detection and control system utilized a microphone to detect chatter and employed online spindle speed and feed rate modification strategies for its suppression. The system was capable of being interfaced with the KBS in order to transfer and store local knowledge unique to a particular toolmaker's operations.