P Srinivasa Pai
Articles written in Sadhana
Volume 33 Issue 3 June 2008 pp 227-233
Acoustic Emission (AE) has been widely used for monitoring manufacturing processes particularly those involving metal cutting. Monitoring the condition of the cutting tool in the machining process is very important since tool condition will affect the part size, quality and an unexpected tool failure may damage the tool, work-piece and sometimes the machine tool itself. AE can be effectively used for tool condition monitoring applications because the emissions from process changes like tool wear, chip formation i.e. plastic deformation, etc. can be directly related to the mechanics of the process. Also AE can very effectively respond to changes like tool fracture, tool chipping, etc. when compared to cutting force and since the frequency range is much higher than that of machine vibrations and environmental noises, a relatively uncontaminated signal can be obtained.
AE signal analysis was applied for sensing tool wear in face milling operations. Cutting tests were carried out on a vertical milling machine. Tests were carried out for a given cutting condition, using single insert, two inserts (adjacent and opposite) and three inserts in the cutter. AE signal parameters like ring down count and rms voltage were measured and were correlated with ﬂank wear values (VB max). The results of this investigation indicate that AE can be effectively used for monitoring tool wear in face milling operations.
Volume 40 Issue 2 April 2015 pp 515-535 Mechanical Sciences
Dual fuel engines are being used these days to overcome shortage of fossil fuels and fulfill stringent exhaust gas emission regulations. They have several advantages over conventional diesel engines. In this context, this paper makes use of experimental results obtained from a dual fuel engine for developing models to predict performance and emission parameters. Conventional modelling efforts to understand the relationships between the input and the output variables, requires thermodynamic analysis which is complex and time consuming. As a result, efforts have been made to use artificial intelligence modelling techniques like fuzzy logic, Artificial Neural Network (ANN), Genetic Algorithm (GA), etc. This paper uses a neuro fuzzy modelling technique, Adaptive Neuro Fuzzy Inference System (ANFIS) for developing prediction models for performance and emission parameter of a dual fuel engine. Percentage load, percentage Liquefied Petroleum Gas (LPG) and Injection Timing (IT) have been used as input parameters, whereas output parameters include Brake Specific Energy Consumption (BSEC), Brake Thermal Efficiency (BTE), Exhaust Gas Temperature (EGT) and smoke. In order to further improve the prediction accuracy of the model, GA has been used to optimize ANFIS. GA optimized ANFIS gives higher prediction accuracy of more than 90% for all parameters except for smoke, where there is a substantial improvement from 46.67% to 73.33%, when compared to conventional ANFIS model.