MARC View

000			03924nam a2200253 a 4500
001			vtls000067736
003			KUKTEM
005			20251117113433.0
008			121212t2012 my a f m 000 0 eng d
020			_aTHE0008059(Local)
039		9	_a201906181324 _bamy2 _c201810031020 _dhuda _c201710091627 _daishah _y201212121512 _zFida
040			_aUMP
090			_aTJ1191 .A84 2012 rs Thesis
100	0		_aMohammad Ashikur Rahman Khan
245	1	3	_aAn experimental investigation and modelling of the electrical discharge machining performance on titanium alloy / _cMohammad Ashikur Rahman Khan
260			_aKuantan, Pahang : _bUMP, _c2012
300			_axxii, 204 p. : _bill. (some col.) ; _c30 cm. + _e1 CD-ROM
502			_aThesis (Doctor of Philosophy in Mechanical Engineering) -- Universiti Malaysia Pahang - 2012
504			_aBibliography: p. 177-195
520	3		_aThis thesis deals with an experimental investigation and modelling of the electrical discharge machining (EDM) performance on titanium alloy Ti-5Al-2.5Sn. Despite enormous applications of lightweight and high-strength titanium alloy, a key problem in machining using conventional machining processes arises. The non-conventional technique, EDM, can machine difficult-to-cut materials effectively. However, in EDM, a complete and clear theory has not yet been established. The proper selection of EDM parameters for the best process performance is still a challenge. Thus, the purpose of the present work is to develop the mathematical models to predict performance characteristics (material removal rate, tool wear rate and surface roughness) along with the optimal parametric set-up of EDM on Ti-5Al-2.5Sn titanium alloy. The peak current, pulse-on time, pulse-off time, servo-voltage, polarity (positive and negative), and electrode material (copper, copper-tungsten and graphite) are considered as process variables. The experimental work was performed based on an experiment design (central composite design). The mathematical models, using the response surface method, and the artificial neural network (ANN) model, using the multilayer perception method, were developed. Analysis of variance (ANOVA) has been performed to verify the fit and adequacy of the developed mathematical models. A confirmation test was conducted to obtain the accuracy of the developed models. In addition, the surface topography of the workpiece was analysed by scanning electron microscopy (SEM). The results evidence that the developed mathematical model can predict the performance characteristics of EDM successfully. The average errors of the mathematical model in predicting material removal rate, surface roughness and tool wear rate were 4.34%, 4.17% and 4.50% respectively. While, the average errors were 2.61%, 2.77% and 3.05% for the ANN model. Thus, the ANN model is more precise than the mathematical model. The negative graphite electrode provides the highest material removal rate. However, it maximizes the tool wear rate, and causes the poorest surface finish. The positive copper-tungsten electrode becomes the best choice in respect of all performance characteristics. It was very difficult to achieve single settings of the process parameters for all the best performance characteristics. In addition, the multiple objectives were incompatible. The surface topography for negative polarity demonstrates larger craters, wider and deeper cracks and greater amounts of globules when compared to positive polarity. The obtained results lead to desirable process output, and cost-effective machining. Therefore, it becomes a precise tool, making the EDM process cost-effective and efficient in the die, mould, tool and other industries.
650		0	_aLaser beam cutting
650		0	_aCarbon dioxide lasers _xIndustrial applications
999			_aVIRTUA40 _c8635 _d8641
999			_aVTLSSORT0010003000500080020003900400090010002450260030005020504052006500650199909992