Modelling and optimization of nano powder mixed micro WEDM process using artificial neural networks and genetic algorithm /
Micro Wire Electro Discharge Machining (µ -WEDM) is a non-conventional machining process which is used for machining complex structural design and achieving net-shape machining. This machining method is mainly used for conductive materials. However, semiconductor materials like Silicon (Si) can not...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kuala Lumpur :
Kulliyyah of Engineering, International Islamic University Malaysia,
2018
|
| Subjects: | |
| Online Access: | Click here to view 1st 24 pages of the thesis. Members can view fulltext at the specified PCs in the library. |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Micro Wire Electro Discharge Machining (µ -WEDM) is a non-conventional machining process which is used for machining complex structural design and achieving net-shape machining. This machining method is mainly used for conductive materials. However, semiconductor materials like Silicon (Si) can not be effectively machined due to its high resistivity. For this requires some advanced technique to enhance the machining process and efficiency. One technique could be the conductive coating on the workpiece material and use of nano powder mixed dielectric fluid. So far not much research has been conducted to machine Si like materials by using nano powder mixed dielectric fluid. Moreover, there is no intelligent system available that can help the users to select optimal parameters to achieve specific machining goal. One aim at this study is to carry out nano powder assisted micro WEDM for temporarily coated Silicon samples to achieve improved surface finish with more machining efficiency. For this purpose, three different type of nano powders like Aluminium(Al), Silicon (Si) and Graphite (C) were used for machining highly doped Silicon workpiece material to observe the effect of nano powders on the machining process. Before machining the workpiece material (Si) was coated temporarily by a highly conductive material like gold (Au) metal to make the workpiece more conductive during the machining process. The research showed that by using nano powder mixed µ-WEDM process, Material Removal Rate (MRR) was improved by almost ~48% than traditional machining process. However, Spark Gap (SG) was also increased by ~28% for nano powder assisted WEDM as compared to dielectric EDM oil used machining. Further, Al powder mixed WEDM process have resulted higher MRR but less SG than any other powder. It was found that at specific condition (at 80V,13 pF, 0.2g/L powder concentration, 320 nm gold thickness) the Al nano powder mixed dielectric used machining can produce the lowest surface roughness as 26 nm. It was also observed that at lower powder concentration and specific parametric conditions C, Al can easily produce nano range surface roughness where Si powder produces comparatively worse surface roughness than other powders. Therefore, it can be concluded that average surface roughness (ASR) can be improved by maximum ~65% for nano powder assisted machining as compared to conventional WEDM. Another main purpose of this research is to establish an intelligent system that can suggest suitable parameters for nano powder assisted µ-WEDM operation (for Si machining) to achieve certain machining goal. The experimental datasets of this study are used carefully to create a successful predictive model using artificial neural network (ANN). On the basis of the established predictive model, some experiments have been further conducted to assess the validity of the model. Then ANN model has been further optimized by using genetic algorithm (GA) to get required input for optimum output results. Finally, the accuracy of the modelling has been calculated by measuring the error percentage which is less than 5-10% for the model. This infers the modelling efficiency up to 90%. |
|---|---|
| Physical Description: | xxii, 173 leaves : colour illustrations ; 30cm. |
| Bibliography: | Includes bibliographical references (leaves 138-144). |