Please use this identifier to cite or link to this item: http://localhost:8081/xmlui/handle/123456789/6636
Authors: Singh, Sehijpal
Issue Date: 2002
Abstract: The processing of hard and tough materials to meet stringent product quality requirements has put a real challenge to technologists to develop newer techniques. Abrasive Flow Machining (AFM) is one of the latest non-traditional machining processes, which possesses excellent capabilities for finish machining of inaccessible regions on a component. AFM has been successfully employed to deburr, radius, and remove recast layer of precision components. In AFM, a semisolid media, consisting of an abrasive and a polymer based carrier a typical proportion, is extruded under pressure through or across the surface to be machined. The visco-elastic media acts as a deformable grinding tool whenever and wherever it is subjected to any restriction. Notwithstanding the several outstanding capabilities of most non-conventional machining processes, their efficiency in terms of volumetric material removal rate is low. This limitation is also shared by AFM. Literature survey reveals that efforts have been made in the past for the improvement of efficiency and capabilities of many non-traditional machining processes such as ECM, EDM, AJM, and ECG, etc but no effort has hitherto been directed towards the enhancement of efficiency of AFM. Recently, the concept of cross or hybrid machining process (HMP) was introduced with the aim of achieving better performance of the advanced machining processes. Some research studies have led to the successful outcome of this composite technology such as Abrasive water jet machining, Abrasive electro discharge machining, and Electrochemical honing, etc. Orbital flow machining process has been recently claimed to be an improvement over conventional AFM, which performs three-dimensional machining of complex components. The modification of AFM by applying ultrasonic wave in media for machining blind cavities is reported to be another improvement in this field. The present research introduces a new hybrid machining technique, which is the combination of AFM and magnetic abrasive finishing (MAF). This composite process is termed as Magnetically Assisted Abrasive Flow Machining (MAFM). The magnetic field has been considered as one of the input process parameters of ii AFM. The research work has been focused on the following aspects: • Development of experimental set up which should be capable of providing varying range of parameters of MAFM. • Experimental work to study the effect of various process parameters on performance characteristics and also to develop the empirical relationships between important process parameters and response characteristics. • Development of models to characterize the surfaces produced by AFM and MAFM by using Scanning electron microscopy (SEM) and Data Dependent System (DDS) approach. • Simulation study of abrasive action in the presence of magnetic field. • Multi-objective optimization of process parameters of MAFM. An indigenous experimental setup is developed which consists of two media cylinders coupled with pistons. Two flanges clamp the fixture in between the media cylinders. The fixture holds the workpiece and generates the restricted path to the flow of media. A DC electromagnet has been designed and fabricated so that it is integrated to the AFM set up. A cylindrical work piece with central bore was selected as the experimental specimen. The media used in this work contains a mixture of a polymer (polyborosiloxane) and a hydrocarbon gel as a carrier compound. The magnetic A1203 abrasive particles are uniformly mixed in the media in the required proportion. Design of experiment is employed for conducting the experiments and analyzing the data. Various process parameters are grouped according to their nature and the objectives of a particular set of experiment. The different sets of parameters and the techniques used for the experimentation are as follows: Process Parameters o Magnetic flux density, number of cycles, and material of workpiece - Full-factorial (Randomized block design) o Magnetic flux density, concentration of abrasives in media, media viscosity, and polymer to gel ratio - Taguchi Method (La orthogonal array) • Magnetic flux density, extrusion pressure, abrasive grain size, media flow volume, and material of workpiece - Taguchi Method (L27 orthogonal array) o Magnetic flux density, media flow rate and number of cycles - Response Surface Methodology. 111 Response/Performance Characteristics • Material removal • Surface roughness • Scatter of surface roughness Analysis of variance (ANOVA) was performed for all the sets of experiments. The significant parameters were identified and their effect on response parameters was studied. The optimal setting of process parameters was carried out by using Taguchi method. Empirical relations were developed between response parameters and some of the important input process parameters by employing RSM. The experimental results indicate that the abrasive concentration, extrusion pressure, media flow volume and rate, number of cycles, and material of the work piece significantly affect the material removal for a given processing time (MR), and % improvement in IR, (ARO. Abrasive concentration and size, media flow volume, media viscosity, and workpiece material has significant effect on scatter of surface roughness (SSR). Abrasive grain size shows significant effect on (ARa) but not on MR. It was further seen that the application of magnetic field around the work piece while being processed by AFM enhances the MR for non-ferromagnetic work materials. As a consequence, less number of cycles is required for removing the same amount of material from the component, if processed in the presence of magnetic field. The media viscosity and the media flow rate interact with magnetic flux density while affecting MR and AR,. The effect of magnetic field is dominant for low viscosity media and at lower media flow rates. The interaction between extrusion pressure and magnetic flux density was found to significantly affect ARa. The magnetic field is more influential at lower extrusion pressure. Furthermore, the interaction between magnetic field and abrasive grain size is significant. The combination of lowest grain size and highest magnetic flux density causes lowest scatter of surface roughness. The rate of material removal from brass work pieces is higher as compared to that from aluminum work pieces in AFM as well as in MAFM. It was interesting to note that the application of magnetic field around the workpiece improves the surface roughness of brass work pieces but it does not appreciably improve surface roughness of aluminum work pieces. The scatter in roughness values on the work surface remains nearly uniform if the work piece is processed by AFM at varying number of cycles. It however varies iv with the number of cycles in the case of MAFM, particularly during initial cycles. Interestingly, a noticeable decrease in scatter values was observed at higher number of cycles in case of brass work pieces. The employed abrasives and number of process cycles could not cause appreciable MR and AR, for mild steel and stainless steel specimens. The modeling of AFM and MAFM was focused on two aspects: In the first aspect, the characterization of surfaces processed by AFM and MAFM was carried out to investigate the effect of applied magnetic field in AFM. The SEM photographs of the processed surfaces were examined to understand the mechanism of material removal in AFM and MAFM. The SEM photographs of workpiece surfaces of alUminium and brass processed with AFM suggest that metal smearing occurs in case of aluminium while nearly pure abrasion takes place on brass. The SEM photographs of work surfaces processed by MAFM indicate the presence of deep scratches on both brass and aluminium workpieces, particularly during initial cycles. Based upon the examination of SEM photographs, the mechanism of material removal process in AFM has been proposed. A time series based modeling technique (DDS methodology) was used to statistically characterize the surfaces processed by AFM and MAFM. DDS modeling of surfaces suggest that AFM smoothened out the 'Green's function' and it tends to drop to zero rather rapidly as compared to that for unprocessed workpiece. The application of magnetic field resulted in further smoothening of the 'Green's function' and the equilibrium position is acquired much rapidly. The reduction in the order of ARMA models and decay patterns of Green's function indicate that the application of magnetic field around the workpieces being processed by AFM causes enhanced improvement in surface roughness. DDS modeling further suggests that number of 'dynamic active grains' in media increases due to the application of magnetic field around the workpiece being processed with AFM. The second aspect is related to the understanding of the abrasive action during MAFM. An analytical model is proposed which depicts the path of abrasive particles in the workpiece passage during MAFM. The media velocity and the magnetic field strength distribution in the workpiece passage (essential for the analytical model) have been obtained by FEM modeling. This model depicts the path of an abrasive particle, as it moves in the workpiece passage. The simulated path of movement of abrasive particles in the workpiece passage can be used for predicting the angle of impingement and the striking velocity of an abrasive particle on the workpiece surface. For optimizing more than one quality characteristics simultaneously, the multi-objective optimization of the process parameters of MAFM was carried out. In one model, the input parameters of Taguchi based experiment were included. A composite multi-objective optimization model based upon Utility theory and Taguchi method was developed. The unified response consisting of all the response parameters was expressed as a common index (Utility) and studied for the optimal settings using Taguchi approach. The input parameters of the RSM based experiment were included in the second multi-response optimization model. This model is based upon the fuzzy set theory. The response parameters were combined into a single response function, which was solved by using a 'modified random search technique' for obtaining an optimal solution. The developed multi-response optimization models provide the optimal settings of process parameters. The models were validated by conducting the experiments for the predicted optimal settings and then comparing the predicted results with the actual ones. The models were found to be valid within the selected range of process parameters. vi
Other Identifiers: Ph.D
Research Supervisor/ Guide: Shan, H. S.
Kumar, Pardeep
metadata.dc.type: Doctoral Thesis
Appears in Collections:DOCTORAL THESES (MIED)

Files in This Item:
File Description SizeFormat 
TH MIED G11559.pdf25.57 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.