Sood, Anoop Kumar (2011) Study on Parametric Optimization of Fused Deposition Modelling (FDM) Process. PhD thesis.
![[img]](http://ethesis.nitrkl.ac.in/style/images/fileicons/application_pdf.png) | PDF 5Mb |
Abstract
Rapid prototyping (RP) is a generic term for a number of technologies that enable fabrication of physical objects directly from CAD data sources. In contrast to classical methods of manufacturing such as milling and forging which are based on subtractive and formative principles espectively, these processes are based on additive principle for part fabrication. The biggest advantage of RP processes is that an entire 3-D (three-dimensional) consolidated assembly can be fabricated in a single setup without any tooling or human intervention; further, the part fabrication methodology is independent of the mplexity of the part geometry. Due to several advantages, RP has attracted the considerable attention of
manufacturing industries to meet the customer demands for incorporating continuous and rapid changes in manufacturing in shortest possible time and gain edge over competitors. Out of all commercially available RP processes, fused deposition modelling (FDM) uses heated thermoplastic filament which are extruded from the tip of nozzle in a prescribed manner in a temperature controlled environment for building the part through a layer by layer deposition method. Simplicity of operation together with the ability to fabricate parts with locally controlled properties resulted in its wide spread application not only for prototyping but also for making functional parts. However, FDM process has its own demerits related with accuracy, surface finish, strength etc. Hence, it is absolutely necessary to understand the shortcomings of the process and identify the controllable factors for improvement of part quality. In this direction, present study focuses on the improvement of part build methodology by properly controlling the process parameters. The thesis deals with various part quality measures such as improvement in dimensional accuracy, minimization of surface roughness, and improvement in mechanical properties measured in terms of tensile, compressive, flexural, impact strength and sliding wear. The understanding generated in this work not only explain the complex build mechanism but also present in detail the influence of processing parameters such as layer thickness, orientation, raster angle, raster width and air gap on studied responses with the help of statistically validated models, microphotographs and non-traditional optimization methods.
For improving dimensional accuracy of the part, Taguchi‟s experimental design is adopted and it is found that measured dimension is oversized along the thickness direction and undersized along the length, width and diameter of the hole. It is observed that different factors and interactions control the part dimensions along different directions. Shrinkage of semi molten material extruding out from deposition nozzle is the major cause of part dimension reduction. The oversized dimension is attributed to uneven layer surfaces generation and slicing constraints. For recommending optimal factor setting for improving overall dimension of the part, grey Taguchi method is used. Prediction models based on artificial neural network and fuzzy inference principle are also proposed and compared with Taguchi predictive model. The model based on fuzzy inference system shows better prediction capability in comparison to artificial neural network model.
In order to minimize the surface roughness, a process improvement strategy through effective control of process parameters based on central composite design (CCD) is employed. Empirical models relating response and process parameters are developed. The validity of the models is established using analysis of variance (ANOVA) and residual analysis. Experimental results indicate that process parameters and their interactions are different for minimization of roughness in different surfaces. The surface roughness responses along three surfaces are combined into a single response known as multi-response performance index (MPI) using principal component analysis. Bacterial foraging optimisation algorithm (BFOA), a latest evolutionary approach, has been adopted to find out best process parameter setting which maximizes MPI.
Assessment of process parameters on mechanical properties viz. tensile, flexural, impact and compressive strength of part fabricated using FDM technology is done using CCD. The effect of each process parameter on mechanical property is analyzed. The major reason for weak strength is attributed to distortion within or between the layers. In actual practice, the parts are subjected to various types of loadings and it is necessary that the fabricated part must withhold more than one type of loading simultaneously.To address this issue, all the studied strengths are combined into a single response known as composite desirability and then optimum parameter setting which will maximize composite desirability is determined using quantum behaved particle swarm optimization (QPSO).
Resistance to wear is an important consideration for enhancing service life of functional parts. Hence, present work also focuses on extensive study to understand the effect of process parameters on the sliding wear of test specimen. The study not only provides insight into complex dependency of wear on process parameters but also develop a statistically validated predictive equation. The equation can be used by the process planner for accurate wear prediction in practice. Finally, comparative evaluation of two swarm based optimization methods such as QPSO and BFOA are also presented. It is shown that BFOA, because of its biologically motivated structure, has better exploration and exploitation ability but require more time for convergence as compared to QPSO.
The methodology adopted in this study is quite general and can be used for other related or allied processes, especially in multi input, multi output systems. The proposed study can be used by industries like aerospace, automobile and medical for identifying the process capability and further improvement in FDM process or developing new processes based on similar principle.
manufacturing industries to meet the customer demands for incorporating continuous and rapid changes in manufacturing in shortest possible time and gain edge over competitors. Out of all commercially available RP processes, fused deposition modelling (FDM) uses heated thermoplastic filament which are extruded from the tip of nozzle in a prescribed manner in a temperature controlled environment for building the part through a layer by layer deposition method. Simplicity of operation together with the ability to fabricate parts with locally controlled properties resulted in its wide spread application not only for prototyping but also for making functional parts. However, FDM process has its own demerits related with accuracy, surface finish, strength etc. Hence, it is absolutely necessary to understand the shortcomings of the process and identify the controllable factors for improvement of part quality. In this direction, present study focuses on the improvement of part build methodology by properly controlling the process parameters. The thesis deals with various part quality measures such as improvement in dimensional accuracy, minimization of surface roughness, and improvement in mechanical properties measured in terms of tensile, compressive, flexural, impact strength and sliding wear. The understanding generated in this work not only explain the complex build mechanism but also present in detail the influence of processing parameters such as layer thickness, orientation, raster angle, raster width and air gap on studied responses with the help of statistically validated models, microphotographs and non-traditional optimization methods.
For improving dimensional accuracy of the part, Taguchi‟s experimental design is adopted and it is found that measured dimension is oversized along the thickness direction and undersized along the length, width and diameter of the hole. It is observed that different factors and interactions control the part dimensions along different directions. Shrinkage of semi molten material extruding out from deposition nozzle is the major cause of part dimension reduction. The oversized dimension is attributed to uneven layer surfaces generation and slicing constraints. For recommending optimal factor setting for improving overall dimension of the part, grey Taguchi method is used. Prediction models based on artificial neural network and fuzzy inference principle are also proposed and compared with Taguchi predictive model. The model based on fuzzy inference system shows better prediction capability in comparison to artificial neural network model.
In order to minimize the surface roughness, a process improvement strategy through effective control of process parameters based on central composite design (CCD) is employed. Empirical models relating response and process parameters are developed. The validity of the models is established using analysis of variance (ANOVA) and residual analysis. Experimental results indicate that process parameters and their interactions are different for minimization of roughness in different surfaces. The surface roughness responses along three surfaces are combined into a single response known as multi-response performance index (MPI) using principal component analysis. Bacterial foraging optimisation algorithm (BFOA), a latest evolutionary approach, has been adopted to find out best process parameter setting which maximizes MPI.
Assessment of process parameters on mechanical properties viz. tensile, flexural, impact and compressive strength of part fabricated using FDM technology is done using CCD. The effect of each process parameter on mechanical property is analyzed. The major reason for weak strength is attributed to distortion within or between the layers. In actual practice, the parts are subjected to various types of loadings and it is necessary that the fabricated part must withhold more than one type of loading simultaneously.To address this issue, all the studied strengths are combined into a single response known as composite desirability and then optimum parameter setting which will maximize composite desirability is determined using quantum behaved particle swarm optimization (QPSO).
Resistance to wear is an important consideration for enhancing service life of functional parts. Hence, present work also focuses on extensive study to understand the effect of process parameters on the sliding wear of test specimen. The study not only provides insight into complex dependency of wear on process parameters but also develop a statistically validated predictive equation. The equation can be used by the process planner for accurate wear prediction in practice. Finally, comparative evaluation of two swarm based optimization methods such as QPSO and BFOA are also presented. It is shown that BFOA, because of its biologically motivated structure, has better exploration and exploitation ability but require more time for convergence as compared to QPSO.
The methodology adopted in this study is quite general and can be used for other related or allied processes, especially in multi input, multi output systems. The proposed study can be used by industries like aerospace, automobile and medical for identifying the process capability and further improvement in FDM process or developing new processes based on similar principle.
| Item Type: | Thesis (PhD) |
|---|---|
| Uncontrolled Keywords: | Rapid prototyping; Polymers; Extrusion; Quality control testing; Dimensional accuracy; Surface roughness; Mechanical strength; Wear; Design of experiment; Grey Taguchi; Principal component; Composite desirability; Artificial neural network; Fuzzy logic; QPSO; BFOA; Multi objective optimization. |
| Subjects: | Engineering and Technology > Mechanical Engineering > Machine Design |
| Divisions: | Engineering and Technology > Department of Mechanical Engineering |
| ID Code: | 3004 |
| Deposited By: | Hemanta Biswal |
| Deposited On: | 01 Feb 2012 14:53 |
| Last Modified: | 13 Jun 2012 17:08 |
| Supervisor(s): | Mahapatra, S S and Ohdar, RajKumar |
Posts
0 comments:
Post a Comment