Process parameter optimization for fused deposition modeling using response surface methodology combined with fuzzy inference system

Fused deposition modeling (FDM) is gaining distinct advantages because of its ability to fabricate the 3D physical prototypes without the restrictions of geometric complexities, while when it comes to accuracy and efficiency, the advantages of FDM is not distinct, and so how to improve them is worth...

Full description

Saved in:
Bibliographic Details
Published in:International journal of advanced manufacturing technology 2014-07, Vol.73 (1-4), p.87-100
Main Authors: Peng, Anhua, Xiao, Xingming, Yue, Rui
Format: Article
Language:English
Subjects:
Artificial neural networks
CAE) and Design
Computer-Aided Engineering (CAD
Deformation
Deposition
Design of experiments
Engineering
Experiments
Extrusion
Fitness
Fused deposition modeling
Fuzzy systems
Genetic algorithms
Industrial and Production Engineering
Inference
Mathematical models
Mechanical Engineering
Media Management
Modelling
Optimization
Original Article
Penalty function
Process parameters
Rapid prototyping
Response surface methodology
Thickness
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fused deposition modeling (FDM) is gaining distinct advantages because of its ability to fabricate the 3D physical prototypes without the restrictions of geometric complexities, while when it comes to accuracy and efficiency, the advantages of FDM is not distinct, and so how to improve them is worthy of study. Focusing on process parameter optimization, such parameters as line width compensation, extrusion velocity, filling velocity, and layer thickness are selected as control factors, input variables, and dimensional error, warp deformation, and built time are selected as output responses, evaluation indexes. Experiment design is assigned according to uniform experiment design, and then the three output responses are converted with fuzzy inference system to a single comprehensive response. The relation between the comprehensive response and the four input variables is derived with second-order response surface methodology, the correctness of which is further validated with artificial neural network. Fitness function is created using penalty function and is solved with genetic algorithm toolbox in Matlab software. With confirmation test, the results are obtained preferring to the results of the experiment 1 with the best comprehensive response among the 17 experiment runs, which confirms that the proposed approach in this study can effectively improve accuracy and efficiency in the FDM process.
ISSN:0268-3768
1433-3015
DOI:10.1007/s00170-014-5796-5