Abstract
The main objective of this research was to study the relationship between green density and compaction pressure in powdered metallurgy. Powder metallurgy has gained popularity and importance because of its near net shape, cost effectiveness and its ability to reduce the complexity of multileveled engineering components. However, powder metallurgy poses challenges that are yet to be fully understood. There are many works performed to address challenges such as the effect of friction, the tool kinematics, handling component prior to sintering and fracture under compaction. This work concentrates on the relationship between green density distribution and compaction pressure. In order to measure the relative density of compacted components, Electron Scanning Microscope was utilized. One can intuitively conceive that the relative density requires more than intuition. It was determined that highest relative density occurs at the center of the specimen and reduces toward the die-powder or punch-powder boundary. For completeness, the application of artificial neural network (ANN) and finite element (FE) model in estimation of green relative density was studied. The results of this research signify that ANN is an excellent technique to determine the relative density distribution of un-sintered compacted specimen. Moreover, finite element method can accurately estimate the average relative density of compacted specimen.
Highlights
1.1 MotivationPowder metallurgy (PM) is a popular processing technique to fabricate ceramic, metallic or composite components, typically of small size and of complex shape
The goal in this chapter was to evaluate neural network modeling as a method to reproduce density variations during compaction process in powder metallurgy
Neural network was used to predict the green density as the function of compaction pressure and determination of density distribution in any desired location
Summary
1.1 MotivationPowder metallurgy (PM) is a popular processing technique to fabricate ceramic, metallic or composite components, typically of small size and of complex shape. A considerable research effort has gone into the development of ANN data analysis, function approximation, sensor processing, and control [51, 52] Their ability to perform complex non-linear mappings can be used for approximating multiple input, multiple output relationships. This chapter outlines the development of a Finite Element (FE) model to simulate the powder compaction process. A surface model of the die and punch was first developed This matched the size of the test specimens used in this study. The surface model of the powder compact was used to develop a three dimensional model which was subsequently discretized by meshing (Figure 6-2). The mesh sensitivity analysis is the process by which the discrete element size is chosen such that the dependency of the results on mesh size is eliminated. Once the optimum size was determined, the mesh size was consistently used in all other FE analyses
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
