Alberto M. Cuitino and Gustavo Gioia
Department of Mechanical and Aerospace Engineering
Rutgers--The State University of New Jersey
Piscataway, NJ 08854
Abstract
We report results of a study aimed at ascertaining how
particle cohesion and pouring bias influence the final compact via
the pouring process. We simulate pouring by two
different methods for cohesive and non-cohesive particles. In the
cohesive case, a pre-compaction structure obtains composed of a closed-packed
phase; a low-density, disordered phase; and relatively large
pores. Upon compaction, we expect a stage of particle accommodation
to occur at very low pressure. (This stage has been extensively
documented in experiments.) The final compact should involve a closed-packed
phase mixed with a relatively dense disordered phase.
Recent experimental work with cohesive powders (Naito et al., Powder
Tech., 95, 214--219) shows that an extensive close-packed phase is
indeed present in the final compact. This structure derives
directly from the pouring process, and may therefore be interpreted as
a relic of this process. (We are developing a
particle dynamics compaction program to further verify this result.)
In the non-cohesive case, the pre-compaction structure turns out to be
`polycrystalline', with low-density cubic `grains' separated by high-density
interfaces. Here, no accommodation phase should be present; compaction
must involve particle deformation. We expect the pre-compaction
structure to be preserved in the final compact. We are currently endeavoring
to verify this conjecture.
We also address the influence of small biases in the direction of pouring.
Even a very small bias leads to greatly magnified boundary effects
in the cohesive case, whereas the non-cohesive case remains
relatively unaffected. In conclusion, these preliminary results indicate
that details of the pouring process strongly influence the obtained pre-compaction
structures. We conjecture that the compact is likely to preserve
most of the features directly derived from the pouring process. Although
this conjecture awaits verification, existing experimental results indicate
that it is correct in the case of cohesive particles.