Feb/March_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | F E B R U A R Y / M A R C H 2 0 2 1 2 4 tomography (µCT) was performed on cut sections from the centers of printed and partially sintered samples. UNDERSTANDING PORE NETWORKS Filter performance can be inferred by evaluating critical aspects of the flow channels. In the studied metal parts, the flow channels consist of large, con- nected networks of porosity, which are tortuous—full of twists and turns. This tortuous nature of the pore network can be quantified using tortuosity, a media transport descriptor to describe flow behavior [11] . For a permeable binder jet- ted sample intended for use as a filter, the tortuosity describes the flow path- way for air through the pore network. Quantitatively, tortuosity is the ratio of the distance a real particle would trav- el through the network from one point to another over the shortest direct dis- tance between those two points. High tortuosity means more twists and turns for the example particle to traverse, while a tortuosity of unity represents a theoretical, fully porous sample with the lowest possible tortuosity. Image sequences reconstructed using µCT data were thresholded and binarized before import to a Perl-based tortuosity calculator developed by Na- kashima and Kamiya [12] . As anticipat- ed, increased sintering temperatures generally resulted in an increase in tor- tuosity, seen in Fig. 2. Densification ki- netics accelerate with higher sintering temperatures, more quickly pinching off and isolating pore networks and in- creasing the tortuosity as ease of travel decreases for the fluid. Continuing work is focused on developing a MATLAB program that, in addition to tortuosity, Fig. 1 — Flow chart describing the process steps and procedures for iterative conceptualization and fabrication of metal filters with integrated characterization and simulation. will calculate the pore volume fraction of the bulk sample and percentage of total porosity that contributes to the largest pore network. SIMULATING PRESSURE DROP, FLOW AND FILTRATION To accurately model flow and fil- tration characteristics using computa- tional fluid dynamics (CFD), an accurate digital representation of the porous microstructure with fine details is re- quired. This digital representation was created by first importing a µCT image stack into the Synopsys Simpleware ScanIP module, which segmented vox- els associated with solid material to create a 3D model of the porous struc- ture. CFD meshes consisting of tetrahe- dral elements (between 8.9 million and 15.8 million depending on resolution and scan size) were then formed with-