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 5 in the segmented solid structures us- ing the Synopsys Simpleware FE mod- ule (e.g., Fig. 3). The tetrahedral meshes represented the bases for single- phase flow and pres- sure drop simula- tions that assumed steady, laminar flow with pressure spec- ified boundaries and flow rates, which spanned the range expected for N95 masks in normal operation. Pressure drop ( ∇ P) versus velocity characteris- tics (v) were used to determine a porous resistance (R) corresponding to a simple viscous relationship of the form: - ∇ P = Rv (Eq 1) Flow resistance ranged from 4.0-5.0 x 10 7 kg/m 3 s for the various 3D-printed and partial- ly sintered filters. These values were used as simulation inputs for the full mask geometry with the filter modeled as a porous region in Ansys Fluent. Particle filtration simula- tions were performed for mesh- es refined by eight times (a factor of two in the three spatial directions) relative to the original mesh count using Lagrangian particle track- ing for discrete phase particles ranging in size from 0.6 to 5 µm. The percent- age of particles trapped as a function of particle size and air flow rate were compared to NIOSH guidelines for N95 masks to determine which metal pow- der/sintering combination adheres to acceptable filtration standards. Results of a typical particle filtration simulation are shown in Fig. 4a where the trajecto- ries of 1 µm particles are superimposed on the computational mesh and show a 57.6% filtration efficiency. Finally, the air flow velocities were simulated for full mask models consist- ing of a porous media (i.e., the porous metal structure) with resistances deter- mined from the resolved pore simula- tions. In Fig. 4b, the uniformity of flow velocity is shown for a candidate mask model with pins to increase the surface area, reducing the overall flow resis- tance and improving the breathability characteristics of the mask. The infor- mation about high and low flow regions of the mask can inform further design iterations to improve filtration efficien- cy and breathability characteristics. OUTLOOK The use of binder jetting technolo- gy and iterative optimization of printing and post-processing parameters to pro- duce reusable, sterilizable metal N95 fil- ters for masks is promising. The filter design, characterization, and optimiza- tion loop can be expedited by using the 3Dmicrostructural characterization and subsequent flow and filtration charac- teristics simulation methods present- ed here. Continuing work is focused on developing the MATLAB program for an expanded view of tortuosity and related parameters, as well as optimizing filter geometry for a balance of breathabil- ity and particle filtration. As the fight against COVID-19 evolves or even disap- pears, the need for effective, reusable mask filters remains. 3D-printed metal filters can be an alternative to current- ly used unwoven or woven masks, and 3D characterization and modeling can help accelerate their development and design. ~AM&P Fig. 2 — Tortuosity plotted against the sintering temperatures for samples from three different stainless steel and copper powder suppliers. Fig. 3 — Example tetrahedral mesh created from µCT data of the porous microstructure for CFD simulation. Fig. 4 — (a) Trajectories of 1 µm particles for a typical sample powder (57.6% efficiency). (b) Velocity contours for a candidate pin filter design.