About this Abstract |
| Meeting |
2026 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2026)
|
| Symposium
|
2026 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2026)
|
| Presentation Title |
Hardware-Aware DMD Mask Optimization for Uniform Microscale Selective Laser Sintering of Copper Nanoparticle Films |
| Author(s) |
Farzana Tasnim, Hsiu-Tsu Shui, Pratik Koirala, Michael Cullinan |
| On-Site Speaker (Planned) |
Farzana Tasnim |
| Abstract Scope |
Digital Micromirror Device (DMD)-based microscale selective laser sintering (µ-SLS) enables area-wide processing of metal nanoparticle films in a single exposure, but achieving spatially uniform thermal conditions remains challenging because optical delivery is non-uniform and heat loss is greater near feature boundaries than in the interior. Direct inversion of the measured optical response is additionally limited by the finite dynamic range of an 8-bit DMD: once the commanded grayscale value reaches 255 counts, further local compensation is not physically realizable. This paper presents a hardware-aware computational pipeline for copper µ-SLS that explicitly accounts for these optical, thermal, and actuation limitations during mask design. Measured beam-profile data are used to construct a reduced-order thermal model and a linear-programming (LP) irradiance optimizer. The resulting coarse-grid power field is then converted into a physically realizable 8-bit DMD mask through hardware-aware correction, cold-spot exclusion, edge compensation, interior-intensity floor control, and an exposure-amplification safeguard. Each generated mask is subsequently evaluated using a higher-fidelity transient thermal model before release for sintering. The LP-optimized field produced a predicted heated-region temperature distribution with a mean deviation of 0.627 K from the target. In transient verification, all four hardware-aware masks achieved complete 450 °C sintering-onset coverage within their retained regions and reduced peak-temperature spatial coefficients of variation to 6.40–6.54%, compared with 8.01% for the uncorrected all-on baseline. These results demonstrate a computational framework for identifying hardware-infeasible regions, generating realizable DMD masks, and screening their thermal performance before fabrication. |
| Proceedings Inclusion? |
Planned: Post-meeting proceedings |