AI driven identification and parameter adjustment of self-supporting directwrite features
March 8 – 12, 2020
Direct-ink-write provides the capability to produce self-supporting “spanning” features however, the range of print parameters that lead to an acceptable spanning geometry vary with the geometry of the gap to span (i.e. width, height) and material. To analyze spanning segments, an image processing routine is developed and applied to a set of training samples to obtain a set of standardized image representation of the deviation from the ideal span. This standardized representation allows for classification of any linear spanning segment regardless of gap geometry or filament thickness.
Please click Additional Files below to see the full abstract.
Marshall Johnson, Kevin Garanger, Surya Kalidindi, Eric Feron, Dan Berrigan, and James Hardin, "AI driven identification and parameter adjustment of self-supporting directwrite features" in "Innovative Materials For Additive Manufacturing (IMAM)", Daniel Schmidt, Luxembourg Institute of Science and Technology, Luxembourg Nikhil Gupta, New York University, USA Chua Chee Kai, NTU, Singapore Brett G. Compton, University of Tennessee, USA Eds, ECI Symposium Series, (2020). https://dc.engconfintl.org/imam/22