The Woodruff School

SUMMARY

https://teams.microsoft.com/l/meetup-join/19%3ameeting_MTdjOTM2OTAtZTc4OC00YjI1LWFmMzctYTY3MGZiNmM3YjY1%40thread.v2/0?context=%7b%22Tid%22%3a%22482198bb-ae7b-4b25-8b7a-6d7f32faa083%22%2c%22Oid%22%3a%22791f3f21-9eee-4e87-9d72-7539450f63c8%22%7d

Machine learning allows for the ability to predict an output from a diverse hyperspace of inputs. In the context of additive manufacturing, this class of approach could be useful in determining whether a specific measured defect field meets a given qualification requirement, this being particularly relevant for the aerospace and medical industries. The present study will investigate the effect of porosity surface determination on predicting the mechanical properties of AlSi10Mg processed by laser powder bed fusion and characterized using micro-computed tomography. A range of machine learning models will be applied that include support vector machine, neural networks, decision trees, Bayesian classifiers, etc. The effects of isosurface and local thresholding approaches for porosity segmentation, as well as image filtering schemes, on precision performance of the machine learning models will be evaluated for samples produced under differing levels of global energy density (GED).

SUBJECT:	M.S. Thesis Presentation

BY:	Caroline Massey

TIME:	Wednesday, April 21, 2021, 11:00 a.m.

PLACE:	MS Teams, Virtual

TITLE:	Understanding the Role of Segmentation on Structure-Property Predictions Made via Machine Learning

COMMITTEE:	Dr. Christopher Saldana, Chair (ME) Dr. Katherine Fu (ME) Dr. Thomas Kurfess (ME) Mr. David G. Moore (Sandia National Laboratories)

SUMMARY https://teams.microsoft.com/l/meetup-join/19%3ameeting_MTdjOTM2OTAtZTc4OC00YjI1LWFmMzctYTY3MGZiNmM3YjY1%40thread.v2/0?context=%7b%22Tid%22%3a%22482198bb-ae7b-4b25-8b7a-6d7f32faa083%22%2c%22Oid%22%3a%22791f3f21-9eee-4e87-9d72-7539450f63c8%22%7d Machine learning allows for the ability to predict an output from a diverse hyperspace of inputs. In the context of additive manufacturing, this class of approach could be useful in determining whether a specific measured defect field meets a given qualification requirement, this being particularly relevant for the aerospace and medical industries. The present study will investigate the effect of porosity surface determination on predicting the mechanical properties of AlSi10Mg processed by laser powder bed fusion and characterized using micro-computed tomography. A range of machine learning models will be applied that include support vector machine, neural networks, decision trees, Bayesian classifiers, etc. The effects of isosurface and local thresholding approaches for porosity segmentation, as well as image filtering schemes, on precision performance of the machine learning models will be evaluated for samples produced under differing levels of global energy density (GED).