Additive Manufacturing of Ti6Al4V: Statistical Modeling  
Author D. Gary Harlow




Abstract Additive manufacturing or 3D printing of metallic structural components is a complex, multi-parameter process. Electron beam additive manufacturing (EBAM) consists of multiple layers of deposited metal and exhibits significant variability in key aspects of the build. In an attempt at establishing consistent and predictable material properties from these builds, one of the key variables identified for consideration is the deposited bead width (BW). It is critical for both the build geometry and microstructure. The desired range for the BW is between 1.0 cm and 1.3 cm; however, it exhibits variations ranging from 0.8 cm to 2.0 cm. The BW is estimated by optical microscopy, and it is statistically modeled with a multimodal cumulative distribution function (cdf) where the underlying modes are characterized with two–parameter Weibull cdfs. Also, the dependence of the BW on other process variables is considered. Some inconsistent recordings for the BW are reported due to discrepancies in the gray scale in close proximity to the extremely hot deposited metal source. In order to eliminate inaccurate BW values, filtering of recorded measurements is investigated. Suggestions for future probabilistic modeling are also made.


Keywords Additive manufacturing, Multimodal distribution, Process parameters, Weibull distribution
    Article #:  2112
Proceedings of the 21st ISSAT International Conference on Reliability and Quality in Design
August 6-8, 2015 - Philadelphia, Pennsylvia, U.S.A.