Quantification and prediction of lack-of-fusion porosity in the high porosity regime during laser powder bed fusion of Ti-6Al-4V

Quantification and prediction of lack-of-fusion porosity in the high porosity regime during laser powder bed fusion of Ti-6Al-4V

Journal of Materials Processing Technology

10.1016/j.jmatprotec.2021.117426

http://dx.doi.org/10.1016/j.jmatprotec.2021.117426

Patcharapit Promoppatum, Raghavan Srinivasan, Siu Sin Quek, Sabeur Msolli, Shashwat Shukla, Nur Syafiqah Johan, Sjoerd van der Veen, Mark Hyunpong Jhon

Mechanics of Materials, additive manufacturing, porosity

09 November 2021

Promoppatum, P., Srinivasan, R., Quek, S. S., Msolli, S., Shukla, S., Johan, N. S., van der Veen, S., & Jhon, M. H. (2022). Quantification and prediction of lack-of-fusion porosity in the high porosity regime during laser powder bed fusion of Ti-6Al-4V. Journal of Materials Processing Technology, 300, 117426. https://doi.org/10.1016/j.jmatprotec.2021.117426

Although lack-of-fusion porosity due to incomplete melting of powder can limit the mechanical properties of additively manufactured metals, quantification and prediction of these defects remains challenging. We compare three common strategies to measure porosity: the Archimedes, micrograph-based, and micro-computed tomography approaches. We find that while these methods work equally well at low void fraction, their predictions diverge at higher void fractions (> 5%). We find that the disparity comes since the Archimedes method measures the total amount of solid in the sample, while micrograph-based approach neglects loose powder trapped inside the samples that can be removed during the preparation process. We conclude that these two methods make use of divergent definitions of porosity. While the Archimedes method measures a “total porosity” defined by the total volume fraction of void in the material, the micrograph method measures an “effective porosity” that only accounts for the continuous material. On the other hand, the resolution of micro-computed tomography is limited by voxel size, leading to ambiguity of trapped powder being identified as solid or void. Consequently, the number density of defects from micro-computed tomography are noticeably smaller than that from micrograph-based approach. A geometric model for porosity prediction is implemented and used to evaluate different models for melt pool geometry. Our numerical predictions of melt pool profiles are surprisingly insensitive to our choice of heat source models. Finally, an analytical geometric model is developed for fast estimation of total and effective porosities and shows good agreement with both numerical simulations and experimental measurement.

Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

This research / project is supported by the A*STAR - Structural Metals and Alloys Programme
Grant Reference no. : A18B1b0061

This research / project is supported by the A*STAR Singapore Aerospace Programme Consortium and Airbus - Prediction of Cracking in Metal Powder-Bed Additive Manufacturing
Grant Reference no. : A1815a0068

This research / project is supported by the Ministry of Higher Education, Science, Research, and Innovation - N/A
Grant Reference no. : RGNS 63-092

https://oar.a-star.edu.sg/communities-collections/articles/20873

0924-0136

Institute of High Performance Computing