Stochastic fracture of additively manufactured porous composites

Özgür Keleş*, Eric H. Anderson, Jimmy Huynh, Jeff Gelb, Jouni Freund, Alp Karakoç

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

21 Citations (Scopus)
165 Downloads (Pure)


Extrusion-based fused deposition modeling (FDM) introduces inter-bead pores into dense materials, which results in part-to-part mechanical property variations, i.e., low mechanical reliability. In addition, the internal structure of FDMed materials can be made porous intentionally to tailor mechanical properties, introduce functionality, reduce material consumption, or decrease production time. Despite these potential benefits, the effects of porosity on the mechanical reliability of FDMed composites are still unclear. Accordingly, we investigated the stochastic fracture of 241 FDMed short-carbon-fiber-reinforced-ABS with porosity ranging from 13 to 53 vol.% under tensile load. Weibull analysis was performed to quantify the variations in mechanical properties. We observed an increase in Weibull modulus of fracture/tensile strength for porosity higher than ~40 vol.% and a decrease in Weibull modulus of fracture strain for an increase in porosity from 25 to 53 vol.%. Micromechanics-based 2D simulations indicated that the mechanical reliability of FDMed composites depends on variations in bead strength and elastic modulus of beads. The change in raster orientation from 45°/−45° to 0° more than doubled the Weibull modulus. We identified five different types of pores via high-resolution X-ray computed tomography. A 22% and 48% decrease in carbon fiber length due to extrusion was revealed for two different regions of the filament.

Original languageEnglish
Article number15437
JournalScientific Reports
Issue number1
Publication statusPublished - 1 Dec 2018
MoE publication typeA1 Journal article-refereed


Dive into the research topics of 'Stochastic fracture of additively manufactured porous composites'. Together they form a unique fingerprint.

Cite this