Loughborough University
1-s2.0-S2214860422007606-mmc1.docx (600.46 kB)

Supplemental information files for Are classical fibre composite models appropriate for material extrusion additive manufacturing? A thorough evaluation of analytical models

Download (600.46 kB)
posted on 2023-02-02, 14:10 authored by Jiongyi Yan, Emrah DemirciEmrah Demirci, Andy GleadallAndy Gleadall


Supplementary files for article Are classical fibre composite models appropriate for material extrusion additive manufacturing? A thorough evaluation of analytical models

To improve mechanical properties, fibre reinforcement has been used in material extrusion additive manufacturing (MEAM). However, there are few reports about fibre composite mechanics and property prediction in MEAM. This study evaluated seventeen classical composite numerical models to assess their prediction capabilities and applicability in MEAM. Four carbon fibre reinforced polymer composites (PLA, ABS, PA, and PETG) were used. Longitudinal strength and modulus were predicted via the rule of mixtures (RoM), modified rule of mixtures (MRoM), Kelly model, Cox shear lag model, and modified Cox models. Transverse strength was predicted via the RoM transverse model and bridging model. Transverse modulus was predicted via the RoM transverse model, Halpin-Tsai, and modified Halpin-Tsai models. These models are widely accepted in the composite field but have not been assessed for MEAM. The predicted strengths and moduli were normalised by experimental values. For all materials, all models overestimated the longitudinal strength. The best predictions were found for the Kelly model, Cox-Krenchel model, and modified Cox model (with porosity) for longitudinal strength: normalised strengths were 2.02, 1.54, 1.63, respectively, averaged for all materials, compared to 2.46–8.82 for other models. Longitudinal modulus was well-predicted by the Kelly model and Cox-based models (normalised modulus of 0.87–2.05). The RoM transverse model and bridging model accurately predicted transverse modulus and strength, respectively (normalised strength 0.99 and normalised modulus 1.14). Additionally, model efficiency also varied between materials. PA was more predictable for longitudinal strength than PLA, PETG, and ABS (normalised strengths: 3.19, 3.98, 4.56, and 4.81, respectively) and in modulus (normalised moduli: 2.64, 3.01, 2.84, and 2.99), but less predictable in transverse strength (1.74, 2.02, 1.29, and 1.64) and transverse modulus (1.90, 2.02, 0.92, and 1.13). Fractography evidenced non-uniform fibre length, non-uniform fibre orientation, and weak fibre-matrix interface, which likely caused the discrepancies between theoretical and experimental properties. Potential revisions to improve model accuracy were discussed. Factors including fibre orientation, fibre length, fibre clustering, porosity, and fibre-matrix interface were highlighted for better prediction, which facilitates better understanding in mechanics and modelling of 3D printing fibre composites. 



  • Mechanical, Electrical and Manufacturing Engineering