N. Yamamoto, R. Braga Nogueira Branco, O. Cook, G. Mu, A.P. Argüelles, C.E. Bakis
Pennsylvania State University,
Keywords: magnetic assembly, ultrasonic testing, nanocomposite
Summary:Development of multi-scale engineered materials for aerospace and other applications still faces a challenge to control and evaluate nanofiller morphology across macro-sized structures. For instance, carbon nanotubes (CNTs) are expected to provide effective mechanical and transport reinforcement and have been integrated to fiber reinforced plastics (FRPs) in the past 20+ years. Analytical and simulation studies indicate that CNT morphology and associated CNT-CNT/CNT-epoxy interphase properties dominantly determine composite properties. Yet, reinforcement effects by CNTs, when studied collectively, are rather puzzling; for example, while aligned and surface-functionalized CNTs are expected to provide more toughening based on modeling, randomly oriented, pristine CNTs have been experimentally shown to provide higher toughening effect. This experiment-vs-model discrepancy is partially due to lack of inspection of CNT morphology across large enough volumes to properly represent their contribution to reinforcement. More specifically, CNTs can be wavy, non-uniformly dispersed, and potentially not in contact with the matrix within agglomerations. With 2D inspection, high resolution imaging across a large volume can become expensive, and still cannot provide 3D morphology information. 3D inspection by microCT is not suitable because the material contrast between CNTs and polymer-based matrices is poor and the resolution is generally not high enough for CNTs. In this work, we introduced two techniques to control and evaluate CNT morphology across CNT-epoxy nanocomposite samples (~cm order). First, magnetic assembly was used, together with surface functionalization, to dissociate CNT agglomeration from CNT volume fraction. Multi-walled CNTs were synthesized (~6-9 nm diameter, wall number of 2-4) by chemical vapor deposition, magnetized by e-beam evaporation of nickel (~80 nm target thickness), surface-functionalized by diazotization, dispersed within a bisphenol-F based epoxy (EPON 862), and applied with unidirectional static magnetic fields. Application of smaller field (180 G) enabled constant toughness increase (up to by ~71 %) with CNT volume fraction (up to 0.5 vol%), indicating that this CNT assembly method can delay inevitable CNT agglomeration with increasing fraction. Second, ultrasonic immersion testing (20 MHz) was used to capture heterogeneous structural features impacting acoustic properties across the composites. Wave speed and attenuation were mapped across the CNT-epoxy nanocomposites (0.25 mm step size); these parameters have been associated with the effective properties such as stiffness, which originate from the combined effects of CNT morphology and interface properties, and porosity existence. The heterogeneity of these acoustic parameters across the sample was observed to increase with magnetic field application. This work is currently being extended further. The magnetic CNT assembly method is applied to the out-of-autoclave fabrication of CNT-integrated CFRPs. Methods are being contemplated to virtually create 3D CNT morphology based on 2D TEM images using deep-learning algorithms, as a more precise element for modeling.