Fiber dispersion as a quality assessment metric for pultruded thermoplastic composites

Abstract

Fiber-reinforced polymer composites are often assessed by measuring void content via thresholding of optical microscope cross-sectional images, relying on contrast to differentiate between voids and surrounding material. This is effective when dealing with primarily closed voids, a characteristic common to thermoset matrix composites. However, composites that exhibit fiber impregnation issues, often due to high matrix viscosity as in thermoplastic composites, will typically contain open pores which are infiltrated by the mounting resin used to embed the sample for grinding and polishing. Voids filled with this resin can be difficult to optically distinguish from the matrix, leading to inaccurate void content measurements. To circumvent this, a strategy for assessing impregnation quality based on fiber dispersion is proposed. During impregnation processes, the fiber network exhibits a tendency to become locally compacted by flow fronts, leading to low permeability regions with fiber clusters that are less likely to become fully impregnated. Fiber dispersion quantifies the percentage of non-clustered fibers from positional data and provides information on degree of impregnation and the locations of interfiber voids. The effects of processing speed and temperature on fiber dispersion are investigated in the production of continuous carbon fiber-reinforced polyetherimide (PEI) additive manufacturing feedstock material from commingled yarn using multi-die pultrusion. Fiber dispersion is observed to increase with slower pultrusion speeds, correlating with higher flexural strength and stiffness properties. This increase in fiber dispersion with longer die residency time indicates that the decompaction of fiber clusters is highly time-dependent and occurs by migration of fibers away from cluster boundaries.