Numerical Modelling of Manufacturing of Graphene-Modified Hybrid Composites for Structural Applications

Abstract

This thesis develops a flow model that simulates resin impregnation of complex fabric architectures (e.g., woven) to examine the impact of yarn (tow) waviness or undulation on the resin flow. This, furthermore, investigates the feasibility of characterising the micro-mesoscale permeability via a representative volume element (RVE), and for what extent can such a RVE of a single fibre ply yields a value representative of multiple plies in light of distinct stacking sequence and ply orientations. In the case of greater moulding temperatures, the present research develops a thermo-chemo-flow model to quantify convection-dominated species transport phenomena during filling and post-filling stages for a controlled and an optimised LCM manufacturing process. The effect of graphene-based nanofillers on polymerisation progress and heat transfer characteristics are also studied. The developed numerical framework uses ANSYS-Fluent to solve Stokes-Brinkman, energy, and species equations with supplemental source terms accounting heat generation (polymerisation) and micro-permeability. The methodology adopts volume of fluid (VOF) approach based on finite volume method (FVM) discretisation scheme. User-Defined Functions (UDFs) are created along with User-Defined Scalers (UDSs) to extend the ability of the standard ANSYS-Fluent modules for modelling micro-pore (inside undulated yarn regions), cure, and rheology using DEFINE macros. The numerical findings suggest that for a high fibre volume fraction (> 42%) and a higher curvature of woven fibre bundles, the proposed model should be adopted, since the resin flow is affected by a mesoscopic tow curvature that could bring about 14% error in predicting permeability. With reference to multi-scale characteristics of single- and multi-ply fabric(s), it appears that in-plane and through-thickness permeabilities of a single-ply fabric cell are different than that corresponding to a multi-ply cell, making them only applicable for a micro-meso-scale (dual-scale) permeability analysis. Nevertheless, with a RVE of multiply preforms, e.g. 9-ply for in-plane and 5-ply for through-thickness, the dual-scale permeabilities can be a representative of the macro-flow making them applicable at all scales (multi-scale flow). Further, the thermo-chemo-flow solution shows its potential to provide details of flow progression, viscosity variation, chemical conversion, and rate of reaction at once. It also makes it possible to allow monitoring resin infusion (e.g. flow front), controlled boundary conditions, and injection techniques during mould filling and curing stages, and in general, digitally twinning the LCM process. The computational model has been validated with previous research findings for permeability and polymerisation in resin liquid moulding of fibrous porous structures.

Date of Award	17 Jan 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Masoud Jabbari (Supervisor) & Chamil Abeykoon (Supervisor)