Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Mechanical and Materials Engineering

Supervisor

Straatman, Anthony G.

2nd Supervisor

Henning, Frank

Affiliation

Karlsruhe Institute of Technology

Co-Supervisor

Abstract

Increased pressure on automotive manufactures to reduce overall greenhouse gas emissions from their vehicles has driven research into replacing conventional steel components with carbon fiber reinforced plastics (CFRP). This is due to their high weight specific properties and design freedom. High pressure resin transfer molding (HP-RTM), a derivative of resin transfer molding, is a manufacturing process for CFRP’s that offers high automation potential and low cycle times; critical properties of a manufacturing process for the automotive sector.

The goal of the present work is to increase the accuracy and reduce the computational overhead of current RTM infiltration and curing models. These models allow for the reduction in development and testing costs, as they provide critical insight into the flow characteristics and cure development during said stages. First, a geometry dependent orientation calculator for the permeability tensor is proposed to quickly and accurately orient the fiber direction solely using the mold geometry and mesh required by the infiltration solver. The proposed calculator is verified against an 1D analytical solution of Darcy’s law and then validated against fiber orientation data generated from a constitutive draping simulation. Secondly, a local thermal non-equilibrium (LTNE) energy model is proposed. Due to the addition complexity and computational overhead, a non-dimensional analysis is performed to determine the appropriate operating conditions where the added accuracy of the LTNE model is required. Finally, the advancements in energy modelling approaches are highlight. A study is run on a complex, floor geometry where the three main energy modelling approaches are compared (i.e., isothermal, equilibrium, and LTNE). The results showcase the variability in the predicted cure degree development throughout both the infiltration and curing stage, depending on the energy model used. This highlights the importance of accurately capturing the temperature development, as it can lead to significant variations in the cure degree and cure degree rate development. To further reduce the computational cost an adaptive time step formulation, dependent on the maximum cure rate and minimum cure degree, is proposed. This formulation is shown to reduce the overall computational time of the curing simulation by up to 10 times.

Summary for Lay Audience

The automotive sector experiences continuous pressure to reduce the carbon emissions of their vehicles. This can be done effectively by reducing vehicle weight through replacing conventional steel components with carbon fiber (CF). The drawback to this transition, however, is increased costs given higher material, infrastructure, and testing costs associated with using CF. Another critical consideration is the time required to make each part, as longer manufacturing times correspond to lower production rates. A new method of making CF parts, high pressure resin transfer molding, reduces the manufacturing time by injecting resin at very high rates into the mold. This process is still relatively expensive; therefore, effort is still required to reduce the cost through optimization during the design stage. This is accomplished using simulation techniques, where each stage of the manufacturing process is modelled and then optimized to ensure part consistency. This project will focus on the stages where resin is injected and then left to cure. These are critical stages as they ultimately dictate the quality of the final part.

First, a new method of quickly and accurately modelling the orientation of the carbon fiber sheets within the mold was developed, which has a profound effect on how the resin fills the mold. This method drastically reduces the simulation time normally spent on this stage and the results have been validated against current, more comprehensive methods. Secondly, a more sophisticated model to predict the temperature changes during the injection stage was proposed. The temperature history of the resin plays a significant role is how long the resin cures and can provide valuable insight into if the resin will cure prior to the mold being filled. Finally, the capabilities of the proposed models are tested on a complex, car floor geometry. This study compares the previous temperature modelling approaches in literature to the proposed model to highlight the improvements in accuracy. The results showcase the importance of the proposed model as well as methods that can be readily applied to substantially reduce the computational time.

Download

Share

COinS