The resin transfer molding (RTM) process has been identified as a cost-effective fabrication technique for producing composite materials from geometrically complex reinforcements. Processing models can be used to determine the temperature and pressure cycles which will produce a finished part with the best properties in the shortest time. This work involved the development and verification of a processing model for RTM. The processing model is based on the assumption that infiltration can be described as flow through a porous medium. Flow through porous media, as governed by D’Arcy’s law, depends on the viscosity of the fluid and the microstructure of the interconnected pores. Infiltration by thermosetting resin systems is assumed to behave as a Newtonian fluid with a time and temperature dependent viscosity. The kinetics of the resin can be described by mathematical expressions determined from standard thermal analysis techniques. The reinforcement is assumed to be a homogenous, anisotropic material which exhibits strain stiffening, hysteresis and plastic deformation. D’Arcy’s law describes the porous material in terms of the material permeability. Kozeny-Carman’s relationship is used to relate the porosity to the permeability. Solution of D’Arcy’s law is accomplished in a quasi-steady state manner by an evolving mesh finite element technique. After infiltration is completed, the model continues to predict the temperature, degree of cure and viscosity of the resin. The equations governing the unsteady heat transfer are solved with an existing cure model by the finite difference method. Results of the processing model include estimates of infiltration, gel and cure times as well as the cured thickness and fiber volume fraction. Test laminates were fabricated, mechanically tested, and compared to prepregged laminate results. Construction of one of the test laminates was simulated with the processing model to verify the accuracy of the simulation. / Master of Science

Viscoelastic polymer materials are being actively considered as a novel material for semiconductor packaging applications as a result of their ability to develop strong adhesive bonds at lower temperatures. Viscoelastic thermoset materials are impacted by the stresses generated during the curing process, which is also accompanied by a dissipation of thermal energy. There is a need to develop a generic modeling formulation that is applicable to any material of interest in order to enable the study of different bonding materials and develop optimized curing cycles. This study reports a numerical formulation to evaluate the stress generated and energy dissipated during the cure of viscoelastic polymers. A generalized method to define the transient variation of degree of cure was developed using a 4th order Runge Kutta approximation. The mathematical formulation was implemented using a novel evaluation methodology that helped reduce the computational power requirement. The commercially-available 3501-6 resin was simulated as a characteristic material in this study. The numerical model was validated against analytically derived solutions for both a single Maxwell model, and a Generalized Maxwell Model (GMM) for cases of constant-strain inputs, and subsequently for sinusoidal strain inputs, wherein, material properties were considered to be constant or varying linearly with degree of cure. A good agreement was obtained between the present model and analytical solutions.

Thesis (M. Tech. - Engineering: Electrical, Department Electronic Engineering, Faculty of Engineering and Technology)--Vaal University of Technology. / Curing of materials in order to obtain different properties has been a practice for many years. New developments in composite materials increase the need to control certain variables during the curing process. One very significant variable is temperature. Temperature control by itself is an old practice, however when the need for repeatedly controlling the process accurately over long periods of time arises, a system is required that outperforms normal manual control. One of the aspects within such a system that needs to be considered is the ability to replicate the temperatures within an oven which were originally used for a specific material’s curing profile. This means that a curing profile would need to be defined, saved for later and finally be interpreted correctly by the controlling system. Different control methods were simulated to enable the system to control the temperature which has been defined by literature. This dissertation introduces a variation on the standard control methods and shows improved results. Switching the oven on and off in order to increase or decrease internal oven temperature seems simple, but can cause switching devices to decrease their operational life span, if not designed carefully. A combination switch was introduced which harnesses the advantages of two very common switching devices to form an improved combination switch. Software for the personal computer environment, as well as software for the embedded environment were developed and formed a control system that produced acceptable results for temperature control. Accuracies of 98% and more were achieved and found to be acceptable according to standard engineering control practices. An accurate temperature profile controller was designed, simulated and built in order to control the temperature inside a specific curing oven which, in turn, determined the curing properties of specific materials. The overall results were satisfactory which lead to achieving the objectives outlined in this dissertation.