Curing of thick thermosetting composites has been studied. The study consists of three parts:​ experiment, simulation, and scaling (modeling). A one-dimensional through-the-thickness experiment using manufacturer recommended cure cycle was performed on a 330-layer laminate. Temperature overshoot and incomplete through-the-thickness consolidation were observed. A two-dimensional experiment was performed to reduce the temperature overshoot. Prebleeding method was used to obtain complete through-the-thickness consolidation.

One-dimensional through-the-thickness and two-dimensional simulations were developed. Heat conduction, kinetic, viscosity, and flow (squeezed sponge model) equations were solved as a coupled system of equations. Control-volume method combined with Alternating Direction Explicit (ADE) method were employed in the solution. Composite physical properties were considered as functions of fiber volume fraction. Experimental results for temperature history and compaction were compared with simulation results. Good agreement was obtained. The simulation was used to study the temperature history, composite compaction, and bleeder effect during cure process. It was found that the prebleeding technique is the most promising method for fabrication of thick composites.

Performing the experiment requires expensive prepreg materials and labor work. Using a simulation is economical. However, simulation requires an accurate expression for kinetic equation which is not available for most resins. To solve those problems, scaling (modeling) has been introduced as an alternative to study the curing process.

Model laws for curing of thermosetting composites were established. The transient heat conduction coupled with kinetic equations and initial and boundary conditions were non-dimensionalized. Dimensionless parameters were extracted from those non-dimensional equations and accordingly model laws were constructed. It is shown how one can use model laws to design and fabricate a less expensive model composite structure for an expensive prototype one. The only feasible way to design a model is to use the same resin, but different fibers. For this case, it was shown that for same initial temperature in the material and same cure cycle for both model and prototype, the temperature and degree of cure at homologous points are the same. Also to transform the model experimental results to prototype for the same resin, it is not necessary to know the resin kinetic equation.