Vacuum Assisted Resin Transfer Molding (VARTM): Model Development and Verification
In this investigation, a comprehensive Vacuum Assisted Resin Transfer Molding (VARTM) process simulation model was developed and verified. The model incorporates resin flow through the preform, compaction and relaxation of the preform, and viscosity and cure kinetics of the resin. The computer model can be used to analyze the resin flow details, track the thickness change of the preform, predict the total infiltration time and final fiber volume fraction of the parts, and determine whether the resin could completely infiltrate and uniformly wet out the preform.
Flow of resin through the preform is modeled as flow through porous media. Darcy's law combined with the continuity equation for an incompressible Newtonian fluid forms the basis of the flow model. During the infiltration process, it is well accepted that the total pressure is shared by the resin pressure and the pressure supported by the fiber network. With the progression of the resin, the net pressure applied to the preform decreases as a result of increasing local resin pressure. This leads to the springback of the preform, and is called the springback mechanism. On the other side, the lubrication effect of the resin causes the rearrangement of the fiber network and an increase in the preform compaction. This is called the wetting compaction mechanism. The thickness change of the preform is determined by the relative magnitude of the springback and wetting deformation mechanisms. In the compaction model, the transverse equilibrium equation is used to calculate the net compaction pressure applied to the preform, and the compaction test results are fitted to give the compressive constitutive law of the preform. The Finite Element/Control Volume (FE/CV) method is adopted to find the flow front location and the fluid pressure. The code features the ability of simultaneous integration of 1-D, 2-D and 3-D element types in a single simulation, and thus enables efficient modeling of the flow in complex mold geometries.
VARTM of two flat composite panels was conducted to verify the simulation model. The composite panels were fabricated using the SAERTEX multi-axial warp knit carbon fiber fabric and SI-ZG-5A epoxy resin. Panel 1 contained one stack of the carbon fabric, and Panel 2 contained four stacks of the fabric. The parameters verified included the flow front location and preform thickness change. For Panel 1, the flow front locations were accurately predicted while the predicted resin infiltration was much slower than measured for Panel 2. The disagreement is attributed to the permeability model used in the simulation, which failed to consider the interface flow in the unstitched preform containing more than one stack of the fabric under very low compaction force. The predicted transverse displacements agree well with the experimental measurement qualitatively, but not quantitatively. The reasons for the differences were discussed, and further investigations are recommended to develop a more accurate compaction model. The simulation code was also used to investigate the VARTM of a new form of sandwich structure with through-the-thickness reinforcements, which is being considered for use in primary aircraft structure. The infiltration of three foam core sandwich preforms with different stitch densities was studied. The objective of the study was to determine whether the preforms could be completely infiltrated and how the stitch density affects the infiltration process. The visualization experiments were conducted to verify the simulation. The model accurately predicted the resin infiltration patterns. The calculated filling times underpredicted experimental times by 4 to 14%. The model revealed the resin flow details and found that increasing the stitch spacing shortens the total filling time, but increases the nonuniformity of the flow front shape. Extreme nonuniformity of the flow front shape could result in the formation of the voids.