In-Situ Observation of Bubble Formation in Neat Resin during the Curing Process by Means of X-Ray Computed Tomography

Because of their outstanding specific strength carbon fibre-reinforced polymers (CFRP) are the favourite material for many lightweight applications, especially in the aeronautic and automotive industries. However, defects like pores in CFRP cannot be completely avoided during manufacturing. For example, moisture in prepreg material can lead to void formation during processing: moisture diffusion forming small water bubbles within epoxy resin that grow or shrink in dependence of pressure and temperature conditions. This work deals with bubble growth in neat out-of-autoclave (OoA) epoxy resin during processing. OoA prepregs can be processed in conventional ovens without high pressure application. The major concern is that the lack of pressure results in voids: entrapped gases/moisture may not be dissolved in the resin. They can conglomerate and form bubbles which finally reduce the laminate's quality. There are multiple models describing bubble growth and collapse due to diffusion and perfect gas law. The aim of this work is to validate the expected bubble growth by means of X-ray computed tomography (XCT). Therefore a simple in-situ setup was built which cures neat epoxy resin within several hours while providing vacuum and the required temperature. Before the curing process water was injected into the neat resin to induce an entrapped water bubble which can be monitored and measured with a laboratory XCT-System. During the investigations a full XCT-scan was performed every five minutes to observe water bubble behaviour during the curing process. This work shows that laboratory XCT-Systems are feasible to observe slow curing processes. Usually synchrotron radiation facilities are used for in-situ observations. In our work, the achieved resolution was (58.7 µm)³ voxel size. Results are discussed in the context of analytical models and help to improve understanding of bubble formation during a curing process.