Abstract

In 2011, Tohoku earthquake and subsequent tsunami hit Fukushima Daiichi Nuclear Power Plant, and they lead to Fukushima nuclear disaster. For more than a decade after the disaster, concrete materials in submerged structures of nuclear power plants have been chronically in contact with water and resulted in calcium leaching. To assess the alteration of the concrete property, it’s necessary to evaluate the permeability change due to calcium leaching and its effect on radioactive ion diffusion. In this study, small scale cylindrical cement paste specimens with 3mm in diameter and 6mm in height were prepared with water to cement ratios of 0.5 and 0.6 respectively, and they were tested in the static leaching. Specimens were subjected to carbonation before immersed into deionized water for different periods. CT image of each specimen at a resolution of 2.46μm/voxel was acquired in SPring-8, Hyogo, Japan. Combining CT images and X-ray diffraction data, dissolution front of portlandite under different leaching periods was determined, and the time dependent development law of dissolution front was evaluated. Introducing Buil’s model as a local equilibrium, time and position dependent porosity of cement paste due to leaching was calculated and correspondingly diffusion coefficient of Sr ion in the numerical simulation was modified. Simulative results on Sr ion diffusivity for 10-year showed the impact of carbonation followed by leaching on radioactive ion diffusion.

Abstract

This study investigates the fracture toughness of short-fiber Sheet Molding Compound (SMC) laminates —including carbon, glass, and hybrid types— under various strain rates (quasi-static, intermediate, and high-speed). Compact Tension (CT) tests were conducted using in-situ X-ray radiography at the European Synchrotron Radiation Facility (ESRF) to monitor internal crack evolution. The results reveal a clear dependency of fracture toughness on fiber type and loading rate, with glass fiber laminates showing the highest sensitivity and carbon fiber ones remaining largely unaffected. Hybrid laminates demonstrated favorable mechanical synergy, combining strength and ductility, with peak toughness observed at intermediate strain rates. Due to the abrupt nature of crack growth, a neural network model integrated with finite element simulations was developed to accurately predict fracture toughness. This work introduces an innovative methodology for characterizing the fracture behavior of composite materials under dynamic conditions, with significant implications for aerospace and automotive applications.