Date of this Version
Concrete can be subject to various type of damages. Some damages such as Alkali Silica Reaction (ASR) tend to start from the inside, where there is no easy way to identify or evaluate them in the early stage. The study of ultrasonic waves scatterings in non destructive testing (NDT) methods play a major role in identifying such damages. In an ultrasonic test, high frequency ultrasonic waves are used to interrogate the internal structure of concrete material, where the coarse aggregates and microcracks cause multiple scattering of ultrasonic waves. Experimental studies have demonstrated that diffuse ultrasonic waves scattered by boundaries, coarse aggregates, and microcracks are very sensitive to microstructural change in concrete. There are extensive studies implementing various experimental ultrasonic methods to detect and monitor the small changes within the concrete structure. However, there are not many numerical simulations of wave propagation in concrete to study the effects of such small changes in concrete in the receiving signals. In this Thesis, a numerical method to model concretes with microcracks in different damage stages is proposed. First, a finite element model of a concrete sample that includes mortar and a random set of aggregates is simulated. For each damage stage, a series of randomly sized and oriented cracks that are partially filled with the ASR gel is added to the sample. Each damage stage can be quantified based on the number of cracks in a normalized surface area. At each stage, an elastic wave is sent through the sample, and a series of Coda Wave Interferometry (CWI) and ultrasound diffusion approximation is then used to compare the velocity change, diffusivity, and dissipation of the receiving signals. Results suggest there is a direct relationship between the damage stages and the mentioned ultrasonic wave factors. The proposed method can be used for nondestructive evaluation and quantification of the damage such as ASR in concrete structures.
Advisor: Jinying Zhu