Commercial titanium alloy is widely used in the rotating components of aircraft engines. To ensure the safety and longer lifetime of these critical parts, the demand to detect smaller defects becomes more and more important. However, the detection of smaller defects by ultrasonic method in such materials is made difficult by the complicated ultrasound-microstructure interactions, such as the high backscattered grain noise levels and serious signal fluctuations. The objective of this research is to develop a more complete understanding of these phenomena to guide solutions that would address those problems.;In Chapter 1, the relationships between ultrasonic properties and the microstructure are investigated for a series of Ti-6Al-4V forging specimen. Close correlation between the ultrasonic properties and the forging deformation parameters are observed. A model was developed to correlate backscattered grain noise levels with microstructural variations (grain orientation, elongation and texture) due to the inhomogeneous plastic deformation during forging. The model predictions and experiments agree reasonably well.;In Chapter 2, an existing backscattered grain noise theory is extended, leading to a formal theory predicting the spatial correlation of the backscattered grain noise. A special form of the theory for a Gaussian beam is also presented to demonstrate that the material microstructure and the overlap of the incident beam are the important physical parameters controlling the grain noise spatial correlation. The developed theory is validated by the excellent agreements between the predictions and experiments. Physical insights of the results for different setups were discussed.;Ultrasonic signal fluctuations are studied in Chapter 3. The microstructure-induced beam distortions are first explicitly demonstrated. An analytical relationship is then derived to correlate the back-wall P/E spectrum at one transducer location to the through-transmitted field. Based on the analytical relationship and the statistical descriptions of various beam distortion effects, a quantitative Monte-Carlo model is developed to predict the back-wall amplitude fluctuations seen in ultrasonic P/E inspections. The predictions are shown to be in good agreements with experiments. The same modeling approach is used to simulate the flaw (small reflector) signal fluctuation and the results are compared with an independent modeling study. Qualitative agreements are observed.