Theoretical and experimental studies of real-time interrogation of fiber Bragg grating (FBG) sensors based on chirped pulse compression with increased interrogation resolution and signal-to-noise ratio are presented. Two interrogation systems are proposed in this thesis. In the first interrogation system, a linearly chirped FBG (LCFBG) is employed as the sensing element. By incorporating the LCFBG in an optical interferometer as the sensor encoding system, employing wavelength-to-time mapping and chirped pulse compression technique, the correlation of output microwave waveform with a chirped reference waveform would provide an interrogation result with high speed and high resolution. The proposed system can provide an interrogation resolution as high as 0.25 μ at a speed of 48.6 MHz. The second interrogation system is designed to achieve simultaneous measurement of strain and temperature. In this system, a high-birefringence LCFBG (Hi-Bi LCFBG) is employed as a sensing element.

The goal of this study is to determine the intentional movement direction based on the neural signals recorded from the motor cortex using optical brain imaging techniques. Towards this goal, we developed a cross-correlation analysis technique to determine the movement direction from the hemodynamic signals recorded from the motor cortex. Healthy human subjects were asked to perform a two-dimensional hand movement in two orthogonal directions while the hemodynamic signals were recorded from the motor cortex simultaneously with the movements. The movement directions were correlated with the hemodynamic signals to establish the cross-correlation patterns of firings among these neurons. Based on the specific cross-correlation patterns with respect to the different movement directions, we can distinguish the different intentional movement directions between front-back and right-left movements. This is based on the hypothesis that different movement directions can be determined by different cooperative firings among various groups of neurons. By identifying the different correlation patterns of brain activities with each group of neurons for each movement, we can decode the specific movement direction based on the hemodynamic signals. By developing such a computational method to decode movement direction, it can be used to control the direction of a wheelchair for paralyzed patients based on the changes in hemodynamic signals recorded using non-invasive optical imaging techniques.

Indiana University-Purdue University Indianapolis (IUPUI) / Adaptive Optics Optical Coherence Tomography (AO-OCT) is a high-speed, high-resolution ophthalmic imaging technique offering detailed 3D analysis of retina structure in vivo. However, AO-OCT volume images are sensitive to involuntary eye movements that occur even during steady fixation and include tremor, drifts, and micro-saccades. To correct eye motion artifacts within a volume and to stabilize a sequence of volumes acquired of the same retina area, we propose a stripe-wise 3D image registration algorithm with phase correlation. In addition, using several ideas such as coarse-to-fine approach, spike noise filtering, pre-computation caching, and parallel processing on a GPU, our approach can register a volume of size 512 x 512 x 512 in less than 6 seconds, which is a 33x speedup as compared to an equivalent CPU version in MATLAB. Moreover, our 3D registration approach is reliable even in the presence of large motions (micro-saccades) that distort the volumes. Such motion was an obstacle for a previous en face approach based on 2D projected images. The thesis also investigates GPU implementations for 3D phase correlation and 2D normalized cross-correlation, which could be useful for other image processing algorithms.

No description available.

Wave propagation analysis in 1-D and 2-D composite structures is performed efficiently and accurately through the formulation of a User-Defined Element (UEL) based on the wavelet spectral finite element (WSFE) method. The WSFE method is based on the first order shear deformation theory which yields accurate results for wave motion at high frequencies. The wave equations are reduced to ordinary differential equations using Daubechies compactly supported, orthonormal, wavelet scaling functions for approximations in time and one spatial dimension. The 1-D and 2-D WSFE models are highly efficient computationally and provide a direct relationship between system input and output in the frequency domain. The UEL is formulated and implemented in Abaqus for wave propagation analysis in composite structures with complexities. Frequency domain formulation of WSFE leads to complex valued parameters, which are decoupled into real and imaginary parts and presented to Abaqus as real values. The final solution is obtained by forming a complex value using the real number solutions given by Abaqus. Several numerical examples are presented here for 1-D and 2-D composite waveguides. Wave motions predicted by the developed UEL correlate very well with Abaqus simulations using shear flexible elements. The results also show that the UEL largely retains computational efficiency of the WSFE method and extends its ability to model complex features. An enhanced cross-correlation method (ECCM) is developed in order to accurately predict damage location in plates. Three major modifications are proposed to the widely used cross-correlation method (CCM) to improve damage localization capabilities, namely actuator-sensor configuration, signal pre-processing method, and signal post-processing method. The ECCM is investigated numerically (FEM simulation) and experimentally. Experimental investigations for damage detection employ a PZT transducer as actuator and laser Doppler vibrometer as sensor. Both numerical and experimental results show that the developed method is capable of damage localization with high precision. Further, ECCM is used to detect and localize debonding in a composite material skin-stiffener joint. The UEL is used to represent the healthy case whereas the damaged case is simulated using Abaqus. It is shown that the ECCM successfully detects the location of the debond in the skin-stiffener joint.

No description available.

Cross-correlation based analysis of electron backscatter diffraction (EBSD) patterns and the use of simulated reference patterns has opened up entirely new avenues of insight into local lattice properties within EBSD scans. The benefits of accessing new levels of orientation resolution and multiple types of previously inaccessible data measures are accompanied with new challenges in characterizing microscope geometry and other error previously ignored in EBSD systems. The foremost of these challenges, when using simulated patterns in high resolution EBSD (HR-EBSD), is the determination of pattern center (the location on the sample from which the EBSD pattern originated) with sufficient accuracy to avoid the introduction of phantom lattice rotations and elastic strain into these highly sensitive measures. This dissertation demonstrates how to greatly improve pattern center determination. It also presents a method for the extraction of grain boundary plane information from single two-dimensional surface scans. These are accomplished through the use of previously un-accessed detail within EBSD images, coupled with physical models of the backscattering phenomena. A software algorithm is detailed and applied for the determination of pattern center with an accuracy of ~0.03% of the phosphor screen width, or ~10µm. This resolution makes it possible to apply a simulated pattern method (developed at BYU) in HR-EBSD, with several important benefits over the original HR-EBSD approach developed by Angus Wilkinson. Experimental work is done on epitaxially-grown silicon and germanium in order to gauge the precision of HR-EBSD with simulated reference patterns using the new pattern center calibration approach. It is found that strain resolution with a calibrated pattern center and simulated reference patterns can be as low as 7x10-4. Finally, Monte Carlo-based models of the electron interaction volume are used in conjunction with pattern-mixing-strength curves of line scans crossing grain boundaries in order to recover 3D grain boundary plane information. Validation of the approach is done using 3D serial scan data and coherent twin boundaries in tantalum and copper. The proposed method for recovery of grain boundary plane orientation exhibits an average error of 3 degrees.

No description available.

No description available.

No description available.