First, we investigate the Bouguer-Lambert-Beer (BLB) law as applied to the transmission of ultrashort pulses through water in the linear absorption regime. We present a linear theory for propagation of ultrashort laser pulses, and related experimental results are in excellent agreement with this theory. Thus we conclude that recent claims of the BLB law violations are inconsistent with the experimental data obtained by our group. Second, we study the dynamics of ultrashort pulses in a Lorentz medium and in water via the saddle point method. It shows that the saddle point method is a more efficient and faster method than the direct integration method to study one-dimensional pulse propagation over macroscopic distances (that is, distance comparable to the wavelength) in a general dielectric medium. Comments are also made about the exponential attenuation of the generalized Sommerfeld and Brillouin precursors. By applying the saddle point method, we also determined that the pulse duration estimated by the group velocity dispersion (GVD) approximation is within 2% of the value computed with the actual refractive index for a propagation distance of 6 m in water.

The research focuses on the critical review and prediction of flash points of high molecular weight compounds used mainly in the specialty chemical area. Thus far this area of high molecular weight specialty chemicals has not been thoroughly reviewed for flash point prediction; therefore critical review for accuracy of experimental values is difficult. Without critical review, the chance of hazards occurring in the processing and handling of these compounds increases. A reliable method for making predictions is important to efficiently review experimental values since duplicate experimentation can be time consuming and costly. The flash point is strongly correlated to the normal boiling point (NBP) but experimental NBP is not feasible for chemicals of high molecular weight. The reliability of existing NBP prediction methods was found inadequate for our compounds of interest therefore a new NBP prediction method was developed first. This method is based on ten simple group contributions and the molecular weight of the molecule. The training set included 196 high molecular weight C, H, N and O compounds. It produced an average absolute error (AAE) of 13K, superior to any other model tested so far. An accurate NBP is essential for critical review and new method development for flash point. A preliminary data analysis based on chemical family analysis allowed for detection of erroneous data points. These compounds were re-tested at a Huntsman facility. With a predicted normal boiling point, a new FP method that differentiates strong and iv weak hydrogen bonding compounds was developed. This was done because of the differences in entropy of vaporization for hydrogen bonding compounds. The training set consisted of 191 diverse C, H, N, O compounds ranging from 100 to 4000 g/mol in molecular weight. The test set consisted of 97 compounds of similar diversity. Both data sets produced an AAE of 5K and maximum deviation of 17.5K. It was also found that no substantial decomposition was found for these compounds at flash point conditions. These compounds appear to follow the same physical trends as lower molecular weight compounds. With this new method it is possible to critically review this class of chemicals as well as update NBP and other physical property data. / PDF file replaced 4-20-2012 at request of Thesis Office.

In this thesis, the PC base and DSP base speech keyword retrieval and recognition systems could work. The keywords and describing sentences will not have the limit of word length and could be any languages. Besides, training speech models is not needed anymore. It means that the database gets its expansibility without training speech models again. We can establish the system on the PC base, and calculate the program with fixed-point DSP board. In the processing of speech signal, lots of mathematical functions will be required. We must reach its immediately effect, so that the system could be useful. In addition, compared with floating point, the fixed point DSP cost much less; it makes the system nearer to users. After being tested by experiments, the speech keyword retrieval and recognition system got great recognition and efficiency.

Open site measurement on the electromagnetic interference is the most direct and universally accepted standard approach for measuring radiated emissions from an equipment or the radiation susceptibility of a component or equipment. In general, if the NSA measurements we recorded at different frequencies do not exceed the ideal value +-4dB, we would regard this site as a normalized site, otherwise it is not a normalized site as long as there is one measurement exceeds the range. A one change point model had been used to fit observed measurements. For each set of observations as well as the corresponding ideal values, we have the estimated regression parameter for a one change point model. Our ideal is using the difference of regression parameters between ideal values and observations to assess whether a site is qualified for measuring EMI or not. The assessment tool for whether the testing site is normalized or not is referred to the confidence region for the regression model parameters. Finally, according to the data collected in this experiment, the estimated parameters obtained from the observations will be used to do further statistical analyses and comparing the qualities of the four different testing sites.

Please read the PDF file of my thesis.

none

Software size estimation has been long a challenging task over a software development process. This paper presents an approach that uses the function point analysis to estimate program coding and testing effort in a MIS department, which maintains an ERP system with low employee transfer rate. The method first analyzes the historical data using regression analysis, and then builds a software estimation model with elaborated coefficients for related parameters. The estimation model is tested with the remaining set of historical data to evaluate its predict accuracy. It is shown that the size estimation model can be as accurate as about 90% correctness. Thus it is useful not only in company-wide information resource allocation, but also in performance evaluation of software engineers.

Many oceanographers have postulated that internal wave form inversion would take place at the turning point where the thickness of the upper and bottom layer are equal in a stratified two-layer fluid system. This implies that an internal wave of depression may convert into elevation as the wave propagates over a continental margin comprising continental slope and shelf. Laboratory experiments were conducted on the propagation of a depression ISW over a trapezoidal obstacle in a stratified two-layer fresh/brine water system in a steel framed wave tank of 12m long with cross section of 0.7m high by 0.5m wide. The relative difference in water depth between the upper and lower layer and the initial ISW amplitude were the main controlling parameters, among others. The water depth in the stratified two-layer system on the horizontal plateau of the trapezoid obstacle fell into one of the following case: (1) the upper layer larger than lower (H1>¢Ö2'); or (2) equal depth in the upper and lower layer (H1=¢Ö2'); or (3) the upper layer less than lower layer (H1<¢Ö2'). In addition of the depth ratio, the difference in the length of the horizontal plateau and the thickness of the phycnocline above if were also parameters affecting the outcome of the experiments. In these experiments, three different type of the height and length of the trapezoidal obstacle were used, including long (4.8x0.37m), medium (1x0.35m) and short (0.5x0.35m) types. A full account on the characteristics of the ISW evolution observed during this experimental study is presented in this thesis. As an ISW propagated on the fronting slope, were run-down, vortex motion, internal hydraulic jump (IHJ) and run-up were occurred. Once the wave passed the turning point (where the depth of upper and lower layer equal), the wave form became elevation on the plateau above the obstacle. Based on the laboratory data available, the effect on internal wave evolution can be evaluated by the relative fluid thickness (H1/¢Ö2') on the plateau. The outcome can be classified into three categories: (1) H1>¢Ö2', the relative layer thickness on the plateau unfits for depression ISW propagation and waveform behaves like elevation type; (2) H1=¢Ö2', wave boluses containing mixed fluid propagating on the plateau after breaking on the slope; (3) H1<¢Ö2', ISW propagated over trapezoidal obstacle subjected to shoaling and viscosity effect, without change in waveform. As a depression ISW propagated over the variable length of the plateau, another important factor affecting the intensity of the internal hydraulic jump was the water volume drawn from the plateau. In the case of long horizontal plateau, the interaction range was large, and the IHJ was strong. Consequently, the thickness of the increased which caused the IHJ to move upward along the fronting slope. However, the amplitude and phase speed of the resulting internal wave decreased as if propagated further.

This paper adopts a dual-point measurement technique to predict the resonance frequency and operation modes of a high speed rotor at different operating speed. The effect of balance policy on the operating mode shapes of a rotor has also been investigated. Displacement signals measured at two different points of a rotating shaft are used to estimate the operation mode shapes at the main operating frequency based on the definition of transmissibility function. To verify the feasibility of this technique, the dynamic parameters, i.e. natural frequencies, damping ratio and normal modes of a stationary rotor are measured in advance by applying the traditional frequency response method for comparison. The values of modal assurance criterion (MAC) and mode shape diagrams indicate that both methods are in a good agreement on the normal modes when the system is undamped or lightly damped. However, a significant difference is observed for highly damped modes. The dual-point measurement technique is applicable to investigate the effect of rotating speed on measured operation modes at different speed. Results indicate that the rotating speed may affect the dynamic parameters of a rotating shaft significantly. Furthermore, the dynamic unbalance effect on the response of a rotating shaft has also been studied in this thesis. Results indicate that the position of the unbalance is quite sensitive to the response of a rotating shaft. The sensitivity of unbalance and its position is dependent on the operation mode correspondence to the driving speed. The measured results indicate the dual-point measurement is quite available to study the dynamic responses of a rotating shaft or rotor.

<p> </p><p>The aim of this thesis is to design a web base system which provides functionality of comparison between two java files on the basis of point-to information (P2I). User will upload Java files and analysis of Java files called point-to analysis (P2A). System will store the files in the file system for reference and download in later time. System will extract the information called P2I from P2A and it will store that information in the database.</p><p>Database should be flexible to accommodate the changes in P2A file and system should be able to extract the P2I and store it in database with minimum support of system administrator.</p><p> </p>