Thesis (D. Min.)--Gordon-Conwell Theological Seminary, South Hamilton, Mass., 2001. / Abstract and vita. Includes bibliographical references (leaves 113-115).

In recent decade growing concerns of CO2 emissions from power plants have increased, which led to development of technologies like oxy-fuel combustion process. P91 steel is profoundly used in power plants, but oxy fuel combustion exacerbates corrosion due to recycling of flue gas. This paper studied the kinetics of the corrosion rate on the boiler tubes and furnace and help achieve a corrosion resistant coating over it. Refractory metal diffusion coating is created and tested at high temperature in corrosive atmosphere. This was done by forming Ti and Zr diffusion coating on P91 steel using pack cementation. Coating thickness of 12 and 20 µm were obtained for Ti and Zr respectively. These samples were tested in thermo-gravimetric system by heating at 950˚C for 24 hours in 5% oxygen in Helium gas. Heating in an oxidizing environment lead to exfoliation corrosion on uncoated P91 steel. TGA procedure confirmed less mass change of Ti and Zr coated samples, than that of uncoated P91 steel sample. SEM and depth profiling confirms oxygen penetration is 2.7mm in uncoated P91 steel sample, whereas the Ti and Zr Coated samples oxygen penetration is just 16 and 56 µm respectively.

The thesis utilized physical testing and computer modeling to determine the feasibility of identifying a change to the mass or stiffness of a steel frame. Physical testing was performed using an accelerometer, linear shaker, and arbitrary function generator. Two methods of laboratory testing were developed: ambient vibration testing (AVT) and forced vibration testing (FVT). AVT was able to preliminarily identify the natural frequencies and mode shapes of the frame. FVT was able to precisely identify four distinct natural frequencies, mode shapes, and damping ratios. The baseline frame then underwent two physical changes: the addition of mass to its roof, and the addition of braces along one of its sides. FVT was used again to determine the natural frequencies, mode shapes, and damping ratios of the newly changed structure. An ETABS computer model was developed to represent the frame. This baseline model produced natural frequencies and mode shapes that closely matched the values determined by FVT. The mass and stiffness of this baseline model were then changed multiple times through the addition of mass and braces at various locations on the model. The frequencies and mode shapes were recorded for each change. Two methods were developed to identify the changes to the steel frame. The first method was able to determine which one of the models best represented a single change to the structure (adding mass to its roof). The second method was able to determine the combination of models that best represented the two concurrent changes to the structure (adding mass to its roof and braces to its sides). Both methods utilized the percent differences of each altered computer model relative to the original, and each method satisfactorily identified its respective physical alteration.