Thesis (Ph. D.)--Cornell University, 1922. / "Reprinted from Physical review, vol. XX, no. 2, August, 1922."

Thesis (M.S.)--University of Wisconsin--Madison, 1956. / Typescript. eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references (leaves 52-53).

Thesis (Ph. D.)--University of Texas at Austin, 2004. / Supervisor: L.F. Demkowicz. Vita. Includes bibliographical references.

Thesis (Ph. D.)--Drexel University, 2006. / Includes abstract and vita. Includes bibliographical references (leaves 68-70).

Thesis (PH.D.) -- Syracuse University, 2006 / "Publication number AAT 3251813."

- / -

A finite difference algorithm for solving the forward modelling problem of geo-electromagnetic induction in three-dimensional structures has been developed in this thesis. Novel features of the method include the incorporation of a thin sheet of anomalous conductance at the surface of an otherwise quite general three-dimensional structure in which the anomalous region is allowed to approach two-dimensional configurations at infinity; the use of magnetic rather than the electric field components for obtaining the solution; the use of integral boundary conditions at the top and bottom of the model; and the application of new cell-integral finite difference equations to the main body of the model. The algorithm has been tested for synthetic models against results delivered by existing two and three dimensional modelling programs which are already well established. The results are found to be very satisfactory. Applications of the algorithm have been shown for two cases. First, the dependence of the induction vectors on the period ranging from 10 to 10000 s has been studied for a model with two perpendicular lateral conductivity contrasts; the directions of induction vectors vary from site to site reflecting the combined effect of the two perpendicular contrasts. In the second case, the distortion effect due to small surface inhomogeneities over a buried 2D anomaly was studied using induction vectors and difference vectors. There is evidence of mutual coupling in a certain region which invalidates a simple subtraction of the vectors to reveal the form of the buried anomaly, but elsewhere the procedure appears to be quite valid. Since surface anomalies can be simulated by an anomalous thin sheet over the general 3D structure, it is suggested that this algorithm could be very useful for testing the validity of existing schemes for impedance tensor decompositions used in MT studies when surface anomalies are thought to be distorting the real data. / Graduate

Mold behavior, mold-related quality and the columnar-to-equiaxed transition influenced by in-mold EMS were examined by performing trials at two steel companies, metallurgical examination of the billet samples and mathematical modelling. The thermal fields in the walls of billet molds (102 X 102mm and 127 X 178mm) and in the cooling water were monitored by a system of thermocouples as the EMS was switched on and off during the continuous casting of several heats. The effect of electro-magnetic stirring on mold heat extraction was found to be negligible. The mold temperatures and cooling water temperatures are strongly dependent on the mold/billet gap which is affected by dynamic distortion of the mold tube. In the case of the square mold, the time-dependent mold distortion resulted from boiling adjacent to the cold face due to low water velocity and poor water quality. In the rectangular mold, differential expansion of the wide and narrow faces of the mold led to periodic wall movement at the midface causing cycling in the mold and water temperature. Both effects completely dominated any potential influence of EMS on mold heat extraction. Cooling water velocities measured in separate experiments and the mold temperature profiles were input to a two-dimensional heat-flow model to establish mold heat-flux profiles. A steep taper of 2.6 %/m in the upper regions of the mold increased heat extraction compared to previously published heat-flux data in 0.8 %/m tapered-molds. However, due to the periodic wall movement in the rectangular mold, the heat flux declines to lower values periodically. The calculated heat flux profiles were employed in a one-dimensional transient heat flow model to predict superheat removal from the liquid pool under a variety of assumed fluid flow conditions. The major heat flow effect of EMS was inferred to be one of increasing the convective heat flow at the solidification front leading to earlier superheat extraction from the liquid steel. Solidification structures in billet samples collected during the trials were examined. The columnar-to-equiaxed transition in continuous casting takes place provided all the superheat is removed from the melt and there is a sufficient density of nuclei present in the pool. At superheats of <20° C in the tundish, high heat extraction in the mold and remelting of the mold generated nuclei facilitate the removal of the superheat well within the mold and the columnar-equiaxed transition is triggered after 10-15 mm of shell growth on both the inside and outside radius faces. At higher superheat in the tundish, the liquid pool leaves the mold with residual superheat which takes longer to remove because of the declining fluid flow. Even though all the superheat is removed lower in the machine, the columnar-equiaxed transition occurs only if dendrite debris generated in the vicinity of the mold has survived in their descent through the superheated liquid. The effect of carbon on the columnar-to-equiaxed transition appears to stem from its influence on facilitating dendrite arm remelting and the survival of the dendrite fragments till the pool reaches sub-liquidus temperature. EMS extracts more superheat by maintaining a steep temperature gradient in the thermal boundary ahead of the solidification front and achieves an earlier columnar-equiaxed transition. Electro-magnetic stirring appears not to affect either the average depth or the variation of depth of oscillation marks across a given face. However, the electro-magnetically driven flow dominates the turbulance at the meniscus due to the input stream and stabilizes a meniscus shape with the result that the oscillation marks are also of a well-defined shape unlike the unstirred billets. No influence of EMS was found on the formation of "hooks" or the fine equiaxed crystal zone near the surface. The influence of EMS on inclining the growing dendrites appears to not come into effect until about 1 mm of shell has formed. It appears that the existence of the momentum boundary layer where the velocity of the rotating steel falls to zero at the surface is the reason for the absence of the influence of EMS on the subsurface solidification. Rhomboidity and off-corner crack formation were found to depend, as reported by previous researchers, on mold distortion and its dynamic nature. The absence of any effect of EMS on these defects is due to its lack of effect on mold heat transfer and thus mold distortion. / Applied Science, Faculty of / Materials Engineering, Department of / Graduate

Various heterospin oligoradicals were prepared. The intramolecular spin coupling of these radicals was studied by ESR spectroscopy. Oligonitroxides and oligophenoxys were synthesized and investigated as potential molecular magnetic materials. Heterospin biradical 3,5-di-tert-butyl-3′ -(N-tert-butyl-N-aminoxy)-4-oxybiphenyl was synthesized by Suzuki reaction. It showed reversible temperature dependent behaviors in both ESR and UV-vis spectra. Heterospin radicals 3-( N-tert-butyl-N-aminoxy)-nitrenobenzene and 4-(N-tert-butyl-N-aminoxy)-nitrenobenzene were prepared by photolysis of the corresponding azido precursors. The relationship between the intramolecular exchange coupling and the geometry of the molecular structures was studied by ESR spectroscopy. Trinitroxide 2,4,6-tris(4-N-tert-butyl- N-aminoxylphenyl)-s-trazine was synthesized by either Suzuki reaction or direct coupling of aromatic anions with cyanuric chloride. This triradical has a quartet ground state and is stable in air. The crystalline trinitroxide was studied by magnetic analysis and showed a intramolecular ferromagnetic exchange with J/kB = 9.0 K and a intermolecular antiferromagnetic exchange with J/kB = −1.45 K. Oligophenoxys 2,3,5-tris(3,5-di-tert-butyl-4-oxyphenyl)- s-triazine and 2,3-bis(3,5-di-tert-butyl-4-oxyphenyl)-5-( N,N-dimethylamino)-s-triazine were prepared to investigate the properties of s-triazine as an exchange linker. Tetraradical tetrakis(4-N-tert-butyl- N-aminoxylphenyl)silane was designed and synthesized as a potential organic magnet. This tetraradical has a tetrahedral geometry with a silicon atom at its center. Its solution shows a clean nonet in ESR spectrum at room temperature. Its crystal structure was also studied. Black precipitates were observed immediately after it was added in a solution of Mn(II)(hfac) 2 in heptane.

Bis(aryloxy)phosphine azides (BAPAs) were prepared from corresponding phenols. The synthesis of BAPAs was limited to sterically hindered aryloxy moieties. Photolysis of the BAPAs under a variety of conditions yielded phenoxyl radicals. During photolysis of BAPAs in the solid state at cryogenic temperatures, triplet state radical pairs were observed by EPR spectroscopy. Studies were carried out to determine the identity of these radical pairs. In most cases the radical pairs obtained consisted of the corresponding aryloxy groups in the BAPA being studied. Co-crystallization of bis(2,6-di-t-butyl-4-methoxyphenoxyl)phosphine azide with 2,6-di-t-butyl-4-methoxyphenol resulted in solid state reactions between the 2,6-di-t-butyl-4-methoxyphenoxyl radicals generated upon photolysis and the host matrix at cryogenic temperatures. This resulted in new radical pairs whose identity was determined by EPR spectroscopy. Annealing photolyzed samples of BAPAs caused only slight changes in the zero field splitting parameters (zfs). Photolysis of BAPAs in potassium chloride matrices was developed as a new method for the analysis of radical pairs in the solid state. Curie law studies on the radical pair formed between two 2,6-di-t-butyl-4-methoxyphenoxyl radicals indicates that the radical pair may be a ground state triplet. Three conformationally inflexible dinitrenes, 1,8-dinitrenoanthracene, 2,6-dinitrenodibenzofuran and 2,6-dinitrenobiphenylene were prepared as tests for the spin dipolar model. The spin dipolar model predicts that the zfs parameters in a quintet state molecule comprised of two triplet subunits will depend on the vector interaction angle between the triplet subunits. In order to test the spin dipolar model unambiguously, conformationally rigid dinitrenes with unusual vector interaction angles were prepared. In each case, a quintet state spectrum was observed which was well simulated using the eigenfield EPR simulation method. The zfs parameters extracted from the EPR spectra closely matched those predicted by the spin dipolar model. These results appear to prove that the spin dipolar model may be quantitatively as well as qualitatively correct. Curie law analyses of the dinitrenes indicated that all three were ground state quintets.