<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:dc="http://purl.org/dc/elements/1.1/"><rdf:Description rdf:about="https://repozitorij.uni-lj.si/IzpisGradiva.php?id=113218"><dc:title>Random thermal loads in the thermal fatigue assessment of nuclear piping</dc:title><dc:creator>Costa Garrido,	Oriol	(Avtor)
	</dc:creator><dc:creator>Cizelj,	Leon	(Mentor)
	</dc:creator><dc:creator>Tiselj,	Iztok	(Komentor)
	</dc:creator><dc:subject>material science</dc:subject><dc:subject>thermal fatigue</dc:subject><dc:subject>structural damage</dc:subject><dc:subject>nuclear safety</dc:subject><dc:subject>spectral methods</dc:subject><dc:description>Thermal fatigue is a structural damage mechanism which has been among the causes of reactor coolant leakages in nuclear power plants. A well-recognized source of thermal fatigue in the safety related nuclear piping is the turbulent mixing of fluids at different temperatures.
The thermal fatigue assessment of components requires an interdisciplinary approach. This involves the evaluation of the fatigue driving loads from the fluid temperatures, heat transfer to the pipe and stress amplitudes within the pipe wall. The fatigue analyses are then needed to predict the fatigue life using the estimated driving loads and experimental fatigue resistance data. It is clear that the accuracy of fatigue life estimates strongly depends on the accuracy of the estimated fatigue driving loads, especially fluid temperature histories. These can be obtained from complex and extremely time-consuming experiments or computational fluid dynamic (CFD) simulations of turbulently mixing fluids. These could only offer spatially incomplete temperature histories, which are typically limited to minutes or hours at the best, and reported in compact terms of the first two statistical moments combined with power spectral densities.
Extrapolation of compact and comparatively short fluid temperature histories to the expected fatigue life time of months or years is inherently uncertain and lacks deeper understanding and proper treatment in the open literature. The major goal of the dissertation is therefore to contribute to the prediction and understanding of the uncertainties involved in the assessment of the safe life of pipes exposed to thermal fatigue due to the turbulently mixing fluids.
To this end, the dissertation explores a range of simplified and complex state-of-the-art numerical methods, including one-dimensional (1D) approximations with one- and multi-frequency (spectral) approaches and turbulent three-dimensional (3D) CFD models with and without conjugate heat transfer. An improved spectral loading approach for the generation of synthetic fluid temperatures at affordable computational costs is proposed. Synthetic fields of temperatures acting near the surface of pipes during turbulent fluid mixing are employed in 3D structural analyses of pipes to verify the persistence of nearly equi-biaxial stress states in the pipe surface away from geometrical, structural and material discontinuities. Then, the improved spectral approach is employed to generate a wide variety of fluid temperature histories resulting in a set of estimated fatigue lives. These are further analyzed to quantify the magnitudes of inherent uncertainties. Linear 1D heat diffusion and thermal stress estimates are used together with fatigue assessment codified rules in the uncertainty analysis.</dc:description><dc:publisher>[O. C. Garrido]</dc:publisher><dc:date>2016</dc:date><dc:date>2019-12-13 12:37:20</dc:date><dc:type>Doktorsko delo/naloga</dc:type><dc:identifier>113218</dc:identifier><dc:language>sl</dc:language></rdf:Description></rdf:RDF>
