Made available in DSpace on 2016-04-25T16:45:24Z (GMT). No. of bitstreams: 1 Ricardo Jose da Silva Fontes.pdf: 470636 bytes, checksum: 7f23f0e0e4e6d7eb793ec362590a7fa4 (MD5) Previous issue date: 2009-12-14 / The objective of this Dissertation is to analyze the beta parameter and the Equity Cost of Capital of a company s portfolio from different segments of the Brazilian Economy, calculating its beta and its equity cost of capital. The covered period of analysis is from January 2000 to may 2009, with comparative intervals in December 2002, dec 2005 and dec 2008. The analysis was based on CAPM (Capital Assets Pricing Model) and was carried out using data from historical basis of stock price, Market Index like Ibovespa and S&P 500, Risk Free rate and inflation index, from several secondary market information sources available in the market. The analysis showed that the beta parameter and the equity cost of capital can be found having its beta computed through the regression of local stocks and the Ibovespa index. When compared the beta parameter and its stock return through the period of time above mentioned we will find correlation statistically significant, especially before 2008 / O objetivo deste trabalho é analisar o parâmetro beta e o custo de capital dos acionistas de uma carteira de empresas de diferentes segmentos da economia brasileira, calculando seus betas e o custo de capital dos acionistas. O período de análise esta compreendido entre jan/2000 e maio/2009, com intervalos comparativos entre 2002, 2005 e 2008. A fundamentação teórica foi baseada na Teoria de Precificação de Ativos de Capital (CAPM) e as análises foram executadas com base em dados históricos como preço da ação, preço dos índices de mercado como Ibovespa e S&P500, retorno da taxa livre de risco e expectativa de inflação, obtidos em diferentes fontes de informação mercadológica disponíveis no mercado. Os estudos realizados mostraram que o cálculo do beta e do custo de capital dos diferentes segmentos pode ser feito com base no beta local e não apresenta autocorrelação. Observou-se ainda, que quando comparados individualmente o beta das empresas e o retorno oferecido pelas ações ao longo do intervalo de tempo supracitado, apresenta-se uma correlação em valores estatisticamente relevantes, principalmente nos períodos que antecedem a crise de liquidez ocorrida no 2º semestre de 2008

The ability to predict fatigue crack growth rates is essential in safety critical systems. The discovery of fatigue crack closure in 1970 caused a flourish of research in attempts to simulate this behaviour, which crucially affects crack growth rates. Historically, crack tip plasticity models have been based on one-dimensional rays of plasticity emanating from the crack tip, either co-linear with the crack (for the case of plane stress), or at a chosen angle in the plane of analysis (for plane strain). In this thesis, one such model for plane stress, developed to predict fatigue crack closure, has been refined. It is applied to a study of the relationship between the apparent stress intensity range (easily calculated using linear elastic fracture mechanics), and the true stress intensity range, which includes the effects of plasticity induced fatigue crack closure. Results are presented for all load cases for a finite crack in an infinite plane, and a method is demonstrated which allows the calculation of the true stress intensity range for a growing crack, based only on the apparent stress intensity range for a static crack. Although the yield criterion is satisfied along the plastic ray, these one-dimensional plasticity models violate the yield criterion in the area immediately surrounding the plasticity ray. An area plasticity model is therefore required in order to model the plasticity more accurately. This thesis develops such a model by distributing dislocations over an area. Use of the model reveals that current methods for incremental plasticity algorithms using distributed dislocations produce an over-constrained system, due to misleading assumptions concerning the normality condition. A method is presented which allows the system an extra degree of freedom; this requires the introduction of a parameter, derived using the Prandtl-Reuss flow rule, which relates the magnitude of slip on complementary shear planes. The method is applied to two problems, confirming its validity.