A competing risks situation arises when a unit can fail due to several distinct failure types. In a competing risk situation, standard techniques from survival analysis may lead to incorrect and biased results. In this thesis, the theory of competing risks is used to identify possible risk factors for breast, uterine and ovarian cancer. This has been done by regression on the cause specific hazard functions, the subdistribution hazard functions and two approximate methods. Cox regression is used for a complete analysis of the medical data.By following 61457 women over approximately 50 years, it has been observed 3407 cases of breast cancer, 934 of uterine cancer and 843 of ovarian cancer. Summarized, it has been found that several births decrease the risk of breast, uterine and ovarian cancer. Obesity is associated with increasing risk of ovarian cancer for postmenopausal women, but not premenopausal. A long reproductive period (early menarche and/or late menopause) and high BMI increases the risk of breast and uterine cancer. Late first and last birth decreases the risk of uterine cancer, while it increases the risk of breast cancer. The data used in the analysis is selected from a screening program organized by the Norwegian Cancer Society for early diagnosis of breast cancer. postmenopausale women, but not premenopausale. A long reproductive period (early menarche and/or late menopause) and high BMI increases the risk of breast and uterine cancer. Late first and last birth decreases the risk of uterine cancer, while it increases the risk of breast cancer. The data used in the analysis is selected from a screening program organized by the Norwegian Cancer Society for early diagnosis of breast cancer.

This thesis is concerned with 4D AVA seismic inversion problems. By comparing two seismic surveys done over the same area, but at different times, one hopes to discover untapped pockets of oil or gas. Using the full likelihood to analyse 4D AVA seismic data is impossible in practice due to memory and computational restrictions. The goal of the thesis is to find a useful framework for parameter estimation and predictions for 4D AVA seismic data, and the composite likelihood is introduced as a possible solution. The composite likelihood method takes in pairs of data points and sums over them instead of taking in all the data as is the case for the full likelihood. This makes calculations fast while avoiding matrix operations on large matrices.The composite likelihood method is tested on a data set from the Norne field for parameter estimations and predictions. Eight variations of the model are tested, the variations being the exponential or Matern correlation function, one or two data columns used as a data point in the composite likelihood, and a simple or wavelet convoluted noise term. The composite likelihood method is shown to perform well; it is fast and the estimates found agree well with previous experience. Comparison of the different models indicate that the choice of correlation function has little effect on the results, that the noise term should be kept simple, and that it is sufficient to use one data column.

This thesis describes a new instrumental approach designed to measure the complex Young’s modulus on cylindrical samples 1 inch in diameter 2 inch long and typically in the range of 1-70 GPa. Excitation frequencies are from 11 Hz to 167 Hz. The setup is based on the Forced Deformation Method (Batzle et al., 2006) and is capable of measuring phase and magnitude of the dynamic stress and strain. An actuator provides an oscillating force from one end of the plug. Strain is measured on the cylindrical side with three strain gages evenly spaced around the circumference. The lowest recordable magnitude of strain is in the order of 10E-8. Force is measured by a piezoelectric transducer. Values are < 10 N. Plug sample measurements of the Young’s modulus values for Berea sandstone, Castlegate sandstone, Pierre shale, PEEK, and aluminium alloys (ALU-7075 and ALU-6061) are included. Reference material results are in accordance with published values. Some of the instrumentation needed is also given in detail in the previous work (Fintland, 2010).

The effect of clustering during natural aging of three different Al-Mg-Sialloys with equal Mg/Si-ratios, but different Mg+Si contents have beenstudied through hardness measurements and transmission electron mi-croscopy (TEM). Hardness measurements were performed both duringnatural aging (NA) and after various NA times followed by subsequentartificial aging for all alloys, while TEM-investigations were performedon selected samples. The results from the experiments were comparedand the existence of three different clustering processes were discovered.The first process was established to be Si-Si clustering causing a slighthardness increase during natural aging and a strong hardness increaseafter subsequent artificial aging. The second process was found to beMg-Mg clustering causing a significant hardness increase during naturalaging, and a clear decrease in hardness after subsequent artificial aging.The third process was found to be Mg-Si clustering and simultaneoustransformation of mono-clusters into co-clusters. This combined processcauses a hardness increase both during natural aging and after subse-quent artificial aging. The hardness increase is caused by a refinementof alloy microstructure including a higher number of smaller ” needles.This precipitate type was observed for all alloys and conditions. Thenegative effect of Mg-Mg clusters on precipitation hardness is suggestedto be related to the repulsion of Si from these clusters. This hinders theco-clustering of Mg and Si in the areas surrounding these clusters.

Ozone and temperature data from Troll and Rothera research stations has been analyzed in order to trace planetary waves. Seasonal variations were removed and spectral analysis was performed in order to find dominating frequencies. Possible planetary wave oscillations were found in the 4, 5, 8.5, 10-11, 11-13, 14-15, 17-20, 21-24 and 26-45 day bands, some corresponding to the atmospheric normal modes. These were found consistently in both the ozone and temperature data, indicating that the millimeter wave spectroscopy used to measure the ozone is a viable technique for tracing planetary waves in a region of the atmosphere where other measurements are sparse.

Computational fluid dynamics (CFD) was be used to investigate aspects of interaction of flow over two airfoils in relatie motion in detail.%Problem statement2D tandem airfoil setups were studied, where the leading airfoil was performing an oscillating motion in the vertical direction while the trailing airfoil was kept stationary. The NACA 0012 and the S809 airfoils were considered.%ApproachAnsys Fluent v13.0 was used as the CFD solver, and Gambit v2.4.6 was employed for grid generation. All simulations were transient at a Reynolds number of either 2*10^4 (laminar flow) or 3*10^6 (turbulent flow). The Transition SST turbulence model was chosen to model turbulence, and Fluent's sliding grid technique was used to achieve the relative motion between the airfoils.%ResultsThe tandem setup was found able to outperform a single airfoil for similar conditions. The presence of a trailing airfoil did not significantly affect the leading airfoil's performance, whereas it did affect the wake structures significanlty. The suction peak near the nose of the airfoil was found to be the most important factor determining the airfoil's propulsive efficiency. Therefore, leading edge vortex (LEV) shedding was found to be of higher importance than trailing edge vortex (TEV) shedding when airfoil performance was concidered.The asymmetric S809 airfoil provided similar results as the symmetric NACA 0012 airfoil. However, the NACA 0012 airfoil achieved slightly higher propulsive efficiencies for the cases investigated, indicating that a symmetric airfoil is desired for flapping airfoil setups. For the tandem setup the highest propulsive efficienies were 0.766 and 0.742 for the NACA 0012 airfoil and the S809 airfoil, respectively. Both peaks were found at k = 0.3 and h = 0.6 (Sr = 0.11) for the leading airfoil. A maximum thrust coefficient of 2.32 was found for the tandem S809 airfoil setup at k = 1.5 and h = 0.5 (Sr = 0.48).The Strouhal number was found to be an important describing parameter, but additional information about the reduced frequency or the oscillating amplitude was needed in order to fully describe the setup.

In this thesis the strain properties of two InAs/GaAs quantum dot intermediate band solar cell materials have been explored. Both samples were thin films grown on a (100) GaAs substrate. The quantum dot material was InAs, and the bulk material was GaAs. One sample had AlAs-cap, while the other had GaAs-cap. Geometrical phase analysis was used to study the strain. A higher degree of strain was found in the AlAs-capped sample. Negative strain was observed in directly above and below the quantum dots in both samples. A stacking fault in a quantum dot in the AlAs-capped sample was found to relax all the strain. Analysis of the chemical composition of the AlAs-capped sample was performed using HAADF-STEM and multislice analysis. This analysis found an average indium concentration inside the quantum dots of 25% +- 10%, with peaks up to 50%.

Flexible data regression is an important tool for capturing complicated trends in data. One approach is penalized smoothing splines, where there are several mainstream methods. A weakness is, however, the quantification of uncertainty. We will in thesis present two mainstream smoothing spline methods, P-splines and O'Sullivan splines, and the RW2 model; a Bayesian hierarchical model based on a latent field. The Bayesian priors are specified by a stochastic Poisson equation, and spline estimates are approximated along a finite element Galerkin approach. We evaluate the three methods using integrated nested Laplace approximations (INLA) for a full Bayesian analysis supplying credible bands. The methods give fairly similar results and we investigate the theoretical motivates behind the methods. As an extension of the Bayesian models, the smoothing parameter is incorporated in latent field. This gives an adaptive smoothing method, which better estimates jumps and quick curvature changes. Further, the close relationship between O'Sullivan splines smoothing splines is discussed, revealing O'Sullivan splines to be a finite element Petrov-Galerkin approximation of smoothing splines. The main results are the possibility of credible bands, the extension to adaptive smoothing and the finite element understanding of O'Sullivan splines.

In recent years, stochastic partial differential equations (SPDEs) have been shown to provide a usefulway of specifying some classes of Gaussian random fields. The use of an SPDEallows for the construction of a Gaussian Markov random field (GMRF) approximation, which has verygood computational properties, of the solution.In this thesis this kind of construction is considered for a specificspatial SPDE with non-constant coefficients, a form of diffusion equation driven by Gaussian white noise. The GMRF approximation is derived from the SPDE by a finite volume method. The diffusion matrixin the SPDE provides a way of controlling the covariancestructure of the resulting GMRF.By using different diffusion matrices, itis possible to construct simple homogeneous isotropic and anisotropic fields and more interesting inhomogeneous fields. Moreover, it is possible to introduce random parametersin the coefficients of the SPDE and consider the parametersto be part of a hierarchical model. In this way onecan devise a Bayesian inference scheme for theestimation of the parameters. In this thesis twodifferent parametrizations of the diffusion matrixand corresponding parameter estimations are considered.The results show that the use of an SPDE with non-constant coefficients provides a useful way of creating inhomogeneousspatial GMRFs.

The performance of steady-state tower shadow models for a wind turbine truss tower have been evaluated. The Reynolds-Averaged-Navier-Stokes (RANS) approach, in conjunction with the $k-omega$ Shear-Stress-Transport (SST) model, was used to simulate transient flows past cross sections of a truss tower. The objective was to compare numerical results with Powles', Blevins' and Schlichting's tower shadow models and evaluate their performance on a multimember structure. Parameters for each model have been estimated. It will be shown that the RANS model was able to reproduce realistic results when used in transient simulations on high Reynolds number flows (supercritical regime). The importance of considering unsteady motion when calculating the turbulence intensity, using RANS with transient simulations, will be explained. The multimember extension used for the tower shadow models reproduces the mean velocity profiles quite well, and by using a suitable estimation method, global parameters were found for all models. Additionally, turbulent inflow has been implemented with a user-defined function in Fluent. The results have been evaluated, and show that such such sophisticated inflow modeling is not necessary to predict realistic mean velocity profiles.