Ostergaard, Erik Barr2013-03-222013-03-222013-03-21vt_gsexam:431http://hdl.handle.net/10919/19294Digital Image Correlation is used to generate high-spatial-density full-field displacement<br />and strain data of a connection box outer diameter for use in the calibration of finite element<br />make-up models. Image acquisition and data processing techniques are discussed and best<br />practice recommendations are made. 3D-wedge models consisting of a twenty-degree sweep of<br />the connection geometry are generated from manufacturer supplied profiles. Deformation<br />plasticity material models are developed from identified minimum strength material coupons.<br />Axisymmetric and 3D meshing schemes are used to capture the geometric complexity, supply<br />enough resolution to represent seal performance, and provide a solution in an acceptable<br />timeframe. Several techniques for achieving good contact resolution are presented. The<br />mechanics of the full 3D connection makeup are decomposed into simple idealized<br />representations. Finite element boundary conditions are developed to adequately represent the<br />360-degree make-up mechanics in a wedge section. The wedge model is loaded to achieve a<br />torque-rotation coupling which satisfies the experimental make-up conditions. This model<br />displays a much improved ability to capture box outer diameter strain and displacement fields,<br />and thus better represents the mechanics of a connection make-up. A 3D inspired axisymmetric<br />pretension loading scheme is developed which enables the 3D-wedge seal conditions to be<br />replicated in a computationally efficient axisymmetric form for connection performance<br />evaluation. Seal metrics are developed and converged to evaluate connection sealing capabilities<br />in the power-tight configuration. Modeling error metrics are developed, and the final 3D-wedge<br />model is evaluated relative to the experimental DIC data.ETDIn CopyrightOil Country Tubular GoodsConnection Make-upFinite element methodDigital Image CorrelationThesis