Exploration for oil and gas is moving towards deeper waters. Steel has been the most common riser material. Related to deep water concepts titanium has become an alternative to steel for these applications. Several codes exist today for predicting collapse loads for marine pipes. However, the capacity formulas are developed for steel. If the formulas are applied directly to titanium several parameter uncertainties will be unknown. Ideally, extensive model testing of titanium pipes is required. This thesis discusses and investigates utilisation of experimental material test data and a supplementary numerical approach based on finite element analysis. The relationship between material model parameters as input to the analysis and the collapse capacity is investigated by performing a series of nonlinear FEM analyses. Statistical models for the input material model parameters are established based on tests on small specimens cut from titanium pipes. These models are subsequently combined with results from the FEM analyses by application of response surface methods. As output from the analysis, the probability distributions of the pipe capacity with respect to local buckling/collapse are obtained. Finally, the data from the nonlinear finite element analyses are compared to a relevant design code. Suggestions for a possible basis for design formulas to check for the local collapse capacity of deep water titanium risers are provided.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:ntnu-454 |
| Date | January 2000 |
| Creators | Bjørset, Arve |
| Publisher | Norges teknisk-naturvitenskapelige universitet, Fakultet for ingeniørvitenskap og teknologi, Fakultet for ingeniørvitenskap og teknologi |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Doctoral thesis, monograph, info:eu-repo/semantics/doctoralThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | Dr. ingeniøravhandling, 0809-103X ; 1999:100 |
Page generated in 0.0021 seconds