Skip to main content
European Commission logo print header

Measurements on structures in ice (STRICE)

Deliverables

Pressure ridge mechanical material properties are keel and consolidated layer full-scale measurements, data analysis, constitutive model, numerical simulation, and parameter extraction. Full-scale ridge mechanical properties testing were conducted in winters 2001 and 2003. The earlier LOLEIF testing equipment was improved by manufacturing a more rigid hydraulic chain saw blade and drive mechanism, and utilizing modified soil trench cutter to cut slots through the consolidated layer above the ridge keel to be tested. Keel deformation displacement measurement was improved by adding four more potentiometer type displacement transducers for continuous recording. A total of 16 different in-situ full-scale field tests were carried through successfully: two circular plate punch tests, eight beam bending-type punch shear tests, five consolidated layer beam bending tests, and one consolidated layer uplift test. The FEM model for 3-D ridge keel failure simulation has been refined after LOLEIF. Iteration loops in UMAT-procedure constitutive model have been reprogrammed. Convergence problems were solved. A new constitutive model - shear cap model - was developed to observe volumetric strain effects during rubble shearing. To describe different deformation mechanisms, two yield functions are combined together for a smooth and continuous yield surface. The first describes the shearing failure and the second one the rubble compaction (cap part). Two types of hardening laws define the yield surface evolution. The first one defines the cohesive softening due to material distortion. The second one defines the rubble compaction depending on the plastic dilatation. The new 3-dimensional constitutive model was implemented as the UMAT-procedure into the commercial ABAQUS Finite Element Code. Studies for extracting proper volumetric material properties from test data, and for element mesh independence in localization, continues. In-field tests have been simulated in order to find out optimal set for model parameters and their evolution laws. The model calibration was made by comparing the ridge keel punch test numerical simulations with the experimental full-scale tests. The model parameters are adjusted to simulate similar load-displacement response compared to the measured one. The comparison was also made for the deformations in the keel. The new constitutive model for ice rubble describes the measured ridge load events successfully. Mechanical properties of the rubble were solved from the punch shear test analysis numerical simulation comparisons. The bending test simulations give means to solve the bending strength of the consolidated layer as well as the interaction effects with the rubble. Initial computer intensive simulations indicate satisfactory agreement with in-field tests.
Data from former field and laboratory ice load measurements were studied and compared with common ice load algorithms. The data and theories were also compared with the Norstromsgrund field data. The studies confirm that the structure width (D), ice thickness (h), strength (sigma_ c) and the ice speed (v) are the major parameters that determine the ice/structure interaction process and ice load. The analysis of experimental data in STRICE leads to the conclusions that the mean peak pressure decreases with the increase of both aspect ratio (defined as structure width/ ice thickness (D/h)) and contact area. The disadvantage of the majority of experimental and theoretical dependencies is that they consider the pressure dependence on these parameters separately, whereas both of them influence the pressure. Further, this task claims that the dependency of the global ice load on the structure width and ice thickness should be written in a way that contains structure width and ice thickness in some exponents. It was found that the exponent for D is almost independent of aspect ratio. Probably, taking into account the accuracy of the methods of investigation and existence of approximations, this parameter can be considered as a constant. Whereas the exponent for ice thickness changes its sign during the transition from small to large aspect ratios. The influence of ice thickness on average pressure is not significant for structure widths and ice thickness relevant for design of offshore structures. Thus a formula is proposed that only contains the structure width in some exponent. When comparing the historical data with the measurements done at the Lighthouse Norstromgrund during the LOLEIF and STRICE period, we find that the load data from the Lighthouse is 4-5 times lower than the historical data.
Both theoretical and experimental results on the ice-structure contact processes were obtained. Very detailed observations were made on the lighthouse Norstomsgunde concurrently with the measurements. These observations revealed several new features on different ice failure processes. These results concern ice failure by crushing and bending and also the behaviour of ice ridges acting on the structure. The observations were documented in several "Observation"-documents that supplement the main log-book of the full-scale measurements. A POAC'03 paper was also published. A new numerical model was derived on the ice-structure contact phenomena. This model was implemented using the Abaqus FE program. As a new feature in modelling, the unloading phase of the ice-structure interaction process was described using a softening material model. Further new results were obtained during the clustering activities between this project and another EC funded project NEST. Accordingly, the dynamic ice forces acting on conical structures can now be described and used in the design of such structures.
Throughout the project and beyond its operative phase a variety of data and information management activities were conducted in support of the project. These incorporated set-up, maintenance and operation of the project web site, editorial work, general and project specific guidelines for data and information production and processing as well as documentation and document producing work. Data and information management including affiliated activities has increasing relevance in the scientific community as well as in consulting activities for a wide range of our project and clients. Besides gathering further technical, operational and related information technology experience the RTD work in the project provided technical and applied scientific experiences in respect to ice technology, ice processes and ice forces/actions. The results of the project data and information management work support general expertise, references and marketing chances in a variety of consultancy areas. They establish new and strengthen existing links and contacts within the scientific, administrative and industrial communities active in the fields covered by the project's RTD work and objectives.
During its first year the EU-STRICE project started to develop a draft for European Ice Load Design Code. Objectives were announced in POAC’01 conference in Ottawa, and a support group of international specialist was assembled. In the meantime, a multinational ISO subcommittee had been preparing, for over ten years, an international code for offshore structures in oil and gas related industry. It decided at the beginning of 2002 to expand its scope to observe also Arctic Offshore Structures with main emphasize on ice loads. The planned ISO Code 19906 will be above national codes, and, according to Vienna Agreement, the European CEN codes will be harmonized with ISO codes. The schedule for the ISO code will be: Working Draft in October 2004, Draft International Standard March 2007, and Final International Code February 2009. Due to its origin, the ISO 19906 was strictly restricted to Arctic Offshore Structures in oil and gas related industry. Needs in Europeand elsewhere pushed forward by the STRICE ice load code effort put pressure to expand the ISO code applicapability range also to subarctic conditions, and to almost all structures in ice infested waters. A list of contents by the STRICE Draft European Ice Load Code will be used as reference to cover sub-arctic European needs. STRICE activities towards codes provided an assembly of LOLEIF and STRICE full-scale measurement data in addition to other published data to be used as background for the codes. The content for the structure and guidelines of the Draft Code was prepared. To avoid duplicate code development work, and because CEN codes will be harmonized to ISO codes, STRICE code development objectives were adjusted to support ISO Ice Load Code development. It is especially important to observe the European interest in the ISO code. To achieve this, LOLEIF and STRICE researchers are active members in the ISO code development work. There are three researchers in the international administrative board, and in Technical Panels, where the actual code writing is conducted, there are two STRICE members as chairmen, and three others as members.
A spectral model was derived for a dynamic analysis of vertical offshore structures. The model concerns ice actions in conditions where level ice is acting on the structure and fails by crushing. The model was derived directly from the full-scale data provided by the STRICE project and concerns conditions where ice fails by continuous crushing. In these conditions the ice failure process can be considered as a stationary stochastic process. The main input parameters of the new model include the standard deviation and the mean value of the local ice forces. The standard deviation is used to obtain the autospectral density functions of the local ice forces that act normal to the contact surface. These functions are obtained from a non-dimensional spectral function, which depends slightly on the ice velocity but is otherwise independent on ice parameters. Coherence functions were also derived using the full-scale data on the local forces. The coherence functions are used to define the cross-spectral density functions between the local forces. They are used to consider the correlation of the local forces both in the space and in time. Estimates of the peak values of the local ice forces are needed in the preparation of the input parameters for this model. Therefore, a separate data analysis was done to obtain these peak forces. The peak forces were studied in terms of the effective pressure, which is the peak force divided by the nominal contact area. A new formula was found for the effective pressure as a function of the contact width and the ice thickness. This formula predicts that the ice force has scale effects separately with respect of the ice thickness and the contact width. An exponential term was proposed to consider how the effective pressure reduces with an increase of the contact width. Full-scale data on the lighthouse Norstomsgund as well as from several other offshore structures were used to evaluate the relationship between the effective pressure and the ice thickness. It was shown that a function, which is inversely proportional to the square root of the ice thickness, provides an upper bound for the effective pressure. Fransson et al. (1993, 1994a, 1994b, 1995) found exactly the same relationship in a study where they applied a simple fracture mechanical approach on their data. Subsequently, Dempsey et al. (2001) and Palmer & Dempsey (2002) made a more detailed analysis of the ice-structure contact phenomena known as "contact line" and "hot spots". They also showed that the functional relationship between the effective pressure and the ice thickness is the same as found in the present data analysis. Non-linear effects of ice-structure interaction were not considered in the new spectral model. Therefore, the model has two limitations. First, the phenomenon of self-excited vibration is not incorporated. Second, the model applies only for structures that are relatively stiff at the waterline. This restriction should be considered because it is deemed that the ice forces will increase if the waterline displacement exceeds a level where the continuous crushing mode changes into intermittent crushing. Based on other tests in laboratory conditions, it is estimated that this critical level of the displacement is about 1% of the ice thickness. While using the new model, a full response analysis is done in the frequency domain. FE programs can be used in the same way as while analysing the response to the actions or wind or waves. The resultant spectral densities of various response parameters can be analysed further to obtain various statistical values. Hence, the model can be used in the design for limit states posed by fatigue, serviceability and the ultimate strength of the structure.
Measurements of Ice Forces and Ice Force Effecting Parameters were carried out during the winters 2001, 2002 and 2003 on the lighthouse Norstromsground, which is located in the northern part of the Gulf of Bothnia abt 30nm south of the town of Lulea at the edge of the drifting ice. During the three occupation periods complete sets of full scale ice force and structure response data, ice force effecting parameters, time-lapse video records and still photographs were collected. In a logbook, spreadsheet file, all valuable information was documented. The measuring campaign of the winter 2001 started in February 24th and ended in April 12th. This winter could be classified as a mild one, but nevertheless the measuring campaign was very successful. An overview of ice interaction events, collected data and recorded video sequences during that period: - About 130 events; - About 3.6 GB of raw data were collected; - 1100 h of recorded time-lapse video; - 4 h real time video; Several events of steady-state vibration were observed and recorded. The measuring campaign of the winter 2002, started on February 12th and ended on April 11th. This winter could be again classified as a mild one, but the measuring campaign was as successful as the first one. An overview of ice interaction events, collected data and recorded video sequences during that period: - About 179 ice load events occurred; - About 4 GB of raw data were collected; - 1200 h of recorded time-lapse video; - 10 h real time video; Several events of steady-state vibration were observed and recorded.

Searching for OpenAIRE data...

There was an error trying to search data from OpenAIRE

No results available