Objectives

Background

The interaction of ice with coastal structures is a complex process involving several possible modes of failure and deformation behaviour of ice. It is influenced by numerous factors related to the structure, to the ice and the physical environment. A considerable amount of work has been done to measure and predict the forces which ice features can exert on a structure. There have been several approaches to this problem including analytical and numerical predictions, and laboratory and field studies. Both deterministic and probabilistic approaches have been used and they have been integrated together in certain cases. 

Despite substantial efforts in the last three decades, the available methods and models for prediction of ice forces on structures are far from satisfactory. In most empirical approaches the effective indentation pressure is related to the uniaxial strength of the ice through a set of multiplicative factors representing different phenomena which are observed to influence the relationship. The ice strength values are mostly taken from small scale tests. After all, the majority of methods predict global ice forces which are 10 times or more higher than measured in the field.
 

Task description

Within this task both pioneer and recent approaches to the estimation of global forces on fixed structures during ice - structure interaction are critically reviewed. Ice - structure interactions such as limit momentum, limit force, limit stress and limit fracture as limiting global ice load scenarios are considered. The evaluation of existing ice force prediction methods considers vertical structures such as bridge piers, lighthouses, column-supported or caisson-retained structures. Interaction with both rigid and compliant structures is investigated. 

The modes considered, were: 

• Indentation with emphasis on evaluation of empirical approaches based on small- and medium-scale testing, and full-scale field measurements, analytical and numerical approaches.

• Buckling of ice sheet: For high aspect ratios buckling loads have been investigated as a possible limit on the ice - structure interaction force. The solutions are typically closed-form or numerical evaluations. 

• Ridge failure: The forces are in the literature often reported as the force necessary to fail the ridge. In these studies the ridge is usually idealised as a linear beam where the interaction is assumed to continue until the ridge fails. This is not necessarily true and can lead to substantial overestimation of forces. 

• Rubble failure: When floating sea ice sheets are driven by environmental forces such as wind or current against structures, they can break into small blocks. These blocks, which in the period immediately after their formation behave as a granular material with little or no cohesion, may accumulate to form a variety of features. Design requirements of structures has motivated a number of studies aimed at the determination of the mechanical properties of bulk rubble and the loads it can exert on structures. 

• Discrete impacts: Single floes and broken ice driven by environmental forces can exert substantial forces on structures. The global load can arise either from single impacts or from interaction with an ensemble of discrete ice floes. Several laboratory and numerical studies are made of this ice loading mode. A thorough evaluation of the different approaches is conducted. 

Deliverables

The evaluation of existing ice force prediction methods highlights on

• ranking of parameters and conditions that affect the global ice force on a structure,
• weaknesses in existing theories/methods,
• reasons for discrepancies between global ice force predictions and measured global forces in the field,
• recommended prediction approaches for the different cases and ice load modes,
• recommendations and guidelines for the full scale measurement program
• and a strategy for improvement of the ice force prediction methods.

Interdependence and links with other tasks

Results of the evaluation of existing ice force prediction methods and models, even intermediate results were distributed to all partners by the data management service (Task 3). This information was required as soon as possible by the partners VTT, HUT and LGGE for the numerical modelling (Task 4) because their work based on these earlier findings.