The repeated passage of ships through ice-infested waters create a field of broken ice pieces. The typical size of the broken ice pieces is generally <2.0 m. This area might be referred to as a brash ice field. The movement of ships and vessels leads to the transportation and accumulation of broken ice pieces in a brash ice field. A better understanding of the properties and behavior of brash ice could improve the estimates of ice load that are associated with shipping in a brash ice field. An in situ test referred to in this study as a pull up test will be performed in Luleå harbor, Luleå, Sweden. An attempt will be made to estimate the mechanical and physical properties of a brash ice field based on the in situ test results. The test setup, procedure, and test results will be described in detail. Furthermore, the test will be simulated using the smoothed particle hydrodynamics (SPH) formulation. The numerical simulations will calibrate the numerical and material model of brash ice using the pull up test measurements. In this numerical model, a discrete mass-spring-dashpot model will be used to simulate buoyancy and drag. The continuous surface cap model (CSCM) will be used as a material model for the brash ice. The elastic modulus and the fracture energy of brash ice as a material model input will be estimated by an ad hoc scaling formula. The parameters, such as void fraction (Vf), cohesion (c), and angle of internal friction (φ) will be altered to assess their influence on the test data. The analysis of the in situ test results and the simulation results provide a preliminary approach to understand the brash ice failure process that could be further developed into modeling techniques for marine design and operations.

Winter navigation in the North Sea is expanding with respect to vessel size and traffic volume. Icebreakers create routes for ice-going vessels by breaking the level ice cover. Repeated vessel passages in the fairways and harbors initiate the formation of brash ice. The brash ice has the ability to refreeze quickly. In the current work, a field study was conducted on a refrozen brash ice ship channel located in Marjaniemi harbor in Bay of Bothnia. Aim of this study is to evaluate the structure and the strength of ice in the fully refrozen ship channel. Ship channel geometry, ice temperature and salinity were assessed in the field. The ice thickness was in average 45 cm covered by a snow layer with an average of 20 cm. The temperature profiles showed approximately -15ᴼC at the ice surface and close to 0ᴼC in the depths above 10 cm. Salinity varied from 0 to 1.5 ppt. Ice texture, density and compressive strength of refrozen brash ice were measured in the laboratory on 200 mm diameter cores. The behavior of refrozen brash ice with random ice texture was more ductile and stronger in uniaxial compression compared to the adjacent level ice.

Brash ice growth in frequently navigated areas like fairways or ports is quick due to the ‘freezing – breaking’ cycle induced by sub-zero temperatures and ship traffic. This problem is very acute in ports in Arctic areas where the temperatures are very low for long durations and the ship traffic is frequent. In order to take adequate action in managing the brash ice, the forecasts of the amount of brash ice expected should be reliable. The aim of this work is to develop and validate these prediction methods.

The growth model developed is based on extension of earlier growth models which modify the Stefan type growth modelling. The improvement on the earlier models is that the brash ice layer is divided into three layers (instead of two in earlier models): The consolidated layer just below the water level, the brash ice over the water level and the unfrozen brash ice below the consolidated layer. The thermodynamic model follows the Stefan formulation including only the heat flux from latent heat release upon freezing (Stefan, 1891 and e.g. Anderson, 1961). The modelling includes the cyclic breaking and refreezing.

The validation of the model is made using measurements carried out in winter 2013 in Luleå port and in winter 2015 in Sabetta in the Yamal peninsula. Luleå data suggests that the sideways motion of brash ice due to ship motion and wake should be taken into account when assessing the brash ice thickness. The analytical calculation over-estimates the brash ice thickness in the actual channel but under-estimates the total amount of broken ice. When applied to Sabetta data, the analytical calculation predicts well the observed brash ice thickness. It can be concluded that the analytical method that does not take into account any radiation heat fluxes can be applied in the high Arctic where solar radiation plays a minor role and ice surface is clearly below zero.

The behavior and properties of brash ice are important issues for the design of ice-going vessels. Heavy brash ice conditions may cause vessels to be dependent on ice-breaker assistance and time delays in the shipping schedule. Brash ice properties are not well studied and full-scale field data are missing in order to verify numerical models on brash ice and broken sea ice in general. The recent study describes new field equipment for testing brash ice and its functionality is tested on brash ice produced by the Swedish Ice-breaker Oden during ice management operations in the Barents Sea. The equipment consists of a big collector, connected to a crane, which is lowered below the brash ice cover. The brash ice mass above is pulled up by the crane and the force required for pulling is measured. A series of 18 field tests were performed and presented. Strengths and weaknesses of the method were evaluated. Ice blocks sizes were measured. The peak load during pull-up was often at least twice the weight of the lifted ice blocks when the blocks were interlocked. For free floating blocks, the peak load conformed to the weight of the blocks.

First-year ice ridges are one of the main load scenarios that off-shore structures and vessels operating in ice-covered waters have to be designed for. For simulating such load scenarios, the knowledge gap on ice mechanical properties from the consolidated part of first-year ridges has to be filled. In total 410 small-scale uniaxial compression tests were conducted at different strain rates and ice temperatures on ice from the consolidated layer of 6 different first-year ridges in the sea around Svalbard. For the first time uniaxial tensile tests were performed on ice from first-year ridges using a new testing method. Ice strength was evaluated for different ice type, which are determined for each specimen based on a proposed ice classification system for ice from first-year ridges. 78% of all samples contained mixed ice with various compounds of brecciated columnar and granular ice. Ice strength of mixed ice showed isotropy, except for the samples containing mainly columnar ice crystals. For horizontal loading, mixed ice was stronger than columnar and granular ice. The residual strength of ductile ice depended on the strain rate. At 1.5% strain remained 70% of peak strength at 10−4 s−1 and 50% at 10−3 s−1. Ductile failure dominated for 75% of all mixed ice tests at 10−3 s−1 and − 10 °C. Ductile compressive strength was generally higher than brittle compressive strength for mixed ice. Brine volume was the main parameter influencing the tensile strength of the mixed ice which was between 0.14 MPa and 0.78 MPa measured at constant ice temperature of −10 °C.

The results from 3 years of comprehensive field investigations on first-year ice ridges in the Arctic are presented in this paper. The scopes of these investigations were to fill existing knowledge gaps on ice ridges, gain understanding on ridge characteristics and study internal properties of ice. The ability of developing reliable simulations and load predictions for ridge-structure interactions is the final principal purpose, but beyond the scope of this paper. The presented data comprise ridge geometry, ice block dimensions from ridge sails, ice structure in the ridge and values on the ridge porosity and the degree of consolidation. The total ridge thickness conformed to other ridges studied in the same regions. The consolidated layer thickness was on average 2–3 times the level ice thickness. Minimum 33% and in average 90% of the ridge keel area was consolidated. The distribution of ice block sizes and block shapes within a ridge appears to be predictable. A new approach for deriving a possible ridging scenario and ridge age is presented. Different steps of the ridge building process were identified, which are in good agreement with earlier simulated ridging events. After formation of very thin lead ice between two floes deformation occurs through rafting and ridging until closure of the lead. Subsequently the adjacent level ice floe fractures proceeding ridge formation until ridging forces exceed driving forces. A time span of 10 days could be assessed for a possible ridge formation date, estimating the ridge age of the studied ridge located east of Edgeøya at 78° N to be 7 to 8 weeks.

During three expeditions in March 2011, March 2012 and April/May 2013 with the Norwegian coastguard vessel "KV Svalbard", two pressure ridges around Svalbard, one ridge in Fram Strait and three ridges in Olga strait have been studied with respect to ridge geometry and physical and mechanical properties. With a discrete measurement, which is of most interest here, information can be obtained on the overall size and shape of an individual ridge at one particular location, both above and below the waterline. This approach provides detailed quantitative information on both the sail and keel size and shape, of a specific ridge, as well as its porosity. It does not supply any information on ridge spacing. The results of individual profile measurements are discussed. Measurements of vertical profiles along the spine and transects perpendicular to the spine are presented for these ridges. The sail height, sail width, keel depth and keel width, consolidated layer thickness, rubble block sizes and porosities are examined for each ridge. The results presented in this paper contribute to more data in terms of geometry and morphology of the first-year ridge off Svalbard, in Fram Strait and Barents Sea. Compression and tensile strength properties are important input data for constitutive modelling. Still strength properties of ridged ice are not yet sufficiently investigated. Uniaxial compressive tests were performed with horizontal and vertical loading directions and the results from the testing of level ice and consolidated layer are summarized with consideration of total porosity.

Issituationen i Luleå hamn vintern 2012/13 var normalsvår vilket innebar en ökad istillväxt motsvarade 150 cm ren is i områden som bröts kontinuerligt. Detta ska jämföras med det obrutna istäckets tjocklek som var ca 60 cm. Isen i rännan bestod av klotformade isblock (10 – 120 cm) omgivna av issörja. Under mars månad var isblockens storlek mätt vid vattenytan i genomsnitt 45 cm och andelen vatten eller finfördelad issörja var ca 30%. Analyser av isborrkärnor visar att isblocken hade en hållfasthet som ökade med antalet brytningar och var i paritet med det ostörda istäcket i början av mars. Blocken bestod då till 70% av finkrossad is eller frusen snösörja.Isproduktionen i området med bruten is tycks vara linjärt beroende av antalet negativa graddagar där tillväxttakten uppmättes till 0,235 cm per grad och dygn. En numerisk modell för beräkning av istillväxt föreslås där frysning av issörja vid ytan och under isblock ingår. Modellen stämmer bra överens med uppmätta värden från en ränna nära hamnen som bröts kontinuerligt två gånger i veckan. Mer fältstudier av isbildning och uppbrytning är önskvärd för att öka modellens tillförlitlighet under förhållanden som skiljer sig väsentligt från de som rådde i testrännan. Samtliga mätprotokoll från ismätningarna i rännan finns bifogade i rapporten.Tillgänglig statistik visar att antalet negativa graddagar efter isläggningen i Luleå hamn vid en svår isvinter är ca 1000. Våra mätningar tyder på att det under en sådan vinter bildas 2,4 m ren is om ett vändområde används kontinuerligt. Beräkningar med den numeriska modellen resulterade i en möjlig istjocklek på 3 m under en svår vinter om medeltemperaturen är 50% kallare än under den aktuella mätperioden 2012. I nuläget klarar hamnisbrytaren Viscaria att operera och vända fartyg i bruten is motsvarande 1,4 m. Om tjockare is bildas är det väsentligt att ett nytt vändområde med ostörd is kan tas i bruk. Under en svår isvinter krävs därför att minst tre åtskilda vändområden är tillgängliga i Luleå hamn.