Enabling prediction of ice loads on structures in the Arctic

The aim of the Russian-Norwegian collaboration project ICELOAD is to enable the prediction of ice loads on structures and vessels in the Arctic. Collaborating partners are SINTEF and Skolkovo Institute of Technology (Skoltech), Moskow. Samples of seawater from three locations in the Barents Sea have been brought in by Skoltech. This water, as well as distilled water, was tested in experiments with single drops. Here, details about the freezing of droplets in flight, as well as the collision of cold droplets against various surfaces were studied. A special experimental rig with associated instrumentation was developed and built. Data from these experiments have been used to validate detailed models of droplet freezing. In 2021 Skoltech carried out a trip to the Barents Sea. Here, optical methods for quantifying the structure of ocean waves and impact-generated sea spray were tested. The voyage resulted in significant video material from ocean waves and sea spray. A second voyage will be carried out in 2023, but then just after the Norwegian part of the project has been completed. A pragmatic one-dimensional model for the freezing of water droplets (pure water) moving in cold air is developed. The crystallization model allows an analytical solution if one assumes a uniform temperature distribution in the liquid-filled core of the droplet. The model was validated using 3D commercial CFD software. Collision of partially frozen droplets against a solid wall, assuming plastic deformation of the ice shell, was also studied. The relationship between shell deformation and shell thickness was evaluated. It was assumed that if this ratio was greater than 1.0, the droplets would stick to the cold surface due to the formation of ice bridges. This model has been extended to treat seawater, which is a multi-component system. This type of approximate analytical model can be used in the large-scale flow model. This allows rapid simulation of droplet temperature and the state of the drops when they hit cold surfaces. Detailed calculations of the thermodynamic state of the droplets require too large computing resources to be used in real applications. The flow model has been developed by SINTEF. The model can predict ocean waves ( initialized with JONSWAP spectrum) and interaction with dynamic structures. Drops can be produced from wave-wind interaction or direct impact of waves against a structure. The principles was previously tested in a 2D version of SIMCOFLOW (https://github.com/lovfall/simcoflow). Drops can be generated via two methods: a) Fluid structures that are pulled out of the flow and are no longer resolved on the computational mesh and b) Entrainment of drops from the surface due to local physics. Data to adapt and validate the model will be produced by Skoltech in 2023. The droplets that are produced are treated as packets consisting of many individual droplets. The droplets that form can be cooled and deposited on cold surfaces, or they can fall back into the sea. A methodology for loading *.stl geometry files into SIMCOFLOW has been developed and published. The read geometry file will result in a "level-set" field that represents the object geometry throughout the simulation. The 3D version of SIMCOFLOW has been bug-fixed and documented. An STL geometry of the ship that was on the voyage has been created and imported into SIMCOFLOW. The ship's motion is represented by a Lagrange model that solves for translation and rotation. An important goal has been to arrive at the coarsest calculation network, with an underlying droplet entrainment model, which is able to do a satisfactory job of calculating droplet states and mass fluxes of drops hitting cold surfaces on structures or vessels. The models that have been developed can be used to evaluate the parameters that affect wave-induced spray. By running parameter studies, building a database, and interpolating in this, one can produce quick predictions of the time history for the arrival of wave-induced spray for a given offshore structure. Source code for the 3D CFD model is published on github (https://github.com/SINTEF/Simcoflow3D). The 2D version is available at https://github.com/SINTEF/simcoflow.

In the project we have developed simplified methods to compute the state of water droplets at the time of hitting some cold surface. Based on this, the specific droplet's contribution to ice accretion is assessed. The thermal state of the droplet at impact is here computed from residence times, sea water and air temperatures, as well as other static data. In this way the detailed thermodynamic history of each droplet is not needed. The droplet state at impact can be computed from look-up tables or an analytical model. This allows fast calculation of trajectories of entrained droplets. Theories for droplet entrainment from ocean waves were explored and good qualitative prediction results were produced. We were invited into a meeting with the SINTEF Ocean led SLADE KPN project (wave slamming) and we had good discussions. We started to involve our Russian partner but this could not develop further due to the war on Ukraine. The project has led to interest into applying the results and technology into renewable energy sources. In ocean wind projects the understanding of forces on structures under extreme weather conditions is critical for design of durable installations. This applies to both effects of waves, wind and droplets. In addition, icing possibilities must be considered. Similarly, the same phenomena may be active in any solar-on-water arrangements. Using this new technology, it is possible to access coupled phenomena in a way that previously has been very difficult. The project has established links between Norway and Russia that are promising and could lead to extensive collaboration in this area between the two countries. Unfortunately, the war on Ukraine has set back this promising collaboration for a long time.

Ice accretion on vessels and constructions in the arctic is a significant operational challenge. The main source for water droplets come from wave impacts, and therefor there is a focus on wave induce sprays as the main source for icing. Current prediction models are extremely rough. Some more detailed attempts have been done recently by use of computational fluid dynamics (CFD). Unfortunately, the spray source cannot be related to waves, wind and geometries by existing models and this far rough source models, ignoring wave and geometry interaction, had to be applied. To be able to predict more fundamentally what happens in a given situation new physics and mathematical approaches need to be introduced. This project is a collaboration between SINTEF and Skoltech, Moscow. In this project Norway has the lead on the model development and computational work, while the Russian team has a strong focus on experimental work and providing data for model validation. The project will run over approximately 3 years and will: 1) Further develop method to simulate spray generation due to wave impacts and wind. The large scale wind-wave-structure interaction is simulated (Computational Fluid Dynamics) and based on simulation data the spray detailed source is modeled and droplet flow is simulated as Lagrangian droplets. 2) Extend existing Cartesian Cut-Cell computational method to dynamically simulated 3D vessels or structures, and their interaction with waves and wind. A particular important outcome is simulation of wave impacts and the spray source. 3) Develop method to compute full energy coupling between fields (air, water, droplets) and simulate the droplet deposition 4) Provide open source computational code that can be applied by industry to assess ice load on vessels and structures 5) For the first time, provide experimental field data for model validation (both spray source and ice accretion) . Data will be open. This is work by our Russian partners.