Browsing by Author "Suortti, Juho"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Puu- ja teräslaivat – mitä seuraavaksi?
    (2020-05-03) Suortti, Juho; Romanoff, Jani; Insinööritieteiden korkeakoulu; Mikkola, Tommi
  • Thumbnail Image
    Item
    Validation and comparison of turbulence models in ship wind load prediction
    (2023-05-15) Suortti, Juho; Mikkola, Tommi; Arjava, Jyri-Pekka; Insinööritieteiden korkeakoulu; Ruponen, Pekka
    Safe and efficient operation of ships requires accurate knowledge about the loads exerted by the stochastic environment. Wind loads affect the propulsion power requirement, manoeuvring, stability and auxiliary machinery. Current methods of predicting wind loads on ships through wind tunnel testing are slow and expensive, while computational and empirical methods are highly inconsistent. A particular difficulty is associated with a highly turbulent inflow from all directions, a complex model geometry and a prevalent flow separation. In this thesis, the predictive capacity of Computational Fluid Dynamics (CFD) is assessed from the perspective of the turbulence model. A model-scale validation procedure is carried out in an idealized numerical wind tunnel. Thorough sensitivity studies are performed in attempts to isolate the influence of the turbulence model and optimize the computational resource allocation. A total of six steady-state Reynolds-Averaged Navier-Stokes (RANS) models are compared against two unsteady Detached-Eddy Simulation (DES) models. Following the validation, selected turbulence models are applied to full-scale computations. Current investigations revealed that k-omega SST RANS turbulence model with modified coefficients can provide improved precision over its RANS peers. In addition, k-omega SST Improved Delayed DES (IDDES) model provided promising results with a reasonable computational burden at wind angles where RANS models struggled. Both RANS and DES exhibited major differences in the surge force coefficients once applied to full-scale. Insufficient turbulent kinetic energy and the following delay in resolved fluctuations are identified as an explanation for the highly erroneous IDDES predictions at 90 degrees wind angle.