OIL & GAS DNV GL s 16th Technology Week Advanced Simulation for Offshore Application: Application of CFD for Computing VIM of Floating Structures 1 SAFER, SMARTER, GREENER
OUTLINE Introduction Elements of Computational Fluid Dynamics Solution Process & Quality Measures VIM & VIV Problem Description Case Studies: Model Scale VIM and Full Scale Model Scale Full Scale Single Column Summary 2
INTRODUCTION Availability of fast computers & robust software has enabled use of CFD for complex problems like VIM CFD of VIM is challenging (flow separation, FSI, mooring damping & stiffness effects and Re dependence Little published full-scale data for validation Questions remain concerning spatial and temporal resolutions in full-scale CFD simulations 3
Elements of Computational Fluid Dynamics Basic Elements Flow Classification Flow Equations Numerical Techniques Errors & Uncertainties Verification & Validation Key Techniques Turbulence Models Rigid Body Motion Fluid-Structure Interaction 4
Solution Process & Quality Measures Geometric Modeling Grid Generation Initial & Boundary Conditions Computation Post-Processing 5
VIM & VIV Problem Description Unsteady Strong fluid/body coupling Flow separation Larger turbulent scales must be resolved Re dependence 6
CASE STUDY Model Scale VIM VIM of deep draft semisubmersible platform Comparison of CFD against scaled model tow test data Used OpenFOAM TM Performed convergence study Compared 3 turbulence models (URANS, DES, SAS) HOE Paired Column Semi 7
NUMERICAL METHODOLOGY Single-phase flow 2nd order implicit scheme for time derivatives 2nd order upwind scheme for spatial derivatives Coupled 6DoF solver Arbitrary Lagrangian- Eulerian (ALE) method to handle rigid body motion Location of platform 8
BASE MESH DETAILS Three mesh resolutions tested (2, 5 & 10M cells) 10 prism layers, y+ < 1 9
Drag Test Validation Heading 0 U = 2 m/s Three turbulence models: URANS komegasst Hybrid URANS-LES SA-IDDES agrees well within 3 10
Drag Test Q-Criterion Drag Test Q-Criterion SA-IDDES (LES) komegasst (URANS) 11
Drag Test Eddy Viscosity Eddy Viscosity = Turbulent Viscosity transfer of momentum SA-IDDES (LES) komegasst (URANS) 12
Decay Test Natural Period Comparison of estimated natural period in sway and yaw Sway Yaw Experiment 15.5 s 9.3 s OpenFOAM 15.3 s 9.4 s Acusolve 15.2 s 9.3 s Fluent 15.4 s 9.5 s Star CCM+ 15.3 s 9.6 s 13
VIM Q-criterion & Velocity Mag. 14
Experimental & CFD Results 0 heading: U r = 4, 6, 8, and 10 15
CASE STUDY Full Scale Single Column Flow around full scale fixed column Single Column at 0,8,22.5 and 45 angle of attacks Comparison of CFD against scaled model tow test data Used OpenFOAM Performed mesh sensitivity Used Spalart-Allmaras IDDES turbulence model 16
FS vs MS Contour of Velocity Magnitude 0 angle of attack Re=3.3x10 4 (Model-scale) Re=14x10 6 (Full-scale) 17
Drag and Lift Coefficients 0,8,22.5 and 45 headings: Comparison of time averaged drag and lift coefficient for a fixed square column with rounded corners 18
SUMMARY CFD is increasingly being used as an alternative to model tests in the assessment of fluid dynamics aspects of offshore structures. This study demonstrates the effectiveness and accuracy of free, open source CFD software, OpenFOAM to predict VIM response. DNV GL attempts to provide a best practice guideline for applying CFD to the investigation of VIM and VIV problems of offshore structures. DNV GL has as a long-term goal to evolve these guidelines into a Recommended Practice. 19
Thank You Mustafa Kara Mustafa.Kara@dnvgl.com +1-281-396-1635 www.dnvgl.com SAFER, SMARTER, GREENER 20