Application of STAR-CCM+ in Marine and Off-Shore Engineering: State-of- the-Art and New Developments S. Enger, M. Perić , F. Schäfer, E. Schreck, J. Singh
Important Features of STAR-CCM+ • Easy process automation for maximum productivity • High-resolution interface-capturing scheme for free surfaces (sharp interfaces, avoiding mixing) • Different wave models, wave damping • Cavitation modelling, user calibration • Dynamic fluid-body interaction (6 DoF body motion), superposition of motions • Overset grids for maximum flexibility in handling body motion • Implicit fluid-structure interaction
High-Resolution Interface-Capturing, I • The scheme combines upwind and downwind discreti- zation to obtain optimal resolution of free surface (typically in one cell). • All fluids involved can be compressible (liquids and gases). • Users can modify some parameters for specific control: – Avoid blending with upwind when marching toward steady- state solution (raise CFL-limits); – Activate anti-diffusion to avoid dilution of liquid in gas through violent sloshing, wave overturning, splashing etc. (sharpening factor > 0).
High-Resolution Interface-Capturing, II Simulation of sloshing in a tank due to a sinusoidal sway motion: one-cell sharp interface before wave overturns and smearing after splashing, when the grid is not fine enough to resolve liquid sheets and droplets . Sharpening prevents dilution and the interface becomes sharp again…
Waves, I • STAR-CCM+ offers several wave models (for initialization and boundary conditions; arbitrary direction of propagation): Linear 1st-order wave theory (for small-amplitude waves); Non-linear Stokes 5th-order wave theory (after Fenton, 1985); Pierson-Moskowitz and JONSWAP spectra (long-crested irregular waves); Superposition of linear waves with an arbitrary direction of propagation, amplitude and period (irregular sea states)... • Accurate wave propagation (with a minimum damping of amplitude) is achieved by 2 nd - order time discretization… • … which imposes a limit on time -step size (wave propagation by less than half a cell per time step).
Waves, II • Any experimental means of wave generation can be easily simulated in STAR-CCM+, e.g. using an oscillating flap: • “Beach” is simulated by applying exponentially growing resistance to vertical fluid motion over a prescribed distance towards boundary.
Waves, III • Wave train initialized using Stokes 5 th order theory over 1002 m (8 wavelengths); Wave damping applied over the last 300 m; Wave period 8.977 s, wave height 5 m • 20 cells per wave height, 80 cells per wave length, 2nd-order discretization in time and space (recommended set-up...) Wave profile after 100 s of simulation time (> 11 periods). Note: 1 cell resolution, very small reduction in amplitude… Scaled 10 times in vertical direction…
Cavitation Modeling, I • The homogeneous two-phase model is used, in which both phases are considered components of a single effective fluid. • The equation for volume fraction of vapor has a source term which describes the growth and collapse of cavitation bubbles – based on Rayleigh equation: Saturation pressure Bubble radius Local pressure Liquid density
Cavitation Modeling, II The model has two parameters: Seed bubbles, uniformly distributed in liquid ( n 0 bubbles per unit volume of liquid); All seed bubbles have the same initial radius. Volume fraction of vapor in a control volume: The growth rate of bubble volume: The source term in equation for vapor volume fraction:
Cavitation Modeling, III • A multiplier of the source term is provided for user to set up (default is 1.0): – Either as a constant or field function; – May be different for positive (bubble growth) and negative (bubble collapse) source term. → • This allows implementation of a new (user) cavitation model by making the multiplier such that the existing source term cancels out:
Superposition of Motions • Superposition of vessel motion, propeller rotation, and oscillatory motion of each blade: easy set-up through GUI, no user programming needed...
Overset Grids, I Optimization of tidal turbine design using overset grids…
Overset Grids, II Simulation of lifeboat launching using overset grids…
Patrol Vessel, Validation Study, I Detailed simulation of flow, resistance, trim and sinkage were performed at the towing tank facility “ Brodarski Institut ” in Zagreb, Croatia…
Patrol Vessel, Validation Study, II Experiments were performed in the towing tank of “ Brodarski Institut ” in Zagreb, Croatia, after simulations were finished. Resistance, trim and sinkage obtained in experiments agree well with simulation, both qualitatively and quanti- tatively, over the whole range of Froude numbers.
Examples of Industrial Application, I Solving a problem with an existing barge, which did not follow the tug… The barge was deviating from the course by up to 250 m…
Examples of Industrial Application, II 5 modified designs tested in simulation – the best one was implemented… Best modified design Original aft shape Course deviation in simulation reduced to The barge was deviating ~1 m – modified vessel from the course by up to behaved similarly… 250 m… Original design Modified aft shape
Examples of Industrial Application, III
Examples of Industrial Application, IV
Examples of Industrial Application, V ORACLE TEAM USA sailing in San Francisco Bay (A merica’s Cup 2013) ORACLE TEAM USA sailing in a high-performance computer cluster (100 million cells, 256 cores; powered by STAR-CCM+, steered by Mario Caponnetto and his CFD analysis team)
Examples of Industrial Application, VI ORACLE TEAM USA: The boat was designed and optimized solely by using simulation – no model experi- ments done… Simulations accompa- nied the race, guided changes to vessel (the night before the last race some modifications to rudder were done based on simulation results) and provided performance data to the crew…
Examples of Industrial Application, VII
STAR-CCM+: New Developments • Additional motion models (prescribed in-plane motion + additional DoF) • Virtual propeller model (using performance curves, theories or coupling to external solvers for propeller flow) • Fluid-Structure-Interaction: Implementing FE-modelling into STAR-CCM+ (see presentation by Alan Mueller) • Custom tool for an automatic set-up of standard tests: resistance, trim+sinkage (in future also PMM, circle, zig-zag …) • Internal wave generation by mass source terms • Coupling to potential flow solver for waves and propellers… • Further developments of overset grids, automatic refinement… • Hydro-acoustics modelling, etc …
New DFBI Motion Types, I New DFBI body motion options: - Four-DoF Maneuvering - Planar Motion Carriage Pure sway Pure yaw
New DFBI Motion Types, II Circle test
Virtual Propeller Model, I SVA Propell ller Virtu tual l Disk Momentum source terms are added to cells within a speci- fied disk zone (grid does not have to be fitted to disk).
With virtual propeller, free Virtual Propeller Model, II surface and hull resistance are well predicted with low cost… Full-scale hull, propeller and Virtual Propeller rudder, Free surface Fixed hull Froude-number 0.21 Rotating Propeller
Internal Wave Generation • Waves generated by mass sources/sinks (injection and suction of water) • Waves reflected off a structure can pass through the internal wave generator • Damping applied at all solution domain boundaries, except where reflection off walls is allowed…
Future Trends • More powerful and affordable computers = higher demands from simulation: More complete system analysis, with all geometrical details; More transient simulations (URANS, DES and LES), predicting pressure fluctuation and noise sources (turbulence, cavitation); More fluid-structure-interaction (slamming, sloshing) and other multi-physics (wind, fire, pollution etc.) applications; Simulation of manoeuvring tests (circle, zig-zag, PMM etc.) and other experiments in the design phase... Simulation of interaction (ship + ice, ship + platform, ship + ship etc.). More automatic optimization studies...
Recommend
More recommend
