towards multiscale green sea loads simulations in
play

Towards Multiscale Green Sea Loads Simulations in Irregular Waves - PowerPoint PPT Presentation

Towards Multiscale Green Sea Loads Simulations in Irregular Waves with the Naval Hydro Pack Inno Gatin, Vuko Vuk cevi c, Hrvoje Jasak Faculty of Mechanical Engineering and Naval Architecture, Zagreb, Croatia 25/July/2017 FSB 1/20


  1. Towards Multiscale Green Sea Loads Simulations in Irregular Waves with the Naval Hydro Pack Inno Gatin, Vuko Vukˇ cevi´ c, Hrvoje Jasak Faculty of Mechanical Engineering and Naval Architecture, Zagreb, Croatia 25/July/2017 FSB 1/20 Motivation Procedure and methods Conclusion 25/July/2017

  2. Different procedures and methods for different scales are presented. Motivation, Procedure and methods, Preliminary results. FSB 2/20 Motivation Procedure and methods Conclusion 25/July/2017

  3. The objective is a complete numerical framework for green sea load calculation. • A multiscale framework comprising CFD and a large scale method, • The large scale method takes into account the statistical nature of wave loads , • CFD uses the results from the large scale method to compute highly nonlinear wave loads due to green sea. FSB 3/20 Motivation Procedure and methods Conclusion 25/July/2017

  4. Three–scale procedure is proposed in this work. 1. Linear seakeeping → exploration of multiple sea states and heading angles → selection of the most adverse condition, 2. Coarse CFD → conducting a three hour seakeeping simulation for the selected sea state → detecting green sea events , 3. Fine CFD → conducting a detailed green sea simulation on the critical part of the deck structure → loads on deck structures. FSB 4/20 Motivation Procedure and methods Conclusion 25/July/2017

  5. Procedure and methods FSB 5/20 Motivation Procedure and methods Conclusion 25/July/2017

  6. Step 1: Hydrodynamic coeffs obtained 1 with linearised free surface solver. • Single–phase simulations with linearised free surface model, • Efficient wave diffraction and radiation simulations. 1.534e+09 1.532e+09 1.53e+09 F Z , N 1.528e+09 1.526e+09 1.524e+09 0 50 100 150 Time, s FSB 6/20 Motivation Procedure and methods Conclusion 25/July/2017

  7. Linearised free surface solver agrees 1 well with potential flow methods. 4 3.6 1.4e+08 3.2 Naval Hydro 1.2e+08 Naval Hydro HydroSTAR HydroSTAR 2.8 1e+08 2.4 φ Fz , rad F Z , N/m 8e+07 2 1.6 6e+07 1.2 4e+07 0.8 2e+07 0.4 0 0 0.2 0.3 0.4 0.5 0.6 0.2 0.3 0.4 0.5 0.6 ω , rad/s ω , rad/s 6 6e+09 5 Naval Hydro 5e+09 HydroSTAR 4 Naval Hydro HydroSTAR 4e+09 φ My , rad M Y , Nm/m 3 3e+09 2 2e+09 1 1e+09 0 0 0.2 0.3 0.4 0.5 0.6 0.2 0.3 0.4 0.5 0.6 ω , rad/s ω , rad/s FSB 7/20 Motivation Procedure and methods Conclusion 25/July/2017

  8. Step 2: Use linear seakeeping methods 2 to assess green water probability. • Using the calculated hydrodynamic coefficients, calculate the green water probability for a large number of sea states , • Select the most adverse sea state, which will serve as a starting point for the next step. FSB 8/20 Motivation Procedure and methods Conclusion 25/July/2017

  9. Step 3: Calibrate the input spectrum 3 using Higher Order Spectrum method. Higher Order Spectrum (HOS) method: • Pseudo-spectral method for solving nonlinear boundary conditions for free surface waves , • Takes into account nonlinear wave–wave interaction and modulation , • Appropriate for efficient nonlinear irregular sea state propagation, • Applicable for coupling with CFD, • Low CPU expense. FSB 9/20 Motivation Procedure and methods Conclusion 25/July/2017

  10. Directional wave spectrum is efficiently 3 propagated using HOS. 3 hours of real time simulated in 5 minutes of CPU time. FSB 10/20 Motivation Procedure and methods Conclusion 25/July/2017

  11. Low CPU expense of HOS enables 3 fast calibration of input spectrum. • The selected wave energy spectrum is calibrated using HOS in order to produce the target spectrum, • Up to 100 three–hour realisations using HOS needed for the calibration → a bit difficult with CFD. . . 200 Target: JONSWAP 150 HOS result after calibration HOS result before calibration 2 s S ζζ , m 100 50 0 0.3 0.4 0.5 0.6 0.7 ω , rad/s FSB 11/20 Motivation Procedure and methods Conclusion 25/July/2017

  12. 3 • Comparing HOS and CFD wave spectrum reveals that minimal wave damping occurs in CFD. 200 HOS CFD 150 2 s S ζζ , m 100 50 0 0.2 0.3 0.4 0.5 0.6 0.7 ω , rad/s FSB 12/20 Motivation Procedure and methods Conclusion 25/July/2017

  13. Step 4: Perform a three hour CFD 4 seakeeping simulation. • SWENSE method is used to couple potential flow and CFD, • Fast and robust simulations with coarse temporal (200 time–steps/period) and spatial (600 000 cells) resolution, • Small number of nonlinear iterations per time–step is enabled using enhanced 6–DOF–fluid flow coupling . 3e+06 2e+06 1e+06 F X , N 0 -1e+06 3e+06 2e+06 -2e+06 1e+06 0 -3e+06 -1e+06 1000 2000 3000 4000 5000 6000 7000 Time, s -2e+06 3900-3e+06 3650 3700 3750 3800 3850 FSB 13/20 Motivation Procedure and methods Conclusion 25/July/2017

  14. Enhanced coupling reduce 4 CPU time by a factor of 4. Strongy coupled 6 23 -0.28 Enhanced Ref. solution 5.8 22.5 -0.29 γ RT1 , rad R T0 , N R T1 , N 5.6 22 -0.3 5.4 21.5 -0.31 5.2 21 5 20.5 -0.32 0 4 8 12 0 4 8 12 0 4 8 12 N N N -0.16 -1.6 0.93 -0.165 -1.7 0.9 γ z1 , rad -0.17 z 0 , m z 1 , m -1.8 0.87 -0.175 0.84 -1.9 -0.18 -0.185 -2 0 4 8 12 0 4 8 12 0 4 8 12 N N N -0.065 1 -0.07 -0.54 0.995 γ φ 1 , rad φ 0 , rad φ 1 , rad -0.075 -0.55 0.99 -0.08 -0.56 0.985 -0.085 -0.57 0 4 8 12 0 4 8 12 0 4 8 12 N N N FSB 14/20 Motivation Procedure and methods Conclusion 25/July/2017

  15. Step 5: Detect green water events and 5 perform detailed CFD simulation. • Customised post–processing tools detect the situations where water on deck occurred in the three hour simulation , • Select the green water incident which is considered the most dangerous, • Conduct a detailed CFD simulation with fine spatial and temporal resolution, including complex geometries. FSB 15/20 Motivation Procedure and methods Conclusion 25/July/2017

  16. Detailed V&V of green sea loads has been performed. Experiments performed for a static FPSO model at SNU: Lee, H.H., Lim, H.J. and Rhee, S.H.: Experimental investigation of green water on deck for a CFD validation database (2012). • Green water pressure is compared at ten locations on deck , • Nine incident waves are considered, • isoAdvector geometric VOF method is used for interface tracking, • Grid, temporal and periodic uncertainty is assessed . FSB 16/20 Motivation Procedure and methods Conclusion 25/July/2017

  17. isoAdvector preserves a sharp interface for green water simulations. Author: Dr. Johan Roenby, DHI . FSB 17/20 Motivation Procedure and methods Conclusion 25/July/2017

  18. Pressure peaks and pressure impulses are compared to experiments. A detailed verification study has been performed for nine waves: • Numerical uncertainties = periodic + discretisation uncertainties, • Experimental uncertainties = periodic + measuring uncertainties, • 20 wave periods simulated to achieve periodic convergence, • Four grid levels used: from ≈ 200 000 to ≈ 4 000 000 cells. 700 8000 600 500 6000 400 p, Pa p, Pa 4000 300 200 2000 100 0 0 6 8 10 12 14 16 14 16 18 20 22 24 26 Time, s Time, s Location further from the Location near the breakwater. breakwater. FSB 18/20 Motivation Procedure and methods Conclusion 25/July/2017

  19. Results agree well with experiments. H = 13.5 cm, λ = 2 . 25 m H = 15.0 cm, λ = 3 . 0 m 600 CFD CFD 800 EFD EFD 500 700 600 400 p max , Pa p max , Pa 500 300 400 300 200 200 100 100 0 0 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Pressure gauge label Pressure gauge label Pressure peaks, Pressure peaks, 300 300 CFD CFD EFD EFD 250 250 200 200 P, Pa s P, Pa s 150 150 100 100 50 50 0 0 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Pressure gauge label Pressure gauge label Pressure impulses Pressure impulses (time integrals). (time integrals), FSB 19/20 Motivation Procedure and methods Conclusion 25/July/2017

  20. All the components of the procedure have been thoroughly validated. A comprehensive procedure for green water load assessment includes: • Stochastic nature of ocean waves, • Complicated geometries. • Validated numerical methods : � Higher Order Spectrum method, � Linearised free surface solver, � SWENSE method with enhanced 6–DOF–fluid flow coupling algorithm, � isoAdvector and Ghost Fluid Method for highly resolved green water simulations. Future work: • Conduct the complete procedure for an example vessel. FSB 20/20 Motivation Procedure and methods Conclusion 25/July/2017

  21. All the components of the procedure have been thoroughly validated. A comprehensive procedure for green water load assessment includes: • Stochastic nature of ocean waves, • Complicated geometries. • Validated numerical methods : � Higher Order Spectrum method, � Linearised free surface solver, � SWENSE method with enhanced 6–DOF–fluid flow coupling algorithm, � isoAdvector and Ghost Fluid Method for highly resolved green water simulations. Future work: • Conduct the complete procedure for an example vessel. Questions? FSB 20/20 Motivation Procedure and methods Conclusion 25/July/2017

Recommend


More recommend