COUPLING BETWEEN DUALSPHYSICS and SWASH MODELS and LATEST APPLICATIONS to COASTAL ENGINEERING PROBLEMS Dr

Home , Coastal engineering

COUPLING BETWEEN DUALSPHYSICS AND SWASH MODELS AND LATEST APPLICATIONS TO COASTAL ENGINEERING PROBLEMS Dr. Corrado ALTOMARE (Flanders Hydraulics Research / Ghent University)

1 Numerical modelling: WHAT FOR?

Physical modelling Flooding Damage 3 basic concepts…

Which reality do we want to simulate?

Which are the coupling techniques we explored?

How efficient and accurate is the coupling? We (I) bet on SPH: why?

Violent Non-linearity Wave-object and hydrodynamics Object-object interactions

Courtesy of Prof. Peter Troch Disadvantages of (WC)SPH

Accuracy Computational cost

Noise in density and pressure field NS equations (affecting wave propagation)

Boundary conditions (GRAND Weakly Compressible Nature CHALLENGE)

Convergence (GRAND Number of Neighbors CHALLENGE) What do we aim at?

Accurate wave generation Good wave transformation

Reasonable computational cost Accurate and detailed modelling of wave-structure interaction What do we aim at?

recently…. DualSPHysics (+ MoorDyn) What do we aim at?

Impacts of wave overtopping flows on coastal defenses and buildings along the Flemish coast Layer thickness Physical model 0.06 SWASH

0.04

h h (m) 0.02

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 What do we aim at? time (s) Layer speed 4

u u (m/s) 0

-2 200 400 600 800Large1000 domains1200 1400 1600 1800 2000 2200 2400 time (s) Instantaneous wave overtopping discharge 100

Detailed wave50- Multi-scaled modeling

structure interaction0 q q (l/s/m)

-50 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Free surfaceLong elevation events at St.6 (toe of dike) 0.2

0.1 eta eta (m) 0

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time(s) Large domains: DONE!! Long events: 1000 waves DONE!!

23.6h runtime (on Tesla K20c ) for 82,541 fluid particles

Err (wave height)=-3.5%, Err (spectral period)= +4.7% COUPLING basic principle

DualSPHysics Wave propagation model (SWASH)

overlap

= shifting the generation boundary in DualSPHysics Layer thickness Physical model 0.06 SWASH

0.04

h h (m) 0.02

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Layer speed COUPLING4 with SWASH 2  SWASH=Simulating u (m/s) 0 WAve till Shore

-2  NLSW equation model200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Instantaneous wave overtopping discharge  Time domain100 model for simulating non-hydrostatic, free-surface and rotational flow.

50  It simulates accurately surface wave and velocity field from deep water. 0 q (l/s/m) -50  Not suitable to deal200 with400 abrupt600 800 changes1000 1200 of 1400shape1600 of 1800the coastal2000 2200 structures2400 . time (s) Free surface elevation at St.6 (toe of dike) 0.2

0.1 eta eta (m) 0

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time(s) COUPLING with SWASH

(1)

(2) Relaxation Zone technique

(3) COUPLING with SWASH (1) MB (Moving boundary): Altomare et al (2015) COUPLING with SWASH (1)

MB (Moving boundary): Altomare et al (2015) COUPLING with SWASH (2)

Relaxation zone (RZ) + damping

푧 [푚]

C(푥) damping 푥 [m]

푊푅푍

푣 푥, 푧, 푡 푅푍 = 퐶 푥 푣푡ℎ푒표푟푦 +(1- 퐶 푥 )푣푆푃퐻

COUPLING with SWASH (2)

2푥 2푥 푡푎푛ℎ + 훼 훽 − 푡푎푛ℎ − 훼 훽 − 푡푎푛ℎ 1 + 훼 훽 − 푡푎푛ℎ 1 − 훼 훽 푊푅푍 푊푅푍 퐶 푥, 훼, 훽 = 푡푎푛ℎ 훼훽 − 푡푎푛ℎ −훼훽 − 푡푎푛ℎ 1 + 훼 훽 − 푡푎푛ℎ 1 − 훼 훽

1 Previous function

0.9 a=0.5, ß=1 0.8 a=0.5, ß=9 a=0.5, ß=5 0.7 a=0.1, ß=1 0.6 a=0.9, ß=10

0.5 C 0.4

0.3 0.2

0.1

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 ퟐ풙

푾푹풁 Different C functions

α=0.1, β=5

α=0.3, β=5

α=0.5, β=5

α=0.9, β=5 RZ-SWASH

푧 [푚] Velocity field SWASH model

Input=H,T,d Run

푥 [m] 0 SPH model 퐷 Input=SWASH model Run

damping

푊푅푍 A CASE OF STUDY Wave overtopping of sea dikes with very shallow foreshores

DualSPHysics stand-alone

SWASH

MB (Altomare et al., 2015)

Relaxation Zone A CASE OF STUDY

DualSPHysics stand-alone

MB (Altomare et al., 2015)

Relaxation Zone A CASE OF STUDY 퐺푒퐹표푟푐푒퐺푇푋 푇𝑖푡푎푛 푋

DualSPHysics stand-alone 푛푝푎푟푡𝑖푐푙푒 = 3,389,266 푅푢푛푡𝑖푚푒 = 95.62ℎ

푛푝푎푟푡𝑖푐푙푒 = 494,388 푥푐표푢푝푙𝑖푛𝑔 = 31. 25푚 푅푢푛푡𝑖푚푒 = 9.95 ℎ MB (Altomare et al., 2015)

푛푝푎푟푡𝑖푐푙푒 = 1,269,820 Relaxation Zone 푥푐표푢푝푙𝑖푛𝑔 = 24.00 푚 푅푢푛푡𝑖푚푒 = 22.65 ℎ A CASE OF STUDY

MB vs RZ-SWASH Layer thickness Physical model 0.06 SWASH

0.04

Conclusions h (m) 0.02 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Layer speed 4

u u (m/s) 0

-2 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Instantaneous wave overtopping discharge Multi scale 100

0 q q (l/s/m)

-50 Multi phase 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time (s) Free surface elevation at St.6 (toe of dike) 0.2

0.1 eta eta (m) Long duration 0

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 time(s) Large domain Remind our target Conclusions

Wave Generation COUPLING Wave Absorption Relaxation Zone + SWASH

Speed up Accuracy

To be continued…