13-7-2015

Challenge the future

DelftUniversity ofTechnology

Pilot application of Delft3D Flexible Mesh

Assisting a field campaign at the Sand Motor

Arjen Luijendijk, Lisanne Braat, Niek Waagmeester, Freek Scheel,Matthieu de Schipper, Marcel Stive

2

Sand Motor (ZM) at the Dutch coast

Area of interest: Sand Motor

Mega-nourishment of 21 mln m3

Scale of intervention:• 2 km alongshore• 1 km cross-shore

Large signal-to-noise ratio

3

Aerial photograph 3D view of measured bed levels

Sand Motor – Morphological evolution

4

Motivation – Field campaign MegaPEXMulti-disciplinary field measurements for 6 weeks – Sept14

Goal: Operational hydrodynamicpredictions for field measurements

Dye experiment

5

Coastal processes around the ZM

Hydrodynamics• Tidal currents• Wind- and wave driven• Fresh river outflow

Waves• From SW and NW

Sediments• Sandy shores• Cohesive sediments from river

Rotterdam

Delft

World Forum

Curvilinear - Zeedelta model

No 3D operational model availablecovering both river mouth and ZM

6

Computational mesh

Boundary conditions• Tidal conditions• Fresh river discharges• Wind

Resolution• Deeper water: ~300 m• Sand engine: ~17 m

Numerics• ~100,000 nodes• 5 layers à 15 layers

7

Delft3D - Flexible Mesh

Benefits using Flexible Mesh compared to Curvilinear

• Optimal resolution where needed• Optimisation of #grid cells (~50% of curvi grid)

• Variable time step• Parallelization• Finite volume• Implicit time integration

• Both BMI compliant – communication with other software• Both are Open Source

8

Model results – Surface salinities

Validation on remote observations

9

Model validation - fronts

• X-Band Radar images

A good qualitative match was found andhence model results were used thereafter

10

Findings

• ZM is influenced by hydrodynamics on multiple scales• System understanding on multiple scales is a prerequisite when conducting

local measurements

• Flexible Mesh allows modeling this interaction on multiple scales• Validation is taking place with MegaPEX data; first results very promising

• Large added value compared to Delft3D-curvilinear• Flexible resolution, # of grid cells, parallel

11

Outlook

• Waves (unstructured SWAN) and morphology are being added• beta version in September

• First step towards seamless modelling for coastal morphology

• Forcing-dependent processes

• Forcing-dependent level ofprocess representation

• Optimised predictions enablesensemble runs

Seamless modelling

12

Meirelles: session 3D

De Vries: session 15a

Model combines multi-disciplinary researchValuable tool in STW-NatureCoast

Transports onlower shoreface

Sedimentgrading

Tidalcharacteristics

Aeoliantransports

HydrologyCohesivesediments

Radermacher: session 8F

Hydrodynamics

Benthos

13

Questions / suggestions?

Acknowledgements:• Lisanne Braat (UU PhD)• Niek Waagmeester (MSc)• Freek Scheel (Deltares)• Rijkswaterstaat (data)• STW – NatureCoast• NEMO