manualWaves2Foam v090

See
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/319160515
waves2Foam Manual
Technical Report August 2017
CITATIONS READS
0 1,577
1 author:
Niels Gjl Jacobsen

Deltares
29 PUBLICATIONS 474 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Generating waves for OpenFoam: waves2Foam View project
All content following this page was uploaded by Niels Gjl Jacobsen on 17 August 2017.
The user has requested enhancement of the downloaded file.

Niels G. Jacobsen
waves2Foam Manual
August 2017, version 0.9 (SVN-revision 2113)
DRAFT
WAVES2FOAM MANUAL
Niels G. Jacobsen
Copyright
c Niels Gjl Jacobsen, 2017
All right reserved
Front matter: Own picture (Ireland, 2016)

The manual is published digitally
Release history:
Date Version Latest SVN-revision Aliation

August 2017 0.9 (DRAFT) 2113 Deltares, The Netherlands
([email protected])
Disclaimer:
This oering is not approved or endorsed by OpenCFD Limited, producer and distributor
of the OpenFOAM software via www.openfoam.com, and owner of the OPENFOAM
R and
OpenCFD
R trade marks.
Contents
Contents 3
Code Frames 6
Nota Benes 7
Warnings 7
1 Introduction 9
1.1 Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2 Background and application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 Compabitility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.1 Branches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.2 Naming terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.3 Supported versions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4 Installation of waves2Foam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.1 Third-party dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.2 Download and installation . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.3 The bashrc le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.4 How to get the updates to waves2Foam? . . . . . . . . . . . . . . . . . . . 11
1.4.5 How to recompile parts of waves2Foam . . . . . . . . . . . . . . . . . . . 12
1.5 Referencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2 Mathematical Description 15
2.1 Coordinate system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 Momentum and continuity equations . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1 Modications for permeable structures . . . . . . . . . . . . . . . . . . . . 15
2.3 Advection of the indicator eld . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.1 Modications for permeable structures . . . . . . . . . . . . . . . . . . . . 16
2.4 Turbulence modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.5 Reection compensation at the boundary . . . . . . . . . . . . . . . . . . . . . . 16
2.6 Numerical beach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7 Relaxation zone technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7.1 Flavour of relaxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7.1.1 Explicit relaxation . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.7.1.2 Implicit relaxation . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7.2 Specifying the relaxation zone . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7.2.1 Relaxation scheme . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7.2.2 Relaxation weight . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.7.2.3 Relaxation shape . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.8 Specication of resistance due to a permeable structure . . . . . . . . . . . . . . 20
2.8.1 Resistance formulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.8.1.1 Native OpenFoam . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.8.1.2 Engelund (1953) . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.8.1.3 Van Gent (1995) . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3 Utilities 23
3.1 Pre-processing utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.1 waveGaugesNProbes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.1.1 Input: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.1.2 Output: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1.1.3 Run-time sampling . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1.2 setWaveParameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1.2.1 Global parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.1.2.2 Sub-dictionary information . . . . . . . . . . . . . . . . . . . . . 26
3
4 CONTENTS
3.1.2.3 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.1.3 relaxationZoneLayout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.1.4 setWaveField . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1.5 faceSetToSTL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.1.6 sampleIncidentWaveField . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2 Run-time sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.1 Numerical wave gauges (surfaceElevation) . . . . . . . . . . . . . . . . . . 29
3.2.2 Overtopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.2.1 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.2.2 Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2.2.3 Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.3 Post-processing utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.3.1 Numerical wave gauges (surfaceElevation) . . . . . . . . . . . . . . . . . . 30
3.3.2 postProcessWaves2Foam . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4 Wave Theories 33
4.1 Validity of wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2 Algebraic wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2.1 Regular wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2.1.1 Stokes rst order . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2.1.2 Standing Stokes rst order . . . . . . . . . . . . . . . . . . . . . 33
4.2.1.3 Stokes second order . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.2.1.4 Modulated second-order Stokes wave . . . . . . . . . . . . . . . 34
4.2.1.5 Stokes fth order . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.2.1.6 First-order cnoidal theory . . . . . . . . . . . . . . . . . . . . . . 35
4.2.1.7 Stream function wave . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2.2 Bichromatic wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2.2.1 First-order bichromatic wave . . . . . . . . . . . . . . . . . . . . 35
4.2.2.2 Second-order bichromatic wave . . . . . . . . . . . . . . . . . . . 36
4.2.3 First-order irregular waves . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.2.3.1 Spectral shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.2.3.2 Spectral discretisation . . . . . . . . . . . . . . . . . . . . . . . . 38
4.2.3.3 Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.2.3.4 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.2.4 Second-order irregular waves . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2.5 Potential current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2.6 Solitary wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2.6.1 First-order solitary wave . . . . . . . . . . . . . . . . . . . . . . 39
4.2.6.2 Chappelear (1962) . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.7 Combined waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.8 External source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.3 External wave theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.3.1 Empty external method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.3.2 Fast summation of irregular waves . . . . . . . . . . . . . . . . . . . . . . 41
4.3.3 OceanWave3D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.3.3.1 Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.4 Extensions with new theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.4.1 Extension: Algebraic wave theory . . . . . . . . . . . . . . . . . . . . . . . 42
4.4.2 Extension: External source . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5 Solvers 45
5.1 waveFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.2 porousWaveFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.3 waveDyMFoam (moving meshes) . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.3.1 Modications to interDyMFoam . . . . . . . . . . . . . . . . . . . . . . . . 45
5.3.1.1 Changes to les . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.3.1.2 Details on Make/options . . . . . . . . . . . . . . . . . . . . . . 46
5.3.1.3 Details on Make/files . . . . . . . . . . . . . . . . . . . . . . . . 47
6 Tutorials 49
6.1 Execute a tutorial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.2 waveFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.2.1 standingWave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.2.2 waveFlume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
6.2.3 bejiBattjes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.4 couplingOceanWave3D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.5 periodicSolitary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.6 3Dwaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.7 squarePile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.3 porousWaveFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.3.1 porousDamBreak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.4 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.4.1 relaxationZoneLayout . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7 Validation 53
7.1 Loads on bridge decks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.2 Interaction with coastal structures . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7.3 Breaking waves on a beach prole . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7.4 Modelling of oating wave energy converters (WECs) . . . . . . . . . . . . . . . . 55
7.5 Wave-structure-seabed interaction . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Bibliography 57
A Overview of Input Files 61

A.1 Files native to OpenFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
A.1.1 system/controlDict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
A.1.2 system/fvSchemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.3 system/fvSolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.4 constant/dynamicMeshDict . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.5 constant/g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.6 constant/transportProperties . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.7 constant/turbulenceProperties . . . . . . . . . . . . . . . . . . . . . . 62
A.1.8 constant/RASProperties . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.9 constant/LESProperties . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.1.10 constant/porosityZones . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.2 Files related to waves2Foam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.2.1 constant/waveProperties.input . . . . . . . . . . . . . . . . . . . . . . 62
A.2.2 constant/waveProperties . . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.2.3 constant/probeDefinitions . . . . . . . . . . . . . . . . . . . . . . . . . 62
A.2.4 constant/postProcessingProperties . . . . . . . . . . . . . . . . . . . . 62
A.2.5 constant/triSurface/stlDefinitions . . . . . . . . . . . . . . . . . . . 62
B Treatment of Irregular Waves 63

B.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
B.2 Mathematical description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
B.2.1 Repetition of the signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
B.2.1.1 Equidistant discretisation . . . . . . . . . . . . . . . . . . . . . . 64
B.2.1.2 Non-equidistant discretisation . . . . . . . . . . . . . . . . . . . 64
B.2.2 Discrete representations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
B.2.2.1 Equidistant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
B.2.2.2 Cosine stretching . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
B.2.2.3 The discrete representations . . . . . . . . . . . . . . . . . . . . 65
B.3 Spectral discretisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
B.3.1 Time series and autocorrelation . . . . . . . . . . . . . . . . . . . . . . . . 65
B.3.2 Spectral wave properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
B.3.3 Probability distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
B.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
B.4 Accelerated evaluation of irregular waves . . . . . . . . . . . . . . . . . . . . . . . 68
B.4.1 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
B.4.2 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
C Source Code History 71
5
6 Code Frames
Code Frames
1.1 Download and installation instructions for the default installation directory ($ refer
to a command). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2 Updating the SVN-repository and recompilation of waves2Foam. . . . . . . . . . . . 11
1.3 Reference to the original paper accompanying the release of waves2Foam. . . . . . . 12
1.4 Reference to the paper on the porosity implementation in waves2Foam. . . . . . . . 12
1.5 Reference to the coupling between waves2Foam and OceanWave3D. . . . . . . . . . . 13
1.6 Reference to OceanWave3D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Fundamental denition of a relaxation zone . . . . . . . . . . . . . . . . . . . . . . . 17

2.2 Empty relaxation scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Spatial relaxation scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Exponential weight function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Free polynomial weight function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.6 Third order polynomial weight function . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.7 Rectangular relaxation shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.8 Semi-cylindrical relaxation shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.9 Cylindrical relaxation shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.10 Frozen relaxation shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.11 The native relaxation zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.12 Resistance formulation according to Engelund (1953) . . . . . . . . . . . . . . . . . . 20
2.13 Resistance formulation according to Van Gent (1995) . . . . . . . . . . . . . . . . . . 21
3.1 Common settings for denition of wave gauges and probes . . . . . . . . . . . . . . . 23

3.2 A circular point distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 A linear point distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.4 A quadrilaterial point distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.5 User-dened point distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.6 Example of the control of the surface elevation sampling. . . . . . . . . . . . . . . . 25
3.7 Globale settings inwaveProperties.input . . . . . . . . . . . . . . . . . . . . . . . 26
3.8 Overview of the sub-dictionary in waveProperties.input. . . . . . . . . . . . . . . . 26
3.9 Example of waveProperties.input. . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.10 Warning when Tsoft diers from 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.11 Two methods to generate an STL-surface of a unit square. . . . . . . . . . . . . . . . 28
3.12 Instructions for sampleIncidentWaveField into waveProperties.input . . . . . . . 29
3.13 Use of sampleIncidentWaveField . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.14 Run-time sampling of overtopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.15 Face zone for overtopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.16 Formatting of the overtopping data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.1 Settings Stokes First Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4.2 Standing Stokes First Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.3 Stokes Second Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.4 Modulated Second Order Stokes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.5 Stokes Fifth Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.6 First Order Cnoidal Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.7 Stream Function Theory. Note the EITHER-OR structures. . . . . . . . . . . . . . . . 36
4.8 First Order Bichromatic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.9 Second Order Bichromatic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.10 Irregular waves: Common settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.11 Irregular waves: JONSWAP settings . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.12 Irregular waves: Pierson-Moskowitz settings . . . . . . . . . . . . . . . . . . . . . . . 37
4.13 Irregular waves: Frequency discretisation . . . . . . . . . . . . . . . . . . . . . . . . 38
4.14 Irregular waves: Random phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.15 Irregular waves: Focused phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.16 Irregular waves: Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.17 Potential current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.18 Solitary First Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.19 Solitary wave (Chappelear, 1962) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.20 Linear Superposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.21 Use of externalSource . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.22 Empty external source (Default) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.23 Fast summation of irregular wave theories . . . . . . . . . . . . . . . . . . . . . . . . 41
4.24 Coupling with OceanWave3D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.25 Create a new algebraic wave theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.26 Create a new wave theory - Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.1 Example code to compare les in order to understand modications made from
interFoam to waveFoam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2 How to execute the comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.3 Additional lines in Make/options for EXE_INC. . . . . . . . . . . . . . . . . . . . . . . 46
5.4 Additional lines in Make/options for EXE_LIB. . . . . . . . . . . . . . . . . . . . . . . 47
5.5 Modications to Make/files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6.1 Executing a tutorial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Nota Benes
1.1 bashrc after updates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1 Regarding estimate of overtopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.1 A note on phase focusing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.2 Always specify an external source (if used) . . . . . . . . . . . . . . . . . . . . . . . 41
4.3 Installation of OceanWave3D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Warnings
1.1 Numerical wave gauges does not compile on recent OpenFoam versions (OF and OF+). 10
3.1 Denition of wave gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.2 Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1 Online documentation to create waveDyMFoam. . . . . . . . . . . . . . . . . . . . . . . 45

5.2 Warning concerning the denition of the gravity vector. . . . . . . . . . . . . . . . . 46
7
CHAPTER 1
Introduction
1.1 Origin
The waves2Foam toolbox was originally developed at the Technical University of Denmark by
Niels Gjl Jacobsen under the supervision of Prof. Jrgen Fredse. The porosity module was
developed in collaboration between Bjarne Jensen and Niels Gjl Jacobsen, both then at the
Technical University of Denmark.
As of the employment of Niels G. Jacobsen at Deltares, The Netherlands (2013-present), the
maintenance and further developments takes place at Deltares.
1.2 Background and application

1,2
The waves2Foam toolbox is a plug-in to the open-source general purpose CFD-package OpenFoam
R .
waves2Foam is available as open-source under the GNU General Public License (GNU GPL).
waves2Foam was initially only released as a toolbox for the generation and absorption of free
surface waves (Jacobsen et al., 2012) in November 2011. In 2014, the toolbox was extended with
the possibility of modelling the interaction between free-surface waves and a permeable medium
such as breakwaters, scour protection, etc (Jensen et al., 2014).
The waves2Foam toolbox is used extensively in a large number of references and it ranges
from validation to application of the toolbox and the free surface capabilities of OpenFoam.
Examples of the usage of waves2Foam is given here:
Jacobsen et al. (2014); Jacobsen and Fredse (2014a,b) applied the model to study the
evolution of cross-shore morphological development under breaking waves. This involved
a coupling between the hydrodynamics, sediment transport and the resulting bed level
change.
Paulsen et al. (2014a,b) used the toolbox to study wave loads on a monopile. The work
by Paulsen et al. (2014a) furthermore included a coupling between waves2Foam and a
solution to the full non-linear wave problem, i.e. the Laplace equation with non-linear
boundary conditions; more details are given in Section 4.3.3.
The resonance of the surface elevation within moon pools (Moradi, 2015).
The study of wave loads on bridge decks, where comparions with experimental work were
presented (Seiert, 2014; Seiert et al., 2014; Hayatdavoodi et al., 2014).
Jacobsen et al. (2015) validated waves2Foam against numerous laboratory experiments

with normal incident irregular waves. This showed that the relaxation zone approach for
the generation and absorption in waves2Foam compensates for the additional mass ux
due to the Stokes drift without a increase in the water level.
Stahlmann (2013) coupled the wave generation in waves2Foam with a method for the
deformation of the sediment bed around a tripod steel structure; the purpose was to
evaluate the eects of this structure on scour patterns.
1 This oering is not approved or endorsed by OpenCFD Limited, producer and distributor of the Open-
FOAM software via www.openfoam.com, and owner of the OPENFOAM R and OpenCFD R trade marks.
2 OPENFOAM R is a registered trade mark of OpenCFD Limited, producer and distributor of the Open-
FOAM software via www.openfoam.com.
9
10 Manual v. 0.9 (Draft) - 1 Introduction
Elsafti and Oumeraci (2013) coupled waves2Foam with a geotechnical model to study the
residual pore pressure build-up and dissipation underneath a caisson breakwater.
The list of application is steadily expanding, why the most up-to-date information is to
look at the citing papers on e.g. Google Scholar (https://scholar.google.com) or ResearchGate
(www.researchgate.net).
1.3 Compabitility
1.3.1 Branches
OpenFoam comes in three main branches:
OpenFoam (OF) as distributed through www.openfoam.org.
OpenFoam+ (OF+) as distributed through www.openfoam.com.
foam-extend (FE) as distributed through the foam-extend community.
The abbreviation in the parenteses are used throughout this document.
1.3.2 Naming terminology

waves2Foam is supported on all three branches of OpenFoam and on a long list of version
numbers. Over the years, the various branches have modied the naming terminology. For
instance, the void fraction ratio, F, is termed alpha1 in some versions and alpha.water in
other versions. Similarly, the excess pressure is termed pd in some versions and p_rgh in other
versions. These dierences in naming terminology are supported by waves2Foam through the
class <waves2Foam_src>/waves2Foam/include/crossVersionCompatibility.[C,H].
1.3.3 Supported versions

waves2Foam compiles on the following (recent) versions, however, please refer to Warning 3.1:
OF-3.0, OF-4.0
OF+-3.0, OF+-1606, OF+-1612
FE-3.1, FE-3.2, FE-4.0
Warning 1.1: Numerical wave gauges does not compile on recent OpenFoam ver-
sions (OF and OF+).
The surface sampling utility (numerical wave gauges) is a vital part of modelling
free surface waves with OpenFoam. Due to a code reorganisation, the wave gauges
do not work for OF-4.0 (and newer) and OF+-1612 (and newer).
This issue has to be resolved in a future update of waves2Foam.
1.4 Installation of waves2Foam

waves2Foam is straightforward to install, if the necessary third-party packages are already avail-
able on the computer. This section details the installation procedure.
1.4.1 Third-party dependencies

The following third-party dependencies are required to install waves2Foam. These dependencies
are needed in addition to those required to compile OpenFoam itself.
gfortran is needed to compile OceanWave3D and Fenton's code for the stream function
coecients.
git is required to access the source code for OceanWave3D.
subversion (SVN) is required to access the source code for waves2Foam.
GNU Scientic Library (GSL) is required to compile waves2Foam. GSL provides various
mathematical functions not natively supported in OpenFoam.
Manual v. 0.9 (Draft) - 1.4 Installation of waves2Foam 11
Please refer to the package management system on your avour of Linux/UNIX for the instal-
lation of these third-party packages. Compilation of waves2Foam is unsuccessful, if any of these
packages are missing.
1.4.2 Download and installation

The download and installation instructions are for a default installation location (Code Fragment
1.1). If the installation is in a non-default location, please refer to Section 1.4.3 for additional
actions before $ ./Allwmake.
Code Fragment 1.1: Download and installation instructions for the default in-
stallation directory ($ refer to a command).
## First, source your OpenFoam version

$ mkdir -p $FOAM_RUN/../applications/utilities
$ cd $FOAM_RUN/../applications/utilities
svn co http://svn.code.sf.net/p/openfoam-extend/svn/trunk/\
Breeder_1.6/other/waves2Foam
$ cd waves2Foam
$ ./Allwmake
1.4.3 The bashrc le

waves2Foam is distributed with a control le called bashrc.org, which is located in the directory
bin. In case of a default installation, this le is automatically copied to bin/bashrc. The
le controls the linking to GNU Scientic Library and the target directories for libraries and
executables.
If a non-standard installation is required, the following set of user-dened variables may be
modied:
WAVES_DIR: This is the full path to the waves2Foam-installation.
WAVES_APPBIN: This is the location for the executables. The default value is the user-
directory ($FOAM_USER_APPBIN), though an alternative is $FOAM_APPBIN, which is useful
for cluster installations with multiple users.
WAVES_LIBBIN: This is the location for the libraries. The default value is the user-directory
($FOAM_USER_LIBBIN), though an alternative is $FOAM_LIBBIN, which is useful for cluster
installations with multiple users.
WAVES_GSL_INCLUDE: Header les for GNU Scientic library (gsl is appended during com-
pilation).
WAVES_GSL_LIB: Path to the GNU Scientic libraries.
It is recommended never to modify bashrc.org. Always make a copy and make modi-
cations in this le. Modications to bashrc.org could cause problems, when updating the
SVN-repository.
1.4.4 How to get the updates to waves2Foam?

Updates to waves2Foam are obtained by updating the SVN-repository, followed by recompila-
tion. This is outlined in Code Fragment 1.2.
Code Fragment 1.2: Updating the SVN-repository and recompilation of

waves2Foam.
$ svn update
$ ./Allwmake
Updates to the SVN-repository are normally announced through CFD-Online in the thread
called Release of a Wave Generation and Absorption Toolbox for OF. The easiest is to subscribe
to the thread and receives email-notications.
All revisions of waves2Foam are detailed in Appendix C.
12 Manual v. 0.9 (Draft) - 1 Introduction
Nota Bene 1.1: bashrc after updates

Please note that the le bashrc.org might get updated with new versions of
waves2Foam. If this happens, it is recommended to use the most up-to-date version
of bashrc.org, see instructions in Section 1.4.3.
1.4.5 How to recompile parts of waves2Foam

Sometimes it is needed to recompile parts of the code, if modications or additions have been
made. It is not possible to only recompile parts of the code, due to the temporary denition of
environmental variables by the inclusion of bin/bashrc. Therefore, waves2Foam should always
be compiled using the Allwmake script.
1.5 Referencing
The users are reguested to support the work put into the waves2Foam toolbox by providing
proper referencing to the academic work that has been supporting the various releases:
Any use of waves2Foam: Please make reference to the original work (Jacobsen et al., 2012)
as specied in Code Fragment 1.3.
Code Fragment 1.3: Reference to the original paper accompanying the release
of waves2Foam.
@article{jacobsenFuhrmanFredsoe2012,
Author = {Jacobsen, N G and Fuhrman, D R and Freds\o{}e, J},
Title = {{A Wave Generation Toolbox for the Open-Source CFD Library:
OpenFoam\textregistered}},
Journal = {{International Journal for Numerical Methods in Fluids}},
Year = {{2012}},
Volume = {{70}},
Number = {{9}},
Pages = {{1073-1088}},
DOI = {{10.1002/fld.2726}},
}
Use of the porosity module: Please make reference to the derivation of the Navier-Stokes
equations in terms of lter velocities for permeable structures (Jensen et al., 2014) as specied
in Code Fragment 1.4.
Code Fragment 1.4: Reference to the paper on the porosity implementation in

waves2Foam.
@article{jensenJacobsenChristensen2014,
Author = {Jensen, B and Jacobsen, N G and Christensen, E D},
Title = {{Investigations on the porous media equations and resistance
coefficients for coastal structures}},
Journal = {{Coastal Engineering}},
Year = {{2014}},
Volume = {{84}},
Pages = {{56-72}},
}
Coupling with OceanWave3D: When the coupling between waves2Foam and OceanWave3D
is used, please make reference to Paulsen et al. (2014a) for the coupling (Code Fragment 1.5) and
reference to Engsig-Karup et al. (2009) for the development of OceanWave3D (Code Fragment
1.6).
The coupling was later streamlined into the waves2Foam framework and it ts into the frame-
work as an external source, see Section 4.2.8).
Manual v. 0.9 (Draft) - 1.5 Referencing 13
Code Fragment 1.5: Reference to the coupling between waves2Foam and Ocean-
Wave3D.
@article{ paulsenBredmoseBingham2014,
Author = {Paulsen, B. T. and Bredmose, H. and Bingham, H. B.},
Title = {{An efficient domain decomposition strategy for wave loads on
surface piercing circular cylinders}},
Journal = {{Coastal Engineering}},
Year = {{2014}},
Volume = {{86}},
Pages = {{57-76}},
}
Code Fragment 1.6: Reference to OceanWave3D.

@article{ engsigKarupBinghamLindberg2009,
Author = {Engsig-Karup, A. P. and Bingham, H. B. and Lindberg, O.},
Title = {{An efficient flexible-order model for 3D nonlinear water waves}},
Journal = {{Journal of Computational Physics}},
Year = {{2009}},
Volume = {{228}},
Number = {{6}},
Pages = {{2100-2118}},
DOI = {{10.1016/j.jcp.2008.11.028}},
}
CHAPTER 2
Mathematical Description
2.1 Coordinate system
The coordinate system is dened by the direction of the gravity vector given in the le constant/g.
This means that the vertical coordinate can be any direction; theoretically. Only denition of
the gravity vector along the y- or z -axis is extensively tested.
2.2 Momentum and continuity equations

The momentum and continuity equations follow the standard implementation in OpenFoam for
Volume of Fluid (VOF) type simulations. The momentum equation is written in the excess
pressure that is dened as the pressure in excess of the hydrostatic pressure. The momentum
equation reads:
u
+ uuT = p + g (x xr ) + tot u (2.1)
t
Here, is the density of the uid, u is the velocity vector in Cartesian coordinates, t is time,
= (/x, /y, /z) is the dierential operator, p is the excess pressure, g is the vector
of acceleration due to gravity, x = (x, y, z) is the Cartesian coordinate vector, xr is a reference
location (dened at sea level, see Section 3.1.2) and tot is the total dynamic viscosity, see
Section 2.4.
The incompressible continuity equation reads:
u=0 (2.2)

The excess pressure is dened as p = p g x, where p is the total pressure. The solution
procedure for the discretised equations are outlined in Rusche (2002).
2.2.1 Modications for permeable structures

In the case of permeable structure, such as breakwaters, scour protection, a permeable sea
bottom, etc, the velocity is dened as the lter velocity. The lter velocity is related to the pore
velocity through the porosity, n, as follows:
u = nup (2.3)
up is the pore velocity. The derivation of the implementation of the correction due to permeable
layers is presented in Jensen et al. (2014). The continuity equation is the same as in Eq. (2.2),
while the momentum equation takes the following modied version:
u 1 1
(1 + Cm ) + uuT = p + g (x xr ) + tot u Fp (2.4)
t n n n n
This form of the momentum equation deviates from the one presented in Higuera et al. (2014),
however, in a later work the main author (Higuera, 2015) aligned himself with the formulation
in Eq. (2.4). The quantities Cm and Fp are discussed in Section 2.8. The continuity equation
is still given by Eq. (2.2).
This modelling framework has successfully been applied to study the interaction between waves
and permeable structure (Jensen et al., 2014; Jacobsen et al., 2015; Van Gent et al., 2015). These
works have validated quantities such as reection, wave dissipation, wave overtopping and wave
induced pressures and setup inside permeable structures.
15
16 Manual v. 0.9 (Draft) - 2 Mathematical Description
2.3 Advection of the indicator eld

The advection of the VOF-eld, F, is done as follows:
F
+ uF + ur (1 F )F = 0 (2.5)
t
Here, ur is a relevative velocity, see e.g. Berberovi et al. (2009) for details. F is used to evaluate
the spatial variation in density and viscosity as follows:
= F 1 + (1 F )0 and = F 1 + (1 F )0 (2.6)
Subscript 0 refers to material properties corresponding to F = 0 and subscript 1 refers to

material properties corresponding to F = 1. It is assumed thatF = 1 for water and F = 0 for
air in waves2Foam.
2.3.1 Modications for permeable structures

The presence of a permeable structure means that the amount of uid in a cell is limited to the
voids between the permeable material. It is still desirable to have F [0, 1], why the advection
equation eectively takes this work:
F 1
+ ( uF + ur (1 F )F ) = 0 (2.7)
t n
n is the porosity.
2.4 Turbulence modelling

No turbulence models are distributed with waves2Foam. There are plenty of turbulence models
available in the various distributions of OpenFoam, however, please be aware of recommendations
given in the literature (some examples are Mayer and Madsen, 2000; Jacobsen et al., 2012; Brown
et al., 2016; Zhou et al., 2017; Devolder et al., 2017)
2.5 Reection compensation at the boundary

Techniques are known to perform reection compensation directly at the boundary (Wellens,
2012; Higuera, 2015), though such methods are not available for waves2Foam, which is developed
around the use of relaxation zones.
2.6 Numerical beach

waves2Foam is prepared for the implementation of a numerical beach. This implementation has
never been completed. Some tutorials contains a keyword beachType in the dention of the
relaxation zone, though a default value is assigned, why this keyword is not required.
2.7 Relaxation zone technique

The relaxation zone technique is meant to remove spurious reection from numerical simulations.
The relaxation zone technique is based on a weighting between the computed solution of the
velocity eld and the indicator eld with a target solution. The relaxation zone technique is
divided into two avours, namely explicit and implicit relaxation, where explicit/implicit refers
to the time integration.
Furthermore, the specication of the weighting and location of the relaxation zones in the
computational domain are described.
2.7.1 Flavour of relaxation

2.7.1.1 Explicit relaxation
The explicit approach in the relaxation zones is simply given as:
= (1 wR )target + wR computed (2.8)
Here, the weighting function wR [0, 1] can be dened in various ways as will be described in
the following sections.
Manual v. 0.9 (Draft) - 2.7 Relaxation zone technique 17
This method explicitly corrects the elds and u according to Eq. (2.8) each time step
prior to the solution to the pressure-velocity coupling. The method was originally described in
Jacobsen et al. (2012) for OpenFoam, though it is a fairly common approach for e.g. Boussinesq
type wave modelling or solution to the full Laplace problem (see e.g. Fuhrman et al., 2006;
Engsig-Karup et al., 2009).
2.7.1.2 Implicit relaxation

An implicit relaxation zone implementation has been proposed by Vukcevi et al. (2016a,b),
though the method is currently not available in waves2Foam.
2.7.2 Specifying the relaxation zone

The relaxation zone is specied in the le waveProperties.input inside a sub-dictionary, see
Section 3.1.2 for more details. The relaxation zone is dened as in Code Fragment 2.1. Each of
these controls are described individually in the following sections.
Code Fragment 2.1: Fundamental denition of a relaxation zone

relaxationZone
{
relaxationScheme <word>;
relaxationWeight <word>;
relaxationShape <word>;
courantCorrection <bool>; // Default: false
}
2.7.2.1 Relaxation scheme

Relaxation scheme is essentially the way, the relaxation is implemented.
relaxationSchemeEmpty This method eectively turns of the relaxation (Code Fragment

2.2).
Code Fragment 2.2: Empty relaxation scheme

relaxationZone
{
relaxationScheme Empty;
}
relaxationSchemeSpatial The spatial relaxation scheme performs the relaxation according

to Eq. (2.8) (Code Fragment 2.3).
Code Fragment 2.3: Spatial relaxation scheme

relaxationZone
{
relaxationScheme Empty;
}
2.7.2.2 Relaxation weight

There are three relaxation weights available and each of these can be corrected for the global
Courant number. The following weights are available:
Exponential weight (default)
Free polynomial weight
Third order polynomial weight

These methods produce a weight, wR , which can be corrected based on the local Courant
number as follows:
Co/Comax
wR = 1 (1 wR ) (2.9)

where wR = 1 wR and wR is used in Eq. (2.8). Here, Co is the local Courant number and
Comax is the maximum Courant number. The method is implemented following the work by
Seng (2012). The eect is activated by adding the keyword courantCorrection in the denition
of the relaxation zone, see Code Fragment 2.1.
In all cases, the weights are a function of a local coordinate system in the relaxation zone,
such thatwR ( = 1) = 0 and wR ( = 0) = 1. Here, is a local coordinate within the relaxation
zone, where the coordinate depends on the shape of the relaxation zone.
Exponential weight The exponential weight distribution is taken from Fuhrman et al. (2006)
and it has the following form:
exp p 1
wR = 1 (2.10)
exp 1 1
The exponent p is set to 3.5 as default. The method is default and there is no need to specify
it, but it can be explicitly included as in Code Fragment 2.4:
Code Fragment 2.4: Exponential weight function

relaxationWeight Exponential; // Default
exponent <scalar>; // Default value = 3.5
Free polynomial weight The free polynomial weight is encountered in Engsig-Karup (2006)
and the weight function is given as:
wR = 1 p (2.11)
where p is an integer exponent. The weight can be specied as in Code Fragment 2.5. There is
no default value for the exponent p.
Code Fragment 2.5: Free polynomial weight function

relaxationWeight FreePolynomial;
exponent <scalar>;
Third-order polynomial weight The third order polynomial weight is encountered in Engsig-
Karup (2006) and the weight function is given as:
wR = 2 3 + 3 2 (2.12)
where = 1 . The weight can be specied as in Code Fragment 2.6
Code Fragment 2.6: Third order polynomial weight function

relaxationWeight ThirdOrderPolynomial;
2.7.2.3 Relaxation shape

The relaxation shape species, where in the computational domain a certain relaxation zone is
applied. There are currently four dierent relaxation shapes available. These are:
Rectangular;
Semi-cylindrical;
Cylindrical;
Manual v. 0.9 (Draft) - 2.7 Relaxation zone technique 19
Frozen
All of the relaxation zones update the aected cell indices every time step, if the mesh is
changing (moving or mesh renement). The exception to this is the frozen relaxation zone, see
below. The settings for each of these are as follows. For all case, the vertical axis is dened
based on the direction of gravity.
Rectangular The rectangular shape is the most useful shape, because it works equally well
for 2D and 3D simulations. It is dened based on the coordinates of the two diagonal corner
points of a rectangle and the direction of one of the lateral sides (and direction of relaxation).
The formatting of the relaxation shape is given in Code Fragment 2.7
Code Fragment 2.7: Rectangular relaxation shape

relaxationShape Rectangular;
relaxType <word>;
startX <point>;
endX <point>;
orientation <vector>;
The keyword relaxType takes either of two words, namelyINLET and OUTLET. The former is
used to dened that the direction of the vector orientation describes the direction of relaxation,
whereas the use of OUTLET means that the direction of relaxation is opposite to that prescribed
by orientation, see more in Section 3.1.3.
All points must be in the same horizontal plane.
Semi-cylindrical A semi-cylindrical shape can be dened based on the specications in Code

Fragment 2.8. This method is perhaps not particularly useful, but it shows an example of how
alternative shapes may be dened.
Code Fragment 2.8: Semi-cylindrical relaxation shape

relaxationShape SemiCylindrical;
centre <point>;
rInner <scalar>;
rOuter <scalar>;
zeroAngleDirection <vector>;
angleStart <scalar>; // [degrees]
angleEnd <scalar>; // [degrees]
Cylindrical The cylindrical relaxation shape dened a relaxation zone with an inner and
outer radious (Code Fragment 2.9). The target solution is enforced at the outer edge of the
cylindrical relaxation zone. Until now, the only usage of this relaxation zone appears to be the
work by Arrighi et al. (2015).
Code Fragment 2.9: Cylindrical relaxation shape

relaxationShape Cylindrical;
centre <point>;
rInner <scalar>;
rOuter <scalar>;
Frozen The frozen relaxation shape is special in the way that is mimics any of the other
relaxation shapes, except for one aspect: It does not update the cell indices for the relaxation
zone, if the computational mesh is moving. The method is not applicable on meshes with
topological changes. The relaxation shape is dened as given in Code Fragment 2.10
Code Fragment 2.10: Frozen relaxation shape

relaxationShape Frozen;
actualRelaxationShape <word>;
// Additional lines for the actual relaxation shape
2.8 Specication of resistance due to a permeable structure

The resistance force in waves2Foam is of the form (Eq. (2.4)):
Fp
= au + bukuk2 (2.13)

where a and b are resistance coecients. In additional to this, the added mass coecient Cm is
given as follows:
1n
Cm = p (2.14)
n
Here, p is a closure coecient is taken as 0.34.
2.8.1 Resistance formulations

There are three resistance formulations implemented in waves2Foam. The required coecients
for each of these are detailed in the following subsections.
2.8.1.1 Native OpenFoam

The native formulation reads two parameters d and f (Code Fragment 2.11).
Code Fragment 2.11: The native relaxation zone

d d [0 -2 0 0 0 0 0] <vector>;
f f [0 -1 0 0 0 0 0] <vector>;
porosity <scalar>;
gammaAddedMass <scalar>;
2.8.1.2 Engelund (1953)

The resistance coecients a and b take the following form in the formulation by Engelund (1953):
(1 n)3 1n 1
a= , b= (2.15)
n2 d250 n3 d50
Here, is the kinematic viscosity and d50 is the nominal, median diameter of the grain material.
The input parameters are specied in Code Fragment 2.12.
Code Fragment 2.12: Resistance formulation according to Engelund (1953)

resistanceFormulation engelund1953;
porosity <scalar>;
d50 d50 [0 1 0 0 0 0 0] <scalar>;

alpha alpha [0 0 0 0 0 0 0] <scalar>;
beta beta [0 0 0 0 0 0 0] <scalar>;
Manual v. 0.9 (Draft) - 2.8 Specication of resistance due to a permeable structure 21
2.8.1.3 Van Gent (1995)

The resistance coecients a and b take the following form in the formulation by Van Gent
(1995):
(1 n)2

7.5 1 n 1
a= , b= 1+ (2.16)
n3 d250 KC n3 d50
Here, is the kinematic viscosity and d50 is the nominal diameter of the grain material. The
input parameters are specied in Code Fragment 2.13.
Code Fragment 2.13: Resistance formulation according to Van Gent (1995)

resistanceFormulation vanGent1995;
porosity <scalar>;
d50 d50 [0 1 0 0 0 0 0] <scalar>;

alpha alpha [0 0 0 0 0 0 0] <scalar>;
beta beta [0 0 0 0 0 0 0] <scalar>;
KC KC [0 0 0 0 0 0 0] <scalar>; // Default: 10000
The inserted values are merely examples. For choice of , and KC , see e.g. Jensen et al.
(2014); Jacobsen et al. (2015); Losada et al. (2016). Furthermore, please note that the eect of
the KC number is being discussed in Jacobsen et al. (In preparation).
CHAPTER 3
Utilities
This chapter discusses the range of pre- and post-processing utilities.
3.1 Pre-processing utilities

3.1.1 waveGaugesNProbes
The utility allows for the denition of probes and wave gauges. The denition is outlined in the
following.
3.1.1.1 Input:
The le constant/probeDefinitions is used to specify the location of the wave gauges and
probes. The common settings are dened in Code Fragment 3.1.
Code Fragment 3.1: Common settings for denition of wave gauges and probes
<name of wave gauges>
{
type waveGauge; // Definition of the type
add (0 0 1); // Length and orientation of gauge

axis z; // Vertical axis
// Insert point distribution

}
<name of probes>
{
type probeGauge;
outputInterval <label>; // The frequency of outputting

fields (<name0> <name1>); // The fields to sample
// Insert point distribution

}
The point distributions to be provided (Code Fragment 3.1) are dened in the following.
Warning 3.1: Denition of wave gauges

All wave gauges must be dened, such that they are at least partly inside of the
computational domain. No checks are made on the location of the wave gauges
relative to the computational domain and the computation will crash without any
explanation, if one or more wave gauges are completely outside of the computational
domain.
23
24 Manual v. 0.9 (Draft) - 3 Utilities
circularDistribution Denes a circular distribution with a certain radius around a given

centre. The point distribution is dened as in Code Fragment 3.2.
Code Fragment 3.2: A circular point distribution

pointDistribution circularDistribution;
N 4;
centre (0 0 0);
radius 1;
lineDistribution Denes a linear distribution along a line dened by a starting and an end
point. It is possible to have the distribution stretched along the line; resulting in a non-
equidistant point distribution (Code Fragment 3.3).
Code Fragment 3.3: A linear point distribution

pointDistribution lineDitribution;
N 4;
linestart (0 0 0);
lineend (1 0 0);
stretch 1.;
quadrilateralDistribution Denes a quadrilateral point distribution, i.e. it is possible to

dene a rectangular distribution of the wave gauges with e.g. 33 = 9 wave gauges (Code
Fragment 3.4).
Code Fragment 3.4: A quadrilaterial point distribution

pointDistribution quadrilateralDistribution;
N0 10;
N1 11;
linestart0 (0 0 0);
lineend0 (1 0 0);
lineend1 (1 1 0);
stretch0 0.9; // Default = 1.0
stretch1 1.1; // Default = 1.0
userDenedDistribution The user dened distribution is conveniently used, when the wave
gauges should be distributed according to locations in a physical experiment (Code Fragment
3.5).
Code Fragment 3.5: User-dened point distribution

pointDistribution userDefinedDistribution;
N 4;
xValues nonuniform List<scalar> 4(0 1 2 3);
yValues nonuniform List<scalar> 4(0.1 0.2 0.3 0.4);
zValues uniform 0;
All of the locations (xValues, yValues and zValues) can be dened according to either
nonuniform or uniform formatting.
3.1.1.2 Output:
There are two (probe gauges) or four (wave gauges) outputs from waveGaugesNProbes per sub-
dictionary in constant/probeDefinitions. All output are written to the folder 'waveGauges-
NProbes' in the case folder. The les are detailed here:
Manual v. 0.9 (Draft) - 3.1 Pre-processing utilities 25
<name of set>_controlDict: This le contains the part that should be copied to the
controlDict for run-time sampling with the wave gauges, see e.g. Section 3.1.1.3 for a
description.
<name of set>_sets (only wave gauges): This le contains the denition of the wave
gauges, which is essentially the denition of vertical lines in the standard format for the
'sample' utility in OpenFoam.
<name of set>surfaceElevationDict (only wave gauges): This le is to be used for sam-
pling of the wave gauges as a post-processing step. Simply copy the le to the system
folder and execute the command surfaceElevation (Section 3.3.1).
<name of set>.vtk: This le can be opened with ParaView/paraFoam in order to visu-
alise the the location of the probe and wave gauges.
3.1.1.3 Run-time sampling

The run-time sampling is exemplied with the wave gauges. Similar approach holds for the
probe gauges.
The surface elevation can be sampled with the pre-dened wave gauges as either a post-
processing step or as a run-time evaluation. The latter is discussed here and the former is
described in Section 3.3.1. A part of the output from waveGaugesNProbes is a le called <name
of set>_controlDict (Code Fragment 3.6)
Code Fragment 3.6: Example of the control of the surface elevation sampling.
surfaceElevation
{
type surfaceElevation;
functionObjectLibs ( "libwaves2Foam.so" );
outputControl timeStep; // Alternative: outputTime

outputInterval 1;
//Additional output controls in waves2Foam

//samplingStartTime -1;
//surfaceSampleDeltaT 0.025;
setFormat raw;
interpolationScheme cellPointFace;
fields (alpha1);
#includeIfPresent "../waveGaugesNProbes/surfaceElevationAnyName_sets";
}
Note that the name of the indicator eld alpha1 is automatically adjusted to the default
naming a given version of OpenFoam.
The ordinary outputControls work for this utility, but it is also possible to sample the surface
elevation at an approximately equidistant time step, which is much smaller than the output
time of the elds (see commented lines in Code Fragment 3.6). Note, irrespectively of the value
of surfaceSampleDeltaT, the sampling utility is only visited based on the standard OpenFoam
controls (outputControl and outputInterval). It is recommended to use the default settings.
3.1.2 setWaveParameters
The utility setWaveParameters is a pre-processing utility, which computes all the necessary
wave parameters based on physical meaningful properties, e.g. setWaveParametes converts
information on water depth and wave period into a wave number for rst order Stokes wave
theory.
All input is written to the le <casePath>/constant/waveProperties.input and the pro-
cessed data is outputted in <casePath>/constant/waveProperties. This sub-division into two
les comes from historical reasons.
The input le waveProperties.input contains global information and information related
to each wave boundary and/or relaxation zone. The global parameters and the setup for a
boundary condition and/or relaxation zone is given in the following two sections.
Code Fragment 3.7: Globale settings in waveProperties.input

// Only relevant for porousWaveFoam
porosityModel jensenJacobsenChristensen2014;
// Global definition of still water level

seaLevel <scalar>;
sealevelAsReference <boolean>;
// Names of relaxation zones. The wave theories are defined in the

// subdictionaries name0Coeffs, name1Coeffs, etc.
relaxationNames (name0 name1 ...);
// Wave theory used to initiate the simulation.

initializationName outlet;
// Not necessary. Default value is 'emptyExternal'.

externalForcing emptyExternal;
3.1.2.1 Global parameters

The global settings are specied in Code Fragment 3.1.2.1. The number of relaxation zones
are dened through the list relaxationNames. It also means that a relaxation zone can be
dened in the sub-dictionaries, e.g. name0Coeffs, but if the string name0 is not written in the
list relaxationNames, then the relaxation zone is not active.
The external forcing is an option to provide the simulation with external wave data, see Section
4.3 for adescription of the options.
3.1.2.2 Sub-dictionary information

Each set of wave properties are given in a sub-dictionary, which is called <name> + "Coeffs".
An example of the structure of such a sub-dictionary is given in Code Fragment 3.8 for the
relaxation zone identied by name0:
Code Fragment 3.8: Overview of the sub-dictionary in waveProperties.input.

name0Coeffs
{
// Insert all required data for any given wave type
// Insert all required data for any given relaxation zone definition
}
The details on wave types can be found in Section 4. The details on relaxation zones and
their denition can be found in Section 2.7.
3.1.2.3 Example
An example of a computational domain with two relaxation zones is shown in Code Fragment
3.9. The leftmost relaxation zone is used to generate a second order wave and the rightmost
relaxation zone is used to absorp the wave.
The applied solver is waveFoam, why the keywork porosityModel is commented.
3.1.3 relaxationZoneLayout
The utility relaxationZoneLayout can be used to study the layout of the relaxation zone. The
utility outputs three elds to the 0-folder:
relaxationZoneLayout: This eld shows the spatial extend of the relaxation zones. The
value of 1 is outside of any relaxation zone. The rst relaxation zone is given a value of
0, the second a value of 1, etc.
relaxationZoneSigmaValue: This is the spatial distribution of the -weight for the relax-
ation zones.
relaxationZoneWeightOnComputed: The shows the resulting weight that is applied to the

computed eld, i.e. a value of 1 means that the computed eld is not altered.
Manual v. 0.9 (Draft) - 3.1 Pre-processing utilities 27
Code Fragment 3.9: Example of waveProperties.input.

FoamFile
{
version 2.0;
format ascii;
class dictionary;
object waveProperties.input;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
//porosityModel jensenJacobsenChristensen2014;
seaLevel 0.00;
relaxationNames (inlet outlet);
initializationName outlet;
inletCoeffs
{
waveType stokesSecond;
height 0.1;
period 3.0;
depth 1.0;
phi 0.0;
direction (1.0 0.0 0.0);
debug off;
Tsoft 2.0;
relaxationZone
{
relaxationScheme Spatial;
relaxType INLET;
startX (-15.0 0.0 -1.0);
endX (-10.0 0.0 1.0);
orientation ( 1.0 0.0 0.0);
}
};
outletCoeffs
{
waveType potentialCurrent;
U (0 0 0);
Tsoft 2;
relaxationZone
{
relaxationScheme Spatial;
relaxType OUTLET;
startX ( 10.0 0.0 -1.0);
endX ( 15.0 0.0 1.0);
orientation ( 1.0 0.0 0.0);
}
};
3.1.4 setWaveField
The utility setWaveField is used to set the initial conditions according to a user dened wave the-
ory. The latter is dened by the keyword initializationName in the le waveProperties.input,
see Section 3.1.2.1. This can be any wave theory, but if the ramping time diers from 0, then
the prescribed initial condition will be a horizontal free surface at still water level and the water
will be stagnant. If this is the case, the warning in Code Fragment 3.10 is written to standard
output:
Code Fragment 3.10: Warning when Tsoft diers from 0

--> FOAM Warning :
From function setWaveField::setWaveField(const fvMesh& mesh, \
volVectorField& U, volScalarField& alpha, volScalarField& p)
in file preProcessing/setWaveField/setWaveField.C at line 77
The specified value of Tsoft is non-zero in the waveType: \

`stokesFirst'
specified in the sub-dictionary waveProperties::inletCoeffs
Consequently, the initialised `wave field' is set to a horizontal \

free surface with zero velocity.
3.1.5 faceSetToSTL
The modelling of permeable structures, such as breakwaters or open lters, requires that the user
denes the location of the permeable structure is located. There are several ways of identifying
the cell indices, one of these is the usage of (non-convex) STL-surfaces and the cellSet (in
foam-extend) or topoSet (in OpenFoam) utilities. An STL-surface is essentially a triangulation
of a closed surface. The utility faceSetToSTL is a tool that creates such a STL-surface. The
tool is only easily applicable for simple structures.
The input to the case is the le stlDefinitions which is placed in <casePath>/constant/-
triSurface. There is an example located in <waves2Foam>/applications/utilities/preProcessing/-
faceSetToSTL.
There are two ways of dening the STL-surface, namely by dening all the faces of the sur-
face in a manner similar to blockMesh or by extruding one face along a given direction (Code
Fragment 3.11).
Code Fragment 3.11: Two methods to generate an STL-surface of a unit square.

FoamFile
{
version 2.0;
format ascii;
class dictionary;
object stlDefinitions;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
unit1
{
points 4((0 0 0) (1 0 0) (1 1 0) (0 1 0));
faces 1((0 1 2 3));
extrude true;
extrudeVector (0 0 1);
}
unit2
{
points 8((0 0 0) (1 0 0) (1 1 0) (0 1 0)
(0 0 1) (1 0 1) (1 1 1) (0 1 1));
faces 6((0 3 2 1) (4 5 6 7) (2 3 7 6)
(0 1 5 4) (1 2 6 5) (4 7 3 0));
}
The STL-surfaces are written to the folder, where stlDefinitions is placed.
3.1.6 sampleIncidentWaveField
The utility sampleIncidentWaveField allows for a fast evaluation of the incident wave eld
based on its algebraic form. The code snippet shown in Code Fragment 3.12 is inserted into
waveProperties.input, followed by execution of the commands shown in Code Fragment 3.13.
The output is an evaluation of the theoretical form of the incident wave eld evaluated in the
specied points. The surface elevation is evaluated for all active relaxation zones. The data is
Manual v. 0.9 (Draft) - 3.2 Run-time sampling 29
Code Fragment 3.12: Instructions for sampleIncidentWaveField into

waveProperties.input
sampleIncidentWaveField
{
deltaT <scalar>; // [s]
endTime <scalar>; // [s]
points N((x1 y1 z1) (x2 y2 z2) ... (xN yN zN));

}
outputted in the folder syntheticWaveField and the formatting is identical to the formatting
from the regular surface elevation tool.
Code Fragment 3.13: Use of sampleIncidentWaveField

$ setWaveField
$ sampleIncidentWaveField
3.2 Run-time sampling

3.2.1 Numerical wave gauges (surfaceElevation)
For run-time sampling of the surface elevation, see Section 3.1.1.3.
3.2.2 Overtopping
The process of overtopping is the amount of water that is overtopping a structure. Overtopping
at impermeable and permeable structures is studied with waves2Foam by Jensen et al. (2014).
This functionality should be used in run-time, because the process of overtopping is very rapid,
so it is very unlikely that the process is captured at a post-processing step. Consequently, it is
recommended to evaluate the overtopping every single time step.
The following sections describe the implementation, the usage and the output.
3.2.2.1 Implementation
During the simulation the following types of face uxes are available:
3
in [m /s] is the ux of uid across a face;
in [kg/s] is the ux of uid across a face multiplied the density of the uid;
3
F in [m /s] is the ux of uid cross a face multiplied with the indicator function.
While a combination of and F would be perfect to evaluate the ux of water across a face,
this is not possible, because F is not available throughout the entire time step; consequently, it
is not available when the function objects are evaluated. Therefore, the ux of water is estimated
with the use of and instead.
In the solution to the advection of the indicator function, the following relationship is used:
= (F =1 F =0 )F + F =0 (3.1)
Consequently, F can be estimated as follows
F =0
F = (3.2)
F =1 F =0
Knowing the ux of water (assuming that the uid is water, when F = 1), if is now possible
to evaluate the overtopping over a set of faces, F:
X Sf
q= F,f (3.3)
kSf k2
f F
where q is the volume ux in [m3 /s] and Sf is the non-unit normal vector to the face. Here,
F is positve in the direction of the normal vector and negative in the opposite direction, so the
combination gives the directional overtopping over a set of faces.
Nota Bene 3.1: Regarding estimate of overtopping

It is stressed that F is only estimated, because and are not based on the
same volume ux at the time of evaluating F for the overtopping.
3.2.2.2 Usage
The function-object denition in Code Fragment 3.14 must be inserted in the controlDict.
Code Fragment 3.14: Run-time sampling of overtopping

overtopping
{
type overtopping;
functionObjectLibs ("libwaves2FoamSampling.so");
outputControl timeStep;
outputInterval 1;
}
There is no explicit statement on which faceZones to operate, because the functionality

operates and all faceZones called overtopping*, i.e. the name merely needs to begin with
overtopping.
The face sets are created with the utility faceSet for older versions of OpenFoam and (cur-
rently) all versions of the foam-extend branch. For newer versions of OpenFoam the utility
topoSet should be used to create face sets. The face sets are subsequently turned into zones
(Code Fragment 3.15).
Code Fragment 3.15: Face zone for overtopping

> setsToZones -noFlipMap
The reader is referred to the templates for faceSetDict and topoSetDict that are available
with the distribution of the source code on how to use faceSet and topoSet respectively.
3.2.2.3 Output
The overtopping is outputting in the le <casePath>//overtopping/<startTimeName>/overtopping.dat.
The use of <startTimeName> in the location means that it is possible to restart a simulation
without loosing data from the rst part of a simulation.
The formatting of overtopping.dat is given in Code Fragment 3.16.
Code Fragment 3.16: Formatting of the overtopping data

Time: <faceZoneName_0> <faceZoneName_1> ... <faceZoneName_N>
t0 <vector> <vector> ... <vector>
t1 <vector> <vector> ... <vector>
... ... ... ... ...
tM <vector> <vector> ... <vector>
Here, the quantity t0 is the time instance and <vector> is the instantaneous overtopping
volume q, i.e. it is not an accumulated quantity.
3.3 Post-processing utilities

3.3.1 Numerical wave gauges (surfaceElevation)
Discrepancies have been observed between run-time and post-processing evaluation of the surface
elevation with the surfaceElevation utility. The run-time results appear to be the correct
values, consequently, it is discouraged to use surfaceElevation as a post-processing utility.
Manual v. 0.9 (Draft) - 3.3 Post-processing utilities 31
3.3.2 postProcessWaves2Foam
The utility postProcessWaves2Foam is intended for easy post-processing of the run-time sampled
properties such as the surface elevation, forces and overtopping.
postProcessWaves2Foam consists of three overall elements, namely (i) read the raw data
streams, (ii) interpolate the data stream to an equidistant time axis and (iii) process the data.
The interpolation is needed for some of the post-processing elements such as the decomposition
into incident and reected wave components for irregular waves, which is essentially based on a
FFT.
Warning 3.2: Post-processing

The output format of probes, forces and moments changes signicantly between
individual versions of OpenFoam. Consequently, it is recommended only to apply
postProcessWaves2Foam for the processing of output les native to waves2Foam
(surface elevastion and overtopping).
The reading of other sources will be removed in future versions of waves2Foam.
The following processing types are currently available:
Reading raw data
rawSurfaceElevation reads the surface elevation signal on a - potentially - non-

equidistant time axis.
rawAlphaProbes reads the data from probes of the void-fraction ratio on a - poten-
tially - non-equidistant time axis.
rawVelocityProbes reads the data from probes of the velocity on a - potentially -

non-equidistant time axis.
rawOvertopping reads the overtopping data on a - potentially - non-equidistant time

axis.
rawForcesAndMoments reads the data on forces and moments on a - potentially -

non-equidistant time axis.
Reading interpolated data
interpolateSurfaceElevation reads and interpolates the surface elevation signal

onto an equidistant time axis.
interpolateAlphaProbes reads and interpolates probes of the void-fraction ratio

onto an equidistant time axis.
interpolateVelocityProbes reads and interpolates probes of the velocity onto an

equidistant time axis.
interpolateOvertopping reads and interpolates the overtopping data onto an equidis-

tant time axis.
interpolateForcesAndMoments reads and interpolates the forces and moments onto

an equidistant time axis.
Data handling
write2Ascii writes the data stream to an easily accessible ASCII-format.

writeIndexLocation writes the index location of the surface elevation data.
removeData removes data from intermediate steps
removeSubDirs removes subdirectories from processing steps.
Data processing
trapz performs a trapezoidal integration in time.

cumTrapz performs a cummulative trapezoidal integration in time.
zeroCrossing performs a zero-crossing analysis of a given time series.
ensembleAverage performs an ensemble averaging analysis of a given time series.
powerSpectraLS performs a power spectrum analysis based on a least-squares tech-
nique; in particular useful for regular waves.
reflectionAnalysis2DLS performs a reection analysis based on a least-squares tech-

niques; in particular useful for regular waves (Jacobsen et al., 2012).
powerSpectraFFT performs an evaluation of the power spectrum for a given time

series with the use of the Fourier transform; in particular useful for irregular waves.
reflectionAnalysis2DFFT performs a decomposition into incident and reected waves

based on a Fourier transport; in particular useful for irregular waves (Zelt and Skjel-
breia, 1992).
Besides this list, the utility is only documented through its application in some of the utilities.
CHAPTER 4
Wave Theories
This chapter discusses the wave theories available in waves2Foam. The information for the wave
theories are given in the le waveProperties.input and the required set of les to run external
wave theories from third-party software source (Sections 4.2.8 and 4.3).
The required information in waveProperties.input are either given in a sub-dictionary for a
given name of a boundary or as global information. The overall organisation of waveProperties.input
is described in Section 3.1.2.
4.1 Validity of wave theories

The choice of an appropriate wave theory should be based on a few considerations:
Is the wave theory valid in the range of wave heights, wave periods and water depths to
be studied?
Is an experimental study sought reproduced? If yes, which wave theory did they apply in
the experimental study?
With respect to the rst question, then there are several sources available, which treat the
validity of wave theories. Regarding the second question, then it is important to stay as close as
possible to the experimental study to obtain accurate results. Here, either the appropriate wave
theory should be chosen or the actual paddle signal should be used as a boundary condition (see
e.g. Higuera et al., 2015; Jacobsen et al., In review, for the latter).
4.2 Algebraic wave theories

The algebraic wave theories are covered in this section. This refers to wave theories that are
given in algebraic form. Each subsection contains information on the required input data for a
particular wave theory and one or more references to the implementation.
The settings are procided in Code Fragments, where both the mandatory and non-mandatory
parameters are provided. The default values for non-mandatory parameters are specied.
4.2.1 Regular wave theories

4.2.1.1 Stokes rst order
The rst order Stokes wave theory is encountered in a multiple of sources (Dean and Dalrymple,
1991; Svendsen and Jonsson, 1982). The latter source is used the waves2Foam-implementation.
The settings are provided in Code Fragment 4.1.
4.2.1.2 Standing Stokes rst order

The rst order Stokes standing wave is a superposition of two waves of equal amplitude that
propagate in opposite directions. The implementation into waves2Foam follows Dean and Dal-
rymple (1991). The settings are provided in Code Fragment 4.2.
33
34 Manual v. 0.9 (Draft) - 4 Wave Theories
Code Fragment 4.1: Settings Stokes First Order

waveType stokesFirst;
depth <scalar>; // [m]
period <scalar>; // [s]
direction <vector>; // [-]
phi <scalar>; // [radians]
height <height>; // [m]
Tsoft <scalar>; // [s] Default value: period
Code Fragment 4.2: Standing Stokes First Order

waveType stokesFirstStanding;
height <scalar>; // [m]
4.2.1.3 Stokes second order

The second order Stokes theory is a perturbation extension to the rst order theory. The
extension allows for a nite wave height and thus extends the range of validity of the wave
in terms of nonlinearity. The implementation follows the description in Svendsen and Jonsson
(1982). The settings are provided in Code Fragment 4.3.
Code Fragment 4.3: Stokes Second Order

waveType stokesSecond;
debug <bool>; // [-]
4.2.1.4 Modulated second-order Stokes wave

The modulated second order Stokes wave theory is an ad.hoc. adjustment to second order
wave theory, with a wave height that is modulated in time. It should be stressed that this wave
theory is not an accurate second order wave theory. It has been applied to study the eect of
a small irregularity in a incident wave on cross shore morphodynamics (Jacobsen and Fredse,
2014a,b). The modulation to the wave height is as follows:

1
Hm = H0 1 + sin (t k x) (4.1)
N
Here,Hm is the modulated wave height, H0 is the mean wave height, is the magnitude of the
modulation, N is an integer, is the cyclic frequency of the primary wave and k is the wave
number vector. The settings are provided in Code Fragment 4.4.
4.2.1.5 Stokes fth order

The fth order Stokes theory further extend the validity for larger nonlinearities of the waves
in comparison with e.g. the second order Stokes theory, see (Fenton, 1985, 1990). The largest
dierence is that the net ux of water in terms of the Stokes drift is needed in order to evaluate
the wave properties. Please note that the perturbution theory is not always applicable (Section
4.1. The settings are provided in Code Fragment 4.5.
Manual v. 0.9 (Draft) - 4.2 Algebraic wave theories 35
Code Fragment 4.4: Modulated Second Order Stokes

waveType stokesSecondModulation;
epsilon <scalar>; // [-]
modN <scalar>; // [-] Number of waves per modulation
debug <bool>; // [-]
Code Fragment 4.5: Stokes Fifth Order

waveType stokesFifth;
stokesDrift <scalar>; // [m/s]
4.2.1.6 First-order cnoidal theory

Cnoidal wave theory is intended for shallower waters than class of Stokes theories and the
underlying assumption in the derivation is that the ratio h/L (h is the water depth and L is
the wave length) is a small number. The waves2Foam-implementation follows the description
in Svendsen and Jonsson (1982). The settings are provided in Code Fragment 4.6.
Code Fragment 4.6: First Order Cnoidal Theory

waveType cnoidalFirst;
The key parameter for cnoidal wave theory is the shape parameter m, which needs to be
strictly less than 1. Consequently, the utility setWaveParameters automatically increases the
write precision of m to 14 digits to ensure that m is written to waveProperties as a value
strictly smaller than 1.
4.2.1.7 Stream function wave

Stream function wave theory is a nonlinear solution to the steady wave problem in arbitrary
depth. The theory is - amongst others - described in Rienecker and Fenton (1981); Fenton (1988).
This wave theory is generally applicable as long as the wave height is below the breaking criterion
(Section 4.1). The settings are provided in Code Fragment 4.7.
The utility used to evaluate the coecients is based on the original source code provided in
1
Fenton (1988) .
4.2.2 Bichromatic wave theories

4.2.2.1 First-order bichromatic wave
A bi-chromatic wave is a wave with two frequency components. The rst order wave is a simple
linear super-position of two Stokes rst order waves with two dierent wave heights (Madsen
and Fuhrman, 2006). The settings are provided in Code Fragment 4.8.
1 The present version was originally typed by Jesper Skourup (DHI) and it was later ammended by Harry
Bingham (Technical University of Denmark). The source code was made available to waves2Foam by Henrik
Bredmose (Technical University of Denmark).
Code Fragment 4.7: Stream Function Theory. Note the EITHER-OR structures.
waveType streamFunction;
N <scalar>; // [-] Number of components
Niter <scalar>; // [-] Number of iterations
EITHER:
specifyPeriod false; // [boolean]
waveLength <scalar>; // [m]
OR:
specifyPeriod true; // [boolean]
EITHER:
specifyEuler true; // [boolean]
eulerVelocity <scalar>; // [m/s]
OR:
specifyEuler false; // [boolean]
stokesVelocity <scalar>; // [m/s]
Code Fragment 4.8: First Order Bichromatic

waveType bichromaticFirst;
period1 <scalar>; // [s]
direction1 <vector>; // [-]
height1 <scalar>; // [m]
phi1 <scalar>; // [radians]
Tsoft <scalar>; // [s] Default value: max(period1, period2)
4.2.2.2 Second-order bichromatic wave
The second order bichromatic wave is similar to the rst order variant, but it includes all the
wave-wave interactions between two waves with dierent periods and propagation directions,
(Madsen and Fuhrman, 2006). The settings are provided in Code Fragment 4.9.
Code Fragment 4.9: Second Order Bichromatic

waveType bichromaticSecond;
Tsoft <scalar>; // [s] Default value: max(period1, period2)
4.2.3 First-order irregular waves

The rst order irregular wave theory is a simple linear superposition of rst order Stokes waves,
where the rst order Stokes theory is extrapolated to the free surface:
N
X
= ai cos(i t ki x + i ) (4.2)
i
Here, is the surface elevation, N is the number of wave components, ai is the amplitude of the
i'th wave component, is the cyclic frequency, k is the wave number vector and is a phase
(random or focused).
In order to give the information for the irregular waves, there are currently three classes of
information needed:
Spectral shape
Spectral discretisation
Phases
The shape determines the variation of ai with i , the discretisation determines the discrete
representation of i
and the phases determines the values of i .
It is possible to extend the present options by modications of the run-time selectable classes
located in <waves2Foam_src>/src/waves2FoamProcessing/preProcessing/setWaveProperties/-
irregular/waveSpectra. The three classes of information are implemented in the sub-directories
spectra, frequencyAxis and phases.
The following paragraphs describe the various settings needed to prescribe an incident irregular
wave eld. Some common settings are provided in Code Fragment 4.10.
Code Fragment 4.10: Irregular waves: Common settings

waveType irregular;
N <label>; // [-] Number of wave components
Tsoft <scalar>; // [s] Ramping time
4.2.3.1 Spectral shape

The following spectral shapes are currently implemented:
JONSWAP
Pierson-Moskowitz
Information on these spectra can be found in many standard introductions to irregular wave
theory, e.g. Holthuijsen (2007). The settings for JONSWAP are provided in Code Fragment
4.11 and the settings for Pierson-Moskowitz are provided in Code Fragment 4.12.
Code Fragment 4.11: Irregular waves: JONSWAP settings

spectrum JONSWAP;
Hs <scalar>; // [m] Spectral wave height
Tp <scalar>; // [s] Peak wave period
gamma <scalar>; // [-] Peak enhancement factor
Code Fragment 4.12: Irregular waves: Pierson-Moskowitz settings

spectrum PiersonMoskowitz;
Hs <scalar>; // [m] Spectral wave height
Tp <scalar>; // [s] Peak wave period
4.2.3.2 Spectral discretisation

There are currently two discretisation methods available for the frequequency axis:
equidistantFrequencyAxis
cosineStretchedFrequencyAxis
The latter produces a stretching of the frequency axis towards the peak of the spectrum. This
stretching (and any other non-equidistant stretching) greatly improves on the statistical prop-
erties of resulting time series for the surface elevation, see Appendix B for details.
The frequency discretisation is described in a sub-dictionary inside of the denition of the
irregular wave (Code Fragment 4.13). The keyword frequencyAxis refers to the discretisation
options metioned in the list above.
Code Fragment 4.13: Irregular waves: Frequency discretisation

frequencyAxis
{
discretisation <frequencyAxis>;
lowerFrequencyCutoff <scalar>; // [Hz]. Default: 1/(3Tp)

upperFrequencyCutoff <scalar>; // [Hz]. Default: 3/Tp
writeSpectrum <bool>; // Write the target spectrum (f, S)

// and not just (f, amp)
}
4.2.3.3 Phases
There are two types of phasing methods available and these are is either a random phase or
a phase focusing for a certain location, x0 at a certain time instance t0 . The settings for the
random phases are provided in Code Fragment 4.14 and the settings for the phase focusing are
provided in Code Fragment 4.15.
Code Fragment 4.14: Irregular waves: Random phases

phaseMethod randomPhase; // Default value
seedForRandomPhase <label>; // No required
If a value for the seeding is not provided (Code Fragment 4.14), then the seeding for the
random number generator is based on the clock time. Applying a xed seeding ensures that the
results can be reproduced.
Code Fragment 4.15: Irregular waves: Focused phases

phaseMethod focusingPhase;
focusTime 10.0; // [s]
focusPoint (0 0 0); // [m]
Nota Bene 4.1: A note on phase focusing

The phases are evaluated to obtain phase focusing based on linear wave theory.
Due to amplitude dispersion, a bit of tweeking of the focus point and focus time
must be expected for large wave heights.
4.2.3.4 Example
A compilation of all of the various components is exemplied in Code Fragmentcode:irregularExample.
This example produces an irregular wave with 100 wave components described with the JON-
SWAP spectrum with a peak enhancement factor of 3.3. The spectral wave height is 1 m, the
peak wave period is 10 s and the water depth if 10 m. The phases are random (but repro-
ducible) with a xed seeding of 10. The frequency axis is equidistantly discretised and user
dened frequency cut-os are specied.
Code Fragment 4.16: Irregular waves: Example

// Define the wave discretisation and ramping
waveType irregular;
N 100;
Tsoft 10;
// Define the phases

phaseMethod randomPhase;
seedForRandomPhase 10;
// Define the spectrum

spectrum JONSWAP;
Hs 1;
Tp 10;
gamma 3.3;
depth 10;
direction (1 0 0);
frequencyAxis
{
discretisation equidistantFrequencyAxis;
lowerFrequencyCutoff 0.01;
upperFrequencyCutoff 0.4;
writeSpectrum false;
}
4.2.4 Second-order irregular waves

Second order irregular waves are currently not implemented, because the computational eorts
are likely to be prohibitively expensive (Appendix B). Second order irregular wave theory in-
cludes wave-wave interaction terms and the theory are described in e.g. Sharma and Dean
(1981); Madsen and Fuhrman (2012).
In the work by Madsen and Fuhrman (2012) it was shown that some of the coecients in the
perturbation solution to the third order irregular wave theory become unbounded and violates
the perturbation assumption. This means that there does not exist a (bounded) perturbation
solution for the irregular wave problem beyond second order. Therefore numerical models are
required to specify nonlinear, irregular incident waves (Section 4.3.3).
4.2.5 Potential current

The wave theory potentialCurrent is a means of introducing a current that is uniform over
the depth, hence the name. This is typically used for outlet relaxation zones, where the velocity
vector is set of 0. The settings are provided in Code Fragment 4.17.
Code Fragment 4.17: Potential current

waveType potentialCurrent;
U <vector>; // [m/s]
Tsoft <scalar>; // [s]
4.2.6 Solitary wave theories

4.2.6.1 First-order solitary wave
A solitary wave is a single wave crest and the current implementation follows that presented in
Svendsen and Jonsson (1982). The settings are provided in Code Fragment 4.18.
Here, the keyword x0 is the location of the crest at t=0 s.
Code Fragment 4.18: Solitary First Order

waveType solitaryFirst;
x0 <vector>; // [m]
4.2.6.2 Chappelear (1962)

The higher-order solitary wave theory according to Chappelear (1962) is also implemented. The
settings are provided in Code Fragment 4.19.
Code Fragment 4.19: Solitary wave (Chappelear, 1962)

waveType chappelear1962;
x0 <vector>; // [m]
Here, the keyword x0 is the location of the crest at t=0 s.
4.2.7 Combined waves

The use of the wave theory combinedWaves allows for the combination of any of the wave
theories described in this chapter. Merely specify a word list of other sub-dictionaries in the le
waveProperties.input. The settings are provided in Code Fragment 4.20.
Code Fragment 4.20: Linear Superposition

waveType combinedWaves;
combinedWaveNames <wordList>;
Please note that a combination of e.g. potentialCurrent with stokesFirst does not aect
the evaluation of the wave number, i.e. the Doppler Eect is not included. Tailored wave
theories are required to correctly model wave-current interaction.
The combined waves is convenient, if the same properties are required on multiple boundaries.
4.2.8 External source

It is possible to couple waves2Foam to an external source that provides the information on the
velocity eld and the surface elevation.
This wave theory has a default value, which means that no external wave source is dened, i.e.
an external wave theory is always constructed, but it is not used. It is only possible to have one
external source in a simulation. The settings require both a globel setting and settings required
for the individual boundaries (Code Fragment 4.21). Besides these settings, there are settings
specic to the external wave model and these are specied in Section 4.3.
Code Fragment 4.21: Use of externalSource

// Global setting
externalForcing <string>; // Default: emptyExternal;
// Settings for individual boundary

<name>Coeffs
{
waveType externalSource;
}
Manual v. 0.9 (Draft) - 4.3 External wave theories 41
Nota Bene 4.2: Always specify an external source (if used)

If externalSource is given as a waveType, a valid external source has to be pro-
vided, since the default type emptyExternal is not accepted.
4.3 External wave theories

One of the wave theories is called externalSource (see Section 4.2.8). This theory allows for a
smooth coupling between waves2Foam and any algebraic or non-algebraic method for computing
the wave eld; this could for instance be the coupling between an external numerical solution
of the nonlinear wave problem and waves2Foam. The available methods are described in the
following, however, the limitations is largely due to the limitations in interfacing with other
models.
All external wave theories have one thing in common: The control of the coupling is spec-
ied through (i) les native for the external wave theory and (ii) the sub-dictionary called
externalForcingCoeffs.
4.3.1 Empty external method

The method is a dummy method that cannot be used. It is needed to construct an external
source method, when the algebraic methods are used, which explains the existence of this theory.
It is not necessary to specify the settings (Code Fragment 4.22) in waveProperties.input, since
it is the default value for externalForcing.
Code Fragment 4.22: Empty external source (Default)

// Global setting
externalForcing emptyExternal;
4.3.2 Fast summation of irregular waves

This external wave theory produces identical results to the ordinary irregular wave implementa-
tion described in Section 4.2.3; the denitions of frequency discretisation, phasing methods and
spectral shape are identical.
The main dierence between the methods is that irregularFast performs a lot of pre-
processing to lower the amount of time spend to evaluate trigonometric and hyperbolic functions.
These pre-computed sets of data are stored in arrays for unique values of x, y and z . If needed, it-
nerpolation will be performed between these values. The consequence is a considerable speed-up
(Appendix B).
The global settings are provided in Code Fragment 4.23. Furthermore, the sub-dictionary
information must be specied in a sub-dictionary called externalForcingCoeffs, where the
content is the denition of the irregular wave as exemplied in Code Fragment 4.16.
Code Fragment 4.23: Fast summation of irregular wave theories

// Global settings
externalForcing irregularFast;
searchTolerance <scalar>; // [m]
ignoreMeshMotion <boolean>; // [-]
The method is not intended to be used with mesh motion, but in certain cases the mesh motion
is the relaxation zone is so small that it is safe to ignore that eect. If ignoreMeshMotion is set
to false/no, then the simulation will stop in case of mesh motion.
The keyword searchTolerance is related to the interpolation routine between the unique
values of x, y and z . If a unique point within the distance of searchTolerance is found, then
no interpolation is executed.
4.3.3 OceanWave3D
waves2Foam is coupled with the third-party software OceanWave3D, which is able to evaluate
fully nonlinear wave propagation (Engsig-Karup et al., 2009). The coupling between Ocean-
Wave3D and waves2Foam is described in Paulsen et al. (2014a) and the current implementation
is streamlined with the general external source functionality in waves2Foam by Bo T. Paulsen
and Niels G. Jacobsen. The settings are provided in Code Fragment 4.24.
Nota Bene 4.3: Installation of OceanWave3D

OceanWave3D is installed alongside waves2Foam, why the user does not have to
do any additional steps. The downloaded source code also contains a GUI, which
allows for easy setup of 2DV simulations with OceanWave3D. The GUI is not
compiled as part of waves2Foam.
Code Fragment 4.24: Coupling with OceanWave3D

// Global setting
externalForcing oceanWave3D;
externalForcingCoeffs
{
waveType oceanWave3D;
// Define the intervals for the OpenFoam calculations

nIntervals 1;
startTimes nonuniform List<scalar> 1(5);
endTimes nonuniform List<scalar> 1(20);
// Should the interval be ramped?

rampInterval off;
// Name of the sub-dictionary (without Coeffs), where the external-

// Source definition is given. Is needed, when the mapping OCW3D to
// OpenFoam is carried out.
mappingZone inlet;
};
4.3.3.1 Output
The output from OceanWave3D consists of three elements:
Kinematics* and *.fort les. The output in these les cannot be used, because the time
stepping is controlled by OpenFoam, hence the time stamps on these les are wrong.
waveGauges.dat contains ASCII-formatted wave gauge data, which comes in the same
format as that for the standard surfaceElevation data in waves2Foam. The time stamp is
correct.
Hot-start les are written in synchronisation with the OpenFoam results. The les are out-
putted to the folder <casePath>/OCW3Dhotstart. The naming convection is OceanWave3D.-
<timeIndex>, where <timeIndex> is the time index in the OpenFoam simulation. This
index can be retrieved in the le <timeName>/uniform/time.
4.4 Extensions with new theories

Extension of waves2Foam with new wave theories is relatively straightforward for both the
algebraic and the external wave theories. The following two sections describes the approach to
develop new wave theories for each of these two classes.
4.4.1 Extension: Algebraic wave theory

In the directory <waves2Foam_src>/doc/templateWaveTheory there is a small script called
makeNewWaveTheory. Execute this script (Code Fragment 4.25) to create a new wave theory
called myWaveFormulation:
The script creates output as shown in Code Fragment 4.26.
It is merely a matter of following this step-by-step guideline to implement a new wave theory.
Manual v. 0.9 (Draft) - 4.4 Extensions with new theories 43
Code Fragment 4.25: Create a new algebraic wave theory

> ./makeNewWaveTheory myWaveFormulation
Code Fragment 4.26: Create a new wave theory - Output

A directory called 'myWaveFormulation' with [C,H]-files is created.
Steps to finalise:
1. Copy the directory to the correct wave type in waves2Foam/src/ \
waveTheories
2. Insert the needed variables and expressions for eta, U, etc.
3. Add a line in waves2Foam/src/Make/files to include the new theory in
the compilation.
4. Done.
4.4.2 Extension: External source

There is no easy development guide for the coupling with an external source. This is because
the class will depend on the interface to the external source. Some relevant thoughts are:
Is it possible to start the external source at a time dierent from t=0 s? If not, it should
be possible to take several time steps in the external source before activating OpenFoam
simulation, if the latter needs restarting;
If it is possible to restart the external source, then make sure to save data for the external
source at the same time instances as data is saved in the OpenFoam simulation;
For a proper integration of the external source with waves2Foam, it should be possible to
control the time stepping of the external source through OpenFoam;
For an optimal integration of the external source with waves2Foam, the external source
should be compatible with a variable time steps;
Alternatively, a time integration scheme between an equidistant time stepping in the ex-
ternal source and OpenFoam should be implemented and consequently a time control of
the external source that allows for the external force to be ahead in time by one or more
time steps.
CHAPTER 5
Solvers
5.1 waveFoam
The waveFoam solver is a dedicated solver to the study of wave interaction and propagation
in a static computational domain. There may be one or more xed structures, but permeable
layers are not included.
5.2 porousWaveFoam
The porousWaveFoam solver is similar to waveFoam (Section 5.1) with the addition that one
can specify one or more permeable layers.
5.3 waveDyMFoam (moving meshes)

waves2Foam is not distributed with a solver for moving meshes and there is similarly not any tu-
torials that shows the use of waves2Foam with moving meshes. There is, however, no limitations
in the waves2Foam framework that prevents a merging of the interDyMFoam and waves2Foam.
While there is no solver available, waves2Foam does take account of the following:
Moving meshes: The weights in the relaxation zone are re-calculated every time step in
case of moving meshes.
Topological changes: If there are topological changes in the computational mesh due to
e.g. mesh renement or addition or removal of computational cells, the location of the
relaxation zones are recomputed every time step.
waves2Foam has already been used with moving meshes in several works (see Section ??).
Warning 5.1: Online documentation to create waveDyMFoam.

Do not use the guide on the Wiki-page or on https://sites.google.com/site/
jordimuela/openfoam/how-to-couple-waves2foam-with-dynamic-mesh-motion,
since the guides are outdated.
5.3.1 Modications to interDyMFoam

The following steps can be taken in order to create a solver waveDyMFoam:
Make changes to the interDyMFoam les, see Section 5.3.1.1 for details;
Update the le Make/options, see Section 5.3.1.2 for details;
Update the le Make/files, see Section 5.3.1.3 for details;
45
46 Manual v. 0.9 (Draft) - 5 Solvers
5.3.1.1 Changes to les

The Code Fragment 5.1 may be used to display the dierences between the interFoam and
waveFoam solvers. This information can be used to make the required modications to create
waveDyMFoam:
Code Fragment 5.1: Example code to compare les in order to understand mod-
ications made from interFoam to waveFoam
#!/bin/bash
list=`ls *C *H`
for i in $list
do
echo "============ BEGIN: "$i" ============"
if [ "$i" == "waveFoam.C" ]
then
diff $i $FOAM_APP/solvers/multiphase/interFoam/interFoam.C
else
diff $i $FOAM_APP/solvers/multiphase/interFoam/$i
fi
echo
echo "============= END: "$i" ============="
echo
done
Place the script in the waveFoam directory for the installed (and sourced) version of waves2Foam.
Execute the commands given in Code Fragment 5.2.
Code Fragment 5.2: How to execute the comparison

> chmod a+x compareFiles.sh
> ./compareFiles.sh
Warning 5.2: Warning concerning the denition of the gravity vector.

It is necessary to create the gravity vector before creating the elds in the
createFields.H le. In some versions of OpenFoam, the gravity vector is cre-
ate in createFields.H after the constructions of the volume elds. The code will
compile, but it will not be possible to execute it.
This explains why the gravity vector is moved outside of createFields.H and into
waveFoam.C.
After the required changes have been identied, copy the les belonging tointerDyMFoam to a
folder next to waveFoam, incorporate the changes and rename interDyMFoam.C to waveDyMFoam.C.
5.3.1.2 Details on Make/options

Additional lines are needed in the EXE_INC-section of Make/options (Code Fragment ??) and
additional lines in the EXE_LIBS-section of Make/options (Code Fragment ??). Take care to
add the necessary backslash:
Code Fragment 5.3: Additional lines in Make/options for EXE_INC.

-DOFVERSION=<Numerized version number> \
-DEXTBRANCH=<OF-release = 0. FE-release = 1> \
-I$(WAVES_SRC)/waves2Foam/lnInclude \
-I$(WAVES_SRC)/waves2FoamSampling/lnInclude \
-I$(WAVES_GSL_INCLUDE)
Manual v. 0.9 (Draft) - 5.3 waveDyMFoam (moving meshes) 47
Code Fragment 5.4: Additional lines in Make/options for EXE_LIB.

-L$(WAVES_LIBBIN) \
-lwaves2Foam \
-lwaves2FoamSampling \
-L$(WAVES_GSL_LIB) \
-lgsl \
-lgslcblas
5.3.1.3 Details on Make/files

The le Make/files should look as in Code Fragment 5.5.
Code Fragment 5.5: Modications to Make/files

waveDyMFoam.C
EXE = $(WAVES_APPBIN)/waveDyMFoam
CHAPTER 6
Tutorials
6.1 Execute a tutorial
The handling of cross-version compatibility means that several actions are required prior to
execution of the tutorial. Due to this, all cases come with an Allrun-script. It is simply a
matter of executing this script according to Code Fragment 6.1 to run the tutorial. Opening
the script furthermore outlines the used commands.
Code Fragment 6.1: Executing a tutorial

$ ./Allrun
6.2 waveFoam
6.2.1 standingWave
This tutorial shows how to handle wave generation and a fully reecting sea wall utilising the
relaxation zone technique.
The following utilities and solver relevant for waves2Foam are used:
setWaveParameters
setWaveField
waveFoam
6.2.2 waveFlume
This tutorial shows the user how to use the waveProperties.input dictionary to set-up the
wave properties, relaxation zones and initialization specications. Furthermore, the tutorial
illustrates the physical wave transformation that occurs if an invalid wave theory is used at the
inlet (see e.g. experimental work by Chapalain et al., 1992).
waveGaugesNprobes
setWaveParameters
setWaveField
waveFoam
postProcessWaves2Foam
49
50 Manual v. 0.9 (Draft) - 6 Tutorials
6.2.3 bejiBattjes
This tutorial provides the experimental setup from Beji and Battjes (1993) and a matlab-script,
which can be used to compare the experimental data and the simulation. This tutorial is not
optimised to work as validation case.
waveGaugesNprobes
setWaveParameters
setWaveField
waveFoam
6.2.4 couplingOceanWave3D
This tutorial shows how waves2Foam can be coupled with the nonlinear potential wave model
for wave transformation: OceanWave3D.
waveGaugesNprobes
setWaveParameters
setWaveField
waveFoam
6.2.5 periodicSolitary
This tutorial initialises the numerical wave tank with solitary wave. The computational domain
is cyclic in the direction of wave propagation and there are no relaxation zones. Note that the
simulation generates a spurious wave propagating in the opposite direction of the solitary wave.
This is thought to be due to the low order of the solitary wave theory.
setWaveParameters
setWaveField
waveFoam
6.2.6 3Dwaves
This tutorial shows the use of relaxation zones in 3D. There are 3 inlet relaxation zones and
one outlet relaxation zone. In the middle of the wave tank there is a vertical internal wall on
which the waves are reected. The waves, which are generated in the 3 inlet relaxation zones,
have dierent directions relative to the x-axis.
setWaveParameters
setWaveField
waveFoam
6.2.7 squarePile
This tutorial shows how to use the cylindrical relaxation zone. A unidirectional current is
the target function all the way around, hence one relaxation zone acts as both generation and
absorption. The relaxation zone removes the internally generated surface disturbances. It could
be interesting to have it tested for ships waves. Alternatively, a better shape for that purpose,
e.g. an elliptical relaxation zone shape could be implemented.
setWaveParameters
setWaveField
waveFoam
Manual v. 0.9 (Draft) - 6.3 porousWaveFoam 51
6.3 porousWaveFoam
6.3.1 porousDamBreak
This tutorial gives an example of the use of porousWaveFoam. The case is of a dambreak
through a permeable structure and there are post-processing scripts available to compare with
experimental data by Liu et al. (1999). The comparison is reported in Jensen et al. (2014).
setWaveParameters
waveGaugesNProbes
porousWaveFoam
postProcessWaves2Foam
6.4 Miscellaneous
6.4.1 relaxationZoneLayout
This tutorial will show how to use the utility to obtain a visual representation of the relaxation
zone(s). This could be helpful to detect potential problems in the setup.
The following utility relevant for waves2Foam are used:
relaxationZoneLayout
CHAPTER 7
Validation
This chapter is intended as a community eort, where all users of waves2Foam can provide
small text pieces. The text pieces should put the work with waves2Foam into perspective of the
existing literature and present cases, where OpenFoam and waves2Foam are validated. Each
section shows the list of author(s) and their aliation.
How to contribute? Please send text (LaTeX-format), references (bib-format) and graphics
1
(eps-format) through my ResearchGate account:

Niels Gjl Jacobsen (https://www.researchgate.net/prole/Niels_Jacobsen3).
7.1 Loads on bridge decks

Contributor: Betsy Seiert (Florida Atlantic University). June 2017
A group of researchers at the University of Hawaii did an extensive study on wave loads on coastal
bridges during a study on the vulnerability of coastal bridges to hurricane and tsunami. In Seiert et al.
(2014), Hayatdavoodi et al. (2014), Hayatdavoodi et al. (2015b), and Seiert et al. (2015), calculations
for surface elevation, as well as vertical and horizontal force due to solitary and cnoidal wave impact
on a at plate and bridge model with girders was compared with experimental measurements, for a
range of fully elevated to fully submerged bridge elevations. Numerical calculations were performed
using OpenFOAM, and Euler's equations were solved in the uid domain by setting viscosity to ap-
proximately zero. Solitary and cnoidal waves were generated using the wave generation and absorption
toolbox waves2Foam developed by Jacobsen et al. (2012) and using solutions of the Level 1 Green-Naghdi
equations for surface elevation, given in Ertekin and Wehausen (1986), Ertekin and Becker (1998), and
Sun (1991), for example. Additionally, eects of air entrapment and venting on forces on a bridge deck
was studied in Hayatdavoodi et al. (2014) by rst comparing experiments and calculations where air is
fully trapped between the girders, and then by adding air relief holes to the bridge deck in numerical
calculations such that air can fully escape between the girders and comparing with experiments. For each
of these studies, OpenFOAM was able to accurately calculate reection, transmission and dispersion of
waves, as well as vertical and horizontal forces on the at plate and bridge model.
Using validations from this study, Hayatdavoodi et al. (2015a) assessed the vulnerability of four selected
bridges on the island of Oahu, Hawaii, to estimated maximum hurricane waves loads based on simulated
hurricane landfall at each bridge location. Calculations of vertical and horizontal force on each selected
bridge geometry are made by solving for Euler's equations in OpenFOAM, and by another code which
solves for force using the Green-Naghdi equations, and which simplies the bridge as a fully submerged
at plate. Calculations are compared with theoretical calculations made using linear long-wave theory
(Patarapanich, 1984) and some simplied design-type force equations (Douglass et al., 2006; McPherson,
2008; Kulicki and Mertz, 2008, (AASHTO)).
Eects of the compressibility of air trapped between bridge girders, and scaling was studied by Seiert
et al. (2015) by solving for both the compressible and incompressible Euler's equations in OpenFOAM and
comparing with experimental measurements. For this study, Seiert et al. (2015) successfully integrated
the waves2Foam toolbox into the compressible version of interFoam, namely compressibleInterFoam.
Bricker and Nakayama (2014) performed numerical calculations to study the cause of failure of the
Utatsu bridge during the 2011 Great East Japan Tsunami. Time series for ow depth and velocity at
the bridge location were taken from a large-scale Delft shallow water simulation and applied to a 2D
bridge model in OpenFOAM. Using these calculations, the authors determined the cause of failure was
1 Please make sure that graphics (gures, photos, etc) are not under any copyright agreement. If graphics
are provided, it is an implicit statement from the authors that the graphics may be included in this manual
without any violation of copyrights.
53
54 Manual v. 0.9 (Draft) - 7 Validation
likely due to a combination of factors, include the presence of a nearby sea wall, the seaward inclination
of the bridge, entrained sediment and entrapped air between the girders.
7.2 Interaction with coastal structures

Contributor: Niels Gjl Jacobsen (Deltares). May 2017
The interaction of waves and coastal permeable structures has been studied with several VOF-based
numerical models. Examples are those of Skylla (Van Gent et al., 1994), COBRAS (Liu et al., 1999),
IH2VOF (Losada et al., 2008; Guanche et al., 2009), ComFlow (Wellens et al., 2010) and lately Open-
Foam, where either of the two wave generation toolboxes IHFoam (Higuera et al., 2014) and waves2Foam
(Jensen et al., 2014; Jacobsen et al., 2015; Van Gent et al., 2015; Jacobsen et al., In review) were applied.
The modelling of permeable coastal structures has been subject to some debate, and dierences in the
formulation of the porous medium are observed between the formulations of for instance Wellens et al.
(2010); del Jesus et al. (2012); Higuera et al. (2014); Jensen et al. (2014). The dierences are amongst
other related to the choice of representative velocity, where Wellens et al. (2010) choose the pore velocity,
while e.g. Jensen et al. (2014) has given the momentum equation in lter velocities. Furthermore, there
are dierences concerning the scaling with the porosity on the various terms in the momentum equation.
waves2Foam was validated for the dambreak through a permeable block in Jensen et al. (2014) and
they also validated the overtopping rates against experimental data. Both comparisons showed a good
predictive skill of the numerical model. Subsequent works by Jacobsen et al. (2015); Van Gent et al.
(2015); Jacobsen et al. (In review) validated waves2Foam against additional experimental data. Several
aspects were covered and validated successfully: (i) Wave reection from hard and porous structures. (ii)
Wave damping within a rubble-mound breakwater. (iii) Pressure distribution at the sand-rock interface
in an open lter. (iv) Loads on a crest wall element on top of a rubble mound breakwater.
In the work by Jacobsen et al. (2015), an analytical model for the wave-induced setup within a
permeable core was derived. The analytical result that the wave-induced setup is a function of the
reection coecient was conrmed with waves2Foam. waves2Foam was subsequently used to developed
a numerically based empirical formula for the wave-induced setup outside the validity range of the
analytical model.
Finally, waves2Foam was used to predict the eect of ventilation of a bridge deck element with girders
against the experimental data by Seiert et al. (2015) (Jacobsen et al., In review). The considerable
eect of even small ventilation holes as observed in the experiments was predicted by the numerical
model.
7.3 Breaking waves on a beach prole

Contributor: Niels Gjl Jacobsen (Deltares). May 2017
A popular study is that of breaking waves on a straight or barred cross-shore prole. The rst study
with waves2Foam was reported in Jacobsen (2011); Jacobsen et al. (2012) with validation against the
experimental data by Ting and Kirby (1994) (spilling breaker). They used a k turbulence closure
with a modied production term following Mayer and Madsen (2000). The comparison is fair, but the
dissipation in the inner surfzone was too rapid. The studies furthermore found a strong dependency of
the breakpoint with the aspect ratio of the computational cells.
The physical experiments by Ting and Kirby (1994) were later studied by Brown et al. (2014, 2016)
(spilling and plunging cases), who considered a multiple of turbulence closures and the inuence of the
gravity in the turbulence formulation. They saw large dierences in the results as a function of the
chosen turbulence model and they found it important to include density in the turbulence equations. A
ranking of the turbulence closures is provided in Brown et al. (2016) and they recommend the nonlinear
k closure.
The measurements of wave height, mean ow and turbulence properties over a xed concrete bar
(Scott, 2005; Scott et al., 2005a,b) were used for validation in Jacobsen et al. (2014); Zhou et al. (2017).
While Jacobsen et al. (2014) used a k model as turbulence closure and performed the simulations in
2DV, Zhou et al. (2017) performed the simulations in 3D with an LES turbulence closure. The results by
Jacobsen et al. (2014); Zhou et al. (2017) were compared directly in Zhou et al. (2017) and the dierences
were not large. Zhou et al. (2017) found that the use of a 3D-model does not solve the problem with
excessive turbulence levels, which seems to be a recurring problem for this type of simulations (Bradford,
2000; Christensen, 2006; Brown et al., 2016; Fernandez-Mora et al., 2016).
Finally, it is worthwhile to mention the validation by Fernandez-Mora et al. (2016), who validated
the surface elevation signal, mean ow, boundary layer processes, turbulence levels and sediment trans-
port rates against recent experimental work (Van der Zanden et al., 2016, 2017). The general picture
conrms the maturity of this type of numerical modelling, though the sediment transport still deserves
attention. It should be mentioned, however, that the experimental data is for an evolving morphology,
why the evolution of the bathymetry could explain some of the discrepancies between the numerical
and experimental results. The importance of the evolving morphology could be addressed by using the
experimental data over a frozen bathymetry as reported in Van der A et al. (2017). This is yet to be
seen in the literature.
Manual v. 0.9 (Draft) - 7.4 Modelling of oating wave energy converters (WECs) 55
7.4 Modelling of oating wave energy converters (WECs)

Contributor: Edward Ransley (University of Plymouth). August 2017
Despite an R&D process reliant on computationally-ecient numerical tools; the higher-order inter-
actions between the waves and the body, the coupled nature of complete wave energy converter (WEC)
systems and, the `extreme' environment at desirable oshore development sites; dictates that high-delity
numerical methods are needed in the development of oating WECs.
Davidson et al. (2017) summarises the use of OpenFOAM in wave energy applications including
examples in which OpenFOAM simulations of xed and oscillating WECs have been validated against
experimental data. In terms of oating WECs: Devolder et al. (2016) used OpenFOAM to simulate the
heave of a oating point-absorber and found good agreement with experimental data for the motion and
radiated wave elds in both decay and regular wave tests; Palm et al. (2016) presented a method for
coupled mooring analysis combining an OpenFOAM/waves2Foam numerical wave tank (NWT) with an
in-house mooring code, MooDy. They show the coupled model is able to capture the non-linear wave
height dependence in the response of a moored oating vertical cylinder and, for the surge and heave
motion, the numerical results compare well with experimental measurements. For the pitch motion,
however, Palm et al. (2016) found the response was under-predicted at resonance but attribute this to
uncertainties and simplications in the buoy properties and geometry; Raee and Fivez (2015) used
an OpenFOAM/waves2Foam-based NWT to study wave interaction with a submerged point-absorber.
They show that, in moderate regular waves, the Navier-Stokes solver is noticeably more accurate than
a linear time-domain model when compared with physical data; Ransley (2015) constructed a NWT,
using native OpenFOAM/waves2Foam code, to assess the validity of such methods when analysing the
survival characteristics of WECs. Utilising physical measurement data from the experimental campaigns
of Jakobsen et al. (2014), Hann et al. (2015) and Gteman et al. (2015), he showed the NWT was capable
of reproducing the loading and pressure distribution on xed WECs as well as the large amplitude
motions of constrained devices in large regular waves (Ransley et al., 2017b). Ransley (2015) added a
linear generator model as a restraint to the sixDoFRigidBodyMotion solver and showed that the NWT was
capable of reproducing the complex behaviour of a point-absorbing WEC that undergoes full submersion
and breach as well as the load in the mooring caused by the generator's limited stroke length. Sjkvist
et al. (2017) used a similar method and showed that, for the same point-absorber, the NWT could
predict the motion of the buoy and the load on the generator throughout a focused wave event that
had been embedded in a regular wave background. Ransley (2015) showed the NWT was capable of
reproducing a design wave event known as NewWave, correctly predicting the combination of
dispersive focusing and energy exchange between frequency components. He went on to show that the
NWT could successfully predict the forces and run-up on xed generic WEC geometries (Ransley, 2015)
as well as the motion and mooring loads of simple WEC systems when subject to design waves (Ransley,
2015; Ransley et al., 2017a).
7.5 Wave-structure-seabed interaction

Contributor: Hisham Elsafti (Technical University of Braunschweig). July 2017
The waves2Foam toolbox was applied to study wave-structure-seabed interaction of caisson break-
waters by Elsafti and Oumeraci (2017). The toolbox was semi-coupled in a one-way fashion with the
geotechFoam solver (Elsafti and Oumeraci, 2016) via the data2TimeVaryingMappedPressure boundary
condition. The geotechFoam solver can simulate fully dynamic, coupled poro-elstoplastic seabed with
frictional contact between the structure and the underlying soil. The developed model system was
applied to reproduce the large-scale caisson breakwater model tests from the EU research project LI-
MAS (LIquefaction around MArine Structures) as described fully in Kudella et al. (2006). The setup
of the large-scale model tests contained wave gauges, pressure transducers, pore pressure transducers,
displacement meters and soil stress transducers. A comparison of the breaking wave pressure on the
front face of the caisson breakwater (produced with waves2Foam) and the large-scale tests may be found
in Elsafti and Oumeraci (2017). Furthermore, the vertical displacement of the front and back sides of
the caisson breakwater subject to breaking wave impact as calculated by geotechFoam (coupled with
waves2Foam) using the fully dynamic Biot formulation and the u p approximation against the results
of the large-scale tests are also found in Elsafti and Oumeraci (2017).
Bibliography
Arrighi, C., Alcrraca-Huerta, J. C., Oumeraci, H., Castelli, F., 2015. Drag and lift contribution to the
incipient motion of partly submerged ooded vehicles. Journal of Fluids and Structures 57, 170184.
Beji, S., Battjes, J., 1993. Experimental investigation of wave-propagation over a bar. Coastal Engineer-
ing 19 (1-2), 151162.
Berberovi, E., Van Hinsberg, N. P., Jakirli, S., Roisman, I. V., Tropea, C., 2009. Drop impact onto
a liquid layer of nite thickness: Dynamics of the cavity evolution. Physical Review E - Statistical,
Nonlinear, and Soft Matter Physics 79 (3), Art.no: 036306.
Bradford, S. F., 2000. Numerical simulation of surf zone dynamics. Journal of Waterway, Port, Coastal,
and Ocean Engineering 126 (1), 113.
Bricker, J. D., Nakayama, A., 2014. Contribution of trapped air, deck superelevation, and nearby struc-
tures to bridge deck failure during a tsunami. Journal of Hydraulic Engineering 140 (5), 05014002.
Brown, S. A., Greaves, D. M., Magar, V., Conley, D. C., 2016. Evaluation of turbulence closure models
under spilling and plunging breakers in the surf zone. Coastal Engineering 114, 177193.
Brown, S. A., Magar, V., Greaves, D. M., Conley, D. C., 2014. An evaluation of RANS turbulence closure
models for spilling breakers. Proceeding to Coastal Engineering Conference, 115.
Chapalain, G., Cointe, R., Temperville, A., 1992. Observed and Modeled Resonantly Interacting Pro-
gressive Water-Waves. Coastal Engineering 16 (3), 267300.
Chappelear, J. E., 1962. Shallow-water waves. Journal of Geophysical Research 67 (12), 46934704.
Christensen, E. D., 2006. Large eddy simulation of spilling and plunging breakers. Coastal Engineering
53 (5-6), 463485.
Davidson, J., Windt, C., Giorgi, G., Genest, R., Ringwood, J. V., 2017. Evaluation of energy maximising
control systems for wave energy converters using OpenFOAM. In: Nbrega, J. M., Jasak, H. (Eds.),
OpenFOAM R : Selected papers of the 11th Workshop. Springer International Publishing AG.
Dean, R. G., Dalrymple, R. A., 1991. Water Wave Mechanics for Engineers and Scientists, 1st Edition.
Vol. 2 of Advanced Series on Ocean Engineering. World Scientic.
del Jesus, M., Lara, J. L., Losada, I. J., 2012. Three-dimensional interaction of waves and porous coastal
structures. Part I: Numerical model formulation. Coastal Engineering , 5672.
Devolder, B., Rauwoens, P., Troch, P., 2016. Numerical simulation of wave-induced roll of a 2-D rect-
angular barge using OpenFOAM. In: Proceedings of the 2nd International Conference on Renewable
Energies Oshore, RENEW. Lisbo, Portugal.
Devolder, B., Rauwoens, P., Troch, P., 2017. Application of a buoyancy-modied k SST turbulence
model to simulation wave run-up around a monopile subjected to regular waves using OpenFOAM R.
Coastal Engineering 125, 8194.

Dimakopoulos, A. S., Cuomo, G., Chandler, I., 2016. Optimized generation and absorption for three-
dimensional numerical wave and current facilities. Journal of Waterway, Port, Coastal and Ocean
Engineering 142 (4), 113.
Douglass, S. L., Chen, Q., Olsen, J. M., Edge, B. L., Brown, D., 2006. Wave forces on bridge decks.
Tech. rep., Oce of Bridge Technology, Washington, DC.
Elsafti, H., Oumeraci, H., 2013. Modelling sand foundation behaviour underneath caisson breakwaters
subject to breaking wave impact. In: Proc. of the ASME 32nd 2013 Int. Conf. on Ocean, Oshore
and Arctic Eng. Nantes, France.
Elsafti, H., Oumeraci, H., 2016. A numerical hydro-geotechnical model for marine gravity structures.
Computers and Geotechnics 79, 105129.
Elsafti, H., Oumeraci, H., 2017. Analysis and classication of stepwise failure of monolithic breakwaters.
Coastal Engineering 121, 221239.
57
58 BIBLIOGRAPHY
Engelund, F., 1953. On the laminar and turbulent ows of ground water through homogeneous sand.
Vol. 3 of Transactions of the Danish Academy of Technical Sciences. Danish Academy of Technical
Sciences.
Engsig-Karup, A., August 2006. Unstructured Nodal DG-FEM Solution of High-Order Boussinesq-Type
Equations. Ph.D. thesis, Technical University of Denmark.
Engsig-Karup, A. P., Bingham, H. B., Lindberg, O., 2009. An ecient exible-order model for 3D
nonlinear water waves. Journal of Computational Physics 228 (6), 21002118.
Ertekin, R. C., Becker, J. M., 1998. Nonlinear Diraction of Waves by a Submerged Shelf in Shallow
Water. J. Oshore Mech. Arct. Eng. 120, pp. 212220.
Ertekin, R. C., Wehausen, J. V., 1986. Some Soliton Calculations. In: Webster, W. (Ed.), Proc. 16th
Symp. on Naval Hydrodynamics, Berkeley. National Academy Press, pp. 167184.
Fenton, J. D., 1985. A 5th-order Stokes theory for steady waves. Journal of Waterway, Port, Coastal and
Ocean Engineering 111 (2), 216234.
Fenton, J. D., 1988. The Numerical-Solution of Steady Water-Wave Problems. Computers & Geosciences
14 (3), 357368.
Fenton, J. D., 1990. Nonlinear Wave Theories. Wiley, Ch. 1, pp. 325, from The Sea. Ideas and Obser-
vations on Progress in the Study of the Seas. Part A. Editors: B. Le Mhaut and D. M. Hanes.
Fernandez-Mora, A., Ribberink, J. S., Van der Zanden, J., Van der Werf, J. J., Jacobsen, N. G., 2016.
RANS-VOF modelling of hydrodynamics and sand transport under full-scale non-breaking and break-
ing waves. Proceeding to Coastal Engineering Conference, 115.
Fuhrman, D. R., Madsen, P. A., Bingham, H. B., 2006. Numerical simulation of lowest-order short-
crested wave instabilities. Journal of Fluid Mechanics 563, 415441.
Gteman, M., Engstrm, J., Eriksson, M., Hann, M., Ransley, E., Greaves, D., Leijon, M., 2015. Scale
tests on the survivability of point-absorbing wave energy converters - part i. Journal of Ocean and
Wind Energy 2, 176181.
Guanche, R., Losada, I. J., Lara, J. L., 2009. Numerical analysis of wave loads for coastal structure
stability. Coastal Engineering 56, 543558.
Hann, M., Greaves, D., Raby, A., 2015. Snatch loading of a single taut moored oating wave energy
converter due to focussed wave groups. Ocean Engineering 96, 258271.
Hayatdavoodi, M., Ertekin, R. C., Robertson, I. N., Riggs, H. R., 2015a. Vulnerability assessment of
coastal bridges on oahu impacted by storm surge and waves. Natural Hazards 79 (2), 11331157.
Hayatdavoodi, M., Seiert, B., Ertekin, R. C., 2014. Experiments and computations of solitary-wave
forces on a coastal-bridge deck. Part II: Deck with girders. Coastal Engineering 88, 210228.
Hayatdavoodi, M., Seiert, B., Ertekin, R. C., 2015b. Experiments and calculations of cnoidal wave
loads on a at plate in shallow-water. Journal of Ocean Engineering and Marine Energy 1 (1), 7799.
Higuera, P., 2015. Application of computational uid dynamics to wave action on structures. Ph.D.
thesis, Universidad de Cantabria.
Higuera, P., Lara, J. L., Losada, I. J., 2014. Three-dimensional interaction of waves and porous coastal
structures using OpenFoam R . Part I: Formulation and validation. Coastal Engineering 83, 243258.
Higuera, P., Losada, I. J., Lara, J. L., 2015. Three-dimensional numerical wave generation with moving
boundaries. Coastal Engineering 101, 3547.
Holthuijsen, L. H., 2007. Waves in Oceanic and Coastal Waters, 1st Edition. Cambridge University Press.
Jacobsen, N. G., 2011. A full hydro- and morphodynamic description of breaker bar develop-
ment. Ph.D. thesis, Department of Mechanical Engineering, Technical University of Denmark,
http://orbit.dtu.dk/recid/313371.
Jacobsen, N. G., Borsboom, M., Van Gent, M. R. A., In preparation. Investigation of the KC -dependency
in the extended Darcy-Forchheimer resistance formulation. (), .
Jacobsen, N. G., Fredse, J., 2014a. Cross-Shore Redistribution of Nourished Sand Near a Breaker Bar.
Journal of Waterway, Port, Coastal and Ocean Engineering 140 (2), 125134.
Jacobsen, N. G., Fredse, J., 2014b. Formation and Development of a Breaker Bar under Regular Waves.
Part 2: Sediment Transport and Morphology. Coastal Engineering 88, 5568.
Jacobsen, N. G., Fredse, J., Jensen, J. H., 2014. Formation and Development of a Breaker Bar under
Regular Waves. Part 1: Model Description and Hydrodynamics. Coastal Engineering 88, 182193.
BIBLIOGRAPHY 59
Jacobsen, N. G., Fuhrman, D. R., Fredse, J., 2012. A Wave Generation Toolbox for the Open-Source
CFD Library: OpenFoam R . International Journal for Numerical Methods in Fluids 70 (9), 10731088.
Jacobsen, N. G., Van Gent, M. R. A., Capel, A., Borsboom, M., In review. Numerical prediction of wave
loads on crest walls on top of rubble mound structures. Coastal Engineering , .
Jacobsen, N. G., Van Gent, M. R. A., Wolters, G., 2015. Numerical analysis of the interaction of irregular
waves with two dimensional permeable coastal structures. Coastal Engineering 102, 1329.
Jakobsen, M. M., Iglesias, G., Kramer, M., Vidal, E., 2014. Experimental study of forces on point
absorber. In: Proceedings of the 5th International Conference Coastlab 14. Varna, Bulgaria.
Jensen, B., Jacobsen, N. G., Christensen, E. D., 2014. Investigations on the porous media equations and
resistance coecients for coastal structures. Coastal Engineering 84, 5672.
Kudella, M., Oumeraci, H., De Groot, M., Meijers, P., 2006. Large-scale experiments on pore pres-
sure generation underneath a caisson breakwater. Journal of Waterway, Port, Coastal, and Ocean
Engineering 132 (4), 310324.
Kulicki, J. M., Mertz, D. R., 2008. Guide Specications for Bridges Vulnerable to Coastal Storms.
AASHTO.
Liu, P. L. F., Lin, P., Chang, K. A., Sakakiyama, T., 1999. Numerical modeling of wave interaction with
porous structures. Journal of Waterway, Port, Coastal and Ocean Engineering 125, 322330.
Losada, I. J., Lara, J. L., Del Jesus, M., 2016. Modeling the interaction of water waves with porous
coastal structures. Journal of Waterway, Port, Coastal and Ocean Engineering 142 (6), 118.
Losada, I. J., Lara, J. L., Guanche, R., Conzalez-Ondina, J. M., 2008. Numerical analysis of wave
overtopping of rubble mound breakwaters. Coastal Engineering 55, 4762.
Madsen, P. A., Fuhrman, D. R., 2006. Third-order theory for bichromatic bi-directional water waves.
Journal of Fluid Mechanics 557, 369397.
Madsen, P. A., Fuhrman, D. R., 2012. Third-order theory for multi-directional irregular waves. Journal
of Fluid Mechanics 698, 304334.
Mayer, S., Madsen, P. A., 2000. Simulation of Breaking Waves in the Surf Zone using a Navier-Stokes
Solver. Proceeding to Coastal Engineering Conference I, 928941.
McPherson, R. L., 2008. Hurricane Induced Wave and Surge Forces on Bridge Decks. Ph.D. thesis, Texas
A&M University.
Moradi, N., 2015. Numerical simulation of uid resonance in the narrow gap of twin bodies in close
proximity. Ph.D. thesis, School of Civil, Environmental and Mining Engineering, The University of
Western Australia.
Palm, J., Eskilsson, C., Paredes, G. M., Bergdahl, L., 2016. Coupled mooring analysis for oating wave
energy converters using CFD: formulation and validation. International Journal of Marine Energy 16,
8399.
Patarapanich, M., 1984. Forces and moment on a horizontal plate due to wave scattering. Coastal
Engineering 8 (3), 279301.
Paulsen, B. T., Bredmose, H., Bingham, H. B., 2014a. An ecient domain decomposition strategy for
wave loads on surface piercing circular cylinders. Coastal Engineering 86, 5776.
Paulsen, B. T., Bredmose, H., Bingham, H. B., Jacobsen, N. G., 2014b. Forcing of a bottom-mounted
circular cylinder by steep regular water waves at nite depth. Journal of Fluid Mechanics 755, 134.
Raee, A., Fivez, J., 2015. Numerical prediction of extreme loads on the ceto wave energy converter.
In: Proceedings of the 11th European Wave and Tidal Energy Conference (EWTEC). Nantes, france.
Ransley, E., 2015. Survivability of wave energy converter and mooring coupled system using cfd. Ph.D.
thesis, Plymouth University.
Ransley, E., Greaves, D., Raby, A., Simmonds, D., Hann, M., 2017a. Survivability of wave energy
converters using CFD. Renewavle Energy 109, 235247.
Ransley, E., Greaves, D., Raby, A., Simmonds, D., Jakobsen, M., Kramer, M., 2017b. RANS-VOF
modelling of the wavestar point absorber. Renewavle Energy 109, 4965.
Rienecker, M. M., Fenton, J. D., 1981. A Fourier Approximation Method for Steady Water-Waves.
Journal of Fluid Mechanics 104, 119137.
Rusche, H., December 2002. Computational Fluid Dynamics of Dispersed Two-Phase Flows at High
Phase Fractions. Ph.D. thesis, Imperical College of Science, Technology and Medicine, available at:
http://powerlab.fsb.hr/ped/kturbo/OpenFOAM/docs/HenrikRuschePhD2002.pdf.
60 BIBLIOGRAPHY
Scott, C. P., 2005. Large-Scale Laboratory Observations of Wave Breaking Turbulence on a Fixed Barred
Beach. Master thesis.
Scott, C. P., Cox, D. T., Maddux, T. B., Long, J. W., 2005a. Large-scale laboratory observations of
turbulence on a xed barred beach. Measurement, Science & Technology 16 (10), 19031912.
Scott, C. P., Cox, D. T., Shin, S., Clayton, N., 2005b. Estimates of surf zone turbulence in a largescale
laboratory ume. Proceeding to Coastal Engineering Conference, 379391.
Seiert, B., Ertekin, R. C., Robertson, I. N., 2015. Wave loads on a coastal bridge deck and the role of
entrapped air. Applied Ocean Research 53, 91106.
Seiert, B., Hayatdavoodi, M., Ertekin, R. C., 2014. Experiments and computations of solitary-wave
forces on a coastal-bridge deck. Part I: Flat Plate. Coastal Engineering 88, 194209.
Seiert, B. R., 2014. Tsunami and storm wave impacts on coastal bridges. Ph.D. thesis, University of
Hawaii.
Seiert, B. R., Hayatdavoodi, M., Ertekin, R. C., 2015. Experiments and calculations of cnoidal wave
loads on a coastal-bridge deck with girders. European Journal of Mechanics - B/Fluids 52, 191205.
Seng, S., 2012. Slamming and whipping analysis of ships. Ph.D. thesis, Department of Mechanical
Engineering, Technical University of Denmark.
Sharma, J. N., Dean, R. G., 1981. 2nd-Order Directional Seas and Associated Wave-Forces. Society of
Petroleum Engineers Journal 21 (1), 129140.
Sjkvist, L., Wu, J., Ransley, E., Engstrm, J., Eriksson, M., Gteman, M., 2017. Numerical mod-
els for the motion and forces of point-absorbing wave energy converters in non-linear waves. Ocean
EngineeringManuscript in review.
Stahlmann, A., 2013. Experimental and Numerical Modeling of Scour at Oshore Wind Turbines. Ph.D.
thesis, Gottfried Wilhelm Leibniz Universitt Hannover.
Sun, X., 1991. Some theoretical and numerical studies on cnoidal-wave-diraction problems. Ph.D. thesis,
Masters thesis, University of Hawaii at Manoa, Honolulu.
Svendsen, I. A., Jonsson, I. G., 1982. Hydrodynamics of Coastal Regions, 1st Edition. Den Private
Ingenirfond.
Ting, F. C. K., Kirby, J. T., 1994. Observation of Undertow and Turbulence in a Laboratory Surf Zone.
Coastal Engineering 24 (1-2), 5180.
Van der A, D. A., Van der Zanden, J., O'Donoghue, T., Hurther, D., Cceres, I., McLelland, S. J.,
Ribberink, J. S., 2017. Large-scale laboratory study of breaking wave hydrodynamics over a xed bar.
Journal of Geophysical Research Oceans N/A, N/A.
Van der Zanden, J., Van der A, D. A., Hurther, D., Cceres, I., O'Donoghue, T., Ribberink, J. S., 2016.
Near-bed hydrodynamics and turbulence below a large-scale plunging breaking wave over a mobile
barred bed prole. Journal of Geophysical Research Oceans 121 (8), 64826506.
Van der Zanden, J., Van der A, D. A., Hurther, D., Cceres, I., O'Donoghue, T., Ribberink, J. S., 2017.
Suspended sediment transport around a large-scale laboratory breaker bar. Coastal Engineering 125,
5169.
Van Gent, M. R. A., 1995. Porous ow through rubble mound material. Journal of Waterway, Port,
Coastal and Ocean Engineering 121 (3), 176181.
Van Gent, M. R. A., Herrera, M. P., Molines, J., Jacobsen, N. G., 2015. Rock slopes on top of sand:
Modelling of open lters under wave loading. Proceeding to Coastal Structures, 112.
Van Gent, M. R. A., Tnjes, P., Petit, H. A. H., Van den Bosch, P., 1994. Wave action on and in
permeable structures. Proceeding to Coastal Engineering Conference, 17391753.
Vukcevi, V., Jasak, H., Malenica, S., 2016a. Decomposition model for naval hydrodynamic applications,
Part I: Computational method. Ocean Engineering 121, 3746.
Vukcevi, V., Jasak, H., Malenica, S., 2016b. Decomposition model for naval hydrodynamic applications,
Part II: Verication and validation. Ocean Engineering 121, 7688.
Wellens, P. R., January 2012. Wave Simulation in Truncated Domains for Oshore Applications. Ph.D.
thesis, Technical University of Delft.
Wellens, P. R., Borsboom, M. J. A., Van Gent, M. R. A., 2010. 3D simulation of wave interaction with
permeable structures. Proceeding to Coastal Engineering Conference, 115.
Zelt, J. A., Skjelbreia, J. A., 1992. Estimating Incident and Reected Wave Fields Using an Arbitrary
Number of Wave Gauges. Proceeding to Coastal Engineering Conference I, 777789.
Zhou, Z., Hsu, T.-J., Cox, D., Liu, X., 2017. Large-eddy simulation of wave-breaking induced turbulent
coherent structures and suspended sediment transport on a barred beach. Journal of Geophysical
Research Ocean 122 (2016JC011884).
APPENDIX A
Overview of Input Files
There is a number of les necessary to make a successful setup of a wave simulation with waves2Foam.
Some of these les are native to OpenFoam, while others are additional input les (pre- and postpro-
cessing and run-time). All les are briey discussed in this chapter to give the user an overview of the
origin and meaning of the les.
Roughly speaking, the les placed in the system-folder are related to the time control and numerical
schemes of the simulation, while the les placed in the constant-folder are related to the physical
properties of the simulation.
The following les are native to OpenFoam:
system-folder:
controlDict
fvSchemes
fvSolution
constant-folder:
dynamicMeshDict
g
transportProperties
turbulenceProperties
RASProperties
LESProperties
porosityZones
The following les are additional les native to waves2Foam:
system-folder:
None
constant-folder:
waveProperties.input
waveProperties
probeDefinitions
postProcessingProperties
wavesPorosityProperties
triSurface/stlDefinitions
In section A.1 les native to OpenFoam are described, while les native to waves2Foam are described
in section A.2
A.1 Files native to OpenFoam

A.1.1 system/controlDict
The controlDict controls the duration of the simulation, the output interval and the adaptive time step
(if activated). This is also the le, where function objects are activated in the simulation. Function
objects allow for run-time evaluation of additional properties: The total force on a structural element,
the surface elevation, or the pressure in specic locations with the probe tool.
61
62 Manual v. 0.9 (Draft) - A Overview of Input Files
A.1.2 system/fvSchemes
fvSchemes controls the numerical schemes used to discretise the system. This relates to both the temporal
and spatial discretisations.
A.1.3 system/fvSolution
fvSolution controls the solution procedure for the linear systems of equation. This includes the spec-
ication of solution tolerances, linear solvers, under-relaxation and the number of iterations in the
pressure-velocity coupling.
A.1.4 constant/dynamicMeshDict
The dynamicMeshDict controls mesh motion, dynamic mesh renement or topological changes. The
keywords in this le depends on the avour of the OpenFoam-branch.
A.1.5 constant/g
This les is used to dene the direction and magnitude of the acceleration due to gravity.
A.1.6 constant/transportProperties
transportProperties is used to dene the viscosity and density of the uids. Also, the surface tension
coecent between two uids is dened in this le. In the case of free surface waves, the surface tension
is typically neglected.
A.1.7 constant/turbulenceProperties
turbulenceProperties controls, whether a laminar, RANS or LES turbulence closure is applied in the
simulation; see section 2.4 for more details on turbulence modelling and waves.
A.1.8 constant/RASProperties
If a RANS-type turbulence closure is used, the actual turbulence closure is specied in RASProperties.
A.1.9 constant/LESProperties
If a LES-type turbulence closure is used, the actual turbulence closure is specied in LESProperties.
A.1.10 constant/porosityZones
The resistance properties of porous zones are specied in porosityZones.
A.2 Files related to waves2Foam

A.2.1 constant/waveProperties.input
waveProperties.input is the main input le for wave modelling. Here, the specications are given in
terms of wave theory, environmental properties, etc.
A.2.2 constant/waveProperties
waveProperties is a result of pre-processing of waveProperties.input. It can also be generated by
hand, but it is generally discouraged.
A.2.3 constant/probeDefinitions
probeDefinitions is used to specify the locations of wave gauges and probes.
A.2.4 constant/postProcessingProperties
postProcessingProperties is used to specify post-processing steps, e.g. zero-crossing analysis of the
evaluation of the reection of a regular wave.
A.2.5 constant/triSurface/stlDefinitions
This le is used to dene simple STL-surfaces. STL-surfaces may be used for the meshing with
snappyHexMesh or for the denition of porous zones.
APPENDIX B
Treatment of Irregular Waves
This appendix discusses the impact of two dierent methods to discretise the frequency axis for a
wave spectrum. In addition to this, the appendix describes the irregularFast approach to evaluate
an irregular wave signal (Section 4.3.2). The combination of these two components oers a way to
accelerate the evaluation of irregular incident waves; the method diers from the work by Dimakopoulos
et al. (2016), who applied Taylor expansions of the trigonometric functions.
The content of this work is realised in collaboration with Bo Terp Paulsen. The content has not been
published elsewhere. The style of the appendix has the avour of a journal contribution as this was
the original ambition, however, the lack of time prevented the nalisation of the manuscript, why it is
adapted to the manual for future reference.
B.1 Introduction
In the eld of coastal engineering it is common practice to evaluate the design of coastal structures by
means of physical experiments and numerical modelling. For the representation of a natural sea state,
irregular waves are described by a spectral distribution of the wave energy. Several wave spectra based on
sea state measurements have become standard practice both for physical experiments and for numerical
modelling, e.g., the JONSWAP spectrum and the Pierzon-Moscowitch spectrum.
For numerical models, the volume-of-uid (VOF) type models are often applied to track the free water
surface in combination with Navier-Stokes models. Also Boussinesq type models and non-linear potential
wave models are applied when a numerical approach is taken towards the evaluation of wave-structure
interaction.
In order to generate an irregular sea state based on a wave spectrum, both experimentally and nu-
merically, the incident wave spectrum must be discretised into a number of discrete frequencies. Each
discrete frequency represent a regular wave component, and the superposition of all discrete regular wave
components forms the irregular sea state.
In physical laboratory facilities the spectral discretisation is typically chosen as f = 1/Te , where f is
the frequency and Te is the duration of the experiment. Hereby an equidistant discretisation is achieved
which adds the same number of regular wave components both at the part of the wave spectrum that
contains a relatively high amount of energy and a low amount of energy.
This approach can also be used for the numerical wave modelling, however, when the combined wave
generation and reection compensation is achieved with the use of relaxation zones (see e.g. Jacobsen
et al., 2012) it become computationally expensive to evaluate thousands of wave components at every
time step. For practical applications a limited number of discrete wave components are usually applied,
e.g., in the order of 100 components. Therefore it becomes important how the discrete components are
distributed along the frequency range. As will be shown, both the discretisation and the total number
of wave components aects the return period of the signal and the wave height distribution.
First, some mathematical considerations are presented with respect to the return period of the time
series as evaluated from a spectrum with a given frequency discretisation. Secondly, the choice on
frequency discretisation and resolution is studied with an emphasis on the integrated spectral properties
and the exceedance distribution of the wave heights and wave periods. Finally, an alternative approach
to evaluate the irregular wave signal is presented, which accelerates the execution of the solvers in
waves2Foam.
B.2 Mathematical description

A set of generated irregular incident waves must avoid a repetition of the signal within the modelled
duration of the experiment (physical or numerical). The period of repetition of the signal can be evaluated
63
64 Manual v. 0.9 (Draft) - B Treatment of Irregular Waves
by a calculation of the autocorrelation; either using the time series of the surface elevation or by applying
the Wiener-Khinchin theorem on the real-valued spectrum:
1
Z
R( ) = S(f ) cos 2f df (B.1)
m0 0
Here R( ) is the autocorrelation function, S(f ) is the spectrum, f is the frequency, is the correlation
time and m0 is the zero order moment of the spectrum.
The discrete version of Eq. (B.1) is given as follows using the trapezoidal integration rule:
N 1
1
(B.2)
X
Rd ( ) ' [Sd (fn ) cos(2fn ) + Sd (fn+1 ) cos(2fn+1 )] fn
2md,0 n=0
Here, sub-index 'd' refers to the discrete values. The discretisation is such that fn = fn+1 fn . N is
the number of discrete wave components in the time domain representation of the surface elevation (see
below).
The spectral discretisation of the surface elevation is transformed into the time domain by the
following linear super-position:
N
(B.3)
X
(t, x) = am cos (m t km x + m )
m=1
am is the amplitude corresponding to the cyclic frequency m , km the wave number vectors, x is the
Cartesian coordinate vector and m the random phases in the interval [0, 2]. Here,
(B.4)
p
am = [Sd (fm1 ) + Sd (fm )]fm1
and
m = (fm + fm1 ) (B.5)
B.2.1 Repetition of the signal

The signal is repeated, when > 0 s and all of the cosines are in phase, i.e. cos(2fn ) = 1 for all
values of n. If the period of repetition, Tr , is small relative to the mean wave period, Tm01 , it means
that there can only be a limited number of wave heights in the signal. Consequently, the wave height
distribution will deviate considerably from a Rayleigh distribution; especially for small values of the
exceedance probability. The eect of the discretisation of the spectrum on Tr is considered here for an
equidistant and a simple (linear) non-equidistant discretisation.
B.2.1.1 Equidistant discretisation

The equidistant discretisation is given by fn = nf0 and fn = f0 for n = 0, 1 . . . , N . It is easy to
obtain the well established relationship
1 N
Tr = = (B.6)
f0 fN
by inserting the equidistant discretisation into Eq. (B.2). fN is the upper cut-o frequency. It is seen
that the larger N (or smaller fN ), the larger the value of Tr .
B.2.1.2 Non-equidistant discretisation

A simple non-equidistant discretisation is considered in this section, namely
fn = (1 + nf )f0 (B.7)
Here

n+1
fn = nf0 + f f0 n 1 = f0 (n + f An ) (B.8)
2
for 1/(1 N ) < f (the equidistant discretisation is re-found for f = 0). Inserting this discretisation
into Eq. (B.2), the cosine terms can be expanded as follows:
cos(2fn ) = cos(2nf0 ) cos(2f An f0 )
(B.9)
sin(2nf0 ) sin(2f An f0 )
It is observed that the autocorrelation for the equidistant discretisation has been overlayed by a second
harmonic, which diers for each value of n. In order to obtain a repetition of the signal it must be
required for the cosine terms that
nf0 = K and f An f0 = L (B.10)
at the same time; here the sinusoidal terms are ignored, because they are out of phase with the cosine
terms. K and L are integers. Since is identical in both equalities, it is found that
K L
f = 1 (B.11)
n An
Manual v. 0.9 (Draft) - B.3 Spectral discretisation 65
Consequently, if is not a rational number, there will never be a repetition of the signal, when a
repetition is dened as the value of for which Rd = 1.
The inuence of f on Tr is exemplied in the following. For = 1/4, Tr = 1/f0 , while Tr = 4/f0
for = 4. Note, however, than the value of f also aects the magnitude of f0 , see Eq. (B.8), such
that f0 decreases with increasing f .
This analysis has shown that Tr can be aected strongly by the discrete representation of the spectrum.
This attribute is further analysed in the following sections. The simple linear expansion of the frequency
axis (Eq. (B.7)) will, however, not be analysed further, because it cannot eciently represent the peak
of the spectrum, which is typically in the middle of the discretised frequency interval.
B.2.2 Discrete representations

In the following, two discrete representations of the spectrum S are analysed with respect to the repetition
of the wave signal, the integrated wave properties and the exceedance distribution of the wave heights
and wave periods. The discrete representations are described in the following:
B.2.2.1 Equidistant
The equidistant discretisation is already presented aboute in Section B.2.1.1. This discretisation is
included in the analysis, because it is the commonly adopted discretisation.
B.2.2.2 Cosine stretching

The cosine stretching is a non-equidistant discretisation of the frequency axis, such that the discretisation
is ner adjacent to the peak of the spectrum. This approach is simple to prescribe, but it is dicult to
apply on e.g. double-peaked spectra. The method works with a lower, fL , a peak, fP and an upper, fU ,
frequency.
For the interval [fL , fP [ the spectrum is discretised with the following number of frequencies:

fP fL
NL = (N + 1) (B.12)
fU f P
The number of frequencies in the interval [fP , fU ] is consequently NU = N + 1 NL . The frequency
discretisation is given as follows for the lower part of the spectrum

2
fn = fL + (fP fL ) sin n for n = 0, . . . NL 1 (B.13)
4NL
while the upper part of the frequency discretisation is given as

2
fNL +n = fP + (fU fP ) 1 cos n for n = 0, . . . NU (B.14)
4NU
B.2.2.3 The discrete representations

The variation in f with f for the applied discretisations is depicted in Figure B.1A. Here, fL = 0.03
Hz, fU = 1.4 Hz, Tp = 3 s and N = 200.
The discrete representation of a JONSWAP spectrum with a peak enhancement factor of 3.3 with
N = 35 is plotted in Figure B.1B for the two discretisation methods. It is clearly seen that for a
small N , the equidistant discretisation cannot capture the peak, whereas the non-equidistant method
is tailored to capture the peak. There are no marked visual dierences for N = 200 between the two
discretisation methods.
B.3 Spectral discretisation

B.3.1 Time series and autocorrelation
The time series of the surface elevation and autocorrelation for a spectrum discretised with N = 20 are
discussed in this section. The remaining parameters are the same as in Figure B.1 and the spectrum is
a JONSWAP spectrum with a peak enhancement factor of 3.3. The small value for N was chosen for
visual reasons. The surface elevation is depicted for the two discretisations in Figure B.2 and while it
is easy to observe a repetition in the signal for the equidistant discretisation (Figure B.2A), there is no
obvious repetition of the irregular wave signal for the rst 100 s for the non-equidistant discretisation.
The diculty in observing a repetition of the surface elevation in the signals based on a non-equidistant
frequency discretisation was addressed by evaluating the autocorrelation using Eq. B.2. The autocorre-
lations are depicted in Figure B.3 for the spectral discretisations in Figure B.2. It is appearent that while
Tr ' 15 s for the equidistant discretisation, then Tr is much larger for the non-equidistant discretisation
with the same value of N . The value of |Rd | hardly exceeds 0.5 during the rst 500 s of the autocor-
relation, i.e. an increase in Tr by a factor of 30 was obtained through a simple change in discretisation
method. The eect of the discretisation method on the distribution of the individual wave heights and
wave periods is presented in Section B.3.3
A. 10 -1
10 -2
f, [Hz]
10 -3
10 -4
0 0.2 0.4 0.6 0.8 1 1.2 1.4

f, [Hz]
B.
Equidistant
0.6 Cosine stretching
S, [m2 s]
0.4
0.2
0
0.2 0.25 0.3 0.35 0.4 0.45 0.5
f, [Hz]
Figure B.1: The variation of f as a function of f for the two discretisations of the frequency
axis. fL = 0.03 Hz, fU = 1.4 Hz, Tp = 3 s. A: N = 200. B: N = 35 and a
JONSWAP spectrum with a peak enhancement factor of 3.3.
B.3.2 Spectral wave properties

The validity of the newly proposed discretisation methods of the spectra requires that the spectral wave
properties converge. The convergence of the spectral wave properties is analysed. The evaluated wave
properties are the spectral wave height H m0 = 4 m0 and the three wave periods Tm01 = m0 /m1 ,
Tm02 = m0 /m2 and Tm10 = m1 /m0 ; they are evaluated as a function of the discretisation and N .
p
Here, mi is the i'th moment of the spectrum and it is dened as

Z
mi = f i S(f )df (B.15)
0
Its discrete approximation was evaluated by a trapezoidal integration.

The utilised spectrum was the same as the one dened in conjunction with Figure B.1. A range of N
from 10 to 10,000 was used to calculate the spectra and the discrete wave properties. For N = 10, 000
the spectral wave properties deviated less than 0.1% between the discretisation methods and the spectral
wave properties for N = 10, 000 were adopted as the converged solution.
The results are depicted in Figure B.4 and it is clear that even for N as small as 30, an accuracy
within a few percentages was obtained (for this specic choice of spectrum). The noise in the equidistant
discretisation was attributed to the discretisation of the peak.
B.3.3 Probability distribution

The integrated spectral wave properties were seen to be insensitive to the spectral discretisation, but
what is the impact on the discretisation on the exceedance propability of the wave heights and wave
periods?
A time series analysis was conducted on 1,000 realisations of the same spectrum, i.e. 1,000 dierent
sets of random phases. This was done for the two frequency discretisations and four values of N : 100,
200, 500 and 1,000. The surface elevation signals were generated from a linear super-position as in Eq.
(B.3). The time series were generated with t = 0.02 s and they had a duration of Te = 5, 000 s, which
is much longer than Tr for all of the equidistant discretisations (Tr = 730 s for N = 1, 000).
The wave height distribution based on a zero-down crossing is presented in Figure B.5 and it is seen
that this distribution is very sensitive to N for the equidistant stretching, whereas the non-equidistant
discretisation is less sensitive to N . It is seen that even for as few as N = 100, the non-equidistant
discretisation is close to the Rayleigh distribution for an exceedance probability of 0.1%, while the
equidistant discretisation does not even represent the 1.0% exceedance probability for N = 100. There
is an uncertainty band around the mean exceedance distribution for all three discretisation methods,
Manual v. 0.9 (Draft) - B.3 Spectral discretisation 67
A. 1
0.5
, [m]
0
-0.5
-1
0 20 40 60 80 100
B. 1
0.5
, [m]
-0.5
-1
0 20 40 60 80 100
t, [s]
Figure B.2: One realisation of the surface elevation. A: Equidistant discretisation. B: Cosine
stretching.
A. 1
R, [-]
-1
0 10 20 30 40 50
B. 1
R, [-]
-1
0 100 200 300 400 500
t, [s]
Figure B.3: The discrete autocorrelation of a given spectral discretisation. A: Equidistant dis-
cretisation. B: Cosine stretching.
and the uncertainty band is dened by one standard deviation . is the largest for the equidistant
discretisation.
The results from the zero-crossing analysis were also used to evaluate the distribution of the wave
periods. The distribution of the wave periods are depicted in Figure B.5 as well. The probability distri-
bution of the wave periods led to the same conclusions, namely that the non-equidistant discretisation
produces a time series signal that is less sensitive to the number of wave components.
A. 1.02
Rel. Hm0 , [-]

1
0.98
10 1 10 2 10 3 10 4
B. 1.02
Rel. Tm01 , [-]
0.98
10 1 10 2 10 3 10 4
C. 1.02
Rel. Tm02 , [-]
0.98
10 1 10 2 10 3 10 4
D. 1.02
Equidistant
Rel. Tm10 , [-]
Cosine stretching
0.98
10 1 10 2 10 3 10 4
N, [-]
Figure B.4: The variation in the spectral wave properties as a function of N. 'Rel.' means
normalisation with the quantity for N = 10, 000.
B.3.4 Discussion
It appears straightforward that the non-equidistant frequency discretisation has many benets over the
equidistant discretisation. This is true, if the main focus is to produce lightweight time series, which
exhibit the correct exceedance properties. However, if the analysed system has resonance frequencies
within the wave frequency band, then one should be careful in applying the lightweight discretisation
without critical analysis of the results and sensitivity to the number of frequency components.
B.4 Accelerated evaluation of irregular waves

One element of a fast evaluation of an irregular wave signal is the non-equidistant discretisation as
discussed above. A second element is a formulation of the problem, where a signicant part of the work
can be done as a pre-processing step.
First of all, consider the simple superposition of the free surface elevation (direct evaluation):
N
(B.16)
X
= 0 + an cos(n t kn x + n )
n=1
Here, is the surface elevation, 0 is the oset from still water level, N is the number of components,
is the cyclic frequency, t is time, k is the wave number vector, x is the coordinate and is a phase. It is
apparent that the sum (Eq. B.16) must be evaluated for every coordinate in the relaxation zone every
time step, since the cosine depends on both space and time. It is known that trigonometric function are
notoriously expensive to compute.
A simple manipulation of the equation leads to this expression:
N
(B.17)
X
= 0 + an [cos(n t + n ) cos kn x sin(n t + n ) sin kn x]
n=1
Manual v. 0.9 (Draft) - B.4 Accelerated evaluation of irregular waves 69
Equidistant Cosine stretching

10 0 N = 100 10 0 N = 100
Exceedance
Rayleigh
10 -2 10 -2
10 -4 10 -4
0 2 4 6 8 0 2 4 6 8
10 0 N = 200 10 0 N = 200
Exceedance
Wave height
10 -2 10 -2
10 -4 10 -4
0 2 4 6 8 0 2 4 6 8
0 0
10 N = 500 10 N = 500
Exceedance
Wave period
10 -2 10 -2
10 -4 10 -4
0 2 4 6 8 0 2 4 6 8
0 0
10 N = 1000 10 N = 1000
Exceedance
10 -2 10 -2
10 -4 10 -4
0 2 4 6 8 0 2 4 6 8
(H/Hrms )2 , (T/Tp )2 , [-] (H/Hrms )2 , (T/Tp )2 , [-]
Figure B.5: The probability distribution for both (H/Hrms )2 and (T /Tp )2 as a function of the
discretisation of the frequency axis (columns) and the number of wave components
(rows). Full lines is the mean values and dashed lines the mean .
This shows that it is possible to reformulate Eq. (B.16) such that each trigonometric function only
depends on space or time. Consequently, all trigonometric functions dependent on x can be evaluated
as a post-processing step prior to the time stepping. All trigonometric functions dependent on t must
be re-evaluated each time step (before relaxation), but it is cheap, as there are merely 2N evaluation,
while the number of spatial coordinates is typically much, much larger.
A similar splitting can be performed for the velocity vector, where the vertical coordinate (z ) is already
split from the remainding arguments: t, x and y .
At rst glance, Eq. (B.17) may look more computationally heavy, but since all components are pre-
computed, the nal summation is considerably faster, because multiplication and memory lookup is
outperforms evaluation of trigonometric and hyperbolic functions.
This approach is termed the split method in the following.
B.4.1 Implementation
The split method in waves2Foam IS termed irregularFast (Section 4.3.2). The implementation is
based on the external wave theories class (Section 4.3), which allows for the necessary pre-processing
steps within a time step.
The implementation is divided into three components, namely (i) global pre-processing step, (ii) local
pre-processing step within a time step, and (iii) evaluation of boundary conditions and relaxation zones:
Global pre-processing
Make a list of the unique x, y and z coordinates.
Evaluate all trigonometric and hyperbolic functions in these coordinates for all frequqency
components and store the values. These lists are eectively matrices with dimensions of the
number of coordinates and the number of frequencies.
Local pre-processing
Evaluate the trigonometric functions at the current time instance and store the values. These
lists are column vectors with the length of the number of frequencies.
Evaluate the surface elevation in all combinations of (x, y) (assume z to be vertical), since
the surface elevation is required for the relaxation zone evaluation. This step reduces the
computational requirements further.
Evaluation
Look-up the pre-calculated values and perform the necessary multiplications.
The method is only implemented for static grids, but there is a ag to ignore mesh motion, if it is
known that the mesh deformation in the relaxation zone is small (Section 4.3.2).
B.4.2 Example
An example of the computational gain is presented in Figure B.6. The execution time was measured for
both the direct evaluation of the irregular wave signal (see Eqn. (B.16)) and the proposed split method
(see Eqn. (B.17)). Here, the relative execution time was obtained by normalising by the execution time
for N = 1.
It is seen that there is a steady increase in time consumption for the direct method, while the split
method at-lines up to N = 100 (Figure ??). The at-lining essentially shows that the costs related to
the relaxing method is more expensive that the evaluation of the wave signal itself.
A. 10 3
Direct method
Split method
TN /TN =1 , [-]
10 2
10 1
10 0
10 0 10 1 10 2 10 3 10 4
B. 30
Direct over split, [-]
20
10
0
10 0 10 1 10 2 10 3 10 4
N, [-]
Figure B.6: A: The execution time of the irregular wave summation relative to the execution
time for N = 1. B: The increase in execution time between the direct and the split
methods.
The decrease in computational requirements between the two evaluation methods is a factor of 5 for
N = 100 and the requirements decreased by more than an order of magnitude for N = 1, 0001 . Conse-
quently, if 1,000 wave components are required for the generation of an irregular wave with equidistant
discretisation of the spectrum, then a computational gain of 50 can be obtained by utilising a non-
equidistant frequency discretisation in combination with the split method. Besides a faster execution
of the numerical wave ume, this will also allow for the inclusion of second order eects in the incident
wave eld (Sharma and Dean, 1981; Madsen and Fuhrman, 2012), which would otherwise have been
prohibitively expensive, because the number of second order components is N 2 .
1 It was observed that these ratios are hardware and/or compiler dependent.
APPENDIX C
Source Code History
This list is a reproduction of the information given at up to the revision covered by this document (see
front page).
Table C.1: Summary of all code changes given per revision
Date Revision Description

2017-07-23 19:46:17 2113 Continue adding pressures to wave theories: Now chappelear and cor-
rects for still water level
2017-07-15 22:07:26 r2112 Added solitary wave theory (Chappelear) and xed bug in solitaryFirst
2017-07-14 18:01:19 r2111 Adding pressure distribution to stokesSecond
2017-05-17 17:12:03 r2107 Added a small interface correction to waveTheory.H
2017-05-07 21:53:13 r2106 Added utility to compute the (analytical) wave eld in a multiple of
points: sampleIncidentWaveField
2017-04-09 20:23:26 r2105 Added breeder version of catenary mooring line (v1606). Also coded
pressure to StokesFirst
2017-01-15 18:50:23 r2104 Support for foam-extend-4.0
2016-11-08 18:41:24 r2103 waves2Foam available for v1606+
2016-10-13 16:00:22 r2102 Preparing for better treatment of KC; no functional change
2016-09-27 07:31:45 r2101 porousWaveFoam for OpenFoam-v3.0+ (only explicit advection of al-
pha.water)
2016-08-12 15:48:56 r2100 Added compatibility with OpenFoam-v3.0+ and solvers for foam-extend-
3.2
2016-07-08 10:40:35 r2096 Minor change to pEqn.H for foam-extend-3.1 to support on-going devel-
opments. No functional change.
2016-07-01 22:41:30 r2095 Added support for OpenFoam-4.0.0. porousWaveFoam not created and
sampling library broken in OpenFoam-4.0.0
2016-05-19 22:05:52 r2094 Support for stream function wave theory
2016-05-19 10:27:57 r2093 Debug overtopping: results of rhoU2A were bugged, not the overtopping
rate itself.
2016-04-29 11:11:26 r2092 Minor cosmetic changes. Only functional change: Extruded faceSet-
ToSTL always have outward pointing normal.
2016-04-04 14:52:42 r2091 Changed source location for OceanWave3D. New version resolves bugs
with respect to wave generation with the use of a paddle signal. Other-
wise no functional eects are expected.
2016-04-04 09:48:18 r2090 Resolved need for dual compilation, if WAVES_LIBBIN does not
exist at the beginning of the compilation. WAVES_LIBBIN and
WAVES_APPBIN are made manually.
2016-03-08 11:05:46 r2089 Minor downstream dependencies. No functional change
2016-02-03 10:58:03 r2088 Minor bug corrected in compilation procedure; eect on foam-extend
branch
2016-02-02 10:50:34 r2087 Corrected minor bug in oceanWave3D-coupling.
2016-01-08 17:42:13 r2086 Update tutorial porousDamBreak + porousWaveFoam for OF2.3.0
2015-12-18 18:29:10 r2085 Added time control checks to oceanWave3D.C
2015-11-23 07:27:38 r2084 Added solver for OpenFoam-3.0.0
2015-11-20 18:50:30 r2083 Preparation of compatibility with OpenFoam-3.0.0 - do not use this re-
vision
2015-10-15 22:09:38 r2082 Debugging of ThirdParty/Allwmake compile script
2015-10-15 21:17:44 r2081 Additional of OceanWave3D coupling + related tutorial + documenta-
tion in 'doc'
2015-10-01 20:00:40 r2080 src+utilities compiles with foam-extend-3.2. Solvers to follow
2015-09-16 09:27:48 r2079 Minor issues with options and le for solvers
2015-09-09 16:07:37 r2078 Added the possibility of using the sea level as vertical reference. Tuto-
rials and setWaveParameters modied accordingly.
2015-08-24 17:10:42 r2077 Added '- phi' to the argument in stokesFifth.C
2015-07-27 09:44:25 r2073 Minor changes related to external wave forcing (e.g. nicer closing in
solver + works with setWaveField)
2015-07-12 13:48:09 r2072 waves2Foam compiles on OF2.4 + waveFoam solver
71
72 Manual v. 0.9 (Draft) - C Source Code History
2015-06-06 18:22:46 r2068 Added zero-crossing functionality + porosity model can be used for non-
wave related things
2015-04-14 21:24:03 r2066 Streamlined the code (removed duplication) for the setWaveProperties
and related classes. Change in input format in waveProperties.input.
2015-03-25 09:05:40 r2064 surfaceElevation now gives correct results on moving meshes in runTime
(tested on foam-extend-3.1). Increase computational time.
2015-03-16 14:38:06 r2063 Small change to porosityCoecients. NO functional change. Also, cor-
rected bug in surfaceElevation.C, which aected foam-extend branch.
2015-03-12 15:14:46 r2062 Small change in the interface for the porosity module. No functional
change.
2015-02-05 13:28:26 r2060 Added porousWaveFoam for foam-extend-3.1
2014-10-30 11:01:34 r2057 Resolving naming match between waves2Foam and 2+ OF-versions [DO
NOT USE REVISION 2056]
2014-10-30 10:18:49 r2056 Made the porosity runTime-selectable. Currently only one method avail-
able. Commit includes derived modications.
2014-10-24 22:03:27 r2055 Directional check on wave directions in wave theories. 3Dwaves tutorial
corrected accordingly
2014-09-09 14:41:16 r2051 Small interface change in the external wave classes (constructs now with
fvMesh as well)
2014-09-04 14:43:15 r2050 Modied options les so the compilation works with local installations
of GSL
2014-09-04 11:36:01 r2049 Changes to the output location of some applications and libraries. Now
uses WAVES_APPBIN/WAVES_LIBBIN
2014-06-21 18:04:47 r2045 Compiling on foam-extend-3.1, waveFoam 3.1 and waveFoam/-
porousWaveFoam added directly from OF2.2.1 to OF2.2.2
2014-05-31 22:02:48 r2044 waveFoam support to foam-extend 3.0 + minor bug-xes + new relax-
ation zone shape
2014-05-17 11:28:34 r2043 Minor maintanence, i.e. waveProperties.input in waveFlume tutorial.
2014-04-12 17:35:52 r2042 Corrected OF2.3.0 issue with the test case bejiBattjes
2014-04-06 12:33:09 r2041 Release to facilitate planned future open-source contribution(s) [No
functional change in execution].
2014-03-27 21:17:38 r2040 Additional corrections to the tutorials to run in 2.3.0. Also corrected
periodicSolitary, so the tutorial runs for recent cyclic denition.
2014-03-26 19:22:46 r2039 Corrected tutorials, such that they can actually execute (adding the
0.org-directory)
2014-03-20 11:11:40 r2038 Added a single post-processing functionality (writeIndexLocation)
2014-03-11 20:57:55 r2037 Updated structure of tutorials to easier accommondate future changes.
Now runs for OF2.3 with alpha.water.
2014-03-01 19:05:46 r2036 Adjusted to be used with OF2.3. Furthermore, the porousDamBreak
case now oers comparison with experiments. The tutorials are not yet
adjusted to work with the changes from OF2.2 -> OF2.3.
2014-01-26 16:13:27 r2026 Compatibility of waves2Foam libraries and utilities with foam-extend-
3.0. Still no solver (waveFoam)
2013-12-28 11:43:19 r2025 Porosity module with solvers (OF1.6, 2.1.0, 2.1.1, 2.2.0, 2.2.1) and a
single tutorial.
2013-11-25 19:35:06 r2024 Modied the fatal error message in the compilation, such that it is easy
to understand how to resolve the issue with a wrong WAVES_DIR en-
vironmental variable.
2013-10-24 09:11:15 r2022 Added a new relaxationScheme and a new relaxationShape
2013-10-07 20:36:24 r2021 Modied src/waves2Foam*/Make/options.* to include XVERSION ag
2013-09-15 09:37:28 r2020 Added phase focusing as an option for irregular waves.
2013-09-05 13:31:41 r2019 Trivial bug-x, not aecting results (Thanks: Bjarne Jensen)
2013-09-03 10:38:12 r2018 Increased exibility for setting wave spectra for JONSWAP.
2013-08-12 14:40:19 r2017 Minor bug-x in src/waves2Foam/waveTheories/irregular/irregular/-
irregular.C
2013-08-07 23:34:24 r2016 Faster evaluation of irregular waves
2013-08-06 21:59:56 r2015 A bit of code maintenance/simplication
2013-07-28 12:15:35 r2014 Fourth revision on code style change according to:
www.openfoam.org/contrib/code-style.php. [No functionality change]
2013-07-27 22:03:15 r2013 Third revision on code style change according to:
2013-07-27 19:31:35 r2012 Second revision on code style change according to:
2013-07-22 22:53:39 r2011 First revision on code style change according to:
2013-07-15 13:09:08 r2010 Corrected missing les in src/waves2FoamProcessing/Make/les
2013-07-13 11:32:48 r2009 waves2Foam now compiles under OF2.2.1
2013-07-09 15:12:25 r2008 Added an additional point distribution (user-dened) for wave gauges
2013-07-07 19:28:45 r2007 Release of pre- and post-processing utilities, overtopping sampling and
modied tutorials.
2013-07-07 18:37:43 r2004- Temporary revision: Restructuring the src-directory
r2006
2013-07-07 17:41:18 r2003 Bug in bichromatic second.
2013-06-11 20:46:14 r2000 Resolve incompatibility between version 2.2.0 and 2.2.x for waveFoam
header les
2013-06-09 09:13:39 r1997 Added a small script for the creation of new wave theories in
waves2Foam/doc/templateWaveTheory
2013-06-08 21:11:23 r1996 Updated compilation system
2013-06-07 21:38:02 r1991- Temporary commit during updating to new SVN
r1995
2012-12-03 12:52:40 r1984 Added second order bichromatic waves. Set numerical beach type to
"Empty" as defalt.
2012-11-29 15:10:18 r1983 Added wave theory. Automatic enhanced write precision for cnoidal
waves parameters. Slight change to waveTheory interface.
73
2012-11-23 15:52:01 r1982 Changed setWaveParameters so (i) parameters are read from waveProp-
erties.input and written to waveProperties and (ii) it does not depend
on the mesh. The latter speeds up the execution (a lot!).
2012-11-18 21:48:17 r1981 Added a check for validity of second order stokes in the properties le
2012-10-31 15:29:27 r1980 Bug in compiling setWaveParameters for OFVERSION >= 17. Added
pre-processor statement
2012-10-31 14:38:18 r1979 (i) Bug in bichromaticFirstProperties.C. (ii) Cosmethics in the *Prop-
erties les, when writing waveProperties-le
2012-10-18 17:00:21 r1977 (i) Added solvers for 2.1. (ii) Changed the approach used by setWavePa-
rameters to write waveProperties le. (iii) Updated headers with publi-
cation data on the journal article. (iv) Added the le sourceCodeStruc-
ture_r1923.svg in the doc directory.
2012-10-17 12:04:01 r1976 Removed all pre-processing if's related to version 15. Altered the
OFVERSION pre-processor to work on e.g. 211 and 160 rather than
21 and 16.
2012-10-01 11:49:20 r1975 Modied the tutorials, so they are also running under OF2.1
2012-09-10 21:29:21 r1973 Corrected innite loop in the computation of the wave number, when
kh > 60 (extreme deep water).
2012-08-16 09:44:50 r1972 Added an option for a local sealevel in potentialCurrent
2012-07-13 20:51:52 r1969 Added the following: (i) runTime selection of relaxation zone weights,
(ii) a local correction to relaxation weight based on the local Courant
number, (iii) update of relaxationZoneLayout to show the weight and
(iv) added interface for numerical beach in UEqn.H but still not func-
tional.
2012-06-08 12:09:21 r1967 Added some post-processing utilities to be used in matlab and modied
the misc/matlab le structure. Minor bug-x in generateStreamFile.m
so the output le is consistent with the format in waves2Foam.
2012-05-16 10:44:43 r1966 Corrected bug in matlab script for stream function theory (previous
results correct, if the program could nalise!)
2012-05-15 12:01:18 r1965 Additional output time control for the surfaceElevation utility. Espe-
cially needed under functionObject functionality
2012-05-10 14:37:47 r1961 Added a optional starting time for the surface elevation sampling
2012-05-09 14:18:23 r1960 Added phase-lag (phi_) in streamFunction.C
2012-05-03 13:22:43 r1953 Added functionObject functionality for sampledSurfaceElevation. In-
cluded in the tutorial waveFlume.
2012-05-02 10:11:01 r1952 Modication to Allwmake and src/Allwmake so "sed" also works under
Mac OSX.
2012-04-30 21:28:20 r1951 Added a post-processing utility, which can be used to extract the surface
elevation. As of now, version 1.5 is no longer supported.
2012-04-11 10:13:10 r1947 Bug in Allwmake script
2012-03-26 21:45:20 r1945 Changed the way setWaveProperties output irregular wave properties.
Adjusted irregular waveTheory accordingly.
2012-03-17 15:56:02 r1944 Solved bug in convexPolyhedral. Furthermore added an addition relax-
ation shape (semiCircular)
2012-03-13 16:13:42 r1943 Minor changes
2012-02-19 14:45:14 r1940 Added the possibility of a wind vector, which is constant in space and
time.
2012-01-29 19:15:41 r1938 Changed the Make/les and Make/options to have the output les in
FOAM_USER_LIBBIN and FOAM_USER_APPBIN.
2012-01-28 15:55:34 r1937 Added a Pierson-Moskowitz spectrum
2012-01-27 13:28:00 r1936 Added framework for irregular wave spectra - properties and wave theory
2012-01-24 13:20:58 r1935 Small changes to src/Make/les and the overall Allwmake script in order
to avoid the recompilation of waves2Foam on non-dev/ext versions of OF
2012-01-19 16:50:48 r1934 Modied the code, so PI can be used both in pre- and post-2.0 versions of
OpenFoam. It is replaced by PI_(4.0*atan(1.0)) a few neccesary places
as a protected member variable.
2011-12-04 20:07:35 r1932 Added a tutorial showing how to generate a standing wave from a fully
reecting sea wall
2011-11-22 10:02:44 r1929 Added a tutorial utilising the cylindrical relaxation zone
2011-11-19 01:01:19 r1928 Added "waveFoam" for 1.7, modied Allwmake script and added neces-
sary les to tutorials due to above change.
2011-11-16 19:22:22 r1937 setWaveParameters now also works for <tt>cnoidalFirst</tt>
2011-11-14 20:04:08 r1926 Added matlab tools for cnoidal and stream function waves
2011-11-14 17:52:52 r1925 Added a 3D wave tank tutorial
2011-11-10 11:30:30 r1923 Initial release of waves2Foam - all les
View publication stats

manualWaves2Foam v090

Uploaded by

Copyright:

Available Formats

manualWaves2Foam v090

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

manualWaves2Foam v090

Uploaded by

Copyright:

Available Formats

What is the purpose of this manual?

What is the purpose of this manual?

What mathematical models are described in the manual?

What mathematical models are described in the manual?

See

Technical Report August 2017

Niels Gjl Jacobsen

Generating waves for OpenFoam: waves2Foam View project

The user has requested enhancement of the downloaded file.

Front matter: Own picture (Ireland, 2016)

Date Version Latest SVN-revision Aliation

A Overview of Input Files 61

B Treatment of Irregular Waves 63

C Source Code History 71

2.1 Fundamental denition of a relaxation zone . . . . . . . . . . . . . . . . . . . . . . . 17

3.1 Common settings for denition of wave gauges and probes . . . . . . . . . . . . . . . 23

4.1 Settings Stokes First Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

6.1 Executing a tutorial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.1 Regarding estimate of overtopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

4.1 A note on phase focusing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3.1 Denition of wave gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

5.1 Online documentation to create waveDyMFoam. . . . . . . . . . . . . . . . . . . . . . . 45

1.2 Background and application

Jacobsen et al. (2015) validated waves2Foam against numerous laboratory experiments

OpenFoam (OF) as distributed through www.openfoam.org.

OpenFoam+ (OF+) as distributed through www.openfoam.com.

foam-extend (FE) as distributed through the foam-extend community.

The abbreviation in the parenteses are used throughout this document.

1.3.2 Naming terminology

1.3.3 Supported versions

OF+-3.0, OF+-1606, OF+-1612

FE-3.1, FE-3.2, FE-4.0

This issue has to be resolved in a future update of waves2Foam.

1.4 Installation of waves2Foam

1.4.1 Third-party dependencies

git is required to access the source code for OceanWave3D.

subversion (SVN) is required to access the source code for waves2Foam.

1.4.2 Download and installation

## First, source your OpenFoam version

1.4.3 The bashrc le

WAVES_DIR: This is the full path to the waves2Foam-installation.

WAVES_GSL_LIB: Path to the GNU Scientic libraries.

1.4.4 How to get the updates to waves2Foam?

Code Fragment 1.2: Updating the SVN-repository and recompilation of

Nota Bene 1.1: bashrc after updates

1.4.5 How to recompile parts of waves2Foam

Code Fragment 1.4: Reference to the paper on the porosity implementation in

Code Fragment 1.6: Reference to OceanWave3D.

2.2 Momentum and continuity equations

2.2.1 Modications for permeable structures

2.3 Advection of the indicator eld

Subscript 0 refers to material properties corresponding to F = 0 and subscript 1 refers to

2.3.1 Modications for permeable structures

2.4 Turbulence modelling

2.5 Reection compensation at the boundary

2.6 Numerical beach

2.7 Relaxation zone technique

2.7.1 Flavour of relaxation

= (1 wR )target + wR computed (2.8)

Date Version Latest SVN-revision Aliation

2.1 Fundamental denition of a relaxation zone . . . . . . . . . . . . . . . . . . . . . . . 17

3.1 Common settings for denition of wave gauges and probes . . . . . . . . . . . . . . . 23

3.1 Denition of wave gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

1.4.3 The bashrc le

WAVES_GSL_LIB: Path to the GNU Scientic libraries.

2.2.1 Modications for permeable structures

2.3 Advection of the indicator eld

2.3.1 Modications for permeable structures

2.5 Reection compensation at the boundary

Code Fragment 2.1: Fundamental denition of a relaxation zone

relaxationSchemeEmpty This method eectively turns of the relaxation (Code Fragment

where = 1 . The weight can be specied as in Code Fragment 2.6

Semi-cylindrical A semi-cylindrical shape can be dened based on the specications in Code

2.8 Specication of resistance due to a permeable structure

Warning 3.1: Denition of wave gauges

circularDistribution Denes a circular distribution with a certain radius around a given

quadrilateralDistribution Denes a quadrilateral point distribution, i.e. it is possible to

Code Fragment 3.5: User-dened point distribution

Code Fragment 3.10: Warning when Tsoft diers from 0