lheea / hos-nwt Goto Github PK

Press_press(ii) = 0.0_rp-0.0_rp*(zpress(ii))-0.5_rp*(vitx2_press(ii)+(phixadd_press(ii))**2 & +vity2_press(ii)+(phiyadd_press(ii))**2+vitz2_press(ii)+(phizadd_press(ii))**2 & + 2.0_rp*vitx_press(ii)*phixadd_press(ii) + 2.0_rp*vity_press(ii)*phiyadd_press(ii) & + 2.0_rp*vitz_press(ii)*phizadd_press(ii))-(phitd_press(ii)+phitadd_press(ii))

Second term on RHS: 0.0_rp*(zpress(ii))

Hello,

I am attempting to run a simulation through the input of the motion from a wave maker from a file. The reference data that I am attempting to compare with my simulation however was generated in a wave tank that has a double-hinged wave maker. Can the HOS-NWT code account for a double-hinged wave maker, maybe through the definition of two separate (d_hinge) rotation axis distances?

Thank You

Hello, I am a student from Hohai university. I have installed the ubuntu system and downloaded the HOS-NWT-1.2 package from this website, but I don't know how to run HOS-NWT. Maybe I am a beginner. Can you tell me the detailed steps for running this program？How to compile and link？I am really looking forward to your reply, thank you very much！

SIGFPE generated for input_HOS-NWT.dat in the the bin folder for iprobes=2.
No problems if iprobes=1
Program compiled with flags -fcheck=all -ffpe-trap=invalid,zero,overflow in gfortran 5.4.0 20160609

Console output just before the exception

CPU time for Output time step 2.76 9 0
volume = 2.3105E-04 t= 5.4000E+00 timestep: 1.1926E-02 486 0
H2 operator
KFSBC/DFSBC max. initial = 6.8966151695545354E-004 , reconstructed = 1.1740512391557623E-002
Max relative error = 0.22180840568248397 ... direct method

SIGFPE reads as

Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.

Backtrace for this error:
#0 0x7FAC399DDE08
#1 0x7FAC399DCF90
#2 0x7FAC3910E4AF
#3 0x4FF3CA in __velocities_MOD_creatematrix_coscos_sinh at velocities.f90:1517 (discriminator 56)
#4 0x5034B3 in __velocities_MOD_velocity_modes_direct at velocities.f90:1299
#5 0x5038EA in __velocities_MOD_hosvel2_direct at velocities.f90:1252
#6 0x45E8B6 in runge_kutta_MOD_runge4_adapt2 at runge_kutta.f90:342
#7 0x464D25 in MAIN at HOS-NWT.f90:520

input_HOS-NWT.txt

An error occurs when compiling HOS-NWT with gfortran 4.4
The error is relative to the definition of derived type (cf. type(wave_def):: wav in the code) and the possible necessity to use SAVE attribute.

Compilation is fine with later versions (tested with gfortran 4.9.x)

Hello,
Thank you for sharing this source to the public. I plane to use HOS-NWT to generate focused wave for my research, which is carried out at Osaka University Japan. I am using ifort to compile this program. To be honest, I am not familiar with the Linux/Unix environment, I have some problems with installation.

1. I can not find makedepf90 and where should I put this file? Here is some problem when compiling.

[earlyup@pc1-64hpc wave]$ make Debug -B
exists
exists
exists
rm -f obj/.depend
makedepf90 -b obj/ sources/main/HOS-NWT.f90 sources/main/wavemaking.f90 sources/main/filtering.f90 sources/main/dealiasing.f90 sources/main/resol_HOS.f90 sources/main/resol_wmkr.f90 sources/main/initial_condition.f90 sources/variabledef/variables.f90 sources/variabledef/common_vars.f90 sources/variabledef/type.f90 sources/wave/definition2.f90 sources/wave/config_def.f90 sources/wave/read_ocean_txt.f90 sources/wave/wave_def.f90 sources/wave/linear_wave.f90 sources/fourier/fourier_FFTW.f90 sources/timeintegration/runge_kutta.f90 sources/timeintegration/variablechange.f90 sources/IO/input.f90 sources/IO/vol_energy.f90 sources/IO/velocities.f90 sources/IO/output.f90 > obj/.depend
/bin/sh: makedepf90: command not found
make: *** [obj/.depend] Error 127

2. I plane to study the interaction between focused wave and a fixed FPSO by fully nonlinear potential flow theory. In such case, could I take the focused wave field as an explicit input and only compute the disturb wave field? Regarding embedding HOS-NWT on my time-domain BEM solver and the methodology itself, could you give some comments or advice? Thank you very much!

Line 310 in resol_wmkr.f90 reads as-

-5.d-1*(phix1st(i1,i2)**2+phiy1st(i1,i2)**2+phiz1st(i1,i2)**2)&

Is this correct? How does 5 enter the equation?

Hello,

I tried using the HOS-NWT Code. I encountered a problem while using the Post_processing program. I used the following input file:

    --- Choice of post-processing options
    VP-card output               :: i_card           :: 1
    Starting time analysis       :: T_start          :: 0.0
    Stoping time analysis        :: T_stop           :: 50.0
    --- Velocities/pressure cards
    Minimum x in VP-card         :: x_min            :: 0.0
    Maximum x in VP-card         :: x_max            :: 50.0
    Minimum y in VP-card         :: y_min            :: 0.0
    Maximum y in VP-card         :: y_max            :: 10.0
    Minimum z in VP-card         :: z_min            :: -5.0
    Maximum z in VP-card         :: z_max            :: 0.5
    Number of points in z        :: i_zvect          :: 20
    --- Input files
    Tecplot version              :: tecplot          :: 11
    Name of modal description    :: file_mod         :: ../../../bin/Results/modes_HOS_SWENSE.dat

The following error occurred:

  At line 58 of file output_post_process.f90 (unit = 31)
  Fortran runtime error: Cannot open file 'Results/VP_card.dat': No such file or directory

I took a look through the program and can't find this file as well. Is it supposed to be created in the process of running Post_processing?

Kind Regards

Felix

1. At first, I created a new folder named build in the cmake directory
2. And, I opened a terminal, changed the working directory into build, typed the command cmake .. -DCMAKE_BUILD_TYPE=Release
3. Everything was going well before I typed make, the information are as follows

Then I checked the code, I found that p2 has been defined as an integer parameter variable with value 1 in common_vars.f90, So the compiler reported an error.

recently, i want to couple the HOS-NWT with CFD software, this one-way coupling allows the mapping of a solution from one simulation to use as inputs for another simulation, via a CFD General Notation System (CGNS) file. it means the results (including mesh, time step, wave height, velocity, et.al) from HOS-NWT needs to be writed into CGNS format. how to do this? or how to transfer the .dat file into CGNS file?

Hello!
I'm currently working on a wave generation and propagation subject and so I'm really interested in your HOS method.
After doing several simulations with the attached benchmark cases and in 2D, I'm willing to use the swense Output as an input in an other software. My problem is that in 3D, I have non numerical results in the .dat, which make it unusable with the Post processing program.
Here is the input_HOS-NWT.dat :
----------------------- Tank dimensions -----------------------
Length(m) of the wavetank :: xlen :: 46.4d0
Beam(m) of the wavetank :: ylen :: 29.74d0
Depth(m) of the wavetank :: h :: 5.d0
------------------------ Select case --------------------------
Choice of computed case :: icase :: 31
------------------ Sloshing case (icase = 1) ------------------
Number of the natural mode :: islosh :: 3
Amplitude (m) of the mode :: aslosh :: 0.2d0
--------------- Monochromatic case (icase = 2) ----------------
Amplitude (m) of wave train :: amp_mono :: 0.1d0
Frequency (Hz) of wave train :: nu_mono :: 0.546875d0
Angle (deg) from x-axis :: theta_mono :: 0.d0
Phasis (rad) of wave train :: ph_mono :: 0.d0
Directional wmk type :: ibat :: 2
Wave target distance (ibat=3):: xd_mono :: 18.d0
-------------------- File case (icase = 3) --------------------
Name of frequency file :: file_name :: CAT1__22
Frequency cut-off :: i_cut :: 0
Low cut_off frequency (Hz) :: nuc_low :: 0.3d0
High cut_off frequency (Hz) :: nuc_high :: 3.d0
------------------ Irregular wave (icase=4) -------------------
Significant wave height (m) :: Hs :: 0.05d0
Peak period (s) :: Tp :: 1.d0
Shape factor (Jonswap) :: gamma :: 3.3d0
Seed number for random numb. :: iseed :: 513
-------------------- Wavemaker definition ---------------------
Nonlinear order of wavemaker :: i_wmk :: 2
Type (1: piston, 2: hinged) :: igeom :: 2
Rotation axis distance :: d_hinge :: 2.147d0
Time ramp :: iramp :: 1
Time ramp duration (s) :: Tramp :: 3.0d0
----------------------- Numerical beach -----------------------
Absorption numerical beach :: iabsnb :: 1
Beginning front num. beach :: xabsf :: 0.8d0
Absorption strength front :: coeffabsf :: 1.d0
------------- Elevation/Velocity-Pressure probes --------------
Use of probes :: iprobes :: 1
Filename of probe positions :: pro_file :: prob.inp
---------------------- Time-integration -----------------------
Duration of the simulation :: T_stop :: 10.0d0
Time tolerance: RK 4(5) :: toler :: 1.e-4
Output frequency :: f_out :: 10.d0
--------------------------- Output ----------------------------
Output: 1-dim. ; 0-nondim. :: idim :: 1
free surface plot :: i3d :: 1
modes plot :: imodes :: 0
wavemaker motion plot :: iwmk :: 1
Swense output 1='yes',0='no' :: i_sw :: 1

and the beginning of the common_vars file:
! Number of modes
INTEGER, PARAMETER :: n1 = 30
INTEGER, PARAMETER :: n2 = 30
INTEGER, PARAMETER :: n3 = 32
! HOS nonlinearity order
INTEGER, PARAMETER :: mHOS = 3
! Dealiasing parameters
INTEGER, PARAMETER :: p1 = mHOS
INTEGER, PARAMETER :: p2 = mHOS

I try to change the value of n1,n2 and n3 but I still have the same kind of outputs... Do you have an idea to help me solve this problem?
And thank you very much for your work and for sharing it!

Many thanks,

Max.

"Hello, I encountered an error while attempting to compile in a Windows environment using Visual Studio Code. What could be the cause of this error?"
makefile:109: .depend: No such file or directory
make: *** [makefile:146: .depend] Error 255

Hello,

I am the learner of the HOS-NWT. I am trying to simulate a 0.29m height and T = 3s (freq = 0.3333) regular wave. When I run the package, there is a SIGFPE error (floating-point exception) and it seems happened in routine wnk_ini. Anyone hase the same problem? How to deal this?

"librefblas.a" for you too?
follow my answer here:

https://stackoverflow.com/a/64260946

Hello

I'm using the HOS-NWT model for a while and I was wondering if there is some kind of upper-limit for the wave height? For instance I wanted to plot a wave field with a frequency of 1 Hz and a wave height of 0.5 m. This gave me some problems because the model didn't finished with calculating. I have added my input file as a picture. Thank you in advance!

Kind regards
Louis

For now icase=3 uses:
- file_name.cfg for description of wavemaker configuration
- file_name.dat for description of linear waves generated by wavemaker motion

It may be useful to allow the input of direct wavemaker motion (even if it is just a matter of multiplying by a transfer function?)

makedepf90 is a program for automatic creation of Makefile-style dependency lists for Fortran source code and is needed within HOS-NWT compilation, apparently the link of it on the wiki is not working, I tried to install it manually by the same other links but doesn't worth it, one needs to solve several other problem which is waste of time.
It is already on 'universe' repository of ubuntu, so go with this:

sudo apt install makedepf90

Hello,professsor
Recently I was working on the code you shared.I'd like to reproduce the wave time calendar,which is obtained by real Experimental data.What I'm curious about is how to control the wave basically focusing on 20 to 40 seconds.


I try to change Tp,Hs and gamma,but it doesn't work.Can you give me some guidance?

Hello recently I got an error about changing the value of order of HOS.
I decided to calculate a 2D irregular wave with mHOS=6, however, it didn't work and showed the error message with 'initiate_parameters: forbidden value of M.'
I tried to change p1 or wave condition but it still didn't work.
What should I do to deal with it?

What is the use of dphis_add_rkramp?

Line 151 resol_wmkr.f90 reads as

dphis_add_rkramp_loc(i3,i2) = - alphat*dpos1stdt(i3,i2) - alphatt*pos1st(i3,i2)

If this is velocity on the rhs then it should be-
dphis_add_rkramp_loc(i3,i2) = - alpha*dpos1stdt(i3,i2) - alphat*pos1st(i3,i2)

Is it used anywhere?

This is for gfortran 7.3.0 and relevant source file dealiasing.f90 line 59,72 and 77.

It seems that during compile time gfortran refuses to accept value p2=1. Complaining that for MOD intrinsic fortran 77 function that

 IF (((p1 /= 1).AND.((MOD(order-1,p1-1) == 0))).OR.((p2 /= 1).AND.((MOD(order-1,p2-1) == 0)))) THEN
                                                                               1
Error: Argument ‘P’ of MOD at (1) shall not be zero
../source/main/dealiasing.f90:72:18:

Usage of MODULO (fortran 95 standard) instead of MOD makes it go away.

Hello,
I recently started a project at the Maritime Research Institute in the Netherlands (MARIN). In this project, I will take a look at a number of NWTs and possibly recommend one for further use at MARIN.
One of the codes I am considering is HOS-NWT. Thank you for making it available to everyone.

I compiled both HOS-NWT1.0 and HOS-ocean1.1 with ifort. The latter is working fine but when I start a simulation with the former, I get the following error message:

Wave first order
At first order
required x-pic mode 19 enabled 512
required y-pic mode 0 enabled 1
At second order
required x-pic mode 79 enabled 512
required y-pic mode 0 enabled 1
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
HOS-NWT 0000000000752CDD Unknown Unknown Unknown
HOS-NWT 0000000000632BAA fourier_fftw_mp_e 485 fourier_FFTW.f90
HOS-NWT 00000000006312F9 fourier_fftw_mp_d 442 fourier_FFTW.f90
HOS-NWT 0000000000566F56 initial_condition 396 initial_condition.f90
HOS-NWT 00000000004168AB MAIN__ 408 HOS-NWT.f90
HOS-NWT 0000000000404FAC Unknown Unknown Unknown
libc.so.6 0000003FAF41ED5D Unknown Unknown Unknown

HOS-NWT 0000000000404EA9 Unknown Unknown Unknown

Do you have any ideas for me so that I can resolve this issue?

All the Best,
Bulent

Hello,
I have some confusions of the wavemaker in HOS-NWT, I got the motion of wavemaker from experiment, but after I multiple the TF function and got the wave component, the calculation broke. Is that matter that the wave is a shallow water wave (with wave length 6m and wave depth 2.8m)? And should I take care about the interval of frequencies in the input of wavemaker?
Thanks a lot.
huaqian.

This is with regards to resol_HOS.f90, subroutines solveHOS and solveHOS2, lines 178-191 and lines 534-547, respectively. They read as
! phix_add = space_2_Fourier(phix_add,'cos','cos') phiy_add = space_2_Fourier(phiy_add,'cos','cos') phiz_add = space_2_Fourier(phiz_add,'cos','cos') ! phix_add_ext = extend(phix_add) phiy_add_ext = extend(phiy_add) phiz_add_ext = extend(phiz_add) ! phix_add = Fourier_2_space(phix_add,'cos','cos') phiy_add = Fourier_2_space(phiy_add,'cos','cos') phiz_add = Fourier_2_space(phiz_add,'cos','cos') phix_add_ext = Fourier_2_space_big(phix_add_ext,'cos','cos') phiy_add_ext = Fourier_2_space_big(phiy_add_ext,'cos','cos') phiz_add_ext = Fourier_2_space_big(phiz_add_ext,'cos','cos') !
Why are all FFT and their inverses cos-cos type? Is it inconsequential since phix_add & others are reproduced back by same technique? Is it related to the extend subroutine or will this change if instead of cos-cos type, sin-sin type or sin-cos type is implemented?

Moreover when reducing back from extended space to physical space, why is reduce not used, instead filter_ext is used?

Dear Author,

I am writing to seek your advice on an issue I have encountered while using HOS-NWT to generate irregular waves. Specifically, when setting the calculation duration to 500 seconds, the FFT results of the wave height time series exhibit "amplitude aliasing" (one frequency corresponds to multiple amplitudes, as shown in Figure 1). Could you please help me understand the cause of this issue? The detailed input file is shown in Figure 2.

Thank you very much for your assistance.

Sincerely,
Ban Zhiqing.

Hi, @gducrozet
I have encountered some problems when using HOS-NWT. Could you please help me to solve them? The issues are lists below.

i) "n1" and "n2" are interpreted as "number of modes/points". If we run a 3D wave simulation, should "n2" (the number of discrete points in the y-direction) be set to greater than 1? However, the web page still allows n2=1 in 3D conditions. So I'm really confused about "n2". Also, could you tell me what the setting criteria for "n2" is? How does it relate to wave length and height?

ii) In 2D simulation, p1=M, p2=1; In 3D p1=p2=M. In general, can I keep these settings unchanged and get a relatively normal result?

iii) I set up the wave height probes in the 3D domain, but the "prob_ele" file is empty, and in the "Results/probes.dat" the elevation data is missing, and only the time series exists. How to get the probe monitoring data? (see Figure 1 & 2, please.)

Have a nice day!
Best regards,
Jiaming Liu

I am trying to generate JONSWAP spectrum irregular wave with Hs=17.3 and Tp.16.65 and gamma=3.15 for. model scale is 50 and Hs and Tp are scaled.
I guess this spectrum should be generated because the waves can be easily generated in normal wave basin test. However, simulation generate NaN after some running.

input is as follows:
1 ----------------------- Tank dimensions -----------------------
1 Length(m) of the wavetank :: xlen :: 120.0d0
2 Beam(m) of the wavetank :: ylen :: 100.d0
3 Depth(m) of the wavetank :: h :: 5.d0
4 ------------------------ Select case --------------------------
5 Choice of computed case :: icase :: 4
6 ------------------ Sloshing case (icase = 1) ------------------
7 Number of the natural mode :: islosh :: 3
8 Amplitude (m) of the mode :: aslosh :: 0.2d0
9 --------------- Monochromatic case (icase = 2) ----------------
10 Amplitude (m) of wave train :: amp_mono :: 0.1d0
11 Frequency (Hz) of wave train :: nu_mono :: 0.5588d0
12 Angle (deg) from x-axis :: theta_mono :: 0.d0
13 Phasis (rad) of wave train :: ph_mono :: 0.d0
14 Directional wmk type :: ibat :: 2
15 Wave target distance (ibat=3):: xd_mono :: 18.d0
16 -------------------- File case (icase = 3) --------------------
17 Name of frequency file :: file_name :: Hs4_Tp10
18 Frequency cut-off :: i_cut :: 0
19 Low cut_off frequency (Hz) :: nuc_low :: 0.3d0
20 High cut_off frequency (Hz) :: nuc_high :: 3.d0
21 ------------------ Irregular wave (icase=4) -------------------
22 Significant wave height (m) :: Hs :: 0.346d0
23 Peak period (s) :: Tp :: 2.3547d0
24 Shape factor (Jonswap) :: gamma :: 3.15d0
25 Seed number for random numb. :: iseed :: 513
26 -------------------- Wavemaker definition ---------------------
27 Nonlinear order of wavemaker :: i_wmk :: 2
28 Type (1: piston, 2: hinged) :: igeom :: 2
29 Rotation axis distance :: d_hinge :: 1.d0
30 Time ramp :: iramp :: 1
31 Time ramp duration (s) :: Tramp :: 5.0d0
32 ----------------------- Numerical beach -----------------------
33 Absorption numerical beach :: iabsnb :: 1
34 Beginning front num. beach :: xabsf :: 0.8d0
35 Absorption strength front :: coeffabsf :: 1.d0
36 ------------- Elevation/Velocity-Pressure probes --------------
37 Use of probes :: iprobes :: 1
38 Filename of probe positions :: pro_file :: prob.inp
39 ---------------------- Time-integration -----------------------
40 Duration of the simulation :: T_stop :: 1528.0d0
41 Time tolerance: RK 4(5) :: toler :: 1.e-4
42 Output frequency :: f_out :: 30.d0
43 --------------------------- Output ----------------------------
44 Output: 1-dim. ; 0-nondim. :: idim :: 1
45 free surface plot :: i3d :: 0
46 modes plot :: imodes :: 0
47 wavemaker motion plot :: iwmk :: 0
48 Swense output 1='yes',0='no' :: i_sw :: 0

Hi, when I compile 2D HOS-NWT (n2 = 1 p2 = 1), it works well, but if I try to compile 3D HOS-NWT (n2 > 1 and p2 > 1) the error shown as:
/usr/bin/ld: failed to convert GOTPCREL relocation; relink with --no-relax
collect2: error: ld returned 1 exit status
make: *** [makefile:105: bin/HOS-NWT] Error 1
How do I fix this?

Hi,

Do you have an update of the code, so it can be installed with modern compilers on modern hardware? I tried to compile the HOS-NWT in the HPC, but it failed because the compiler of the code was old. Many thanks.

Best regards
Xiwu

Hello Guillaume,
Recently I read the codes of wave spectrum in https://github.com/LHEEA/HOS-NWT/blob/master/sources/main/initial_condition.f90#L261，I have some confusions.

1. For JONSWAP spectrum, compared with JONSWAP equation, the source code missed a parameter alpha
2. For Bretchneider spectrum, the source code missed Hs.

Why does there exist difference?

Reference:
https://www.orcina.com/webhelp/OrcaFlex/Content/html/Waves,Wavespectra.htm
Sincerely,
huaqian

Hi everyone

I'm trying to use HOS-NWT for a basin with a wavemaker which has a limited length. I read in the PhD thesis of Félicien Bonnefoy that he adapted the code for the specific case of the B600 towing tank of DWA. His adaptations should be described in the study report 'Etude numérique de la faisabilité de la génération de houles obliques au B600 avec un batteur latéral'. However, I can't find that report anywhere on the web. By going through the code, I saw that the implementation is still present in the different subroutines, but that the necessary input parameters need to be added to the input file and read by the algorithm. If I'm not mistaken, the situation of a wavemaker on only a part of the side can be activated by setting 'iwidth' to two. Doing so, eight more parameters need to be defined: ywmk, yl, yr, y_rmp, yabsl, yabsr, coeffabsl and coeffabsr.

I'm not sure what the variable y_rmp stands for. I assume that it has something to do with the shape of the absorbing beaches next to the wavemaker in the y-direction, but I haven't a clue on its exact definition. Anyone who can help me with this? Any other tips on how to use HOS-NWT for a wavemaker with limited length are welcome as well.

Thanks in advance!
Tim

Attached input_HOS-NWT.txt. If nu_mono>1.5 Hz no paddle motions are registered. Happens even if icase is not equal to 2 but nu_mono is set to be greater than 1.5 Hz.

Actually, it has been observed that even this value of 1.5 Hz is not consistent as it seems to depend upon tank dimensions. This is one such example. If such a situation is encountered, no matter what value of icase is, the paddle motions won't register any movement.

The dependence on tank dimensions is making debugging difficult.
input_HOS-NWT.txt

Hello! I have been following the Grid2Grid Manual for the installation of both HOS-ocean and HOS-NWT. I managed to install HOS-ocean without any issues but am currently experiencing difficulties in installing HOS-NWT. The error I am experiencing occurs when I initiate the make command into the terminal window. I have attached a screenshot showing the error that is stopping the installation which is occurring in the dealiasing.f90 script. I am attempting to install HOS-NWT onto a Mac High Sierra Operating System. Any help would be greatly appreciated.

Thank You

Hello,Thank you for sharing this source to the public .I'm using this model and I'm having some problems: when I simulated long time evolution from the case3-file , the wave elevations has been a marked decline and oblique rise, when I increase the n1 the phenomenon improved obviously but there are still a slight decline (Fig. 1 -2), I guess as the effect of n1 or n3 led to a decline or rise, but I did not find in your papers about how to determine the value of n1.
I am looking forward to your reply