Re: [swanmodel-users] Frequency spectrum unstable?

[Mathew T. <Mat...@jb...>]

Taylor, it's just that, thanks. I ran the model with OFF REFrac and now the spectrums look fine. The challenge is finding the problem cells now, I guess, so I can turn it back on again.

Thanks a million for your help,

Mat

Dr Mathew Topper
Marine Renewables Technical Specialist | Marine and Coastal Risk Management


JBA are WEM Framework Suppliers 2013 - 2017<http://www.jbaconsulting.com/news/wemframework>


JBA Consulting, Port Neuk, 1 Longcraig Road, South Queensferry, Edinburgh, EH30 9TD
Telephone:
JBA Group<http://www.jbagroup.co.uk/> supports the JBA Trust<http://www.jbatrust.org/>, Follow us on Twitter @JBAConsulting<http://twitter.com/JBAConsulting>

This email is covered by JBA Consulting's email disclaimer<http://www.jbaconsulting.com/disclaimer>

JBA Consulting<http://www.jbaconsulting.com/> is the trading name of Jeremy Benn Associates Limited, registered in England, company number 03246693


From: Taylor Asher [mailto:tay...@gm...]
Sent: 28 March 2014 18:48
To: Mathew Topper
Subject: Re: [swanmodel-users] Frequency spectrum unstable?


Mat,

In my experience in unstructured meshes, this arises from non physical spectral propagation (usually from refraction in shallow water) arising from an under resolved mesh.  But it's hard to say.

-Taylor
On Mar 28, 2014 11:37 AM, "Mathew Topper" <Mat...@jb...<mailto:Mat...@jb...>> wrote:
Hi,

I'm having some trouble running an unsteady unstructured swan simulation with OpenMP and I wonder if anyone might be able to help?

It appears as if the frequency spectrum is unstable, with large energies appearing in the low frequencies very early on in the simulation. For instance, this is a 1D spectrum from a time step:

Freqs:

    0.0400
    0.0440
    0.0484
    0.0532
    0.0586
    0.0644
    0.0709
    0.0779
    0.0857
    0.0943
    0.1037
    0.1141
    0.1255
    0.1381
    0.1519
    0.1671
    0.1838
    0.2022
    0.2224
    0.2446
    0.2691
    0.2960
    0.3256
    0.3582
    0.3940
    0.4334
    0.4767
    0.5244
    0.5768
    0.6345
    0.6980
    0.7678
    0.8446

Values:

  0.1196E+02  151.4   25.8
  0.1142E+01  148.2   23.2
  0.5511E+01  154.6   17.8
  0.5508E+00  146.6   23.2
  0.5600E+00  153.1   15.2
  0.5211E+00  152.3   13.8
  0.1793E+00  151.6   12.6
  0.4544E-01  150.9   12.3
  0.2899E+01  146.4   21.9
  0.2118E+01  151.2   13.4
  0.8698E+00  149.8   25.5
  0.3000E+00  168.2   49.1
  0.1921E+00  225.5   51.6
  0.1750E+00  247.5   43.6
  0.1529E+00  255.9   37.4
  0.1293E+00  260.5   43.3
  0.8851E-01  263.9   28.7
  0.6261E-01  264.5   29.9
  0.4405E-01  263.6   36.9
  0.4272E-01  238.6   66.7
  0.5973E-01  139.0   62.3
  0.8355E-01  131.9   50.2
  0.1023E+00  138.3   44.5
  0.9005E-01  148.2   44.5
  0.6431E-01  157.2   48.2
  0.4386E-01  163.6   52.0
  0.3197E-01  164.7   52.6
  0.2448E-01  160.0   54.9
  0.1876E-01  159.7   57.9
  0.1368E-01  164.0   59.0
  0.9751E-02  169.5   59.7
  0.6773E-02  166.0   61.6
  0.4624E-02  167.1   62.4

As you can see the spectrum is dominated by the lowest frequency value. Has anyone come across this before?

Many Thanks,

Mat

Dr Mathew Topper
Marine Renewables Technical Specialist | Marine and Coastal Risk Management


JBA are WEM Framework Suppliers 2013 - 2017<http://www.jbaconsulting.com/news/wemframework>


JBA Consulting, Port Neuk, 1 Longcraig Road, South Queensferry, Edinburgh, EH30 9TD
Telephone:
JBA Group<http://www.jbagroup.co.uk/> supports the JBA Trust<http://www.jbatrust.org/>, Follow us on Twitter @JBAConsulting<http://twitter.com/JBAConsulting>

This email is covered by JBA Consulting's email disclaimer<http://www.jbaconsulting.com/disclaimer>

JBA Consulting<http://www.jbaconsulting.com/> is the trading name of Jeremy Benn Associates Limited, registered in England, company number 03246693



------------------------------------------------------------------------------

_______________________________________________
swanmodel-users mailing list
swa...@li...<mailto:swa...@li...>
https://lists.sourceforge.net/lists/listinfo/swanmodel-users