Obtaining RAO from Power spectral density

Discussion in 'Software' started by tea_floss, Jul 6, 2016.

  1. tea_floss
    Joined: Oct 2013
    Posts: 16
    Likes: 1, Points: 3, Legacy Rep: 34
    Location: USA.

    tea_floss Junior Member

    I am working on motions of Asymmetrical catamarans using CFD; I am particularly investigating heave and pitch motions using Star CCM+.

    The issue is, the output is in the time domain, and when I perform Fast Fourier transform to convert it into the frequency domain, the output I get is frequency v/s Power spectral density. My question is how to turn my" frequency v/s Power spectral density" data into "frequency v/s ROA of heave and pitch."

    can anyone point me in the right direction?

    [​IMG]

    Heave in time domain

    [​IMG]

    Power spectral density v/s frequency
     
  2. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    If you are producing a PSD you are going too far. It's a step past the FFT. You need to compute the FFT to produce the PSD.

    It is tediously reversible but rather than computing the amplitude from the PSD you should be able to get whatever software you are using to just produce the FFT. I'd ask the software producers for some help.
     
  3. b1ck0
    Joined: Mar 2010
    Posts: 128
    Likes: 3, Points: 18, Legacy Rep: 40
    Location: Varna/Hamburg

    b1ck0 Senior Member

    When you are computing PSD via FFT things are not 100% accurate due to FFT itself. I normally use the Welch's method and applying a Hann window to it which will significantly improve your PSD in terms of noise, note that ANSYS AQWA and DNV SESAM - PostResp Time are doing the same.

    Now back to your question. What you are doing is computing the PSD of the response of the vessel to a given environment. This is really cool if you want to do some statistical post-processing and get maximum responses but if you want to find the RAO itself you need to remember how the RAO is related to the response PSD. More specific PSD = |RAO|^2 x Wave Spectrum ...
     
  4. tea_floss
    Joined: Oct 2013
    Posts: 16
    Likes: 1, Points: 3, Legacy Rep: 34
    Location: USA.

    tea_floss Junior Member

    Also, My heave and pitch responses are in negative, What does that say about my hulls seakeeping characteristic ?
     
  5. b1ck0
    Joined: Mar 2010
    Posts: 128
    Likes: 3, Points: 18, Legacy Rep: 40
    Location: Varna/Hamburg

    b1ck0 Senior Member

    It says nothing about them ... I think the volume which is calculated based on the mesh is less than the mass defined for the vessel ... that's why it sink a little ...
     
  6. jehardiman
    Joined: Aug 2004
    Posts: 3,762
    Likes: 1,152, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Trying to run a time domain solution to get a frequency domain answer is a very poor method. You will need to run thousands of different runs and then compile them for any useful irregular seaway predictions. While Star CCM+ is a fair flow solver, WAMIT is the program you need for this.

    If you want to go ahead and do this, get a copy of Dynamics of Marine Vehicles by Bhattacharyya and read the first 9 chapters. It lays out all the issues with this method.

    And FWIW, if you really want to do this you will need sample frequencies twice the shortest spectral bin and at least an n^2 sample where n is the total number of spectral bins of interest. So for a normal irregular seaway with a spectral bin size of 0.5 seconds and wave periods between 0.5 seconds and 30 seconds you will need a minimum run length of 225 seconds (30^2*0.25). However, because most FFT routines requires a sample size of 2^m you will more likely have a minimum run length of 256 seconds. Note these run lengths are after all initialization transients have died out (which is why your data may be negative; your initialization is not in equilibrium...so start the model with no flow or waves and see what happens).
     
  7. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member


    tea
    Are you doing this to work out added resistance or just an MSI study ?
     
  8. tea_floss
    Joined: Oct 2013
    Posts: 16
    Likes: 1, Points: 3, Legacy Rep: 34
    Location: USA.

    tea_floss Junior Member

    Just MSI study
     

  9. tea_floss
    Joined: Oct 2013
    Posts: 16
    Likes: 1, Points: 3, Legacy Rep: 34
    Location: USA.

    tea_floss Junior Member

    Thanks a lot people, you have been helpful.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.