Boat Design Forums  |  Boat Design Directory  |  Boat Design Gallery  |  Boat Design Book Store  |  Thanks to Our Site Sponsors

 Boat Design Forums puzzles about AQWA DRIFT

#1
03-11-2012, 12:15 PM
 godfrey_wong Junior Member Join Date: Mar 2012 Rep: 10 Posts: 6 Location: China

Hello Everyone:

Recently I'm studying on the Mannual of AQWA DRIFT. Accually I'm feeling frustrated cause the lack of theory on second-order hydrodynamic as well as the related cases. Here I list some of my questions and hope for your help:

1. We do not take the effects of mooring into consideration in AQWA LINE. But we need to input parameters about mooring in Deck 14. Is there problem if we just use the RAOs of Line in Drift?

2. When it comes to second-order force, we are asked to input slow added mass and slow damping in Deck 9. I'v never seen the two terms before. I regard it as a defenition similar with added-mass and damping in the theory of linearized theory(Just replace the normal frequency with the slow frequency). Is this right?

3. There are so many coefficiences in Deck 9~14. How can we get these parameters(by estimation or sources from other software)?

Thanks a lot for your attention and help.


#2
03-12-2012, 10:19 PM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
I'll need to recheck my notes when I go into work tomorrow, but, no, it's not valid to use the RAOs calculated in LINE without the effects of articulation. There's a number of ways you can solve this, e.g. running AQWA-FER with the articulations in between the AQWA-LINE and AQWA-DRIFT (I was told the exact way to do this, I just can't remember but I know I wrote it down).

For the quadratic forces part of the calculation is resolved by doing an extra set of added mass and damping matrices. These are similar to the traditional added mass and damping matrices but at ultra low frequencies (drift frequencies).
#3
03-13-2012, 02:48 AM
 godfrey_wong Junior Member Join Date: Mar 2012 Rep: 10 Posts: 6 Location: China

Quote:
 Originally Posted by CWTeebs I'll need to recheck my notes when I go into work tomorrow, but, no, it's not valid to use the RAOs calculated in LINE without the effects of articulation. There's a number of ways you can solve this, e.g. running AQWA-FER with the articulations in between the AQWA-LINE and AQWA-DRIFT (I was told the exact way to do this, I just can't remember but I know I wrote it down). For the quadratic forces part of the calculation is resolved by doing an extra set of added mass and damping matrices. These are similar to the traditional added mass and damping matrices but at ultra low frequencies (drift frequencies).

Your suggestion to connect LINE and DRIFT with FER is quite reasonable.

By comparing added mass and damping matrices of quadratic forces to the counterpart of wave frequence forces, we can easily explain the phonomena. But I still fail to kown how to input the exact number of the second force parameters(or we can just use the LINE again with the difference-frequency?).

Anyhow, I really appreciate your attention and help.
#4
03-13-2012, 09:29 AM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
Hello,

I may have slightly misspoke in my earlier reply. What I wrote in my notes is simply:

"If the RAOs are wrong, then the drift coefficients are wrong, among other things."

So, if you've somehow directed AQWA-LINE to calculate incorrect RAOs, e.g. your mesh has too many bad panels, then there's no way the subsequent analyses will be correct. I guess in retrospect this is intuitive.

Running AQWA-FER to recalculate the RAOs with the articulations added in would work but evidently isn't necessary. For example, when you run a drift analysis from Workbench, you can go directly from the LINE analysis, which is run with the calculate quadratic transfer functions, to the drift analysis in a time response, which handles the stiffness from articulations (there may be a LIBRIUM run in between there if you've directed the starting condition be program controlled).

Also from my notes, which may or may not be of use to you:

The LINE analysis can use the near field solution (NQTF) for calculation of the QTFs, which is in general a more rigorous treatment of the drift forces but is also more sensitive to bad panels, and absolutely requires no gaps (if gaps are present, they must be closed with non-diffracting panels).

Difference frequency effects are most prevalent in most cases, except TLPs, and WAMIT may do a better job of resolving sum-frequency effects because it meshes the free surface.

I will be glad to clarify any of the above points, I realize my writing is largely incoherent mumbo jumbo.

- CWTeebs
#5
03-13-2012, 02:10 PM
 oliversnowvon Join Date: Mar 2012 Rep: 10 Posts: 2 Location: Shanghai, China
Quote:
 Originally Posted by CWTeebs Hello, Running AQWA-FER to recalculate the RAOs with the articulations added in would work but evidently isn't necessary. - CWTeebs
You mean it's unnessary to recalculate the RAOs with the articulations added in AQWA-FER? My understanding is that in most cases(except for TLPs), mooring lines can hardly affect the results of RAOs in AQWA-LINE because RAOs stand for wave frequency motions which is 6 degrees of freedom and of small amplitude. So there's no need to add articulations in AQWA-LINE or AQWA-FER.

And I am very grateful if you can provide the concrete values of the drift (low frequency) added mass and damping from your cases. It'll be better if you can also tell us (me and Godfrey) how you got these values.

__________________
love is wise, hatred is foolish---by Bertrand Russell
#6
03-13-2012, 03:41 PM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
You're right, but in some instances the stiffness from articulation is very important, and AQWA-LINE is free-floating only, which is where AQWA-FER comes in. AQWA-FER will recalculate the RAOs taking into account the stiffness from articulations and mooring connections. For example, I used AQWA-FER to do a frequency (RAO) based analysis of a wave energy converter that was hinged at the seabed, the result was a new set of RAOs that reflected the articulation stiffness and damping about the hingepoint. In this manner you are able to get a very fast and reasonably accurate description of the motions, before going into a more computationally expensive time domain analysis.

The information required by AQWA-DRIFT is mostly be calculated in a preceeding AQWA-LINE analysis (AQWA-LINE run with the CQTF option calculates the quadratic transfer functions, for example).

Hope that clarifies some points.

- CWTeebs

EDIT:
A problem I've had with AQWA-FER is that it does not let you control at what frequencies the RAO is calculated, which makes it somewhat difficult if you are trying to find the frequency at which a maximum response occurs. Unfortunately to do this I've had to write my own scripts that recalculate the RAOs, by recalculating the transfer function manually, which was a bit of a pain. The frequencies at which the RAOs are calculated are based on how the spectrum energy is divided up, and does not necessarily contain the maximum RAO in the output.
#7
03-14-2012, 06:20 AM
 godfrey_wong Junior Member Join Date: Mar 2012 Rep: 10 Posts: 6 Location: China
Quote:
 Originally Posted by CWTeebs The information required by AQWA-DRIFT is mostly be calculated in a preceeding AQWA-LINE analysis (AQWA-LINE run with the CQTF option calculates the quadratic transfer functions, for example). - CWTeebs
Yeah, your solution is quite feasible. When utilizing the AQWE, I'm frustrated by both the operation and the theory. Would please give us beginners some suggestions on the software learning as well as working in offshore enginnering?

Regards!
#8
03-14-2012, 07:54 AM
 wgingenieria Naval Architect Join Date: Dec 2009 Rep: 10 Posts: 3 Location: Ecuador
Quote:
 Originally Posted by godfrey_wong Yeah, your solution is quite feasible. When utilizing the AQWE, I'm frustrated by both the operation and the theory. Would please give us beginners some suggestions on the software learning as well as working in offshore enginnering? Regards!
ABOUT THE LOW FREQ DAMPING IN DECK 9
For a typical ship shape vessel (170x36x18m) the low freq damping coeff could be: (note that the fourth value corresponds to Roll and it is much smaller than the viscous damping you add in deck 7 (about 1E8kgm2/s))
09FIDD 6.0000E5 1.1000E6 1.7000E6 1.100E4 8.000E06 3.3300E9
As you see here for the low freq motions the added damping in surge and sway and yaw are very important (unfortunately the are found from decay tests). For the other ones you may just simply take the radiation damping at a low freq (say 0.2rad/s) from aqwa line.
The RAO's in LINE FER and DRIFT should be the same if the system is linear and not coupled with various structures. If the system is coupled then use AQWA FER and if the system is coupled and nonlinear (drag induced damping in a typical subsea installation) then use AQWA DRIFT. The QTF from AQWA LINE from the AGS may be used directly to compute the mean drift forces or AQWA DRIFT to compute the mean and the dynamic forces (as you realize they will depend on the stiffness of the mooring lines and the viscous damping input for surge, sway and yaw in deck 9).
#9
03-14-2012, 08:41 AM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
I agree with wgingenieria's reply. The only point I would make is that FIDD (and FIDA) are frequency independent coefficients, not low frequency coefficients, meaning they are applied at all frequencies regardless.

Sections 3.4 and 3.4.2 of AQWA-LINE.PDF for ANSYS version 14 does a good job of explaining the theory.
#10
03-23-2012, 12:34 AM
 godfrey_wong Junior Member Join Date: Mar 2012 Rep: 10 Posts: 6 Location: China
Quote:
 Originally Posted by CWTeebs Sections 3.4 and 3.4.2 of AQWA-LINE.PDF for ANSYS version 14 does a good job of explaining the theory.
Thanks a lot for your guidance. I have got some ideas about the calulation.
#11
03-23-2012, 12:36 AM
 godfrey_wong Junior Member Join Date: Mar 2012 Rep: 10 Posts: 6 Location: China
Quote:
 Originally Posted by wgingenieria ABOUT THE LOW FREQ DAMPING IN DECK 9 For a typical ship shape vessel (170x36x18m) the low freq damping coeff could be: (note that the fourth value corresponds to Roll and it is much smaller than the viscous damping you add in deck 7 (about 1E8kgm2/s)) 09FIDD 6.0000E5 1.1000E6 1.7000E6 1.100E4 8.000E06 3.3300E9 ......
Your explaination is quite comprehensive and sound. Thanks a lot!!!
#12
01-04-2013, 07:28 AM
 khaled79 Junior Member Join Date: Nov 2012 Rep: 10 Posts: 7 Location: Egypt
RAO's from AQWA DRIFT ?

Hi.. Is there any chance to produce the RAO's using AQWA DRIFT? I need to know the RAO's of a (TLP) or a (semi-submersible where mooring lines stiffness effect the pitch and roll RAO's).. can i do this from the Workbench directly or from AQWA supervisor ? or modifications in the DAT files ?

Thanks

Quote:
 Originally Posted by godfrey_wong Hello Everyone: Recently I'm studying on the Mannual of AQWA DRIFT. Accually I'm feeling frustrated cause the lack of theory on second-order hydrodynamic as well as the related cases. Here I list some of my questions and hope for your help: 1. We do not take the effects of mooring into consideration in AQWA LINE. But we need to input parameters about mooring in Deck 14. Is there problem if we just use the RAOs of Line in Drift? 2. When it comes to second-order force, we are asked to input slow added mass and slow damping in Deck 9. I'v never seen the two terms before. I regard it as a defenition similar with added-mass and damping in the theory of linearized theory(Just replace the normal frequency with the slow frequency). Is this right? 3. There are so many coefficiences in Deck 9~14. How can we get these parameters(by estimation or sources from other software)? Thanks a lot for your attention and help.
#13
01-04-2013, 10:14 AM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
AQWA-DRIFT is a time domain program and doesn't recompute RAOs.

If you want to see how the mooring stiffness affects the RAOs you can use AQWA-FER from AQWA-supervisor using the CRAO option.
#14
01-04-2013, 03:31 PM
 khaled79 Junior Member Join Date: Nov 2012 Rep: 10 Posts: 7 Location: Egypt
Thanks CWteebs. I am new in Aqwa. Can you please explain how I can run the option CRAO from Aqwa supervisor, and how to recompute the raos again.
Thanks
#15
01-04-2013, 05:26 PM
 CWTeebs AnomalyGenerator Join Date: Apr 2011 Rep: 171 Posts: 232 Location: Maine
What I normally do is create a data file from workbench and modify it into an aqwa-fer file then execute it from AQWA supervisor.

There's an example aqwa fer file in the aqwa/training directory called aftakbuy.dat.

.dat files get generated by Workbench. You need to copy the Analysis.dat file created by workbench to a new directory, convert it into an AQWA FER file, add the CRAO option and run from aqwa supervisor by the run->batch file command. The new RAOs will be generated in the output .LIS file. Here is what the AQWA help says about the CRAO command (excerpt from the AQWA Reference.chm 14 file):

Quote:
 Originally Posted by AQWA Reference v14.chm (F) Instructs AQWA-FER to calculate and output the RAOs for each structure, INCLUDING the mooring lines, but assuming each body is independently moving. These may be used to assess the effect of the coupling of the complete system by comparing these RAOs with those for the fully coupled system (see OPTION PRRI). This is done for the first spectrum of each mooring line combination only, unless the PRRP option is used.