Mesh creation in ANSYS AQWA for Sea Keeping Analysis

[neodesigner]

Hi all !

This is my first post on this forum. On the first instance, let me begin by thanking the person who runs this website, this has helped me many number of times.Thanks again..Now to the problem -

I am trying to model a 60m long OSV into ANSYS AQWAfor Sea Keeping analysis. This is the first time I am doing it, and find this a bit tricky. I load the LIN file in AQWA GS, managed to get the profile scan completed, and generated the mesh. I give relevant info like CG, drafts etc and then save this as a DAT file .When I try to run AQWA on the DAT file , it pops up with numerous errors like aspect ratio incorrect, very colse to another element etc etc.

I am confused because, the mesh created by AQWA fails to run with AQWA itself ? Or is it that I am doing something really stupid?

If there is someone experienced with AQWA, please advise.

Thanks in advance..

[Ad Hoc]

If this is your first time..then you should start with a simple shape and compare the results against known tested values. Then slowly build up to bigger and more complex models, to verify/validate the results.

Just putting in the largest model you have and expecting good results and accepting them as absolutes, without knowing their limitations to begin with, is asking for trouble.

[neodesigner]

Hi Ad Hoc,

Thanks for the quick reply. Infact, I have the computed RAOs from a consulting firm for this specific model. I am trying to learn by generating the RAOs myself and validating against these values.

But I do agree that I have to start with a simple model first. I tried modelling a barge with 4 coordinates inside aqwa using the LIN file, but it just fails to complete the profile scan. Have been trying all different tricks since last week.
I will keep trying, but any advise will be of use.

Thanks..

[Ad Hoc]

well, there you are...you know this and the results suggest as such. If you are unable to get a simple 4 co-ord model to work...how else do you expect to import a complex model and have this working?

baby steps....

Have you thought of generating the ROAs by hand, from the data you already have?

[neodesigner]

Hi Ad Hoc,

Thanks again..

I have to clarify here - The problems I face are different in both cases.
in the box shape, it doesn't complete a profile scan, probably because the stations are far apart & there is no way to manuallly guide a profile scan.
Hence I am not able to generate a mesh.
But in the OSV's case, it completes a profile scan and I am able to generate the mesh.But this mesh is of no use because AQWA finds the mesh not proper.

Yep.. I am a baby right now & look forward to seniors like yourself to guide me in the proper direction.

Regarding hand calculation, I referred the PNA several times before setting out to try the stuff & I have to admit the Mathematics is complicated. Even though I am confident of solving equations, there is very little guidance available on how to proceed.

Will appreciate any guidance in this direction.

Many thanks..

[Ad Hoc]

RAOs are not that tricky, especially if you use standard ITTC formulae for a developed sea. But again, better to obtain a spectrum of the intended sea conditions.

But you need to know the unit response of "bodies" at certain frequencies and amplitudes. That is the tricky part. Model tests are using the best.

[wgingenieria]

Hi man, I just read your post and perhaps it could be useful. When creating the mesh from the line file, you must comply with certain geometric requirements such as ratio between each side of the elements of the mesh, avoid having nondiffracting elments above the still water line, etc. Most of this information is given in the AQWA reference manual or the user's manual. If you guarantee that your model does not have diffracting points or plates above still water line, the geometric errors could be ignored by using the option GNON (in the options card) moreover. That's all then you can run aqwa line and plot your RAOS. After that you have to correct the raos of the vessel for roll (because of the viscous roll damping) you can do it adding a 5% approx of the critical damping!!. Then you RUN again the model by increasing this critiacl damping in the DECK nNo.7. That is the end of the story (I strongly recommend you to read the aqwa referenc manual). Good luck

[HReeve]

Hi,

The simple answer is, yes, a mesh developed by AQWA for a .lin file can have errors.

The errors you are seeing are due to adjacent elements having too large a size difference. You may also have aspect ratio errors, or elements too large at a hard corner (turn of the bilge or at the stern).

You need to look at the mesh in the AGS, as well as the text output file, to see which elements are causing the problems, and adjust the node locations accordingly. It is a slow and tedious process. Most modeling is.

You can turn the diffracting elements a different color in the AGS to identify if they are above the waterline.

I do not allow my engineers to use the Go On feature, as too liberal use of it results in an invalid result.

[neodesigner]

Thanks for all your words of wisdom..Especially to wgingenieria & HReeve. It is definitely good to know what can be done as well as what can go wrong. I finally managed to get Aqwa running. Thanks again..

[frank2526244]

It always no problem if just a few elements cause this problem, just add GOON
at the option card. By the way , the mesh generation in AGS is designed only for ships.If you come with a model with complex geometry and import the mesh by yourself, this option is very useful. Some error elements may cause prssure error, but it's just has local effect and the RAOs turn out ok. You can find modeling limits at LINE manual. And I recommed that you can use some FEA software (such as Patran). Best wishes!!

[ssaw005]

I am using AQWA with workbench. I have exported model from Rhino to designmodelar, and export was sucessfull. When I try to mesh in AQWA the mesh seems to have error. I had a close look to mesh then I shaw very fine mesh at the edge of bigger mesh. Can any tell me what could be problem with model.

[ssaw005]

I am carrying out a hydrodynamic analysis of a offshore supply vessel using ANSYS-AQWA.

Is it necssary to close the deck if the deck is above WL?.. can any body help me on this?

[HReeve]

Not to be mean, but if you need to ask that question, then you probably should go back and do a few AQUA tutorials, or find someone to closely guide your work for your analysis.

You can't just blindly jump into using a programme like AQWA and expect to get any kind of reasonable results. There are way to many pitfalls and tricks that if you get wrong will completely invalidate your results.

[quartet]

Hi,
I don't have the tutorial for ANSYS AQWA.
How can i reach them?
I am using ANSYS 12.1 version.
Regards

[CWTeebs]

I realize this thread is old now, but, I've been using AQWA for the past few months, and, now that I'm actually GOOD at it I'm interested in helping others get started, as a hobby. It's a sophisticated program but extremely difficult to use.

Anyway, if anybody wants help, email cthibault@mainemarinecomposites.com

I do not allow my engineers to use the Go On feature, as too liberal use of it results in an invalid result.

Again, I realize that this is all very old topics, but there's also not much of an AQWA user community out there that I've found, but I've found that for all but the simplest shapes (cylinders, box barges), it's impossible to satisfy all of the meshing criteria. This is okay in most cases as the simple shapes are pretty good indicators of wave loadings. The 'GOON' option also doesn't really constitute true 'errors', they are really warnings about the geometry as the relative size of the panels can cause the calculated element strengths to be erroneous.

I'm messing around with Excel's optimizer in VBA to jiggle the points to satisfy the mesh on a per-body basis.