[colinB]

Hi all,
I’m currently doing an internship in a ship design office. My task is it to evaluate different cfd programs in order to perform initial hull optimization inhouse.
We want to use potential theory algorithms for wave resistance analyzes in particular wave pattern, pressure/velocity distribution over the hull and wave cuts.
Our first choice is OpenFOAM since financial aspects play a major role.
Also the fact that you can customize the code for your needs looks attracting to us.

So last week I made a big research upon this topic and I figured out that there are many things to observe. Though I have some basic experience with OpenFOAM (I once was allowed to play with calculating parameters and see their impact on the final result, with icofoam for a term paper) some questions come up to me:

Meshing is one of the crucial points and I wonder if there is a proper mesh generator/converter/manipulator included in OpenFOAM. I know that there are a lot of codes included for these tasks, but are they also capeable of modelling more complex meshes including ships appendices such as bilgekeels, nozzles, rudders, bow thrusters and so on. Further I read that SALOME has been used for mesh generation also for ships, but how reliable is that and can it meet the requirements?

As mentioned above we want as a start just use the potential theory. Is it therefore convenient to use potentialFOAM or are there other solvers that are more appropriate especialy when it comes to hull optimization. Or is it even better to create an own solver that fullfills the special needs we demand? I saw one test case of Kevin Maki and Bill Rosemurgy from the university of Michigan published on the 4th OpenFOAM Workshop in Montreal, Quebec in 2009 what looked quite interesting to me. But this "experiment" is just validated for slender hull forms with the thin ship theory and we would like to extend this for general hull forms using the more simple potential theory. So my question is if you have some more codes validated with test cases, such as the Kriso VLCC or Containervessel KCS. Another problem that occures, how does OpenFOAM deal with free surfaces?

I hope you can help me out and maybe give me some further hints.
If you have some further questions don’t hesitate to ask me!
Thanks for your trouble!


Colin

[daveatstyacht]

Colin,
In answer to your question about meshing, OF already contains a mesher called SnappyHexMesh which is quite capable of producing complex (ie ships and associated forms) shapes. SALMONE has to my knowledge not continued to be developed for OF (last version I know of it supported was 1.4).
I have also looked into using potentialFoam to do initial runs of ship hulls, unfortunately as far as I can tell it is unable to handle free surfaces (please someone correct me if I am wrong since that would be wonderful). There is a quite capable VOF solver called interDyMFoam (dynamic version of interFoam capable of dealing with the free-surface) that is very suitable as a RANS solver with 6 degree of freedom capability (which you can constrain to do only trim and heave and has a built in dynamic mesh and remeshing capability). OF is definitely capable of being used as a design tool for ship design, but you may have to do verifications based on your own available data (in our case tank and starCCM data), additionally it has a very steep learning curve in some aspects (I have been learning how to use it for 2 months and I am just starting to feel comfortable with it)

Good luck in your studies of OF ,
Dave

[elvis]

Originally Posted by colinB

Meshing is one of the crucial points and I wonder if there is a proper mesh generator/converter/manipulator included in OpenFOAM. I know that there are a lot of codes included for these tasks, but are they also capeable of modelling more complex meshes including ships appendices such as bilgekeels, nozzles, rudders, bow thrusters and so on. Further I read that SALOME has been used for mesh generation also for ships, but how reliable is that and can it meet the requirements?

There are several high quality "commercial" meshers that support Openfoam directly!
ANSA http://www.beta-cae.gr/ansa.htm.
One of the best commercial mesher comes from pointwise
http://www.pointwise.com/pr/pr-08g-jasak.shtml
hypermesh could also be a commercial candiate http://www.cfd-online.com/Forums/ope...hypermesh.html as well. You will find more mesher like harpoon etc. http://www.cfd-online.com/Forums/openfoam-meshing/

opensource take a look at engrid http://engits.eu/cms/
salome (uses via plugin "netgen" and "gmsh", if no additional commercial mesher is bought)
calculix http://www.calculix.de/ supports exporting to Openfoam.

study http://www.openfoamworkshop.org/08/i...and_generation and the years before and later on meshing ;-)

You should take a look at http://openfoamwiki.net/index.php/Si..._Hydrodynamics

BMW-Oracle contacted Wikki for some validation calculations.(slide 12)
http://www.ihs.uni-stuttgart.de/file...t_5Nov2009.pdf

as a commercial company you might want to invest some money and get a solver with capabilaties that are crucial for you.
A big advantage is, you have the choice of several "consulting" companies (icon, wikki, opencfd, engys) to name a few with a very good reputation, specialized in the field of cfd/openfoam, that could do this job for you.
=> I am not affiliated with any name mentioned here.

elvis

[colinB]

Thank you guys for your reply some other thoughts came up to me when I read your posts.

@elvis: of course I contacted this SIG already especially since two of the members of this group are highly involved in SH and OF.
Thank you also for the meshing instructions which I surely will consider.
When it comes to the right meshing software we also want to rely on OpenSource software since it seems interesting to get everything for free.
Of course we are expecting that this will take some longer time and will be a little more work than buying a commercial software and support, but since cooperations with academic institutions are a good resource this won't rise the costs either.
Validation cases are not the problem for us since there are also some around in the Internet.

@dave: the features such as RANS solvers and so on are actually not what we want. We are looking for a small handy tool to get a first estimation of the wave resistance. Therefore we will only focus on the foreship, since the accuracy of the results will decrease when there is a point of seperation such as the aft shoulders.

So finally I installed now OF I try to get more acquainted with it and try to optimize it for my purposes. If you have some further hints feel free to contact me.

regards
Colin

[kolloff]

I have read your posts regarding OpenFoam and ship towing resistance, which were posted half a year ago. I work in a company where we start to discuss the same thing and I am curious to know what you found out and what is the situation regarding OF for ship design.

Looking forward for your respond.