Maxsurf to ANSYS

[xtrememachineuk]

Hello,

I was wondering if anybody had any help or advice about exporting a model generated in Maxsurf in order to perform an FEA in ANSYS. A problem I have at the moment is how to model the stiffeners and transverese. Because of the complex geometry it would be difficult to model them from scratch in ANSYS and add them to the model, so I was hoping to model the necessary lines in MaxSurf and then export it across. Maxsurf is not my strong point (and it is my team mate who has developed the hullform) but I'm pretty good with ANSYS.

Any help, hints or tips would be gratefully received.

[MikeJohns]

Is this for a full global analysis?

I dont use that particular combination of software but models that I have been sent for analysis and usually I find that the solids need some further tweaking in a CAD package , for example the framiing and stiffeners may need to be merged with the plating by more than your tolerance is set to to ensure no 'trapped' surfaces that occur with adjacent surfaces. It is also nice to add some extra 'tags' to act as restraints and for easy point loads.

But if you send a PM or email to the member here 'Andrew Mason' he may be able to give you some direction.

[Ad Hoc]

If you have curves which lay on the boundaries or extent of the members, ie frames and stiffener spacings, then you can use a beam element to define the stiffness located on the curve that is define.

I do all my modelling in the FEA i use, not in "another". I also recommend to mesh manually not automatically. May take longer but you are controlling the mesh, its density and the 'pattern' (which can affect the results) and have greater confidence in the results.

[xtrememachineuk]

Thank you very much for your help guys, it's very useful and Mike, I'll make sure I send the email tomorrow.

It is for a global linear analysis of the hull, yes, but unfortunately being at University I'm limited with the software that we have. ANSYS is the University's preferred platform for FEA as I believe the Civils and Mechanicals like it, and being a smaller department we Naval Archs have to make do. I've heard that Maestro is good for ship and boat modelling though.

I'll play around, but at least I know now which angle to approach things from and that certain functions are possible.

[Ad Hoc]

I use Cosmos/m, it is excellent. I've been using it for 15 years, very 'user' freindly and easy to model too.

The global analysis you're doing, is it a monohull or catamaran?...what are you looking to 'get' from the FE...that will dicate what you need to model and what is not necessary.

[alidesigner]

Use offset beam elements for your frames.

As a first pass just ignore your stringers. You will still see problem areas in the boat.

If you later want to see the effects of stringers you can either model every second one and adjust the properties, or even better, dont model them at all and just adjust your plate properties to allow for increased shear area and different stiffness x & y direction.

Owen Hughes book ship structural design explains how to do it. There is an update out now
http://marinelink.com/en-US/News/Art...ed/327141.aspx

But if you really want to model every stringer (and have a really fine mesh as a result) run them in Workshop. Then export the side and bottom views with just the stringer lines shown as dxf. Then import these into the relevant planes in your model. Then use the curves to extrude surfaces that intersect your hull surface. Then split your hull surface at the intersection and you now have edges that match the workshop stringer runs that you can use to model offset beam elements for your stiffeners.

Do not try and model using solids, you need to use a combination of shell and offset beam elements. Also get your surface geometry right before you start meshing.

But remember, even if you are using class loads, they are not exact, so your results can not be exact. This is why I find it better to ignore stringers and just see how the boat is going to behave. Find the problem areas and then do a local model of them. It will be much faster and just as reliable.

And remember to speed things up you only need to model half teh boat and use symetry constraints.

Cosmos was great but you cant buy it anymore in its old form and the new version that comes as part of solidworks doesnt have offset beam elements.
I use Femap/Nastran.

Good luck.

[MikeJohns]

I like Cosmos for detailed local analysis of solids there is never any associated error then of whether beam elements have been correctly applied.

For bigger vessels the dedicated ship analysis tools make life simpler. But node and beam modeling is so simple relative to a solids model.

[Ad Hoc]

I have tried modelling just half the model with symmetry constraints, but often found the constraints effect the results too much. I always model the whole vessel, when doing global analysis. It is easy, well in cosmos anyway, as just model half then copy via symmetry commands.

I didn't know cosmos is no longer available, as a stand alone...pity, it is an excellent FEA.

[alidesigner]

I just noticed you are in Japan Ad Hoc, and a Cosmos user for 15 years - you wouldnt happen to known anything about the crane base on Negeen would you ?

[Ad Hoc]

Negeen...now there is a name i havent heard for years
Bloody excellent boat...still running too

I just might know a thing or 2 about the boat...

[alidesigner]

Ah so it is you then... hello Mr K

Long time no speak

Her twin sisters werent half bad either.

(sorry to bore those that have subscribed with this personal stuff)

SC.

[Ad Hoc]

Hi SC...i thought it was you

Yes been a while!

Negeen had 2 sister vessels??...well, the designer knew what he was doing is all i can say...hahaha

I was in Iran recently, they still use them and love them to bits, they were impressed when i told thme the history of the vessels