Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
Is Mecway friendly for ship structure modelling/analysis?
I am in Ship Building industry and I have been using LISA for a while for simply FEA (for complex one we will engage consulting company for professional FEA work). I am mostly dealing with simple stiffened plate structure (like a small section from ship consists of stiffened plates and T-beam). I noticed that in LISA it is very limited to do such a modelling because 1) you cannot specify L shape as beam element, so I have to model the angle bar as plate also; 2) The T beams intersection requires a lot of efforts to model. I know that Mecway is based on LISA but much stronger. So just wondering if this can be easier done in Mecway? Thank you!
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Comments
I have done some testing and MECWAY & CCX seems perfectly capable.
I have obtained this result and its being tricky, at least for me, but result looks pretty good and coherent.
I suppose it will need some more work in case your shape is curved.
I posted a beam sections library some time ago in this forum which could be useful. It has all common European sections including T's.
There are contacts involved and T sections are solid elements, but overall computational time has been 26 seconds on a laptop for you to have an idea.
Not sure what kind of accuracy and complexity do you need but for simple geometries I think it can be perfectly suitable.
I haven't used Lisa so I can't compare.
This is the tricky part.
I have start from the section profiles. You can download them from the forum in the post “Library of Steel Beam Sections according to EN”.
Import the ones you are going to use and delete all the rest.
Position all the beam sections in the 3D space before meshing by means of basic operations. Rotate, move, copy,… Everything will move faster and selections will be easier.
Mesh the beam sections with the Mecway auto-mesh tool and then extrude them.
This may seem rudimentary but I ‘m getting used to build many parts of my meshes manually and they look good.
If the stiffeners cross one another, I don’t waste my time cleaning the intersection. It represents some extra material in the common volume but… in fact there is some extra welding material in the real world you are not representing. (don’t merge nodes)
Plate has been the last thing to do as the beams are aligned in one face.
Take four nodes on the beams, move/copy up and build a quad element. Finally refine the shell plate as desired. You could also build a solid if desired.
Select all the contact faces on the beams (S1) and apply contact to the plate (S2). Plate (S2) must be the Master.
Hope it works for you and you join the community.
The tools for manually building plate meshes are much the same too.
Thanks Victor for the clarification. Will try a demo version first.
It is a small problem with typical ship structure. Normally, due a ship specify, a model is builded as shell elements for a shell (side, decks , bulkheads etc) with profiles (L, T, HP etc) modelled as beam (not solid). An - in addition, beam is located not as it is shown in scereenshot above, but in overside direction - a flange is not lying on plate, web is welded to plate an flange is at oposite site (for better section modulus). There are a 3 problem with modelling in these way in Mecway (and in other software also
In my experience in Mecway a stiffener modeling is a biggest problem. Normally I decide to use not a beam, but modelling stiffeners also as shell elements. (Using a solid elements is not acceptable (due a Ship classification requirements about a mesh size etc)). But it is a many problem with consistent mesh when I try to build it in these way. Normally, after import shell model from CAD software mesh generated at edges are not consistent and it is many manual works to update it. This problem (how to add beam elements to shell structures) is a main problem for me. I know - i can use a Salome or other software - but it is not solution in real work (very time consuming).
Your issues are probably due to the standard procedure in the industry. Meshes are built from CAD models. Have you tried to build meshes from Scratch? I mean , manually. I’m talking seriously.
Extrusion is an underrated operation. This is a fully conformal mesh build in 20 min using your picture as reference. I'm, using T's and L sections as reference which are typically more difficult to manage.
Fully done in MECWAY but the strategy could be applied to any software.
We are talking about simple plates with maybe some detail. If there is a client in the middle with some additional requirements I understand there are other professional software to deal with them but for simple checks , internal use or pre-dimensioning operations I think it is perfectly suitable.
Just one other suggestion - there is shell offset which you might be able to use instead of all the stiffeners have the same offset from the plate. But I imagine it would get tricky to be sure the meaning is correct for more complex structures than just a flat deck.
I agree with disla that you can do a surprising amount manually. Here's a solid part I built entirely with the manual meshing tools. Mostly extrude, hole, fit to curved surface, and refine custom.
By the way - it is a most weak point of Mecway - how to build a consistent mesh based on imported shell model. If tis problem will be updated/solved - mecway will be very strong competitor for other very expensive software.
I hope msy288 you are not planning to build a ship.
I agree, simple problems, simple solutions, big problems, prepare your wallet.
And again.. Mecway is so fantastic software. It is a only a small step to be a perfect software for engineer ...
btw. For more information about use FE in ship structure analisis pleas see e.g. :
https://rules.dnv.com/docs/pdf/DNV/RU-SHIP/2021-07/DNV-RU-SHIP-Pt3Ch7.pdf
https://rules.dnv.com/docs/pdf/DNV/CG/2021-08/DNV-CG-0127.pdf