Shell-solid interface

[jpblasques]

Hi all

I am modeling a composite marine propeller blade. I am using shell elements for the skins using shell99 and I have included a solid tip using solid 187 (the blade is hollow except for the tip region). The question is that the solid element does not share the rotational degrees of freedom with the solid elements so their connection works like a hinge (only ux, uy and uz are transfered from shell to solid).
I have been told to use constraint equations but this are not so easy to implement as the mesh is not mapped. I read about the MPC184 element but I don't think it is useful for this purpose. Both shell and solid elements share nodes so the idea of using contact elements does not seem to suit also.

Does anyone has any suggestion on how to do this?

Thanks in advance.
Regards.

Zé

 

[Stringmaker]

Ze,
I have two suggestions for you:

1) Why not model the tip of the blade with shells too? I assume that it is still sufficiently thin that shells should be fairly appropriate for that region. Also, I'm guessing it's not one of the critical regions on the blade either. Can you obtain the midsurface geometry for this region as well? Just a thought...use your judgement here.

2) Back to your original question I've used the SHSD command for interfacing shells and solids in the past and it's worked pretty well. See this command in the documentation and have a look at Section 8.2 in the Contact Technology portion of the documentation.

-Brian

 

[cbrn]

Hi,
in addition, you could investigate if the "mixed-formulation" SOLSH190 is appropriate for you.

Regards

 

[Stringmaker]

Just to add on what Cbrn said, if you were to mesh the propeller tip with SOLID186's and then mesh the primarly portion of the blad with 190's that would work beautifully. Nodal compatability does not exist between 187's and 190's so a little change in strategy would be required.

Good luck,
-Brian

 

[jpblasques]

Well, first of all thanks for the help! I am sorry I have not answered before... and sorry for the long post below.

The main reason why I have modeled the blade this way lies in the fact that I do not know any of the thicknesses to start with. This will be the result of an optimization routine which is the next phase of this project. Hence it is easier to use shell elements as this are easier to handle when controlling the thickness.

Since the sections at the tip are very thin I have decided that it would be easier to make it solid to avoid the risk of having the shell thicknesses from pressure and suction side so high that they would overlap. Another problem with the volume at the tip is that it can only be modeled with tetahedra. I have tried layered solid elements and it does not accept a mapped mesh...

From my understanding, you suggest that I model the main body of the blade using solid elements. Well this poses a problem as I would have to create volumes with the right thickness which is something I do not know yet. But... at this point I have a question for you:

Would it be correct to use solid layered elements with a certain thickness, and then make some of the top/bottom layers to have zero thickness/strength/stiffness and in this way "control" the laminate thickness? In this case i would create the volume with "exaggerated" thickness. mesh it and then just "turn off" some of the layers.

In this case the use of SOLSH190 is clearly a tempting option, although I do have a remark. The element has no mid-side nodes and as such I am not sure how suitable is it to model curved surfaces... once again, what do you think?

Nevertheless, and back to the original question. I have been considering all the possibilities to solve the solid-shell interface problem. Honestly the most straight forward of them all seems to be to simply overlap some shell elements on the solid surface. I am not worried about local stresses as my intention is mainly to observe the blade behavior when using unbalanced laminates. What do you think about this solution?

Once again, I am very grateful for all your help.
Regards
Zé