Effect of SHELL elements out of plane

[enriquelvarez]

Hello,

I've been working on simulating two numerical samples in solid elements. Both samples are in different materials that have the same Coefficient of Thermal Expansion (CTE)in three directions.
In addition, a thin skin of shell elements in aluminium (with a CTE different from the one of the solids) is put on the top and on the bottom of the samples. So samples are a sandwich: shell(CTE1) + solid(CTE2) + shell(CTE1)

When heating the samples, expansion behaviour is not the same for the samples: maximal and minimal displacement in out of plane direction (plane of the shells) is not the same, even without considering border effect. I suppose that this effect is because of the difference of CTE of the solid and the CTE of SHELL elements.

What I can't understand is how shell elements which have defined CTE just in plane can have an effect in displacements out of plane...

Can anyone help me?


Thank's very much

Enrique

 

[40818]

Firstly, why would you have different coefficient of thermal expansion for each orthogonal direction?
I have probably had to much wine by now, but are you intimating that your calculating an exoansion normal to the shell??
I dont undertsand that as it is infinitly thin.
If your shells are connected to a central solid (honeycomb example), then your solid will expand in all 3 directions, which will move the shells "outwards".
Maybe i'm just confused....

 

[enriquelvarez]

Thanks for your answer 40818,

But as you said, maybe you had too much wine.

An orthotropic material can have different CTEs in each direction...That's one of the differences with isotropic materials. There are some others that can be found at basic engineering books.

Calculation of expansion normal to the shell is for solid elements and not for shell elements. Of course, shell elements were created in order to do calculations in plane.

This is a forum to share ideas and not to demonstrate how wise you're.

But maybe I'm just confused...

Anyway, thanks for updating my knowledges

Enrique

 

[GBor]

Enrique,

I will take a shot, but it is purely a guess without knowing what specific software you are using and how it calculates thermal expansion:

Remember that poisson's ratio impacts the expansion along a particular direction based on the orthogonal direction. For instance, you have v12, v13, and v23. These ratios relate the expansion of your structure for inplane and out-of-plane directions.

In the case of your skins, they will impact the expansion inplane by either limiting the expansion (if their coefficient is smaller than your core) or trying to increase expansion (if their expansion coefficient is greater). Since this inplance expansion must be consistent with the core surface expansion, and the core expansion out of plane is related by poisson's ratio, so the end results can be affected by the face sheets.

 

[40818]

Think i was confused, i thought you were talking about the facing skins and as you said they were aluminium i could'nt make the step to the 3 CTE for each orthogonal direction for the aluminium skins themsleves.

 

[enriquelvarez]

Hi Gbor,

Thank's to the point you gave I have solved the problem.

I'm using NASTRAN 2005.

I should clarify some other conditios of the problem: samples are isostatically restrainted and cores have different E,G but same CTE in three orthogonal directions.

The main doubt I couldn't clarify was that if the problem is isostatically restrainted, deformations should just be defined by the CTE's.

However, if aluminium shells have a CTE in plane different from the ones of the cores, stress will be generated in plane and will be transmited to out of plane direction through poissons ratio. Deformation under these conditions in out of plane direction will be then determined by mechanical parametres as E an G of the cores. As they're different for each samples, deformations will be different too.

Finally, the problem was not so hard

Thank's very much.

Enrique