Learning Forum

[Rameez_ul_Haq]

Hello. I am trying to solve a honeycomb structure, with a top plate thickness of 0.57 mm, core of 1.5 mm, and the bottom thickness of again 0.57 mm. The model is completely shell elements. However, I encounter the warnings as shown below while solving.

[Rahul Kumbhar]

You can try by adding command /nerr, ,99999999 under the analysis setting. Please check the this link for more details.nhttps://ansyshelp.ansys.com/account/secured?returnurl=/Views/Secured/corp/v202/en/ans_cmd/Hlp_C_NERR.htmln

[peteroznewman]

nWhen you say your model is all shell elements, how did you treat the honeycomb core? Did you construct actual hexagonal cells?nRead this discussion for that describes how to replace the hexagonal cells with a thick solid that is assigned orthotropic material properties that represent how the core behaves in bulk. This is a more efficient way to analyze a large honeycomb structure.n/forum/discussion/415/honeycomb-structure-adhesive-contactn

[Rameez_ul_Haq]

Actually, I treated the whole honeycomb structure as a single plate, with 3 plys on the top, then a ply of core in the middle, and then again 3 plys in the bottom. Is this a good approach? The properties of the plys and the core are adjusted in the Engineering data beforehand.n

[Rameez_ul_Haq]

But I don't understand if I use the approach above, or use a solid instead, then the cell gaps that are present inside the core will be neglected. And gaps mean stress concentrations near those gaps in the core, which we are basically neglecting. Am I correct or not? n

[peteroznewman]

ArraynUsing a single ply as the core gives you the warning you got because the element thickness is large compared to the shell element edge length. Shell elements were originally designed to model thin walled structures where the shell thickness would be less than the shell element edge length.nYes, when you use a solid block and assign material properties, you are ignoring all the stress concentrations that can occur down at the honeycomb cell walls and corners. This smeared orthotropic material core model is good for predicting the whole structure stiffness and deformation response.n

[Rameez_ul_Haq]

,but the stiffness, deformations and stresses, all of them actually depend on the geometric shape of the structure. How can you claim that this orthotropic core model (kind of like a core ply in the middle) will be a good approach to predict the overall structural stiffness and deformational response?nEven if it is a good model to predict the before mentioned, but still it won't be a good enough model to accurately predict the failure of the core, right?n

[peteroznewman]

nOf course the stiffness, deformations and stresses depend on the geometric shape of the structure.nAn orthotropic core model is a good way to predict the overall structural stiffness and deformation response because it matches experimental data. Checking that a model matches experimental data is called Validation.nI purchased a honeycomb core and cut rectangular samples with the long dimension aligned with the W and L directions of the core. These were put in a tensile testing machine to obtain the shear modulus G of the core, then pulled to failure to obtain the shear stress at failure. If you have those values, then the model can be used to predict the applied load that will result in a core failure.n

[Rameez_ul_Haq]

Arrayand Array, I tried using the APDL COMMAND /NERR, ,99999999 under the static structural analysis tree, but still the problem is not solved and the solution aborts before progressing.n

[Rameez_ul_Haq]

Since some messages were suppressed in the solution output file, I redirected myself to the .err file and found this error there, maybe the warnings limit within ANSYS has changed after the APDL command I input, but it is not working because of this error:nThere are two elements side by side which corresponds to the two errors I received in my output file. Can anyone guide me on how does the ansys decide on the limit of the radius/thickness ratio of an element, that below or above which it will return an error?n

[ramgopisetti]

I have gone through the case, can you try to post an image of what are your handling. It would be best if you try to putdown an snapshot of your BC aswell,so i can help you much better.nCheers,nRamn

[Rameez_ul_Haq]

I am still mystified because of this one. The reason being that for some elements, the radius/thickness ratio gives me a warning and for a bunch of elements, it returns me an error (notifying me of the reason that it violates the assumption of a shell element). I mean, what is the assumption of a shell element? As has already pointed out in one of his previous comments that the shell elements were originally designed to have thickness less than the element length, but where is this 'radius' coming from within the warnings and errors. I mean how can an element have a radius, either it can straight (if linear) or there can be a sharp angle within the element itself (if quadratic). What does this element radius actually mean?nWhy does some elements only return a warning because of radius/thickness ratio and why does some return an error? What is the standard used by ANSYS to decide that this element will only result in a warning and this element will result in an error (which would make my analysis to come to a halt). Can it be because of geometry?nHow can I solve this problem?n,unfortunately I cannot share any snaps of the model currently. The boundary conditions are fixed supports. But this error is occurring far from the fixed supports, without any force applied, without any contact, and I believe it is solely because of the element size and thickness given to that shell element, since solution doesn't even start to solve and straight away stops, giving this error.n