Aganin about link top/bottom facesheets of sandwich concept.

[cmcho]

I have been trying to size a sandwich panel structure fully symmetric with unsymmetric facesheets such as [45/-45/core/-45/45], /45/0/core/0/45] etc. in order to Bij=0  Trying several times with various options, I found a solution finally by using backdoor parameters, which Hypersizer answered in user forum in 2014. The successful procedure is as follows:
1) backdoor data set
     Geometry Check-Beam/TrussCore Sandwich-Generate only Symmetric Cross-Sections: Yes
2) Layup Rules(Top & Bottom face)
     Unchecked: Symmetry, Balance
     top: first plies [45/-45..], [45/0..] etc
     bottom: last plies [-45/45..], [0/45..] etc
3) Tooling rules: unchecked link option
4) Rapid sizing
5) Detailed sizing set(necessary for design variables set)
    a. applied set variables, then facesheet link option was automatically set to "checked" in tooling rules
    b. I did "unchecked" facesheet link option again(important!!)
6) Detailed sizing
7) Confirm sizing with Bij=0

Only this process makes possible to size symmetric whole sandwich structure.

HOWEVER,
when I expand this procedure to 6 plies of top and bottom facesheets with only two first and last plies set, I failed. Hypersizer gave me same thickness for top and bottom facesheet, but not the symmetric cross section nor ply percentage. Of course, Bij is not zero.

Could you check this, and suggest any solution for symmetric cross section design of sandwich panel?

Thanks.

Changmin.
   

[max.weintraub]

Hi Changmin,

I have a couple of questions to clarify your process so that I can help you get it working again.

It sounds like you are using rapid sizing so that you can run set variables for detailed sizing, and then running detailed sizing with discrete laminates. Is that correct?

Do you have any other stacking rules other than the facesheets be mirrored about the core?

Thank you,

-Max

[cmcho]

Hi, Max
Thank you for your assistance.

1. Yes. Rapid sizing is for setting variables in detailed sizing. As I mentioned, when I ran set variables, then top and bottom faces was automatically linked again in tooling rules. After removing link again, I got the discrete laminates for each top and bottom faces separately.

2. No. I used only first plies in top face and last plies in bottom face in layup rule, and the others such as symmetry, balance etc. were not employed. In ply angle rules, min. 2 plies and 10% of 45 deg. Minimum zero percentage for 0 and 90 deg were applied.

While I have been waiting for your reply, I have found a solution for this by using "material family" for top and bottom surfaces, but I would like to look for a more general way.

Thanks.
Changmin

[max.weintraub]

Changmin,

Sorry for the initial slowness responding to your question. Glad to hear that you found a solution that is working for you!

Unfortunately, the backdoor option you referred to in your initial post, "TrussCore Sandwiches - Generate Only Symmetric Cross-Sections", does not work for honeycomb or foam sandwich concepts.

There is not a perfect solution to this at the moment. Your likely best option is to use laminate families, with the facesheets having mirrored sequencing. This will ensure that as long as the facesheets are the same thickness that you will have a symmetric cross-section.

Hope this helps!

-Max

[cmcho]

Thanks, Max.

If we use symmetric laminate for each facesheet, we don't need to consider this case. In practical use, however, we sometimes meet this condition for very low loading sandwich structures. In general we seldom use sandwich structure with a thick facesheet.
I hope this mirror facesheet sizing is possible in Hypersizer sometime.

Pleased to communicate with you.

Changmin