Can loads for orthogrid panels be entered at the panel midplane?

[HyperSizer User]

We are obtaining user loads from LS-DYNA, which calculates loads at the midplane rather than at HyperSizer's expected "midplane of the upper facesheet". 

I know that I can transform the forces and moments from the mid-plane to HyperSizer's reference plane.  However, it would be better if they could be entered directly into HyperSizer at the midplane.  How can this be done in HyperSizer?

[Phil]

Are your load primarily in-plane, membrane loads, or are your panels actually bending?

I ask this because if you are entering loads through the User loads and your loads are primarily membrane loads, you might be able to get away without shifting the reference plane. 

If you leave the Mx, My and Mxy as “constrained” on the FBD tab and just enter the Nx, Ny, Nxy from your solution, then HyperSizer will apply your membrane loads and then will calculate the necessary bending loads to keep the panel from bending.  This would be like a uniform end shortening solution.  When you shift the reference plane of a panel, the membrane loads (Nx, Ny, etc) do not need to change, but the Mx and My do.    However, if your panels are assumed to be loaded primarily in membrane and are not bending, then this approach could work quite nicely.

If you have substantial panel bending going on, however, this will not work.  If that is the case, then you will have to shift the reference plane by calculating new moments at HyperSizer’s reference plane, or, we will have to change HyperSizer to allow you to shift the reference plane.

Phil

[HyperSizer User]

Unfortunately, some of the panels do have noticeable bending.  One of our primary driving failure modes is skin stability, which can be influenced noticeably by bending moments.

The loads could be transformed to the appropriate (HyperSizer) reference plane.  That would involve recalculating the moments, as well as including additional Nx and Ny components needed for the transformation.  While feasible, this is not an optimal solution, as it creates the need for a standalone transformation code and naturally introduces additional opportunities for error.

In my opinion, the best options would be to either:
1.   Change HyperSizer to allow use of different reference planes using user-input loads. 

2.   “Trick” HyperSizer into thinking it is using FEM loads rather than user-input loads.  Granted, it would require a dedicated code to reformat the results, but it may be feasible to create a ‘fake’ .f06 file containing the desired loads. 

[Phil]

I think I understand the situation.  As HyperSizer is today, you can enter a user specified reference plane ONLY for unstiffened and sandwich panels, not for any stiffened panel concepts (including orthogrid).  This is true for both user entered loads and FEM loads.  In both cases, HyperSizer assumes that the reference plane is mid-plane of the upper facesheet.  Therefore, even if we had a way to read in loads from a “fake” F06 file, there would still need to be a load transformation to get the forces and moments into the reference plane that HyperSizer expects. 


I think the best course of action would be for us to implement a user-defined reference plane for orthogrid panels, and then give it the ability to read in this faked F06 file format.  It might be easier to use a simpler file format without needing all the extra stuff that goes into the F06 file.  I'll post back here when I know more about when this can happen.

Phil