Failure Indices (Max. and Ply) do not match 1

[pcompg]

I've solved a composite structure in MSC Nastran with 4 plies below the meshed surface (PCOMP). On top of the meshed surface is honeycomb core modeled as solid elements and again 2 plies on top of core. Shell elements on top of core and below the core are in separate groups. The non-core region has 4+2=6 plies which are in a different group. I've set up Tsai-Wu failure theory. The failure indices extracted by "Failure indices, Maximum indices" and the maxumun of "Failure indices, Ply indices" do NOT match.

When I model a structure with constant number of plies throughout without the core, the failure indices extracted by the 2 methods match exactly.

Do you know the reason why failure indices do not match in the first case with core present.

 

[Spirit]

Ameyad,


I assume you are using MSC PATRAN. In this case, consider that the "Failure indices, Maximum indices" plots the maximum between 'Ply' and 'Bonding' failure indices. So if the latter is higher, this one is plotted instead of the lamina index. If you use the "Failure indices, Ply indices", the maximum is calculated only among lamina indices. If you use a very high interlaminar bonding value (field 5 on PCOMP card) the two options should actually give the same output.


I hope this helps.

Regards




'Ability is 10% inspiration and 90% perspiration.'

 

[pcompg]

Spirit,

Thanks a lot for the info.

Another question is when checking for failure indices, is is good to go for FI indicated in fringe results with avaraging domain set to "all entities" or with averaging domain "none"? FI in second case are higher that match with FI in f06 file for each element.

 

[SWComposites]

You should NEVER average failure indices. You should also be very careful when averaging stresses/strains in contour (fringe) plots as peak values can be suppressed. The best approach in general is to turn off all averaging in contour plots (unless making pretty pictures to show to management).

 

[pcompg]

Thanks SWComposites!