News: Need training? HyperSizer Training Videos are available now! Learn more here:

Recent Posts

Pages: [1] 2 3 ... 10
Material Strength Composites: Ply Approach / LaRC03 Fiber Failure Theory
« Last post by Ömer on Yesterday at 06:56:22 AM »
First of all, we calculated toughness ratio (g) by using thin ply formulas as mentioned in "Ply Based Composite Strength Document" to obtain fiber failure.

Because of the lack of fracture toughness terms (G1c, G11c), we assume Ytis= Yt and Sis=S. Also, the angle of fracture is taken as 53 degree as shown in "Composite Strength,Ply Based" Document".

For the our case of fiber compression (Sigma11<0), ply stresses are transformed into the new m coordinates and Margin of Safety(MoS) values has been calculated. In our case, Sigma22m is bigger than 0.

The example case is given in Attachments. We calculated MoS as 0.7, but HyperSizer shows 0.134.

We have been seen that the results of HyperSizer and the aforementioned document are very different.

What is the reason of this difference? Which one is more reliable? And, if HyperSizer result is more acceptable, how can we calculate the same value analytically? Could you please check it by using any values that are necessary for LaRC03?

Thank you.
Thanks James

Is the segment approach designed around the using the summed object loads?  I asking because I believe the preferred approach to sized a discretely stiffened panel is to use Plate and Bars concepts first, which does not use the summed object loads and then do a final sizing with stringer segments.  As this approach is deemed quicker.

However if the Plate and Bar approach doesn't use the summed object load, and the uses the individual object loading, then stringer segments will want to but more loading within the skin instead of the stiffener.

Is there a preferred process for sizing a discretely stiffened panel?

It's okay to keep the segment loads flag set = Yes for final MS.

It depends whether you want to use the FEA-computed local object loads, or allow HyperSizer to compute the local object loads from the total segment load. Perhaps on the final pass you may want to use the FEA-computed local object loads because the FEM and the dimensions/materials in HyperSizer are the same. If you're okay with always using the segment loads approach to determine the object loads, then keep it = Yes.

FYI.. there's good documentation on this load summation approach. See:

I hope this is helpful.


The Hypersizer help states that this Backdoor data tab should be switched 'ON' for sizing, but yet switched 'OFF' for final margins.  Is there a reason for this, as switching this setting off and on makes a big difference to the final panel sizes.

Panel Buckling / Re: Orthotropic Flat Plate SSSS Shear Buckling
« Last post by Stephen on January 17, 2020, 10:09:34 AM »

It turns out that the HME document you are referring to actually has a couple of typos that are resulting in the difference you are seeing there. The actual reference equations are correctly reproduced in Word stress reports from the software, and attached below for reference. I've noted this so we can address it in a future revision of the HME.

An additional consideration if you are attempting to replicate the HyperSizer result is that we always use compressive stiffnesses when doing buckling calculations, even for shear.

Panel Buckling / Orthotropic Flat Plate SSSS Shear Buckling
« Last post by Ömer on January 17, 2020, 02:38:46 AM »
We calculated shear buckling critical load, Nxy_cr, both in HyperSizer and according to Method Documentation of HyperSizer (Plate and Shell Equations.HME in Buckling of Failure Analysis). We made this comparison for a few component and we saw that these two result which belong to HyperSizer and Method Document are very different in some cases, although they are similar sometimes.

For the parameters below:

D is the stiffness matrix,
D11: 1050575
D12: 184745
D22: 428207
D33: 211547.5

Buckling Spans,
a: 538.3 mm
b: 1405.5 mm

Shear Load,
Nxy: -40 N/mm

Method Result: 108.76 N/mm
Result is showed by HyperSizer: 158.6 N/mm

What is the reason of this difference? Which one is more reliable? And, if HyperSizer result is more acceptable, how can we calculate the same value analytically?

Thank you
Thanks Stephen

Post processing the updated FEM before the Nastran run seems a possible solution.  I'll try that.


Analyzing & Optimizing Stiffened Panels / Re: CBAR pin flags removal
« Last post by Stephen on January 16, 2020, 12:50:06 PM »

Interesting, I'll have to look into that. Thanks for the update and I'm glad it seems to be working as expected now!

Analyzing & Optimizing Stiffened Panels / Re: CBAR pin flags removal
« Last post by Gawain on January 16, 2020, 12:49:08 PM »

Yes it only worked once I reimport the FEM.  It didn't recognise the pin flags until I reimported the FEM.


This issue was actually previously identified and fixed in a more recent version of HyperSizer. It was indeed a bug and not the intended functionality.

If updating to the latest version is going to be a challenge, I think the remaining options would be to use DSM from the beginning (which will have implications on the speed of sizing and analysis), or to have a script that post-processes the updated FEM and corrects the bar offsets. The latter is definitely possible but would be somewhat involved (as you would need to map between property IDs and CBAR elements to determine the appropriate element offset).

Apologies that I can't really think of a better solution right now..

Pages: [1] 2 3 ... 10