Sol106 limitations

[Nix52]

Hi all,
I've created the model in NX...and wanted to solve thermal stress analysis problem using sol106...for my problem following is the description-
1. Used nonlinear material curve
2. Large displacement on
3. Elements used- lower order hex, pyramid, tetra
The problem is first solved using contact and without contact in second time. But in both cases the problem was not converging. I also tried this problem by simply using sol101 and sol400 (MD nastran) and in both solvers the problem was converging. So I wanted to know is their any limitation of sol106 for thermal stress analysis (considering thermal gradient) or for elements or for materials Or their is some thing else which I need to consider?

Need urgent help...and thanks in advance.

 

[nlgyro]

I'm a bit surprised that you got a pyramid element running in MD Nastran. The last that I checked MSC/MD Nastran do not support pyramid elements. MSC choose to go the way of glued contact to connect hexa and tetra meshes.

If I can ask, which version of MD Nastran did you use to get your successful completion of your job?

 

[nlgyro]

Quoting:

> 1. Used nonlinear material curve

Supported

> 2. Large displacement on

Supported but note that SOL106 is a large disp, small strain assumption.

> 3. Elements used- lower order hex, pyramid, tetra

Supported, although I would recommend against using a tet4 element. Use a tet10 with the corresponding pyramid element order to connect to the Hexa8.

In addition to this I would suggest you run your model using SOL601. This is more robust that the traditional SOl106, and supports large disp/large strain assumption.

 

[Nix52]

Just a small correction it's not pyramid element, it's prism element which we have used along with hex and tet. So you were right about pyramid
The version I am using is 2013.

The fatal error which I am getting is-
Fatal non-linear element error.

I hope you can guide me better with this discription.

 

[nlgyro]

Can you post the portion of the f06 file which has the FATAL msg?

 

[Nix52]

Please find the below attachment.

 

[nlgyro]

Is this msg obtained with or without contacts?

 

[Nix52]

Without contact.
Also when I used hex and penta elements only and not tet elements same error is occurred.

 

[nlgyro]

The error msg. doesn't look so obvious, one has a look at your model to understand what you're trying to simulate.

Looking at the NLITER table at load-step 0.9625, the code is having a hard time achieving convergence. Have you tried using a full-newton raphson method? If not, please do as this helps.

Meanwhile, the iteration table that you have shown is not for a standard SOL106, looks more like for a SOL400.

 

[Nix52]

Thanks for the input.
Just one more que if I need to use glue contact in sol106 (MSc)then how to define glue contact. Do u have any material to read about glue parameters in msc.

 

[nlgyro]

SOL106 is not capable of handling glue contact. This has to be achieved with SOL400.

The MSC Nastran Non-linear users guide is good starting point. You can download this at:

https://simcompanion.mscsoftware.co...4&cat=MSC__MD_NASTRAN_DOCUMENTATION&actp=LIST

You'll need to register on the site.

 

[Nix52]

Thanks nlgyro
That is a great help from you...

 

[Nix52]

How NX glue contact is different from msc glue contact for sol 106 and sol400?