Wrong element axes in curved volumes Hypermesh with Abaqus interface

Hello, experts.

I have been trying to create a composite laminate with several plies with Abaqus by using Hypermesh as the preprocessor. The problem is that when I try to implement the cohesive model with shell elements, it does not work. So I must use solid elements.

My geometry comes from a cut of an sphere, so it has that curvature. When I create the 2D mesh in the face, the normals are right, but when I use that 2D mesh to create the 3D one and I display the element axes, they are wrong oriented and the local z does not correspond with a normal of the face.

Due to this issue I cannot implement the cohesive model correctly.

Any ideas?

Thank you.

Find more posts tagged with

English

HyperMesh

v2023.1

Accepted answers

All comments

Jason_Craanen

Hello,

Can you please verify your 2d normals are correct

And then explain which tool you are using to create the 3d?

and verify their normals?

Rafal Blasczykowski

Hello,

Can you please verify your 2d normals are correct

And then explain which tool you are using to create the 3d?

and verify their normals?

Hello, Jason. Thank you for your answer.

Yes, we've verified that the 2D normals and the 3D normals are correct, as you can see in the following images.

We have tried several tools for creating the 3D mesh:

-Solidmap

-Hexa

-Thinsolid

-Extrussion of the surface and one of the prior methods

We've also meshed in Abaqus CAE and we obtain always the same error on the elemental axes:

Orientation review tool to change elemental axes does not work and neither does material orientation tool.

Thanks in advance for your time

Community-Pasted-Attachment-1712220298520.png

Jason_Craanen

Hello, Jason. Thank you for your answer.

Yes, we've verified that the 2D normals and the 3D normals are correct, as you can see in the following images.

We have tried several tools for creating the 3D mesh:

-Solidmap

-Hexa

-Thinsolid

-Extrussion of the surface and one of the prior methods

We've also meshed in Abaqus CAE and we obtain always the same error on the elemental axes:

Orientation review tool to change elemental axes does not work and neither does material orientation tool.

Thanks in advance for your time

Are they simple plates skewed from the global axis? I know there is a visualization issue with the triads in this case. Can you verify the vector is correct in the deck?

Jason_Craanen

Are they simple plates skewed from the global axis? I know there is a visualization issue with the triads in this case. Can you verify the vector is correct in the deck?

I recommend you use the Material Orientation tool through the Composites Browser

Rafal Blasczykowski

I recommend you use the Material Orientation tool through the Composites Browser

Thank you for the reply, Jason.

Our composite is made of several plies, modelled with 3D elements SC8R, but the geometry is curved (convex). Apart from the curvature, the plate is deviated from the global axes, because it is a sided quarter from another bigger plate. We are using the material orientation tool, but the problem is that in one corner the elemental (or material) axes orientation is correct, but in the other it maintains the same orientation as in the previous corner (see images), not keeping the curvature.

Correct corner:

Other side corner:

About your image, we cannot select in the page Reference Orientation the element type, only property, so in X direction method we can only choose between Curve or System axis, not vector...

Community-Pasted-Attachment-1712302588414.png

Jason_Craanen

Thank you for the reply, Jason.

Our composite is made of several plies, modelled with 3D elements SC8R, but the geometry is curved (convex). Apart from the curvature, the plate is deviated from the global axes, because it is a sided quarter from another bigger plate. We are using the material orientation tool, but the problem is that in one corner the elemental (or material) axes orientation is correct, but in the other it maintains the same orientation as in the previous corner (see images), not keeping the curvature.

Correct corner:

Other side corner:

About your image, we cannot select in the page Reference Orientation the element type, only property, so in X direction method we can only choose between Curve or System axis, not vector...

As I mentioned before I have noticed a visualization issue with the triads that are skewed from global. Can you verify in the solver deck the orientation is correct?

Jason_Craanen

As I mentioned before I have noticed a visualization issue with the triads that are skewed from global. Can you verify in the solver deck the orientation is correct?

Are you able to share the model, because I am not able to reproduce

Rafal Blasczykowski

Are you able to share the model, because I am not able to reproduce

Hello, thank you for the reply.

We do not know how to print the triads in the solver deck, so we are attaching the model.

str_helpb.inp

Jason_Craanen

Hello, thank you for the reply.

We do not know how to print the triads in the solver deck, so we are attaching the model.

This is what I see at the ends at the corners.

Is this what it should be?

Community-Pasted-Attachment-1712757400545.png

Community-Pasted-Attachment-1712757420291.png

Rafal Blasczykowski

This is what I see at the ends at the corners.

Is this what it should be?

No, they are OK in one corner, but in the other they do not follow the curvature.

The thing is that we have been using a .inp without geometry and there is no way that the elements orient well.

We have redone the model, but this time keeping the Hypermesh file with the geometry. There we have oriented the material system with lines and surfaces from the geometry, and the material system now follows the curvature and automatically changes the elemental system too, so that they coincide.

This way we have achieved the correct orientation.

Thank you for the help!