3D meshing

[kulandaivel_palanisamy]

Hello all,

I am using Hypermesh for meshing. I used do only surface mesh (Dual domain mesh) in Hypermesh and import the meshed file (nas) in to moldflow. After importing into moldflow, when i check the mesh match and reciprocal mesh match ratio, it shows less values. when i remesh the same mesh in molflow (dual domain) with the same edge length, the mesh match is getting improved. Is this method is ok to do Dual domain analysis?

Also, If i wish to use 3D analysis, is it necessary for me to improve the mesh match ratio before remeshing into tetras? or we can remesh the same hypermesh mesh (low mesh match or reciprocal mesh match) in moldflow in to 3D mesh?

Thanks in advance for sharing your views.

JK

[M.Giro84]

Hi,

I don't have Hypermesh but, for what I've seen, I think that there is an option dedicated to mesh for Moldflow but only for 3D. If you try to mesh with a DualDomain mesh in Hypermesh, you don't have the "match" and "reciprocal match" options. So you want to use the Hypermesh's mesh for a Dual Domain analysis, often it will not be good.

But, when you want to create a 3D model, the match is unsignificant, so you can start from the Hypermesh mesh and mesh in 3D in Moldflow or use directly the 3D mesh generated by Hypermesh (only tetras).

The mesh generated by Hypermesh is also useful in a Midplane analysis.

BR

Marco

[kulandaivel_palanisamy]

Hello Marco,

Thank you for sharing your information.

Regards

JK

[fukushige]

Hello, JK-san

Let me share one more thing with you.

We use other mesher "SimLab" for Moldflow as well as Hypermesh.

And, I agree with Giro's advice.

But you need to be careful that when you generate 3D Tet-mesh in Hypermesh and import it into Moldflow, flow-front temp result at the corner sometimes would be wrong.  That's happen when all of the nodes consist of one Tet-element surfacing the cavity surface (means four nodes on the part surface), heat transfer calculation would be wrong.

So, my opinion is that you can use the way to generate d/d mesh in Hyperworks then generate 3D mesh in Moldflow.

Regards,

Shige

[kulandaivel_palanisamy]

Hello Shige,

Yes i used to follow the same what u said. i do d/d mesh in Hypermesh and export as nas and import into moldflow.

and creat the 3D mesh in moldflow.

Do you have any information for the procedure to do 3D mesh of a symmetry part? Is it ok to do d/d domain mesh for one half of the part in hypermesh and bring the same into moldflow and create  a 3D mesh. and reflect the one half to form a full part to do simulation??

Thanks for your information.

Regards

JK

[fukushige]

Hi, JK

I'm sorry I don't know the best way to make a symmetry part.

I think at this time, it's very difficult to get the coumpletely symmetrical mesh model of part.

Actually, once 3D Tet-mesh was generated in Moldflow, the surface mesh pattern would not be such of the cross section of the part. That' way the mesh arrangement on the mirroring plane would be wrong when you reflect a half part to the other side.

Regards,

Shige

[kulandaivel_palanisamy]

hello Shige,

As of my experience, i do d/d mesh for one half of the symmetry part in hypermesh and i guess we have to make sure its a completely enclosed volume and export the mesh as nas. And import the d/d domain in mesh and we can again check is there any free edges here in moldflow before converting in to 3D mesh. importantly while reflecting make sure u carefully merge all the nodes in the mirroring surface. in which case you can organize all the mirroring nodes in one layer and make note of the number of nodes. and after reflecting make sure all the nodes are merged to have a proper mesh.

I'll appreciate any comments

[fukushige]

Hi, JK

Exactlly.  You are right.

But what I worried, as shown in the figure attached, the inside mesh of joined part would be strange sometimes.

Because the Tet-mesh on model surface isn't arranged as layer, that is way layer mesh looks like end at the joined plane.

Although the mesh itself is symmetry, I worry about these disarranged layer mesh might influence on the fill result.

Regards,

Shige