Discussion Groups

Simulation Moldflow Insight

Reply
Active Member
andreeccel
Posts: 10
Registered: ‎10-21-2011

3D mesh x Dual domain mesh

580 Views, 10 Replies
03-21-2012 06:27 AM

Hello,

 

I'm working with a specymen and Ive made the simulation with a 3D mesh and a Dual domain mesh. Both the mesh follow the mesh criterias of the Moldflow. The problem is that the cycle time result of the 3D mesh is 45% higher.

Anyone know why is that happenning?

 

Thanks,

 

André Eccel Vellwock

Please use plain text.
Valued Contributor
nishit78
Posts: 82
Registered: ‎08-05-2009

Re: 3D mesh x Dual domain mesh

03-21-2012 12:06 PM in reply to: andreeccel

Check the thickness diagnostic in Dual domain and make sure it does not under-estimate the thickness distribution. For example the place where two wall intersect is actually a thick area but dual domain may underestimate the thickness. Whle the 3D mesh will more accurately estimate the true volume.

 

I'd trust 3D more, in general.

 

That being said, in reality the may be able to eject the part sooner that the 3D estimates if the wall is adequately rigid.

Please use plain text.
Active Member
andreeccel
Posts: 10
Registered: ‎10-21-2011

Re: 3D mesh x Dual domain mesh

03-22-2012 06:19 AM in reply to: nishit78

Thanks for your answer.

I dont have this problem because the thickness of the wall is constant.

You can see in the printscreen attched that the part is very simple, so the dual domain's volume is the same as the 3D's.

 

Please use plain text.
Valued Contributor
nishit78
Posts: 82
Registered: ‎08-05-2009

Re: 3D mesh x Dual domain mesh

03-22-2012 06:26 AM in reply to: andreeccel

I'd still check the thickness diagnostic for Dual-Domain just to be sure that the thickness assignments are not faulty.

 

Other than that, i can only advise to re-check you process parameters and resin data between the two different setups.

 

Perhaps you could also see which area is controlling the cooling time in both the setup and narrow down your problem area.

Please use plain text.
Product Support
nordhb
Posts: 453
Registered: ‎03-23-2009

Re: 3D mesh x Dual domain mesh

03-22-2012 08:12 AM in reply to: nishit78

Hi,

looking at the picture of specimen part, the thinner passage is almost quadratic in cross section.
Then, the mesh will be too coarse for 3D tetras.
Have a look of mesh inside.

A finer mesh overall, and maybe and some layers for 3D (default 6, maybe change to 8 or 10).
This should improve the result.


3D mesh coarse vs fine.jpg


Regards,
Berndt



Berndt Nordh
Please use plain text.
Active Member
andreeccel
Posts: 10
Registered: ‎10-21-2011

Re: 3D mesh x Dual domain mesh

03-22-2012 09:15 AM in reply to: nordhb

Thanks for the reply.

I've improved the mesh quality for the 3D but the cycle time havent changed.

And the thickness is OK in the dual domain mesh.

Please use plain text.
Employee
INACTIVE_AstroJohn
Posts: 492
Registered: ‎03-25-2010

Re: 3D mesh x Dual domain mesh

03-22-2012 10:07 AM in reply to: andreeccel

Hi,

 

I am not a Moldflow user nor an injection molding engineer, so someone please correct me if I am wrong. I think the part is not suitable for the dual domain analysis. The documentation says that one requirement for dual domain is "the minimum length and width of any local region should be greater than four times the local thickness. A more conservative estimate of ten time the thickness ensures more accurate results".

 

I believe the dual domain analysis does not consider some (or any?) effects along the edge of the model, so the resin flows more easily through the narrow section in the dual domain analysis than it does in reality.

 

 

 

Sincerely,
John Holtz, P.E.
Senior User Experience Designer, Simulation
Autodesk, Inc.

Current version of Mechanical & Multiphysics: 2013 SP1 (2013.01.00.0012 28-Jun-2012)
Please use plain text.
Active Member
andreeccel
Posts: 10
Registered: ‎10-21-2011

Re: 3D mesh x Dual domain mesh

03-22-2012 11:22 AM in reply to: INACTIVE_AstroJohn

Sorry, in what documentation have you found this?

 

I checked in the Moldflow 2011 and the recommendation that Ive found is:

Dual domain mesh -" These layers enable an accurate representation of thin, cross-sectioned parts, where there is a rapidly changing characteristic profile, for example, temperature and flow-front velocity."

 

But Ive found another interesting thing, check the printscreen, that can be the answer for this.

Actually, the software is prediciting a heat loss in the z direction (the thickness direction). It can be estimating uncorrectly. If this has a relation with the thickness, I cant afirm, because i dont know how the software calculate this heat loss in z direction.

Please use plain text.
Active Member
andreeccel
Posts: 10
Registered: ‎10-21-2011

Re: 3D mesh x Dual domain mesh

03-22-2012 11:40 AM in reply to: andreeccel

Refreshing:

Ive tried to rotate all the model to see if there is a relation with the coordinate system, but nothing happenned.

Please use plain text.
Employee
INACTIVE_AstroJohn
Posts: 492
Registered: ‎03-25-2010

Re: 3D mesh x Dual domain mesh

03-22-2012 11:50 AM in reply to: andreeccel

Hi,

 

The passage that I found is from the online 2012 Help: "Choosing an analysis technology".

 

I presume the passage about estimating the heat losses in the "Z" direction are referring to a local coordinate system where X and Y are in the plane of the element and Z is perpendicular. I would be very surprised if it was referring to the global X, Y, Z directions. So rotating the model should not change the results.

 

Sincerely,
John Holtz, P.E.
Senior User Experience Designer, Simulation
Autodesk, Inc.

Current version of Mechanical & Multiphysics: 2013 SP1 (2013.01.00.0012 28-Jun-2012)
Please use plain text.