Moldflow Insight Forum
Welcome to Autodesk’s Moldflow Insight Forums. Share your knowledge, ask questions, and explore popular Moldflow Insight topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Ear Flow

10 REPLIES 10
Reply
Message 1 of 11
Maudet.Francois
1463 Views, 10 Replies

Ear Flow

I’m disappointed by a ‘bad’ Moldflow filling result. You can see briefly attached the behavior of the physical flow and the simulation.

Today there is a weld line which was not detected by Moldflow.

I tried to remesh the tetras along the edge of the component but the result is not better.

It’s dangerous because I can’t explain to my colleagues why we can’t trust the software only for this case. I hope that they will not think that is a generality.

What is your feeling about this topic?

 

PS : It's an Outer Lens Component for rear lamp product injected by PMMA 8N

10 REPLIES 10
Message 2 of 11

look like your parts are diffrent...can you upload you part?

Message 3 of 11

Unfortunatly it's not possible to upload the part.

But the observation is very simple. There is a weld line which is not intented by Moldflow because the flow front behavior is not the same that the physical process. The velocity is higher along the edge of the part.

Message 4 of 11

I would suggest looking at the mesh in that area.  If it's dual-domain, make sure the thickness is correct to show the thicker edge.  Not matter what the mesh type is, make sure there are enough elements along the edge to pick up the flow/weld line.  (ie, if your mesh size is 1", and the flow leader along the edge is only 1/2", then it won't pick up the flow difference)

 

You could always try turning inertia effects on.  Sometimes this helps.  But it makes it tricky when you need to simulate things correctly in the first place, not after the fact.

Tim

Message 5 of 11

Thanks for your advices Tim,

 

This simulation was done by 3D model. I remeshed localy the edge to 10 elements (see the picture attached) and a ratio close to 1 but there was no effect on result.

Concerning the inertia effect I already tested it but I guess that it's not really significative for a thin part (2.5 mm). But I can launch one for test.

 

François

Message 6 of 11

Hmm.  Your mesh looks more than adequate to pick up any flow differences.  Again, you could try inertia, but I agree with you that it typically doesn't help unless your fill speed is really fast.

 

Best of luck.

Message 7 of 11
phammer
in reply to: Maudet.Francois

Hello, 

 

its complicated without seeing the whole mesh, but sometimes it can be usefull to mesh 3d with 12 and more layer. And it could be better in your case to mesh the whole part with finer mesh density. What moldflow version do you use?

 

What quality is your material file. Gold?

 

if you send me the cad file by peronal mail, iI would test it for you. Should be fast enaough my maschine. 2 processors 96GB.

 

Philipp

Message 8 of 11
Sandriver_Eagle
in reply to: phammer

Hi Francois,

 

I need a little bit more information flow to my ear because you are not able to publish the model itself.

 

  • What is the actual thickness ratio in the model - flow leader outside, thickness in the center?
  •  How far is the distance of the gate location from the problem area?
  •  Which PMMA you are using from the Moldflow database?
  • Filling time/melt temperature and mold temperature?

This inputs can help me to capture the problem conditions more then yet and brings us hopefully to an helpful answer.

 

br, Anton

Message 9 of 11

Hi Francois,

It's the first time I see this kind of bad flow results.

It seems that your study doesn't considers cooling lines. What I suggest is checking the cooling parameters and temperature distribution during the trial. In fact the mold temperature may be not so uniform, this could explain such big difference from moldflow study.

Message 10 of 11

I think this trouble can come from the shear heating.

Due to the simplified approach of most common mold-filling simulation software, the shear influence on the polymer flow is not diagnosed properly.

 

Look at this :

http://www.beaumontinc.com/files/documents/mf-9-flowanalysis-742007.pdf

http://www.beaumontinc.com/rheological-control/meltflipper-design-services/meltflipper-max/

http://aucache.autodesk.com/au2011/sessions/4447/class_handouts/v1_MA4447-P-J.Beaumont%20Handout.pdf

http://www.sigmasoft.de/casestudy-category/plastics-applications/

Signature: "Maybe Moldflow does not work properly, but the real world neither" my son...6 years old 😉

Message 11 of 11

Hi Francois,

Even I have experienced similar results for one of my ABS parts.
When I did the simulation at first with 3D mesh, I was not able to capture the flow perfectly, as we saw in the actual molding. It was the same weldline problem which we were not able to aniticipate in simulation.

After some experimentation we were able match with the actual flow. The first problem was mesh type. 3D was not able to capture that flow pattern, we were able to see it in Dual Domain mesh type. Secondly it was the material grade, we did not use the material which is used in actual. But it was same family plain ABS material. We got the actual flow after changing the mesh type and using the actual material grade.

So, I suggest you to try DD mesh and use exactly the same material grade for simulation as used in actual.

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report