Simulation Forums

Find Simulation answers, share expertise, and connect with your peers

Simulation Moldflow Insight


Very long times for "time to reach ejection temperature" - material or solver?

78 Views, 5 Replies
04/12/2018 8:24 PM

I compared two materials - BASF A3WG6 and BASF B3WG6, so PA 66 GF 30 to PA 6 GF 30, both have good thermal data and I would expect what I get in the first "sim-estimations".

With the same part I got about 20 s cooling time estimation from "Molding window" for the A3WG6 and 28 s
for the B3WG6 - expected I would say.

Then , with a 3D Fill, pack, warp analysis - based on the molding window findings - but without an actual tool geometry or even cooling lines - I get 28 s time to reach ejection temperature with the A3WG6 and 35s with the B3WG6.

-> More or less expected and somehow in a common range.

 

The same part in a tool with simple cooling lines (and based on the same meshes, just changed material and the most necessary parameters) gives me a result of 35 s time to reach ejection temperature for the A3WG6 and 600 (!!) s
time to reach the injection temperature for the B3WG6.
There are only a few dozens elements that shows this time, but the average of the part would also need about 80 s,
- which is too much difference for a good gut feeling.

 

Any ideas if the mistake is more related to the user or anything else (I assume it is a mistake).

 

Thanks in advance

Harald


Report Inappropriate Content
Message 1 of 5 ( Views: 78 )

Re: Very long times for "time to reach ejection temperature" - materia

04/13/2018 8:34 AM

Hello @harald.goetz,

 

I also tried to compare the two materials with/without cooling curcuit in 2018.2.

However I could not reproduce your situation.

(I also checked "time to reach ejection temperature, part" result which is in cool result.)

 

compare.jpg

 

Does your model contain any runner system?

(As you know, it is recommended that the gate beams are over 3 elements for the better prediction of "time to reach ejection temperature" result.)

 

If possible, can we look into your models?

 

Regards,

Mayumi



Mayumi O

Report Inappropriate Content
Message 2 of 5 ( Views: 68 )

Re: Very long times for "time to reach ejection temperature" - materia

04/13/2018 8:36 AM

Hi @harald.goetz,

 

Thanks for posting!

 

Does your part have metallic  inserts? Do you use a cool(BEM) for the analysis? We have seen in the past, that small metallic inserts together with our BEM method (mold defined also as steel) can create iteration issues and that may explain why you get some elements with higher cooling times.

 

If you want we can handle this as a support case!

 

I hope it helps!

Cheers,

Javier


______________________________________________________________

If my post answers your question, please click the "Accept as Solution" button. This helps everyone find answers more quickly!


Javier Jubierre

Report Inappropriate Content
Message 3 of 5 ( Views: 66 )

Re: Very long times for "time to reach ejection temperature" - materia

04/14/2018 6:20 PM

Thanks for your posts.

I couldn't answer from the workplace so I have to do it by memory...

 

I don't think the behavior is related to the material (there are no inserts!).

I started with molding window - the "educated guess" from it was

19,5 s for A3WG6

29 s for B3WG6

28 s for B3WG8 HS (another PA 6 with higher fluidity)

 

So after the molding window I used the same geometry - just new imported as "volume" and 3D meshed.

I did a Fill, pack and warp sequence with (hopefully) senseful parameters from the molding window analysis.

 

The numbers where other than in the molding window analysis - but the difference about the same.

28 s for A3WG6

35 s for B3WG6

34 s for B3WG8 HS

Makes sense to me - with ideal temperature around the part....


But there are long fingers to cool - so I did a analysis with simple cooling layout, a moldblock and I used the same mesh for all materials (FEM Cool analysis, not BEM - but maybe I should try that, too?)

 

35 s for A3WG6 (which makes sense to me! And it shows the potential when using conformal cooling or different insert materials).

But with both PA 6 I got about 600 (!) s - there where only a few dozen of elements that high, but even if I cut the range I did get a time about 100 s for time to reach ejection temperature - and this is nonsense...

 

I can and will exchange the file with the Autodesk Moldflow team, - just let me know who I should contact.

That might be the fastest way to find out what went wrong or if I made a mistake.
I might test an older version of Moldflow, too - if I find the time.

 

Have a nice weekend

Harald


Report Inappropriate Content
Message 4 of 5 ( Views: 56 )

Re: Very long times for "time to reach ejection temperature" - materia

04/18/2018 1:27 PM

Solution found (or Mistake found) by Javier!

 

I was a bit in a hurry it seems.

The main mistake that I made was to keep more or less the same settings with that kind of analysis.

So the overall process time was long enough for PA 66 - but not for PA 6.

(Only valid for the calculation *with* tool and cooling included).

 

The kind of simulation is meant to optimize in the end - not in the beginning.

So the time really needed to reach ejection temperature is estimated by the solver in case not everything is frozen at the end of the simulation - that leads to significantly wrong (*long*) results here.

 

Just by "giving more time", I ended up with a reasonable result of 48 s for the B3WG8 HS.
This makes absolute sense compared to the 35 s for A3WG6.
Both results can be optiomized with conformal cooling (of course).

 

Hope Jay Shoemaker doesn't want my certification back ...
(I do see John Oliver in front of me stating "RTFM Harald, RTFM!") :)

 

Thank you Javier for pointing out the problem and solution so quick.

Harald


Report Inappropriate Content
Message 5 of 5 ( Views: 31 )