topic Re: [Question]Difference of temperature result between fill and cool(fem) in Moldflow Insight Forum

[Question]Difference of temperature result between fill and cool(fem)

sea512 — Fri, 07 Feb 2020 06:31:48 GMT

Hi, all

I checked the temperature result using Cool(fem)+Fill+Pack sequence.

When I checked the results which are 'temperature, part(transient):XY Plot and Temperature:XY Plot, the results are different.

I do not know why it can happen. What is the exact solution?
Is there anybody who knows about this?

Re: [Question]Difference of temperature result between fill and cool(fem)

bernor_mf — Fri, 07 Feb 2020 14:58:36 GMT

Hi @sea512 ,

this difference is expected, I would say.

The results are produced by the different solvers for Cool and Flow.

Cool : Temperature, part : the solver initially instantly fills the cavity with melt at melt temperature to get a starting point for iterations. So the temperature max should be set melt temperature.

Fill or Fill+Pack(Flow): Temperature : this is the temperature of melt during filling and packing of cavity.
Increase could indicate shear heating.

Cool: Temperature, mold-cavity interface : review this result too.

Please, also review the on-line help for explanation:
Cool results:
Temperature, part (transient) result
Temperature, mold-cavity interface (transient) result
Flow results:
Temperature result (3D)

Hope this helps.

Regards,

Berndt

Re: [Question]Difference of temperature result between fill and cool(fem)

sea512 — Fri, 28 Feb 2020 04:32:16 GMT

Dear.

It was so late to see your reply.(Sorry about that)

I would need more deeply thinking about this.

Thank you so much.

Re: [Question]Difference of temperature result between fill and cool(fem)

bernor_mf — Fri, 28 Feb 2020 09:39:10 GMT

Hi @sea512 ,
no worries.
Maybe I do not understand you fully, what you would like to know and understand here.

First, you need to understand that the temperature results you discuss and show in picture are results produced by different solvers, hence different solutions and results.
So this happens as the results comes from the different solvers Cool (Temperature, part(transient):XY Plot)
and Flow (Temperature: XY Plot (3D)).
A difference is expected.

Cool solver:
Temperature, part(transient):XY Plot : this result is produced by Cool solver.
The main objective for Cool is to calculate mold temperature from cooling circuit design.
As a starting point for Cool solver iterations is that it uses cavity instantly filled with melt
at set melt temperature in process settings.
Cool solver does not consider the filling of cavity as such, and hence no shear heating.
So max temperature will be set melt temperature, as set in process settings.
This is seen on your picture, to the left.

Flow solver:
Temperature result (3D), Temperature: XY Plot (3D) : this result is produced by Flow solver
In your picture it seems to be node inside skin.
The melt seems to be shear heated as temperature increases by time in filling, then cools down.
As flow solver considers melt flow and shear heating of melt the change of temperature is seen,
and Cool solver does not consider filling of cavity as such, and hence no shear heating.

To me this is expected, and by design.

If you run Cool FEM, transient, Flow at every iteration (takes a long time), you will not get

Temperature, part(transient):XY Plot result.

This is how I would reply to your questions and concerns.

Let's see if anyone in community has another comment to this.

Thanks.

Regards,
Berndt

Re: [Question]Difference of temperature result between fill and cool(fem)

sea512 — Wed, 04 Mar 2020 00:05:18 GMT

Thank you so much.

I understood that cool(fem) temperature result has some assumption so, the temperature result has a limitation as melt temperature(key in value in injection condition). And Fill temperature result do not. Am I right?

Through this inquiry, I want to know about the difference or approximation between CAE temperature and Real injection molding temperature.
Because through this knowledge, I want to predict the ejection time for thick part molding.
The attached file shows that the temperature results of CAE and the experiment are different.

Of course, I think that there are just so many reasons to make a difference.

But, I want to make the criteria result in CAE for comparing with the experiment result.
I think that the result of temperature through fill solving is a suitable choice for this case.
Please reply to my opinion.

Re: [Question]Difference of temperature result between fill and cool(fem)

bernor_mf — Wed, 04 Mar 2020 08:16:35 GMT

Hi @sea512 ,
you are welcome.

Thank you for providing the details,
I better understand your issue.

Yes, you got it right:
Cool assumption : starting point for iterations is melt temperature as keyed-in in process settings.
This temperature will not change during simulation.
Flow: melt temperature a starting point, but could get shear heated during filling, increasing melt temperature.

Yes, I agree that the temperature from fill results should be used.

Some comments:

Be careful of used node to trace temperature.
So not in shear region, as this might show higher shear heating that thermocouple actually measures.
You could to an Temperature: Probe XY plot at same position as node, to understand what happens
with temperature through thickness over time.

The starting melt temperature in your case: 210C
The real value from thermocouple shows starting around 210-212C : a little shear heating
Moldflow: shows more shear heating at point/node at distance from thermocouple 225 - 239C : so quite a difference.

So the shear heating in Moldflow is higher than expected.

Could depend also depend on trace node position used.

Now, the shear heating of melt seems to happen upstream from node in plot.
Try to follow it back stream to understand where and why it starts.
Is it in sprue, runner, gate, and adding up?
Check temperature at sprue start, and then at gate: has temperature already increased 10-15C?

Then the runner system shear heating is increasing the melt temperature.

The mesh: how many tetra layers?
How thick is the part?
You might need more than default 10 tetra layers through thickness to capture what is going on.

Further, I noticed that the temperature slope from thermocouple trace is slightly steeper than Moldflow.
This could be due to Mold-melt Heat Transfer Coefficient (HTC) should be different for thick wall.
You can change this to investigate, if needed.
In Process settings > Advanced options.. Solver parameters > Edit...
tab Mesh ; default:
Filling HTC 5000 W/m^2
Packing HTC 2500 W/m^2
Detached HTC 1250 W/m^2 (part is detached: i.e., pressure is zero)

Hope this helps.

Regards,
Berndt

Re: [Question]Difference of temperature result between fill and cool(fem)

sea512 — Thu, 05 Mar 2020 04:52:22 GMT

Thank you so much.

It is very helpful advice.