Correlation of moldflow simulation results with real measurement

[Anonymous]

Dear Moldflow Users,

           I am doing moldflow simulation on clutch master cylinder body which is of Technyl A218 V30 material.It is a cylindrical component with 4-5mm thick, height around 150 mm, outer diameter will have two flanges and inner diameter will vary along the height from 24 t0 15mm. Kindly imagine the part.

           The objective of the project is to correlate the deformation (inner diameter) results of simulation with the real measurement.The simulation result shows much deaviation from the real measurement. This is because the measurement in real is taken in room temperature. But in moldflow simulation I have specified the cycle time as 50 sec. At the end of this cycle time the part has a temperature range of 70 to 130 deg C.

My doubt - Is it possible to bring the part to room temperature in the mold flow simulation.
                  In real, around 150 C the part get ejected and sourrounding air cools the part to room temperature.
                  We have to bring this effect inside moldflow simulation.

Process parameter
As in real
     Mold temperature - 80 C
     Injection temperature -295 C
     Injection time - 0.81 Sec
     Injection pressure - 894 bar
     Packing time (profile)- 11 sec
     Packing pressure (profile) - 500 bar
Assumed
     coolant inlet temp - 65 C
     cycle time   - 50 sec

Thanks and Regards

Ramesh

[Anonymous · Accepted solution]

Moldflow warpage is supposed to predict the shape of the particle at room temperature after cooling.

Either there is something wrong with your setup of the simulation or Moldflow software does not work for your settings very well.

Just a few common sources of simulation errors:

• Have you adequately inputted packing profile/cooling time (never use the defaults!)?
• Have you inputted filling profille?
• Is your V/P transfer realistic?
• Have you modelled runners (sometimes modelling the barrel and injectors of the injection molding machine helps)
• Have you modelled all the significant features of the part?
• Is your mesh correct (low aspect ratios, no intersecting elements, etc.)?
• Have you modelled cooling?
• Do you measure moldings after the number of shots is enough to have the mold in the steady-state conditions?
• Is you material shrinkage-tested? (for 2.5D)
• If the polymer is water absorbing (like PA6, eg.) have you dried it enough?
• Does the plastication is good for the part (the air shot is clean and does not need much pressure)?
• Is the injection pressure predicted well? 

[raalteh · Accepted solution]

Hi,

What Alex says is correct; the warpage results that Moldflow predicts are at room temperature, so it is already taking into account the cooling down from 'right afte ejection' to room temperature.

There is a longer explanation how it is done, but I'll spare you those details.

Hanno

[Anonymous]

Hi Hanno ,

                I'm simulating the same part with different gate locations and timings to understand the warpage behavior and choose the best among them .Can you share the details of theory behind how moldlfow will take into account the  temperature difference between ejection temperature and room temperature in the Warpage calculation. 

Thanks and regards 

Ram

---

@raalteh wrote:

Hi,

What Alex says is correct; the warpage results that Moldflow predicts are at room temperature, so it is already taking into account the cooling down from 'right afte ejection' to room temperature.

There is a longer explanation how it is done, but I'll spare you those details.

Hanno

---

[raalteh]

The solver will before the warpage analysis takes place, cool down the part to room temperature, and it will do this instantaneously.The cooling down to room temprature will increase the stress levels in the part before the warpage analysis starts. This is the quenching assumption.

Another way to look at this is to assume that the end of the molding process, the part cools down to room temperature inside the mold, and is then being ejected and warps.

The assumption is pretty reasonable as long as the material (in reality) is pretty cold by the time the part is ejected.

I would expect that with time we will move away from the quenchinc approach and account for the cooling down path between ejection and reaching final cooled state, but this will be computationally intense.

Hanno

[harald_goetz]

Hello Hanno,

"me too", please. I would like to know if that is truth for rubber as well?
Some of your competitors have exactly that in their (marketing) focus, but I'm more interested in results (especially real world results).

Regards
Harald