CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)

 Near April 21, 2012 10:53

about choosing Turbulence model for multiphase flow problem.

hi everybody,

i am now dealing with a multiphase flow problem (gas-assisted injection molding, as the pic show, i just focus on "the second step") by vof.
Seems that the gas phase and the air phase (white in the pic(second step)) are Turbulence flows(Re>6000), and the melting plastic phase(black in the pic, second step))is laminar flow (Re<1).

so,the melting plastic was set as laminar flow(By udf), and the Turbulence model was choosen as k-e, SST, Re Stress. But, all these simulation was no so match with experiment result--the wall thickness was much smaller than that of experiment result.

PS:i am not so good at the parmetres-setting thing.
http://www.jeco-tech.com/en/upImgFil...2221219673.jpg

 LuckyTran April 21, 2012 17:50

Quote:
 Originally Posted by Near (Post 356016) hi everybody, i need your help, i am now dealing with a multiphase flow problem (gas-assisted injection molding, as the pic show, i just focus on "the second step") by vof. Seems that the gas phase and the air phase (white in the pic(second step)) are Turbulence flows(Re>6000), and the melting plastic phase(black in the pic, second step))is laminar flow (Re<1). so,the melting plastic was set as laminar flow(By udf), and the Turbulence model was choosen as k-e, SST, Re Stress. But, all these simulation was no so match with experiment result--the wall thickness was much smaller than that of experiment result. you got any advise? PS:i am not so good at the parmetres-setting thing.
I don't think this type of problem is appropriate to solve using Fluent or turbulence model in general.

1) The role that turbulence plays is at best debatable for this problem. I think it is better to use 0D or 1D calculations and not rely on CFD. You can probably choose any turbulence model (including laminar) and get very close to the correct result.
2) injection molding is a well known non-Newtonian flow. Throw on top of it a solidification front and CFD will not like that.

Again, I think simple 0D and 1D calculations are more appropriate, and probably more accurate for this problem.

 Near April 21, 2012 20:55

Quote:
 Originally Posted by LuckyTran (Post 356112) I don't think this type of problem is appropriate to solve using Fluent or turbulence model in general. 1) The role that turbulence plays is at best debatable for this problem. I think it is better to use 0D or 1D calculations and not rely on CFD. You can probably choose any turbulence model (including laminar) and get very close to the correct result. 2) injection molding is a well known non-Newtonian flow. Throw on top of it a solidification front and CFD will not like that. Again, I think simple 0D and 1D calculations are more appropriate, and probably more accurate for this problem.
# 1) you've got a very good understanding about this problem, to find out the role that turbulence plays is the most critical and difficult thing .
# 2) the non-Newtonian flow phase was set by UDF (the same model as Moldflow (a software only for injection simulation)...).."solidification front" thing maybe just the answer to the difference between my simulation and experiment.

1) about the solidification front, do you got advise in FLUENT(any alternative solution?)
2) or other SOFTWARE is ok~

 LuckyTran April 21, 2012 23:16

Quote:
 Originally Posted by Near (Post 356124) Thaks for reply , Tran. # 1) you've got a very good understanding about this problem, to find out the role that turbulence plays is the most critical and difficult thing . # 2) the non-Newtonian flow phase was set by UDF (the same model as Moldflow (a software only for injection simulation)...).."solidification front" thing maybe just the answer to the difference between my simulation and experiment. 1) about the solidification front, do you got advise in FLUENT(any alternative solution?) 2) or other SOFTWARE is ok~
Sorry i can't offer much help on this one. Not sure what you are trying to do, but Fluent will give you many nasty headaches with no real benefit. Why does the role of turbulence even matter to you? This seems like a classical injection molding problem.

I do not have any experience with VOF. With multiphase however, I've helped someone with a solidification/melting problem for a phase-change heat exchanger / heat storage device. The results were great, but that was for a stationary front.

I highly recommend sticking to pure injection molding solvers. I've worked with some that are great. You mentioned MoldFlow, the SolidWorks cad package has an injection molding software. I-DEAS also has injection molding support (I-DEAS is now Unigraphics NX). If you are trying to focus on the turbulence then you should focus on just the gas injection process and throw away everything else. I seriously doubt turbulence would have any effect since you are just pumping gas into a flexible tank. Anyway, you probably know better than I do.

 Near April 22, 2012 09:55

Quote:
 Originally Posted by LuckyTran (Post 356127) Sorry i can't offer much help on this one. Not sure what you are trying to do, but Fluent will give you many nasty headaches with no real benefit. Why does the role of turbulence even matter to you? This seems like a classical injection molding problem. I do not have any experience with VOF. With multiphase however, I've helped someone with a solidification/melting problem for a phase-change heat exchanger / heat storage device. The results were great, but that was for a stationary front. I highly recommend sticking to pure injection molding solvers. I've worked with some that are great. You mentioned MoldFlow, the SolidWorks cad package has an injection molding software. I-DEAS also has injection molding support (I-DEAS is now Unigraphics NX). If you are trying to focus on the turbulence then you should focus on just the gas injection process and throw away everything else. I seriously doubt turbulence would have any effect since you are just pumping gas into a flexible tank. Anyway, you probably know better than I do.
1) the Re of gas flow is lager than 10000, so i choose turbulence instead of laminar.is that right? what i focus on is the distribution of the plastic melt after the gas-assisted injection molding process, the distribution is largely result from pushing of the high pressure gas. again , the high pressure gas and the high speed it bring make the process no longer a classical injection molding problem...
2) thanks for ur doubting thing ("...turbulence would have any effect...... "), in fact the flow time is about 0.5 sec or less, i will think about the effect of gas from the begining.
3) Moldflow is the best pure injectoin solver, but it cannot satisfy me, and it's lack of flecibility.

may i have your e-mail, i wanna consult something about solidification/melting molding~~~

thanks

 All times are GMT -4. The time now is 00:00.