|
[Sponsors] |
October 23, 2015, 04:43 |
|
#21 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
Hi,
I would like first to ask you a theoretical question. In my process normally is room temperature.. So I donīt have temperature changes. Based on that... I think I cannot use ideal gas law because as Iīm working with low pressure automatically it will decrease too much the temperature and then I will face error... For example.. in the inlet I found 30K temperature... this is unreal for my process... So I need to fix a density... maybe calculate the density of Nitrogen in 1Pa and room temperature... And then run simulation as incompressible. What do you think? |
|
October 23, 2015, 04:48 |
|
#22 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
How do you know that the temperature decreases that much? I though you don't have any converged solution...
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 04:50 |
|
#23 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
This value is what appears when I set Initial solution from the inlet... Also the velocity is generated in this way (760 m/s) supersonic...
|
|
October 23, 2015, 05:05 |
|
#24 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
This is just some initial guess - where the numerical solver starts to iterate. This is not the solution. It will change once you have a stable solution.
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 05:59 |
|
#25 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
But can I change it for more realistic values (or maybe what would I expect to see?)
Because if it starts with such low temeprature, i would not have a flow.. because it would be freezed... In another hand.. with such high velocity (17000 m/s) itīs not possible in practice... |
|
October 23, 2015, 06:00 |
|
#26 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
Yes, you can change it by hand in fluent, just type another number under "initialization"
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 06:50 |
|
#27 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
I changed to a more simple problem..
Just the pipe (closed in both sides) with 14 linear holes on the top surface and one inlet in the middle. I set the same parameters: Laminar model, ideal gas Operating Pressure= 1Pa Mass flow inlet = 3,81293E-06 kg/s Pressure outlet = 30Pa (each one of the holes I called outlet. Ex: Outlet1, Outlet2 ... Outlet14. As you can see in the attach the residuals are not decreasing so much.. It will not converge...and the velocity is still high 180 m/s and pressure inlet 95Pa.... |
|
October 23, 2015, 06:55 |
|
#28 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
1) First of all, it at least converged to some more or less stable state, which is good.
2) What is the reason of setting both "operating pressure" and the "outlet pressure"? I did not study the manual for compressible cases, but I guess you will get the same as if you set the operating pressure to 31 Pa and outlet pressure to zero, right? 3) You manually set all outlets to the same pressure. Maybe that is not physically meaningfull. Some of them are closer to the inlet as others. This could be one reason why the residuals don't drop too much. 4) If inlet pressure is 95 Pa, then this might be the result. You "press" some amount of gas through a small pipe. Of course you will get some kind of pressure drop.
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 07:00 |
|
#29 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
Sorry.. I inverted.. Actually I set the Operating pressure as 30Pa and outlet pressure as 0Pa.
Regardin your 3) comment... I identified each outlet because than itīs easier to receive from the software the mas flow rate through each hole. The main purpose of this simulations is to evaluate the mass flow through outlets. I would like it it to be the more homogeneous as possible. But what you said make sense. What would you suggest? |
|
October 23, 2015, 07:01 |
|
#30 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
Run the full simulation, not just the pipe, but with the plasma chamber.
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 07:03 |
|
#31 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
But then, how would I measure the mass flow through each hole? Because later I will change the pipe section area.. the hole diameter.. and finally realize what is the best configuration for my case..
|
|
October 23, 2015, 07:09 |
|
#32 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
You can do that in fluent, you need to define the areas inside the volume. But, I just thought they are all in the same chamber... maybe that will be sufficient for the assumption that the pressure is the same.
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 07:13 |
|
#33 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
The problem is that you messed up this thread .
You give just too little information from post to post to get good help. Each answer post causes a new post by you with a little new information what you are actually trying to do. It's hard to help. In one of your first posts you show a picture of what seems to be your setup. Now you write that the chamber isn't part of it... You understand? (Don't take it personal or too serious.)
__________________
The skeleton ran out of shampoo in the shower. |
|
October 23, 2015, 07:14 |
|
#34 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
Yes you are right... But the mass flow depends on the section area and also in the hole diameter. Just because they are all inside the chamber doesnīt means in practice that they have the some pressure drop... For example.. if you have a very small section area the pressure drop will be much higher than a higher section area... Thatīs why I need to try different configurations. And for that, I need to measure the mass flow through holes...
|
|
October 23, 2015, 07:16 |
|
#35 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
the pipe with holes is inside the chamber...
|
|
October 30, 2015, 03:58 |
|
#36 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
Hi,
Iīm trying to rescue this thread. So... I attached some images. Hope it can be clear. 1st image: Error "divergence temperature" with the conditions described in the information below. 2nd image: Residuals plot 3rd Image: Model / Mesh Information: Laminar model, ideal gas Operating Pressure= 1Pa Mass flow inlet = 1,14388E-05 kg/s Pressure outlet = 0Pa Hybrid Initialization Gas- Nitrogen Objective: Run some simulations with different Mass flow and different pipe geometries to evaluate the homogeneity of the gas feed through the holes. |
|
October 30, 2015, 04:04 |
|
#37 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
Two things:
1) On your third picture, the small pipe in the upper right corner: It looks like you set the cell size of the inlet face to some small value, but the cell size of the adjacent pipe wall to some large value. If this is the case, why? This doesn't look right. 2) If the temperature makes problems, this could be due to bad initialization. You could try to start your case with different temperature values (use the "patch" button to set some other value), or you could start with a smaller flux, than increase the inlet flux step by step.
__________________
The skeleton ran out of shampoo in the shower. |
|
October 30, 2015, 04:19 |
|
#38 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
Thank you for your comment:
1)Regarding the meshing I used automatic. Just set face sizing to the face inlet and face outlet. Should I specify something different? Would you have any suggestion? 2)I never used Patch. I can search in this forum or in Ansys Manual. Maybe I can start with a small flux and then increase slowly. What would you suggest? Increase 20% after each 200 iterations? Is is a good initial plan? Additional question: 3) If my simulation shows no errors but the Residuals plot were not flat. After each iterations changes a little bit but in the same little range. (Ex: 1,076e-3, 1,056e-3, 1,065e-3, 1,080e-3, 1,077eī6... etc) but the velocity inlet/pressure inlet and velocity outlet/pressure outlet were converged to a constant value. May I rely on this results? |
|
October 30, 2015, 04:42 |
|
#39 |
Senior Member
Philipp
Join Date: Jun 2011
Location: Germany
Posts: 1,297
Rep Power: 27 |
1) As I understand it, you set the wrong values. If you set the face inlet / outlet to small sizes, then (as you can see from the picture) only this small face has small cells, but everything else is larger. In your case the cells on the small pipe are too large to accuratly represent a pipe. You need to set the cell size on the wall of the pipe, not on the inlet. You can easily see your mesh in the pipe by creating a cut through the pipe in "Ansys Meshing".
2) After your hybrig initialization click on "Solution Initialization->Patch" then choose a variable, choose a "zone to patch" and type in a certain value. Then click "patch" and the old value (e.g. of temperature) is overwritten by the new one. Choose a new value that is close to what you expect. If you run the small flux, just try what converges and if it does increase the inlet a bit. What you suggest sounds good. 3) You mean the residuals are constant but large or what?
__________________
The skeleton ran out of shampoo in the shower. |
|
October 30, 2015, 04:45 |
|
#40 |
Senior Member
Gustavo
Join Date: Jun 2015
Posts: 121
Rep Power: 11 |
1) and 2) I understood . Thank you so much. I will try it.
3) Continuity residuals in the order of 1e-2 (But if a see the report flux the imbalance is very low, e.g 1e-12). The energy normally is small than 1e-6. But velocities (x,y,z) in the order of 1e-4. Whatīs your opinion? |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Issues on the simulation of high-speed compressible flow within turbomachinery | dowlee | OpenFOAM Running, Solving & CFD | 11 | August 6, 2021 06:40 |
mass flow inlet and pressure outlet with target mass flow rate | Zigainer | FLUENT | 13 | October 26, 2018 05:58 |
Convergence problem with target mass flow rate | ADL | FLUENT | 2 | May 29, 2012 21:11 |
mass flow | Wenbin Song | FLUENT | 0 | September 27, 2005 13:00 |
Mass Flow Inlet | Pravir Kumar Rai | FLUENT | 0 | February 19, 2003 14:03 |