|
[Sponsors] |
April 2, 2015, 08:06 |
centrifugal compressor outlet overflow
|
#1 |
New Member
陈思成
Join Date: Oct 2014
Posts: 7
Rep Power: 11 |
hello everyone,
Someone had used by CFX simulation transonic centrifugal compressor? I am conducting a centrifugal compressor simulation,just have diffuser and impeller. And I have got stuck for a long time. when I calculated ,the CFX-solver is all normal except notice : A wall has been placed at portion(s) of an OUTLET | | boundary condition (at 29.5% of the faces, 5.6% of the area) | | to prevent fluid from flowing into the domain. and it notice all the time until convergence. these are the details: Impeller: 9 main blade and 9 splitter blade Diffuser: 16 blade 111700 rev/min steady state air ideal gas ref. pressure 0 atm heat transfer : total energy turbulence : SST Inlet: Stat.Frame Tot.Press : 100kpa Stat.Frame Tot.Temp : 298k Outlet: Average static Pressure : 150kpa Solver: Specified Blend Factor : 1 RMS:1e-5 The mesh y+<1 When this situation convergenced and increase the outlet pressure ,this notice also has all the time. And when the outlet pressure is increased more high just like 500kpa ,has notice too .It must influence the massflow because inlet and outlet massflow is not change at all the situation. I do not know where is not true. I want to used the two methods: (1) Move the outlet to a region well away from this influence by, perhaps, extending the flow domain further downstream. (2) Use the opening boundary. And I want to konw: (1) Extending the flow domain is not influneced the flow?(I used it and the pressure ratio and Total To Total Device Efficiency (Isentropic Compression (%)) is too low) And how to obtain the pressure ratio and efficiency on the ture outlet position(Previous position) (2) If I change boundary(outlet to opening),how to choose the opening temperature and calculate the efficiency I just want to draw the compressor performance characteristic diagram. Can someone help me ? And I would deeply appreciate for this problem . Thx the CFD-online sorry for my bed english. |
|
April 2, 2015, 09:08 |
|
#2 |
Senior Member
Thomas MADELEINE
Join Date: Oct 2014
Posts: 126
Rep Power: 11 |
the wall placed at the outlet means that you have a recirculation at this position.
I suggest you check ANSYS help to see about that (just search outlet on CFX Help) they have good explanation of what happens. The best solution is to move your outlet (a simple extrusion will be enough I think). If you are afraid that it will modify the efficiency of your pump you can set free slip walls adiabatic (not recommended). you can always create a plan in CFD-Post to measure the real efficiency. |
|
April 3, 2015, 03:44 |
|
#3 | |
New Member
陈思成
Join Date: Oct 2014
Posts: 7
Rep Power: 11 |
Quote:
But I have been seen the ansys help on CFX,and it said " A good rule of thumb is to place the boundary a distance downstream that is at least 10 times the height of the last obstacle." I think it is too long. And I had used the simple extrusion just like 5 times the height of the ;ast obstacle and the notice problems persist. |
||
April 3, 2015, 03:46 |
|
#4 |
New Member
陈思成
Join Date: Oct 2014
Posts: 7
Rep Power: 11 |
No one encountered this kind of situation ?
I am really worried about it |
|
April 3, 2015, 04:49 |
|
#5 |
Senior Member
Thomas MADELEINE
Join Date: Oct 2014
Posts: 126
Rep Power: 11 |
I think almost every one have one day encountered this problem. It happens regularly if you have complex geometry...
You have to expend your outlet, it is the best you can do. The proof that the help is right you tried with 5 time and it wasn't enough. You should try to expend it a little more. |
|
April 3, 2015, 05:47 |
|
#6 | |
New Member
陈思成
Join Date: Oct 2014
Posts: 7
Rep Power: 11 |
Quote:
I will to expend the outlet to try again. But this distance will not affect the result ? And I will get data by intercepting plane. I do not know this plan is practical but I will try it. Thanks again |
||
April 3, 2015, 09:22 |
|
#7 |
Senior Member
Thomas MADELEINE
Join Date: Oct 2014
Posts: 126
Rep Power: 11 |
It will affect the results... but by a really thin effect (I don't think it matters here).
if you want to avoid that you have two solutions: 1 get the experimental data of your outlet on the new outlet position (it is the best since that the recirculation problem applies to the CFD and the experimentation). 2 the new walls created set to free slip and adiabatic or create a big box at the outlet to negate any effect on the wall. |
|
April 4, 2015, 02:12 |
|
#8 | |
New Member
陈思成
Join Date: Oct 2014
Posts: 7
Rep Power: 11 |
Quote:
Do you konw if I use the opening boundary replace the outlet, the results will be what effect ? |
||
April 7, 2015, 06:56 |
|
#9 |
Senior Member
Thomas MADELEINE
Join Date: Oct 2014
Posts: 126
Rep Power: 11 |
then it will not add a wall at the boundary but let the flow re-enter inside your domain with the specified characteristics (temperature and pressure)
|
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
centrifugal compressor-diffuser simulation : Overflow | insisivus | CFX | 7 | April 22, 2014 07:41 |
centrifugal pump | Chalghoum | CFX | 18 | April 16, 2014 07:37 |
Autoblade centrifugal pump template | mahdi balali | Fidelity CFD | 3 | February 18, 2014 09:56 |
Centrifugal Pump Cavitation problem or not. | ismael.s | CFX | 13 | February 27, 2012 09:00 |
How to model flow of centrifugal fan? | Peter | Main CFD Forum | 0 | April 2, 2008 07:07 |