[feizaghaee]

i'm modeling free surface flow in a rectangular channel by CFX 12. CFX calculates secondary flows inverse. this false calculation result in false velocity field (my last thread). dose anybody know waht's the reason of this wrong result? i attached a pic.

[feizaghaee]

please help

[Tajnesaie]

Salam
I dont have this idea! Beacause in my simulation, it is correct. But the number of it is incorrect(Instead of 3 in half of section,it is only 2). Perhaps your bundary condition is incorrect.

[feizaghaee]

what kind of boundry condition are you using?

[Tajnesaie]

I use static pressure in inlet stead Normal speed.

[Tajnesaie]

Originally Posted by feizaghaee

i'm modeling free surface flow in a rectangular channel by CFX 12. CFX calculates secondary flows inverse. this false calculation result in false velocity field (my last thread). dose anybody know waht's the reason of this wrong result? i attached a pic.

Salam
You must see Tominaga's Paper. I think yours is incorret, Both of direction and location of them,base on this paper I tell it.

[feizaghaee]

i found that CFX calculates secondary currents like close channels (ducts). i dont know why!!!!!!!!!!!!

[ghorrocks]

? What are you saying? If the Navier Stokes equations say secondary flows exist then if CFX is running well it will predict secondary flows. So what are you asking?

[feizaghaee]

directions of secondary currents in ducts (close channel) and open channel are different. this different is related to free surfce effect. the direction of ducts' secondary currents are in contrast with open channel and CFX dosen't calculate the free surface effect.

[zandi]

Originally Posted by feizaghaee

i found that CFX calculates secondary currents like close channels (ducts). i dont know why!!!!!!!!!!!!

salam
if we can not get a right answer to our problem that doesn't mean cfx is in a wrong way! (if it happens we can challenge with ANSYS cfx team) of course there is something with our modeling (or ghoflshekastegi barname bekhatere aberu finglish neveshtam!!!).
check all things with a sample that exists (catched from someone who had good answers) or change the conditions and data and test them in a simple run to catch right answer then use conditions.
good luck

[feizaghaee]

hi miss.
doctor bonakdary said that he has worked with a version of CFX before version 10. and it wasn't able to calculate the channel and roughness of the channel. he reported that problem to Ansys and they sent him version 10 beta to be examined by him. no one has obtained such answer i need. why do you think these software always give right answers

[zandi]

ok
you know miss Tajnesaie. she has the similar project, if i'm right. did you test the static pressure in the inlet instead of normal velocity?
if you insist on wrong answering of cfx why don't you share it with ansys, too?
anyway I’m looking forward and will be glad to hear from your solution

[feizaghaee]

yes i know her. she used static pressure at inlet and outlet. what about flow's velocity. in that way we can't control velocity of inflow or outflow. i investigating an experimental research and i need to put a certain velocity at inlet or outlet. and about reporting to ANSYS, it's reason is exactly ghofl shekastegi.
regards

[alex-bishop]

hi,can you tell me how to get secondary flow vector filed in cfx?

[Gert-Jan]

- Create a plane perpendicular to main flow.
- Plot vectors on this plane
- Set projection to tangential
- Change symbol size (and colour) of you vectors to get something feizaghaee generated.

[alex-bishop]

ok,i will try it. thanks a lot