CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   2D Laminar Pipe Flow(Urgent) (http://www.cfd-online.com/Forums/cfx/22006-2d-laminar-pipe-flow-urgent.html)

tnpmember January 6, 2006 13:17

2D Laminar Pipe Flow(Urgent)
 
Hi All I m facing problem to simulate a circular pipe with laminar flow in CFX.I m taking 45 deg section of pipe and trying to model laminar flow.I am imposing inlet,outlet and two syymetry boundary conditions on circumference along with a wall on top. I dont know why solver is giving an error in symmetry plane not in strict parellel.Can anybody suggest how to overcome and is it possible to define a symmetry axis in CFX.Because i doing it in ICEM but its not visible in CFX. Please help.

regards

kirti

tnpmember January 6, 2006 13:25

Re: 2D Laminar Pipe Flow(Urgent) *NM*
 

tuks January 6, 2006 14:52

Re: 2D Laminar Pipe Flow(Urgent)
 
Hi friend, Are you really using 45 deg section? Then it should work. Sometimes you may get problem because two symmetric planes are merging(having common) onto one line(centre) and according to definition of the symmetry plane vector's can't cross it. That means for one symmetry plane vectors are parallel but for other it may cross it which is not allowed by definition of symmetry. I think the error also indicates same.. but anyway

Try with 180, 90 and less than 45 deg sections and see what happenes.. if problem persists, generate geometry accurately you may be able to solve the problem.

Regards

Jeff January 7, 2006 11:14

Re: 2D Laminar Pipe Flow(Urgent)
 
Two things to check:

1. Make sure you're doing a build topology in ICEM (under Repair Geometry) to get the edges on the ends of the pipe. If you don't, you may get elements that "round off" the sharp edges and result in faces that don't lie in the symmetry plane. This happens because there's no hard curve on which to constrain the element edges. You can see these type of defects in the mesh by zooming in and visually inspecting the sharp corners of the mesh looking for round off elements.

2. Slight bumps in the mesh surface on the symmetry plane can result in element faces which have a normal that is more than 5 degrees off normal of the symmetry plane. There's an expert parameter in CFX that allows you to set the tolerance (in degrees) for symmetry faces. Set this to something like 15 or 20 degrees and see if this fixes the problem.

Jeff

tnpmember January 7, 2006 11:42

Re: 2D Laminar Pipe Flow(Urgent)
 
Hi jeff Thanks for your help.I have some queries as i am new to all this stuffs. should i take a radius at the end of pipe?or i can model it with sharp corner. What about the boundary conditions.Should i take periodic type bc or simple symmetry would be enough. Please reply

regards

Jeff January 7, 2006 15:59

Re: 2D Laminar Pipe Flow(Urgent)
 
If it's the inlet or outlet, you probably want a sharp corner at the end of the pipe.

If there is any swirl component to the flow, you want periodic, i.e. what goes out one face comes in the other. However if flow is all in the axial direction use plain symmetry.

Jeff

tnpmember January 9, 2006 05:40

Re: 2D Laminar Pipe Flow(Urgent)
 
Hi jeff how can i define a symmetry axis in CFX.However if i am taking radius at the end it is running well but i want the velocity exactly at the centre of pipe. Should i use expert control parameter?How can i use them? kindly reply

regards


All times are GMT -4. The time now is 08:43.