CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Turbulence in a pipe flow (

JM December 19, 2006 08:28

Turbulence in a pipe flow

I am trying to model turbulence in a pipe flow. I am using a program which came as an add on to my CAD program. I have a 1 meter long pipe of 10 cm diameter, then a 90 degree bend and then again a 1 meter long pipe (like a right angled lines).

Since, I am experimenting with this program, I introduced a water flow with very high velocity of 25 m/s, that is inlet velocity condition, and a static pressure outlet condition. What I see in the 90 degree bend is, the velocity near the wall is zero. In my experiments, I see the highest velocity in the same position. I can imagine that when a flow of so high velocity comes in a bend after travelling through a long pipe, it will directly hit the wall and turn, something which was well seen in the experiment.

am I missing something in my simulations? I have turbulent flow switched on. I dont know what model is this program using but I saw their documentation and it appears like k epsilon (I mean, well tested model). is the result logical in your opinion?

Thanks once again


opaque December 19, 2006 09:58

Re: Turbulence in a pipe flow
Dear Jan,

Just for starters, the mass flow in your pipe seems to be

m_dot = density * mean velocity * area cross section

for water

= 1000 [kg/m^3] * 25 [m/s] * pi * (0.1[m])^2 *.25

= 196.3 [kg/s]

Are we real on this?

On the Reynolds number side,

Re = density * mean velocity * Diameter / dynamic viscosity

= 1000 [kg/m^3] * 25 [m/s] * 0.1 [m] / 855E-06 [N s/m^2]

= 2.92E+6

fairly high Reynolds number.. Do you have the "right" mesh?

Good luck,


PS.. Don't they have support staff? They can probably help you better than we do..

JM December 19, 2006 14:58

Re: Turbulence in a pipe flow
Dear Opaque,

I initially started with a low velocity (5 m/s) and then increased it to 10 and 25m/s. the reason to do so was to see why I have a zero wall velocity when it should be the other way around. but no luck even with higher flowrate. From my mesh side, i have a fairly decend mesh. The support staff are all on vacation (european christmas holidays)


Chr December 20, 2006 03:08

Re: Turbulence in a pipe flow
Hi Just to understand your problem better:

1) At the wall the velocity is always zero for viscous flow. Is your velocity zero AT the wall or NEAR (how far away) the wall? Is the velocity zero in the bend or after the bend, and which wall are you refering to (inner or outer wall)? 2) How were your experiments conducted (eg. type of measuring instrument, how far from the wall did you measure, where did you measure relative to the bend)

msureshkumar December 21, 2006 05:04

Re: Turbulence in a pipe flow
Can you tell the name of the programme? Is any wall calculator there?

All times are GMT -4. The time now is 14:04.