Coanda Effect

Henrike01 · October 8, 2019, 04:58

Hello, everybody,
I'm about to simulate a Coanda nozzle. However, speeds of up to 1600 m/s always occur in the narrow gap, which does not agree with the corresponding literature. On the left side is the inlet with a boundary condition of 0 bar pressure, on the right the outlet with a boundary condition of 0 bar and from above the air flows in with a pressure of 6 bar.
Does anyone have any idea what I can change in the simulation that the velocity significantly decreses?
Best regards
Henrike

BlnPhoenix · October 8, 2019, 05:06

What kind of speeds are you expecting?
What kind of solver are you using (transient or steady state)?
Is the solver crashing?
What about mesh resolution in the narrow gap?

Henrike01 · October 8, 2019, 05:10

I expect speeds of about 50-70 m/s.
I am using a steady state solver SST.
No, the calculation converges.
The mesh in the narrow gap should be okay as I have 5 Inflation layers.

BlnPhoenix · October 8, 2019, 05:28

Ok. Can you use a velocity inlet instead of pressure inlet?
Have you checked that 6 bar inlet pressure is correct? It looks like your inlet pressure may be too high for your expected speeds.

Henrike01 · October 8, 2019, 05:30

Unfortuatley, I have to use a pressure inlet.
Even with 2 bar the velocity is about 500 m/s which is way too high.

BlnPhoenix · October 8, 2019, 05:33

Can you give rough estimate of mass flow and gap diameter?

Henrike01 · October 8, 2019, 05:42

The gap has a width of 40 µm and the mass flow should be around 0.25 m^3/s.

Gert-Jan · October 8, 2019, 07:16

If you look in standard literature you will find that when the pressure ratio Pin/Pout >2 (absolute pressures, for air, atmospheric conditions) the velocity will be around Mach=1=340m/s. So for a ratio of 6, your results look fine. Provided you have compressible conditions.

If you expect 50-60 m/s then your pressures are incorrect, or you are dealing with a multiphase system.

Henrike01 · October 10, 2019, 05:22

Thank you for your reply. Can you name the literature that you are referring to?

flotus1 · October 10, 2019, 05:45

Which fluid are you trying to simulate? Air? Water? Something else?
Are you using compressible or incompressible fluid in the solver?
How did you ensure that the results we are discussing here are actually from a converged solution?

Henrike01 · October 10, 2019, 05:47

Hello I am simulating air that is incompressible. I also carried out simulations were the air was modelled as compressible but also there the velocities are a lot higher than expected (800 m/s).

flotus1 · October 10, 2019, 07:57

So that answers the first part of my question. For air you absolutely need a compressible formulation for this type of flow. E.g. the compressible ideal gas model.
Now back to the ignored part of my question: are you absolutely certain that your solutions are converged? How did you check for convergence?

Edit: and what's the speed of sound in the fluid you are simulating?
maybe related: Supercritical outflow from an orifice

October 8, 2019, 07:16		#8
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,928 Rep Power: 28	If you look in standard literature you will find that when the pressure ratio Pin/Pout >2 (absolute pressures, for air, atmospheric conditions) the velocity will be around Mach=1=340m/s. So for a ratio of 6, your results look fine. Provided you have compressible conditions. If you expect 50-60 m/s then your pressures are incorrect, or you are dealing with a multiphase system. BlnPhoenix likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Setup for a 2D Coanda Effect UAV on CFD	Erfann	FLUENT	0	May 8, 2017 10:15
Coanda Effect to produce lift.	Austin	Main CFD Forum	4	May 23, 2011 08:51
Coanda effect, condensation on sphere	technophobe	Main CFD Forum	4	February 17, 2010 11:04
Coanda effect	nacaairfoil	Main CFD Forum	11	October 23, 2009 08:11
coanda effect	Craig Robbins	Main CFD Forum	5	September 9, 1998 00:21

October 8, 2019, 05:06		#2
BlnPhoenix Senior Member Join Date: Aug 2014 Location: Germany Posts: 292 Rep Power: 14	What kind of speeds are you expecting? What kind of solver are you using (transient or steady state)? Is the solver crashing? What about mesh resolution in the narrow gap?

October 8, 2019, 05:10		#3
Henrike01 New Member Henrike Join Date: Oct 2019 Posts: 6 Rep Power: 7	I expect speeds of about 50-70 m/s. I am using a steady state solver SST. No, the calculation converges. The mesh in the narrow gap should be okay as I have 5 Inflation layers.

October 8, 2019, 05:28		#4
BlnPhoenix Senior Member Join Date: Aug 2014 Location: Germany Posts: 292 Rep Power: 14	Ok. Can you use a velocity inlet instead of pressure inlet? Have you checked that 6 bar inlet pressure is correct? It looks like your inlet pressure may be too high for your expected speeds.

October 8, 2019, 05:30		#5
Henrike01 New Member Henrike Join Date: Oct 2019 Posts: 6 Rep Power: 7	Unfortuatley, I have to use a pressure inlet. Even with 2 bar the velocity is about 500 m/s which is way too high.

October 8, 2019, 05:33		#6
BlnPhoenix Senior Member Join Date: Aug 2014 Location: Germany Posts: 292 Rep Power: 14	Can you give rough estimate of mass flow and gap diameter?

October 8, 2019, 05:42		#7
Henrike01 New Member Henrike Join Date: Oct 2019 Posts: 6 Rep Power: 7	The gap has a width of 40 µm and the mass flow should be around 0.25 m^3/s.

October 10, 2019, 05:22		#9
Henrike01 New Member Henrike Join Date: Oct 2019 Posts: 6 Rep Power: 7	Thank you for your reply. Can you name the literature that you are referring to?

October 10, 2019, 05:45		#10
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,427 Rep Power: 49	Which fluid are you trying to simulate? Air? Water? Something else? Are you using compressible or incompressible fluid in the solver? How did you ensure that the results we are discussing here are actually from a converged solution?

October 10, 2019, 05:47		#11
Henrike01 New Member Henrike Join Date: Oct 2019 Posts: 6 Rep Power: 7	Hello I am simulating air that is incompressible. I also carried out simulations were the air was modelled as compressible but also there the velocities are a lot higher than expected (800 m/s).

October 10, 2019, 07:57		#12
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,427 Rep Power: 49	So that answers the first part of my question. For air you absolutely need a compressible formulation for this type of flow. E.g. the compressible ideal gas model. Now back to the ignored part of my question: are you absolutely certain that your solutions are converged? How did you check for convergence? Edit: and what's the speed of sound in the fluid you are simulating? maybe related: Supercritical outflow from an orifice