2D Venturimeter Analysis Problems

henriktheking · November 30, 2015, 17:06

Hi guys,

I have a problem to model the flow of water through a ventuimeter using ANSYS WORKBENCH 14; FLUENT, however I am a bit of a beginner and am having some problems setting up my solutions.

I had set up the analysis as 2D, and modelled the top half of the geometry (the line of symmetry is set up as an axis in the set-up) and created a bias in the mesh towards the venturi throat.

In my set up I use the k-epsilon model as the flow is turbulent and have enhanced wall treatment enabled. The fluid is water and is defined as such, and enters the pipe at a velocity of 4.77 m/s as defined in ANSYS.

The inlet is set as a velocity inlet, centreline as an axis, wall as a wall and outlet as a pressure outlet.

However when I run my calculation the solution does not converge and I am continually faced with being told that the flow is reversed?

Any help at all on how to successfully setup my problem and obtain solutions would be kindly received.

Cheers

Cocofeliz · November 30, 2015, 18:45

Hi!

Did you elongate the outlet to avoid backflow?

henriktheking · November 30, 2015, 19:30

Quote:

Originally Posted by Cocofeliz

Hi!

Did you elongate the outlet to avoid backflow?

I just read an older post after I started this thread. Back in tomorrow afternoon so will try that and update. Thanks!

Cocofeliz · December 1, 2015, 01:48

Perfect! Good luck and keep us updated!

henriktheking · December 1, 2015, 16:00

Quote:

Originally Posted by Cocofeliz

Perfect! Good luck and keep us updated!

That sorted it I think, thanks for the help! Is there a reason why I was having that problem with a shorter length?

Cocofeliz · December 1, 2015, 16:49

Yep. The flow probably detaches at the venturi's wall. This creates a backflow that difficults the convergence.

Since this is probably your case, you may want to check the k omega sst model, instead of the k epsilon ewt. In models with detached flow like venturis, it may underpredict the pressure losses. But i'm new also in this, so dont trust me much

November 30, 2015, 17:06	2D Venturimeter Analysis Problems	#1
henriktheking New Member Sean Join Date: Nov 2015 Posts: 3 Rep Power: 10	Hi guys, I have a problem to model the flow of water through a ventuimeter using ANSYS WORKBENCH 14; FLUENT, however I am a bit of a beginner and am having some problems setting up my solutions. I had set up the analysis as 2D, and modelled the top half of the geometry (the line of symmetry is set up as an axis in the set-up) and created a bias in the mesh towards the venturi throat. In my set up I use the k-epsilon model as the flow is turbulent and have enhanced wall treatment enabled. The fluid is water and is defined as such, and enters the pipe at a velocity of 4.77 m/s as defined in ANSYS. The inlet is set as a velocity inlet, centreline as an axis, wall as a wall and outlet as a pressure outlet. However when I run my calculation the solution does not converge and I am continually faced with being told that the flow is reversed? Any help at all on how to successfully setup my problem and obtain solutions would be kindly received. Cheers

November 30, 2015, 18:45		#2
Cocofeliz New Member Join Date: Nov 2015 Posts: 4 Rep Power: 10	Hi! Did you elongate the outlet to avoid backflow? henriktheking likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problems Running Seakeeping Analysis in Headwave	JeffStrath	STAR-CCM+	0	April 28, 2015 15:37
Acoutic fan analysis	arunraj	FLUENT	0	October 4, 2012 06:54
Eigenfrequencies static and modal analysis	Laura_mecheng	ANSYS	1	May 15, 2012 04:40
How to recover unfinished CFX analysis with backup file	DiddyPablo	CFX	3	December 18, 2011 07:14
(urgent)parallel analysis problems	Karteek	Siemens	1	July 18, 2008 20:04

December 1, 2015, 01:48		#4
Cocofeliz New Member Join Date: Nov 2015 Posts: 4 Rep Power: 10	Perfect! Good luck and keep us updated!

December 1, 2015, 16:49		#6
Cocofeliz New Member Join Date: Nov 2015 Posts: 4 Rep Power: 10	Yep. The flow probably detaches at the venturi's wall. This creates a backflow that difficults the convergence. Since this is probably your case, you may want to check the k omega sst model, instead of the k epsilon ewt. In models with detached flow like venturis, it may underpredict the pressure losses. But i'm new also in this, so dont trust me much