# Regarding divergence

 Register Blogs Members List Search Today's Posts Mark Forums Read

 March 1, 2015, 11:36 Regarding divergence #1 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Hi i m doing a CFD analysis of thermostat valve . My inlet boundary condition is velocity inlet c=1.25m/s . Outlet is pressure outlet .I am using K-epsilon model and simple algorithm.No energy equation. If i activate laminar zone in cell zone conditions then my problem is getting solved. But i have turbulence in my flow. So i gave turbulence and hydraulic diameter in my BC and i deactivated laminar zone and then i get divergence in epsilon. Can anyone help me out ???

 March 1, 2015, 12:44 #2 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Dear Barathwaj, Turbulent kinetic energy k is basically determined by your turbulent intensity set. And the dissipation epsilon is determined by both length scale and turbulent kinetic energy. Since you have had K converged, it might be the problem of length scale: ie, hydraulic diameter. There could be many reasons for not converging. kai.

 March 1, 2015, 12:53 thanks #3 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 I ve given my pipe diameter as my hydraulic diameter. And I also calculated my turbulent intensity as 5% . So is there any mistake in this value

 March 1, 2015, 12:59 #4 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 these seem to be right, and how about your mesh quality?

 March 1, 2015, 13:02 #5 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Geometry of the valve is complicated .skewness is very high. I couldnt reduce it.I tried both with fine and medium mesh and also I increased smoothness. Yet I had this error.

 March 1, 2015, 13:03 #6 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Or maybe you can decrease the turbulent intensity if you don't have that large Reynolds number. In my opinion, the Intensity here simply provide an initial guess for your equations to be iterated/solved, it has no effect on your result normally.

 March 1, 2015, 13:05 #7 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Decrease intensity, and if it doesn't work, try to decrease you relaxation factors, although i don't believe this will help you too much.

 March 1, 2015, 13:06 #8 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 My Reynolds number is around 180000

 March 1, 2015, 13:07 #9 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 WOW! then your intensity is not high...

 March 1, 2015, 13:08 #10 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 What should I do??

 March 1, 2015, 13:11 #11 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Try to decease your relaxation factor, and btw, how could you have that high Re with a speed of only 1.25m/s initially? Sorry i have no idea of thermostat valve.

 March 1, 2015, 13:13 #12 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 My pipe od is just 0.053 m that's the problem fluent is water dynamic viscosity is also very low . that's why I get such a large Reynolds number

 March 1, 2015, 13:14 #13 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Regarding the relaxation factor should I ve given it as 0.8 . should I reduce it to o.5

 March 1, 2015, 13:17 #14 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Re=density*velocity*D/viscosity=998.2*1.25*0.053/0.01=6625？

 March 1, 2015, 13:18 #15 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 I think the default is 0.8? and yes try 0.5.

 March 1, 2015, 13:21 #16 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Viscosity of water is 0.000338 (kg/(m-s))

 March 1, 2015, 13:23 #17 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Sorry, i am wrong with that , but viscosity is 0.001 from what i see, giving turbulent intensity 66250, are you talking about water vapor maybe?

 March 1, 2015, 13:25 #18 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Not really. I need viscosity of water at 84 deg c for my problem

 March 1, 2015, 13:28 #19 Member   kai Join Date: Jul 2013 Posts: 54 Rep Power: 13 Well, then i can't help more, just try the relaxation.

 March 1, 2015, 13:29 #20 New Member   barath Join Date: Mar 2015 Posts: 18 Rep Power: 11 Okay thanks a lot. I ll tell you the results. Can I get your mail id

 Tags divergence detected, fluent 15, k-e model, turbulence and laminar