NLH simulation - convergence problem

reicht4 · October 23, 2018, 03:39

Hello all,

I have 2 different turbine setups to perform NLH simulations. 1 turbine setup consists of 1 stator and 1 vane only. Since one of them is totally under 0.85 Mach, the other is going to sonic region as high as 1.05M.

They are meshed in autogrid with fine grid detail. There are no problems in steady state simulations. I've completed first turbine's NLH simulation also. I applied requiered mesh detail to rows to obtain 3 harmonics.

But the sonic one is diverging whatever i do so far, about luckly 100 iterations after passing harmonic simulations.

I observed that efficiency is effecting by unsteady effects dramatically and mass balance is diverging instantly. I suspect that numerical solution couldnt handle the strong sonic effects. Do you have any idea to solve this problem?

I tried different things sulch as under-relax but couldnt find any solution so far...

Many thanks,
Fatih

Hamidzoka · October 24, 2018, 05:18

Hi;
Do you initialize the harmonic run with your previously well-converged steady run? This is quite helpful in getting convergence.

reicht4 · October 24, 2018, 06:05

Hello,

Yes of course. ı started from very well converged steady-state case.
And also, i tried decreasing pressure ratio by increasing outlet pressure. So lowered the sonic efects. It worked a bit but not suitable for my case still.

October 23, 2018, 03:39	NLH simulation - convergence problem	#1
reicht4 New Member ybyo Join Date: Feb 2010 Posts: 3 Rep Power: 16	Hello all, I have 2 different turbine setups to perform NLH simulations. 1 turbine setup consists of 1 stator and 1 vane only. Since one of them is totally under 0.85 Mach, the other is going to sonic region as high as 1.05M. They are meshed in autogrid with fine grid detail. There are no problems in steady state simulations. I've completed first turbine's NLH simulation also. I applied requiered mesh detail to rows to obtain 3 harmonics. But the sonic one is diverging whatever i do so far, about luckly 100 iterations after passing harmonic simulations. I observed that efficiency is effecting by unsteady effects dramatically and mass balance is diverging instantly. I suspect that numerical solution couldnt handle the strong sonic effects. Do you have any idea to solve this problem? I tried different things sulch as under-relax but couldnt find any solution so far... Many thanks, Fatih

October 24, 2018, 06:05		#3
reicht4 New Member ybyo Join Date: Feb 2010 Posts: 3 Rep Power: 16	Hello, Yes of course. ı started from very well converged steady-state case. And also, i tried decreasing pressure ratio by increasing outlet pressure. So lowered the sonic efects. It worked a bit but not suitable for my case still. Last edited by reicht4; October 24, 2018 at 07:35.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Convergence and backflow problem in cavitation simulation	burakaltintas	CFX	6	August 11, 2016 11:00
Convergence and backflow problem in cavitation simulation	burakaltintas	CONVERGE	2	August 9, 2016 08:42
Problem with an old Simulation	FrankW	CFX	3	February 8, 2016 05:28
Cyclone Simulation Convergence Problem	Sal	FLUENT	8	December 17, 2014 08:46
Heat Transfer simulation: No convergence problem	fiqs	CFX	2	April 21, 2010 16:47

October 24, 2018, 05:18		#2
Hamidzoka Senior Member Hamid Zoka Join Date: Nov 2009 Posts: 293 Rep Power: 19	Hi; Do you initialize the harmonic run with your previously well-converged steady run? This is quite helpful in getting convergence.