DES / URANS / SAS sampling for convergence study

RodriguezFatz · December 12, 2013, 08:19

Dear all,

I need some advice.
I want to simulate a pipe with a bluff-body in it (vortex flow-meter). Now, before comparing any result with measurements, I want to be sure my simulation is numerically robust / converged. For that, I would like to vary things such as the inlet length, outlet length, time-step, convergence criterion at each time-step, grid size, turbulence model (see title) ...
For each of them I am going to try if the solution changes significantly when I increase / decrease it.
The question is: What parameters should I watch? Since the simulation is highly unsteady, I can not just use some simple pictures. Also time averages will propably not be very usefull, because of the large vorticies. Maybe time averages of u'' / v'' / w'' ? Or an FFT the force on the bluff body?

Philipp.

flotus1 · December 12, 2013, 10:05

The RMS value of the force on the bluff body (especially in cross-streamwise direction) would be another good indicator for convergence, at least for the URANS simulations.
The frequency of the vortex shedding is another good indicator since this is the variable a vortex flow meter actually measures.

Will you present some of the results here?

RodriguezFatz · December 16, 2013, 04:46

Alright, I will try that. Yes, it will take some time, but when I am done, I will show results here.

What is the best way to get the shedding frequency? Using an FFT seems to be not good, because I will need thousands of periods to get an accurate result. Do I just need to do this "by hand"?

flotus1 · December 16, 2013, 09:14

For the URANS methods you can determine the shedding frequency by counting waves in the plot, yet I dont think it will be possible with DES and SAS.
But you dont really need that many periods to get a result. I doubt that the differences will be that small.

December 12, 2013, 08:19	DES / URANS / SAS sampling for convergence study	#1
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 26	Dear all, I need some advice. I want to simulate a pipe with a bluff-body in it (vortex flow-meter). Now, before comparing any result with measurements, I want to be sure my simulation is numerically robust / converged. For that, I would like to vary things such as the inlet length, outlet length, time-step, convergence criterion at each time-step, grid size, turbulence model (see title) ... For each of them I am going to try if the solution changes significantly when I increase / decrease it. The question is: What parameters should I watch? Since the simulation is highly unsteady, I can not just use some simple pictures. Also time averages will propably not be very usefull, because of the large vorticies. Maybe time averages of u'' / v'' / w'' ? Or an FFT the force on the bluff body? Philipp. __________________ The skeleton ran out of shampoo in the shower.

December 16, 2013, 04:46		#3
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 26	Alright, I will try that. Yes, it will take some time, but when I am done, I will show results here. What is the best way to get the shedding frequency? Using an FFT seems to be not good, because I will need thousands of periods to get an accurate result. Do I just need to do this "by hand"? __________________ The skeleton ran out of shampoo in the shower.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
sampling interval and sampling frequency	beauty	FLUENT	13	November 9, 2022 08:23
How to use DES well?	Daniel	Main CFD Forum	0	October 26, 2008 05:59
DES ( SA+LES) vs DES ( kwSST+LES) for bluff body	Phil	Main CFD Forum	0	January 8, 2008 09:04

December 12, 2013, 10:05		#2
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	The RMS value of the force on the bluff body (especially in cross-streamwise direction) would be another good indicator for convergence, at least for the URANS simulations. The frequency of the vortex shedding is another good indicator since this is the variable a vortex flow meter actually measures. Will you present some of the results here?

December 16, 2013, 09:14		#4
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	For the URANS methods you can determine the shedding frequency by counting waves in the plot, yet I dont think it will be possible with DES and SAS. But you dont really need that many periods to get a result. I doubt that the differences will be that small.