Diffuser simulation in Fluent

alicesun · December 4, 2016, 19:15

Hi all,

I'm trying to simulate an annular diffuser (pre+dump diffuser, see pic) of a combustor. The simulation is in 2D axisymmetric. I have used several turbulence models including k-epsilon, k-omega and RSM. It seems to me that the predicted total pressure loss at the pre-diffuser outlet (i.e. from 1-2 in the pic) is significantly lower than the experimental value (about 50% lower). The y+ of the mesh is lower than 2.5 which I think should be okay and finer mesh didn't make any difference. I also used the inlet velocity profile obtained from the experiment as the inlet condition.
What do you think could possibly be the reason? I 've also noticed that by increasing the inlet turbulent intensity, the total pressure loss could increase. If that's the case, what intensity is suitable with a Re=2*10^5 and inlet height=38.1mm. I initially set it to 4% but the predicted total pressure loss was too low.

Thanks for your suggestions.

Alice

Mohy · December 5, 2016, 09:59

Hi alicesun

You can check this link, https://goo.gl/rc8TY4 , as it says:
" High-turbulence case: High-speed flow inside complex geometries like heat-exchangers and flow inside rotating machinery (turbines and compressors). Typically the turbulence intensity is between 5% and 20% "

Also by checking both the Medium-turbulence &Low-turbulence cases, I think that annular diffuser of a combustor is considered as a High-turbulence case with turbulence intensity is between 5% and 20%.

For a sensible change in the value of pressure loss, take the turbulence intensity by 10%, after that compare it with experimental value.

alicesun · December 8, 2016, 05:26

Hi Mohy,

Thanks a lot for the suggestion. As you said, I did increase the turbulence intensity to 10% and the total pressure loss increased too. However, in the experiment it is stated that the flow is a fully developed flow from an intake with a very long settling length, which is not the case of turbomachinery where flow comes from the OGV exit with some distortion. In such circumstances, do you think I=10% is still reasonable? As the equation for fully developed flow in your link says, I obtain an intensity around 4% for a Reynolds number of 2*10^5.

Kind regards,

Alice

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
looking for a smart interface matlab fluent	chary	FLUENT	24	June 18, 2021 09:07
Running UDF with Supercomputer	roi247	FLUENT	4	October 15, 2015 13:41
The fluent stopped and errors with "Emergency: received SIGHUP signal"	yuyuxuan	FLUENT	0	December 3, 2013 22:56
Simulation on fluent 14	NAD	FLUENT	0	September 11, 2013 02:05
LSR (Land speed record) car simulation on FLUENT	Maxime31850	FLUENT	2	May 1, 2013 11:15

December 5, 2016, 09:59		#2
Mohy New Member Join Date: Oct 2015 Posts: 5 Rep Power: 10	Hi alicesun You can check this link, https://goo.gl/rc8TY4 , as it says: " High-turbulence case: High-speed flow inside complex geometries like heat-exchangers and flow inside rotating machinery (turbines and compressors). Typically the turbulence intensity is between 5% and 20% " Also by checking both the Medium-turbulence &Low-turbulence cases, I think that annular diffuser of a combustor is considered as a High-turbulence case with turbulence intensity is between 5% and 20%. For a sensible change in the value of pressure loss, take the turbulence intensity by 10%, after that compare it with experimental value.

December 8, 2016, 05:26		#3
alicesun New Member Join Date: Dec 2016 Posts: 2 Rep Power: 0	Hi Mohy, Thanks a lot for the suggestion. As you said, I did increase the turbulence intensity to 10% and the total pressure loss increased too. However, in the experiment it is stated that the flow is a fully developed flow from an intake with a very long settling length, which is not the case of turbomachinery where flow comes from the OGV exit with some distortion. In such circumstances, do you think I=10% is still reasonable? As the equation for fully developed flow in your link says, I obtain an intensity around 4% for a Reynolds number of 2*10^5. Kind regards, Alice