CFX overpredicting turbine massflow

sherifkadry · January 28, 2010, 12:26

I have been modeling a 2 stage axial turbine, and basing my Total Pressure inlet boundary condition and static pressure exit condition based on my experimental results. The turbulence model used is SST, and its a steady state case that's being run. I notice that at all shaft speeds the massflow is overpredicted by 5%. However, total-to-static efficiencies are within 0.5%. What could be the reason for the overprediction? Here is an image of my
model:
http://i615.photobucket.com/albums/t...Screenshot.png

Any expert opinions would be helpful thanks.

Attesz · January 31, 2010, 15:40

5% is not too much I think. It can caused by the inaccuracy of the measurement too. You can find it out by measuring an other quantity. Or doing some mesh sensitivity check can be helpful..

ghorrocks · January 31, 2010, 16:04

In general, to get the simulation accurate within that last 5% will require very careful and detailed refinement. You will need to do mesh, convergence and timestep sensitivity studies to a much tighter tolerance than you usually would. You will also have to consider the accuracy of your physics to additional precision - for instance does surface roughness contribute? Do you know your gas composition accurately? Are you using real gas models? Is SST suitable for this level of detail? You obviously have a good handle on most of these things if you are within 5% but to get even more accurate you will need to reconsider.

sherifkadry · February 1, 2010, 03:15

attesz, ghorrocks, thanks for your replies. Ghorrocks you banged on an important point here. When I ran the cfd simulations I ignored the effect of humidity, being that I ran my experiments in Texas (very humid), I think humidity plays an important role, now, I need to figure out how to incorporate humidity, which reduces the density and could explain the 5% difference. Anyone have any ideas on how best to incorporate humidity into CFX, should I run a water-vapour+air ideal gas mixture?

January 28, 2010, 12:26	CFX overpredicting turbine massflow	#1
sherifkadry Member Sherif Kadry Join Date: May 2009 Posts: 38 Rep Power: 16	I have been modeling a 2 stage axial turbine, and basing my Total Pressure inlet boundary condition and static pressure exit condition based on my experimental results. The turbulence model used is SST, and its a steady state case that's being run. I notice that at all shaft speeds the massflow is overpredicted by 5%. However, total-to-static efficiencies are within 0.5%. What could be the reason for the overprediction? Here is an image of my model: http://i615.photobucket.com/albums/t...Screenshot.png Any expert opinions would be helpful thanks.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Two stage axial turbine in CFX	sherifkadry	CFX	16	June 8, 2020 07:58
Tesla Turbine analysis in CFX	sivarama1	CFX	2	September 5, 2009 20:37
CFX application to radial inflow gas turbine	Chetan Mistry	CFX	5	June 2, 2007 13:13
re: Massflow reset in CFX 10	Wooster	CFX	5	March 16, 2007 20:08
Help CFX Steam turbine	Yogesh	CFX	2	April 26, 2006 01:38

January 31, 2010, 15:40		#2
Attesz Senior Member Attesz Join Date: Mar 2009 Location: Munich Posts: 368 Rep Power: 17	5% is not too much I think. It can caused by the inaccuracy of the measurement too. You can find it out by measuring an other quantity. Or doing some mesh sensitivity check can be helpful..

January 31, 2010, 16:04		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	In general, to get the simulation accurate within that last 5% will require very careful and detailed refinement. You will need to do mesh, convergence and timestep sensitivity studies to a much tighter tolerance than you usually would. You will also have to consider the accuracy of your physics to additional precision - for instance does surface roughness contribute? Do you know your gas composition accurately? Are you using real gas models? Is SST suitable for this level of detail? You obviously have a good handle on most of these things if you are within 5% but to get even more accurate you will need to reconsider.

February 1, 2010, 03:15		#4
sherifkadry Member Sherif Kadry Join Date: May 2009 Posts: 38 Rep Power: 16	attesz, ghorrocks, thanks for your replies. Ghorrocks you banged on an important point here. When I ran the cfd simulations I ignored the effect of humidity, being that I ran my experiments in Texas (very humid), I think humidity plays an important role, now, I need to figure out how to incorporate humidity, which reduces the density and could explain the 5% difference. Anyone have any ideas on how best to incorporate humidity into CFX, should I run a water-vapour+air ideal gas mixture?