Help Required on CFD for Turbine Blade Analysis

kkjj2008 · July 10, 2009, 11:11

Hi,
i m new here .. nd i m nt familiar with ANSYS flotran.. so can any one help me in doing a turbine blade analysis... i have the same divergent solution error each time ... the solution is not converging
i already use the MIR stability conditions...
i m not sure about the reference conditions will any body explain me how to specify them...
the temperature at the inlet is 900 K and pressure is 1.96 bar.
i m also using artificial Viscosity concept but the results are not with in range and negative temperature error occured..
kindly some one help me ..
thanks in advance for any response......

ghorrocks · July 11, 2009, 09:18

Hi,

Are you using ANSYS Flotran? That package is very old technology and you really should more to a more modern solver. Things like artificial viscosity are so 1990's.

Glenn Horrocks

kkjj2008 · July 11, 2009, 20:17

thanks for ur reply friend... i know its old but i have my project in tht so i have to do it so if anybody can help me in these i will be thankful...

ghorrocks · July 12, 2009, 00:39

Hi,

I don't think you will find many people who know Flotran here but the general concepts of CFD are applicable to all solvers. I would check things like convergence, mesh size and quality, time step size (if transient), physical accuracy of the boundary conditions and validity of the modelled physics.

Start with a coarse grid, tight convergence, double precision numerics and lots of upwinding to have a nice stable solution. Once that is running OK then start moving to more accurate but unstable schemes on finer grids, higher order discretisation. You can also check whether you need double precision numerics and determine how tight convergence is required in your case.

Also consider starting at a lower pressure ratio and increasing it. Shoudl make convergence easier.

Glenn Horrocks

kkjj2008 · July 12, 2009, 17:42

Thanks Friend...
so i m going to explain my problem..
wht i have to do is a turbine blade analysis. the pic below will show u the whole domain.. 6 number of turbine blades and a fluid channel.

the upper side of the domain is the inlet side and the downward side is the exit side...
at inlet the boundary conditions given are:
inlet velocity:
Vx = 31.5
Vy = -43.5
Vz = 0
Pressure = 1.96e5
Temp = 900 K

at exit side the boundary conditions are:
Pressure = 1.932e5

at blade walls the velocity components are zero.
and at the side walls symmetry boundary condition is used.
I will be very thankful to u if u take a look at this.

waiting for ur reply friend.
thanks

July 10, 2009, 11:11	Help Required on CFD for Turbine Blade Analysis	#1
kkjj2008 New Member Kami Join Date: Jul 2009 Posts: 6 Rep Power: 17	Hi, i m new here .. nd i m nt familiar with ANSYS flotran.. so can any one help me in doing a turbine blade analysis... i have the same divergent solution error each time ... the solution is not converging i already use the MIR stability conditions... i m not sure about the reference conditions will any body explain me how to specify them... the temperature at the inlet is 900 K and pressure is 1.96 bar. i m also using artificial Viscosity concept but the results are not with in range and negative temperature error occured.. kindly some one help me .. thanks in advance for any response......

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFD analysis of turbine blades	jayaero	FLUENT	1	June 10, 2009 01:26
CFD Analysis of pump	remith	CFX	7	October 6, 2008 08:57
ASME CFD Symposium	Chris Kleijn	Main CFD Forum	0	August 22, 2001 07:41
Where do we go from here? CFD in 2001	John C. Chien	Main CFD Forum	36	January 24, 2001 22:10
ASME CFD Symposium, Atlanta, July 2001	Chris R. Kleijn	Main CFD Forum	0	August 21, 2000 05:49

July 11, 2009, 09:18		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Hi, Are you using ANSYS Flotran? That package is very old technology and you really should more to a more modern solver. Things like artificial viscosity are so 1990's. Glenn Horrocks

July 11, 2009, 20:17		#3
kkjj2008 New Member Kami Join Date: Jul 2009 Posts: 6 Rep Power: 17	thanks for ur reply friend... i know its old but i have my project in tht so i have to do it so if anybody can help me in these i will be thankful...

July 12, 2009, 00:39		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Hi, I don't think you will find many people who know Flotran here but the general concepts of CFD are applicable to all solvers. I would check things like convergence, mesh size and quality, time step size (if transient), physical accuracy of the boundary conditions and validity of the modelled physics. Start with a coarse grid, tight convergence, double precision numerics and lots of upwinding to have a nice stable solution. Once that is running OK then start moving to more accurate but unstable schemes on finer grids, higher order discretisation. You can also check whether you need double precision numerics and determine how tight convergence is required in your case. Also consider starting at a lower pressure ratio and increasing it. Shoudl make convergence easier. Glenn Horrocks

July 12, 2009, 17:42		#5
kkjj2008 New Member Kami Join Date: Jul 2009 Posts: 6 Rep Power: 17	Thanks Friend... so i m going to explain my problem.. wht i have to do is a turbine blade analysis. the pic below will show u the whole domain.. 6 number of turbine blades and a fluid channel. the upper side of the domain is the inlet side and the downward side is the exit side... at inlet the boundary conditions given are: inlet velocity: Vx = 31.5 Vy = -43.5 Vz = 0 Pressure = 1.96e5 Temp = 900 K at exit side the boundary conditions are: Pressure = 1.932e5 at blade walls the velocity components are zero. and at the side walls symmetry boundary condition is used. I will be very thankful to u if u take a look at this. waiting for ur reply friend. thanks