In this case, the 1st thing..I want to see the air flow simulation inside the empty tunnel with the effect of the rotating fan blade.This is only my first run (empty tunnel). I've made the whole body of the wind tunnel including the fan blade also(refer to the pic).The body of the tunnel has been made using design modeler CFX 10.0

Fan Blade: I made it according to the NACA spec w 16 blades in CATIA V5 and import it inside the tunnel through cutting operation.

Then, I meshed part by part (diffuser, 2 corners, contraction & test section, tunnel with blade) in .gtm.but not a very good one.still have warning to refine it.

Then I imported all the GTM file into CFX-Pre and made two domains: rotating domain (tunnel with fan blade inside) and stationary domain for the rest.I did all the domain interfaces for each assembly part.

In CFX-Pre; Rotating Domainomain Motion-rotating (-2000rpm) according to the axis. I used isothermal 25C, SST, Fully tubulent for the transitional behavior.50 iterations, high resolution.

I solved already and ran normally and got the ok result.I checked in Post processing (in linked pic).At first I assume to get 80m/s in the test section area.but i only got 42m/s.Can anybody comment the way I did...is it the correct way to see the wind tunnel simulation?? or maybe smthing wrong with the fan blade??or smtnhg else.I need comments to learn more.i can share any result if u want more.Thanx

http://rapidshare.com/files/56493467/1st_trial_1.htm

sorrry for the availability of the linked pic.Here the new one....

http://rapidshare.com/files/56503716/cfx_Pre.JPG

http://rapidshare.com/files/56503717/cfx-Post.JPG

Hi,

The information you have given is too general to identify the error. Also the links did not work.

I recommend you read the best practises guide which come with CFX, in the reference manual. The guide for turbomachinery and the guide for numerical accuracy will help I am sure.

Regards, Glenn Horrocks

Hi Glenn, the new second link picture attached..I think you can access.I need anybody to see what is happening and give any comment on it.for your information i didn't use turbomachinery yet.It's just the simple way I did.thanx guys..

Hi Fadly,

You say you expected 80 [m/s]. Can you explain the basis of this assumption?

-CycLone

Exactly, I follow one example of 1 low speed wind tunnel.and its specifiation in the test section can achieve 80m/s.and the tunnel that I made it is approximate same dimension but a litle bit smaller but not too much.It's a complete wind tunnel with vanes, honeycomb and all the guiding parts.So I asssume the speed,it will be more than that right.but the RPM of that tunnel fan is not known (only give the power 430KW).i just give 2000 rpm for fan which the dimension of the blade and hub is roughly sketched according to the fan spec given.is it too high for the rpm??but if it is to0 high, why in the test setion the loss of speed is too much from rotating fan in empty tunnel??

Fadly,

The speed will be a function of the fan geometry and speed. You can't just arbitrarily pick a speed and expect to automagically come up with the velocity.

If you don't know what your fan is doing, just take it out of the simulation and specify a flow rate. If you need your fan, try to understand what you are doing with it before creating your simulation.

-CycLone

ok..thanx cyclone. I try to modify and try some more options after this.But, is meshing problem affects much in the results??How can I aciheve a uniform flow??what criterion sould be taken??

one more question that worrying in my mind...how can I setup velocity or flowrate in a closed body of wind tunnel because there will be 2d regions in one boundary in domain that have already been used.or if the fan is taken out...how can I setup the fan boundary condition in a closed tunnel??