CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   table fan aerodynamic analysis using gambit and fluent (http://www.cfd-online.com/Forums/fluent/84274-table-fan-aerodynamic-analysis-using-gambit-fluent.html)

ziq January 25, 2011 06:50

table fan aerodynamic analysis using gambit and fluent
 
hi all,

i am new here. i was assigned to do an aerodynamic analysis of a table fan weeks ago. i was given the solid model of the fan which can be imported to gambit to do the meshing. and i know this kind of fan analysis will involve sliding mesh or MRF.

can anybody give me some ideas on how to perform this kind of analysis? i have no clue about which step should i start first and what type of setting/scheme is more suitable to this kind of analysis. can anyone help me out? thanks in advance

mettler January 25, 2011 09:24

Fluent has all types of great tutorials. I know they have sliding mesh ones. I would start there.

ziq January 25, 2011 09:42

i've tried a couple of related tutorials.
i tried to mesh the fan, but loads of errors, like problematic geometry, output entities with badly checking geometry, etc... popped out. sometimes, meshing process aborted due to some errors...

mettler January 25, 2011 09:50

if this is your first time trying Fluent you can NOT start with a sliding mesh example. You need to start with the first tutorial and do the proceeding ones in order to learn some of the basics. Then, you can skip to a sliding mesh example. Fluent is tough, but once you understand the software it's not too bad.

ziq January 25, 2011 10:02

okay... i just do not really know where to start the whole thing. i've tried all different the setting, but all popped out errors... can you give me a brief idea on the procedures to perform this kind of analysis? should i mesh the face of the fan? or should i just create boundaries and mesh the volume of them without meshing the face of the fan?

mettler January 25, 2011 10:16

without me actually seeing the fan etc I can tell you how to do it. It also depends on what you are after, but I almost always start with the edges. Go back to the some of the earlier tutorials as they deal with meshing.

ziq January 25, 2011 10:19

can i have ur email address, so that i can send you the solid model and have a clear picture on what i am asking?

ziq February 16, 2011 06:57

hi,

i've finished meshing the fan blade and its boundary. but, there is a problem when i try and run the simulation with Fluent. Error: 1000000: Failed to allocate 38.39MB memory popped out. why is it so? can anyone help?

ziq February 19, 2011 11:35

hi,
i am facing another problem while running a simulation of my mesh on fluent.

the fluent windows popped out:
Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc Mass Flux, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a0, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a1, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a1, size 14236608, t 1, myid 1000000.
why is it so? is anybody out there can help?

sail February 19, 2011 16:33

Quote:

Originally Posted by ziq (Post 296058)
hi,
i am facing another problem while running a simulation of my mesh on fluent.

the fluent windows popped out:
Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc Mass Flux, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a0, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a1, size 23992976, t 8, myid 1000000.

Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc f_a1, size 14236608, t 1, myid 1000000.
why is it so? is anybody out there can help?


it looks like you've run out of ram. how many ram do you have? how big is the mesh? when exactly this error appears? (loading of the mesh? initialization? halfay during calculations?)

ziq February 19, 2011 21:43

thanks for replyiong.
i am currently using 4gb, but only 3.5gb usable as i am running on 32 bit software. the mesh is not as big as it should be, because i already scaled it down for the test run. but, the whole mesh still contains few millions of elements. this error appears during initialization.
i'm curious if i've meshed the fan model correctly. have you ever mesh this kind of model before? may i know which mesh types are best suit for model like this? i am trying 2 combination. first attempt is using tetrahedral for both small volume(sliding mesh) and big volume(flow region), while for the second one, i am using tetrahedral for small volume(sliding mesh) and hex(copper) for the big volume(flow region).

Far February 20, 2011 06:25

mesh size of 0.4-0.6 million should be sufficient for these flows. As i guess that the flow is incompressible

ziq February 20, 2011 06:30

thanks for replying...
ok... another question. i am actually using the big volume split with the small one(rotating volume), then subtract the blade from the small volume. is it correct to do it this way?
may i know what is the best way to mesh the rotating volume that used to enclose the blade?

Far February 20, 2011 06:33

gambit is compulsory to be used in your simulations?

ziq February 20, 2011 06:36

not compulsory... but i only know how to use this meshing software. i have no idea what other software can do that. do you have any recommendation?

Far February 20, 2011 06:39

although gambit has very powerful geometry tools, but also requires time to generate the mesh. on the other hand CFX turbogrid has automatic meshing procedure.

If gambit is the option then I shall guide you step by step

Far February 20, 2011 06:41

definitely there is no stationary part downstream of fan (rotating) so there is no need to use the sliding mesh model. MRF is the recommended choice in this situation

ziq February 20, 2011 06:44

yeah. i am using MRF... my problem is i am not sure whether i meshed the model correctly as everytime i import it in fluent. error popped out. can you guide me on the steps to generate the mesh for this kind of simulation? thanks

Far February 20, 2011 06:52

1. import iges file to gambit

2. put the surface mesh on blade with sufficient point to represent the geomtry accurately.

3. write this mesh file as neutral mesh file

4. import this file into the excel and you will see the coordinates in X,Y,Z format. Keep these and delete all other text thingy

5. copy these coordinate and past them in text file (creat with notepad) and save this file as .dat .

6. Import this file into the gambit. you will see the vertices on the blade (which are at the mesh cross lines)

7. you need the join the point through the spline

I shall tell you other steps and if necessary past the pics in the next post.In the nean time do these steps.

Best Regards
Far

ziq February 20, 2011 06:55

ok... thanks.. i will try this out... do you have any other better way to contact like msn or skype? in which i can send message to you instantly if i have any problem.


All times are GMT -4. The time now is 01:43.