huh? bro, why? sorry a stupid question :confused:
|
because you don't need those volumes.
They are isolated volumes without any fluid-connection to your computationnal domain. |
oh sorry heading for my CNY holiday, get to work. now i know, i deleted it, can i start meshing?
|
? are u there bro?
|
Quote:
|
ok, i meshed it.. Unfortunately, it caused me around 30 minutes, and this time, all the result i save for u to analyse. i get some of the meaning, but some not. Guide me thanks
http://i.imgur.com/W9EDd.png |
Examine face face.3086, you will see that skewed cells are linked with topology "issue" like sharge angle etc...
Use merge tools for fixing this problem. Then same thing with face face.2557, etc.... But as I already told you, I would FIRST try to mesh your domain without all your fans, which will give your lots of "clean-up" work |
Quote:
|
when you started with import: delete fan's volumes before substracting
|
hmm, ok forget about deleting the fan volume first, i will try with examine face, for examine face, what shud i do? where is the examine face button?
|
direct above your windows clock
|
That is examine mesh, fine. now i press it. Where to edit so that i can examine face 3086? This is the place where only i can see the quality of the mesh. http://i.imgur.com/AMLnY.png
|
go to range
select all kind of elements (hexa/pyramid/wedge/tetra) set lower limit to 0.9 set upper limit to 1 and display... all the displayed element should be fixed (especially elements between 0.95 and 1) Then check where are the cells, they should be focused at one topology "issue" like sharp edge etc... Then exit to the examine panel, et display the surfaces where Gambit failed to mes the surfaces. You should recongnize the topology "issues"... |
i presume i done it correctly? but why when i press update, it will reset back lower and upper limit 0 and 1 with no element in there? NOTE THAT i already change the lower and upper limit as u say :(
http://i.imgur.com/nZjtT.png |
your volume mesh failed, that's why you cannot see any volume elements (because you don't have any)
do the same but with 2d elements |
wow, can i did this while i off mesh in display attribute? cause it is annoying freaking long just to see it. and i cant rotate the screen(left click turn) as all hang started again:(
|
with 2 elements, it shouldn't be slow.
With 10miilions volume elements, then yes, you can pray that your session won't crash. |
ya, finnaly at the end, i am NOT able to see it. i will consider removing the fan volume again then if that is the only idea
|
hey, i am back, remeber last time i say i want to remove the fan, i did it. and here is the summary.
I import the STP file and delete the fan partition. Then unite whole volume, subtract it to get fluid domain. Then i mesh using "Mesh Volume" with spacing of 1. Fortunately, 1 equisized skew issue only comes out. But, mesh fail and the transcript state that due to face 43, asking me to mesh face 43 using the "Mesh face" function(i guess i correctly digested the meaning?) Then i do face meshing at face 43, here is the result :( i am running out of idea now. What should i do? http://i.imgur.com/dPeKK.jpg |
you are trying to mesh your face (complex one) with quad.
Forget trying to mesh your volume with hexa. Switch with tetra (tri for surface mesh) |
hey, i already try it, and i able to mesh the face, but while meshing volume, i get this
http://i.imgur.com/z9vu1.jpg |
i manage to mesh the face, it is success. But it is different for the volume, i get this error
http://i.imgur.com/z9vu1.jpg |
you are trying to mesh a complex volume in one shot (you'll never succeed in that way)
you need to split your volume into subvolumes. And you gave an element size which is very small. So split your volume and use size fonction for controling your mesh size |
err, how bro? only 1 volume there
|
yes you only have one volume, that's why I adviced you to split it into several small volumes.
>>>Volume/Split |
so now what the step i have to do?
1) i will spilt the volume. (By the way, how to spilt it? just simply spilt it or spilt it as small and as simple as possible?) 2)mesh those spilted volume 1 by 1. 3)merged them back. is this the way? |
1- split your volume where you think you need to refine your mesh, and where you think you can apply an hexa mesh (simple geometries like cube, cylinder etc...)
2- yes 3- no, don't merge back your volumes (gambit will glue them together) |
meaning i dont have to do step 3? gambit automatically joint aka "glue" them back?
|
no you don't need step 3
|
and the size function is necessary>?
|
and another question, after i spilt the volume, let say not on the exactly on the boundary, eg like the PSU, in my case it is not purely cube, so i spilt a cube slightly larger then the PSU, is it ok?
|
This is my first volume spilted and mesh, is it ok?
http://i.imgur.com/Fw0F5.jpg i am using spacing of 2 in order to save time, is it necessary to change to 1? |
you are free to split your domain as you want.
Regarding the size fonction, it is not necessary to use them, but it is very helpful for controling the mesh refinement |
Where this size function be generally used?
|
grid refinement >> check help
|
I have an idea here, if i can redo all this again, mean before i unite all this volume and subtract to get the fluid domain, can i copy all this volume inividually once? like copy each of PSU, HDD and so on. then after i done all the procedure and want to spilt volume, i just move back the copied item to thier respective locaton, so that i can use spilt volume with volume easy, can i>?
|
you want to split your domain with HDD volume, etc...?
|
This is what i done, at the figure, volume 313 is the volume i subtract (the fluid domain). And u can see that volume 314 is the volume which exactly the same (i guess) for the fan heat sink, this (volume 314) volume is i copy the original volume and place at the same location on it before i perform the spilt. However, spilting is success, as shown in figure, but i still fail to mesh volume 314
http://i.imgur.com/jjtc3.jpg By the way, in the volume spilt option there is biderectional, connected, and retain, what that 3 option for? |
No the volumes are not the same. Only outer surfaces are the same.
From your flow domain you already substracted this fan. I can see it, because when you pick the whole flow domain, outer surfaces from this fan are highlighted, meaning that the surfaces are outer surfaces from fow domain. Now splitting the fan volume from the flow domain should fail, since there is no fluid domain inside the fan (hollow). End even if it worked, I don't understand what you wanted to do... Once again I don't advice you to import fans. If you really want to include fans, then create a cylinder which diameter is greater than fan's one, and split your domain with it. The cylinder will surround your fan and you will be able to pick it for meshing (cylinder substracted with fan). Then try to mesh it. |
Oh, ok bro.. I will try that. Meanwhile, for your knowledge, beside the geometry which guided by u, i have do 1 alternative on my own, and it is the simplified version. And, i managed to mesh as shown in the figure below. However, when i import the case file to FLUENT, test.msh in this case, i got the following error. Why?
Building... grid, Note: Separating wall zone 3 into zones 3 and 1. wall -> wall (3) and wall:001 (1) Warning: ST_Malloc: out of memory.malloc_storage: unable to malloc cell1, size 51779932, t 15, myid 1000000. Error: FLUENT received fatal signal (ACCESS_VIOLATION) 1. Note exact events leading to error. 2. Save case/data under new name. 3. Exit program and restart to continue. 4. Report error to your distributor. Error Object: () http://i.imgur.com/mrmqc.png |
All times are GMT -4. The time now is 06:12. |