Hi, before i start, i have posted here before and found your answers/advice so thankyou very much.

now, another (probably simple) question from me...iv recently imported a solidworks assembly into gambit using an ACIS (*.x_t) format and found that is works fine.however, when i try and mesh the model, the actual time it takes is rather long. i was wondering, would it be better to 'join' or 'add' all the different volumes of the assembly together so that it is all just one. (one volume)?

also, if i dont want the model to take too long to mesh, what would be the best way to limit this? i was thinking of limiting the number of elements used? (and how do you do this by the way.)

thankyou so much for the help. very much appreciated.

corey.

I assume you import an assembly of SOLID parts. If yes, you are meshing solid parts. If your want to solve NS-equations, you aren't on the right way. You shoud mesh, the fluid domain (--> boolean operation to cut out solid parts)

yes it is solid parts in gambit, however, it is several different volumes together, as per the assembly in solidworks. would it be best to COMBINE them together in gambit to form just one volume??

and then i would be subtracting this volume from the domain volume being used? thus being able to mesh the fluid domain now?

thankyou for the heads up.

corey.

I think the best way, is to create first of all the fluid domain. You can extract all your solid parts, by cuting out each one from a ground volume (brick, cylindre) with boolean operation. Your result will be one fluid volume. Then you can decompose this volume with several logical and simple volume for getting a quality mesh (-> hex, etc...)

thankyou for your response. now, i have done the first few steps, however, now when i try and subtract the solid parts from the fluid domain created, when i eventually subtract the last solid part, more volumes seem to appear from nowhere. would this be due to errors in the imported geometry, and if so, how can i check and fix this. Otherwise, do you recommend any ways of getting around this volume so that i can start meshing.

also, when i tried to setup a sizing function etc and begin meshing even with these arbitrary volumes, i eventually got a gambit error message saying something like ' Gambit cannot continue any further, error etc etc. and so i guess its not possible to do this..?

anyways, thanks in advance for any help. corey

don't work too fast. If you aren't sure that your fluid domain is extracted, don't begin with meshing. From cut out operation (boolean, it is posible that some volumes appear. If you want, I can extract your fluid domain, and I will send you back, the .dbs and also the journal file .jou, wou will able to see step by step how I handled the stuff. mAx

thanks for your help..although i think i have finally figured what was wrong. two of my solid parts were hollow which im pretty sure were the cause of the arbitrary volumes being created in gambit when i tried to subtract from the fluid domain. do you know if this is actually the case? I fixed up the assembly in solidworks so that there is now no gaps or hollow parts in the assembly and it seems to be working ok.

im now in the process of meshing the new domain now that i have fixed this problem and ill see how i go.

thanks for your help once again. i will let you know how i go but im sure ill come into more problems shortly and will require assistance again..

regards corey.

hi, as predicted i have come across another problem.haha.hope you can help...as i mentioned in my previous post, i was able to subtract the solid part from the fluid domain such that only the domain was left. from here,i setup a sizing function beginning at the solid part and ending at the (3d) domain walls. then from here, i was able to mesh the fluid domain, very coarsely first as a test to make sure everything was working properly.

However, when i tried to export the model out of gambit and into fluent, i ran into some problems.i am exporting it as a .msh format and i get an error msg in gambit something like - 1).volume domain is not meshed and is not exported & 2).continuum entity domain fluid does not contain any valid entity and is not written.

...and thus obviously getting error messages when i tried to read the .msh file in fluent.

do you have any suggestions as to what is wrong with my mesh?? i have tried to be as descriptive as possible but feel free to ask any other questions about it.

also, would it be possible for me to send you my .dbs and .jou files so you can have a look at it and make any suggestions??

thanks once again.your help is much appreciated. corey

I assume you only have surface mesh and no volume mesh. Yes you can send me .dbs file I will have a look at it. if the file is over 5MB, please let it on a share-server I can handle it tomorrow, then I have a 4 days weekend

sorry, what do you mean by 'let it on a share-server'...like as in upload onto a file share server online? which one do you suggest?

files are about 17mb too.

thanks

something like this: http://rapidshare.com/

thanks. so iv uploaded all the necessary files and here are the links:

.msh file that I tried to export: http://rapidshare.com/files/100901473/CDG1.msh.html

wing & domain with no mesh .dbs file: http://rapidshare.com/files/10090182...ING_DOMAIN.dbs

wing and domain with no mesh .jou file: http://rapidshare.com/files/10090196...ING_DOMAIN.jou

wing and domain with no mesh .trn file: http://rapidshare.com/files/10090203...ING_DOMAIN.trn

wing and domain with attempted mesh .dbs file: http://rapidshare.com/files/10090236...DOMAINMESH.dbs

wing and domain with attempted mesh .jou file: http://rapidshare.com/files/10090250...DOMAINMESH.jou

wing and domain with attempted mesh .trn file: http://rapidshare.com/files/10090256...DOMAINMESH.trn

i hope its not too confusing.thank you so much for having a look at this. iv tried numerous times to try and mesh it properly but for some reason it doesnt seem to work properly. The mesh stays a bright yellow in gambit, so ye, im pretty sure now that it is only meshing the surfaces, but i am not sure why.

thanku once again. corey.

ok, your problem is the following. You have to many short edges (especially on wings) so go and merge faces on the wings --> il will automatically merge short edges. Also I have seen, a small problem, which can be easy fixed under your cad software: the leading edge form one wing is to close from the extremity form the vertical support (see around point (-90.33 / 279.43 / 439.5)) I recommen you to use the Geometry Clean Up Tools in Gambit, it will automatically detect you short edges etc... Have fun with your aero stuff
)

ahh i see, didnt know that the short edges would be too much of a problem, thanks. i will try and merge a few of the aerofoil section faces manually and see how i go. maybe having each aerofoil with only 3-4 faces? also, im not too sure i understand what you mean by: 'the leading edge form one wing is to close from the extremity form the vertical support' could you please elaborate? thanks.

also, just a quick question, how exactly does the geometry clean-up tool in gambit work? i tried using this previously however it looked like it wasnt doing anything.

once again, thankyou so much for your help and general direction.

corey

leading edge is rounded curve of your profile (first to be hitted from airflow) I don t use this Geometry Clean Up Tool, but it returns automatically all the short edges. Once you know where they are, the you can fix your problem with Face Merge

Dear Sir,

Many times when I give the meshing command, GAMBIT shows out of memory. How to solve this problem?

I am working on PC having 1 GB RAM and 160 GB hard disc.Please let me know what are the maximum no. of nodes/cells which GAMBIT can handle. Is it affected by no. of geometries available in particular problem.

Shall I increase the RAM capacity?

Also let me know, whether is it necessary to use TURBO commands while modelling Turbomachinery say Hydro turbine?

With regards

Sanjay jain

When its talking about out of memory its referring to the RAM. increasing it might solve some of your problems.

hi mAx, sorry to keep hassling you, just keep having problems... i have finally tried to merge the faces together however it doesnt allow me to do this, with the faces remaining the same. With this, i mean that the faces do not change and the edge lengths in particular also do not change.

What i attempted was to lower all the geometry to faces, and then manually mesh these wing faces from here.

Would you recommend any other way? I was thinking possible lowering the geometry again to the individual edges and then merging/combining them from here...

Also, would i need to delete some of the edges? And wouldn't this lead to the aerofoil wing section being really inaccurately shaped (boxy)

Thankyou so much for your help.

corey

in adding on to my previous post, i have also attempted lowering all the geometry to edges, then deleting a few and combining them so that only large edges were present at the sides of the aerofoil sections. however, i found that gambit does not allow you to create faces from these 'unreal' (curved) edges....is there a way around this, so that i can increase the shorter edges??

thanks once again. corey

[sivarama1]

Hi all,
i was given quard elements,pave type,
my geometry rectangular type(2000x2000 mm) intervel count 1,
but in that condition some error is comming
errorroblam is perminent face overlap
whats meaning,whats solution any body help in that point of view.
Thankyou for all.