|
Fluent - StarCD
Hi all,
I have to decide wether we get a Fluent or a StarCD licence. We would like to simulate 2 phase flows with VOF and, eventually, moving meshes. I have quite a big experience with Fluent, but I never used Star CD. Does anyone know advantages or disadvantages of one program over the other in these fields? Thank you very much for your help! |
Re: Fluent - StarCD
Hi there CD adaptec has launched a new version of Star-CD, it is called Star-ccm and a limited node version is available for free download from their web site, so enjoy your time with real world experience www.cd-adaptec.com or check their flashing banner on this new CFD-OnLine web site
|
Re: Fluent - StarCD
Sorry I did not spell it correctly, it is CD-adapco
|
Re: Fluent - StarCD
i downloaded the new version of starCCM and was comparing with fluent. out of the cases i did there was one 2d bluff body kind of flow, that is of some interest.
in that case, for simple laminar calculations with same conditions on both the solvers. i observed that when i use second order convection scheme for flow varibles. the results are same. but when i use first order scheme, the results were heavily different. fluents results were more close to second order scheme. where as starCCM failed to predict the buble in first order scheme. so accuracy wise they are similar if u use second order schemes , but with first order i would go with fluent. second, i had a nonorthognal mesh, that i used to check my code that i was writing. on that mesh,, starCCM fails to get much convergence, that is continuity residuals hangs around 2-3, where as fluent was able to converge that to limits 10^-2, so here also i would go with Fluent. further, i took one difficult case with unstructured mesh, and plotted the velocitu vectors with iterations. i observed that when with starCCM, during iteration process the vector at some places may shoots up, that is suddenly it increases to very high values , but solver can fix this in coming iterations. what it means is , starCCM can diverge in difficult situations. So for the same mesh, by fixing AMG cycle to V type and max iterations to 3, i was able to make starCCM to diverge (residuals greater than 10^4) on that particular case. however fluents iteration process does not show these kinds of shoots up, and no matter what i set in AMG i could not make it diverge. it even worked with out AMG. (that is 0 levels ). on calculation speeds periterations, fluent is just a bit faster than starCCM. in all i would go with fluent even though i liked starCCM. p.s. : i have read their documentations, and starCCM decretisation is more accurate, specially the way they calculate the gradients. (fluent only mentions that they use phi_face = 0.5 * (phi_0 + phi_1), but i doubt since with this approaximation solution can not be so stable, and that is why in starCCM after this gradient, they calculate another gradient (using this one) which is more accurate and stable) |
Re: Fluent - StarCD
Hi,
I would chose Fluent or CFX. The learning curve when using Star is low compared to Fluent or CFX. Also VOF simulation if there are a bit advanced for far more quicker on Fluent using the new solver. Regards Jens |
Re: Fluent - StarCD
These observations do not really help when the task is to do a VOF with moving mesh. It might be the best to use COMET, which is also sold from the CD-Adapco group.
|
Re: Fluent - StarCD
When you compare two programmes that are solving the same Navier Stokes equations using the same Finite volume technique, it is a difficult task indeed. In my modest opinion, we all would find some kind of difference between the two knights. Among the advantages of Star-ccm that I terribly like is when you write a UDF, a window pops up with the original function and all what you need to do is to write your own new statements, compare that to FLUENT, another feature I like with Star-ccm is the tapped windows, the model tree and the ease of accessing all the components of your model etc. The ability of Star-ccm to read the native Cad file format (ProE and etc.) is important in the design process of products. In addition to all that you can download a trial version that you can test by yourself and arrive at your own conclusions. Let us be more objective, wait and see what chusmann will decide !
|
Re: Fluent - StarCD
i think all i wanted to say is that fluent is much more stable with respect to starCCM+, and this aspect is very important in VOF. But in the end it boils down to what you like (because underlying procedure for all the solvers is very much same).
|
Re: Fluent - StarCD
STAR-CCM+ is the future code from CD-ADapco group which is still an unfinished product. The main product from CD-Adapco group is STAR-CD. Therefore maybe your conclusion is not quite relevant.
|
Re: Fluent - StarCD
If you can down load an evaluation copy NOW then it is the present not the future, of course they will introduce improvements and modifications in the future but presently they have the product, so again, this is the Present (Now) and not the future (Tomorrow)
|
Re: Fluent - StarCD
|
Re: Fluent - StarCD
Angelina What all that stuff has got to do with a Comparison between two fluid mechanics programmes. CFD means Computational Fluid Dynamics not Certificates of Deposits which is a banking term that is also abbreviated as CFD. You are in the wrong place
|
| All times are GMT -4. The time now is 14:53. |