I am a phd student and new one on cfd programms and I am intersted in thermal analysis of buildings with double skin facades. This means I have to study the air flow in the cavity between the 2 skins(the external skin is usually of glass and the internal skin is the common brick wall and concrete) and the thermal distribution in the building. Which of the programms, CFX and FLUENT, is the best for thermal analysis of buildings and the air flow simulation? Any help will be useful... Thanks a lot to everybody!

Fluent I have seen as the best, in general. But as your problem is much simpler CFX should also do a good job.

Hi, I would say, CFX, because of the intergration between a CAD, Pre, Solver and Post.

The integration of coupled solver and efficient models

The Post processing in Fluent sucks. You will need to buy additional software to do the job.

And do not go for Airpack, you will end up doing the simulation in Fluent.

My few cents.

Regards

Jens

post processing in fluent only sucks when you really do not know how to do it in fluent. And its not your fault, if you do not know it, because fluent manuals hardly make any effort to explain how to fully use their post processing features. Fluent post processing is very good, if we know how to use it.

I agree with zxaar on that point. The surface creation tools and the "scene" menu are the most useful options in post processing, but have almost no valuable information in the manuals. So you end up having to play around with the options to see what they do, but there is some pretty impressive stuff there.

Jason

i have been noticing that whenever we have a post comparing fluent and cfx, always someone says that we should chose cfx since fluent post processign is shady. And sometimes this reasoning affect the decision to buy the software, i wonder why don't fluent people work a little bit and improve the documentation regarding post processing, if it affects the business. I mean if they wish they certainly have manpower to do so. I wonder if fluent people even bother to come to this discussion forum and listen to what people are saying.

Of course fluent know that they need to work on their GUI and their documentation. Has anyone heard of Flowizard? It's a wizard driven cfd tool which is very user friendly but doesn't have all of the complex features of Fluent. It's inevitable that they must be working towards a similar version of Fluent.

I doubt their crappy documentation loses them too much business as they give pretty good training to big customers.

Jimbob

i think that their crappy documentation infact gives them business, because if we understand the things properly we can do it ourselves, if we do not we ask them. So then you have money specially courses conducted by fluent to educate people for the things those could be easily put in docs and then lot of consultancy work. For fluent fuzzily written docs are better, but for users it horrible thing.

Unfortunately Zxaar you're right...sigh!Luca

One thing you have to remember: FLUENT has the largest market share - and hence has the most time and money to invest in improving their code. When a competing code comes out with new features i.e. polyhedral mesing, or transition modelling, you can be almost assured that Fluent will have the features of ALL their competitors in their next version.

...assuming that Fluent invests their money back into development...

By your logic, Fluent should already have all the capabilities of their competitors, which is simply not true. You have to account for the level of code debt that they have and how flexible their architecture is. Not to mention as to whether they have the expertise to pull it off.

Just look at the mediocre improvements between Fluent 6.1 and 6.2, which took 2+ years to develop and you have the answer. Bigger does not mean better.

Josh

actually what you said is correct, bigger does not mean better. But fluent in many ways better than others and others some ways better than fluent. For example i found fluent to be very stable solver compared to others, i only compared with starCCM+, but then starCCM+ can have some model which fluent may not have, or starCCM+ may have some model better implemented than fluent. About the improvement between 6.1 and 6.2, i feel that as the software becomes too large improving it becomes too difficult, i think this is what happening with fluent. ultimately, its difficult to compare two softwares.