[MBttR]

Hi,

The Solidworks dealer passed by our office for a demo on Solidworks Flow Simulation (FloeEFD). As someone who is currently learning OpenFOAM the ease of use and aesthetics of this sure appeal to me, but I am stuck with the thought of this being a limited CFD package, a claim that I mainly got thanks to some users on the internet. I confronted the demo person with this claim and asked him bluntly where this comes from and he gave me two reasons;

- Lack of multiphase flow simulations; this is something that to me is not interesting (unless conjugate heat transfer counts as multiphase but I doubt it does)
- Only one turbulence model, being an "adjusted" k-Omega model: Does anyone know what in Solidworks' case is meant by "adjusted"? I am still inexperienced in choosing turbulence models but I feel like I see k-Omega being used the most. What would I be missing out on, what would be impossible for me? Initially I'd mainly be involved with rotor design and motor cooling (separately).

Right now I think Solidworks Flow isn't worth it for me. After a good week of doing OpenFOAM tutorials and using sHM on my own geometry I'm starting to see some decent snaps, whereas my first and only attempt at automeshing in Solidworks failed. Initially I'd be occupied with two cases that won't vary that much over the course of their design, so I think once I have set them up successfully the difficulty of OpenFOAM will be gone. I also think the cost is Solidworks Flow is quite high (14,000 euro + 3,500 euro anually for updates and support). I don't want to overestimate myself but I feel like the main difficulties with OF lie in meshing and the initial setup of your case, something I think I'll be comfortable with in a matter of weeks. Validation will be equally difficult on both systems.

So basically I made up my mind regarding staying with OpenFOAM, but I was hoping someone could clarify the k-Omega situation for me or give me other arguments to support my case

Cheers!
Bruno

[arjun]

What is that "claim" that you are talking about. Either you forgot to mention or its not clear from your post.


PS: For the price mentioned, what is offered in my opinion is too less. I already do much more than that in FVUS and still cant think of charging this much.

[MBttR]

Hey arjun,

Sorry, claim might not be the best word, but I'm talking about people "claiming" (there it is again) packages such as SolidWorks and Autodesk being lesser packages because they are limited. A 1 minute Google search already points be to several heated discussions in which people make that "claim":

http://www.cfd-online.com/Forums/mai...imulation.html
https://www.reddit.com/r/AskEngineer...o_cfd_at_home/
https://www.reddit.com/r/engineering...ow_simulation/

According to that instructor they are only limited in only feature single phase flow and one turbulence model, and now here I am asking how severe this limitation is (since k-Omega seems to be used a lot) and if anyone knows what the "adjusted" k-Omega model entails in Solidworks Flow Simulation.

Good to hear you also think it's expensive for what it is. I'm not familiar with FVUS, but Google tells me it's something you're developing yourself, so that makes sense

[arjun]

Based on features one could say that a certain package has less features. This in my opinion is a valid point.
Being lesser is however debatable (again in my opinion). Imagine that one has to do just turbulence calculations, then in situation rest of the features are useless to user and their presence or absence makes no real problem for user.
(actually the presence of more features make difference to user, because more feature means more a company has to spend its resources and then that certain users features don’t get that much priority.).
Second being lesser in terms of solver robustness, stability, accuracy could be another way of looking at it. In that situation it is very likely that solid works flow solver is lesser as fluent solver is more robust. (I did not check so correct me if i am wrong).
If someone says that a “lesser” solver can not beat commercial code that he may also be wrong because a specifically tailored solver could very well beat a general purpose code. Quite possible.

Originally Posted by MBttR

Good to hear you also think it's expensive for what it is. I'm not familiar with FVUS, but Google tells me it's something you're developing yourself, so that makes sense

Yes, FVUS wildkatze is work in progress. Still it has more features than what you mentioned, for example as of today, the solver supports vof based multiphase, flow, energy, turbulence (Spalart Almaras, SST KW , LES models), also gravity etc small things.
FVUS already supports proper multiregion support (solved together in one matrix, openfoam struggles here), already has immersed boundary support.
I don’t promote much because still lot to do, and until all multiphase models come in (full eulerian eulerian and population balance), i don’t think it is anywhere near where I want it to be). I cant think of charging 13000 euros or so.

[FMDenaro]

I think you should focus first on the categories of flow problems you want to simulate. Then, decide your budget and consider the possible commercial CFD codes that better are suited for you. Honestly, this code with only one RANS modelling is too expensive.
If you are confortable using OF I think this is defintely your best choice.