I am interested in knowing people's experience with CFX-5 for combusion applications. I have prior experience up to earlier versions (vintage 5.6) and found that most of the advanced models were not available. Has that changed and what really works now....not just pics in an advert?

Additionally, I found that the robust low iteration counts that the coupled AMG gives for typical of "constant or near constant property flows" did not show up for combustion problems. How does the current state compare to a more combustion oriented code like CFX-4, Fluent, CFDRC, etc??

Thanks,

Bak_Flow

Hi again,

so is it safe to assume that nobody out here has used CFX5 for serious combustion work? Please give me any comments?

Bak_Flow

Since 5.6 the partially premixed model (similar to Fluent's Zimont model) is in. My experience was that it was working pretty good. I used it for gas turbine combustion.

There are many combustion models in 5.7/5.7.1.

Single and Multi-step EDM reaction modelling with:

- Chemical Kinetics

- Ignition and Extinction Models

- Pre-defined reaction and materials libraries (user modifiable)

- Coupled solution of multi-step reactions for strong convergence and robustness

- PDF-type combustion models for detailed chemistry

- Flamelet model for diffusion flames

- Zimont model for premixed/partially premixed combustion

- NOx models

- Magnusson Soot model

- Combustion with single or multi-phase (Eulerian or Lagrangian) flow

Compared to CFX-4 I think all that is missing are the ignition models but there are user fortran versions of these around.

I'd be interested to know what is missing?

Convergence problems may not have to do with combustion at all so don't write it off just like that.

Neale

Hi Neale,

thanks for the input...so there has been some development there.

Does CFX have a non-adiabatic flamelet?

How does one generate flamelet libraries? ..other than RIF....I have seen that and waited for it only to be so disapointed at the fuel limitations. Has it been further developed?

So I would like some insight into the convergence issues? I have done and seen many cold flow problems that converge in 100 iterations or less and then the same problem takes thousands with combustion on. In particular the energy global ballace is usually the last to come in....often taking several thousands of iterations?

Thanks........Bak