Nondimensionalization in FIDAP
 

Hello Fluent users,

I have a simulation involving turbulence, compressibility, and multiple species, the equations for which I'm trying to make dimensionless. The Reynold's number for the flow is about 6E5.

I'm interested in what happens to material properties such as specific heat when multiple species are involved in the flow. For flows not involving species, the dimensionless specific heat is typically set equal to unity. If there are multiple species, however, I'm wondering if the dimensionless specific heat for each species is 1) set to unity, or 2) divided by the same reference value (the carrier fluid, for instance).

For instance, if the carrier fluid is air for which Cp is about 1006 J/kgK, and other species are CO (1041) and HCl (800), the normalised values may all be 1, or they may be 1, 1.03 and 0.768 respectively).

I would greatly value some advice on which is the correct way to normalise this and other material properties. Thank you very much for your help.

Regards,

Andrew C. Ward

Department of Chemical Engineering The University of Queensland

Re: Nondimensionalization in FIDAP
 

Of course, the specific heat for each species is divided by the same reference value.

Sincerely, Jinwook

Re: Nondimensionalization in FIDAP
 

Thank you, Jinwook. This is very helpful.

I've since been looking again at the FIDAP Tutorial Manual. To normalise diffusivity when there are multiple species, it suggests setting the diffusivity for species n to 1/Pe_n, where Pe_n is the mass transport Peclet number.

In light of Jinwook's response about specific heat (normalise each species by the same thing), I wonder if what the manual says makes sense.

Any further comments would be more than welcome.

Regards,

Andrew C. Ward

Re: Nondimensionalization in FIDAP
 

Is it necessary to make the equations non-dimensional??

Its possible to simulate problems for various values of non-dimensional groups, by setting the appropriate values of the physical properties and ensuring they form the non-dimensional values you need.

It may also be that the code adds additional scaling to the equations before solving them anyway. For example in the old Fluent, densities were scaled by a reference density in some model.

You may need to be very careful of your scaling in order to get everything right. Perhaps its easier to check that different combinations of individual physical values which give the same non-dimensional groupings yield the same solution. And then just use one such set of values....

Greg