[Gerry Kan]

Howdy Foamers:

From the OpenFOAM on-line documentation, the basic units are [ kg, m, s, K, mol, A, cd ].

However, according to page 27 of this document, chemical reactions are supposed to be in the units of [ m, s, kmol ].

So the question is, what unit should I be using for the Arrhenius preexponential factor, [ cu m / kmol / s ] or [ cu m / mol / s ]?

Thanks in advance, Gerry.

[clapointe]

I've looked into this before -- a long time ago so my memory is a bit fuzzy but I might be able to point to some helpful files. I think at one point I wanted to figure out why a mech in openFoam format (after running chemkinToFoam), was different than the chemkin format. Looking into related files (e.g., https://github.com/OpenFOAM/OpenFOAM...hemkinReader.C) we see a comment about converting from mol/cm^3 to kmol/m^3. I might have stopped there after seeing that this conversion factor matched the observed difference. You could probably also find relevant information in the reaction rate source code.

Caelan

[Gerry Kan]

Dear Caelan:

Thanks for the link. This is also in line with what I read. Also, looking through some of the chemistry tutorials it does seem to confirm that the units for the A factors are indeed in kmol.

I am asking because, for my very simple mechanism, I have converted the reaction constants from [ cm, molecule, s ] to [ m, kmol, s ] and their magnitude seem a bit different from what I expect (about 3 orders of magnitude "smaller"), and the overall reaction also seem to be about 1000 times "slower" compared to measurement data. However I double checked my arithmetic and they are, for the most part, correct.

I will dig through the reaction code to see if I can see anything insightful.

Sincerely, Gerry.

[schuyler]

Hi Gerry,


Did you have any luck with this? I too am a bit confused on this front.

[sqek]

Hello
I'm almost certain it should be kmol (at least in the latest version):
In StandardChemistryModel::calculate the concentrations are got by dividing the species density by its molecular weight

c_[i] = rhoi*Yi/specieThermos_[i].W();

and from specie::W, W is in kg/kmol
the c_[i] from the chemistry model is passed on to the omega/omegaI functions in the chemistry model, and then into reaction which also has omega/omegaI functions.
These call IrreversibleReaction::kf (or reversible reaction, or kb, as needed)
Which eventually works its way down to something like ArrheniusReactionRate:: operator()
There don't seem to be any unit conversions between any of these, so the Arrhenius rate must be in kmol, M^3, s

[Marpole]

I would like to echo this. Yes, OpenFOAM uses SI system [kg, m, s, K, mol, A, cd]. But OpenFOAM chemical reaction part uses [m, s, kmol]. This is very confusing.

In the same document, that Gerry referred to, https://openfoamwiki.net/images/8/8a...FW6_slides.pdf , on page 27, the chemkin pre-exponential factor is 3e13. After conversion with chemkinToFoam, the pre-exponential factor becomes 9.48e11.

There is one thing missing in the above discussion that the conversion is not from to . It acutally is to , where n is the total reactor order. In the example case, n=0.7+0.8 = 1.5.

The conversion factor is then

The conversion factor is .

In the slide, the chemkin pre-exponential facor is 3e13. We can do conversion, . This matches the OpenFoam pre-exponential factor in the slide.

[khaledhmz]

Originally Posted by Marpole

I would like to echo this. Yes, OpenFOAM uses SI system [kg, m, s, K, mol, A, cd]. But OpenFOAM chemical reaction part uses [m, s, kmol]. This is very confusing.

In the same document, that Gerry referred to, https://openfoamwiki.net/images/8/8a...FW6_slides.pdf , on page 27, the chemkin pre-exponential factor is 3e13. After conversion with chemkinToFoam, the pre-exponential factor becomes 9.48e11.

There is one thing missing in the above discussion that the conversion is not from to . It acutally is to , where n is the total reactor order. In the example case, n=0.7+0.8 = 1.5.

The conversion factor is then

The conversion factor is .

In the slide, the chemkin pre-exponential facor is 3e13. We can do conversion, . This matches the OpenFoam pre-exponential factor in the slide.

Hello bro,
I have the same problem I use chimken reduced reaction mechanism and when I import it in fluent I found the pre-exponential factor and activation energy units are changed I try to find a solution to convert it to SI units but I stuk in it, can you help me please?

[Marpole]

Sure. What's the exact problem you have experienced?

[khaledhmz]

Thank you bro for the reply,

the problem is I use reduced mechanism of propane-air and in the article the author gave the pre-exponential factor units as cm3-mol-s and activation energy as j/mol and when I import the chemkin file to Ansys-fluent I found the value of A and Ea is changed, now Im trying to convert them to SI units as fluent reaquire but still have problem with them,
without the correct value of A and Ea the simulation will be wrong, so any suugestion about this or you can help directly because Im confused a lot....

[Marpole]

I heard your frustration and confusion. But I don't know your problem.

[khaledhmz]

the problem is how to convert from cm3-mol-s to kg/m3-s and for activation energy from j/kgmol to cal/mol...

[Marpole]

still not sure what's your problem.

For cm3-mol-s to kg/m3-s, can you do conversion: cm3 -> m^3, mol->kg, and s->s? For mol->kg, you need to use molecular weight.

Activation energy unit is J/kg. Can you do J->cal, kg->mol (same as above)? I'm not sure J/kgmol in your message.

[khaledhmz]

Originally Posted by Marpole

still not sure what's your problem.

For cm3-mol-s to kg/m3-s, can you do conversion: cm3 -> m^3, mol->kg, and s->s? For mol->kg, you need to use molecular weight.

Activation energy unit is J/kg. Can you do J->cal, kg->mol (same as above)? I'm not sure J/kgmol in your message.

bro can you explain more about the pre-exponential factor; how can I use molecular weight?
- for the activation energy this is what fluent required

[Marpole]

Starting from simple, for activation energy, please check out this link.

problem with activation energy in fluent

For pre-exponential factor, we can get it from the following example. Assuming you have a reaction,



The reaction rate is, for elementary reaction,

The pre-exponetial factor has the same unit of reaction rate constant k.

The rate of reaction (r) has a unit of mole/cm^3/s. So the unit of k should be,


So you have to do unit conversion one by one. It would be easier to do it with an excel spreadsheet as you may have many reactions.

[khaledhmz]

so what I understand for activation enrgy to convert from j/mol to cal/ kmol, I can just divided by 4.184 j to convert from joules to cal and then devide by 10E-3 to convert from mol to kmol, then fluent adjust it to j/kgmol (see attached picture)


and for pre-exponential factor, to convert from (mol/cm3/s)/(mol/cm3)-(a+b) to (kmol/m3/k)/(kmol/m3)-(a+b)
where a and b is reaction order.
I do the following, (10E-3 *mol/10E-6*cm3/s)/(10E-3*mol/10E-6*cm3)-(a+b)
is this correct?

[Marpole]

Please pay attention to the formula. The reaction order -(a+b) should be exponent: (mol/cm3/s)*(mol/cm3)^[-(a+b)] or (mol/cm3)^(1-a-b)/s.

[khaledhmz]

Thank you very much Bro

[DaveD!]

Originally Posted by khaledhmz

so what I understand for activation enrgy to convert from j/mol to cal/ kmol, I can just divided by 4.184 j to convert from joules to cal and then devide by 10E-3 to convert from mol to kmol, then fluent adjust it to j/kgmol (see attached picture)


and for pre-exponential factor, to convert from (mol/cm3/s)/(mol/cm3)-(a+b) to (kmol/m3/k)/(kmol/m3)-(a+b)
where a and b is reaction order.
I do the following, (10E-3 *mol/10E-6*cm3/s)/(10E-3*mol/10E-6*cm3)-(a+b)
is this correct?

Dear Khaled,

the kgmol in the Fluent reaction dialogue is a very ancient unit, meaning 1000 mol or 1 kmol:

https://www.aqua-calc.com/what-is/am.../kilogram-mole

So basically, we can consider kgmol to be kmol. I think they should change it to avoid further confusion!


Regards, DaveD