Learning Forum

[abtharpe42]

I need help understanding what units Fluent uses for the pre-exponential factor when using chemical reactions. I understand that the units for pre-exponential factor are implied depending on the rest of the terms in the reaction rate equation and that the values that appear in the reaction window for both pre-exponential factor and activation energy in Fluent are automatically adjusted depending on the units in the Chemkin mechanism import. For an example with a first-order volumetric reaction:

R=k*C_nh3

Where "R" and "C_nh3" are in units of [mol/m^3*s] and [mol/m^3] respectively, meaning "k" and the pre-exponential factor have to be in units of [s^-1] to satisfy the units for "R". Another example, which is actually what I'm currently trying to figure out, is a first-order surface reaction:

R=k*C_nh3

Where "R" and "C_nh3" are in units of [mol/m^2*s] and [mol/m^3] respectively, meaning "k" and the pre-exponential factor have to be in units of [m/s] to satisfy the units for "R". In Fluent, the default units for concentration and reaction rate are [kmol/m^3] and [kmol/m^3*s] OR [kmol/m^2*s], respectively. I've seen while skimming through some older Chemkin documentation that the prexponential factor is prescribed in a mechanism assuming its component units are [mol] and [cm] by default, although I am unsure if that is how it still works current day in Fluent and Chemkin. 

I have am trying to simulate a first-order surface reaction (NH3 = 3/2H2 + 1/2N2, R=k*C_nh3) where the Arrhenius parameters are given as:

A = 3.10*10^3 [m/s]   ,   Ea = 1.16*10^8 [J/kmol]

When I am setting up a text file for a Chemkin mechanism using a research paper as the source for the parameters, should I convert the prexponential from:

 A = 3.10*10^3 [m/s]  >  A = 3.10*10^5 [cm/s]

So that Fluent automatically converts it to [m/s]? And does Fluent automatically convert a [mol] unit in a pre-exponential factor to [kmol] instead? Thanks in advance.

[jcooper]

Hi:

The conventions that Fluent will follow depend on the source of the mechanism.  If the mechanism is being input manually, it is advisable to keep the Fluent working units set to SI.  Fluent will follows the convention  [kmol/m^3] and [kmol/m^3*s] OR [kmol/m^2*s] for reactions.

When a reaction mechanism is read using Chemkin import, the mechanism keyword, UNITS will drive how the Chemkin reader (and therefore Fluent) interpret the units.  Information on this keyword is below:

 

UNITS

Reaction Units - Supersedes the current units for a particular reaction rate fit that may differ from the default units specified for other reaction expressions in the chemistry mechanism. UNITS must be followed by the slash-delimited character-string string, where string is one of the following: EVOL[TS], KELV[INS], CAL/[MOLE], KCAL[/MOLE], JOUL[ES/MOLE], or KJOU[LES/MOLE] for parameters with energy units such as E i, or MOLES or MOLEC[ULES] for pre-exponential factors A i, where the letters in brackets are optional. The inclusion of MOLEC[ULES] would indicate that the reaction rate expression is in units of molecules/cm3 rather than mole/cm3. The UNITS auxiliary keyword allows only one string parameter, but the user can repeat the UNITS as many times as needed for a given reaction.

The UNITS keyword is written as UNITS/string/, where string specifies the desired units.

 

Some other points to consider:

• Common unit strings include EVOLTS, KELVINS, CAL/MOLE, KCAL/MOLE, JOULES/MOLE, KJOULES/MOLE for energy parameters, and MOLES or MOLECULES for pre-exponential factors.

• If the UNITS keywords is specified on the REACTIONS header line, the units apply to all reactions unless superseded by an auxiliary UNITS keyword for a particular reaction.

• The UNITS keyword can be repeated as needed for different reactions within the mechanism file.
• Example below:

! Example: Specify reaction rate units in molecules and activation energy in electron volts
CF3+ + E + #WSIO2(B) => #SIO2 + CF3 0.33 0.0 0.0 BOHM
UNITS/MOLECULES/
UNITS/EVOLTS/

 

 

Details on all the Chemkin mechanism keywords are available in the Chemkin Input Manual in Chapter 3: Auxiliary Reaction Data

https://ansyshelp.ansys.com/Views/Secured/corp/v252/en/chemkin_in/i18070.html

[SamW]

Hello, I can't speak to your specific reactions but in general your information about units looks to be up to date. You can check the Chemkin input guide for current information about units specification. In general, Fluent uses [m, kmol, sec] for units and Chemkin uses [cm, mole, sec]. Automatic conversion will take place on import based on the nature of the reaction as defined in the Chemkin file. You will also want to convert your Ea, since kmol is not avalable for REACTIONS-line options per the link above.

[abtharpe42]

Thanks for the info. For my next question, a surface reaction that I was given goes as follows:

MATERIAL CATALYST

SITE/Ru_SURFACE/    SDEN/1.66E-9/

Ru(S)

END

 

THERMO

Ru(S)                   Ru  1               S    300.0    3000.0  1000.0       1

0.00000000E+00 0.00000000E+00 0.00000000E+00 0.00000000E+00 0.00000000E+00    2

0.00000000E+00 0.00000000E+00 0.00000000E+00 0.00000000E+00 0.00000000E+00    3

0.00000000E+00 0.00000000E+00 0.00000000E+00 0.00000000E+00                   4

END

 

REACTIONS        KJOULES/MOLE   MWOFF              

NH3 + Ru(S) => 1.5H2 + 0.5N2 + Ru(S)    1.5E16  0.0  128.0

END

 

Where R = k*C_nh3 is the provided reaction. "k" would be in units of [s^-1], so there should be no conversions needed for pre-exponential factor "A".

 

However, I noticed when importing the mechanism into Fluent that the value displayed for the pre-exponential factor is given as 1.5E13 instead of the mechanism's 1.5E16. Does this mean that the reaction rate equation shown above is not correct and that there should be a [mol] unit somewhere in "A" to get a factor of 1000 involved? This is extremely important because I'm having issues simulating a 2D axisymmetric laminar flow through a channel with a section of isothermal walls set to 1000 K where the surface reaction rate is unreasonably small for the temperature I've applied. To reiterate, the walls set to 1000K are where the surface reactions are enabled. The inlet composition is a 1:1 molar ratio of NH3 and N2.

[SamW]

It's possible I'm misunderstanding your intent, but it looks like the reaction as you have it written would have a rate expression of R = k*C_nh3*S_Ru, where S_Ru is the surface concentration of Ru (e.g. mol/cm^2 in Chemkin units), so k would need to have different units to account for this. Since R has units of kmol/(m^2*s) for a surface reaction in Fluent, you do get a factor of 1000 once you do the conversion. Maybe you are trying to accomplish something different?

[abtharpe42]

No you had it right. What I was given to simulate was a surface reaction where I was told the reaction rate equation is R = k*C_nh3. During the period of waiting for a response to my latest post, I did ultimately come to the conclusion that the correct reaction rate equation is R = k*C_nh3*S_Ru as you've said. I don't think my colleague realized this mistake, but according to him he had successfully validated that the simulation results using the reaction mechanism with the results in literature, albeit by accident apparently. I've figured out that Chemkin mechanism's default units for surface reactions are:

R = [mol/cm^2*s]     (wall source/sink)

C_nh3 = [mol/cm^3]     (gas-phase concentration)

S_Ru = [mol/cm^2]     (surface site concentration and obtained using the site density in the mechanism and the surface concentration in the wall BC menus)

The units of "k" therefore has to be [cm^3/mol-s], which is what the units of "A" defaults to in the mechanism file. Since Fluent's default units for the same parameters are:

R = [kmol/m^2*s]     (wall source/sink)

C_nh3 = [kmol/m^3]     (gas-phase concentration)

S_Ru = [kmol/m^2]     (surface site concentration)

The units of "A" are therefore [m^3/kmol*s] times a factor of 0.001 after working out the conversion from [cm^3/mol-s], hence the change of 1.5E16 in the mechanism to 1.5E13 in Fluent's reaction window. So this part is figured out, which is great. Now I'm stuck trying to figure out why the other simulation I'm working on seems to not be working with the surface reaction. I can save that for another Forum post later, if that is preferred.

[SamW]

Great! Glad it makes sense.

As you say, if you're dealing with other questions relating to surface reactions it may be best to create a new post with some more details, and me or another community member can weigh in and see if we can help. Thanks!