Learning Forum

[s.sethuraman]

Hi everyone,

I’m trying to understand how Energy Due to Reaction (also known as Heat of Reaction) is calculated in ANSYS Fluent.

According to the Fluent Theory Guide (Eq. 5-11), the energy source term due to reaction is given as:

S_h,rxn = ∑ h⁰_j * R_j / M_j 

where h⁰_j is the enthalpy of formation, M_j is the molecular weight and R_j is the creation rate of species. 

My main question is about R_j: Is R_j the net reaction rate (in [kg/m³·s]) as used in source terms, or is it R̂_{i,r}, the molar rate of creation/destruction of species i in reaction r, as defined in Eq. 7-11 of the Theory Guide? Intuitively, I believe it should be the net reaction rate (mass basis), especially since the units line up, and this is also suggested in the ANSYS Forum article here: Here

However, when I follow the method in that article, the calculated Sh_rxn values differ significantly from those shown in Fluent’s Heat of Reaction contour plot.

If anyone can help explain this discrepancy or clarify how Fluent computes the heat of reaction internally, I’d greatly appreciate it. 

Thanks in advance!

[jcooper]

 

 

Hi:

The heat of reaction source term can be reproduced from the S_rxn term as it is above, but it is best to use integral to verify the calculation:

 

For a simple methane reaction:

rate of production of CH4 = NetReactionRateof(species = ‘ch4’)/16.04303 [kg kmol^-1]

rate of production of CO2 = NetReactionRateof(species = ‘co2’)/44.00995 [kg kmol^-1]

rate of production of H2O = NetReactionRateof(species = ‘h2o’)/18.01534 [kg kmol^-1]

(O2 rate doesn’t matter because the heat of formation of O2 is zero at 298.15 K)

 

S_h_rxn = (-7.49e+07 [J kmol^-1]*rateof_ch4 – 3.935e+08 [J kmol^-1] *rateof_co2 – 2.418e+08 [J kmol^-1]*rateof_h2o)*-1.0 (sign change is due to the convention change S_h_rxn is reported as viewed from reaction rather than from surroundings)

 

We can then integrate S_h_rxn over the volume:

HR = VolumeInt(S_h_rxn,[‘default-domain’])

 

And do the same with the Fluent heat of reaction, which is also reported volumetrically:

VolumeInt(HeatofReaction/vol,['default-domain'])

 

 

The above integrals should be comparable, but results may differ somewhat because the solver used more significant figures and because the solver can access intergration points that are unavailable for post-processing calculations.  For very low reaction rates, the “noise” due to these differences can overpower the calculations with round-off error because the numbers involved are very small.

For multiple reactions, you can do this same calculation for multiple reactions to get the net heat release, but the error may be larger due to accumulated error in the calculations. The heat of reaction reported by Fluent will have the closest correspondence to the solution.

I hope this helps.

Regards,

Judy