vlachosgroup / openmkm Goto Github PK

For the FIM run, is the FIM_Diag value the Fischer information matrix diagonal value of the relevant reaction? If not, please correct the documentation.

The following is a list of reactor models to be implemented. Some of them have common implementation and distinguished only with operating conditions.

1. TPD Surface science
2. Molecular beam flux
3. PFR with T profile imposed
4. PFR with heat and mass transfer
5. Particle model (BVP)
6. Fixed bed reactor model (BVP)
7. Non isothermal PFR
8. TPD in fixed bed

In the above 1) TPD designates operational mode.
2) PFR with T profile imposed and PFR with heat transfer and non-isothermal PFR require energy equation to be solved.

Looks like having more than one surface species in an elementary step is not allowed. Can this be fixed?

Can you please fill out the sections titled “Files with _ss.out extension:” and “Files with _tr.out extension:” with the appropriate information? While doing this, can you please add an explanation in the documentation on why the first volume change for the numerical PFR is different from all other volume changes? Also, can you please add an explanation in the documentation on why the final volume is not equal to the total reactor volume? There are partially written sections on these for you to complete.

Please provide a description in the documentation of the what information is provided in the sensitivity file output in a LSA run.

Chemkin allows for activation energies to be set for each run in a series as specified in Tflow.inp. These activation energies are usually set to allow a temperature-dependent activation energy. Therefore, a separate activation energy in the EA input file aligned with the temperature of the simulation as specified in Tflow.inp accomplishes this task. Chemkin provides two input files for this purpose; one for the surface reactions (EAs.inp) and one for the gas reactions (EAg.inp). In addition, there is a flag in tube.inp to turn on/off the use of a separate EA input file (lEA). In the EA input file EA's are specified by reaction. If the IEA flag is True then Chemkin looks for an EA in the EA input file otherwise it uses the EA specified in surf.inp or gas.inp.

cantera chemkin io reader failed to parse the equation "CH3CH3+2PT(S)<=>CH2CH3(S)+H(S)+2PT(B)"

Please provide the formula used to calculate Kc under the kc.out documentation and and include the units Kc will have for each formula.

It does not seem that gas phase reactions are implemented in omkm, although an example on such a reactor set up is given in the examples: .

With pmutt I am able to write (cti and xml) files that include reactions that occur in the gas phase as well as on the surface. While it is an alternative to stick to pure cantera without openmkm for sole gas phase systems, this is not so easily achievable when I need both surface reactions and gas phase reactions.

It might very well be that I am doing something wrong, so do you have an example on such usage?

The error is get when running OpenMKM is:

'''
Error raised in Cantera. Exiting...

CanteraError thrown by GasKinetics::addReaction:
Unknown reaction type specified: 20

'''
A first test was made by simply adding the direct formation of NH3 from N2 and H2 in the gas phase (in your openmkm example from pmutt) in addition to all the surface interactions.

I would be very grateful on some pointers on how to proceed here.

Can we have an example of what a gas_mole_ss.out and accompanying gas_mole_tr.out look like for a numerically solved pfr? Currently, the transient file is not outputting meaningful values. Once we have these files, we will elaborate the documentation on them.

Hi @mbkumar, we see you added "mass flow rate into the reactor (in kg/s), mass flow rate out of the reactor (kg/s), mass produced from the surface (in kg/s), net rate of change of mass (kg/s) in the reactor" to the output files documentation for the rctr_state output file. However, we do not see these values in our output file. Can you clarify when we should expect to see these values? Are they for an updated version of OpenMKM, or for specific reactors only?

Windows executable is not displaying the error thrown by Cantera or OpenMKM.
An alternative is print the error in the main file and gracefully exit.

I ran an ammonia model at 1atm and then ran it at 100 atm.
The Hform.out values are all the same between the runs, but the Hrxn.out values change. The biggest change is for this reaction:
At 1atm: -9.002899e-02 NH(T) + RU(T) <=> H(T) + N(T) + RU(B)
At 100 atm: -7.910262e-02 NH(T) + RU(T) <=> H(T) + N(T) + RU(B)
All species involved in these reactions are surface species, which don't even have their Sform.out values affected by pressure.

Why do the Hrxn.out values change when the Sform.out and Hform.out values are identical?

Caviness paths with relevant MKMS:
1atm location: /work/ccei_biomass/users/maxrc/MKM/openmkm/NH3_1atm
100atm location: /work/ccei_biomass/users/maxrc/MKM/openmkm/NH3_100atm

Is kf calculated at the reactor inlet conditions or the reactor outlet conditions? Please add the necessary information to the documentation under kf.out.

An YAML format for inputting reactor parameters has been implemented.
Pending Issues:

1. Test all 0d reactors
2. Test 1d reactors
3. Test all operating modes
4. Check heat conducting walls.

Implement CSTR.
CSTR can be implemented directly from cantera, but cantera is missing solver for CSTR. Only ReactorNet has integrator implemented.

1. Attach appropriate components (Inlets, outlets, walls, surfaces) to cantera's Reactor class to define CSTR
2. Validate the governing equations generated by cantera.
3. Implement time integrator and steady state solver. Steady state solver is the one we care about. Integrator is to ride out the fast equations.

Please provide a description in the documentation of the information in the file output in a sensitivity analysis that contains “_state” in the filename.

In the rctr_state file, are the reported density values the density of the gas phase? If not, please correct the documentation.

Dear all,

In the example of homogeneous catalysis provided in JCIM (J. Chem. Inf. Model. 2023, 63, 3377−3391), the pre-exponential factor (A) is a constant value, and the activation energy (Ea) is replaced using Gibbs free energy (ΔGts), indicating that the pre-exponential factor is calculated using the formula for surface reactions. So for homogeneous catalytic reactions, what do the site density (σ) and the number of surface specifications (m) in the formula represent?

Looking forward to everyone's reply！

What pressures are Hform and Sform calculated at? This matters for the gas-phase species. Please add this information to the underlined section in the documentation.

What are the formulas used to calculate kf? Please add this information to the documentation for kf.out and include the units kf will have for each formula.

For the quantities of Hform and Sform, is the temperature these are calculated at the reactor inlet temperature? If not, please correct the documentation.

Currently OpenMKM is based on Cantera 2.4. The newer version, Cantera 2.5, supports YAML format to specify the thermochemistry data. This is easy on the eye compared to the XML format. There is also an added advantage of not having to work with CTI file and the associated conversion.

Things to do:

1. Generate YAML file from pMuTT conforming to dict format required by Cantera for terms already supported by Cantera.
2. Rebase openmkm branch of Cantera to incorporate updates from upstream
3. Support YAML format for thermochemistry inside OpenMKM
4. Define YAML format for BEP and lateral interactions
5. Support BEPs and lateral interactions to be read from YAML file inside Cantera
6. Implement the associated code in pMuTT (as part of 1) for BEPs and lateral interactions

Handlers:

Task-1: Jon
Task-2: Bharat,
Task-3: Bharat
Task-4: Bharat, Jon, and Gerhard
Task-5: Bharat
Task-6: Jon.

Reference: https://cantera.org/tutorials/yaml/defining-phases.html

Hi,

After the latest commits in August I am no longer able to compile OpenMKM with the Cantera from the forked cantera repo as specified in the openmkm instructions.

git clone https://github.com/mbkumar/cantera.git.
cd cantera
git checkout openmkm
```
Has the required cantera installation been updated elsewhere?
With my openmkm fork from prior to August I have no problems with the installation itself. 
I am installing with the Dockerfile as in the pull request I made as a suggestion a few months ago.

Best regards, 
Francesca


```
This is the end of the errors messages during compilation:
#22 102.7 io.cpp:229:18: error: 'class Cantera::Kinetics' has no member named 'getDeltaRefEntropy'; did you mean 'getDeltaEntropy'?
#22 102.7   229 |             mgr->getDeltaRefEntropy(sRxnRef.data());
#22 102.7       |                  ^~~~~~~~~~~~~~~~~~
#22 102.7       |                  getDeltaEntropy
#22 102.7 io.cpp: In function 'void OpenMKM::print_rxn_gibbs(std::vector<Cantera::Kinetics*>, doublereal, std::string)':
#22 102.7 io.cpp:276:18: error: 'class Cantera::Kinetics' has no member named 'getDeltaRefGibbs'; did you mean 'getDeltaGibbs'?
#22 102.7   276 |             mgr->getDeltaRefGibbs(muRxnRef.data());
#22 102.7       |                  ^~~~~~~~~~~~~~~~
#22 102.7       |                  getDeltaGibbs
#22 113.5 scons: *** [io.o] Error 1
#22 113.5 scons: building terminated because of errors.

Do the values in the rates_ss.out file correspond to quantities at the outlet of the reactor? If not, please correct the documentation.

Currently, OpenMKM only prints net production rate of species (both gas and surface species), due to reactions occuring on the surface phase. OpenMKM does't print production rate of species due to gas-phase reactions. Gas-Phase production rates are important and provide valuable insight. This behaviour is inconsistent with CHEMKIN.

Resolution:

1. OpenMKM should write independent production rate files for reactions occuring in each phase, i.e.,

   • gas_sdot_ss/surf_sdot_ss (for net production of gas/surface species from surface reactions)
   • gas_gdot_ss (for net production of gas species from gas-phase reactions)

2. Alternatively, OpenMKM could follow CHEMKIN convention and print just two files species_sdot (combination of gas and surf sdot) and species_gdot which is just gas_gdot.

Hi Bharat,

Quick question. Can ReNView be called from OpenMKM or do users have to run it manually? If yes to the former, what keywords are required?

Can you please clarify that for a numerical PFR set of results is each row in the _ss.out files a set of values corresponding to the quantities entering the node? Or are they the quantities leaving that node?

Please specify the units of the reported values in the gas_msdot file.

Hi @mbkumar, we noticed you added this surf_sdot.out file to the outputs documentation. However, we have not seen it output in our runs of OpenMKM. Can you specify when this file is generated, or has it been removed from the outputs now?

Is it correct that Hrxn and Srxn can be calculated accurately from Hform and Sform and the reaction stoichiometry? If not, please explain the differences in the documentation under Hrxn.out and Srxn.out.

Implement Ideal PFR.
Background: A series of CSTRs is often used to simulate PFR. However an implementation of an ideal PFR is also desired.

Add python code to parse chemkin input files such as tube.inp and others.
Cantera code can be used to parse gas.inp, surf.inp and thermdat files.

Please fill in the necessary information in all underlined sections and words of “BLANK” in the documentation.

PFR is failing at different optimization levels. Understand the issue.

Please provide a description in the documentation of how to use the information in the LSA run output sensitivity file for the sensitivity analysis.

We know from BEP relationships that activation energies are correlated to heats of reaction. In a microkinetic model heats of reaction can change based on a number of factors that change the surface energies of adsorbed species. The most important of these is lateral interactions. As surface concentrations of species change, adsorption energies change, and heats of reaction change. Therefore, in kinetic models where BEP's are used to determine activation energies, changes in surface reaction energies are reflected properly in the activation energies. Reactions using a "fixed" activation energies (fixed wrt surface heats of reaction) as specified in the surf.inp, gas.inp, EAs.inp, or EAg.inp the changes in surface reaction energies are not reflected in the surface reaction activation barriers. To address this Chemin uses a proximity factor as describer by Grabow, L. C. et al. (2008) (‘Mechanism of the water gas shift reaction on Pt: First principles, experiments, and microkinetic modeling’, Journal of Physical Chemistry C, 112(12), pp. 4608–4617).

The proximity factor ranges from 0 - 1 and represents the reaction coordinate of the transition state for that reaction and functions identically to the BEP slope appling a fraction of the change in heat-of-reaction to the activation energy.

Chemkin provides for a default proximity factor in tube.inp (omega) and a flag (Iomega) to enable both proximity factors and a per reaction proximity factor file (omega.inp). The omega.inp file contains a list of reactions and associated proximity factors. If Iomega is True then, for all reactions not using BEP's, Chemkin first looks in the omerga.inp for a proximity factor and, if not found, uses the default value specified in tube.inp.

Is Kc calculated at the reactor outlet temperature and standard pressures of 1 bar for gaseous species? If not, please correct the documentation.

Hi @mbkumar, can you clarify what the Note in the rctr_state file documentation means? It says "Note: Mass flow rate out of the reactor is printed as -ve value. Similarly, if mass produced from surface is -ve, it means surface absorbs more mass than the mass released." What does "-ve" mean? Thank you for the help.

Do all the files under the “Reactor State Data” section have both _ss.out and _tr.out files output for every type of reactor? For example, I do not see a rates_tr.out file for the numerical PFR. I have added an exception for that case to the documentation. Please add any additional exceptions or necessary clarifications.

Is it correct that Grxn can be calculated using the values from Hrxn and Srxn by the formula Grxn = Hrxn – T*Srxn? Is this T the reactor inlet temperature? If any of this is wrong, please correct the documentation for Grxn.out.

Please specify the units of the specific enthalpies reported in the rctr_state file.