Incorporating the LMR-R mixture rule into Cantera for the first time #1710

pjsingal · 2024-06-12T01:11:25Z

Changes proposed in this pull request

LmrRate.cpp and LmrRate.h allow for vastly improved consideration of the mixture-dependent aspect of the rate constants for complex-forming ("pressure-dependent") reactions, using the new LMR-R mixture theory developed by the Burke Lab at Columbia University
This code builds, runs, and produces results that have been validated and peer-reviewed
LMR-R can handle three different representations of the pressure-dependent aspect of complex-forming reactions: Troe, PLOG, and Chebyshev
LMR-R calculates the mixture-dependent aspect using temperature-dependent third-body efficiencies that have been supplied by the user for at least one collider - these parameters can be provided in either relative or absolute terms, as long as the user's choice is consistent for all colliders in a given reaction
Description of the theory and instructions for implementation have been added to rate-constants.md and reactions.rst, respectively
Four new Python examples have been added

Checklist

[x] The pull request includes a clear description of this code change
[x] Commit messages have short titles and reference relevant issues
[x] Build passes (scons build & scons test) and unit tests address code coverage
[x] Style & formatting of contributed code follows contributing guidelines
[x] The pull request is ready for review

codecov · 2024-06-12T11:42:05Z

Codecov Report

Attention: Patch coverage is 0.78740% with 252 lines in your changes missing coverage. Please review.

Project coverage is 72.87%. Comparing base (ec0cfdb) to head (0eb6e58).
Report is 3 commits behind head on main.

Files with missing lines	Patch %	Lines
src/kinetics/LinearBurkeRate.cpp	0.00%	224 Missing ⚠️
include/cantera/kinetics/LinearBurkeRate.h	0.00%	15 Missing ⚠️
interfaces/cython/cantera/reaction.pyx	14.28%	12 Missing ⚠️
src/kinetics/ReactionRateFactory.cpp	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1710      +/-   ##
==========================================
- Coverage   73.23%   72.87%   -0.37%     
==========================================
  Files         381      383       +2     
  Lines       54375    54616     +241     
  Branches     9251     9282      +31     
==========================================
- Hits        39823    39802      -21     
- Misses      11580    11832     +252     
- Partials     2972     2982      +10

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

…'t yet run (the version in previous commit both builds and runs)

…l speed by changing obj defs

This reverts commit d355c9e.

…MRR ref materials

speth

Thanks for the continued work on this, @pjsingal. I would suggest that the first priority at this point is to add some unit tests. I would recommend getting those into place before making any of the other changes that I've suggested below, so you can be sure that making those changes doesn't introduce any regressions.

speth · 2024-09-17T16:18:34Z

data/alzueta.yaml

This mechanism is a good candidate for moving to the separate example_data repository (https://github.com/Cantera/cantera-example-data).

I appreciate the effort to at least share the species data between the two versions of the mechanism, but I'd like to suggest going a step further, and sharing the reactions that they have in common as well. To do this, you would need to divide the reactions into three sections in the input file: one for the shared reactions (named for example common), one for the baseline mechanism's pressure-dependent reactions (named for example baseline-pdep), and one for the reactions implemented using the linear-Burke rate model (named linear-Burke). Then, you can put two phase definitions in the file, one which specifies the reactions as reactions: [common, baseline-pdep] and the other that specifies the reactions as reactions: [common, linear-Burke].

speth · 2024-09-17T16:56:20Z

samples/python/kinetics/ignition_delay_time.py

There are several other examples showing IDT as a function of temperature, so I think we probably don't need this example. Your paper is the main place for demonstrating this result.

speth · 2024-09-17T19:09:30Z

samples/python/kinetics/flame_speed.py

+Two models are compared in this example:
+(i)  A 2023 model of H2 and NH3 chemistry published by Alzueta et al. [2]
+(ii) An adapted version of this model that has applied the reduced-pressure linear mixture
+     rule (LMR-R) and ab initio third-body efficiencies [3]
+


Please make sure that this header can be parsed as reStructuredText and that the output looks like what you are expecting. You can take a look at the generated HTML by downloading the "docs" artifact from the CI builds (go to the "Checks" tab at the top of this PR page, then look through the list of "artifacts").

Also, please wrap content, including the comments, to no more than 88 characters.

speth · 2024-09-17T19:44:26Z

include/cantera/kinetics/LinearBurkeRate.h

@@ -0,0 +1,140 @@
+//! @file LINEARBURKERATE.h


The capitalization here should match the capitalization of the file name -- otherwise, Doxygen issues a warning.

speth · 2024-09-17T19:46:41Z

include/cantera/kinetics/LinearBurkeRate.h

+ * The data container `LmrData` holds precalculated data common to
+ * all `LinearBurkeRate` objects.
+ */
+struct LmrData : public ReactionData


The name of this class should also be changed to something less obscure -- LinearBurkeData would be consistent with the pattern that we use with the other rates and their corresponding data classes.

speth · 2024-09-25T03:31:24Z

src/kinetics/LinearBurkeRate.cpp

+        // Collider has an eps specified, but no other info is provided. Assign it the same rate and data objects as "M"
+        else {
+            rateObjs.push_back(rateObj_M);
+            dataObjs.push_back(dataObj_M);
+            epsObjs1.push_back(epsObj_i);
+            epsObjs2.push_back(epsObj_M);
+        }
+    }


Since this seems to be the most common case (based on the mechanism used in the examples), for the sake of efficiency I think you want to avoid recalculating the value of rateObj_M for each collider when (I think) it will just have the same value for all of them and all you really need is to compute it once and scale it by the sum of the efficiency-weighted mole fractions.

speth · 2024-09-25T03:41:07Z

doc/example-keywords.txt

+mixture rule
 multicomponent transport


I don't think this is going to be a well-understood keyword to most users, and I'd recommend just leaving it out.

speth · 2024-09-25T03:42:26Z

doc/example-keywords.txt

+species profile
+temperature profile


I would recommend dropping these two keywords as well

speth · 2024-09-25T03:46:32Z

interfaces/cython/cantera/reaction.pyx

+            elif input_data:
+                raise TypeError("Invalid parameter 'input_data'")
+            else:
+                raise TypeError("Invalid parameter 'rates'")
+            self.set_cxx_object()


You should be able to get the constructor from input_data working by following the example of the other rate types. This is just a Python data structure containing the same information as the YAML input.

I think you can just remove the rates kwarg -- I'm guessing that's just leftover from copying from the PlogRate implementation.

speth · 2024-09-25T03:51:56Z

include/cantera/kinetics/LinearBurkeRate.h

+    void getParameters(AnyMap& rateNode, const Units& rate_units) const;
+    void getParameters(AnyMap& rateNode) const override {
+        return getParameters(rateNode, Units(0));
+    }


I think you can just implement the version that doesn't take a Units object directly and skip providing this other overload of getParameters.

speth added the Kinetics label Jun 12, 2024

pjsingal changed the title ~~Burkelab clean~~ Incorporating the LMR-R mixture rule into Cantera for the first time Jun 12, 2024

speth mentioned this pull request Jun 13, 2024

Advanced mixing rules for colliders Cantera/enhancements#157

Open

ischoegl added the enhancement label Jul 1, 2024

pjsingal force-pushed the burkelabClean branch from 5b74b9f to 59d89de Compare September 4, 2024 19:38

pjsingal and others added 24 commits September 4, 2024 15:55

synced branch to upstream/main and then added in LMRR files

a9c300b

objs are now declared in setParameters. The code builds, but it doesn…

b6f39ea

…'t yet run (the version in previous commit both builds and runs)

Added LmrRate.h ReactionRateFactory.cpp

677204d

added setContext. Scope issue remains w/ colliderIndices

7b203f3

commented out a test line for colliderIndices left by mistake

8bb6b6b

bug fixed, code now builds

4dc14c5

cleanups and minor debugs

47157b7

Added my local test file dir to the gitignore

56f5687

add bklab files to gitignore

a2aeecf

add bklab test yamls to gitignore

07ca746

updated gitignore again

3771c70

removed kineticsfromscratch_LMRtest.yaml from remote tracking

a16ac2f

tried to fix k_i_plog returning nan, but could not yet find solution

cbcaf98

added node_M to eig0 only case in setParams

9e259d3

debugging

2f340a8

rate units are not being calc properly for all rxns

94c9b28

added "LMR_R" to ReactionPath.cpp

a11bca8

Added LMR_R to Reaction.cpp

b2ac4f9

didn't fix rate_units problem, but dramatically improved computationa…

8d1a9e4

…l speed by changing obj defs

partial code cleanup

694e654

attempt to reduce data obj definitions (doesn't yet work)

0e5d184

reorganized data obj definition and updates

f8811cc

found very inelegant solution to rate_unit problem

aed0da2

fixed eig0 units (now dimless), made sigmaXM subtractive not additive

61a8f57

pjsingal added 6 commits September 4, 2024 19:48

added newlines to ends of examples, fixed whitespace

7fceded

added newlines to ends of lmrr scripts

069c504

fixed flame speed example indentation

c159659

shortened alzueta_LMRR by importing alzueta.yaml

3e9d52c

added newline to end of gitignore

ba12286

added description to alzueta_LMRR.yaml

780c996

pjsingal force-pushed the burkelabClean branch 2 times, most recently from ee0a85f to 780c996 Compare September 5, 2024 17:27

pjsingal added 21 commits September 5, 2024 13:28

added description to alzueta.yaml

1db9b11

fixed whitespace errors in examples

2d861e9

fixed keyword errors in examples

c513999

added titles to each example

d441a43

removed sys.path.append from examples

27ec8dd

replaced NH3/H2 with pure NH3 to simplify fspeed example

3f76be8

updated keywords of IDT and JSR example

4b50e65

changed sec header to "linear-burke" in rate-constants.md

cd26c9a

removed junk comments from alzueta_LMRR.yaml

25a6a05

Added description of linear-burke YAML format

b702e3f

Added link to linear-burke YAML guide in rate-constants.md

4fb48ed

updated YAML description examples

632ecca

fixed formatting of yaml examples

cae5721

Added citation to Eq8 of paper

044cb2b

fixed formatting issues

3e57ba9

reordered yaml description

d355c9e

Revert "reordered yaml description"

b000624

This reverts commit d355c9e.

removed whitespace errors

d4373af

added to example-keywords.txt

bf1e7b3

updated keywords in examples

bd6f480

added singal2024 placeholder to cantera.bib and updated citation in L…

0eb6e58

…MRR ref materials

speth requested changes Sep 25, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Incorporating the LMR-R mixture rule into Cantera for the first time #1710

Incorporating the LMR-R mixture rule into Cantera for the first time #1710

pjsingal commented Jun 12, 2024 •

edited

Loading

codecov bot commented Jun 12, 2024 •

edited

Loading

speth left a comment

speth Sep 17, 2024

speth Sep 17, 2024

speth Sep 17, 2024

speth Sep 17, 2024

speth Sep 17, 2024

speth Sep 25, 2024

speth Sep 25, 2024

speth Sep 25, 2024

speth Sep 25, 2024

speth Sep 25, 2024

Incorporating the LMR-R mixture rule into Cantera for the first time #1710

Are you sure you want to change the base?

Incorporating the LMR-R mixture rule into Cantera for the first time #1710

Conversation

pjsingal commented Jun 12, 2024 • edited Loading

codecov bot commented Jun 12, 2024 • edited Loading

Codecov Report

speth left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pjsingal commented Jun 12, 2024 •

edited

Loading

codecov bot commented Jun 12, 2024 •

edited

Loading