Fix discontinuities in NASA polynomials in thermo libraries

.github/workflows/CI.yml

@@ -37,7 +37,7 @@ concurrency:
 env:
   # if running on RMG-Database but requiring changes on an un-merged branch of RMG-Py, replace
   # main with the name of the branch
-  RMG_PY_BRANCH: main
+  RMG_PY_BRANCH: thermodisc
   # RMS branch to use for ReactionMechanismSimulator installation
   RMS_BRANCH: for_rmg
   # RMS mode used for install_rms.sh

input/thermo/libraries/2-BTP.py

@@ -3974,22 +3974,31 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.13733,0.0264862,-9.05687e-06,-5.53864e-19,2.12819e-22,15710.9,13.5294], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[9.03381,0.00821245,7.1754e-06,-5.88343e-09,1.03439e-12,14335.1,-20.9858], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(
+                coeffs=[5.39212, 0.00298346, 5.22542e-05, -6.64727e-08, 2.56305e-11, 15362.7, -0.196129],
+                Tmin=(200, "K"),
+                Tmax=(1000, "K"),
+            ),
+            NASAPolynomial(
+                coeffs=[7.26055, 0.0180161, -6.47062e-06, 1.04411e-09, -6.24741e-14, 13812.3, -14.8554],
+                Tmin=(1000, "K"),
+                Tmax=(6000, "K"),
+            ),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
-        E0 = (129.221,'kJ/mol'),
-        Cp0 = (33.2579,'J/mol/K'),
-        CpInf = (224.491,'J/mol/K'),
+        Tmin = (200,'K'),
+        Tmax = (6000,'K'),
     ),
-    shortDesc = """A 8/83""",
+    shortDesc = """A 8/83 for H298 & S298. Cp(T) from Burcat 2005""",
     longDesc = 
 """
-A 8/83
-_low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
 CC#CC
-_imported from 2-BTP/2-BTP_thermo.txt.
+Originally imported from 2-BTP/2-BTP_thermo.txt.
+H298 and S298 values from USC-Mech-ii,JetSurF,AramcoMech1.3,etc., including 2-BTP model.
+(Original source most likely estimated by Colkett, August 1983)
+Cp from Burcat 2005, because published polynomials inconsistent and discontinuous.
+New polynomials change G from original values by 1-3 kcal/mol at 1000-2000K
+and very little below 1000K. 
+See https://github.com/ReactionMechanismGenerator/RMG-database/pull/738
 """,
 )
 
@@ -4702,20 +4711,23 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.96963,0.0244422,-9.12514e-06,-4.24669e-18,1.63047e-21,35503.3,12.0361], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[14.5382,-0.00856771,2.35595e-05,-1.36764e-08,2.44369e-12,33259.1,-45.3695], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(coeffs=[2.31011,0.0283747,-1.63837e-05,4.46252e-09,-4.30511e-13,35584.3,14.9106], Tmin=(200,'K'), Tmax=(1385,'K')),
+            NASAPolynomial(coeffs=[10.3231,0.0117626,-4.00005e-06,6.18728e-10,-3.58084e-14,32586.1,-28.8794], Tmin=(1385,'K'), Tmax=(5000,'K')),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
-        E0 = (297.026,'kJ/mol'),
-        Cp0 = (33.2579,'J/mol/K'),
-        CpInf = (199.547,'J/mol/K'),
+        Tmin = (200,'K'),
+        Tmax = (5000,'K'),
     ),
-    shortDesc = """H6W/94""",
+    shortDesc = """H6W/94 from Curran's 2016 Pentane model""",
     longDesc = 
 """
 H6W/94
-_low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
+The original NASA parameters from this source had a discontinuity in the polynomials, causing errors in Cantera.
+The parameters from "Curran Pentane" libary are used here instead. The match below 1000K is identical.
+It even has the same attribution (`H6W/94`). Above 1000K the deltaG differs by 2kcal/mol up to 6kcal/mol at 2000K
+Its full source is "An ignition delay time and chemical kinetic modeling study of the pentane isomers"
+John Bugler, Brandon Marks, Olivier Mathieu, Rachel Archuleta, Alejandro Camou, Claire Gregoire, Karl A. Heufer, Eric L. Petersen, Henry J. Curran
+Combustion and Flame, 2016, 163, 138-156  https://doi.org/10.1016/j.combustflame.2015.09.014
+
 [CH2]C#CC
 _imported from 2-BTP/2-BTP_thermo.txt.
 """,

input/thermo/libraries/Chlorination.py

@@ -1136,13 +1136,13 @@
 """,
     thermo = NASA(
         polynomials=[
-            NASAPolynomial(coeffs=[3.54650,0.00233217,0.523315e-05,-0.973660e-08,0.446729e-11,-0.102996e+05,7.39746e+01],
+            NASAPolynomial(coeffs=[3.54650,0.00233217,0.523315e-05,-0.973660e-08,0.446729e-11,-0.102996e+05,7.3974601],
             Tmin=(200,'K'), Tmax=(1000,'K')),
             NASAPolynomial(coeffs=[4.36649,0.00205137,-0.670877E-06, 0.10131893E-09, -0.57791828E-14, -0.10576070E+05, 0.28049555E+01],
             Tmin=(1000,'K'), Tmax=(6000,'K'))],
         Tmin=(200,'K'), Tmax=(6000,'K')
         ),
-    shortDesc = u"""Gurvich 89 """,
+    shortDesc = u"""Gurvich 89""",
     longDesc =
 u"""
 See http://www.ipd.anl.gov/anlpubs/2005/07/53802.pdf

input/thermo/libraries/JetSurF1.0.py

@@ -1700,16 +1700,22 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.96963,0.0244422,-9.12514e-06,-4.24669e-18,1.63047e-21,35503.3,12.0361], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[14.5382,-0.00856771,2.35595e-05,-1.36764e-08,2.44369e-12,33259.1,-45.3695], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(coeffs=[2.31011,0.0283747,-1.63837e-05,4.46252e-09,-4.30511e-13,35584.3,14.9106], Tmin=(200,'K'), Tmax=(1385,'K')),
+            NASAPolynomial(coeffs=[10.3231,0.0117626,-4.00005e-06,6.18728e-10,-3.58084e-14,32586.1,-28.8794], Tmin=(1385,'K'), Tmax=(5000,'K')),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (5000,'K'),
     ),
-    shortDesc = u"""H6W/94""",
+    shortDesc = """H6W/94 from Curran's 2016 Pentane model""",
     longDesc = 
-u"""
-
+"""
+H6W/94
+The original NASA parameters from this source had a discontinuity in the polynomials, causing errors in Cantera.
+The parameters from "Curran Pentane" libary are used here instead. The match below 1000K is identical.
+It even has the same attribution (`H6W/94`). Above 1000K the deltaG differs by 2kcal/mol up to 6kcal/mol at 2000K
+Its full source is "An ignition delay time and chemical kinetic modeling study of the pentane isomers"
+John Bugler, Brandon Marks, Olivier Mathieu, Rachel Archuleta, Alejandro Camou, Claire Gregoire, Karl A. Heufer, Eric L. Petersen, Henry J. Curran
+Combustion and Flame, 2016, 163, 138-156  https://doi.org/10.1016/j.combustflame.2015.09.014
 """,
 )
 
@@ -1824,16 +1830,29 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.13733,0.0264862,-9.05687e-06,-5.53864e-19,2.12819e-22,15710.9,13.5294], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[9.03381,0.00821245,7.1754e-06,-5.88343e-09,1.03439e-12,14335.1,-20.9858], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(
+                coeffs=[5.39212, 0.00298346, 5.22542e-05, -6.64727e-08, 2.56305e-11, 15362.7, -0.196129],
+                Tmin=(200, "K"),
+                Tmax=(1000, "K"),
+            ),
+            NASAPolynomial(
+                coeffs=[7.26055, 0.0180161, -6.47062e-06, 1.04411e-09, -6.24741e-14, 13812.3, -14.8554],
+                Tmin=(1000, "K"),
+                Tmax=(6000, "K"),
+            ),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (6000,'K'),
     ),
-    shortDesc = u"""A 8/83""",
+    shortDesc = """A 8/83 for H298 & S298. Cp(T) from Burcat 2005""",
     longDesc = 
-u"""
-
+"""
+CC#CC
+H298 and S298 values from USC-Mech-ii,JetSurF,AramcoMech1.3,etc., including 2-BTP model.
+(Original source most likely estimated by Colkett, August 1983)
+Cp from Burcat 2005, because published polynomials inconsistent and discontinuous.
+New polynomials change G from original values by 1-3 kcal/mol at 1000-2000K and very little below 1000K. 
+See https://github.com/ReactionMechanismGenerator/RMG-database/pull/738
 """,
 )
 

input/thermo/libraries/JetSurF2.0.py

@@ -3873,18 +3873,30 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.13733,0.0264862,-9.05687e-06,-5.53864e-19,2.12819e-22,15710.9,13.5294], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[9.03381,0.00821245,7.1754e-06,-5.88343e-09,1.03439e-12,14335.1,-20.9858], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(
+                coeffs=[5.39212, 0.00298346, 5.22542e-05, -6.64727e-08, 2.56305e-11, 15362.7, -0.196129],
+                Tmin=(200, "K"),
+                Tmax=(1000, "K"),
+            ),
+            NASAPolynomial(
+                coeffs=[7.26055, 0.0180161, -6.47062e-06, 1.04411e-09, -6.24741e-14, 13812.3, -14.8554],
+                Tmin=(1000, "K"),
+                Tmax=(6000, "K"),
+            ),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (6000,'K'),
     ),
-    shortDesc = u"""A 8/83""",
+    shortDesc = """A 8/83 for H298 & S298. Cp(T) from Burcat 2005""",
     longDesc = 
-u"""
-A 8/83
-Low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
+"""
 CC#CC
+Originally imported from JetSurF2.0
+H298 and S298 values from USC-Mech-ii,JetSurF,AramcoMech1.3,etc., including 2-BTP model.
+(Original source most likely estimated by Colkett, August 1983)
+Cp from Burcat 2005, because published polynomials inconsistent and discontinuous.
+New polynomials change G from original values by 1-3 kcal/mol at 1000-2000K and very little below 1000K. 
+See https://github.com/ReactionMechanismGenerator/RMG-database/pull/738
 """,
 )
 
@@ -5886,17 +5898,22 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.96963,0.0244422,-9.12514e-06,-4.24669e-18,1.63047e-21,35503.3,12.0361], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[14.5382,-0.00856771,2.35595e-05,-1.36764e-08,2.44369e-12,33259.1,-45.3695], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(coeffs=[2.31011,0.0283747,-1.63837e-05,4.46252e-09,-4.30511e-13,35584.3,14.9106], Tmin=(200,'K'), Tmax=(1385,'K')),
+            NASAPolynomial(coeffs=[10.3231,0.0117626,-4.00005e-06,6.18728e-10,-3.58084e-14,32586.1,-28.8794], Tmin=(1385,'K'), Tmax=(5000,'K')),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (5000,'K'),
     ),
-    shortDesc = u"""H6W/94""",
+    shortDesc = """H6W/94 from Curran's 2016 Pentane model""",
     longDesc = 
-u"""
+"""
 H6W/94
-Low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
+The original NASA parameters from this source had a discontinuity in the polynomials, causing errors in Cantera.
+The parameters from "Curran Pentane" libary are used here instead. The match below 1000K is identical.
+It even has the same attribution (`H6W/94`). Above 1000K the deltaG differs by 2kcal/mol up to 6kcal/mol at 2000K
+Its full source is "An ignition delay time and chemical kinetic modeling study of the pentane isomers"
+John Bugler, Brandon Marks, Olivier Mathieu, Rachel Archuleta, Alejandro Camou, Claire Gregoire, Karl A. Heufer, Eric L. Petersen, Henry J. Curran
+Combustion and Flame, 2016, 163, 138-156  https://doi.org/10.1016/j.combustflame.2015.09.014
 [CH2]C#CC
 """,
 )

input/thermo/libraries/Klippenstein_Glarborg2016.py

@@ -1209,7 +1209,7 @@
         Tmin = (200,'K'),
         Tmax = (6000,'K'),
     ),
-    shortDesc = u"""FORMIC ACID A 5/14""",
+    shortDesc = u"""FORMIC ACID L 8/88""",
     longDesc = 
 u"""
 FORMIC ACID A 5/14
@@ -1221,6 +1221,10 @@
 equil. mix.  REF=ATcT C   {HF298=-378.6 kJ  REF=CHAO & ZWOLINSKI JPCRD 7.(1978),
 363   HF298=-363.9 kJ  REF=Burcat G3B3 calc}  Max Lst Sq Error Cp @6000 K 0.47%
 The polynomials were adjusted by B Ruscic..
+
+The above comment, from Burcat 2014, was used by Klippenstein and Glarborg in their 2016 work.
+But it has a discontinuity in the polynomials.
+We have reverted to the Burcat 2010 values, which omit the "The polynomials were adjusted by B Ruscic" comment.
 O=CO
 """,
 )

input/thermo/libraries/NOx2018.py

@@ -1337,19 +1337,23 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[3.89836,-0.00355878,3.55205e-05,-4.385e-08,1.71078e-11,-46778.6,7.34954], Tmin=(200,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[4.61383,0.00644964,-2.29083e-06,3.6716e-10,-2.18737e-14,-45330.3,0.847884], Tmin=(1000,'K'), Tmax=(6000,'K')),
-        ],
+            NASAPolynomial(coeffs=[3.89836,-0.00355878,3.55205e-05,-4.385e-08,1.71078e-11,-46770.6,7.34954], Tmin=(200,'K'), Tmax=(1000,'K')),
+            NASAPolynomial(coeffs=[4.61383,0.00644964,-2.29083e-06,3.6716e-10,-2.18737e-14,-47514.8,0.847884], Tmin=(1000,'K'), Tmax=(6000,'K')),
+       ],
         Tmin = (200,'K'),
         Tmax = (6000,'K'),
     ),
-    shortDesc = u"""""",
+    shortDesc = u"""L 8/88""",
     longDesc = 
 u"""
-FORMIC ACID A 5/14
-Fabian WMF Janoschek R J Mol Struct THEOCHEM 2005, 713, 227?234
-CL Rasmussen J Hansen P Marshall P Glarborg Int J Chem Kinet 40 (2008) 454-480
-
+FORMIC ACID
+The values used in P. Glarborg, J.A. Miller, B. Ruscic, S.J. Klippenstein
+"Modeling nitrogen chemistry in combustion" Progress in Energy and Combustion Science
+Volume 67, July 2018, Pages 31-68 https://doi.org/10.1016/j.pecs.2018.01.002
+that this thermo library is based on, come from Burcat's 2014 compilation.
+But that has a discontinuity in the polynomials.
+See https://github.com/ReactionMechanismGenerator/RMG-database/issues/703
+We have reverted to the Burcat 2010 values.
 O=CO
 """,
 )

input/thermo/libraries/USC-Mech-ii.py

@@ -1850,17 +1850,22 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.96963,0.0244422,-9.12514e-06,-4.24669e-18,1.63047e-21,35503.3,12.0361], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[14.5382,-0.00856771,2.35595e-05,-1.36764e-08,2.44369e-12,33259.1,-45.3695], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(coeffs=[2.31011,0.0283747,-1.63837e-05,4.46252e-09,-4.30511e-13,35584.3,14.9106], Tmin=(200,'K'), Tmax=(1385,'K')),
+            NASAPolynomial(coeffs=[10.3231,0.0117626,-4.00005e-06,6.18728e-10,-3.58084e-14,32586.1,-28.8794], Tmin=(1385,'K'), Tmax=(5000,'K')),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (5000,'K'),
     ),
-    shortDesc = u"""H6W/94""",
+    shortDesc = """H6W/94 from Curran's 2016 Pentane model""",
     longDesc = 
-u"""
+"""
 H6W/94
-Low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
+The original NASA parameters from this source had a discontinuity in the polynomials, causing errors in Cantera.
+The parameters from "Curran Pentane" libary are used here instead. The match below 1000K is identical.
+It even has the same attribution (`H6W/94`). Above 1000K the deltaG differs by 2kcal/mol up to 6kcal/mol at 2000K
+Its full source is "An ignition delay time and chemical kinetic modeling study of the pentane isomers"
+John Bugler, Brandon Marks, Olivier Mathieu, Rachel Archuleta, Alejandro Camou, Claire Gregoire, Karl A. Heufer, Eric L. Petersen, Henry J. Curran
+Combustion and Flame, 2016, 163, 138-156  https://doi.org/10.1016/j.combustflame.2015.09.014
 [CH2]C#CC
 Imported from USC-Mech ii thermdat.txt.
 """,
@@ -1986,19 +1991,30 @@
 """,
     thermo = NASA(
         polynomials = [
-            NASAPolynomial(coeffs=[2.13733,0.0264862,-9.05687e-06,-5.53864e-19,2.12819e-22,15710.9,13.5294], Tmin=(298,'K'), Tmax=(1000,'K')),
-            NASAPolynomial(coeffs=[9.03381,0.00821245,7.1754e-06,-5.88343e-09,1.03439e-12,14335.1,-20.9858], Tmin=(1000,'K'), Tmax=(3000,'K')),
+            NASAPolynomial(
+                coeffs=[5.39212, 0.00298346, 5.22542e-05, -6.64727e-08, 2.56305e-11, 15362.7, -0.196129],
+                Tmin=(200, "K"),
+                Tmax=(1000, "K"),
+            ),
+            NASAPolynomial(
+                coeffs=[7.26055, 0.0180161, -6.47062e-06, 1.04411e-09, -6.24741e-14, 13812.3, -14.8554],
+                Tmin=(1000, "K"),
+                Tmax=(6000, "K"),
+            ),
         ],
-        Tmin = (298,'K'),
-        Tmax = (3000,'K'),
+        Tmin = (200,'K'),
+        Tmax = (6000,'K'),
     ),
-    shortDesc = u"""A 8/83""",
+    shortDesc = """A 8/83 for H298 & S298. Cp(T) from Burcat 2005""",
     longDesc = 
-u"""
-A 8/83
-Low T polynomial Tmin changed from 300.0 to 298.0 K when importing to RMG.
+"""
 CC#CC
-Imported from USC-Mech ii thermdat.txt.
+Originally imported from USC-Mech ii.
+H298 and S298 values from USC-Mech-ii,JetSurF,AramcoMech1.3,etc., including 2-BTP model.
+(Original source most likely estimated by Colkett, August 1983)
+Cp from Burcat 2005, because published polynomials inconsistent and discontinuous.
+New polynomials change G from original values by 1-3 kcal/mol at 1000-2000K and very little below 1000K. 
+See https://github.com/ReactionMechanismGenerator/RMG-database/pull/738
 """,
 )