Scalability strategies for automated reaction mechanism generation

Vandewiele, Nick; Han, Kehang; Liu, Mengjie; Gao, Chloe Y.; Gillis, Ryan J.; Green, William

doi:10.1016/j.compchemeng.2019.106578

Cited by 11 publications

(17 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Up to four cores were utilized during reaction generation. Recent testing suggests the generation time could be cut by an order of magnitude using a modern supercomputer, whose larger memory allows RMG to use more cores in parallel while running on a single node …”

Section: Methodsmentioning

confidence: 99%

Detailed Reaction Mechanism for 350–400 °C Pyrolysis of an Alkane, Aromatic, and Long-Chain Alkylaromatic Mixture

2022

Self Cite

View full text Add to dashboard Cite

Many technologically important systems involve mixtures of fairly large molecules and relatively unselective chemistry, leading to complex product mixtures. These corresponding reaction networks are quite complex since each molecule in the feed can form many isometric intermediates and a variety of byproducts in addition to its major product. A variety of modeling methods have been developed to attempt to deal with this, but building accurate reaction mechanisms for these complicated systems is challenging, and the methodology is still under development. To showcase the advancements that have been made in automatic generation of large mechanisms, we constructed such a model for a three-component mixture containing species with up to 18 carbon atoms. The generated model is able to predict many of the major and minor products with relatively high accuracy against gold-tube batch pyrolysis data collected for this system. The high fidelity between the predicted species profiles and the experimental data is notable given the low temperature pyrolysis conditions studied, as any errors in ab initio rate parameters become more significant at lower temperatures.

show abstract

Section: Methodsmentioning

confidence: 99%

Detailed Reaction Mechanism for 350–400 °C Pyrolysis of an Alkane, Aromatic, and Long-Chain Alkylaromatic Mixture

2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…In RMG v3.0, parallelization has been completely revamped using the built-in multiprocessing module in Python, providing parallel processing support for reaction generation and quantum calculations for the QMTP (Quantum Mechanics for Thermochemical Properties) module. 39…”

Section: Parallel Computingmentioning

confidence: 99%

“…In RMG v3.0, parallelization has been implemented using the built-in multiprocessing module in Python, providing parallel processing support for reaction generation and quantum calculations for the QMTP (quantum mechanics for thermochemical properties) module. 44 Molecule Comparison. One task which can require substantial computing time in RMG is molecule comparison to determine if two RMG molecules are the same chemical species.…”

Section: ■ Performance Improvementmentioning

confidence: 99%

“…However, there are certain portions of the algorithm which are more amenable to parallelization. In RMG v3.0, parallelization has been implemented using the built-in multiprocessing module in Python, providing parallel processing support for reaction generation and quantum calculations for the QMTP (quantum mechanics for thermochemical properties) module …”

Section: Performance Improvementmentioning

confidence: 99%

See 1 more Smart Citation

Reaction Mechanism Generator v3.0: Advances in Automatic Mechanism Generation

Liu

Dana

Johnson

et al. 2021

J. Chem. Inf. Model.

Self Cite

200

216

View full text Add to dashboard Cite

In chemical kinetics research, kinetic models containing hundreds of species and tens of thousands of elementary reactions are commonly used to understand and predict the behavior of reactive chemical systems. Reaction Mechanism Generator (RMG) is a software suite developed to automatically generate such models by incorporating and extrapolating from a database of known thermochemical and kinetic parameters.Here, we present the recent version 3 release of RMG and highlight improvements since the previously published description of RMG v1.0. One important change is that RMG v3.0 is now Python 3 compatible, which supports the most up-to-date versions of cheminformatics and machine learning packages that RMG depends on. Additionally, RMG can now generate heterogeneous catalysis models, in addition to the previously available gas-and liquid-phase capabilities. For model analysis, new methods for local and global uncertainty analysis have been implemented to supplement first-order sensitivity analysis. The RMG database of thermochemical and kinetic parameters has been significantly expanded to cover more types of chemistry. The present release also includes parallelization for reaction generation and on-the-fly quantum calculations, and a new molecule isomorphism approach to improve computational performance. Overall, RMG v3.0 includes many changes which improve the accuracy of the generated chemical mechanisms and allow for exploration of a wider range of chemical systems.

show abstract

“…(If one considers conformational changes to be reactions, as in protein folding or some polymer problems, the complexity is orders of magnitude worse 54 15,55,56 and be slow enough that it needs parallelization 57 . This sort of problem is often encountered in systems where molecular weight growth (e.g., coke or soot formation, polymerization) is significant, because then the complexity of the molecules in the system, and the number of important reactions per molecule, increases as the reaction proceeds.…”

Section: Technical Issuesmentioning

confidence: 99%