CEMFF: A force field database for cementitious materials including validations, applications and opportunities
Summary¶
Atomistic simulations of cementitious materials—C–S–H, portlandite, C\(_3\)S, and related phases—are scattered across dozens of force-field families with inconsistent validation. This Cement and Concrete Research review introduces CEMFF, a web database cataloging atomistic potentials applied to cementitious minerals, summarizing functional forms, benchmarks, and validation against experiment and quantum references for each entry. The scope spans Born–Mayer–Huggins, ClayFF, IFF, ReaxFF, UFF, and specialized cement parameterizations. Adri C. T. van Duin coauthors alongside an international consortium, positioning reactive options where bond-making/breaking is essential. The local pdf_path is a proof PDF; confirm pagination against the final issue when citing page-level details.
Methods¶
Force-field training / fitting. Same cementitious force-field survey as [[2017mishra-cement-and-c-cemff-force]] (DOI 10.1016/j.cemconres.2017.09.003); this slug tracks proof PDF bytes only.
MD application (atomistic dynamics). Literature-level summary of how surveyed groups run classical and reactive MD for cement phases; no new unified MD dataset accompanies the review text.
Static QM / DFT. DFT and broader QM benchmarks appear as reference data in the surveyed fitting stories summarized in the tables.
Review / database framing. CEMFF registry narrative (http://cemff.epfl.ch) plus community upload guidance; prefer [[2017mishra-cement-and-c-cemff-force]] when citing VOR pagination. The abstract inventories Born–Mayer–Huggins, IFF, ClayFF, C–S–H FF, CementFF, GULP-style ionics, ReaxFF, and UFF as representative families applied to C₃S, portlandite, and C–S–H models.
Findings¶
The review highlights diversity of force-field choices and validation gaps when moving potentials between chemistries, water models, and interface configurations. It explicitly positions ReaxFF among options when reactive chemistry cannot be neglected, while warning that reactive cost and training scope must match the modeling question. The overarching message is pragmatic: practitioners need property-targeted selection guidance rather than a single universal cement force field. For knowledge-base maintenance, treat CEMFF as a living inventory: when new cementitious parameterizations appear in the literature, the database entry (not this static article alone) may update faster than the review text, so verify current entries at the EPFL site when planning simulations.
Limitations¶
Proof ingest may differ in layout from the final article; the database evolves independently of the paper’s static snapshot.
Reader notes (MAS / retrieval)¶
Database-first queries should still verify CEMFF entries at the live site; treat this review as narrative context, not a live parameter dump.
Relevance to group¶
Situates van Duin reactive cement-related developments within community-wide FF governance and tooling.
Citations and evidence anchors¶
- DOI:
https://doi.org/10.1016/j.cemconres.2017.09.003(papers/Mishra_CEMFF_proof_2017.pdf).