EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Deep learning based identification of solvents from post-desiccation patterns

Jesús Israel Morán Cortés, Felipe Pacheco Vázquez

Revista Mexicana de Física · 01.11.2026 · 72 (6) · 1-14

DOI: 10.31349/revmexfis.72.061701 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Revista Mexicana de Física, online erschienen am 01.11.2026. Verfasst von Jesús Israel Morán Cortés und 1 weiteren Autor:innen. Die Arbeit nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We introduce an optimized Deep learning protocol for fracture pattern classification to identify the solvent involved in the desiccation cracking process of starch-liquid slurries, even after its complete evaporation. For this purpose, image analysis techniques were used to characterize patterns obtained from drying suspensions using single solvents (water, ethanol, acetone) and two-component solvents (water-ethanol mixtures at different concentrations). Frequency histograms were generated based on nine morphological features, taking into account their size, shape, geometry and orientational ordering. Subsequently, we used these histograms as input data in different artificial neural network variants to determine which set of features leads to the highest accuracy in solvent identification. An average accuracy of up to 96(±1)% was achieved in the classification when all solvents were considered in the analysis. The highest accuracy was obtained with sets that include the crack area, AcrN , a feature barely considered in the literature. We found that AcrN can be related to the cohesive energy associated with hydrogen bond interactions. Taking this into account, we estimated some physicochemical properties of two-component solvents (specifically, the Hansen and Kamlet-Taft parameters) based solely on the features of patterns obtained from the individual components, achieving an error within 15% compared to the values determined from the literature. The proposed protocol could help to encode some implicit parameters and the optimal combination of features for pattern recognition in other fields of science and engineering.

Zitieren

APA

Cortés, J. I. M., & Vázquez, F. P. (2026). Deep learning based identification of solvents from post-desiccation patterns. Revista Mexicana de Física, 72(6), 1-14. https://doi.org/10.31349/revmexfis.72.061701

Vancouver

Cortés JIM, Vázquez FP. Deep learning based identification of solvents from post-desiccation patterns. Revista Mexicana de Física. 2026;72(6):1-14. doi:10.31349/revmexfis.72.061701

BibTeX
@article{Corts2026,
  author = {Cortés, Jesús Israel Morán and Vázquez, Felipe Pacheco},
  title = {Deep learning based identification of solvents from post-desiccation patterns},
  journal = {Revista Mexicana de Física},
  year = {2026},
  volume = {72},
  number = {6},
  pages = {1--14},
  doi = {10.31349/revmexfis.72.061701}
}
Kostenlos starten →