TY - JOUR
T1 - An efficient tool for modeling and predicting fluid flow in nanochannels
AU - Ahadian, Samad
AU - Mizuseki, Hiroshi
AU - Kawazoe, Yoshiyuki
N1 - Funding Information:
S. Ahadian greatly appreciates Dr. Dimitar I. Dimitrov, Professor Andrey Milchev, and Professor Kurt Binder. The authors sincerely appreciate the staff of the Center for Computational Materials Science of the Institute for Materials Research (IMR), Tohoku University, for its continuous support of the supercomputing facilities. This work was supported (in part) by the Japan Society for the Promotion of Science (JSPS).
PY - 2009
Y1 - 2009
N2 - Molecular dynamics simulations were performed to evaluate the penetration of two different fluids (i.e., a Lennard-Jones fluid and a polymer) through a designed nanochannel. For both fluids, the length of permeation as a function of time was recorded for various wall-fluid interactions. A novel methodology, namely, the artificial neural network (ANN) approach was then employed for modeling and prediction of the length of imbibition as a function of influencing parameters (i.e., time, the surface tension and the viscosity of fluids, and the wall-fluid interaction). It was demonstrated that the designed ANN is capable of modeling and predicting the length of penetration with superior accuracy. Moreover, the importance of variables in the designed ANN, i.e., time, the surface tension and the viscosity of fluids, and the wall-fluid interaction, was demonstrated with the aid of the so-called connection weight approach, by which all parameters are simultaneously considered. It was revealed that the wall-fluid interaction plays a significant role in such transport phenomena, namely, fluid flow in nanochannels.
AB - Molecular dynamics simulations were performed to evaluate the penetration of two different fluids (i.e., a Lennard-Jones fluid and a polymer) through a designed nanochannel. For both fluids, the length of permeation as a function of time was recorded for various wall-fluid interactions. A novel methodology, namely, the artificial neural network (ANN) approach was then employed for modeling and prediction of the length of imbibition as a function of influencing parameters (i.e., time, the surface tension and the viscosity of fluids, and the wall-fluid interaction). It was demonstrated that the designed ANN is capable of modeling and predicting the length of penetration with superior accuracy. Moreover, the importance of variables in the designed ANN, i.e., time, the surface tension and the viscosity of fluids, and the wall-fluid interaction, was demonstrated with the aid of the so-called connection weight approach, by which all parameters are simultaneously considered. It was revealed that the wall-fluid interaction plays a significant role in such transport phenomena, namely, fluid flow in nanochannels.
UR - http://www.scopus.com/inward/record.url?scp=72949095080&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=72949095080&partnerID=8YFLogxK
U2 - 10.1063/1.3253701
DO - 10.1063/1.3253701
M3 - Article
AN - SCOPUS:72949095080
SN - 0021-9606
VL - 131
JO - Journal of Chemical Physics
JF - Journal of Chemical Physics
IS - 18
M1 - 184506
ER -