dc.contributor.author | El Jery, Atef | |
dc.contributor.author | Ramírez Coronel, Andrés Alexis | |
dc.contributor.author | Orosco Gavilán, Juan Carlos | |
dc.contributor.author | Al Ansari, Nadhir | |
dc.contributor.author | Sh Sammen, Saad | |
dc.date.accessioned | 2023-10-25T00:27:51Z | |
dc.date.available | 2023-10-25T00:27:51Z | |
dc.date.issued | 2023-04-15 | |
dc.identifier.citation | El Jery, A., Ramírez, A. A., Orosco, J. C., Al, N., & Sh Sammen, S. (2023). Proposing empirical correlations and optimization of Nu and Sgen of nanofluids in channels and predicting them using artificial neural network. Case Studies in Thermal Engineering, 45(2023), 102970. https://doi.org/10.1016/j.csite.2023.102970 | es_PE |
dc.identifier.other | . | es_PE |
dc.identifier.uri | https://hdl.handle.net/11537/34764 | |
dc.description.abstract | Getting the best performance from a thermal system requires two fundamental analyses, energy
and entropy generation. An ideal mechanism has the highest Nu and the lowest entropy Sgen. As
part of this research, a large dataset of fluid flow via tubes has been collected experimentally. As
well as the inclusion of nanoparticles, analyses are included as well. By using deep learning algorithms, the Nusselt number and total entropy generation are predicted. In both models, the
mean absolute error was lower than 5%. To determine the most accurate model, hyperparameter
tuning is performed. That is adjusting all the settings in the neural network to attain the best
results. The results of the predictive models are compared against experimental and benchmark
results. The study incorporates a massive optimization strategy to fine-tune the predictive capabilities of the models. Furthermore, the model’s predictive abilities are evaluated through the
use of the coefficient of determination R2. For water and nanofluids flowing through circular,
square, and rectangular cross-sections, the proposed models can predict Nu and Sgen. The results
showed remarkable agreement with the experimental results. The models showed an MAE of not
higher than 1.33%, which is a great achievement. Also, empirical correlations are proposed for
both parameters, and double factorial optimization is implemented. The results showed that to
achieve the best results, the Re should be higher than 1600, and the nanoparticle concentration
should be 3%. A thorough justification of selected cases is presented as well. | es_PE |
dc.format | application/pdf | es_PE |
dc.language.iso | spa | es_PE |
dc.publisher | Elsevier Ltd | es_PE |
dc.rights | info:eu-repo/semantics/openAccess | es_PE |
dc.rights | Atribución-NoComercial-CompartirIgual 3.0 Estados Unidos de América | * |
dc.rights.uri | https://creativecommons.org/licenses/by-nc-sa/3.0/us/ | * |
dc.source | Universidad Privada del Norte | es_PE |
dc.source | Repositorio Institucional - UPN | es_PE |
dc.subject | Flujo de fluidos | es_PE |
dc.subject | Nanofluidos | es_PE |
dc.subject | Nanopartículas | es_PE |
dc.title | Proposing empirical correlations and optimization of Nu and Sgen of nanofluids in channels and predicting them using artificial neural network | es_PE |
dc.type | info:eu-repo/semantics/article | es_PE |
dc.publisher.country | SA | es_PE |
dc.identifier.journal | Case Studies in Thermal Engineering | es_PE |
dc.subject.ocde | https://purl.org/pe-repo/ocde/ford#3.01.00 | es_PE |
dc.description.sede | Sede virtual | es_PE |
dc.identifier.doi | https://doi.org/10.1016/j.csite.2023.102970 | |