Browsing by Author "Karakus, Selcan"

Now showing 1 - 2 of 2

Handling Uncertainty in Rheological Properties of Green Eggshell Nanocomposites by a Fuzzy-Hybrid Modeling Approach: A Comparative Study
(IOP Publishing Ltd, 2023) Insel, Mert Akin; Karakus, Selcan; Temelcan, Gizem; Kocken, Hale Gonce; Albayrak, Inci
Fuzzy regression is an attractive application due to its ability to solve problems in the absence of accurate mathematical models. In this study, rheological properties of a bio-nanocomposite were predicted using fuzzy regression. The aim of study is to introduce the methodology and application of fuzziness in detail for handling uncertainty in the experimental data. In this context, rheological properties of green eggshell bio-nanocomposites are investigated using different operating parameters. Since the determination of surface tension is prone to experimental errors, a novel fuzzy-hybrid model is proposed in addition to a crisp model. Statistical error analysis is conducted to assess the validity of the multivariate models proposed. While both models provided satisfactory results, it is observed that the fuzzy-hybrid model outperforms the crisp model in both accuracy and generality. This study highlights that the proposed models have an impressive potential to be used for modeling experimental studies in nanotechnological manufacturing applications.
On Modelling of Surface Tension of Cmc-Α Nanoparticles by Fuzzy-Hybrid Approach: A Comparison Study
(Wiley, 2023) Kocken, Hale Gonce; Insel, Mert Akin; Temelcan, Gizem; Karakus, Selcan; Albayrak, Inci
Surface tension is one of the most important rheological parameters of nanoliquids. It influences the thermophysical and mass transfer properties of nanostructures. Accurate estimation of the surface tension from operating variables is critical for determining optimal production processes. However, the challenges of producing nanoparticles and measuring their properties introduce experimental errors in the data used for mathematical modelling. Crisp regression approaches provide adequate representation of the data, but they do not provide information about the experimental uncertainty. In this study, a fuzzy-hybrid approach is proposed for mathematical modelling of surface tension of carboxymethyl cellulose/chitosan-alpha-Fe2O3 nanoparticles. Then, the proposed model is compared with a crisp model from a previous study. Error analysis is conducted to validate the constructed fuzzy model. It is observed that the fuzzy-hybrid modelling approach has yielded significantly lower error values (a 60%-90% improvement in all error metrics on average), and thus, it is superior to the crisp approach. This study contributes to the subject of modelling rheological properties. It is shown that the fuzzy-hybrid approach has impressive potential to be utilized for modelling the rheological properties of nanostructures.