EXPERIMENTAL STUDIES AND ARTIFICIAL NEURAL NETWORKS (ANN) MODELING OF MOISTURE ABSORPTION CHARACTERISTICS OF POLYESTER/MOMODICALFIBRE REINFORCED COMPOSITE

Adeyi, Abiola and Durowoju, Mondiu and Adeyi, Oladayo (2018) EXPERIMENTAL STUDIES AND ARTIFICIAL NEURAL NETWORKS (ANN) MODELING OF MOISTURE ABSORPTION CHARACTERISTICS OF POLYESTER/MOMODICALFIBRE REINFORCED COMPOSITE. International Journal of Mechanical Engineering andTechnology (IJMET), 9 (11). pp. 1453-1467. ISSN 0976-6359

Text (Full document) IJMET_09_11_150.pdf - Published Version Download (659kB)

Abstract

Nowadays, there are interests around the world concerning the applicability of natural fibers in polymer composite development fortechnological advancement. Natural fiber inclusion in synthetic polymer is termed partially degradable polymer composite which is considered environmentally friendly and acceptable. However, a notable deficiency is their poor moisture resistance behaviour that degrades their mechanical properties over time. This work therefore investigated the effect of Momodical fiber fractions on the moisture absorption properties ofPolyester/Momodical fiber reinforced composite. The Momodical fibers were alkali treatedto improve the fiber properties. The Polyester/Momodical fiber composite were developed by incorporating the alkali treated Momodical fibers in the weight fraction of 10, 20, 30 and 40 % in polyester resin. Water immersion test was used to evaluate the water absorption characteristics from which the water diffusion mechanisms of the developed composites were established. For the sake of system behaviour prediction and control, Artificial Neural Networks (ANN) was used for modeling and prediction of the moisture gain ofthe Polyester/Momodical reinforced composites. Scanning Electron Microscope (SEM) was used to elucidate the morphology of raw and alkali treated Momodical fiber. The results showed that the water absorption process was diffusion controlled and diffusion mechanisms cut across less Fickian / Fickian behavior for investigated composites. The composite weight gained and percentage water absorption increased with increased immersion time and fiber loading. The moisture diffusivity ranged from 1.98 E – 12 to5.38 E - 12. The ANN structure 2-5-1-1 developed using 'tansig' 'tansig' 'purelin'transfer function showed a high capability and reliability in modeling and prediction of moisture gain observed in the developed composites. The results suggested an outdoor application in desert cooler pad and built material development where the moisture diffusion tendencies could be beneficial and loss of mechanical strength is trivial.

Item Type:	Article
Subjects:	T Technology > TP Chemical technology
Depositing User:	Dr Oladayo Adeyi
Date Deposited:	07 Dec 2018 03:56
Last Modified:	07 Dec 2018 03:56
URI:	https://eprints.lmu.edu.ng/id/eprint/1685

Actions (login required)

View Item