Experimental Investigation of the Absorption Behavior of Date Palm Fiber Reinforced Iso-Polyester Composites: Artificial Neuron Network (ANN) Modeling
Riyadh Benyettou, Salah Amroune, Mohamed Slamani, Yasemin Seki, Alain Dufresne
Abstract
The present article attempts to study absorption properties of bio-composites reinforced with date palm fibers. The effect of fiber loading on water absorption at room temperature 25°C was investigated. The weight gain was measured of bio-composites immersed in distilled water, seawater and rainwater, for more than 670 hours, until reaching the saturation with a measurement interval between 24 and 48 hours. To understand absorption phenomenon, scanning electron microscopy was used. Porosity rate was determined using image J software. It was noted the water absorption rate of the bio composites reached 16.20%, 16.33%, 21.94%, 41.99% for seawater, 16.41%, 16.52%, 20.84%, 30.08% for distilled water, and 14.00%, 14.04%, 19.30%, 36.94% for rainwater, respectively. The absorption increases when increasing fiber content. The diffusion coefficient of bio-composites has minimum and maximum values of about 1.94 × 10−6mm2/s and 3.99 × 10−6mm2/s, respectively. Palm fibers are highly porous. The porosity value was higher than 51%. To predict the absorption rate, artificial neural network method was used. The ANN models obtained are very well correlated with the experimental data where the values of the correlation coefficient of the datasets are all beyond 0.99 and the average error value was estimated at 3 × 10−5.