Effect of pretreatments on drying characteristics, rehydration properties, and total energy consumption of carrot slices: comparison between thin layer mathematical modelling and artificial neural network modelling

The study presents the effect of blanching, ultrasound, and ethanol pretreatments on the drying characteristics and rehydration properties of carrot slices, comparison of modelling, and total energy consumption. After the pretreatments, water and solid loss and weight reduction were determined at the samples. The highest water loss (%), solid loss (%), and weight reduction (%) were detected at the samples immersed in 100% ethanol solution for 30 min with the values of 16.45, 4.67, and 21.12, respectively. All pretreatments accelerated drying rate and improved the rehydration rate of the samples. Pretreatment time had also positive contribution to drying rate and rehydration rate. The highest rehydration rate was observed at the samples immersed in 100% ethanol solution for 30 min as 5.81 +/- 0.06. Additionally, moisture diffusion and mass transfer rate increased with the increment in drying rate. The highest moisture diffusivity (D) from Dincer and Dost model, effective moisture diffusivity (Deff) from Crank equation, and mass transfer coefficient (hm) values were calculated in the samples immersed in 100% ethanol solution for 30 min as 7.52 x 10-9 m2 s-1, 8.64 x 10-10 m2 s-1, and 8.15 x 10-7 m s-1, respectively. Pretreatments reduced the total energy consumption. The highest reduction rate was found in samples immersed in 100% ethanol solution for 30 min as 37.50%. Artificial neural network (ANN) modelling gave the best fitting with the highest determination coefficient (R2) values (0.9997-0.9999), lower the root mean square error (RMSE), and lower sum of square error (SSE) values (0.006740690-0.000900453 and 0.000818-0.00000162).