Speaker
Prof.
Antoine F. Mulaba - Bafubiandi
(University of Johannesburg)
Description
The garification process is a food preparation method which depends on process factors such as cassava age, fermentation days and temperatures. These process parameters affect the yield and quality of the gari especially during roasting (garification). This paper report on findings from a study where the convective heat transfer coefficients during garification process were examined. The material losses and yield of gari sample obtained, at different processing stages, from cassava roots of different ages, on fermentation days and at varied garification temperatures and time were evaluated. The effects of cassava processing variables [cassava ages (9, 12 and 15 months), fermentation days (0 - 6 days), garification temperatures (100, 120 and 140oC and time, (t) (0 – 21minutes)] on the thermo-physical properties of the product were determined using standard laboratory methods. The temperature changes of garification process were monitored, under natural and forced convection. The dimensionless numbers associated with convective heat transfer coefficients were estimated. Artificial Neural Networks (ANN) model was employed in this study. The feed forward network structure with input, hidden layer(s) and back propagation network algorithm was utilized in model training. A log sigmoid transfer function was used for input layers while pure-line transfer function was used for output layers for the modeling of the processes. Results showed that mean garification conversion rate achieved was 0.22(wt/wt). Cassava roots of 15 months age of maturity produced higher yields of gari. The dimensionless numbers obtained for garification process predicted optimally (R2>0.9) the relationship between momentum and heat transfer by diffusion. These were used to determine the magnitude of convective and evaporative effects at the surface. The estimated values of convective heat coefficients for the garification process ranged from 4.92 to 38.62 W/m2 oC. Empirical equations developed for heat transfer coefficients [hc=0.017t²-0.388t+3.039] with (R2>0.9) were best described by polynomial relationships and the effectiveness of these results were validated using the ANN model, with mean error of less than 10%. The optimum ANN model produces convective heat transfer coefficients with two hidden layers and twenty five neurons in each hidden layer, with mean square error, mean absolute error and sum square error of 0.000016, 0.0029 and 0.0085%, respectively, with R2 of 0.992. The study showed that optimization of the combined effects of the cassava processing variables gave higher yield and good quality gari, which was achieved using 15 months cassava, fermented for three days with garification temperature and time of 120oC and 15 minutes, respectively.
Level for award<br> (Hons, MSc, <br> PhD, N/A)?
N/A
Apply to be<br> considered for a student <br> award (Yes / No)?
No
Primary author
Prof.
Antoine F. Mulaba - Bafubiandi
(University of Johannesburg)
Co-author
Dr
Sunday S. Sobowale
(University of Johanensburg)
