Results and Discussion
Fitting the response surface models: In this study, multiple
regression analyses were carried out using response surface
analysis to (1) determine regression coefficients and statistical
significance of model terms and fitting (2) the mathematical
models to the experimental data, aiming at and overall optimal
region for the response variables 31. Lasekan and Abbas 35
demonstrated that the response surface analysis allows the
development of an empirical relationship where each response
(Yi) was assessed as a function of time (X1), temperature (X2) and
loading capacity (X3) and predicted as the sum of constant (bo),
three first-order effects (linear terms in X1, X2 and X3), three
interaction effects (interactive terms in X1X2, X1X3 and X2X3) and
three second-order effects (quadratic terms in X1
2, X2
2 and X3
2).
The estimated regression coefficient of response surface models
with the corresponding R2 values and lack of fit test are reported
in Table 3. The R2 values for these response variables were between
0.315 and 0.928. It can be seen from Table 3 that the regression
models for the response variables were significant by the test at
the 5% confidence level (p