Slide 1

Chaughule, Thorat - Statistical Analysis/Design of Experiments

Figure 8. 7. Combined effect of air flow and air temperature on Carr Index (Feed flow –

3.25 ml/min; atomization pressure – 1.88 kg/cm2)

{Adapted from Jangam and Thorat, Optimization of spray drying of ginger extract ac-

cepted for publication in Drying Technology, with permission}

8.9.1.2. Studies on Solar Cabinet Drying of Green Peas (Pisum sativum)

Jadhav et al, (2010) have optimized the pretreatment conditions for green peas prior
to solar cabinet drying. Response surface methodology having a 22 factorial experimen-
tal design with four axial and five central points leading to 13 experiments were per-
formed for the optimization of blanching time in hot water (96°C for 1 - 5 min) and KMS
concentration (0.2-0.5%). Color and hardness of the dehydrated green peas were the
responses. RSM was applied to the experimental data using a commercial statistical
package, Design Expert version 7.0. The details about coded and actual levels of va-
riables can be found elsewhere (Jadhav et al., 2010). RSM has given different combina-
tions of different levels of the two factors studied. The optimum conditions of pretreat-
ment before solar cabinet drying were reported as 4.24-min blanching time and 0.49%
KMS concentration, resulting in 345.38 g hardness and 19.92mg/100 g color (Jadhav et
al, 2010).

CLOSING REMARKS

The application of statistical designs for optimization of drying processes and their
utility in drying R & D is still in its early years. An attempt is made here to compile statis-
tical optimization designs for potential relevance in food processing R & D.

Optimization using factorial design is a rigorous and simple method to find the ade-
quate experimental conditions to produce the best response of the drying system. Fac-
torial designs can be used for screening purposes to identify the factors affecting the se-
lected response and as a tool to explore and model this response as a function of these
significant experimental factors. The use of multilevel designs allows an efficient explo-
ration of the response in the experimental region and the estimation and optimization of

Carr Index (

• )^