Exploratory Response Surface Analysis of Noni Fruit Juice Extraction Using a Hot Dry Steam System

Abstract

This study investigates whether temperature, pressure, and extraction time adequately explain the recovery of noni (

Morinda citrifolia

L.) juice using a Hot Dry Steam (HDS) system. Twenty observations generated from a central composite design (CCD) were analyzed using linear and full quadratic regression models. The response variable was treated as recorded mass because the source dataset reports values in kilograms (kg), although confirmation of the measurement unit is required. The observed responses ranged from 5.20 to 9.10 kg, with a mean of 7.1225 kg. The linear model was not statistically significant (F = 0.2757, p = 0.8420; R² = 0.0492; adjusted R² = −0.1291), and none of the individual factor coefficients were significant. The addition of three interaction and three quadratic terms did not significantly improve model fit (p = 0.6157). Similarly, the full quadratic model was not statistically significant (p = 0.7786; adjusted R² = −0.2398). Within the factorial design space, the linear model predicted a maximum response of 7.6049 kg at 80 °C, 2.0 bar, and 30 min, whereas the experimentally observed response under these conditions was 6.10 kg. This discrepancy, together with the lack of statistical significance and the boundary location of the predicted maximum, indicates that the prediction cannot be considered a validated optimum. Overall, the dataset supports exploratory assessment of the HDS process but does not provide sufficient evidence to establish optimal operating conditions or a comparative advantage of HDS over alternative extraction methods. Verification of operating and measurement protocols, improved control of feed variability, appropriate experimental replication, and independent validation are required before specific process conditions can be recommended.