PREDICTION MODELS FOR THE PERCENTAGE EXTRACT OF OXALIC ACID FROM THE BARK OF EUCALYPTUS CAMALDULENSIS

Abstract

The main aim of this research work was to use i t factorial analysis to predict the percentage extract of oxalic acid obtained from the bark of eucalyptus camaldulensis as a [unction of three operating parameters. Three 2 3 models were used and only two satisfy the G-test. The obtained equations are: First Model: Vel'sion l: y = 2.022 - 0.434xl - 0.411 X2 - 0.423x3 - 0.342xlx2 + 0.338x,x3 Vel'sion: 2: y = 2.022 + 0.434xl + 0.4llx2 + 0.423x3 - 0.342xJX2 + 0.338xJX3 Where Xl= mass, X2= size and X3= time. Second Model: y = l.327 + 0.260Xl - 0.438x3 - O.314xlx2 Where x,= concentration, X2= time and X3= size. It was found out that for the first model mass, time and size are inversely proportional to the percentage extract. Nevertheless, for the second model, concentration is directly proportional to the percentage extract while time and size are inversely proportional to the percentage extract oxalic acid obtained . . The two models that satisfy the G-test were adequate but the first model was more accurate than the second. During the course of the work, a computer program that can implement the model and prediction calculation was developed. The progr~m can then be used to get a model and predict the percentage extract [or three operating variables, which faUs within the range of the data used for the 2k designs. The program is an interactive Pascal program that can be used for 22, 23 and 24 factorial analyses