Emprego da análise multivariada na classificação de polpas de frutas de acordo com as propriedades físicas e a difusividade térmica

Abstract

The production of frozen pulps is a favorable means for the full use of fruits, avoiding problems related to seasonality. Heat treatments are necessary to eliminate contaminating microorganisms, inactivate enzymes and slow down metabolic and microbiological processes under storage conditions. Thermophysical properties values of fruits have already been determined by several methods, with variations of temperature, soluble solids concentration and humidity, but these data are dispersed in the literature. The present work had the objective of using the multivariate statistical analysis to classify fruit pulps of a certain commercial brand according to the physical properties and the thermal diffusivity. The actual density was obtained through the use of pycnometer and analytical balance. The determination of the humidity was determined by Infra-red equipment. The value of soluble solids was determined by refractometry. The thermal diffusivity was obtained through the use of an aluminum tube, sensors and thermostatic bath. Principal Component Analysis (PCA) and Cluster Analysis (CA) were conducted using the Statistica 7.0 computational tool. The thermal diffusivity values varied from 2.63x10-7 m2.s-1 for orange pulp with 23.74 ° Brix, to 1.38x10-7 m2.s-1 for the grape pulp, with 14, 64 ° Brix. The work presented coherent answers with the literature and multivariate statistical analysis was useful in the classification of fruit pulps according to physical properties and thermal diffusivity. The orange pulp presented greater variability among the fruits studied and the pulp density was the parameter that contributed the most to the variability among them.