A composition-based model for prediction of frozen muscle food quality

An effective approach for prediction of frozen food quality following the freezing process is needed to meet consumer demands for high quality frozen food products. The overall objective of this investigation was to develop and demonstrate a composition based model for prediction of frozen food quality, with specific emphasis on attributes impacted by ice crystals size. The proposed model would not require experimental thermophysical properties of the product, and could be applied to any product with a defined composition. A basic component of the model is the prediction of freezing temperature depression using mass fraction and molecular weight of components.Specific food components with significantly influence on freezing temperature depression of food products included low molecular weight water-soluble compounds with molality of 50 mumol/100g food or higher. Based on an analysis of two hundred high-moisture food products, nearly 45% of the experimental data were within an absolute difference (AD) of +/-0.15 °C and standard error (SE) of +/-0.65 °C, when compared to initial freezing temperatures predicted by the model. The predicted relationships between temperature and frozen water fraction, enthalpy, density and thermal conductivity for all analyzed food products were in close agreements to experimental data with +/-0.06 SE, +/-5.4 J/kg SE, +/-10 to +/-260 kg/m 3 AD and +/-0.15 W/m·°C SE, respectively.The predicted temperature distribution histories were successfully verified using experimental local characteristic freezing time data. The experimental and predicted local characteristic freezing times were in close agreements (+/-6.0 min SE). The ice crystal sizes in beef products, predicted based on unfrozen water fraction distribution histories were in closer agreement to the experimental data (e.g. +/-5.0 mum SE) than those predicted (e.g. +/-5.5 mum SE) based on temperature distribution histories. Food dimension, convective heat transfer coefficient and freezing medium temperature had significant influences on all the frozen food quality. Using the proposed composition-based model, the drip loss in a muscle food product was predicted. At a convective heat transfer coefficient of 300 W/m2°·C, freezing medium temperature of -40 °C, product thickness of 1 cm and initial temperature of 5 °C, the drip loss was predicted to be 3.4g exudates/100g food.