Effet du salage et du séchage sur la dynamique d’évolution de la protéolyse, de la structure et de la texture lors de la fabrication d’un jambon sec. Développement d’un modèle de « jambon numérique » couplant transferts d’eau, de sel et protéolyse

摘要

Because of public health problems, the food industry must lower sodium content in all food products, therefore in cured meat products. During the dry-cured ham elaboration process, decreasing salt content may induce microbial safety problems and texture defects due to an excessive proteolysis that could affect later the industrial stage of slicing. On account of that, this work of thesis aims at (1) studying the relationship between proteolysis, structure and texture during the various stages of dry-cured ham manufacture, and (2) building a “numerical ham” model to predict spatially the time course of water and salt content, and thus water activity (a w ), and to couple these variations with proteolysis. This work combines experimental studies and numerical modelling and simulation. Firstly, a new and powerful technique for quantifying proteolysis that uses “Fluorescamine” was developed and validated on pork meat samples and samples extracted from industrial dry-cured hams; a new proteolysis index (PI) was defined. Based on an experimental design, the time course of proteolysis was quantified in laboratory-salted and dried pork meat samples prepared from five different types of pork muscle. Applying multiple linear regression enabled us to build phenomenological models relating, for each pork muscle, PI velocity to temperature, and to water and salt content. Using Comsol ® Multiphysics software, these phenomenological models were then combined with heat and mass transfer models and associated with calculation of a w , thus constituting the “numerical ham” model. In addition, the time course of PI, five textural parameters (hardness, fragility, cohesiveness, springiness and adhesiveness), and four structural parameters (fiber number, extracellular spaces, cross section area, and connective tissue area) was quantified on samples extracted from two different muscles of industrial dry-cured hams removed from the process at five different processing times. Multiple polynomial regression was applied to build phenomenological models relating PI, salt and water content to some textural and structural parameters investigated. These last models could be rapidly incorporated in the “numerical ham” model to constitute a real process simulator. In the future, the “numerical ham” model should be improved in order to take into account (1) the strong decrease in ham volume due to drying and also (2) the decrease in proteolysis velocity with time as a result of the reduction in the amount of protein that can be hydrolysed in the ham. Once completed and improved, the process simulator will be available to professionals to test scenarios allowing sodium content to be reduced in dry-cured hams without altering their final quality.