Sechage des levures en lit a jet conique : Experimentation et modelisation multi-echelles.

摘要

Drying is a unit operation that is practiced since thousands of years and which for scientific knowledge is still very incomplete. However it is more and more necessary to dry " well " in order to reach an optimal combination between drying duration, consumed energy and also final product quality. This is even truer in the case of food drying involving a lot of products that are very sensitive to temperature and dehydration. Baker's yeast often encounters a drastic decrease of its gassing power, which makes for instance inflate bread dough, when drying is operated in wrong conditions. Using special driers such as conical spouted beds should allow improving drying in terms of efficiency and final quality. The main objective of this thesis is to contribute to better understand and model Baker's yeast drying in a conical spouted bed by using a multiscale approach.;Drying of a single Baker's yeast pellet is characterised by using a new thermogravimetric set up where the evolution of the pellet dimensions and surface temperature can be measured during drying. The obtained results are then used to develop and validate a new model of the drying of a Baker's yeast pellet. The model is based on the existence of three types of water in yeast and it allows predicting the evolution of the pellet's moisture content and temperature during drying. The combined analysis of the model and the experimental results permits to put forward that pellet's shrinkage during drying have no significant influence on the drying rate and that it is really essential to correctly model the pellet's temperature.;Solid flow in the conical spouted bed is experimentally characterised by using a radioactive particle tracking technique (RPT). A post-treatment of rough data (time-evolution of the tracer position) is developed in order to predict a series of parameters linked to the solid flow: shape of the bed regions (spout, annulus and fountain), distributions of solid flow in the bed regions, residence time distribution of the solids in the bed, mean solids velocities and flowrates, and voidage in the spout and annulus. Experimental results allowed to show that the shape of the spout is nearly not influenced by inlet air velocity; also, the ratio of volumetric solid flowrate between the different regions of the bed and of the mean solids velocity in the annulus has a constant value for a given static bed height. Empirical correlations are also developed in order to predict the mean solids' velocities and flowrates, and the mean residence time of the solids in each region of the bed.;Gas flow in the conical spouted bed is experimentally characterised by measuring gas residence time distributions (RTD) in the bed through the injection and detection of a radioactive gas tracer into the operated spouted bed. The existence of non-negligible gas flow in the annulus of the bed is highlighted. Mean gas velocities in the spout and annulus, and the part of the total gas flow going to the spout are deducted from the RTD curves. It is identified that gas moves at least twice faster in the spout than in the annulus, which leads to mass exchanges between solid and gas that are more intense in the spout than in the annulus.;Baker's yeast drying experiments are done in a conical spouted bed in order to characterise the effects of operating conditions on drying. A new multiscale model, describing Baker's yeast drying in conical spouted bed, is presented; it is based on experimental results and on the models of a single pellet drying and of gas-solid flows in the conical spouted bed. This phenomenological model has only one unknown parameter and permits reproducing the experimental results of Baker's yeast drying in conical spouted bed. It takes into account the fact that, in a spouted bed, vapour saturation of the air during its residence time in the bed can be a limiting phenomena for the drying rate, especially in the beginning of the drying.;The characterisation of the evolution of the Baker's yeast gassing power in a bread dough during the drying has also been done in the case of a single pellet drying and in the case of spouted bed drying. Similar conclusions are presented for both cases. Indeed, yeast degradation is linked to intracellular water removal (type D, end of the drying below a moisture content of around 0,5 (d.b.)) and it is mostly the rate of this water removal that controls the final quality of the product. Intercellular water removal (type E, beginning of the drying) has no significant influence on yeast degradation. In all the cases, degradation is amplified when solid temperature is higher than 40°C.