Food fluids' flow modelling : complex, thixotropic and multiphases flow.
Details of the research topic
From a rheological point of view, the food product flow modelling, such as
«fruit yoghourt», is a problem that couples many physical aspects.
Controlling the product behaviour during the different stages of the manufacturing
process is an essential issue to control the product quality, its reliability and even
to specify the required tools.
The current modelling methods have two main difficulties:
-
Because of the macroscopic and mesoscopic aspect of the majority of
simulations and expermiental processes, the latter cannot lead to accurate results on
the microscopic structure state. This state is charged and coloured materials' specific.
-
Taking into account the mutual effects, or in other terms, the consequence of
the microstructure, or also the thixotropic behaviour that greatly affects the
flow kinematics at macroscopic scales. In addition, it results that the product
homogeneïty, the particles' dispersion quality (fruits) and all the final product mecanichal
properties are greatly dependant on the mix and agregation state.
Resolving this type of problems only by experimental methods cannot enable as to
achieve consistant knowledge about all the interacting mechanisms. Solving the
problem requires numerical, physical and mechanical skills. Three sets of themes
are specific to this type of problems :
-
First of all, the food product thixotropic behaviour. This behaviour type is very
particular and characterizes a microscopic structure that can incorporate and
dissociate under shearing strains. Existing developped models are based
on macroscopic approaches using a scalar parameter of structure.
If we realise that such a model is not rich enough to give microscopic
information we will use a finer approach based on mutli-population
microscopic model which behaviour of each population is affected by
the flow kinematics and also the transversal exchanges between populations.
Those exchanges charactrize the agregations' state evolution.
-
The second characterizes the dye or jam mixes (yoghourt). This problematic is generally
approached by a a Lagrangian approach in most commercial codes, in which they put
initially a large number of particles and observe their evolution in the system.
The mix quantity is only qualitatively described. In This PhD research we suggest
to describe that mix using a finer approach that give as the amount of the mix using a
factor related to the exchange interface surface between the two phases.
This result can be described by a Fokker Planck formalism in a microscopic space.
-
The third theme is specific to the particles' concentration description (of fruit particles).
A simple description consists of considering the global scalar variable to describe
the concentration evolution continously.
This variable follows a convection diffusion law, the matric coefficients of which
can be eventually anisotropic. If such approach can be insufficient we can consider
a particle specific dynamic study. In this case, each particle is individully
isolated to solve its conservation equations. We can also consider other appoaches that
will be in the medium level between simple and dynamic ones.
Bibliographie sur le sujet
F. Chinesta, A. Ammar, F. Lemarchand, P. Beauchene, F. Boust. «Efficient numericlal tools
for simulating thermal model: model reduction, parallel integration and high resolution
homogenization », Computer Methods in Applied Mechanics
and Engineering.
J. Azaiez, K. Chiba, F. Chinesta. A. Ammar, «On the Steady Regime Assumptions in Flows
Involving Short Fiber Suspensions», International Journal
of Forming Processes.
A. Ammar, B. Mokdad, F. Chinesta, R. Keunings, «New Family of Solvers for Some Classes
of Multidimensional Partial Differential Equations Encountered in Kinetic Theory
Modeling of Complex Fluids», Journal of Non-Newtonian
Fluid Mechanics.
A. Ammar, F. Chinesta, D. Ryckelynck,«Deterministic Particle Approach
of Multi-Bead-Spring Polymer Models » European Journal
of Computational Mechanics.
A. Ammar, D. Ryckelynck, F. Chinesta, R. Keunings, «On the Reduction of Kinetic
Theory Models Related to Finitely Extensible Dumbbells»,
Journal of Non-Newtonian Fluid Mechanics, 134 : 136-147, 2006.,
D. Ryckelynck, F. Chinesta, E. Cueto, A. Ammar,
«On the "A Priori" Model Reduction: Overview and Recent Developments»,
Archives of Computational Methods in Engineering, State of the
Art Reviews, 13/1 : 91-128, 2006.
F. Chinesta, E. Cueto, D. Ryckelynck, A. Ammar,
«Alpha-NEM and Model Reduction: Two New and Powerful Numerical
Strategies for Describe Flows Involving Short Fiber
Suspensions»Revue Européenne des Eléments
Finis, 14/6-7 : 903-923, 2005.
Laboratoire d'accueil
Laboratoire de Rhéologie - UMR 5520 / UJF / INPG
Directeurs de thèse
Amine AMMAR
Francisco Chinesta, Laboratoire : LMSP, ENSAM Paris.
Mikael Berton, entreprise : PCM / FTC / Champtocé sur Loire.
