Solid-Fluid Transfer Group

Michel Pons

Introduction - Research topics - Teaching, Training, Collaborations -

Exemples of Research activities

INTRODUCTION

The group Solid-Fluid Transfer focuses its researches on various kinds of transfer occurring at an interface between a solid and a fluid: mass-, heat- or momentum-transfer. The solid may be porous, microporous, rough, polished, anisotropic, or idealised. One main concern is to highlight the basic phenomena involved in the corresponding transfer, the other concern may also be to model, simulate, and sometime optimise, processes driven by such a transfer.

Our investigations are mainly experimental: adsorption kinetics or equilibriums, Kapitza resistance, heat transfer in pool boiling, growth of vapour bubbles. However, modelling also plays an important role in these works. First reason, the quantities of interest for basic research can rarely be directly measured. We evaluate them by fitting numerical results to the experimental data. In order to be reliable, this approach requires well-validated models. Second reason, the description of the studied system given by numerical models is always more extended than experimental observations, which are always limited. When both are available, information supplied by the model greatly help interpreting the experiments. Our other works are essentially numerical: flow through periodical and non-isotropic porous media, analysis of thermodynamic systems (either for adsorptive cooling or in natural convection) and also phase change liquid-vapour. When the opportunity exists, the numerical results are compared either to experiments, some of them being performed in LIMSI, or to results of other models. They can also be used for designing experimental machines.

The problems studied in the group Solid-Fluid Transfer most often involve a local scale and a global one. In such cases, an accurate correlation between the two scales is absolutely necessary for interpretation to be correct. Here are some examples of coupling between size-scales that we encounter. (1) The diffusion coefficients of a prescribed gas in the macro- and micropores of a given adsorbent are evaluated from macroscopic measurement of temperature, pressure and volume. (2) Experimental values of the Kapitza resistance give insight on the diffusion of phonons by the rough structure of the interface solid-Helium II. (3) When a vapour bubble leaves the wall, it induces a sudden flow within the whole test-pool volume.

As a summary, it can be said that the group works between experiments and computations, between micro- and macro-scales, and between basic and applied research.

In the group are one researcher, four faculty members, one research engineer, one assistant engineer and one PhD student. In addition, one former researcher, yet retired, brings us an important contribution. Damir Juric, formerly Associate Professor at Georgia Tech, is spending one year (October 2002-September 2003) among us, financially supported by CNRS.

The members of the group Solid-Fluid Transfer develop many collaborations, especially with other French laboratories, but also with some industrial companies and foreign laboratories. New collaborations, just begun in 2002, indicate that new activities are appearing in the group.

 

RESEARCH TOPICS

Topic 1: Solid-gas adsorption

V. Bourdin, D. Bisch

In this field, two problems are experimentally investigated: adsorption kinetics of pure gases, and coadsorption equilibriums.

The development in LIMSI of the experimental set-up for measuring adsorption kinetics begun around 1990. The originality of this set-up consists of combining the method of frequency response with an infrared measurement of the temperature at the surface of the adsorbent sample, so that measurements can be performed at high temperature or on rapid systems. Moreover, the strong stability of the signal makes it possible to investigate accurately systems with very slow kinetics. Unfortunately, in 2002 a serious breakdown occurred during measurements performed at 200°C. The temperature regulation circuits and the infrared optical devices are being repaired. The new systems will improve both the ability and the accuracy of this equipment, e.g. the signal disturbance is expected to be significantly reduced at high temperature. Despite that contretemps, our industrial partner, Institut Français du Pétrole, continues our collaboration.

The second set-up, for measuring coadsorption equilibriums, uses the following principle. In addition to a rather classical volumetric system, that only yields the total number of adsorbed moles, a laser-interferometry system measures the refraction index of the gas phase. The refraction index of a gas depends on its concentration and composition; for a given binary mixture, this correlation can be established by a precalibration, or from published data on refraction indexes of pure gases. From that additional experimental data, the gas composition after adsorption can be evaluated, and consequently the actual composition of the adsorbed phase. This point is essential in the field of coadsorption, because now the actual adsorbed quantity of each component can be experimentally determined -with a flexible, fast and non-intrusive method- without the need of any assumption about the adsorbed phase. As a scientific benefit, complete data of coadsorption equilibriums will be available; different theories about coadsorption will be tested in regard to our experimental data; if effects such as interactions between adsorbed molecules exist, they will be evidenced. However, measuring a refraction index in the present conditions (the gas must be circulated in the circuit for sake of uniformity of its composition) and with sufficient accuracy and stability is not an easy task. In 2002, a computer-controlled regulation system for the adsorbent temperature has been implemented, so that the regeneration process is reproducible. Beside, measurements have already be done on one-component systems (zeolite NaX + alkanes or + CO2), in order to test the experimental procedures. The accuracy of our results permits critical comparison with data published in the literature, see the presentation page [1] . This second equipment should also be of great interest for industrial partners.

This topic is also involved in the project PALOMA, driven by the Institut Pierre-Simon Laplace (IPSL, Paris) and aiming to answer the NASA invitation to tender "2009 Mars Lander". The global purpose is to perform a complete isotopic analysis of the rare gases of Mars. Our contribution focuses on the use of an adsorption process for, first trapping the carbon dioxide (which represents 97% of the Mars atmosphere), second separating efficiently and quantifying accurately the other components. We showed that purely thermal processes (Temperature Swing Adsorption) do not work sufficiently well; better performance could be expected from a chromatographic process.

Lastly, within the framework of a network 'Environment' created in 2002 at Université Paris-Sud, the LIMSI began collaborating with Laboratoire de Physique des Gaz et Plasmas (LPGP, Orsay) about a VOC-depollution process by using cold plasmas and a porous catalyst, the cordierite. Our present contribution consists of measuring the adsorption capacity of VOC traces by the cordierite.

Topic 2: Homogeneisation through anisitropic porous media

M. Firdaouss

The long range purpose of this research is to understand the flow pattern and the transport mechanisms inside periodical, two-dimensional, and however complex, porous media. The effects of geometrical anisotropy on the permeability tensor have been studied, and the results of our detailed numerical model have been compared to those obtained by a simple formula. Three problems are presently investigated: (1) how does the internal porosity of the solid grains influence the flow through porous media with two-scales of porosity, (2) which are the transient effects, (3) how does a hydrogen-storage tank with carbon materials behave during loading operation (coupled heat and mass transfer).

Herein, our main tool is a two-dimensional code solving the Navier-Stokes equations with the method of finite elements. The equations are homogenised with the assumption of periodicity, and solved accurately in the fluid volume lying between the closed solid particles. Anisotropy can be due to the particle shape or to the relative positions of the particles.

In collaboration with Jean Prieur du Plessis (University of Stollenbosch, South Africa) we studied the permeability of two-dimensional porous media consisting of rectangular solid particles arranged periodically in rectangular cells. Assuming Poiseuille-type flow between the parallel edges of the solid particles, Jean Prieur du Plessis obtains the permeability of the medium with a simple analytical formula. We compare his results to the average permeability resulting from our Navier-Stokes code. The agreement between the two approaches is excellent as long as the porosity lies under 0.6. This work will be published soon.

The problem of solution transport through a periodical two-dimensional porous medium, tackled in P. Tran's PhD-thesis (defended in 2000) is presently under close study in collaboration with B. Goyeau (Fluides, Automatique et Systèmes Thermiques, FAST, Orsay).

A new collaboration is beginning with Laboratoire d'Ingénierie des Matériaux et Hautes Pressions (LIMHP, Villetaneuse) in the framework of the Action 'Énergie' of CNRS and French Ministry of Research and Technology. It focuses on modelling and simulation of the operation of a hydrogen storage reservoir filled with carbonaceous materials. Such a reservoir could equip vehicles powered with hydrogen. This collaboration also relies on other competence existing in LIMSI about heat and mass transfer in adsorbent packed beds and adsorption processes. It also raises the question of how the adsorbing porosity of the solid grains influences the flow through the inter-granular volume; a question already approached in 1996-1997 with K. Naimi's PhD-thesis. Moreover, the operation of such a reservoir is essentially transient, which brings us to test and run our models for such conditions. Comparison with experimental data obtained in LIMHP will be possible.

Topic 3: Modeling and second law analysis of systems

M. Pons

This topic has two aspects. The first concern is about modelling and second law analysis of energy processes, especially solar-powered adsorption cooling-machines. The second concern, newly developed, is about the thermodynamics of natural convection.

The solar-powered refrigerator, designed in collaboration with the École d'Ingénieurs du Canton de Vaud (EIVD, Yverdon-les-Bains, Switzerland) and tested in Yverdon-les-Bains, achieved its third year of operation. The experimental performance in 2002 confirmed the results of the former years, i.e. an improvement of about 45 % compared to the best values found in the literature. These good results make this prototype a promising basis for technology transfer, with the help of non-governmental organisations, toward developing countries. In 2002, the collaboration with EIVD focused mainly on measurement interpretation and common publications.      
After having established the second law balance of several energy processes with adsorption, a somehow new concept has been developed, that of irreversibility number. This non-dimensional quantification of irreversibilities has the following characteristics:     
- this number can be obtained as well via the entropy analysis as via the exergy analysis;    
- it perfectly keeps the property of additivity that characterises exergy-losses and entropy productions;
- it leads to a nondimensional second law balance. This balance appears to be a robust framework for analysing energy cycles more complex than the basic cycles. Among complex cycles can be found the processes powered altogether by thermal and mechanical energy, the cogeneration processes (producing work & heat, heat & cold, or trigeneration), the four-temperature heat-powered cycles, or any combination of these possibilities. 
This reflection is developed within the PRI CARNOT in the framework of the Action 'Énergie' of CNRS and French Ministry of Research and Technology [2] .

The thermodynamic analysis of natural convection and Boussinesq approximation leads to results described in the presentation page [3] . They can be summarised as follows. In the heat equation must be taken into account, not only the source term due to viscous dissipation, but also the thermal effects of fluid-expansion/-contraction [4] within the hydrostatic pressure gradient. These two source terms make the system thermodynamically consistent: the total energy is indeed conserved, the irreversibility balance is consistent with requirements deduced from global analysis. The latter point is important because it opens questions such as: how does the principle of entropy-production minimisation, introduced by Prigogine, apply in natural convection? Indeed, there are two kinds of irreversibility, the conductive one and the viscous one: which is their topology and which trade-off do they find? This reflection is developed within an AmETh group involving Laboratoire d'Études Thermiques (LET, Poitiers), Laboratoire d'Étude des Phénomènes de Transfert Appliqués au Bâtiment (LEPTAB, La Rochelle), Centre de Thermique de Lyon (CETHIL, Villeurbanne) et Centre d'Études et de Recherche Méthodologique d'Architecture (CERMA, Nantes).

Topic 4: Heat transfer between solid nd superfluid helium

M.-X. François, J. Amrit

In this field, our research is both basic and applied.

The fundamental aspect focuses on the heat transfer resistance located at the interface between solid and superfluid helium, called the Kapitza resistance. Indeed, the disagreement between the experimental values on one side, and on the other side the values predicted by different theories (for instance that of Khalatnikov) can be as large as a factor of 30-40. The reasons of this disagreement are not yet clear.

The applied aspect focuses on thermal optimisation of the materials used in superconducting cavities of particle accelerators. In these superconducting cavities, made of Niobium, the magnetic component of the high-frequency field injected into the cavity generates an electrical current at the internal surface of the cavity. Because of the surface electrical resistance, this current dissipates heat by the Joule effect and the temperature of the cavity tends to increase. On its outer interface, the cavity wall is cooled by superfluid helium. The temperature in the cavity wall, especially at the inner interface, depends on, first the intensity of the internal heat dissipation –related via the electrical current to the intensity of the magnetic field-, second the global heat-transfer resistance toward the helium pool. This global resistance is the sum of the conductive resistance in the Niobium itself plus the Kapitza resistance. The latter resistance appears to be dominating. As it is necessary to maintain the Niobium in superconducting conditions, the magnetic field of the accelerator must lie under a maximal limit. This is the point where the basic and applied aspects of our research meet: studying the Kapitza resistance. Any factor likely to reduce this global resistance makes it possible to enhance the maximal magnetic field allowed in particle accelerators.

With our experimental set-up, we can measure the thermal resistance between a sample (of Niobium or Silicium) and the superfluid Helium pool. The set-up has been significantly modified in 2002: design and implementation of a new experimental cell, implementation of a new temperature regulation and of a new data acquisition system. As accuracy and reproducibility have both been improved, we can now interpret reliably the differences appearing in the measured thermal resistance for various samples. For instance, by comparing data obtained with samples of different thicknesses, the transverse thermal conductivity of the samples can be determined, see the presentation page [5] . Moreover, augmentation of the heat-exchange surface by engravement is likely to reduce the effective resistance. This axis can now be investigated and should result in concrete propositions for enhancing the maximal allowed value of the magnetic field in particle accelerators. These works are conducted in collaboration with C.‑Z. Antoine (DAPNIA/SEA, CEA, Saclay).

Moreover, our measurements of the Kapitza resistance on Silicium samples with surface roughness in the range of nanometers, and our first analysis, support our assumption that phonons are scattered at the solid-helium interface. According to Adamenko and Fuks [6] , this scattering of phonons is likely to explain the above-mentioned discrepancy between experimental and theoretical values of the Kapitza resistance. Our measurements apparently confirm that assumption, at least in the limits of accuracy obtained in the evaluation of surface roughness, yet measured by atomic force microscopy. Indeed, the surface roughness appears to be fractal-like. We are now especially interested in correlating the wavelength of the scattered phonons to the size of the surface roughness at a given scale. The roughness size is obviously fixed, the sample temperature must therefore be changed over a sufficiently large range, so that the phonon wavelength is significantly modified. The measurement cell for this investigation is designed, we must now find a temperature regulation operating below 2 K. This work is supported by an Action  Incitative LIMSI.

 

Topic 5: Boiling heat transfer

M.-C. Duluc, V. Daru, D. Juric, J. Pakleza

The research on this topic is mainly experimental and follows three axes.

The first one is the experimental study of transient pool boiling around a wire. Experiments can be described as follows: starting from given initial conditions, the heat-flux in the wire is increased stepwise and the consequent change in the wire temperature is recorded. In the past years, equilibrium with saturated liquid was the initial condition: the result is a delayed onset of pool boiling, and a transient overshoot of the wire superheat, significantly large compared to one recorded in steady-state with the same heat-flux. One assumption for explaining both delay and overshoot involves activation of nucleation sites. In 2002, systematic measurements have been done with initial conditions maintained out of equilibrium by a small heat-flux. When comparing the latter experiments (initial condition out of equilibrium) to the former ones (initial condition at equilibrium), with a same heating flux, the results can be summarised as follows. When the initial heat-flux is so low that only convection takes place in the fluid, without boiling, then the time delay for reaching the steady state with pool boiling is significantly reduced, but the wire transient superheat is almost unchanged. When the initial condition already involves boiling, even at a very low rate, then the systems reaches steady-state very rapidly and the transient superheat is significantly reduced (e.g. by a factor larger than two). Moreover, the thermal response of the wire is highly reproducible. These results show that two phenomena are involved in the onset of pool boiling: onset of natural convection, activation of nucleation sites. More details are given in the presentation page [7] . Our experimental results, as well as our analysis, are consistent with those developed in Centre de Thermique de Lyon (CETHIL, Villeurbanne) and Laboratoire d’Énergétique et de Mécanique Théorique et Appliquée (LEMTA, Nancy) in the framework of the AmETh sub-network [8] dedicated to transient pool boiling. These results can be very useful for improving the operation procedures of vapour generators with frequent on/off operations.

The second axis, study of laminar free convection, involves several members of the Mechanical Engineering Department. The experimental rig is rather similar to that used for studying transient pool-boiling, except it is combined with the PIV equipment developed by F. Lusseyran and P. Gougat (group Fluid Dynamics and Turbulence) in order to capture the transient velocity fields. These experimental fields are then compared to the corresponding ones numerically computed by S. Xin and P. Le Quéré (group Dynamics of Heat Transfer and Instabilities). The first comparisons resulted in an excellent agreement, considering velocity fields or movement of vortex centres (see last year's presentation page [9] ). In addition, in 2002 a very good agreement has been similarly obtained for the temperatures in the plume. Once all these results are reproduced and confirmed, unsteady configurations with modulated heat-flux will be investigated.

The third axis consists of the study of vapour bubble growth. Our approach is mainly experimental, but nowadays significant progress rely on the use of a high-speed video-camera. Meanwhile, numerical simulation of this problem is under active development, thanks to both presence of Damir Juric (formerly associate professor at Georgia Tech) in LIMSI, and collaboration with V. Daru (group Fluid Dynamics and Turbulence). The initial code, developed by D. Juric, could treat three-dimensional, two-phase with phase change liquid-vapour situations, but with a density ratio between the two phases not larger than ten or twenty. We have introduced into the model the relevant phenomena from the viewpoint of physics, heat-transfer and hydrodynamics, and also we use numerical schemes formerly developed for compressible fluids. This original collaboration of different kinds of know-how results in a radical transformation of the code. Our first results are detailed in two presentation pages Part I [10] and Part II [11] and they are promising. One word about experiments: a large number of configurations investigated in the last years have been classified. This work evidences that the size of the heating spot is an important parameter for the life cycle of the bubble.           
The study of vapour bubbles, either experimentally or numerically, became very active in the last years, so that several French laboratories (CETHIL in Villeurbanne, CEA-Grenoble, IMFT in Toulouse, IUSTI in Marseille, MASTER in Bordeaux and LIMSI) decided to form a new AmETh network dedicated to this problem. The excellent collaboration with T. Kowalewski (IPPT-PAN, Warszawa, Poland), supported by CNRS, must also be mentioned.

 

Specific action: Hydrodynamics and boiling in liquid nitrogen

M.-X. François, G. Defresne, A. Planchette

In response to a demand from Air-Liquide, a BDI thesis began in September 2001 for studying experimentally the interactions between boiling in liquid nitrogen and fluid hydrodynamics in parallel vertical channels. The experimental set-up has been built in the laboratory of Air-Liquide. Analysis aims to determining how the vaporised flow rate and the two-phase flow configuration depend on the experimental conditions, liquid flow rate and heating energy.

TEACHING, TRAINING, COLLABORATIONS

TEACHING ACTIVITIES AND KNOWLEDGE DIFFUSION

Organization of Conferences

- M.-X. François is member of the organizing committee of the International Conference of Cryogenics Refrigeration, scheduled in 2003, Hangzhou, China.

- M.-C. Duluc and M. Pons are members of the organizing committee of Congrès Français de Thermique SFT-2004, Société Française de Thermique, scheduled in May 2004, in Presqu'île de Giens, France.

Participation in Editorial Boards

- M.-X. François : Journal de Physique III, Cryogenics.

Teaching activities or responsibilities (related to research)

- M.-X. François is co-responsible of the DEA Fluid and Transfer Dynamics for Paris 6 University; he teaches Complex and two-phase fluids in this DEA.

- M. Pons teaches Conductive and radiative heat transfer in the DESS Simulations in Fluid- and Transfer-Dynamics for University Paris-Sud.

- M. Pons gives a research lecture about Thermodynamics, natural convection and the Boussinesq approximation in the DEA Mechanics, Paris-6 University.

TraIning

- J. Amrit: teaching about Superfluid helium in Cryogenics.

- V. Bourdin: Combustion, pollution and elements of solution, Évry University (IUT).

- M.-X. François: teaching of Cryogenics.

Participation to seminars

- M. Pons : "Thermodynamics and natural convection", seminar in Laboratoire de Modélisation en Mécanique (LMM), Univ. Paris-6, June 18, 2002, Paris.

- M. Pons : " Thermodynamics: from adsorptive machines to numerical fluid mechanics ", seminar PRI CARNOT, September 16, 2002, Ecole des Mines de Paris.

NATIONAL RELATIONS

Institutional responsibilities

- M.-X. François is member of the bureau of AFF (French association for refrigeration), part of the IFF (French institute for refrigeration).

Scientific relations

- Participation to a network coordinated by Institut Français du Pétrole, with three French laboratories (LRRS in Université de Bourgogne, Dijon, IRC in Villeurbanne, and LIMSI), about alkane diffusion in an industrial zeolite.

- Participation to the project PALOMA, conducted by Institut Pierre-Simon Laplace, Paris, for projects related to the NASA action, "2009 Mars Lander" (full isotopic analysis of the rare gases to be captured on Mars).

- New collaboration with Laboratoire de Physique des Gaz et Plasmas (LPGP, Orsay) about interaction between cold plasma, VOC's and cordierite.

- Collaboration with B. Goyeau, FAST (Orsay), about solute transport through a two-dimensional periodic porous medium.

- Collaboration with LIMHP (Villetaneuse) about modelling and simulation of a hydrogen storage tank on carbonaceous materials.

- Participation to PRI CARNOT (Communauté d'Analyse et de Réflexion sur les Nouvelles Orientations de la Thermodynamique), in the framework of the Action Énergie of CNRS and MNRT.

- Collaboration with C.-Z. Antoine, (DAPNIA/SEA, CEA, Gif sur Yvette, France) about heat transfer between a solid and superfluid helium.

- Participation to the project Transient Boiling in the AmETh network, with CETHIL (Villeurbanne) and LEMTA (Nancy). This project is conducted by M.-C. Duluc.

- Participation to the new project Growth of Vapour Bubbles in the AmETh network, with CETHIL (Villeurbanne), IUSTI (Marseille), IMFT (Toulouse), MASTER (Bordeaux) et CEA-Grenoble.

- Collaboration, about hydrodynamics and boiling of liquid nitrogen, with Air-Liquide.

Contracts

- PRI Materials for Storing Hydrogen, in the action Énergie of CNRS and MNRT, with LIMHP (Villetaneuse), LPT (Orsay), LCMTR (Vitry-Thiais) and ICMCB (Bordeaux).

- Project H2-THERM (Thermal and Mechanical Modelling of Hydrogen Storage by Adsorption), in the action Énergie of CNRS and MNRT, with LIMHP (Villetaneuse), LEGI (Grenoble), LMARC (Besançon), LEMTA (Nancy) and IMP (Perpignan).

INTERNATIONAL RELATIONS

Scientific relations

- Collaboration with Jean Prieur du Plessis (University of Stollenbosch, South Africa), about permeability of periodic porous media.

- Collaboration with Ph. Dind, C. Hildbrand, F. Buchter and J. Mayor (Ecole d'Ingénieurs du Canton de Vaud, Yverdon-les-Bains, Switzerland), about solar-powered refrigeration by adsorption.

- Collaboration with T. Kowalewski (Science Academy, IPPT-PAN, Warszawa, Poland), about growth of vapour bubbles.