SAM
https://sam.ensam.eu:443
The DSpace digital repository system captures, stores, indexes, preserves, and distributes digital research material.Mon, 13 Mar 2023 18:57:56 GMT2023-03-13T18:57:56ZConséquences du confinement dans le transfert de chaleur sur une sphère dans un fluide non newtonien
http://hdl.handle.net/10985/6606
Conséquences du confinement dans le transfert de chaleur sur une sphère dans un fluide non newtonien
DESPEYROUX, Antoine; AMBARI, Abdelhak; BEN RICHOU, Abderrahim; CHAMPMARTIN, Stéphane
Le phénomène de transfert de chaleur ou de masse sur une particule sphérique en situation d’interactions hydrodynamiques et thermiques ou massiques est d’un intérêt majeur pour de nombreux problèmes rencontrés dans les procédés industriels faisant intervenir des particules en suspension [1]. Nous avons contribué par ce travail à l’étude de l’influence du confinement sur les phénomènes de transfert en présence d’un fluide de type loi de puissance. La comparaison des résultats numériques avec les calculs asymptotiques effectués par nous-mêmes et ceux obtenus par Acrivos confirment la validité des calculs. Dans le cas des fluides non newtoniens, il apparaît que la rhéofluidification est favorable aux phénomènes de transfert contrairement au cas rhéoépaississant.
Fri, 01 Jan 2010 00:00:00 GMThttp://hdl.handle.net/10985/66062010-01-01T00:00:00ZDESPEYROUX, AntoineAMBARI, AbdelhakBEN RICHOU, AbderrahimCHAMPMARTIN, StéphaneLe phénomène de transfert de chaleur ou de masse sur une particule sphérique en situation d’interactions hydrodynamiques et thermiques ou massiques est d’un intérêt majeur pour de nombreux problèmes rencontrés dans les procédés industriels faisant intervenir des particules en suspension [1]. Nous avons contribué par ce travail à l’étude de l’influence du confinement sur les phénomènes de transfert en présence d’un fluide de type loi de puissance. La comparaison des résultats numériques avec les calculs asymptotiques effectués par nous-mêmes et ceux obtenus par Acrivos confirment la validité des calculs. Dans le cas des fluides non newtoniens, il apparaît que la rhéofluidification est favorable aux phénomènes de transfert contrairement au cas rhéoépaississant.Identification of the pore size distribution of a porous medium by yield stress fluids using Herschel-Bulkley model
http://hdl.handle.net/10985/19438
Identification of the pore size distribution of a porous medium by yield stress fluids using Herschel-Bulkley model
OUKHLEF, Aimad; AMBARI, Abdelhak; CHAMPMARTIN, Stéphane
In this paper, we present a new method to determine the pore-size distribution (PSD) in a porous medium. This innovative technique uses the rheological properties of non-Newtonian yield stress fluids flowing through the porous sample. In a first approach, the capillary bundle model will be used. The PSD is obtained from the measurement of the total flow rate of fluid as a function of the imposed pressure gradient magnitude. The mathematical processing of the experimental data, which depends on the type of yield stress fluid, provides an overview of the pore size distribution of the porous material. The technique proposed here was successfully tested analytically and numerically for usual pore size distributions such as the Gaussian mono and multimodal distributions. The study was conducted for yield stress fluids obeying the classical Bingham model and extended to the more realistic Herschel-Bulkley model. Unlike other complex methods, expensive and sometimes toxic, this technique presents a lower cost, requires simple measurements and is easy to interpret. This new method could become in the future an alternative, non-toxic and cheap method for the characterization of porous materials.
Wed, 01 Jan 2020 00:00:00 GMThttp://hdl.handle.net/10985/194382020-01-01T00:00:00ZOUKHLEF, AimadAMBARI, AbdelhakCHAMPMARTIN, StéphaneIn this paper, we present a new method to determine the pore-size distribution (PSD) in a porous medium. This innovative technique uses the rheological properties of non-Newtonian yield stress fluids flowing through the porous sample. In a first approach, the capillary bundle model will be used. The PSD is obtained from the measurement of the total flow rate of fluid as a function of the imposed pressure gradient magnitude. The mathematical processing of the experimental data, which depends on the type of yield stress fluid, provides an overview of the pore size distribution of the porous material. The technique proposed here was successfully tested analytically and numerically for usual pore size distributions such as the Gaussian mono and multimodal distributions. The study was conducted for yield stress fluids obeying the classical Bingham model and extended to the more realistic Herschel-Bulkley model. Unlike other complex methods, expensive and sometimes toxic, this technique presents a lower cost, requires simple measurements and is easy to interpret. This new method could become in the future an alternative, non-toxic and cheap method for the characterization of porous materials.Reliability of a Hydrodynamic Journal Bearing
http://hdl.handle.net/10985/10283
Reliability of a Hydrodynamic Journal Bearing
DIOP, Khadim; CHARKI, Abdérafi; CHAMPMARTIN, Stéphane; AMBARI, Abdelhak
Journal fluid bearings are widely used in industry due to their static and dynamic behavior and their very low coefficient of friction. The technical requirements to improve the new technologies design are increasingly focused on the indicators of dependability of systems and machines. Then, it is necessary to develop a methodology to study the reliability of bearings in order to improve and to evaluate their design quality. Few works are referenced in literature concerning the estimation of the reliability of fluid journal bearings. This paper deals with a methodology to study the failure probability of a hydrodynamic journal bearing. An analytical approach is proposed to calculate static characteristics in using the Reynolds equation. The commonly methods used in structural reliability such as FORM (First Order Reliability Method), SORM (Second Order Reliability Method) and Monte Carlo are developed to estimate the failure probability. The function of performance bounding two domains (domain of safety and domain of failure) is estimated for several geometrical configurations of a hydrodynamic journal bearing (long journal bearings with the hypotheses of Sommerfeld, Gümbel and Reynolds, and a short journal bearing with the hypothesis of Gümbel).
Thu, 01 Jan 2015 00:00:00 GMThttp://hdl.handle.net/10985/102832015-01-01T00:00:00ZDIOP, KhadimCHARKI, AbdérafiCHAMPMARTIN, StéphaneAMBARI, AbdelhakJournal fluid bearings are widely used in industry due to their static and dynamic behavior and their very low coefficient of friction. The technical requirements to improve the new technologies design are increasingly focused on the indicators of dependability of systems and machines. Then, it is necessary to develop a methodology to study the reliability of bearings in order to improve and to evaluate their design quality. Few works are referenced in literature concerning the estimation of the reliability of fluid journal bearings. This paper deals with a methodology to study the failure probability of a hydrodynamic journal bearing. An analytical approach is proposed to calculate static characteristics in using the Reynolds equation. The commonly methods used in structural reliability such as FORM (First Order Reliability Method), SORM (Second Order Reliability Method) and Monte Carlo are developed to estimate the failure probability. The function of performance bounding two domains (domain of safety and domain of failure) is estimated for several geometrical configurations of a hydrodynamic journal bearing (long journal bearings with the hypotheses of Sommerfeld, Gümbel and Reynolds, and a short journal bearing with the hypothesis of Gümbel).Reliability of a hydrostatic bearing
http://hdl.handle.net/10985/8530
Reliability of a hydrostatic bearing
CHARKI, Abdérafi; DIOP, Khadim; CHAMPMARTIN, Stéphane; AMBARI, Abdelhak
This paper presents a methodology for evaluating the failure probability of fluid bearings, which are sensitive components for the design of machine rotors, mechatronic systems, and high precision metrology. The static and dynamic behavior of a fluid bearing depends on several parameters, such as external load, bearing dimensions, supply pressure, quality of the machined surfaces, fluid properties, etc. In this paper, the characteristics of a simple geometry hydrostatic bearing are calculated analytically in order to demonstrate the usefulness of the methodology and its pertinence to bearing design. Monte Carlo simulation and first order reliability method (FORM) are used to evaluate the probability of failure.
Tue, 01 Jan 2013 00:00:00 GMThttp://hdl.handle.net/10985/85302013-01-01T00:00:00ZCHARKI, AbdérafiDIOP, KhadimCHAMPMARTIN, StéphaneAMBARI, AbdelhakThis paper presents a methodology for evaluating the failure probability of fluid bearings, which are sensitive components for the design of machine rotors, mechatronic systems, and high precision metrology. The static and dynamic behavior of a fluid bearing depends on several parameters, such as external load, bearing dimensions, supply pressure, quality of the machined surfaces, fluid properties, etc. In this paper, the characteristics of a simple geometry hydrostatic bearing are calculated analytically in order to demonstrate the usefulness of the methodology and its pertinence to bearing design. Monte Carlo simulation and first order reliability method (FORM) are used to evaluate the probability of failure.Wall effects on the transportation of a cylindrical particle in power-law fluids
http://hdl.handle.net/10985/8454
Wall effects on the transportation of a cylindrical particle in power-law fluids
DESPEYROUX, Antoine; AMBARI, Abdelhak; BEN RICHOU, Abderrahim
The present work deals with the numerical calculation of the Stokes-type drag undergone by a cylindrical particle perpendicularly to its axis in a power-law fluid. In unbounded medium, as all data are not available yet, we provide a numerical solution for the pseudoplastic fluid. Indeed, the Stokes-type solution exists because the Stokes’ paradox does not take place anymore. We show a high sensitivity of the solution to the confinement, and the appearance of the inertia in the proximity of the Newtonian case, where the Stokes’ paradox takes place. For unbounded medium, avoiding these traps, we show that the drag is zero for Newtonian and dilatant fluids. But in the bounded one, the Stokes-type regime is recovered for Newtonian and dilatant fluids. We give also a physical explanation of this effect which is due to the reduction of the hydrodynamic screen length, for pseudoplastic fluids. Once the solution of the unbounded medium has been obtained, we give a solution for the confined medium numerically and asymptotically. We also highlight the consequence of the confinement and the backflow on the settling velocity of a fiber perpendicularly to its axis in a slit. Using the dynamic mesh technique, we give the actual transportation velocity in a power-law “Poiseuille flow”, versus the confinement parameter and the fluidity index, induced by the hydrodynamic interactions.
Sat, 01 Jan 2011 00:00:00 GMThttp://hdl.handle.net/10985/84542011-01-01T00:00:00ZDESPEYROUX, AntoineAMBARI, AbdelhakBEN RICHOU, AbderrahimThe present work deals with the numerical calculation of the Stokes-type drag undergone by a cylindrical particle perpendicularly to its axis in a power-law fluid. In unbounded medium, as all data are not available yet, we provide a numerical solution for the pseudoplastic fluid. Indeed, the Stokes-type solution exists because the Stokes’ paradox does not take place anymore. We show a high sensitivity of the solution to the confinement, and the appearance of the inertia in the proximity of the Newtonian case, where the Stokes’ paradox takes place. For unbounded medium, avoiding these traps, we show that the drag is zero for Newtonian and dilatant fluids. But in the bounded one, the Stokes-type regime is recovered for Newtonian and dilatant fluids. We give also a physical explanation of this effect which is due to the reduction of the hydrodynamic screen length, for pseudoplastic fluids. Once the solution of the unbounded medium has been obtained, we give a solution for the confined medium numerically and asymptotically. We also highlight the consequence of the confinement and the backflow on the settling velocity of a fiber perpendicularly to its axis in a slit. Using the dynamic mesh technique, we give the actual transportation velocity in a power-law “Poiseuille flow”, versus the confinement parameter and the fluidity index, induced by the hydrodynamic interactions.Scaling laws explain foraminiferal pore patterns
http://hdl.handle.net/10985/19437
Scaling laws explain foraminiferal pore patterns
RICHIRT, Julien; CHAMPMARTIN, Stéphane; SCHWEIZER, Magali; MOURET, Aurelia; PETERSEN, Jassin; AMBARI, Abdelhak; JORISSEN, Frans
Due to climate warming and increased anthropogenic impact, a decrease of ocean water oxygenation is expected in the near future, with major consequences for marine life. In this context, it is essential to develop reliable tools to assess past oxygen concentrations in the ocean, to better forecast these future changes. Recently, foraminiferal pore patterns have been proposed as a bottom water oxygenation proxy, but the parameters controlling foraminiferal pore patterns are still largely unknown. Here we use scaling laws to describe how both gas exchanges (metabolic needs) and mechanical constraints (shell robustness) control foraminiferal pore patterns. The derived mathematical model shows that only specific combinations of pore density and size are physically feasible. Maximum porosity, of about 30%, can only be obtained by simultaneously increasing pore size and decreasing pore density. A large empirical data set of pore data obtained for three pseudocryptic phylotypes of Ammonia, a common intertidal genus from the eastern Atlantic, strongly supports this conclusion. These new findings provide basic mechanistic understanding of the complex controls of foraminiferal pore patterns and give a solid starting point for the development of proxies of past oxygen concentrations based on these morphological features. Pore size and pore density are largely interdependent, and both have to be considered when describing pore patterns.
Tue, 01 Jan 2019 00:00:00 GMThttp://hdl.handle.net/10985/194372019-01-01T00:00:00ZRICHIRT, JulienCHAMPMARTIN, StéphaneSCHWEIZER, MagaliMOURET, AureliaPETERSEN, JassinAMBARI, AbdelhakJORISSEN, FransDue to climate warming and increased anthropogenic impact, a decrease of ocean water oxygenation is expected in the near future, with major consequences for marine life. In this context, it is essential to develop reliable tools to assess past oxygen concentrations in the ocean, to better forecast these future changes. Recently, foraminiferal pore patterns have been proposed as a bottom water oxygenation proxy, but the parameters controlling foraminiferal pore patterns are still largely unknown. Here we use scaling laws to describe how both gas exchanges (metabolic needs) and mechanical constraints (shell robustness) control foraminiferal pore patterns. The derived mathematical model shows that only specific combinations of pore density and size are physically feasible. Maximum porosity, of about 30%, can only be obtained by simultaneously increasing pore size and decreasing pore density. A large empirical data set of pore data obtained for three pseudocryptic phylotypes of Ammonia, a common intertidal genus from the eastern Atlantic, strongly supports this conclusion. These new findings provide basic mechanistic understanding of the complex controls of foraminiferal pore patterns and give a solid starting point for the development of proxies of past oxygen concentrations based on these morphological features. Pore size and pore density are largely interdependent, and both have to be considered when describing pore patterns.Kinematics of a Cylindrical Particle at Low Reynolds Numbers in Asymmetrical Conditions
http://hdl.handle.net/10985/19436
Kinematics of a Cylindrical Particle at Low Reynolds Numbers in Asymmetrical Conditions
CHAMPMARTIN, Stéphane; AMBARI, Abdelhak; BEN RICHOU, Abderrahim
This paper concerns the hydrodynamic interactions on a cylindrical particle in non-dilute regime at low Reynolds numbers. The particle moves between two parallel walls with its axis parallel to the boundaries. A numerical finite-volume procedure is implemented and a generalized resistance matrix is built by means of the superposition principle. Three problems are solved: the settling of the particle, the transport of a neutrally and of a non-neutrally buoyant particle in a Poiseuille flow. Concerning sedimentation, the settling velocity is maximal off the symmetry plane and decreases when the confinement increases. The particle rotates in the direction opposite to that of contact rolling. The particle induces a high pressure zone in the front and a low pressure zone in the back, the difference of which is maximal in the symmetry plane. For a neutrally-buoyant particle, the hydrodynamic interactions lead to a velocity lag between the particle and the undisturbed flow. The magnitude of the velocity lag increases with confinement and eccentricity. The angular velocity and pressure difference are opposite to the previous case. For a non-neutrally buoyant particle, three situations are found depending on a dimensionless parameter similar to an inverse Shields number. For its extreme low and high values, the particle is respectively either carried by the flow or settles against it whatever its position. For intermediate values, the particle either settles close to the walls or is dragged by the flow close to the symmetry plane. Similar results are obtained for the angular velocity and the pressure difference. All these results question the assumption usually met in particulate transport in which the kinematics of the particle is often supposed to be that of the flow.
Tue, 01 Jan 2019 00:00:00 GMThttp://hdl.handle.net/10985/194362019-01-01T00:00:00ZCHAMPMARTIN, StéphaneAMBARI, AbdelhakBEN RICHOU, AbderrahimThis paper concerns the hydrodynamic interactions on a cylindrical particle in non-dilute regime at low Reynolds numbers. The particle moves between two parallel walls with its axis parallel to the boundaries. A numerical finite-volume procedure is implemented and a generalized resistance matrix is built by means of the superposition principle. Three problems are solved: the settling of the particle, the transport of a neutrally and of a non-neutrally buoyant particle in a Poiseuille flow. Concerning sedimentation, the settling velocity is maximal off the symmetry plane and decreases when the confinement increases. The particle rotates in the direction opposite to that of contact rolling. The particle induces a high pressure zone in the front and a low pressure zone in the back, the difference of which is maximal in the symmetry plane. For a neutrally-buoyant particle, the hydrodynamic interactions lead to a velocity lag between the particle and the undisturbed flow. The magnitude of the velocity lag increases with confinement and eccentricity. The angular velocity and pressure difference are opposite to the previous case. For a non-neutrally buoyant particle, three situations are found depending on a dimensionless parameter similar to an inverse Shields number. For its extreme low and high values, the particle is respectively either carried by the flow or settles against it whatever its position. For intermediate values, the particle either settles close to the walls or is dragged by the flow close to the symmetry plane. Similar results are obtained for the angular velocity and the pressure difference. All these results question the assumption usually met in particulate transport in which the kinematics of the particle is often supposed to be that of the flow.Numerical simulation and experimental study of thrust air bearings with multiple orifices
http://hdl.handle.net/10985/8480
Numerical simulation and experimental study of thrust air bearings with multiple orifices
CHARKI, Abdérafi; DIOP, Khadim; CHAMPMARTIN, Stéphane; AMBARI, Abdelhak
The objective of this paper is to provide a numerical simulation and an experimental study in order to assess stiffness and damping characteristics of thrust air bearings with multiple orifices. Finite element modeling is used to solve the non-linear Reynolds equation while taking into account the movement equation for the bearing. The numerical results obtained show that performance characteristics are related to bearing design type. An experimental investigation allows us to analyze the behavior of thrust air bearings with several orifices as well as that of groove or porous material bearings. Frequency response measurements have been realized in order to compare the dynamic properties of the different bearings. The frequency responses obtained demonstrate that air bearings with multiple orifices have a damping higher than the other types in certain conditions. Air bearings with multiple orifices offer many advantages from a dynamic point of view. Their performance may be characterized not only by flow conditions but also by the number or diameter of the orifices in the bearing surface.
http://dx.doi.org/10.1016/j.ijmecsci.2013.03.006
Tue, 01 Jan 2013 00:00:00 GMThttp://hdl.handle.net/10985/84802013-01-01T00:00:00ZCHARKI, AbdérafiDIOP, KhadimCHAMPMARTIN, StéphaneAMBARI, AbdelhakThe objective of this paper is to provide a numerical simulation and an experimental study in order to assess stiffness and damping characteristics of thrust air bearings with multiple orifices. Finite element modeling is used to solve the non-linear Reynolds equation while taking into account the movement equation for the bearing. The numerical results obtained show that performance characteristics are related to bearing design type. An experimental investigation allows us to analyze the behavior of thrust air bearings with several orifices as well as that of groove or porous material bearings. Frequency response measurements have been realized in order to compare the dynamic properties of the different bearings. The frequency responses obtained demonstrate that air bearings with multiple orifices have a damping higher than the other types in certain conditions. Air bearings with multiple orifices offer many advantages from a dynamic point of view. Their performance may be characterized not only by flow conditions but also by the number or diameter of the orifices in the bearing surface.Levitating spherical particle in a slightly tapered tube at low Reynolds numbers: Application to the low-flow rate rotameters
http://hdl.handle.net/10985/8482
Levitating spherical particle in a slightly tapered tube at low Reynolds numbers: Application to the low-flow rate rotameters
CHAMPMARTIN, Stéphane; AMBARI, Abdelhak; CHHABRA, Raj P.
In this study, a theoretical framework is developed to predict the equilibrium conditions of a non-neutrally buoyant sphere placed in a vertical conical tube as encountered in liquid rotameters. The analysis presented herein is applicable for a sphere heavier than the surrounding fluid, situated on the axis of a slightly tapered tube. The sphere is subject to the laminar flow conditions with the Reynolds numbers ranging between the Stokes type regimes up to values corresponding to slightly inertial regimes. In this work, we assume that the aperture angle of the tube is small and that the drag force is mainly due to the dissipation located in the gap between the tube and the sphere. Under these conditions, it is possible to consider the tube as locally cylindrical and we can use the results previously obtained for the correction factor of the Stokes force on a sphere subject to a Poiseuille flow in a tube of constant cross-section. We obtain an equation relating the flow rate to the vertical position of the sphere in the tube and the validity of this analysis is demonstrated by applying it to a commercially available rotameter. The present study provides a simple but sound theoretical method to calibrate such flowmeters.
Sun, 01 Jan 2012 00:00:00 GMThttp://hdl.handle.net/10985/84822012-01-01T00:00:00ZCHAMPMARTIN, StéphaneAMBARI, AbdelhakCHHABRA, Raj P.In this study, a theoretical framework is developed to predict the equilibrium conditions of a non-neutrally buoyant sphere placed in a vertical conical tube as encountered in liquid rotameters. The analysis presented herein is applicable for a sphere heavier than the surrounding fluid, situated on the axis of a slightly tapered tube. The sphere is subject to the laminar flow conditions with the Reynolds numbers ranging between the Stokes type regimes up to values corresponding to slightly inertial regimes. In this work, we assume that the aperture angle of the tube is small and that the drag force is mainly due to the dissipation located in the gap between the tube and the sphere. Under these conditions, it is possible to consider the tube as locally cylindrical and we can use the results previously obtained for the correction factor of the Stokes force on a sphere subject to a Poiseuille flow in a tube of constant cross-section. We obtain an equation relating the flow rate to the vertical position of the sphere in the tube and the validity of this analysis is demonstrated by applying it to a commercially available rotameter. The present study provides a simple but sound theoretical method to calibrate such flowmeters.Reliability of a Hydrodynamic Journal Bearing
http://hdl.handle.net/10985/10285
Reliability of a Hydrodynamic Journal Bearing
DIOP, Khadim; CHARKI, Abdérafi; CHAMPMARTIN, Stéphane; AMBARI, Abdelhak
Journal fluid bearings are widely used in industry due to their static and dynamic behavior and their very low coefficient of friction. The technical requirements to improve the new technologies design are increasingly focused on the indicators of dependability of systems and machines. Then, it is necessary to develop a methodology to study the reliability of bearings in order to improve and to evaluate their design quality. Few works are referenced in literature concerning the estimation of the reliability of fluid journal bearings. This paper deals with a methodology to study the failure probability of a hydrodynamic journal bearing. An analytical approach is proposed to calculate static characteristics in using the Reynolds equation. The commonly methods used in structural reliability such as FORM (First Order Reliability Method), SORM (Second Order Reliability Method) and Monte Carlo are developed to estimate the failure probability. The function of performance bounding two domains (domain of safety and domain of failure) is estimated for several geometrical configurations of a hydrodynamic journal bearing (long journal bearings with the hypotheses of Sommerfeld, Gümbel and Reynolds, and a short journal bearing with the hypothesis of Gümbel).
Thu, 01 Jan 2015 00:00:00 GMThttp://hdl.handle.net/10985/102852015-01-01T00:00:00ZDIOP, KhadimCHARKI, AbdérafiCHAMPMARTIN, StéphaneAMBARI, AbdelhakJournal fluid bearings are widely used in industry due to their static and dynamic behavior and their very low coefficient of friction. The technical requirements to improve the new technologies design are increasingly focused on the indicators of dependability of systems and machines. Then, it is necessary to develop a methodology to study the reliability of bearings in order to improve and to evaluate their design quality. Few works are referenced in literature concerning the estimation of the reliability of fluid journal bearings. This paper deals with a methodology to study the failure probability of a hydrodynamic journal bearing. An analytical approach is proposed to calculate static characteristics in using the Reynolds equation. The commonly methods used in structural reliability such as FORM (First Order Reliability Method), SORM (Second Order Reliability Method) and Monte Carlo are developed to estimate the failure probability. The function of performance bounding two domains (domain of safety and domain of failure) is estimated for several geometrical configurations of a hydrodynamic journal bearing (long journal bearings with the hypotheses of Sommerfeld, Gümbel and Reynolds, and a short journal bearing with the hypothesis of Gümbel).