Source author record

Jean-Pierre Hulin

Jean-Pierre Hulin appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

7works
3topics
4close collaborators

Actions

Connect this record

Log in to claim

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

7 published item(s)

preprint2013arXiv

Transverse and lateral confinement effects on the oscillations of a free cylinder in a viscous flow

The different types of instabilities of free cylinders (diameter $D$, length $L$) have been studied in a viscous flow (velocity $U$) between parallel vertical walls of horizontal width $W$ at a distance $H$: the influence of the confinement parameters $D/H$ and $L/W$ has been investigated. As $D/H$ increases, there is a transition from stable flow to oscillations transverse to the walls and then to a fluttering motion with oscillations of the angle of the axis with respect to the horizontal. The two types of oscillations may be superimposed in the transition domain. The frequency $f$ of the transverse oscillations is independent of the lateral confinement $L/W$ in the range: 0.055 \le L/W \le 0.94$ for a given cylinder velocity $V_{cx}$ and increases only weakly with $V_{cx}$. These results are accounted for by assuming a 2D local flow over the cylinder with a characteristic velocity independent of $L/W$ for a given $V_{cx}$ value. The experimental values of $f$ are also independent of the transverse confinement $D/H$. The frequency $f_f$ of the fluttering motion is significantly lower than $f$: $f_f$ is also nearly independent of the cylinder diameter and of the flow velocity but decreases significantly as $L/W$ increases. The fluttering instability is then rather a 3D phenomenon involving the full length of the cylinder and the clearance between its ends and the side walls.

preprint2012arXiv

Oscillations and translation of a free cylinder in a confined flow

An oscillatory instability has been observed experimentally on an horizontal cylinder free to move and rotate between two parallel vertical walls of distance H; its characteristics differ both from vortex shedding driven oscillations and from those of tethered cylinders in the same geometry. The vertical motion of the cylinder, its rotation about its axis and its transverse motion across the gap have been investigated as a function of its diameter D, its density s, of the mean vertical velocity U of the fluid and of its viscosity. For a blockage ratio D/H above 0.5 and a Reynolds number Re larger then 14, oscillations of the rolling angle of the cylinder about its axis and of its transverse coordinate in the gap are observed together with periodic variations of the vertical velocity. Their frequency f is the same for the sedimentation of the cylinder in a static fluid (U = 0) and for a non-zero mean flow (U 6= 0). The Strouhal number St associated to the oscillation varies as 1/Re with : St.Re = 3 $\pm$ 0.15. The corresponding period 1/f is then independent of U and corresponds to a characteristic viscous diffusion time over a distance ~ D, implying a strong influence of the viscosity. These characteristics differ from those of vortex shedding and tethered cylinders for which St is instead roughly constant with Re and higher than here.

preprint2012arXiv

Stokes flow paths separation and recirculation cells in X-junctions of varying angle

Fluid and solute transfer in X-junctions between straight channels is shown to depend critically on the junction angle in the Stokes flow regime. Experimentally, water and a water-dye solution are injected at equal flow rates in two facing channels of the junction: Planar Laser Induced fluorescence (PLIF) measurements show that the largest part of each injected fluid "bounces back" preferentially into the outlet channel at the lowest angle to the injection; this is opposite to the inertial case and requires a high curvature of the corresponding streamlines. The proportion of this fluid in the other channel decreases from 50% at 90\degree to zero at a threshold angle. These counterintuitive features reflect the minimization of energy dissipation for Stokes flows. Finite elements numerical simulations of a 2D Stokes flow of equivalent geometry con rm these results and show that, below the threshold angle 33.8\degree recirculation cells are present in the center part of the junction and separate the two injected flows of the two solutions. Reducing further leads to the appearance of new recirculation cells with lower flow velocities.

preprint2009arXiv

Influence of flow confinement on the drag force on a static cylinder

The influence of confinement on the drag force $F$ on a static cylinder in a viscous flow inside a rectangular slit of aperture $h_0$ has been investigated from experimental measurements and numerical simulations. At low enough Reynolds numbers, $F$ varies linearly with the mean velocity and the viscosity, allowing for the precise determination of drag coefficients $λ_{||}$ and $λ_{\bot}$ corresponding respectively to a mean flow parallel and perpendicular to the cylinder length $L$. In the parallel configuration, the variation of $λ_{||}$ with the normalized diameter $β= d/h_0$ of the cylinder is close to that for a 2D flow invariant in the direction of the cylinder axis and does not diverge when $β= 1$. The variation of $λ_{||}$ with the distance from the midplane of the model reflects the parabolic Poiseuille profile between the plates for $β\ll 1$ while it remains almost constant for $β\sim 1$. In the perpendicular configuration, the value of $λ_{\bot}$ is close to that corresponding to a 2D system only if $β\ll 1$ and/or if the clearance between the ends of the cylinder and the side walls is very small: in that latter case, $λ_{\bot}$ diverges as $β\to 1$ due to the blockage of the flow. In other cases, the side flow between the ends of the cylinder and the side walls plays an important part to reduce $λ_{\bot}$: a full 3D description of the flow is needed to account for these effects.

preprint2009arXiv

Miscible transfer of solute in different types of rough fractures: from random to multiscale fracture walls heights

Miscible tracer dispersion measurements in transparent model fractures with different types of wall roughness are reported. The nature (Fickian or not) of dispersion is determined by studying variations of the mixing front as a function of the traveled distance but also as a function of the lateral scale over which the tracer concentration is averaged. The dominant convective dispersion mechanisms (velocity profile in the gap, velocity variations in the fracture plane) are established by comparing measurements using Newtonian and shear thinning fluids. For small monodisperse rugosities, front spreading is diffusive with a dominant geometrical dispersion (dispersion coefficient $D \propto Pe$) at low Péclet numbers $Pe$; at higher $Pe$ values one has either $D \propto Pe^2$ ({\it i.e.} Taylor dispersion) for obstacles of height smaller than the gap or $D \propto Pe^{1.35}$ for obstacles bridging the gap. For a self affine multiscale roughness like in actual rocks and a relative shear displacement $\vecδ$ of complementary walls, the aperture field is channelized in the direction perpendicular to $δ$. For a mean velocity $\vec{U}$ parallel to the channels, the global front geometry reflects the velocity contrast between them and is predicted from the aperture field. For $\vec{U}$ perpendicular to the channels, global front spreading is much reduced. Local spreading of the front thickness remains mostly controlled by Taylor dispersion except in the case of a very strong channelization parallel to $\vec U$.

preprint2006arXiv

Anisotropic self-affine properties of experimental fracture surfaces

The scaling properties of post-mortem fracture surfaces of brittle (silica glass), ductile (aluminum alloy) and quasi-brittle (mortar and wood) materials have been investigated. These surfaces, studied far from the initiation, were shown to be self-affine. However, the Hurst exponent measured along the crack direction is found to be different from the one measured along the propagation direction. More generally, a complete description of the scaling properties of these surfaces call for the use of the 2D height-height correlation function that involves three exponents zeta = 0.75, beta = 0.6 and z = 1.25 independent of the material considered as well as of the crack growth velocity. These exponents are shown to correspond to the roughness, growth and dynamic exponents respectively, as introduced in interface growth models. They are conjectured to be universal.

preprint2006arXiv

Low self-affine exponents of fracture surfaces of glass ceramics

The geometry of post mortem rough fracture surfaces of porous glass ceramics made of sintered glass beads is shown experimentally to be self-affine with an exponent zeta=0.40 (0.04) remarkably lower than the 'universal' value zeta=0.8 frequently measured for many materials. This low value of zeta is similar to that found for sandstone samples of similar micro structure and is also practically independent on the porosity phi in the range investigated (3% < phi < 26%) as well as on the bead diameter d and of the crack growth velocity. In contrast, the roughness amplitude normalized by d increases linearly with phi while it is still independent, within experimental error, of d and of the crack propagation velocity. An interpretation of this variation is suggested in terms of a transition from transgranular to intergranular fracture propagation with no influence, however, on the exponent zeta.