11.07.2015 Views

Diffusion in Deforming Porous Media - Department of Mathematics

Diffusion in Deforming Porous Media - Department of Mathematics

Diffusion in Deforming Porous Media - Department of Mathematics

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

DEFORMING POROUS MEDIA 13which is the exact model <strong>of</strong> the f<strong>in</strong>e-scale structure. It is supplemented with appropriate<strong>in</strong>itial and boundary conditions, and then it follows that there is a uniquesolution p ε 1 (x, t), pε 2 (x, t), uε (x, t), x ∈ Ω for each ε > 0. Moreover, the threesequences <strong>of</strong> functions have two-scale limits as ε → 0, and these are given by thetriple <strong>of</strong> functions p 1 (x, t), P 2 (x, y, t), U(x, y, t), x ∈ Ω, y ∈ Y 2 . Then we showthat this triple is a solution <strong>of</strong> a Darcy-Biot distributed microstructure systemwhich we now describe.The global flow <strong>in</strong> the porous structure is determ<strong>in</strong>ed on the macro-scale by themacroscopic Darcy equation˜c 1 ṗ 1 (x, t) −∂∫A 1 ∂p 1 (x, t) ∂P 2 (x, s, t)ij + κ 2 ds = F 1 (x, t) , x ∈ Ω ,∂x j ∂x i Γ 12∂nwhere ∫ ∂PΓ 12κ 2 (x,s,t)2 ds is the exchange term represent<strong>in</strong>g the flow <strong>in</strong>to the local∂n<strong>in</strong>clusion Y2 ε at the po<strong>in</strong>t x ∈ Ω. The flow and deformation with<strong>in</strong> the re-scaledare given by the local Biot systemY ε2c 2 ˙ P 2 (x, y, t) − ∇ y · κ 2 ∇ y P 2 (x, y, t) + b 0 ∇ y · ˙U(x, y, t) = F 2 (x, y, t) , y ∈ Y 2 ,P 2 (x, s, t) = p 1 (x, t) , s ∈ Γ 12 ,P 2 (x, s, t) and κ 2 ∇ y P 2 (x, s, t) are Y -periodic on Γ 22 ,ρÜi(x, y, t) − ∂∂y j(σyij (U(x, y, t)) − δ ijb 0 P 2 (x, y, t) ) = 0 , y ∈ Y 2 ,U(x, s, t) and σ y ij (U(x, s, t))n j − b 0 P 2 (x, s, t)n i are Y -periodic on Γ 22 ,U(x, s, t) = 0 , s ∈ Γ 12 .The subscript y on the gradient <strong>in</strong>dicates that the derivative is taken with respectto the local variable y. The solution P 2 (·, ·), U(·, ·) <strong>of</strong> the local system depends onthe global pressure p 1 (·) at the po<strong>in</strong>t x ∈ Ω. Because <strong>of</strong> the small size <strong>of</strong> the cells,this pressure is assumed to be well approximated by the “constant” value p 1 (x, t)on the <strong>in</strong>terface Γ ε 12. Note that if the deformation is suppressed <strong>in</strong> this system, i.e.,if U(x, y, t) = 0, then this is precisely the model <strong>of</strong> Arbogast et al [2] for s<strong>in</strong>glephase flow <strong>in</strong> a doubly-porous medium.This Darcy-Biot model is a very special case, <strong>in</strong>tended only to suggest the structure<strong>of</strong> the limit<strong>in</strong>g <strong>in</strong>itial-boundary-value problems that arise, and the micromodelsthat come from the particular applications always lead to considerably morecomplicated distributed microstructure models. One can use Biot systems for eachcomponent and scale the parameters for each component <strong>in</strong> a wide variety <strong>of</strong> ways,the choice be<strong>in</strong>g dependent on the situation. Also, one can start with a Biot systemfor the structure coupled to a fluid flow model either <strong>of</strong> Stokes type or <strong>of</strong> slightlycompressible flow type and then <strong>in</strong>vestigate the limit<strong>in</strong>g form <strong>of</strong> the composite forvarious scal<strong>in</strong>gs <strong>of</strong> the parameters. Similar Biot-Biot models have been constructedfor the mechanics <strong>of</strong> s<strong>of</strong>t biological tissues. These are based on the hypothesis thattissue can be regarded as a composite cellular poroelastic material composed <strong>of</strong> a

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!