Sharp Interface Limit for the Navier Stokes Korteweg Equations

Σχετικά έγγραφα
Phase-Field Force Convergence

Example Sheet 3 Solutions

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Approximation of dynamic boundary condition: The Allen Cahn equation

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Uniform Convergence of Fourier Series Michael Taylor

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Other Test Constructions: Likelihood Ratio & Bayes Tests

Statistical Inference I Locally most powerful tests

D Alembert s Solution to the Wave Equation

Eulerian Simulation of Large Deformations

L p approach to free boundary problems of the Navier-Stokes equation

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Global nonlinear stability of steady solutions of the 3-D incompressible Euler equations with helical symmetry and with no swirl

ST5224: Advanced Statistical Theory II

ES440/ES911: CFD. Chapter 5. Solution of Linear Equation Systems

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

Homework 8 Model Solution Section

forms This gives Remark 1. How to remember the above formulas: Substituting these into the equation we obtain with

Parametrized Surfaces

Partial Differential Equations in Biology The boundary element method. March 26, 2013

Forced Pendulum Numerical approach

A bound from below for the temperature in compressible Navier-Stokes equations

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

Solutions to Exercise Sheet 5

Lecture 34 Bootstrap confidence intervals

Major Concepts. Multiphase Equilibrium Stability Applications to Phase Equilibrium. Two-Phase Coexistence

Solvability of Brinkman-Forchheimer equations of flow in double-diffusive convection

Heisenberg Uniqueness pairs

Reminders: linear functions

4.6 Autoregressive Moving Average Model ARMA(1,1)

Areas and Lengths in Polar Coordinates

5. Choice under Uncertainty

1. (a) (5 points) Find the unit tangent and unit normal vectors T and N to the curve. r(t) = 3cost, 4t, 3sint

The Pohozaev identity for the fractional Laplacian

Modelli Matematici. Gianni Gilardi. Cortona, giugno Transizione di fase solido solido in un sistema meccanico

Lecture 21: Properties and robustness of LSE

Areas and Lengths in Polar Coordinates

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

DERIVATION OF MILES EQUATION FOR AN APPLIED FORCE Revision C

the total number of electrons passing through the lamp.

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

GLOBAL WEAK SOLUTIONS TO COMPRESSIBLE QUANTUM NAVIER-STOKES EQUATIONS WITH DAMPING

Takeaki Yamazaki (Toyo Univ.) 山崎丈明 ( 東洋大学 ) Oct. 24, RIMS

6.3 Forecasting ARMA processes

The circle theorem and related theorems for Gauss-type quadrature rules

Iterated trilinear fourier integrals with arbitrary symbols

Section 8.3 Trigonometric Equations

Matrices and Determinants

( y) Partial Differential Equations

Discontinuous Hermite Collocation and Diagonally Implicit RK3 for a Brain Tumour Invasion Model

The Simply Typed Lambda Calculus

ΗΜΥ 220: ΣΗΜΑΤΑ ΚΑΙ ΣΥΣΤΗΜΑΤΑ Ι Ακαδημαϊκό έτος Εαρινό Εξάμηνο Κατ οίκον εργασία αρ. 2

Credit Risk. Finance and Insurance - Stochastic Analysis and Practical Methods Spring School Jena, March 2009

STABILITY FOR RAYLEIGH-BENARD CONVECTIVE SOLUTIONS OF THE BOLTZMANN EQUATION

ECE Spring Prof. David R. Jackson ECE Dept. Notes 2

C.S. 430 Assignment 6, Sample Solutions

Local Approximation with Kernels

Appendix to On the stability of a compressible axisymmetric rotating flow in a pipe. By Z. Rusak & J. H. Lee

Inverse trigonometric functions & General Solution of Trigonometric Equations

Acoustic Limit for the Boltzmann equation in Optimal Scaling

Divergence for log concave functions

ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =?

Every set of first-order formulas is equivalent to an independent set

Space-Time Symmetries

Bounding Nonsplitting Enumeration Degrees

1 String with massive end-points

wave energy Superposition of linear plane progressive waves Marine Hydrodynamics Lecture Oblique Plane Waves:

Solution Series 9. i=1 x i and i=1 x i.

Survival Analysis: One-Sample Problem /Two-Sample Problem/Regression. Lu Tian and Richard Olshen Stanford University

Geodesic Equations for the Wormhole Metric

MATH423 String Theory Solutions 4. = 0 τ = f(s). (1) dτ ds = dxµ dτ f (s) (2) dτ 2 [f (s)] 2 + dxµ. dτ f (s) (3)

Math 6 SL Probability Distributions Practice Test Mark Scheme

Global weak solution to the viscous two-phase model with finite energy

Higher Derivative Gravity Theories

Lifting Entry (continued)

b. Use the parametrization from (a) to compute the area of S a as S a ds. Be sure to substitute for ds!

Dr. D. Dinev, Department of Structural Mechanics, UACEG

Strain gauge and rosettes

EE512: Error Control Coding

Concrete Mathematics Exercises from 30 September 2016

Lifting Entry 2. Basic planar dynamics of motion, again Yet another equilibrium glide Hypersonic phugoid motion MARYLAND U N I V E R S I T Y O F

Approximation of distance between locations on earth given by latitude and longitude

Reflecting Brownian motion in two dimensions: Exact asymptotics for the stationary distribution

Arithmetical applications of lagrangian interpolation. Tanguy Rivoal. Institut Fourier CNRS and Université de Grenoble 1

Second Order RLC Filters

Fractional Colorings and Zykov Products of graphs

Math 446 Homework 3 Solutions. (1). (i): Reverse triangle inequality for metrics: Let (X, d) be a metric space and let x, y, z X.

Thermodynamics of the motility-induced phase separation

Congruence Classes of Invertible Matrices of Order 3 over F 2

Problem Set 9 Solutions. θ + 1. θ 2 + cotθ ( ) sinθ e iφ is an eigenfunction of the ˆ L 2 operator. / θ 2. φ 2. sin 2 θ φ 2. ( ) = e iφ. = e iφ cosθ.

Markov chains model reduction

Μονοβάθμια Συστήματα: Εξίσωση Κίνησης, Διατύπωση του Προβλήματος και Μέθοδοι Επίλυσης. Απόστολος Σ. Παπαγεωργίου

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8 questions or comments to Dan Fetter 1

MA 342N Assignment 1 Due 24 February 2016

Mathematical theory of compressible viscous. Eduard Feireisl and Milan Pokorný

Homework 3 Solutions

Transcript:

Sharp Interface Limit for the Navier Stokes Korteweg Equations Albert-Ludwigs-Universität Freiburg Johannes Daube (Joint work with H. Abels, C. Kraus and D. Kröner) 11th DFG-CNRS Workshop Micro-Macro Modeling and Simulations of Liquid-Vapour Flows Paris, March 3, 2016

Organisation Introduction The Diffuse Interface Model The Sharp Interface Model The Sharp Interface Limit March 3, 2016 Johannes Daube SIL for the NSK Equations 2 / 21

Diffuse and Sharp Interface Models Diffuse Interface Models interface of non-zero thickness (diffuse interface) system of PDEs on whole domain describes motion of fluid smooth order parameter (here: density) indicates different phases Sharp Interface Models infinitely thin (sharp) interface separates different fluid phases system of PDEs in each bulk phase describes motion of fluid coupling by boundary conditions at interface jump discontinuities at interface β 1 ε 0 Sharp Interface Limits relates diffuse to sharp interface models letting in diffuse models the interface thickness tend to zero leads to corresponding sharp interface models β 1 β 2 March 3, 2016 Johannes Daube SIL for the NSK Equations 3 / 21

Navier Stokes Korteweg Equation (Diffuse Interface Model) extension of the compressible Navier Stokes equations (Korteweg 1901) diffuse interface model for liquid vapour flows which allows for phase transitions a small parameter ε > 0 represents the thickness of an interfacial area, where phase transitions occur phase field like scaling (Hermsdörfer, Kraus and Kröner 2011) (DIM) ε t ρ ε + div(ρ ε v ε ) = 0 in [0,T], t (ρ ε v ε ) + div(ρ ε v ε v ε ) + 1 ε p(ρ ε) = 2div(µ(ρ ε )Dv ε ) + εγρ ε ρ ε in [0,T], where ρ ε density, v ε velocity, p(ρ ε ) (non-monotone) pressure, µ(ρ ε ) viscosity. Boundary conditions: ρ ε ν = 0 and v ε = 0 on [0,T]. Initial conditions: ρ ε (,0) = ρ ε (i) and v ε (,0) = v ε (i) in. March 3, 2016 Johannes Daube SIL for the NSK Equations 4 / 21

Pressure and Double Well Potential non-monotone pressure function p(ρ) = ρ W (ρ) W(ρ) (normalized) double-well potential W(ρ) = W(ρ) l(ρ), where l Maxwell line and 0 < β 1 < β 2 Maxwell points Note: p (ρ) = ρ W (ρ) = ρw (ρ) = p(ρ) = p (ρ) ρ = ρw (ρ) ρ p Ρ W Ρ W Ρ vapour Α1 Α2 liquid Ρ Β1 Α1 Α2 Β2 Ρ Β1 Α1 Α2 Β2 Ρ Growth of the Double-Well Potential. For a = β 1 +β 2 2 and b = β 2 β 1 2, there exist constants C 1 > 0, C 2 (0,b) and p > 2 such that W (z) C 1 z a p 2 for all z R with z a b C 2. Prototypical example (p = 4): W(z) = (z β 1 ) 2 (z β 2 ) 2. March 3, 2016 Johannes Daube SIL for the NSK Equations 5 / 21

Static Case Motivation for phase field like scaling of the Navier Stokes Korteweg equations. Static case: velocity v ε 0, time independent; γ = 1 for simplicity. In the static case (DIM) ε reduces to 1 ε p(ρ ε) = ερ ε ρ ε (p (ρ) = ρw (ρ)) = ερ ε ρ ε = 1 ε p (ρ ε ) ρ ε = 1 ε ρ ε W (ρ ε ) ( ) 1 (assume ρ ε > 0) = ε W (ρ ε ) ε ρ ε = 0. Hence, there is a constant l ε R such that 1 ε W (ρ ε ) ε ρ ε = l ε. Euler Lagrange equation with Lagrange multiplier l ε of { } min E ε (ρ) : ρ H 1 1 (), ρ(x)dx = m, E ε (ρ) = ε W(ρ) + ε 2 ρ 2 dx, where m (β 1,β 2 ) (prescribed total mass). March 3, 2016 Johannes Daube SIL for the NSK Equations 6 / 21

Static Case: Heuristics Heuristic picture from minimization problem { 1 min ε W(ρ) + ε } 2 ρ 2 dx : ρ H 1 (), ρ(x)dx = m. 1 ε W(ρ)dx favours configurations close to pure phases β 1 and β 2 due to the double-well shape of W. minimizing 1 ε W(ρ)dx leads to phase separation. 1 2 ε ρ 2 dx penalizes occurrence of large interfaces, i.e., rapid transitions between {ρ β 1 } and {ρ β 2 }. March 3, 2016 Johannes Daube SIL for the NSK Equations 7 / 21

Aim (DIM) ε t ρ ε + div(ρ ε v ε ) = 0 in [0,T], t (ρ ε v ε ) + div(ρ ε v ε v ε ) + 1 ε p(ρ ε) = 2div(µ(ρ ε )Dv ε ) + εγρ ε ρ ε in [0,T], Sharp interface limit for (DIM) ε : (DIM) ε ε 0 Sharp interface model Main steps 1 Convergence/compactness results for solutions (ρ ε,v ε ) of (DIM) ε 2 Introduction/Justification of the sharp interface model 3 Limit functions of (ρ ε,v ε ) are solutions of the sharp interface model March 3, 2016 Johannes Daube SIL for the NSK Equations 8 / 21

(DIM) ε : Notion of Weak Solutions A couple (ρ ε,v ε ) ( L (0,T;H 1 ()) L (0,T;L p ()) ) L 2 (0,T;H0 1()n ) is a weak solution to (DIM) ε w.r.t. prescribed initial values (ρ ε (i),v ε (i) ), if Conservation of mass: For all ϕ C(0) ( [0,T)), it holds ρ ε t ϕ + ρ ε v ε ϕ dx dt + ρ ε (i) ϕ(0)dx = 0. Conservation of momentum: For all ψ C (0) ( [0,T))n, it holds ρ ε v ε t ψ + ρ ε v ε v ε : ψ + 1 ε p(ρ ε)div(ψ) 2µ(ρ ε )Dv ε : Dψ dx dt + ρ ε (i) v ε (i) ψ(0)dx = γε ρ ε ρ ε : ψ + 1 2 ρ ε 2 div(ψ) + ρ ε ρ ε div(ψ)dx dt. Energy inequality: Let Eε tot 1 (t) = ε W(ρ ε (t)) + ε 2 γ ρ ε(t) 2 + 1 2 ρ ε(t) v ε (t) 2 dx. For almost every t (0,T), it holds t Eε tot (t) + 2 µ(ρ ε ) Dv ε 2 dx ds Eε tot (0) assumption C. March 3, 2016 Johannes Daube SIL for the NSK Equations 9 / 21

(DIM) ε : Compactness of Weak Solutions Family of weak solutions (ρ ε,ε ε ) ε (0,1) to (DIM) ε with uniformly bounded initial energy. Theorem (Compactness of weak solutions to (DIM) ε ) There are functions ρ 0 C 0, 1 29 ([0,T];L 2 ()) L (0,T;BV(,{β 1,β 2 })) and v 0 L 2 (0,T;H 1 0 ()n ) such that (up to subsequences) ρ ε ρ 0 in C 0, 1 29 ([0,T];L 2 ()) and v ε v 0 weakly in L 2 (0,T;H 1 () n ). Limiting sharp interface: ρ 0 (t) BV(,{β 1,β 2 }) indicates phases (t) = {ρ 0 (t) = β 1 } and + (t) = {ρ 0 (t) = β 2 } of finite perimeter, interface Γ(t) = ( (t)). Convergence to a reasonable sharp interface model? March 3, 2016 Johannes Daube SIL for the NSK Equations 10 / 21

(SIM) 0 : Two-phase Incompressible Navier Stokes Equation with Surface Tension Sharp interface model is free boundary problem with unknown interface Γ : [0,T] R n such that = (t) Γ(t) + (t), velocity function v(t) : \ Γ(t) R n, pressure function p(t) : \ Γ(t) R satisfying β 1 t v + β 1 (v )v + p = 2div(µ(β 1 )Dv) in (t), β 2 t v + β 2 (v )v + p = 2div(µ(β 2 )Dv) in + (t), (Navier Stokes) div(v) = 0 in \ Γ(t), (incompressible phases) (SIM) 0 [v] + = 0 on Γ(t), (continuous velocity) (Navier Stokes) V = v ν on Γ(t), (pure transport of interface) [T] + ν = 2σκν on Γ(t). (Young Laplace law) T (v,p) = 2µ(β 1 )Dv pi and T + (v,p) = 2µ(β 2 )Dv pi stress tensors, [ ] + jump across Γ(t), V normal velocity of interface, σ surface tension constant, κ mean curvature. + suitable boundary conditions on and initial conditions March 3, 2016 Johannes Daube SIL for the NSK Equations 11 / 21

(SIM) 0 : Weak Solutions to Sharp Interface Model For (ρ,v) L (0,T;BV(,{β 1,β 2 })) ( L (0,T;L 2 σ ()) L 2 (0,T;H 1 0 ()n ) ) let (t) = {ρ(t) = β 1 } and Γ(t) = ( (t)) (density determines interface). Then (ρ,v) is a weak solution of (SIM) 0, if Transport equation: characteristic function χ(t) = ρ(t) β 2 β 1 β 2 of t χ + v χ = 0 in (0,T). of (t) is a distributional solution Energy inequality: For almost every t (0,T), it holds 1 t ρ(t) v(t) 2 dx + 2σH n 1 (Γ(t)) + 2 µ(ρ) Dv 2 dx ds 2 1 ρ (i) v (i) 2 dx + 2σH n 1 (Γ(0)). 2 Variational formulation: For all divergence free test functions ψ C0 ([0,T);C 0,σ ()), it holds ρv t ψ + ρv v : ψ 2µ(ρ)Dv : Dψ dx dt + ρ (i) v (i) ψ(0)dx = 2σ ν ν : ψ dh n 1 (x)dt. 0 Γ(t) March 3, 2016 Johannes Daube SIL for the NSK Equations 12 / 21

(SIM) 0 : Divergence Free Test Functions Variational formulation ρv t ψ + ρv v : ψ 2µ(ρ)Dv : Dψ dx dt + ρ (i) v (i) ψ(0)dx = 2σ ν ν : ψ dh n 1 (x)dt 0 Γ(t) uses divergence free test functions ψ C 0 ([0,T);C 0,σ ()) = { ψ [0,T) : ψ C 0 ( ( 1,T))n, div(ψ) = 0 }. +: Removes pressure function from weak formulation. Consequence : No passage to the limit of pressure term. -: Reconstruction of pressure function from weak formulation. March 3, 2016 Johannes Daube SIL for the NSK Equations 13 / 21

(SIM) 0 : Reconstruction of Pressure (for Smooth Solutions) Additional smoothness assumptions: interface: Γ = Γ(t) is space-time interface of class C 3,1. velocity: all derivatives of v exist in L 2 ( ) and L 2 ( + ), where ( (t) {t} ) and ( + (t) {t}). t (0,T) t (0,T) Note: No additional regularity across interface is assumed! Theorem (Reconstruction of pressure) For a sufficiently regular weak solution (ρ,v) to (SIM) 0, there exists a unique p L 2 ( (0,T)) with the following properties. 1 p ± L 2 (0,T;W 1,2 ( ± (t))). 2 For almost all t [0,T) it holds p(t)dx = 0. 3 p = β 1 t v + 2µ(β 1 )div(dv) β 1 (v )v a.e. in. 4 p = β 2 t v + 2µ(β 2 )div(dv) β 2 (v )v a.e. in +. 5 [p] + = 2[ µ(ρ)dvν ] + ν + 2σκ in L 2 (0,T;H 1 2 (Γ(t))). Jusitfies notion of weak solutions to (SIM) 0 March 3, 2016 Johannes Daube SIL for the NSK Equations 14 / 21

(DIM) ε (SIM) 0 : The Sharp Interface Limit Aim Weak solutions (ρ ε,v ε ) ε (0,1) of (DIM) ε ε 0 weak solution (ρ 0,v 0 ) of (SIM) 0. Recall: Compactness properties of weak solutions: ρ ε ρ 0 in C 0, 1 29 ([0,T],L 2 ()) with ρ 0 C 0, 1 29 ([0,T];L p ()) L (0,T;BV(,{β 1,β 2 })) v ε v 0 weakly in L 2 (0,T;H 1 () n ) with v 0 L 2 (0,T;H 1 0 ()n ) Question: (ρ 0,v 0 ) weak solution of (SIM) 0? March 3, 2016 Johannes Daube SIL for the NSK Equations 15 / 21

(DIM) ε (SIM) 0 : Mathematical Main Difficulties Task: Passing to the limit in the weak formulation. Momentum balance: Recall variational formulation ρ ε v ε t ψ + ρ ε v ε v ε : ψ + 1 ε p(ρ ε)div(ψ) 2µ(ρ ε )Dv ε : Dψ dx dt + ρ ε (i) v ε (i) ψ(,0)dx = γε ρ ε ρ ε : ψ + 1 2 ρ ε 2 div(ψ) + ρ ε ρ ε div(ψ)dx dt. for any ψ C (0) ( [0,T))n. Challenging terms in the passage to the limit ε 0: 1 T p(ρ ε )div(ψ)dx dt and ε ρ ε ρ ε : ψ dx dt. ε March 3, 2016 Johannes Daube SIL for the NSK Equations 16 / 21

(DIM) ε (SIM) 0 : Treatment of Pressure Term Weak formulation for (SIM) 0 takes into account divergence-free test functions ψ C0 ([0,T);C 0,σ ()) = 1 p(ρ ε )div(ψ)dx dt = 0. ε Resolves problem of passing pressure term to the limit! Variational formulation simplifies to for any ψ C 0 ([0,T);C 0,σ ()). ρ ε v ε t ψ + ρ ε v ε v ε : ψ dx dt + ρ ε (i) v ε (i) ψ(,0)dx = 2µ(ρ ε )Dv ε : Dψ γε ρ ε ρ ε : ψ dx dt March 3, 2016 Johannes Daube SIL for the NSK Equations 17 / 21

(DIM) ε (SIM) 0 : Treatment of Capillary Term Due to energy inequality, 1 ε W(ρ ε) + ε 2 γ ρ ε 2 weak- pre-compact in L 1 (0,T;C 0 ()). Problem: Identification of limit function. Solution: If 1 ε W(ρ ε) + ε 2 γ ρ ε 2 converges to appropriate surface measure. Then control of εγ ρ ε ρ ε : ψ dx dt. March 3, 2016 Johannes Daube SIL for the NSK Equations 18 / 21

(DIM) ε (SIM) 0 : The Sharp Interface Limit (Solving the Sharp Interface Model) Assumption: For every ϕ L 1 (0,T;C 0 ()) it holds ( lim 1ε W(ρ ε ) + ε ε 0 2 γ ρ ε 2) ϕ dx dt = 2σ ϕ dh n 1 (x)dt 0 Γ(t) with σ = β 2 γw(r) β 1 2 dr and Γ(t). Theorem (Convergence of capillary term) For every ψ C0 ([0,T);C 0,σ ()) it holds lim εγ ρ ε ρ ε : ψ dx dt = 2σ ν ν : ψ dh n 1 (x)dt. ε 0 0 Γ(t) Theorem (Solving the sharp interface model) The couple (ρ 0,v 0 ) is a weak solution of (SIM) 0. March 3, 2016 Johannes Daube SIL for the NSK Equations 19 / 21

Conclusion Navier Stokes Korteweg equations (diffuse interface model) notion of weak solution compactness properties Two-phase incompressible Navier Stokes equations with surface tension (sharp interface model) notion of weak solution consistency result Sharp interface limit sufficient condition: convergence of energy to associated surface measure open question: possible to weaken/remove additional assumption? March 3, 2016 Johannes Daube SIL for the NSK Equations 20 / 21

References Abels, H.: On generalized solutions of two-phase flows for viscous incompressible fluids. Interfaces Free Bound. 9, No. 1, 31 65 (2007). Chen, X.: Global asymptotic limit of solutions of the Cahn Hilliard equation. J. Differ. Geom. 44, No. 2, 262 311 (1996). Hermsdörfer, K., Kraus, C., Kröner, D.: Interface conditions for limits of the Navier Stokes Korteweg model. Interfaces Free Bound. 13, No. 2, 239 254 (2011). Korteweg, D. J.: Sur la forme que prennent les équations du mouvement des fluids si l on tient compte des forces capillaires causées par les variations de densite mais continues et sur la théorie de la capillarité dans l hypothèse d une variation continue de la densité. Arch. Néerl. (2) 6, 1 24 (1901). Thank you for your attention! March 3, 2016 Johannes Daube SIL for the NSK Equations 21 / 21