Applied Statistical Mechanics Lecture Note - 13 Molecular Dynamics Simulation

Similar documents
Thermodynamics of solids 4. Statistical thermodynamics and the 3 rd law. Kwangheon Park Kyung Hee University Department of Nuclear Engineering

PHYS 705: Classical Mechanics. Derivation of Lagrange Equations from D Alembert s Principle

Mechanics Physics 151

Energy in Closed Systems

V. Principles of Irreversible Thermodynamics. s = S - S 0 (7.3) s = = - g i, k. "Flux": = da i. "Force": = -Â g a ik k = X i. Â J i X i (7.

Set of square-integrable function 2 L : function space F

Scalars and Vectors Scalar

Physics 11b Lecture #2. Electric Field Electric Flux Gauss s Law

Multistage Median Ranked Set Sampling for Estimating the Population Median

Chapter 8. Linear Momentum, Impulse, and Collisions

24-2: Electric Potential Energy. 24-1: What is physics

Stellar Astrophysics. dt dr. GM r. The current model for treating convection in stellar interiors is called mixing length theory:

Chapter 23: Electric Potential

PHY126 Summer Session I, 2008

Large scale magnetic field generation by accelerated particles in galactic medium

Contact, information, consultations

PHYS Week 5. Reading Journals today from tables. WebAssign due Wed nite

Part V: Velocity and Acceleration Analysis of Mechanisms

Dynamics of Rigid Bodies

Tian Zheng Department of Statistics Columbia University

Review. Physics 231 fall 2007

Learning the structure of Bayesian belief networks

E For K > 0. s s s s Fall Physical Chemistry (II) by M. Lim. singlet. triplet

Notes on Analytical Dynamics

Physics 207 Lecture 16

ALL QUESTIONS ARE WORTH 20 POINTS. WORK OUT FIVE PROBLEMS.

Integral Vector Operations and Related Theorems Applications in Mechanics and E&M

Physics 2A Chapter 11 - Universal Gravitation Fall 2017

One-dimensional kinematics

Review of Vector Algebra and Vector Calculus Operations

PHYS 705: Classical Mechanics. Newtonian Mechanics

Lecture 14: Forces and Stresses

COMPLEMENTARY ENERGY METHOD FOR CURVED COMPOSITE BEAMS

Molecular Dynamics and Monte Carlo Methods

Implicit Integration Henyey Method

4.4 Continuum Thermomechanics

Some Approximate Analytical Steady-State Solutions for Cylindrical Fin

CSU ATS601 Fall Other reading: Vallis 2.1, 2.2; Marshall and Plumb Ch. 6; Holton Ch. 2; Schubert Ch r or v i = v r + r (3.

DYNAMICS VECTOR MECHANICS FOR ENGINEERS: Kinematics of Rigid Bodies in Three Dimensions. Seventh Edition CHAPTER

NVU dynamics. III. Simulating molecules at constant potential energy

MOLECULAR DYNAMICS ,..., What is it? 2 = i i

Slide 1. Quantum Mechanics: the Practice

Army Ants Tunneling for Classical Simulations

2/24/2014. The point mass. Impulse for a single collision The impulse of a force is a vector. The Center of Mass. System of particles

Test 1 phy What mass of a material with density ρ is required to make a hollow spherical shell having inner radius r i and outer radius r o?

Capítulo. Three Dimensions

EE 5337 Computational Electromagnetics (CEM)

Physics 607 Exam 1. ( ) = 1, Γ( z +1) = zγ( z) x n e x2 dx = 1. e x2

COLLEGE OF FOUNDATION AND GENERAL STUDIES PUTRAJAYA CAMPUS FINAL EXAMINATION TRIMESTER /2017

Density Functional Theory I

CFD Investigations of Spatial Arc Kinetic Influence on Fuel Burning- Out in the Tornado Combustor

Physics 181. Particle Systems

Chapter I Matrices, Vectors, & Vector Calculus 1-1, 1-9, 1-10, 1-11, 1-17, 1-18, 1-25, 1-27, 1-36, 1-37, 1-41.

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE

A Brief Guide to Recognizing and Coping With Failures of the Classical Regression Assumptions

LINEAR MOMENTUM. product of the mass m and the velocity v r of an object r r

Event Shape Update. T. Doyle S. Hanlon I. Skillicorn. A. Everett A. Savin. Event Shapes, A. Everett, U. Wisconsin ZEUS Meeting, October 15,

Rotary motion

Electron density: Properties of electron density (non-negative): => exchange-correlation functionals should respect these conditions.

1. Starting with the local version of the first law of thermodynamics q. derive the statement of the first law of thermodynamics for a control volume

THE MANY-BODY PROBLEM

Chapter IV Vector and Tensor Analysis IV.2 Vector and Tensor Analysis September 29,

The Unique Solution of Stochastic Differential Equations With. Independent Coefficients. Dietrich Ryter.

coordinates. Then, the position vectors are described by

Supersymmetry in Disorder and Chaos (Random matrices, physics of compound nuclei, mathematics of random processes)

Instantaneous velocity field of a round jet

Machine Learning. Spectral Clustering. Lecture 23, April 14, Reading: Eric Xing 1

Nuclear Chart. Takashi NAKATSUKASA Theoretical Nuclear Physics Laboratory RIKEN Nishina Center. Real-space, real-time approaches ) Few-body model

UNIT10 PLANE OF REGRESSION

8 Baire Category Theorem and Uniform Boundedness

3. A Review of Some Existing AW (BT, CT) Algorithms

Complex atoms and the Periodic System of the elements

CS649 Sensor Networks IP Track Lecture 3: Target/Source Localization in Sensor Networks

CE 530 Molecular Simulation

Fundamental principles

1. A body will remain in a state of rest, or of uniform motion in a straight line unless it

Physics 202, Lecture 2. Announcements

A. Thicknesses and Densities

Correspondence Analysis & Related Methods

P 365. r r r )...(1 365

Minimization of the Free Energy under a Given Pressure by Natural Iteration Method

User-friendly model of heat transfer in. preheating, cool down and casting

gravity r2,1 r2 r1 by m 2,1

Physical & Interfacial Electrochemistry 2013

LASER ABLATION ICP-MS: DATA REDUCTION

Complementi di Fisica Lecture 19

CHAPTER 6. LAGRANGE S EQUATIONS (Analytical Mechanics)

11. Dynamics in Rotating Frames of Reference

Rotating Disk Electrode -a hydrodynamic method

19 The Born-Oppenheimer Approximation

Module 1 : The equation of continuity. Lecture 1: Equation of Continuity

Physics Exam II Chapters 25-29

THE VIBRATIONS OF MOLECULES II THE CARBON DIOXIDE MOLECULE Student Instructions

Chapter Fifiteen. Surfaces Revisited

Molecular Dynamic Simulations of Nickel Nanowires at Various Temperatures

Multipole Radiation. March 17, 2014

12. The Hamilton-Jacobi Equation Michael Fowler

Mechanics Physics 151

Chapter IV Vector and Tensor Analysis IV.2 Vector and Tensor Analysis September 23,

The Feynman path integral

Transcription:

Appled Statstcal Mechancs Lectue Note - 3 Molecula Dynamcs Smulaton 고려대학교화공생명공학과강정원

Contents I. Basc Molecula Dynamcs Smulaton Method II. Popetes Calculatons n MD III. MD n Othe Ensembles

I. Basc MD Smulaton -MC vs. MD MC MD Pobablstc smulaton technque Lmtatons eque the knowledge of an equlbum dstbuton goous samplng of lage numbe of possble phase-space gves only confguatonal popetes (not dynamc popetes!) Detemnstc smulaton technque Fully numecal fomalsm numecal soluton of N-body system A d N A( d N N exp N { βu ( )} N { βu ( )} )exp

I. Basc MD Smulaton - The Idea Follow the exactly same pocedue as eal expements Pepae sample pepae N patcles solve equaton of motons Connect sample to measung nstuments (e.g. themomete, vscomete, ) afte equlbaton tme, actual measuement begns Measue the popety of nteest fo a cetan tme nteval aveage popetes Example : measuement of tempeatue α mv k B T T ( t) mv ( t) k N B f

I. Basc MD Smulaton - Equaton of Moton Classcal Newton s equaton of moton Thee fomulaton Newtonan Lagangan Hamltonan Hamltonan pefeed fo many-body systems soluton of N dffeental equatons t m t F p p Soluton methods : Fnte Dffeence Method j j j z y x z y x p p p ),, ( ),, ( F F p p

I. Basc MD Smulaton - Velet Algothm Velet (967) : Vey smple, effcent and popula algothm 3 4 m m 3! 3 4 m m 3! ( t + δt) () t + p() t δt + F() t δt + () t δt + O( δt ) ( t δt) () t p() t δt + F() t δt () t δt + O( δt ) m 4 ( t + δt) + ( t δt) ( t) + F( t) δt + O( δt ) m 4 ( t + δt) ( t) ( t δt) + F( t) δt + O( δt ) featue : update wthout calculatng momentum (p)

I. Basc MD Smulaton - Leapfog Algothm Hockeny (970), Potte (97) Half-step leap-fog algothm Mathematcally equvalent to Velet algothm m ( t + δt) ( t) + v( t + δt) δt v( t + δt) v( t δt) + F( t) δt ( t + δt) ( t) + ( t t) v δ + F( t) δtδt m

. Popetes Calculaton n MD - Eneges Potental enegy Can be calculated dung foce calculaton Knetc enegy K mv

. Popetes Calculaton n MD - Pessues In an MD smulaton, calculaton of pessue usng tenso notaton s not the most effcent method. Fo homogeneous systems, thee s smple way to calculate pessue (Ivng and Kkwood, 950) Knetc deal gas tem Confguatonal called Val P ρk V B T + m V v ( t) v j ( t) + j ( t) F j j ( t) j ( t) F j ( t) Calculate when foce update Calculate when velocty update

. Popetes Calculaton n MD - Tanspot Popetes Appoaches fo tanspot popetes Method : NEMD (Non-equlbum Molecula Dynamcs) Contnuous addton and emoval of conseved quanttes Gves hgh sgnal-to-nose ato (good statstcs) Method : Equlbum molecula dynamcs Stat wth ansotopc confguaton of mass, momentum and enegy Obseve natual fluctuatons and dsspaton of mass, momentum and enegy Poo sgnal-to-nose aton (poo statstcs) All tanspot popetes can be measued at once

. Popetes Calculaton n MD - Tanspot Popetes Dffeental Balance Equaton Mass Enegy Momentum c(, t) t + j 0 T (, t) Dv(, t) c p + q 0 ρ + τ 0 t Dt Consttutve Equatons Fck s Law Foue s Law Newton s Law j D c q k T τ v ρv ) xy y ( x

. Popetes Calculaton n MD - Tanspot Popetes Pupose : Obtan tanspot coeffcent by molecula smulaton Not that the laws ae only appoxmaton that apply not-too-lage gadents In pncple tansfe coeffcents depends on c, T and v Geen-Kubo Relaton Relaton between tanspot popetes and ntegal ove tme-coelaton functon.

. Popetes Calculaton n MD - Tanspot Popetes Consde self-dffuson n a pue substance Consde how molecules ae dsspated when ntal confguatons ae gven as Dac delta functon Combne mass balance eqn. Wth Fck s Law Dmensonalty of gven system c(, t) D c(, t) 0 t B.C. c(, t) δ ( ) Soluton c(, t) (πdt) / d exp( ) Dt

. Popetes Calculaton n MD - Tanspot Popetes We do not need concentaton tself c(,t) - just dffuson coeffcent (D) < ( t) > c(, t) d c(, t) d < ( t) t > dd c(, t) D c(, t) t 0 < ( t) > ( ( t)) ddt N

. Popetes Calculaton n MD - Tanspot Popetes < ( t) > ( ( t)) ddt N Slope hee gves D t Plot of t vs. squae of taveled dstance gves dffuson coeffcent In 3D space, < > s mean squae dsplacement (MSD)

. Popetes Calculaton n MD - Tanspot Popetes An altenatve fomulaton usng velocty nstead of patcle poston t ( t) v( τ ) dτ 0 t () t v( τ ) dτ v( τ ) dτ t 0 0 dτ dτ v( τ ) v( τ ) 0 0 t τ dτ dτ v( τ ) v( τ ) 0 0 t τ dτ dτ v(0) v( τ τ ) 0 0 t t dτ dτ v(0) v( τ) 0 0 t ddt t dτ v(0) v( τ) 0 t t

. Popetes Calculaton n MD - Tanspot Popetes t ddt t dτ < v(0) v( τ ) > 0 t D d 0 dτ < v(0) v( τ ) > < v( 0) v( τ ) > < v( 0) v( τ ) >< v( t') v( t'' ) > Autocoelaton functon : popety dffeence between two adjacent tme steps Aea unde the cuve gves the value of self-dffuson coeffcent τ

. Popetes Calculaton n MD - Evaluaton of tme coelaton functons Tme consumng and eque a lot of stoage Altenatve method : FFT (Fast Foue Tansfom), Coase Ganng method A(t 0 ) A(t ) A(t ) A(t 3 ) A(t 4 ) A(t 5 ) A(t 6 ) A(t 7 ) A(t 8 ) A(t 9 ) A( Δt) A(0) A 0 A + A A + A A 3 + A 3 A 4 + A 4 A 5 + A 5 A 6 + A 6 A 7 + A 7 A 8 + A 8 A 9 +... A(t 0 ) A(t ) A(t ) A(t 3 ) A(t 4 ) A(t 5 ) A(t 6 ) A(t 7 ) A(t 8 ) A(t 9 ) A( Δt) A(0) A 0 A + A A 3 + A A 4 + A 3 A 5 + A 4 A 6 + A 5 A 7 + A 6 A 8 + A 7 A 9 +... A(t 0 ) A(t ) A(t ) A(t 3 ) A(t 4 ) A(t 5 ) A(t 6 ) A(t 7 ) A(t 8 ) A(t 9 ) A(5 Δ t) A(0) A 0 A 5 + A A 6 + A A 7 + A 3 A 8 + A 4 A 9 +...

. Popetes Calculaton n MD - Tanspot Popetes Zeo-shea vscosty η VkT 0 dτ < σ xy (0) σ xy ( τ ) > σ xy x y m v v + x j f y ( j ) j Themal Conductvty d λt > q mv + u( j ) VkT dt j dτ < q(0) q( τ ) 0

. Popetes Calculaton n MD - Radal Dstbuton Functon Tme aveaged value of numbe densty Ensemble aveaged numbe densty g ( ) ρ ( ) ρ V g ( ) < δ( j ) N j > Just count the numbe of molecules wthn a ange

3. MD n Othe Ensembles - Constants Wth pope choce of g(), we can calculate useful themodynamc popetes Intenal enegy Pessue Chemcal Potental U c P 0 πnρ φ( ) g( ) d ρkt πρ 3 0 dφ( ) 3 g( ) d d 3 ρλ π dv ' μ kt ln N ' q nt 3 N V V 0 dφ( ) d g( ) 3 d

3. MD n Othe Ensembles Constants Hamltonan fomulaton Consevaton of knetc + potental enegy H K + U (N,V,E) ensemble Cannot be appled to othe ensemble constant T, constant P, fo example we can keep const T whle H s constant dstbuton of K and U Two types of constants Holonomc constants : may be ntegated out of equaton of moton Nonholonomc constants : non-ntegable (nvolves veloctes) Tempeatue, pessue, stess,

3. MD n Othe Ensembles Constants Foce momentum tempeatue to eman constant One (bad) appoach at each tme step scale momenta to foce K to desed value advance postons and momenta apply p new λp wth λ chosen to satsfy epeat equatons of moton ae evesble tanston pobabltes cannot satsfy detaled balance does not sample any well-defned ensemble

3. MD n Othe Ensembles Constants Gauss pncple of least constants Gaussan constants : petubatve foce ntoduced nto the equaton of moton mnmzes the devaton to classcal tajectoes of patcles fom the unpetubed tajectoes Consde a functon f, a functon of patcle acceleaton f ( ) m F m f0 : nomal Newtonan equaton of moton othewse, constaned non-newtonan equaton of moton Gauss pncple : physcal acceleaton f to be mnmum ( f ( ) ζ g( )) 0 ζ : Lagangn (Gauss) Multple

3. MD n Othe Ensembles Constants Constant Tempeatue constants 0 ),, ( m dt dg t g 0 3 ), ( NkT m t G ( ) 0 ) ( ) ( g f ζ 0 j j j j j j j m m m F m m F ζ Newtonan Constant foce

3. MD n Othe Ensembles Constants Modfed equaton of moton m m F p p ζ m F ζ one of good appoach, but tempeatue s not specfed!

3. MD n Othe Ensembles Nose Themostat Extended Lagangan Equaton of Moton N m( s ) N Q LNose U( ) + s gkt lns p p L ms s L s Qs s Ks Qs U gktln s

3. MD n Othe Ensembles Nose-Hoove Themostat Equatons of moton p m p F ξp s ξ s N p ξ gkt Q m Integaton schemes pedcto-coecto algothm s staghtfowad Velet algothm s feasble, but tcky to mplement v F t-δt t t+δt At ths step, update of ξ depends on p; update of p depends on ξ p F ξp N p ξ Q m gkt