다중규모 모사

Multiscale simulation
for adsorption process development
[Ch. 7]

Major: Interdisciplinary program of the integrated biotechnology

Graduate school of bio- & information technology

Youngil Lim (N110), Lab. FACS

phone: +82 31 670 5200 (secretary), +82 31 670 5207 (direct)
Fax: +82 31 670 5445, mobile phone: +82 10 7665 5207
Email: limyi@hknu.ac.kr,
limyi@hknu.ac.kr, homepage:  http://hknu.ac.kr/~limyi/index.htm
http://hknu.ac.kr/~limyi/index.htm
Ch. 7. Computational methods on Mesoscopic/Macroscopic scale

• +Euclidean space in classical mechanics

• +Hilbert space in quantum theory
• +Sobolev space in fluid or solid mechanics
• +-space in thermodynamics
• +non-Euclidean curved space and time in general relativity
theory
• +Minkowski space in special relativity theory
 Equations linkage in CFD

Thermal Decomposition in a Parallel Plate

Reactor
 Introductions

outlet

R OH
R Heated cylinder R OH
O O
R
Inlet

• Chemical Engineering Module

– Fully coupled momentum, energy and mass transport
– Strategies for Multiphysics modeling

k
A F
 Coupled Three phenomena

(Re=4) (CA/C= 0.02)

Momentum Mass
(Navier-Stokes Equations) (Convection and Diffusion, Reaction)
Velocity, pressure Concentration
reaction rate =r(CA)
(exothermic or endothermic)
Density, viscosity
Thermal conductivity
Heat capacity
Reaction rate

Temperature

Energy
(Convection and Conduction, Heat transfer)
 모델링 1 – One-way coupled

• The momentum transport is independent of the energy and

mass transport
• The energy transport depends only on the momentum transport
• The mass transport depends on both the momentum transport
and the energy transport
• The heat of reaction, Q, is neglected

momentum Heat (2)

(1)

Mass (3)
Sequential solution approach
 Momentum : Navier-stokes equations


u
t
 
   u   u    u  u  p  F
T

u
  0

Energy : Convection and Conduction

T
 CP      kT   Q   C P u  T T
t

 E 
Mass : Convection and Diffusion
Ri  A exp ci
R T
ci  g 
     Di ci   Ri  u  ci Ci
t
 모델링 2 – Fully Coupled

• The momentum transport depends on the energy

transport
• The energy transport depends on both momentum and
mass transport
• The mass transport depends on both the momentum
transport and the energy transport

momentum Heat

Mass

Simultaneous solution approach

 Momentum : Navier-stokes equations


u
t
 
   u   u    u  u  p  F
T

u
  0

Energy : Convection and Conduction

T
 CP      kT   Q   C P u  T T
t

 E 
Mass : Convection and Diffusion
Ri  A exp ci
R T
ci  g 
     Di ci   Ri  u  ci Ci
t
 Simulation procedure in COMSOL

지배식설정 Model Navigator

(Momentum / Heat / Mass Transport)

CAD
구조체만들기 (1D, 2D or 3D)
Subdomain / Boundary Settings
물성값 / 경계값 설정 (Momentum / Heat / Mass Transport)

격자나누기 Mesh

계산 Solver (processing)

후처리 Post-processing
 Coupled Three phenomena

Momentum Mass
(Navier-Stokes Equations) (Convection and Diffusion, Reaction)
Velocity, pressure Concentration
reaction rate =r(CA)
(exothermic or endothermic)
Density, viscosity
Thermal conductivity
Heat capacity
Reaction rate

Temperature

Energy
(Convection and Conduction, Heat transfer)
 Results - velocity

Sequential Simultaneous
(constant parameters: , , k, Cp, Q) (time dependent parameters)
 Results - Temperature

Sequential Simultaneous
(constant parameters: , , k, Cp, Q) (time dependent parameters)
 Results - Concentration

Sequential Simultaneous
(constant parameters: , , k, Cp, Q) (time dependent parameters)
 External linage in Multiscale simulation for adsorption problems

Momentum : Navier-stokes equations


u
t
 T

   u   u    u  u  p  F u(t,x,r)

  0

Energy : Convection and Conduction

T T(t,x,r)
 CP      kT   Q   C P u  T
t
Mass : Convection and Diffusion
ci
     Di ci   Ri  u  ci Ci(t,x,r)
t
ni Ri   k m,i (ni*  ni )
  Ri
t ni(t,x,r)
n*  K  C
i i
 External linage in Multiscale simulation for adsorption problems

Momentum : Navier-stokes equations


u
t
 T

   u   u    u  u  p  F u(t,x,r)

  0

Energy : Convection and Conduction

T T(t,x,r)
 CP      kT   Q   C P u  T
t
Mass : Convection and Diffusion
ci
     Di ci   Ri  u  ci Ci(t,x,r)
t
ni Ri   k m,i (ni*  ni )
  Ri Q: adsorption heat
t ni(t,x,r)
n*  K  C
i i
Di: axial/radial diffusivity
km,i: mass transfer coeff.
K: adsorption heat
 External linage in Multiscale simulation for adsorption problems

MS MFS (?) FS PS
Molecular simulation Micro-flow dynamics simulation Fluid dynamics simulation Process simulation

Molecular force field Dissipative particle dynamics Momentum (Mm) Heat (H) Boundary conditions
(MFF, e.g., COMPASS) (DPD) (Node model, NM)

Grand Canonical Monte Carlo Computational fluid dynamics

Mass (M) 1D Mass balance
(GCMC) (CFD)
(1DM)

Internal coupling
NVT molecular dynamics • Velocity, u: Mm-H-M Internal coupling
(Canonical ensemble; NVT) Internal coupling (?) • Fluid density, (T): Mm-H • Operating conditions (, Qi): NM-
• Pressure drops, p: CFD • Viscosity, (T): Mm-H 1DM
Internal coupling • Pore diffusivity, Dp(T): CFD- • Heat capacity, Cp(T): H-H • Design parameters (Lc, Dc, Ni, b):
• Molecular formulation: Predefined input DPD • Thermal conductivity, k(T): H-H NM-1DM
(experiment) • Adsorption kinetics, Ri(T): H-M • Computational parameters (Nmesh, z,
External coupling(?)
• Particle Density, p: MFF • Adsorption isotherms, ni(Ci): • Adsorption heat, Q(Ri): M-H t):NM-1DM
• Specific surface area, A: MFF • Axial/radial diffusivity, Di(T): H-M External coupling (Model
MS-MFS
• Porosity, p: MFF • Pore size distribution: MS-MFS External coupling parameters)
• Adsorption heat, Q(T): MFF-GCMC • Adsorption isotherms, ni(T, Ci): MS- • Adsorption isotherms, ni(T, Ci): MS-
• Adsorption isotherms, ni(T, pi): MFF-GCMC FS PS
• Pore size distribution (PSD): MFF-GCMC • Adsorption heat, Q(T): MS-FS • Mass transfer coefficient, k(T): MS-
• Pore diffusivity, Dpore(T): MFF-NVT • Pore diffusivity, Dpore: MS-MFS-FS MFS-PS
External coupling • Geometry effects of equipments: FS- • Axial dispersion coefficient, Dax: FS-
• Force field parameters from DFT PS PS
(density functional theory) or ab-initio
methods

Material Studio (Accelrys Inc., COMSOL Multiphysics COMSOL Multiphysics FAST-Chrom/SMB

USA)
(Comsol Inc., Sweden) (CFD standard) (CESE PDE solver)
Forcite Plus & Sorption modules
Tools for calculation
 External linage in Multiscale simulation for adsorption problems
MS MFS FS PS
Molecular simulation Micro-flow dynamics simulation Fluid dynamics simulation Process simulation

DFT MC/MD Coarse-grained Lattice-Boltzmann CFD Process model

(GAUSSIAN) (Accelrys) particle simulation simulation (Comsol/Fluent) (FAST-Chrom/SMB)

Molecular force field Boundary conditions

Momentum
(MFF, e.g., COMPASS) Heat (H) (Node model, NM)
(Mm)
Dissipative particle dynamics
(DPD)
Grand Canonical Monte Carlo
Mass (M) 1D Mass balance
(GCMC)
(1DM)
Lattice Boltzmann Method
(LBM)
Internal coupling
NVT molecular dynamics • Velocity, u: Mm-H-M Internal coupling
(Canonical ensemble; NVT) Internal coupling (?) • Fluid density, (T): Mm-H • Operating conditions (, Qi): NM-
• Pressure drops, p: LBM • Viscosity, (T): Mm-H 1DM
Internal coupling • Pore diffusivity, Dp(T): LBM- • Heat capacity, Cp(T): H-H • Design parameters (Lc, Dc, Ni, b):
• Molecular formulation: Predefined input DPD • Thermal conductivity, k(T): H-H NM-1DM
(experiment) • Adsorption kinetics, Ri(T): H-M • Computational parameters (Nmesh, z,
External coupling(?)
• Particle Density, p: MFF • Adsorption isotherms, ni(Ci): • Adsorption heat, Q(Ri): M-H t):NM-1DM
• Specific surface area, A: MFF • Axial/radial diffusivity, Di(T): H-M External coupling (Model
MS-MFS
• Porosity, p: MFF • Pore size distribution: MS-MFS External coupling parameters)
• Adsorption heat, Q(T): MFF-GCMC • Adsorption isotherms, ni(T, Ci): MS- • Adsorption isotherms, ni(T, Ci): MS-
• Adsorption isotherms, ni(T, pi): MFF-GCMC FS PS
• Pore size distribution (PSD): MFF-GCMC • Adsorption heat, Q(T): MS-FS • Mass transfer coefficient, k(T): MS-
• Pore diffusivity, Dpore(T): MFF-NVT • Pore diffusivity, Dpore: MS-MFS-FS MFS-PS
External coupling • Geometry effects of equipments: FS- • Axial dispersion coefficient, Dax: FS-
• Force field parameters from DFT PS PS
(density functional theory) or ab-initio
methods

Material Studio (Accelrys Inc., COMSOL Multiphysics COMSOL Multiphysics FAST-Chrom/SMB

USA)
(Comsol Inc., Sweden) (CFD standard) (CESE PDE solver)
Forcite Plus & Sorption modules
Tools for calculation
 External linage in Multiscale simulation for adsorption problems
Di: axial diffusivity
Mass (process model) : Convection and Diffusion + node model
km,i: mass transfer coeff.
ci 1 b
     Di ci   
K: adsorption heat
Ri  u  ci Ci(t,x,r)
t b Column model
ni
 k m ,i ( K  Ci  ni ) ni(t,x,r)
t

 C
C( z ,0 )  Cinitial ( t  0 , z ) v  C
 L z 0  C in   Dax , t
IC    z z 0
BC  
n( z ,0 )  ninitial ( t  0 ,z )  C  0 , t
 z z  L
 c

Node model

Qin  Qoutnode
 Qinnode  Q feed  Qdesorbent  Qextract  Qraffinate
 node node node node
Cin  Qin  Cin  Qin  C feed  Q feed  Cdesorbent  Qdesorbent  Cin  Qextract  Cin  Qraffinate

Operating conditions: , Qdesorbent, Qextract, Qfeed, Qraffinate

 Ch 7. Mesoscale to Macroscale

Ch 7. Computational methods on
mesoscopic & macroscopic scale
– FEM
– Dissipative particle dynamics
– Theory of elasticity
– …