miet 2394 cfd lecture 11

Upload: cepong89

Post on 06-Jul-2018

242 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    1/52

    Computational FluidDynamics – Lecture 11

    Prof. Jiyuan Tu

    & Dr. Sherman C.P. Cheung

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    2/52

    RMIT University 2

    Some Advanced Topics in CFD

    Demands from real-world problems and industries

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    3/52

    RMIT University 3

    Turbulent Reacting Flows

    ※ Combustion

    Heavy modelling!

    Turbulent flow

    +Chemical reactions

    Persistent Flame

    Intermittent Flame

    Buoyant Plume

     A photographic image of a buoyant fire. 

    Complex Reacting Flow

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    4/52

    RMIT University 4

    Concept of olumetric Heat ource

    Heat

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    5/52

    RMIT University 5

    tec"ler#s $xperiment

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    6/52

    RMIT University 6

    isuali%ation of Computational model

    X

    Y

    Z

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    7/52

    RMIT University 7

    Results from olumetric Heat ource

    X

    Y

    Z

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    8/52

    RMIT University

    &ir Flow 'attern

    X

    Y

    Z

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    9/52

    RMIT University !

    elocity 'rofile at Doorway

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    10/52

    RMIT University "#

    Temperature 'rofile at Doorway

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    11/52

    RMIT University ""

    Temperature Distribution

    X

    Y

    Z

    T

    52#

    5#$42"

    4!6$42

    45$263

    473$64

    462$"#5

    45#$526

    43$!47

    427$36

    4"5$7!

    4#4$2""3!2$632

    3"$#53

    36!$474

    357$!5

    346$3"6

    334$737

    323$"5

    3""$57!

    3##

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    12/52

    RMIT University "2

    &dvantages

    (t is simple

    )ive reasonable prediction both in velocities and

    temperatures

    ave computational time

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    13/52

    RMIT University "3

    Disadvantages

    Fire*flame height should be obtained prior from

    experimental data or analytical approximation

    +,rong prediction at the flame

    .isinterpretation of flame structure

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    14/52

    RMIT University "4

    ,hy needs Combustion .odel

    (mprove prediction at near fire field

    ome of the Fire problems re/uire the shape of

    flame structure

    Flame spread along combustible materials

    Fire suppression by sprin"lers or water mists

    $ddy 0rea" 1p 2$013 or presumed 'DF

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    15/52

    RMIT University "5

    Fire Triangle

    Fuel

    Heat Oxygen

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    16/52

    RMIT University "6

    .ixed is 0urnt Concept

    Fuel 4xidant

    Combustion

    Turbulent .ixing

    'roduct

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    17/52

    RMIT University "7

    Flame hape by Combustion .odel

    X

    Y

    Z

    T

    "7##

    "6##

    "5##

    "4##

    "3##"2##

    ""##

    "###

    !##

    ##

    7##

    6##

    5##

    4##

    3##

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    18/52

    RMIT University "

    Two Compartment Fire

    3$6 m 3$6 m

    2$4 m

    %&en

    'n(

    Burn

    Room

    Adjacent Room

    )P* Burner

    2$# m

    #$ m

    +oor,ay

    $xperimental setup by 5ielsen and Fleischmann 267773

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    19/52

    RMIT University "!

    Temperatures at &d8oining Room

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    3## 35# 4## 45# 5##

    #

    #$5

    "

    "$5

    2

    2$5

    Meaurment

    !rediction "y #om"ution Model

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    20/52

    RMIT University 2#

    Temperatures at Doorway

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    3## 35# 4## 45# 5## 55# 6## 65##

    #$5

    "

    "$5

    2

    Meaurment

    !rediction "y #om"ution Model

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    21/52

    RMIT University 2"

    Temperatures above Fire ource

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    4## 6## ## "### "2## "4## "6## "## 2### 22## 24#

    #$5

    "

    "$5

    2

    $ncorrected experimental data

    !rediction %ith #om"ution Model

    #orrected experimental data (&' *)

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    22/52

    RMIT University 22

    Disadvantages

    $xtra e/uations involved in simulation

    9nowledge of the fuel Thermal Decomposition is

    re/uired 2Detail Chemistry3

    Complicated to implement

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    23/52

    RMIT University 23

     5eeds of Radiation and oot .odel

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    24/52

    RMIT University 24

    oot Formation 'rocess

    Describes fours chemical processes:

     5ucleation

    Coagulation

    urface )rowth

    4xidation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    25/52

    RMIT University 25

     5ucleation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    26/52

    RMIT University 26

    Coagulation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    27/52

    RMIT University 27

    urface )rowthinyl

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    28/52

    RMIT University 2

    4xidation464H

    C46

    H64

    C46

    C46

    H64

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    29/52

    RMIT University 2!

    Radiation and oot Contribution Revisit the two compartment fire experiment

    (ncorporate the radiation and soot model

    Compare the prediction with previous results

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    30/52

    RMIT University 3#

    Temperatures at &d8oining Room

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    3## 35# 4## 45# 5###

    #$5

    "

    "$5

    2

    2$5

    Meaurment

    !rediction %ith only gaeou radiation

    !rediction %ith gaeou and oot radiation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    31/52

    RMIT University 3"

    Temperatures at Doorway

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    3## 35# 4## 45# 5## 55# 6## 65##

    #$5

    "

    "$5

    2

    Meaurment!rediction %ithout radiation

    !rediction %ith gaeou and oot radiation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    32/52

    RMIT University 32

    Temperatures above Fire ource

    Temperature (K)

         H    e     i    g     h     t     (    m     )

    4## 6## ## "### "2## "4## "6## "## 2### 22## 24#

    #$5

    "

    "$5

    2

    $ncorrected experimental data!rediction %ithout radiation

    !rediction %ith gaeou and oot radiation

    #orrected experimental data (&' *)

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    33/52

    RMIT University 33

     A Buoyant FreeStanding Fire

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    34/52

    RMIT University 34

     Advance in Computational!odels( ") ;arge $ddy imulation 2 LES 3

    4nly resolve large eddies < sub-modeling of small eddies-   ubgrid scale turbulent viscosity

    -1nsteady simulation-Depending on grid si%e

    -(ndustrial applications possible

     RANS-LES  Coupling

    Reynolds &veraged-5~ 5ear-wall uses

    ~ (nternal domain uses ;$

    Hybrid Coupling

    SGS 

    T  µ 

    V ∆

      = - RANS    κ ε 

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    35/52

    RMIT University 35

    Bac#ward Facing Step

    +-. .olution /rom .tan/or( University

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    36/52

    RMIT University 36

     Advance in Computational!odels( "")Direct 5umerical imulation 2 DNS 3

    Resolve all scales of turbulent eddies

      For fundamental understanding of turbulence

    Challengesignificant grid numbers

    mall time step

    Higher order discreti%ation1n"nown initial < inlet 0C

    >? Re

     ?   t ∆ ↓↓

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    37/52

    RMIT University 37

    !ultip$ase Flows( ")

    ※ )as olid particle flow

    'hase@ 'hase6

      .ining industry

    'ollutants  $nvironment

    Coal fired power 

    DustA soilB

    $lectricity

    ※ )as Droplet particle flow

    &ir  ;i/uid

      5asal prayer 

      &erosol

    (C $ngines pray

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    38/52

    RMIT University 3

    !ultip$ase Flows( "")

    ※ 0ubble ,ater particle flow

      0oilers

      5uclear 'ower 

    Chemical reactor 

    &ir  ;i/uid

    ※  4il ,ater &ir olid particle flow

    'hase@ 6 >

    ※ &ir ,ater free surface

     5umerical methods

    ※$ulerian-$ulerian .ethod orTwo- fluid .ethod※$ulerian-;agrange .ethodor 'article trac"ing .ethod

    ※olume-of-fluid 2VOF 3Trac"ing free surface

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    39/52

    RMIT University 3!

    Typical Flow be$aviour inbubble columns

     

    Bu00ly Flo, 1a&Bu00ly /lo, .lu Flo, 1urnTur0ulent Flo,

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    40/52

    RMIT University 4#

    Bu00le 1oaleseneBu00le 1oalesene 0ubbles may +merge together

    forming larger bubbles

    Coalescence reduces number  of

     bubbles but increases si%e of

     bubbles

    estern Miian University

    -D p8

    D p

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    41/52

    RMIT University 4"

    Bu00le Brea9aeBu00le Brea9ae

    Revuelta et al$ 2##68 :$F$M$

    0ubbles may brea"-up due toturbulence impact

    0rea"age increases number  of

     bubbles but decrease si%e of

     bubbles

    -D p8

    D p

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    42/52

    RMIT University 42

    Po&ulation Balane ;&&roaesPo&ulation Balane ;&&roaes 'opulation 0alance $/uation 2'0$3

    Three .ain &pproaches has been proposed: .oment of .ethod 2.4.3 ;east /uare .ethod 2;.3

    Class .ethod 2C.3

    ( )( )   +−=⋅∇+

    ∂3A23A23AA2

    AAt  f  t bt r  f  v

    t r  f  ξ ξ ξ 

    ξ 

    −∫ @

    73A23A23AA2   dst  s f  t  sbt  sh ξ 

    +∫ @

    73A23AA23A2   dst  s f  t  sC t  s f     ξ 

    ∫    −−ξ 

    ξ ξ 7

    3A23A23AA26

    @dst  s f  t  s f  t  s sC 

    +eat rate (ue to 0rea9ae

    Birt rate (ue to 0rea9ae

    +eat rate (ue to 1oalesene

    Birt rate (ue to 1oalesene

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    43/52

    RMIT University 43

    1lass Meto(s /or Po&ulation Balane1lass Meto(s /or Po&ulation Balane

    -D p8

    D p

     ;verae -um0er 

    .im&le

    )oss (istri0ution in/ormation

    1om&le< an( e

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    44/52

    RMIT University 44

     Average Bubble %umberDensity &AB%D'

    &dopt a single averaged value to describe the

    changes of local bubble number 

    4nly one additional e/uation is needed

    &dvantages: Fast in terms of computational time

    imple to implement

     Disadvantages: 4nly one averaged value can be obtained

    Fluid +elocity

    n n n n n

    n’ 

    n’ 

    n’ n’ n’ n’  

    ( )   Breakagen

    eCoalesen

    n g n!t 

    nφ φ    +=⋅∇+

    ∂  

    n’ 

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    45/52

    RMIT University 45

    !(ltiple S")e *roup &!(S"*'!odel &"' 1sed 5 si%e group 2scalar3 to describe bubble population

     5 extra e/uations have to be solved 2depends on number of

    group used3

    &dvantages:

    Higher resolution for bubble classes

    Disadvantages: &dditional computational time needed

    ( )   BC  BC " g "  D D #  # n!t 

    n−−+=⋅∇+

    ∂∂  

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    46/52

    RMIT University 46

    !(ltiple S")e *roup &!(S"*'!odel &""'

    Fluid +elocity

    n"   n"   n"   n"   n"

    n"’    n’    n"’ n"’ n"’ 

     #  B   D B

    ( )   $" $ N 

    " $

     B   nvv% #  :@

    ∑+=

    =   "" B   n% D   =

     # C    DC 

     BC  BC n   D D #  # S  " −−+=

    n"   n"

    n"n"

    ∑∑= =

    ="

    "

     $"kl  &"C    nn # @ @6

    @  χ  ∑=

    = N 

     $

     $""$C    nn D@

     χ 

    kl  &"kl  &"   χ  χ    =  if "l k    vvv   =+

      else7=kl  &" χ    "l k    vvv   ≠+

    -n"8

    n"

    -n"8

    n"

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    47/52

    RMIT University 47

    Bu00le MeanismBu00le MeanismBu00le 1oalesene

    Ran(om 1ollision

    Bu00le Brea9ae

    a9e 'ntrainment

    Tur0ulene Im&at

    .earino// .ur/ae insta0ility

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    48/52

    RMIT University 4

    Free Sur+ace Flow in BuildingDrainage system

     

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    49/52

    RMIT University 4!

    Fluid Structure "nteraction

    Fluid flow tructural &nalysis

    .ethod:

    → F  # C

    )eometry  Ne'

    32

    32

    CF( 

     Fl!ent  32 ANS)S 

     

    Fluid .odelling

    2CFD3

    !olid .odelling 

    2F$!.3

    Coupling

    Intera5tion 

    (5T$RF&C$

    2middle-ware3

    •  Turbulence

    •  'ressure

    •  Fluctuations

    •  !tress

    •  Deformation

    •  ibration.4are

    In/ormation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    50/52

    RMIT University 5#

    0iomedical &pplication

    CT

    .R(

    )eometry

    .odel

    2C&D3

    CFD .odel of &ir way

    0lood vessel

    'ressureA hear stressA

    Drug depositionA Temperature etcBB

     

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    51/52

    RMIT University 5"

    Blood ,essel – -all S$earStress &-SS'

    Carotid 0ifurcation

  • 8/17/2019 Miet 2394 Cfd Lecture 11

    52/52

    Blood ,essel – FS"