Download lapn.notice

Back to the list

Display this manual page in
$$$$ LAPN     NOTICE  CHAT      11/09/12    21:16:52     7124           
                                             DATE     11/09/12
                                                                                
  Operateur LAPN                          Voir aussi :



 I Finite Elements Formulation:
 ______________________________


     Syntax EQEX (see EQEX) :


     ... 'EQEX' ... 'OPTI' MOT1 MOT2
                    'ZONE' MOD1
                    'OPER' 'LAPN' OBJ1
                    'INCO' MOT3 (MOT4)

     Description :
     _____________

     LAPN operator deals with the discretisation of diffusive terms
in scalar or vectorial equations. For vectorial equation, equation
of momentum conservation for incompressible flows is considered.

     For a scalar equation as the heath equation
                     dT/dt = div(alpha grad T)
this operator treats the div(alpha grad T) term, where alpha is the
diffusivity (alpha in m2/s [SI]).

     For the momentum equation, this operator deals with the divergence
of viscous stress tensor in incompressible form
                      ->                        ->  t     ->
                     dU/dt + ... = div(nu (grad U +  grad U))
so we have to consider div(nu (grad U + tgrad U)) where nu is the
cinematic viscosity (nu in m2/s [SI]).

     For a system, the discretised term is of the generalised form
div(d grad T) in an equation dealing with the evolution of an unknown
named V. T is called the primal unknown as V is the dual unknown.

     Sign convention is the following: when the coefficient of
diffusion is greater then zero, the maximum of the scalar or vectorial
unknown decreases.

     Transient procedure EXEC use this operator with data constructed
with EQEX operator with the syntax done.


     Contents :
     __________
                                                                                
    'OPTI' : Key word introducing numerical options
     MOT1  : Discretization in space ('EF', 'VF' or 'EFM1')
     MOT2  : Discretization in time ('EXPL' or 'IMPL')
     Actually, VF and EF are treated in IMPL ; EFM1 in EXPL.

    'ZONE' : Key word introducing geometric informations
     MOD1  : MODELE object ; contains geometric data
             associated to the region where exchange is located

    'OPER' : Key word introducing operator's name and physical data
    'LAPN' : Operator's name treated by EQEX
     OBJ1  : Coeff of diffusion d (CHPO SCAL CENTRE, FLOTTANT or MOT)

    'INCO' : Key word introducing primal and dual unknowns' name
     MOT3  : Primal unknown's name (T)
     MOT4  : Dual unknown's name (V)
     When primal and dual names are the same, only one is mandatory.
Note that dual and primal unknowns are always the same when explicit
formulation is used.


     Results :
     _________

     Explicit formulation :
     - The tight hand side is stored in a CHPO and set in order in the
table KIZG at the index MOT3.

    Implicit formulation :
    - The matrix is stored in a MATRIK and set in order in the table
TAB1 at the index of type MOT MATELM.
    - The tight hand side is stored in a CHPO and assembled with the
other's stored in EQEX table at the index of type MOT SMBR. CHPO
component's nane of the LAPN's contribution is MOT4.


     Remarks :
     ___________
                                                                                
     1) When OBJ1 is of type MOT, we look for data in INCO
table at the index OBJ1.

     2) Geometric support (spg) of CHPOINT contains one of those
of the table of subtype DOMAINE. Some compatibilities are verified :
   - For EF or EFM1 formulation, dual unknown's spg contains SOMMET
   - For VF formulation, dual unknown's spg contains CENTRE
   - In implicit formulation, when primal and dual unknowns are not
the same, spg are to be the same
   - Exchange coefficient spg is always CENTRE

     3) Graduate user who develops his own transient procedure calls
the LAPN operator with the following syntax :
     LAPN TAB1 ;
with TAB1 : Table of subtype EQEX ; contains physical and numerical
            informations ; created by the EQEX operator.
 



 II Finite Volume Formulation:
 _____________________________

 IIa : Mono-component perfect gas (constant volume specific heat)
 ___

 Diffusive terms discretization of the compressible Navier-Stokes 
 equations for a calorically perfect gas.

 SYNTAX:

 RMAT1 RCHP1 DELTAT = 'LAPN' 'VF' 'PROPCOST' MOT1 MOT2 MOD1
                     FLOT1 FLOT2 FLOT3 CHPO1 CHPO2 CHPO3 CHPO4 CHPO5
                     (CHAM1 CHAM2 si MOT2 = 'IMPL')
                     ('VIMP' CHPO6) ('TAUI' CHPO7)
                     ('QIMP' CHPO8) ('MIXT' CHP10) 
                     ('TIMP' CHPO9) LMOT ('CLAUDEIS');


  MOT1   : MOT object
           'RESI' if we want to compute the residuum
           'FLUX' if we want to compute the flux

  MOT2   : MOT object
           'IMPL' if we want to compute the residuum jacobian RMAT1
           'EXPL' if we don't want to compute RMAT1

  MOD1   : MODELE object.

  FLOT1  : 'FLOTTANT' object (dynamic viscosity)
 
  FLOT2  : 'FLOTTANT' object (thermal conductivity)

  FLOT3  : 'FLOTTANT' object (constant volume specific heat)

  CHPO1  : 'CHPOINT' 'CENTRE' (mass density, 1 component, 'SCAL')

  CHPO2  : 'CHPOINT' 'CENTRE' (velocity, 2/3 components 'UX','UY',
           ('UZ'))

  CHPO3  : 'CHPOINT' 'CENTRE' (temperature, 1 component, 'SCAL')

  CHPO4  : 'CHPOINT' 'FACE' (velocity gradient 4/9 components,
           'P1DX', 'P1DY',('P1DZ'),'P2DX','P2DY',('P2DZ'), 
           ('P3DX','P3DY','P3DZ'))
 
  CHPO5  : 'CHPOINT' 'FACE' (temperature gradient, 2/3
            components, 'P1DX','P1DY',('P1DZ'))

  CHAM1  : 'MCHAML' (coefficients used for velocity gradient 
                     computation (see 'PENT' operator))

  CHAM2  : 'MCHAML' (coefficients used for temperature gradient 
                     computation (see 'PENT' operator))

  CHPO6  : 'CHPOINT' 'FACE' (imposed velocity, 2/3 components,
           'UX','UY',('UZ'))

  CHPO7  : 'CHPOINT' 'FACE' (imposed viscous stress tensor, 
            3/6 components 'TXX','TYY','TXY',('TXZ''TYZ','TZZ'))

  CHPO8  : 'CHPOINT' 'FACE' (imposed heat flux, 2/3 components, 
           'UX','UY',('UZ'))

  CHPO9  : 'CHPOINT' 'FACE' (imposed temperature, 
            1 component 'SCAL')

  CHP10  : 'CHPOINT'  'FACE' 
           (4 components, lambda1,lmabda2,qlimx,qlimy ) 
           (imposed mixed condition)
           lambda1 (d grad T . n) + lambda2 T = (qlimx*nx) + (qlim*ny) 

  LMOT1  : LISTMOTS object
           It contains  the components names of RCHPO1 in the 
           following order: density, momentum, total energy

  RMAT1  : MATRIK object
           (SPG = 'DOMA' MOD1 'CENTRE')
           (primal unknowns = dual unknowns; their names are into 
           LMOT1)
           In the implicit case, it contains the Jacobian of the
           residuum with respect to the conservative variables.
           Conversely, it is an empty  MATRIK
 
  RCHPO1 : CHPOINT object (components into LMOT1)
           Residuum  if MOT1 = 'RESI' (SPG = 'DOMA' MOD1 'CENTRE')
           Flux  if MOT1 = 'FLUX'    (SPG = 'DOMA' MOD1 'FACE') 
 
  RFLOT1 : FLOTTANT object 
           It is the stability time step for the explicit Euler
           scheme.

  CLAUDEIS : Option to take into account thermal part of the 
             Jacobian of the residuum  
 

 IIb : Multi-component perfect gas (constant volume specific heat
 ___   for the species)

 Diffusive terms discretization of the compressible Navier-Stokes 
 equations for a mixture of calorically perfect gases.
 Unknowns : mass density (of the mixture), momentum,
            total energy per unit volume, species density

 SYNTAX:

 RMAT1 RCHP1 DELTAT = 'LAPN' 'VF' 'PERFMULT' MOT1 MOT2 MOD1 TAB2
                     CHPO1 CHPO2 CHPO3 CHPO4 CHPO5 CHPO6 CHPO7 CHPO8
                     (CHAM1 CHAM2 when MOT2 = 'IMPL')
                     ('VIMP' CHPO9) ('TAUI' CHP10)
                     ('QIMP' CHP11) ('TIMP' CHP12)  ('RIMP' CHP12)
                     LMOT ;



  MOT1   : MOT object
           'RESI' if we want to compute the residuum
           'FLUX' if we want to compute the flux

  MOT2   : MOT object
           'IMPL' if we want to compute the residuum jacobian RMAT1
           'EXPL' if we don't want to compute RMAT1

  MOD1   : MODELE object.

  TAB2   : TABLE object with the species properties :
          * name of the specie not in the Euler equation system :
            TAB2 . 'ESPNEULE' (MOT) ;                                  
          * name of the species explicitly part of the Euler equ. 
            system in TAB2 . 'ESPEULE' (LISTMOTS) ;
          * degree of polynomial for cv_i=cv_i(T), in TAB2 . 'NORD'        
            (it must be 0)                                                
          * properties of each gaz 'ESPI', in table TAB2 . 'ESPI':              
            - TAB2 . 'ESPI' . 'A' (LISTREEL)                             
              the (TAB2.'NORD')+1 coefficients of the polynomial
              cv(T)
            - TAB2 . 'ESPI' . 'R' (J/kg/K in SI system, FLOTTANT)
              perfect gaz constant
            - TAB2 . 'ESPI' . 'H0K' (J/kg, FLOTTANT) mass formation enthalpy
            - TAB2 . 'ESPI' . 'CDIFF' (FLOTTANT) : diffusion coefficient
              of the specie into the mixture
            - TAB2 . 'ESPI' . 'CLYK' :  'CHPOINT' 'FACE' (Dirichlet BC
              for the specie density, 1 component, 'SCAL')
            - TAB2 . 'ESPI' . 'YK' :  'CHPOINT' 'FACE' (specie density 
              , 1 component, 'SCAL') 
            - TAB2 . 'ESPI' . 'CGRYK' :  'MCHAML' (coefficient used 
              for the computation of the gradient of the specie density)
            - TAB2 . 'ESPI' . 'GRADYK' :  'CHPOINT' 'FACE' (gradient of the
              specie density, 2/3 components, 'P1DX','P1DY',('P1DZ'))

  CHPO1  : 'FLOTTANT' object (dynamic viscosity of the mixture)
 
  CHPO2  : 'FLOTTANT' object (thermal conductivity of the mixture)

  CHPO3  : 'FLOTTANT' object (constant volume specific heat of the mixture)

  CHPO4  : 'CHPOINT' 'CENTRE' (mass density, 1 component, 'SCAL')

  CHPO5  : 'CHPOINT' 'CENTRE' (velocity, 2/3 components 'UX','UY',
           ('UZ'))

  CHPO6  : 'CHPOINT' 'CENTRE' (temperature, 1 component, 'SCAL')

  CHPO7  : 'CHPOINT' 'FACE' (velocity gradient 4/9 components,
           'P1DX', 'P1DY',('P1DZ'),'P2DX','P2DY',('P2DZ'), 
           ('P3DX','P3DY','P3DZ'))
 
  CHPO8  : 'CHPOINT' 'FACE' (temperature gradient, 2/3
            components, 'P1DX','P1DY',('P1DZ'))

  CHAM1  : 'MCHAML' (coefficients used for velocity gradient 
                     computation (see 'PENT' operator))

  CHAM2  : 'MCHAML' (coefficients used for temperature gradient 
                     computation (see 'PENT' operator))

  CHPO9  : 'CHPOINT' 'FACE' (imposed velocity, 2/3 components,
           'UX','UY',('UZ'))

  CHP10  : 'CHPOINT' 'FACE' (imposed viscous stress tensor, 
            3/6 components 'TXX','TYY','TXY',('TXZ''TYZ','TZZ'))

  CHP11  : 'CHPOINT' 'FACE' (imposed heat flux, 2/3 components, 
           'UX','UY',('UZ'))

  CHP12  : 'CHPOINT' 'FACE' (imposed temperature, 
            1 component 'SCAL')

  CHP13  : 'CHPOINT' 'FACE' (imposed mass density,
            1 component 'SCAL')

  LMOT1  : LISTMOTS object
           It contains  the components names of RCHPO1 in the 
           following order: density, momentum, total energy,
           specie mass density.

  RMAT1  : MATRIK object
           (SPG = 'DOMA' MOD1 'CENTRE')
           (primal unknowns = dual unknowns; their names are into 
           LMOT1)
           In the implicit case, it contains the Jacobian of the
           residuum with respect to the conservative variables.
           Conversely, it is an empty  MATRIK
 
  RCHPO1 : CHPOINT object (components into LMOT1)
           Residuum  if MOT1 = 'RESI' (SPG = 'DOMA' MOD1 'CENTRE')
           Flux  if MOT1 = 'FLUX'    (SPG = 'DOMA' MOD1 'FACE') 
 
  RFLOT1 : FLOTTANT object 
           It is the stability time step for the explicit Euler
           scheme.

 IIc :  Laplacian discretisation in porous media or on heat equation   
 ___

 SYNTAXE:

 RMAT1 RCHP1 DELTAT = 'LAPN' 'VF' 'CLAUDEIS' MOT2 MOD1
                       CHPO3 CHPO5 
                     (CHAM1  si MOT2 = 'IMPL')
                     ('QIMP' CHPO8) ('MIXT' CHP10) 
                     ('TIMP' CHPO9) ;

  MOT2   : MOT object
           'IMPL' if we want to compute the residuum jacobian RMAT1
           'EXPL' if we don't want to compute RMAT1

  MOD1   : MODELE object


  CHPO3  : 'CHPOINT' 'CENTRE' (temperature, 1 component, 'SCAL')

  CHPO5  : 'CHPOINT' 'FACE' (temperature gradient, 2/3
            components, 'P1DX','P1DY',('P1DZ'))

  CHAM1  : 'MCHAML' (coefficients used for temperature gradient 
                     computation (see 'PENT' operator))

  CHPO8  : 'CHPOINT' 'FACE' (imposed heat flux, 
            1 component, 'FLUX')

  CHPO9  : 'CHPOINT' 'FACE' (imposed temperature, 
            1 component ,'SCAL')

  CHP10  : 'CHPOINT'  'FACE' 
           (3 components, lambda1,lmabda2,qlim) 
           (imposed mixed condition)
           lambda1 (d grad T . n) + lambda2 T = qlim 
 
 
 
 
 
 
 
 
 

© Cast3M 2003 - All rights reserved.
Disclaimer