towards 2d overland flow simulations
play

Towards 2D overland flow simulations Olivier Delestre Laboratory - PowerPoint PPT Presentation

Towards 2D overland flow simulations Olivier Delestre Laboratory J.A. Dieudonn e & Polytech Nice Sophia University of Nice Sophia Antipolis CEMRACS 2013 Problem context Preventing overland flow and erosion From upstream... (Photos


  1. Towards 2D overland flow simulations Olivier Delestre Laboratory J.A. Dieudonn´ e & Polytech Nice Sophia University of Nice – Sophia Antipolis CEMRACS 2013

  2. Problem context

  3. Preventing overland flow and erosion From upstream... (Photos : Yves Le Bissonnais, INRA)

  4. ...to downstream.

  5. Downstream zones modifications (watersheds) ◮ Where is the water coming from ? ◮ Where is it flowing ? Use of physical models is required to : ◮ simulate flow (volumes and location) ◮ suggest changes (grass strip). � to carry out improvements

  6. Shallow Water (Saint-Venant) system z, z + h h v u � u y z O x Data : topography z , rain P , infiltration I Unknowns : velocities u , v , water height h  ∂ t h + ∂ x ( hu ) + ∂ y ( hv ) = P − I  hu 2 + gh 2 / 2 � � ∂ t ( hu ) + ∂ x + ∂ y ( huv ) = gh ( − ∂ x z − S f x ) hv 2 + gh 2 / 2 � � ∂ t ( hv ) + ∂ x ( huv ) + ∂ y = gh ( − ∂ y z − S f y ) 

  7. Strategy ◮ Properties of the 1D Shallow Water system ◮ Choice of the method depending on the properties ◮ Validation : analytical solutions and laboratory experiment ◮ Application : field data

  8. 1D Shallow Water system z, z + h P ( t, x ) q ( t, a ) q ( t, b ) I ( t, x ) O a b x Data : topography z , rain P , infiltration I Unknowns : velocities u , water height h A system of conservation laws � ∂ t h + ∂ x ( hu ) = P − I (1) ∂ t ( hu ) + ∂ x ( hu 2 + gh 2 / 2) = gh ( − ∂ x z − S f )

  9. System properties (I) : Hyperbolicity Setting q = hu � h � � � � P − I � q U = , F ( U ) = , B = , q 2 / h + gh 2 / 2 gh ( − ∂ x z − S f ) q compact form ∂ t U + ∂ x F ( U ) = ∂ t U + F ′ ( U ) ∂ x U = B , Hyperbolicity if h > 0 : � � λ − ( U ) = u − gh , λ + ( U ) = u + gh Saint-Venant gaz dynamic Froude number Fr = | u | Mach number | u | c c c = √ gh free surface waves celerity p ′ ( ρ ) sound speed 1 � c = subcritical Fr < 1 subsonic supercritical Fr > 1 supersonic 1. p ( ρ ) = ρ RT perfect gaz

  10. System properties (II) : Conservation laws Integral of equation (1) in x ∂ t h + ∂ x q = P − I , gives � b � b d h ( t , x ) dx = q ( t , a ) − q ( t , b ) + P ( t , x ) − I ( t , x ) dx , dt a a Mass conservation of water. Second equation : momentum equation

  11. System properties (III) : Steady states

  12. System properties (III) : Steady state � ∂ t h + ∂ x ( hu ) = P − I (2) ∂ t ( hu ) + ∂ x ( hu 2 + gh 2 / 2) = gh ( − ∂ x z − S f ) ∂ t h = ∂ t u = ∂ t q = 0 � ∂ x hu = P − I . ∂ x ( hu 2 + gh 2 / 2) = gh ( − ∂ x z − S f )

  13. System properties (III) : Steady states Lac at rest equilibrium � u = 0 . g ( h + z ) = Cst z, z + h H sur = z + h = Cte O x

  14. Numerical method (I) Finite volume method ∆ x t We integrate ∂ t U + ∂ x F ( U ) = 0 t n +1 on the volume ∆ t n [ t n , t n +1 [ × ] x i − 1 / 2 , x i +1 / 2 [, t n and we set � x i +1 / 2 1 U n U ( t n , x ) dx i = ∆ x x i − 1 / 2 O x i − 1 x i − 1 / 2 x i x i +1 / 2 x i +1 x we get i − ∆ t U n +1 = U n F n i +1 / 2 − F n � � , i − 1 / 2 i ∆ x with the interface flux approximation � t n +1 1 F n i +1 / 2 = F ( U n i , U n i +1 ) ∼ F ( U ( t , x i +1 / 2 )) dt . ∆ t t n

  15. Numerical method (I) ◮ For each choice of F ( U G , U D ) we have a different finite volume scheme : HLL, kinetic, Rusanov, VFRoe-ncv, suliciu, ... ◮ second Order ◮ in space : MUSCL, ENO, modified ENO ◮ in time : Heun

  16. Numerical method (I) ◮ For each choice of F ( U G , U D ) we have a different finite volume scheme : HLL, kinetic, Rusanov, VFRoe-ncv, suliciu, ... ◮ second Order ◮ in space : MUSCL, ENO, modified ENO ◮ in time : Heun ◮ Coupling with the source term (topography ∂ x z ) Necessity : compatibility with steady states

  17. Steady states (II) � ∂ t h + ∂ x ( hu ) = 0 (3) ∂ t ( hu ) + ∂ x ( hu 2 + gh 2 / 2) = − gh ∂ x z ∂ t h = ∂ t u = ∂ t q = 0 � hu = Cst . u 2 / 2 + g ( h + z ) = Cst We consider � u = Cst . g ( h + z ) = Cst

  18. Hydrostatic reconstruction (II) [Audusse et al., 2004] We define z ∗ = max ( z G , z D ) and U ∗ G = ( h ∗ G , h ∗ G u G ) , U ∗ D = ( h ∗ D , h ∗  D u D )  h ∗ G = max ( h G + z G − z ∗ , 0) . h ∗ D = max ( h D + z D − z ∗ , 0)  Thus, we have  � � 0 F G ( U G , U D , ∆ Z ) = F ( U ∗ G , U ∗ D ) +  2 − ( h ∗  G ) 2 ) / 2 g ( h G  , � � 0 F D ( U G , U D , ∆ Z ) = F ( U ∗ G , U ∗ D ) +  2 − ( h ∗  D ) 2 ) / 2 g ( h D  where F ( U G , U D ) is the numerical flux.

  19. Friction treatment Shallow Water system with friction f � ∂ t h + ∂ x ( hu ) = 0 , (4) ∂ t ( hu ) + ∂ x ( hu 2 + gh 2 / 2) + h ∂ x z = − hf , f = f ( h , u ) friction force (on the bottom) Several friction laws possible f = n 2 u | u | ◮ Manning : h 4 / 3 f = F u | u | ◮ Darcy-Weisbach : 8 gh

  20. Friction treatment ◮ Apparent topography [Bouchut, 2004] z app = z + b n We consider : with ∂ x b n = S n f

  21. Friction treatment ◮ Apparent topography [Bouchut, 2004] z app = z + b n We consider : with ∂ x b n = S n f ◮ Semi-implicit [Bristeau and Coussin, 2001] i | q n +1 + F | q n i + ∆ t q n +1 ∆ t = q n i � � F i +1 / 2 − F i − 1 / 2 i i h n +1 8 h n ∆ x i with q n +1 ∗ for the right part, we have i q n +1 ∗ q n +1 i = i 1 + ∆ t F | u n i | 8 h n +1 i

  22. Validation on analytical solutions – SWASHES New test cases : ◮ Saint-Venant/shallow water : ◮ data z ◮ unknowns h et u (and so q )

  23. Validation on analytical solutions – SWASHES New test cases : ◮ Saint-Venant/shallow water : ◮ data z ◮ unknowns h et u (and so q ) ◮ test cases ◮ data h and q (and so u ) ◮ unknown z

  24. Validation on analytical solutions – SWASHES New test cases : ◮ Saint-Venant/shallow water : ◮ data z ◮ unknowns h et u (and so q ) ◮ test cases ◮ data h and q (and so u ) ◮ unknown z ◮ Several possibilities ◮ several friction laws ◮ diffusion source term [Delestre and Marche, 2010] ◮ rain source term

  25. 1 1 topographie topographie surface libre surface libre 0 niveau critique 0 niveau critique -1 -1 -2 z, z+h (m) z, z+h (m) -2 -3 -4 -3 -5 -4 -6 -5 -7 -8 -6 0 200 400 600 800 1000 0 200 400 600 800 1000 x (m) x (m) 1 2 topographie topographie surface libre surface libre 0 0 niveau critique niveau critique -2 -1 -4 z, z+h (m) z, z+h (m) -2 -6 -8 -3 -10 -4 -12 -5 -14 -6 -16 0 200 400 600 800 1000 0 200 400 600 800 1000 x (m) x (m)

  26. Validation on analytical solutions – SWASHES Apparent topography (subcritical-subcritical) 1 topographie surface libre 0 niveau critique -1 -2 z, z+h (m) -3 -4 -5 -6 -7 -8 0 200 400 600 800 1000 x (m)

  27. Validation on analytical solutions – SWASHES Semi-implicit (subcritical-subcritical) 1 topographie surface libre 0 niveau critique -1 -2 z, z+h (m) -3 -4 -5 -6 -7 -8 0 200 400 600 800 1000 x (m)

  28. Validation on analytical solutions – SWASHES Semi-implicit (subcritical-supercritical) 1 topographie surface libre 0 niveau critique -1 z, z+h (m) -2 -3 -4 -5 -6 0 200 400 600 800 1000 x (m)

  29. Validation on analytical solutions – SWASHES Semi-implicit (supercritical-subcritical) 1 topographie surface libre 0 niveau critique -1 z, z+h (m) -2 -3 -4 -5 -6 0 200 400 600 800 1000 x (m)

  30. Summary of the chosen numerical method ◮ Numerical flux : HLL ◮ Second order scheme : MUSCL ◮ Friction : semi-implicit treatment ◮ Shallow Water system with rain P � ∂ t h + ∂ x ( hu ) = P (5) ∂ t ( hu ) + ∂ x ( hu 2 + gh 2 / 2) + h ∂ x z = − hf time splitting/explicit treatment

  31. Validation on experiments – INRA rain simulator

  32. Settings of the experiment z 5 cm P ( t ) L c = 4 m q ( t, L ) x O L 0 ≤ t ≤ 250s � 50 mm/h if ( x , t ) ∈ [0 , L ] × [5 , 125] R ( x , t ) = 0 else

  33. Analytical solutions and simulations 7 f=0.12, numerique f=0.12, cinematique 6 f=0.12, exact f=0.34, numerique f=0.34, cinematique 5 f=0.34, exact 4 q (g/s) 3 2 1 0 0 50 100 150 200 250 Temps de simulation (s)

  34. Water height and velocity at equilibrium 0.12 0.0006 0.1 0.0005 0.08 0.0004 u (m/s) h (m) 0.06 0.0003 0.04 0.0002 f=0.12, numerique f=0.12, numerique f=0.12, cinematique f=0.12, cinematique f=0.12, exact f=0.12, exact 0.02 0.0001 f=0.34, numerique f=0.34, numerique f=0.34, cinematique f=0.34, cinematique f=0.34, exact f=0.34, exact 0 0 0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4 x (m) x (m)

  35. What about reality ? 9 8 7 6 q(.,L) (g/s) 5 4 3 2 1 0 0 50 100 150 200 250 t (s)

  36. ”Calibration” Darcy-Weisbach Manning 9 9 Mesures Mesures f=0.1 n=0.008 8 8 f=0.11 n=0.009 f=0.12 n=0.01 7 7 f=0.13 n=0.011 f=0.14 n=0.012 6 6 f=0.15 n=0.013 f=0.16 n=0.014 5 f=0.17 5 n=0.015 q (g/s) q (g/s) f=0.18 n=0.016 4 4 3 3 2 2 1 1 0 0 0 50 100 150 200 250 0 50 100 150 200 250 Temps de simulation (s) Temps de simulation (s)

  37. A simulation result (Manning) 9 Mesures n=0.013 8 7 6 5 q (g/s) 4 3 2 1 0 0 50 100 150 200 250 Temps de simulation (s)

  38. Parcels in Niger ([Esteves et al., 2000], IRD)

Recommend


More recommend