Some modelling and computational issues

1
Some modelling and computational issues

• Grids
• 1-D, 2-D, 3D problems
• physical space vs computational space
• orthogonal vs non-orthogonal grids
• structured vs unstructured
• grid generation
• spacing and smoothing
• near wall grids

2
Some modelling and computational issues

• Solvers
• have to solve the following steady or transient
  equations
• continuity,
• momentum (x1,2 or 3)
• turbulence (k and e)
• energy
• species
• p-r-T relationship
• Algorithms
• SIMPLE(R)
• QUICK
• fractional step
• Convergence criteria

3
Some modelling and computational issues

• Turbulence
• Lagrangian vs Eulerian view
• Mathematical treatment
• DNS (Direct Numerical Simulation) of Navier
  Stokes equations - computes all turbulent
  structures in time and space
• LES (Large Eddy Simulation) computes the large
  eddies and models the small ones
• RANS (Reynolds Averaged Navier Stokes) solution -
  models all structures typically through a
  turbulent viscosity concept

4
Some modelling and computational issues

• Initial and boundary conditions for v, p, T, c,
  k, e
• Initial condition -fieldfrom experiments, a
  laminar or coarse turbulent simulation
• Boundary conditions
• Inlet from experimental data, entry length
  simulation or cyclic boundary condition
• Outlet from experimental data, zero gradient or
  convective hypothesis
• Solid boundary - wall no-slip or wall functions
• Symmetry zero gradient