Lecture 9 of 'Scientific Computing' (wi4201)

• The following subjects are discussed:
  
  
  • Warning: no sound in the first minute of the Collegerama video
  • smoother and coarse grid correction operators
  • convergence analysis via eigenvalues, slow and fast varying components
  • coarsening of the grid, from h to 2h, no elimination of boundary conditions
  • integrid operators: restriction and prolongation (stencil and matrix notation)
  • coarse grid operator: rediscretization or Galerkin approximation
  • advantages and disadvantages
  • coarse grid operator as iterative solver
  • two grid method
  • pre- and postsmoothing
  • multigrid: use the coarse grid operator recursively
  • V, F, and W cycle
  • amount of work and memory
  • number of iterations independent of h, scalable solutions method
    
    
• Material is described in pages 90-95 of the lecture notes.