The AMG solver [128] builds coarse levels by grouping fine level cells to make coarse level cells, and uses piecewise constant interpolation.
In the Multigrid tab (Figure 28.9: The Multigrid Tab for the Flexible Cycle), you can choose a stabilization method. If desired, you can choose the bi-conjugate gradient stabilized method [9] (BCGSTAB) option or recursive projection method [104] (RPM) in order to improve the convergence of the linear solver. BCGSTAB can be preconditioned, and both BCGSTAB and RPM provide stability for the AMG solver.
ANSYS Fluent usually builds diagonally dominant matrices for the linear solver. However, this is not always possible. A linear system with highly dominant off-diagonal coefficients may occur during discretization of complex physical models such as multiphase cavitation. Using the BCGSTAB or RPM option in such cases can be helpful. In addition, the AMG convergence in parallel can be improved using the BCGSTAB option with AMG.
If you are using the pressure-based segregated solver and the flow is incompressible, an additional stabilization method for the algebraic multigrid solver will appear in the Stabilization Method drop-down list. This is the conjugate gradient method, or CG [9] and will be available only for the pressure equation. This method is typically used in conjunction with an AMG preconditioner (any available AMG method). The CG formulation requires a symmetric system matrix. Such matrices result from the finite volume discretization of steady or transient elliptic operators, such as the pressure correction equation in the incompressible case. The CG method provides a useful extension to BCGSTAB and RPM, since it reduces the memory requirements and the number of floating point operations, especially when compared to BCGSTAB. In addition, in transient pressure-based segregated solvers, as typically used in LES and/or NITA simulations, the solution of the pressure correction equation constitutes a significant share of the overall computational effort.