GMS:SMS Package

From XMS Wiki
(Redirected from GMS:SMS Solver)
Jump to navigationJump to search
MODFLOW
Pre-processing
MODFLOW Commands
Building a MODFLOW Model
Map to MODFLOW
Calibration
Packages Supported in GMS
Saving a MODFLOW Simulation
Importing MODFLOW Files
Unsupported MODFLOW Features
Run MODFLOW
Post-processing
MODFLOW Display Options
MODFLOW Post-Processing Viewing Options
Reading a MODFLOW Simulation
Tutorials
Packages
Flow: BCF6, HUF, LPF, UPW
Solvers:

DE4, GMG, NWT, PCG,

PCGN, LMG, SIP, SOR,

SMS
Other:

BAS6, BFH, CHD1, CLN,

DRN1, DRT1, EVT1, ETS1,

GAGE, GHB1, GNC, HFB1,

HUF, LAK3, MNW1, MNW2,

OUT1, RCH1, RIV1, SFR2,

STR1, SUB1, SWI2, WEL1,

UZF1

Example of the MODFLOW SMS Solver dialog.

The Sparse Matrix Solver (SMS) Package is the only solver package compatible with MODFLOW-USG and is not compatible with other MODFLOW models.

The MODFLOW SMS Solver dialog has the following options:

  • Comments (Text) – Places a comment in the solver.
  • Max head change between outer iterations (L)(HCLOSE)
  • Max head change between inner iterations (L)(HICLOSE)
  • Max number of outer nonlinear iterations for problem (MXITER)
  • Max number of inner linear iterations for problem (ITER1)
  • Print additional info to listing file (IPRSMS)
    • (0) print nothing
    • (1) print summary
    • (2) print detail
  • Nonlinear solution method (NONLINMETH)
    • (-2) Picard with Cooley
    • (-1) Picard with Delta-Bar-Delta
    • (0) Picard
    • (1) Newton with Delta-Bar-Delta
    • (2) Newton and Cooley
  • Linear matrix solver (LINMETH)
    • (1) χMD
    • (2) PCGU
    • (3) SAMG
  • Options (OPTIONS)
    • SPECIFIED
    • SIMPLE
    • MODERATE
    • COMPLEX

More information regarding inputs to the SMS package can be found in the MODFLOW-USG documentation at pubs.usgs.gov.

The SMS package, when used with GMS, also includes the option of using the SAMG or LMG within the SMS package. This option is selected under the Linear matrix solver (LINMETH) option in the MODFLOW SMS Solver dialog.

Nonlinear Solution Method Options

  • Delta-bar-delta learning rate reduction factor (THETA)
  • Delta-bar-delta learning rate increment (AKAPPA)
  • Delta-bar-delta memory term factor (GAMA)
  • Nonlinear fraction history added (AMOMENTUM)
  • Maximum residual backtracking iterations (NUMTRACK)
  • Residual change tolerance (BTOL)
  • Residual change reduction size (BREDUC)
  • Residual reduction limit (RESLIM)

χMD Options

  • Acceleration method (IACL)
    • (0) conjugate gradient
    • (1) ORTHOMIN
    • (2) Bi-CGSTAB
  • Ordering scheme (NORDER)
    • (0) original ordering
    • (1) RCM ordering
    • (2) minimum degree ordering
  • ILU decomposition level of fill (LEVEL)
  • Number of orthogonalizations for ORTHOMIN accel. (NORTH)
  • Reduced system (IREDSYS)
    • (0) do not apply
    • (1) apply
  • Residual tolerance criterion (RRCTOL)
  • Perform drop tolerance (IDROPTOL)
    • (0) do not perform
    • (1) perform
  • Drop tolerance value (EPSRN)

PCGU Options

  • Linear acceleration method (CLIN)
    • CG
    • BCGS
  • Preconditioner (IPC)
    • (0) None
    • (1) Jacobi
    • (2) ILU(0)
    • (3) MILU(0)
  • Matrix scaling approach (ISCL)
    • (0) None
    • (1) POLCG preconditioner
    • (2) l2 norm
  • Matrix reordering approach (IORD)
    • (0) Original
    • (1) reverse Cuthill Mckee
    • (2) minimum degree
  • Convergence flow residual tolerance (L^3)(RCLOSEPCGU)
  • MILU(0) relaxation factor (RELAXPCGU)
  • Only solve active cells (SOLVEACTIVE)
  • Bottom damping procedure used (DAMPBOT)
  • Right-hand vector translated with linear solve (SHIFT)
  • Truncated newton approach used (TRUNCATEDNEWTON)