GMS:GMG Package: Difference between revisions

From XMS Wiki
Jump to navigationJump to search
No edit summary
No edit summary
 
(6 intermediate revisions by 2 users not shown)
Line 1: Line 1:
{{MODFLOW Links}}
{{MODFLOW Links}}
The GMG package is one of the [[GMS:MODFLOW Packages Supported in GMS#Solver Packages|solver packages]] for MODFLOW supported by GMS. The GMG solver is an iterative solver based on a multi-grid approach.
The GMG package is one of the [[GMS:MODFLOW Packages Supported in GMS#Solver Packages|solver packages]] for MODFLOW supported by GMS. The GMG solver is an iterative solver based on a multi-grid approach.  


Parameters include:
Parameters include:
* ''Inner convergence residual (RCLOSE)'' – The algorithm computes the l2-norm of the residual and compares it against RCLOSE.
* ''Inner convergence residual (RCLOSE)'' – The algorithm computes the 12-norm of the residual and compares it against RCLOSE.
* ''Maximum inner iterations (IITER)'' – For each linear solution. The default is 100. Specify a smaller number for nonlinear problems to prevent an excessive number of inner iterations.
* ''Maximum inner iterations (IITER)'' – For each linear solution. The default is 100. Specify a smaller number for nonlinear problems to prevent an excessive number of inner iterations.
* ''Outer convergence residual (HCLOSE)'' – For nonlinear problems. After each linear solve (inner iteration), the maximum norm of the head change is compared against HCLOSE. Can be set to a large number for linear problems. Ignored if ''Maximum outer iterations'' is 1.
* ''Outer convergence residual (HCLOSE)'' – For nonlinear problems. After each linear solve (inner iteration), the maximum norm of the head change is compared against HCLOSE. Can be set to a large number for linear problems. Ignored if ''Maximum outer iterations'' is 1.
Line 9: Line 9:
* ''Damping parameter (DAMP)'' – A value of 1.0 should be used for linear problems. For nonlinear problems, a value less than 1.0 but greater than 0.0 may be necessary to achieve convergence.
* ''Damping parameter (DAMP)'' – A value of 1.0 should be used for linear problems. For nonlinear problems, a value less than 1.0 but greater than 0.0 may be necessary to achieve convergence.
* ''Damping option (IADAMP)'' – Flag that controls adaptive damping. Contains the following options:
* ''Damping option (IADAMP)'' – Flag that controls adaptive damping. Contains the following options:
** "(0) Constant"
<blockquote style="margin-top:0px; margin-bottom:0px;">
** "(1) Cooley adaptive"  
* "(0) Constant"
** "(2) RRR adaptive"
* "(1) Cooley adaptive"  
* Maximum damping (DUP)
* "(2) RRR adaptive"  
* Minimum damping (DLOW)
</blockquote>
* Maximum head change (CHGLIMIT)
* ''Maximum damping (DUP)''
* ''Minimum damping (DLOW)''
* ''Maximum head change (CHGLIMIT)''
* ''Output flag (IOUTGMG)'' &ndash; Flag that controls the output of the GMG solver. Values can be 0&ndash;4.
* ''Output flag (IOUTGMG)'' &ndash; Flag that controls the output of the GMG solver. Values can be 0&ndash;4.
** 0 = Only the solver inputs are printed.
<blockquote style="margin-top:0px; margin-bottom:0px;">
** 1 = For each linear solve, prints the number of PCG iterations, the value of the damping parameter, the l2-norm of the residual, and the maximum norm of the head change and its location (column, row, layer). At the end of a time/stress period, the total number of GMG calls, PCG iterations, and a running total of PCG iterations for all time/stress periods are printed.
* 0 = Only the solver inputs are printed.
** 2 = The convergence history of the PCG iteration is printed, showing the l2-norm of the residual and the convergence factor for each iteration.
* 1 = For each linear solve, prints the number of PCG iterations, the value of the damping parameter, the 12-norm of the residual, and the maximum norm of the head change and its location (column, row, layer). At the end of a time/stress period, the total number of GMG calls, PCG iterations, and a running total of PCG iterations for all time/stress periods are printed.
** 3 = Same as 1 except the output is sent to the terminal instead of the output file.
* 2 = The convergence history of the PCG iteration is printed, showing the 12-norm of the residual and the convergence factor for each iteration.
** 4 = Same as 2 except the output is sent to the terminal instead of the output file.
* 3 = Same as 1 except the output is sent to the terminal instead of the output file.
* Maximum head output (IUNITMHC)
* 4 = Same as 2 except the output is sent to the terminal instead of the output file.
</blockquote>
* ''Maximum head output (IUNITMHC)''
* ''Multi-grid preconditioner smoothing (ISM)'' &ndash; Flag that controls the type of smoother used in the multi-grid preconditioner.
* ''Multi-grid preconditioner smoothing (ISM)'' &ndash; Flag that controls the type of smoother used in the multi-grid preconditioner.
** "(0) ILU smoothing" &ndash; This smoothing requires an additional vector on each multigrid level to store the pivots in the ILU factorization.
<blockquote style="margin-top:0px; margin-bottom:0px;">
** "(1) SGS smoothing" &ndash; Symmetric GaussSeidel (SGS) smoothing.  
* "(0) ILU smoothing" &ndash; This smoothing requires an additional vector on each multigrid level to store the pivots in the ILU factorization.
* "(1) SGS smoothing" &ndash; Symmetric GaussSeidel (SGS) smoothing.  
</blockquote>
* ''Multi-grid preconditioner coarsening (ISC)'' &ndash; Flag that controls semi-coarsening in the multi-grid preconditioner.
* ''Multi-grid preconditioner coarsening (ISC)'' &ndash; Flag that controls semi-coarsening in the multi-grid preconditioner.
** "(0) Row, Col., Layer" &ndash; Coarsens all the rows, columns and layers.
<blockquote style="margin-top:0px; margin-bottom:0px;">
** "(1) Row, Col." &ndash; Coarsens all the rows and columns.
* "(0) Row, Col., Layer" &ndash; Coarsens all the rows, columns and layers.
** "(2) Col., Layer" &ndash; Coarsens all columns and layers.
* "(1) Row, Col." &ndash; Coarsens all the rows and columns.
** "(3) Row, Layer" &ndash; Coarsens all rows and layers.
* "(2) Col., Layer" &ndash; Coarsens all columns and layers.
** "(4) None" &ndash; No coarsening.
* "(3) Row, Layer" &ndash; Coarsens all rows and layers.
* Relocation parameter (RELAX)
* "(4) None" &ndash; No coarsening.
</blockquote>
* ''Relocation parameter (RELAX)'' &ndash; Used to improve the spectral condition number of the ILU preconditioned system.


For more info on the GMG solver look at the following document:  
For more info on the GMG solver look at the following document:  
Line 45: Line 53:
[[Category:MODFLOW Dialogs]]
[[Category:MODFLOW Dialogs]]
[[Category:External Links]]
[[Category:External Links]]
{{stub}}
<!-- Needs more information. This article is giving the reader the run around. -->

Latest revision as of 19:38, 6 March 2023

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

The GMG package is one of the solver packages for MODFLOW supported by GMS. The GMG solver is an iterative solver based on a multi-grid approach.

Parameters include:

  • Inner convergence residual (RCLOSE) – The algorithm computes the 12-norm of the residual and compares it against RCLOSE.
  • Maximum inner iterations (IITER) – For each linear solution. The default is 100. Specify a smaller number for nonlinear problems to prevent an excessive number of inner iterations.
  • Outer convergence residual (HCLOSE) – For nonlinear problems. After each linear solve (inner iteration), the maximum norm of the head change is compared against HCLOSE. Can be set to a large number for linear problems. Ignored if Maximum outer iterations is 1.
  • Maximum outer iterations (MAXITER) – Also referred to as MXITER. For linear problems, set to 1. For nonlinear problems, set to a higher number though it is usually unnecessary to go above 100.
  • Damping parameter (DAMP) – A value of 1.0 should be used for linear problems. For nonlinear problems, a value less than 1.0 but greater than 0.0 may be necessary to achieve convergence.
  • Damping option (IADAMP) – Flag that controls adaptive damping. Contains the following options:
  • "(0) Constant"
  • "(1) Cooley adaptive"
  • "(2) RRR adaptive"
  • Maximum damping (DUP)
  • Minimum damping (DLOW)
  • Maximum head change (CHGLIMIT)
  • Output flag (IOUTGMG) – Flag that controls the output of the GMG solver. Values can be 0–4.
  • 0 = Only the solver inputs are printed.
  • 1 = For each linear solve, prints the number of PCG iterations, the value of the damping parameter, the 12-norm of the residual, and the maximum norm of the head change and its location (column, row, layer). At the end of a time/stress period, the total number of GMG calls, PCG iterations, and a running total of PCG iterations for all time/stress periods are printed.
  • 2 = The convergence history of the PCG iteration is printed, showing the 12-norm of the residual and the convergence factor for each iteration.
  • 3 = Same as 1 except the output is sent to the terminal instead of the output file.
  • 4 = Same as 2 except the output is sent to the terminal instead of the output file.
  • Maximum head output (IUNITMHC)
  • Multi-grid preconditioner smoothing (ISM) – Flag that controls the type of smoother used in the multi-grid preconditioner.
  • "(0) ILU smoothing" – This smoothing requires an additional vector on each multigrid level to store the pivots in the ILU factorization.
  • "(1) SGS smoothing" – Symmetric GaussSeidel (SGS) smoothing.
  • Multi-grid preconditioner coarsening (ISC) – Flag that controls semi-coarsening in the multi-grid preconditioner.
  • "(0) Row, Col., Layer" – Coarsens all the rows, columns and layers.
  • "(1) Row, Col." – Coarsens all the rows and columns.
  • "(2) Col., Layer" – Coarsens all columns and layers.
  • "(3) Row, Layer" – Coarsens all rows and layers.
  • "(4) None" – No coarsening.
  • Relocation parameter (RELAX) – Used to improve the spectral condition number of the ILU preconditioned system.

For more info on the GMG solver look at the following document:

Wilson, John D., and Richard L. Naff. MODFLOW-2000, The US Geological Survey Modular Ground-Water Model—GMG Linear Equation Solver Package Documentation. No. 2004-1261. 2004. [1]
The MODFLOW GMG Solver package dialog.