Solver SAMG crash every time and PPest errors.

[long]

I run a model for an River Plan. I still analyze sensitive some parameter. But every time I run MDOFLOW then Parallel PEST slayer report error on solver and OPTIMISATION RECORD did not show off as it should. I run model with SAMG solver in x64 edition GMS v9.2.3. I already try with v9.1.6 but got the same problem. And I have licence for SAMG solver. The error report by ppest slayer is:

C:\TEMP\GMS_7840\ppest\slave2>"c:\program files\gms 9.2\models\mf2k5\64\mf2k5_h5

_64.exe" 0<v0.00.09.in_1

MODFLOW-2005

U.S. GEOLOGICAL SURVEY MODULAR FINITE-DIFFERENCE GROUND-WATER FLOW MODEL

Version 1.11.00 8/8/2013

Enter the name of the NAME FILE:

This file is not a GMS MODFLOW superfile. Assuming NAM file

Using NAME file: v0.00.09.mfn_1

Run start date and time (yyyy/mm/dd hh:mm:ss): 2013/12/20 18:59:27

WARNING READING LMT PACKAGE INPUT DATA:

[sTANDARD] HEADER NO LONGER SUPPORTED; [EXTENDED] HEADER USED INSTEAD.

Solving: Stress period: 1 Time step: 1 Ground-Water Flow Eqn.

forrtl: error (65): floating invalid

Image PC Routine Line Source

libmmd.dll 000000000059DACC log10.L 474 log10_ct.c

samg_64.dll 000007FEDB4A457F Unknown Unknown Unknown

samg_64.dll 000007FEDB4A0AC7 Unknown Unknown Unknown

samg_64.dll 000007FEDB35E08D Unknown Unknown Unknown

mf2k5_h5_64.exe 000000013FEED1C8 Unknown Unknown Unknown

mf2k5_h5_64.exe 000000014021DF49 Unknown Unknown Unknown

mf2k5_h5_64.exe 0000000140547E5C Unknown Unknown Unknown

mf2k5_h5_64.exe 0000000140543B46 Unknown Unknown Unknown

kernel32.dll 000000007777652D Unknown Unknown Unknown

ntdll.dll 00000000779AC541 Unknown Unknown Unknown

C:\TEMP\GMS_7840\ppest\slave2>exit

Model completed: waiting for command to run model again.

(Press <Ctl-C> if you wish to stop PSLAVE execution.)

I do not know how to solve this problem!!!

[Alan Lemon]

Can you contact support@aquaveo.com so we can get your files. It sounds like it may be a bug in the SAMG solver.

Do any of the other MODFLOW solvers work for your model? PCG or GMS?

[long]

I try with GMG solver and got this " GMG ASSEMBLY ERROR IN SUBROUTINE GMG1AP" and OPTIMISATION RECORD still did not show off.

I try with PCG2 solver and it can run to the end but OPTIMISATION RECORD still did not show off.

I will contact support@aquaveo.com right now. But sad that my maintenance contract time will come to end very soon.

[Alan Lemon]

One other item. Before using PEST you should make sure that your model runs fine in "forward" mode. If your model does not converge then it will not do you any good to try running PEST with it so you will need to edit the model until it will converge.