VUSDFLD - successful datacheck, error during analysis

[Manuel Costa]

Hello guys!

So, I've been facing a wierd error during explicit analysis with a VUSDFLD subroutine. So long I've trying to define my displacement at failure in '*Damage Evolution' as a function of a field variable.

As running the datacheck with a subroutine VUSDFLD, the datacheck shows:

EXPLICIT EXECUTABLE TIME SUMMARY
...
THE ANALYSIS HAS COMPLETED SUCCESSFULLY

But as running the analysis, I have this:

Abaqus/Analysis exited with errors

But my .sta and .dat don't show any sign errors. I am worried maybe my problem is too complex (8825 elements, 10315 nodes), or there is something wrongly defined in my subroutine. I am attaching my subroutine and .sta \ .dat files.

Thank you guys!

Subroutine
.sta file
.dat file

PS. I am using Visual Studio and Intel Fortran Composer 2015.

 

[FEA way]

Did you try running this analysis without the subroutine ? Maybe the problem is not caused by VUSDFLD and there's some error in the model itself. Also check the warnings reported in the sta and dat files. They may give you a hint about what's wrong. Some of them suggest sort of overconstraint.

 

[Manuel Costa]

I tried to run it without subroutine, and it seems to be running fine.
although, something about that .dat file bothers me.
it says:

***WARNING: THE *USER DEFINED FIELD OPTION HAS BEEN USED, BUT THERE ARE NO
USER-DEFINED FIELDS ACTIVE IN THE MODEL. USER-DEFINED FIELDS MAY
BE ACTIVATED BY THE *FIELD OR *INITIAL CONDITIONS,TYPE=FIELD
OPTIONS OR BY DEFINING MATERIALS WITH FIELD VARIABLE DEPENDENCIES

which has me wonder since I put the 'dependencies=1' in *Damage evolution.

 

[FEA way]

Apart from that you should also add User Defined Field and Depvar to material definition (they can be found in the General tab).

 

[Manuel Costa]

yes, the input file contains *user defined field and *depvar options written.

 

[Manuel Costa]

So, I re-wrote the subroutine and apparently fixed it. ONE TIP FOR GUYS WHO ARE TRYING TO GET INTO ABAQUS SUBROUTINES: Don't trust all the examples put into the verification guide and subroutines contained in the files, they are not good for learning purposes. You should start with the basics, i.e., the subroutine interface.

Now, I'm facing another problem, I don't know if I should open a new thread for this, but for now I'll post here.

Im trying to compute the stress invariants by doing the call vgetvrm, but I still thinking that my method is somewhat wrong.
Is there a more correct way of calling these deviatoric and principal invariants (1°, 2° and 3°) in VUSDFLD soubroutine?

This is what I've been trying.

dimension rData(maxblk*nrData),
* jData(maxblk*nrData),
* stress(nblock, nrData)

do 100 k = 1, nblock

call vgetvrm( 'S', stress, rData, jData, cData, jStatus )

S11 = stress(k,1)
S22 = stress(k,2)
S33 = stress(k,3)
S12 = stress(k,4)
S23 = stress(k,5)
S31 = stress(k,6)
FIRST = (S11+S22+S33)

SECOND = (S11*S22)+(S22*S33)*(S11*S33)-(S12**2)-(S23**2)-
* (S13**2)

THIRD = (S11*S22*S33)+2*(S12*S23*S31)-((S12**2)*S33)-
* -((S23**2)*S11)-((S13**2)*S22)

c Setting INV3 variable

inv3 = ((2 / 27)*(FIRST)**3)-(1/3)*(FIRST)*(SECOND)+(THIRD)

 

[FEA way]

Your approach seems correct. Just note that you should use S31 instead of S13 in the equations for SECOND and THIRD.