Search
Close this search box.
Notifications
Clear all

Isotropic viscoplasticity

8 Posts
2 Users
0 Reactions
981 Views
Posts: 28
Topic starter
(@catmandoo)
Eminent Member
Joined: 16 years ago

Dear Jorgen,

I am analysing the isotropic viscoplasticity and I was confused about one thing as follows:

I wrote the main program called HULL.f to check the results of UMAT subroutine. The Hull.f file ran well in the Fortran software as I thought but when I bring UMAT into Abaqus CAE to compile together with this UMAT, I always get the error

Error in job tensile: The executable /opt/abq/6.7-1/exec/standard.exe aborted with system error Illegal memory reference (signal 11). Please check the .dat, .msg, and .sta files for error messages if the files exist. If there are no error messages and you cannot resolve the problem, please run the command abaqus job=support information=support to report and save your system information. Use the same command to run Abaqus that you used when the problem occurred. Please contact your local Abaqus support office and send them the input file, the file support.log which you just created, the executable name, and the error code. Job tensile aborted due to errors.

I checked all parameters of material model. They are no wrong at all.

Actually I also doubt that there is a problem in Newton Raphson method to compute delta_plasticstrain....:confused:

Please let me know your good ideas based on your experience before for isotropic viscoplasticity within Implicit Euler.

Thanks a lot

Topic Tags
7 Replies
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

It sounds like you are on the right track. To figure out the details of the problem that you mentioned I suggest that you add print statements in your Fortran code in order to verify that the calculations work as you intended.

-Jorgen

Topic Tags
7 Replies
Posts: 28
Topic starter
(@catmandoo)
Eminent Member
Joined: 16 years ago

Yes I realized a problem in the agorithm.
You could have a look into a passage. I built two functions Yieldstress and Dyieldstress(derivative of yieldstress). I need to compute the variable dg by Newton iteration method

......
.....
DG = 0.d0
sigF = YieldStress(ep,dg,dtime)
!!! ep is accumulated plastic strain
phi = norm - wrz23*sigF
!!! wrz23=sqrt(2/3)
!!! norm is norm of trial stress
!!! Inelastic increment
if (phi .gt. Zero) then
!!! Compute inelastic multiplier DG
niter = 0
DG = 0.d0
!!! Newton-Procedure
DO WHILE (niter .lt. itmax .and. abs(phi) .gt. TOL)
!!! TOL = 1d-8
!!! Compute d phi/ d DG and update DG:
dphi = -2.d0*mu-(two/three)*DYieldstress(ep,dg,dtime)
!!!mu = = E / 2.0 / (1.0+nu)
dgold = dg
dg = dg - phi/dphi
if (dg .lt. 0.d0) then
dg = dgold/2.d0
end if
!!! Recompute YieldStress and PHI,
!!! Increment number of iterations
SIGF = YieldStress(ep,dg,dtime)
niter = niter + 1
phi = norm - 2.d0*mu*dg - wrz23*sigF
END DO

IF( niter .ge. itmax) THEN
!!! itmax=50
write(6,*)ERROR: No convergence of local Newton iteration
CALL EXIT ! STOP COMPUTATION
ENDIF
.....
....
The problem is that I face a strange thing. for example for the values here.
At first (step 0) with dg=0 we get phi>0,dphi0,dphi<0
step 2 : dg = 4, and compute phi<0,dphi<0
step 3: dg 0,dphi<0 of this dg
step 4: dg=4 and compute phi<0,dphi<0
step 3: dg 0,dphi not convergence to compute dg

I want to know how to solve this problem well? Please if possible, give me some nice ideas..
Thanks Jorgen

Reply
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

I recommend that you slowly debug your code line-by-line. You can do that by implementing it into a separate Fortran program, or as part of the user-material subroutine. Either way, you should verify each step and that it agrees with a pen and paper solution.

Old school approach that will work...

-Jorgen

Reply
Posts: 28
Topic starter
(@catmandoo)
Eminent Member
Joined: 16 years ago

Dear Jorgen,

Please could let me know what kind of error in Abaqus CAE? It made me confused...:frown:

Reply
Posts: 3998
(@jorgen)
Member
Joined: 5 years ago

I am afraid that your image is too small to read 🙁

-Jorgen

Reply
Posts: 28
Topic starter
(@catmandoo)
Eminent Member
Joined: 16 years ago

Dear Jorgen,

for creating a model into Abaqus CAE, when we run the program to model I sometimes see a kind of error as follows: (in the MONITOR)

....

The system matrix has 3 negative eigenvalues.

The system matrix has 3 negative eigenvalues.

The system matrix has 3 negative eigenvalues.

The system matrix has 3 negative eigenvalues.

The system matrix has 3 negative eigenvalues.

...

Please could you explain me about the error? or a certain best way to avoid this error while creating a model into Abaqus CAE.

Thanks

Regards

Reply
Page 1 / 2
Share: