Hey Everyone,
I am running an Abaqus simulation with a subroutine and it terminated itself because of "floating point error". I tried after that running the simulation without the subroutine by giving it a material property, that doesn't require any subroutine. (In my case, the subroutine is for viscosity definition that is dependent on temperature and shear stress). With material property that independent to temperature and shear stress (no subroutine required), the simulation runs smoothly. That's why I think the problem has to be in my subroutine. That's why I would like to run a debugger to my subroutine by following the manual from https://www.ichec.ie/support/tutorials/abaqusdebug.pdf .
Following are my questions:
1. What could be gone wrong in subroutine that leads to "floating point error"? Excuse me, if the answer is obvious, because I have no or less experience with fortran. Following is the subroutine:
subroutine vuviscosity (
C Read only -
* nblock,
* jElem, kIntPt, kLayer, kSecPt,
* stepTime, totalTime, dt, cmname,
* nstatev, nfieldv, nprops,
* props, tempOld, tempNew, fieldOld, fieldNew,
* stateOld,
* shrRate,
C Write only -
* viscosity,
* stateNew )
C
include 'vaba_param.inc'
C
dimension props(nprops),
* tempOld(nblock),
* fieldOld(nblock,nfieldv),
* stateOld(nblock,nstatev),
* shrRate(nblock),
* tempNew(nblock),
* fieldNew(nblock,nfieldv),
* viscosity(nblock),
* stateNew(nblock,nstatev)
C
character*80 cmname
C
parameter ( one = 1.d0)
REAL, DIMENSION(nblock) :: n0, T
C
C VARIABLES
n = 0.2647
Tau = 309343
D1 = 1.68413e+13
D2 = 323.15
A1 = 31.206
A2 = 51.6
C
C Cross-WLF viscosity model
C
do k = 1, nblock
T(k) = tempOld(k) + 273.15
n0(k) = D1 * exp (-(A1 * (T(k) - D2)) / (A2 + T(k) - D2))
viscosity(k) = n0(k) / (one + ((n0(k) * shrRate(k)) / Tau)**(one-n))
end do
C
return
end
2. How could I link the subroutine with a debugger? I don't understand the 2nd step of the manual (login to stokes or stoney on two seperate termials). It would be ideal for debugging purpose, if I could read the values stored in variables for the subroutine.
I really appreciate for any input or advice from Everyone :)
Many regards,
Gunardi
Use this forum. Good luck.
http://www.eng-tips.com/threadminder.cfm?pid=799
In a simple program like this, you can just debug directly by simply printing out the relevant quantities using a write statement.
If I plot your function in gnuplot, I get a singularity at around T(k)=275; may be that's causing the trouble?
Thank you Herr Baeker for your answer.
Do you mean the singularity of the n0(k) when T(k) arround 275, as could be seen in the file img_1762? (If I understood you correctly)
I don't think there is something wrong with this equation, because this equation is a Cross-WLF equation to define the viscosity of a liquid (when the liquid still behave as a newtonian liquid at low shear rate). By knowing the value of n0, we could calculate the viscosity of the same liquid when it flows under higher shear rate by using the last equation (viscosity (k)).
The result of the last equation should look like the last photo (voila_capture...).
Still I really like the idea, that I can easily debug a subroutine by "seeing" directly the values that being transferred from abaqus to subroutine. Because I have to use subroutines in my simulation.
If anyone have experience in this area, some tips or advice would be really grateful.
- Badges
- Science topic