[Birnensuppe]

Hey,

I want to simulate plasma flow with the P1 und DT radiation model. I set up each model as a multiband model. I am using temperature depending absorption coefficients for each band.
The P1 model crashes with the failure “divide by zero”. I checked the absorption coefficients, they are all unequal to zero. For low temperatures I have high absorption coefficient and I observed that if I lower the absorption coefficients for temperatures below 1000K, I do not get this failure.
The message is:
ERROR #001100279 has occurred in subroutine ErrAction.
Message:
c_fpx_handler: Floating point exception: Divide by Zero

The DT model crashes with the failure:
“ERROR #001100279 has occurred in subroutine ErrAction.
Message:
Something went wrong in the secant
+--------------------------------------------------------------------+ followed by
ERROR #001100279 has occurred in subroutine ErrAction.
Message:
Stopped in routine CAL_RAD_TVOL “


I set an user function with the absorption coefficient for every 100 K.

Has anybody perhaps an idea how to solve these problems? I you need further information about my setup, please ask.
Thank you in advance for your response.

Best Greetings,
Lukas

[Vinn]

Hi, I have the same problem.
Did you resolve? Can you explain?
Thanks.

[QUOTE=Birnensuppe;617592]
ERROR #001100279 has occurred in subroutine ErrAction.
Message:
c_fpx_handler: Floating point exception: Divide by Zero

The DT model crashes with the failure:
“ERROR #001100279 has occurred in subroutine ErrAction.
Message:
Something went wrong in the secant
+--------------------------------------------------------------------+ followed by
ERROR #001100279 has occurred in subroutine ErrAction.
Message:
Stopped in routine CAL_RAD_TVOL “

[Birnensuppe]

Hey Vinn,

sadly I didn't solve the problem as I would like to.
I lowered the absorption coefficients for temperatures below 1000 K to 1[m^-1] for two freqency bands with the highest frequency of my setup. This is absolutely not accurate, so I am still trying to find a good solution.

Please keep me updated if you can find a solution.

Kind regards,
Lukas

[Opaque]

May I ask what is the characteristic length scale of your model ? Keep in mind that an optical thickness beyond 100 is not useful since the emitted radiation will be absorbed extremely close to the point of origin; therefore, providing no much difference in solution respect to the solution of 100.

The comment above is very conservative, in practice I find that accounting for values >> 10 do not provide meaningful "engineering differences".

However, it would be interesting to know the source of why the iterative method failed to converge.

Hope the above helps,

[Birnensuppe]

Dear Opaque,

my setup is simplified a cylinder with a height of 1 m and a radius of 0.5 m, therefore my characteristic length should be 0.167 m.
I also think that an absorption coefficient of 100 m ^ -1 should deliver the same results, as you say.
I simulate electric arcs, so you can simplify setup my to a plasma cylinder with a radius of 1 cm.


I also reduced the number of points for the temperature in my user function of the absorption coefficient and checked if there are some nearby temperatures with the same coefficient, this was not the solution to my problem.
I will keep you definitely up to date and will now work with the absorptions coefficient of less than 100 m ^ -1.
Thank you very much for your answer.

[Birnensuppe]

Hey,

the support confirmed a while ago that this is a bug in the software. They are working on fixing it.

Kind regards
Lukas

[Birnensuppe]

Hey,

the bug will be fixed in version 19.0.

Kind regards,
Lukas