[TRNSYS-users] Message 441

[Diego A. Arias]

Dear Constanze,

I would not recommend reducing the convergence criteria as much as you 
are doing. This would make the results meaningless.

Remember that TRNSYS' solver is a numerical procedure, and you should 
always keep that in mind.   Think of your simulation as a system of 
equations. The default solver uses successive substitution in order to 
solve the system of equations.  All iterative numerical methods are 
based on a guess or initial values, that are used as starting point for 
the iterations.  If your system of equations is very complex, you have 
to give good guess values.  Sometimes, this feels like you have to give 
the answer so the program gives you the back the same answer.

My suggestion to debug the system is that you start by making sure that 
you do not get that warning message during one time step.   If you get 
this message, it means that you should change your initial values.  In 
the Studio, when you double click on a component, the initial values are 
those that appear in the tab 'input'.

Another suggestion is to try to find the information flow that is 
causing problems: try deleting the connection between some components, 
and see if your simulation converges.  This will show you the variable 
that is causing the convergence problem.

If you still keep having problems, please feel free to send me the 
project files and I will try to see what may be the problem.
Best regards,

Diego

Constanze.Bongs at csiro.au wrote:
> Hi,
>
> During simulation the following warning occurred:
> Message 441: The inputs to the listed units have not converged at this
> timestep. The values of the last iteration will be used to proceed. 
>
> As I get this message more than thirty times Trnsys sends an error
> message and stops the simulation. 
>
> I already set the tolerance convergence to 0.5 (from 0.001) and
> tolerance integration to 0.01 (from 0.001).
> Can you explain to me what is happening and how I can fix it? Does this
> mean that the system is too complex to be solved and are the simulation
> results still useful? Which maximum values for tolerance convergence and
> tolerance integration still manage to give me reasonable results?
>
> Cheers,
>
> Constanze
>
> _______________________________________________
> TRNSYS-users mailing list
> TRNSYS-users at engr.wisc.edu
> https://www.cae.wisc.edu/mailman/listinfo/trnsys-users
>   
-- 

Diego A. Arias
TRNSYS Coordinator
Solar Energy Laboratory
University of Wisconsin - Madison
1500 Engineering Drive
Madison, WI 53706