[TRNSYS-users] TYPE 56 And convergence tolerance

[Diego A. Arias]

Dear Javier,

It is possible that your system has very different results, depending on 
the values of ventilation that you obtain with the equation.  I guess 
that your equations are working like a controller, that depending on the 
value of the room temperature it gives different values of ventilation.
In order to give some stability to this control strategy, you may try to 
use Type93 (Utility/Input Value Recall).

This is the description of Type93:
"The input value recal component remembers the values NI inputs for the 
preceeding NT timesteps. The values of NI and NT are set as parameters 
to the component. Such "input value recall" devices are useful in adding 
time delays to a system that is unrealistically sensitive to input 
changes or that suffers from control instability. By controlling based 
on values from a previous time step, systems will become much more 
robust. It is understood that in such situations, a small enough time 
step must be used so that the assumption of a delay in sensing the 
control signal is reasonable."

You just need to store 1 input for 1 timestep (these are the two 
parameters in Type93).   My idea is that you get the result of Type56 
(the zone temperature) and use it as input1 to Type93.   Then, you use 
the output from Type93 as the temperature that you use in your equation.

This strategy will give you a more stable simulation.
Best regards,

Diego



javier mota wrote:
> dear all:
> I AM USING TYPE 56 FOR BUILDING SIMULATION.
>     I GET TEMPERATURE OUTPUT FROM TYPE 56,
>     THEN I USE AN EQUATION TO CALCULATE VENTILATION.
>     IN THIS SIMULATION I GET MANY ERROR OF CONVERGENCE.
>     IS ANYTHING I CAN DO.
>     I HEVE PROVED TO CHANGE CONVERGENCE TOLERANCE BUT EVEN WITH A VALUE OF 1
>     THIS DOES NOT SOLVE ANYTHING.
>     WHAT MEANS THE CONVERGENCE TOLERANCE, IS IT  A MEASURE OF THE ERROR IN 
> ALL
>     THE EQUATIONS OF THE PROYECT.RELTIVE CONVERGENCE TOLERANCE MEANS THAT 
> THE
>     VARIATIONS BETWEENS THE SOLUTIONS OF CONSECUTIVE EQUATION (SOLUTION  A 
> STEP
>     N-SOLUTION AT STEP N+1)/(SOLUTION AT STEP N) IS LESS THAT THIS ERROR? 
> FOR
>     ALL THE EQUATIONS AND COMPONENTS IN THE PROJECT.
>     wHAT MAY BE THE MAXIMUN CONVERGENCE TOLERANCE THAT I COUL ADMIT 
> EXPECTING TO
>     GET CREDIBLE RESULTS?
>     THNKS IN ADVENCE.
>     jAVIER MOTA FROM 3I-INGENIERIA, MADRID
>
> _________________________________________________________________
> Dale rienda suelta a tu tiempo libre. Mil ideas para exprimir tu ocio con 
> MSN Entretenimiento. http://entretenimiento.msn.es/
>
> _______________________________________________
> 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