[TRNSYS-users] Convergence error with Type 56 and Type 23

[Technical Support]

Ethan,
  I suspect that what is happening is that your heater has enough capacity that it is overheating the cabin within a single time step. The two most common causes of that situation are that there is either too long a time step (TRNSYS can only make one control decision per time step) or that there isn't enough thermal capacitance in the system. One of the problems that occurs a lot in simulating is that systems and components are simplified in order to write algorithms that predict their performance and one of the simplifying assumptions that get made is that the component is always in steady-state (ie start up effects are ignored).
 
 If you are pretty sure that the heater capacity is modulable down to a level at which it can't overheat the space in one time step, that your time step is appropriate, and that your cabin-zone has enough thermal capacitance then I believe that Type23 has a mode in which it can be set up to make its control decision only after the system has converged. In essence what this does is to delay the controller's output signal by one time step so that its decision does not make the TRNSYS kernel iterate again until the next time step. There are also a few models available that will do the same artificial delay (Type93 for one). 
 Kind regards,
 David
 




Technical Support TeamThermal Energy System Specialists, LLC2916 Marketplace Dr, Suite 104Madison, WI 53719techsupport at tess-inc.com

----- Original Message -----

Subject: [TRNSYS-users] Convergence error with Type 56 and Type 23

From: "Ethan Lust" >;elust at umd.edu>

Date: Mon, May 5, 2008 8:10







Hello,
 
I’m having a convergence error that seems to be rooted
in the PID controller.  I was wondering if anyone
could shed some light on the subject.
 
I’m trying to model a long-haul truck sleeper cabin
climate control system.  I’m using a Type 23 PID controller in
conjunction with a “seasonal switch” as the climate control
unit.  The purpose of the seasonal switch is to  turn the heater on
and the air conditioner off when the temperature drops below some low deadband
temperature and vice versa when the temperature rises above the high deadband
temperature.  In other words, the purpose of this switch is to keep the
air conditioner from coming on during slight over-heating and the heater from
coming on during slight over-cooling.  The deadband in the seasonal switch
is set by default to 5 degrees Celsius.  
 
When just this group of components is run, the output signal
from the PID controller and seasonal switch match just as they should (climate
controller.tpf).  However, when a fuel-fired heater (Type 151, homegrown)
is added, the control signal starts doing some crazy stuff (Climate controller
with heater.tpf).  The control signals, which should align nicely, start diverging considerably.  This (I
assume) creates the convergence error that stops the simulation after a little
more than an hour with the current settings.    
 
Is there something inherent in the way I am trying to string
together these components that is making the solver diverge?  Is there a
better way to go about it?  I’m not looking for a total solution-
simply a pointer or two towards why the program is behaving as it is and what I
might do to get back on track.  Please let me know if you need any other
files.  
 
Thank you in advance for your time,
Ethan Lust
301.405.8672 (desk)
757.636.8727 (cell)
University of Maryland   

_______________________________________________TRNSYS-users mailing listTRNSYS-users at engr.wisc.edu
https://www.cae.wisc.edu/mailman/listinfo/trnsys-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/trnsys-users-onebuilding.org/attachments/20080506/8b79bc72/attachment-0001.htm>