[TRNSYS-users] Strange intermediate values for the inputs/outputs within an iteration timestep

[a8304506 at graduate.hku.hk]

Hello, Trnsys users.

I got an error in one simulation relating to the cooling tower (Type51).  I try
to chase the cause of the error by adding some output statements in the source
code to record particular parameters.  Then I found some strange results.  I
use a thermostat to control a water pump to supply water to the cooling tower. 
Hence, the inlet water flow for the cooling tower should be either zero or
design flow.  However, it seems that in some iterations, the logged water flow
differs from these values.  When it becomes a small value (but not zero), error
occurs within the subroutine Psychromatics.  I wonder if it is because I choose
a value for the minimum relaxation factor other than 1 (I suppose that the use
of a smaller value can enchance the convergence).  If it is so, then the
allowance for selecting the relaxation factor is probably a trap, as the
intermediate values for particular input/output may fall into some dangerous
ranges.  Indeed, I encounter similar problems occasionally.  Does anyone has
other ideas for the cause of the problem?

Best Regards
LEE, Chun Kwong
Research assistant
Division of Building Science and Technology
City University of Hong Kong