[TRNSYS-users] Re : problem of convergence

[Yujun TAO <aleck1314@yahoo.fr>]

It's almost OK for my problem. Sorry for this inconvenient question.

I changed the tolerance convergence. I gave a higher value than 0.001.
But i want to know the exact meaning of this tolerance, for relative and absolute value. What's its influence's emergence for all the types and Trnbuild?
Can some one tell me?

---

De : Yujun TAO <aleck1314@yahoo.fr>
À : TRNSYS USERS <trnsys-users@engr.wisc.edu>
Envoyé le : Mardi, 24 Février 2009, 12h50mn 06s
Objet : [TRNSYS-users] problem of convergence

Hello,
I have a problem of convergence when i simulated a complex building with double skin ventilated. I found when the window "adjacent" has a big surface like 60m² in zone "DPSE", the simulation took many times and is difficult to converge. When the surface reduced like just has 10m², the simulation goes fast and has no problem of convergence. But the window should have 78m² in the reality.
So how could I resolve this problem?
Sorry the files bui and tpf couldn't be attached with mail, blocked by the firewall of yahoo.