Dt in simulation

Remi Cote 270196.1509

To Bill P. (one of the Gods of my adolescence)

I am 31 and willing to learn (from my mistake?) and open-minded.

I have tried what you suggested about dt. As I understand it, it change
the stroke of the given system in the good direction. The stroke are
more numerous and therefore the resulting movement is more gracious, more
rafined. dt=0.0005 and G=1000 it is 5/10000 of G at a time. If G is
calculed for 1000 hours, it means 30 minutes. If G reoresent the gain
of heat per second, you have a very refined system. Very effective...

If I do as you said, at the first iteration, that represent 0.0005 sec
(by convention between me, you and the readers of this list) I got 10.
(initial temp is 0, setpoint=20, coolinfactor=0.3 by sec.)
10 because heat=G* (20-0) - cooling
G for a dt is 1000*0.0005
cooling is 0.3*0.0005=0.00015 (insignifiant)
so it is 0.5*20=10

At the second iteration, it is 15(10*0.5), third=17.5, 18.75,19.37
It reach plateau at 20, very precise, perfect...

I tried it with a cooling of 3000 per second, and then again I found
a plateau at 17. So what I learn is that your desing is efficient small
perturbation. I am only beginning to think about all this. My mind is
not set at all. I am going to think about it and formulate my idea in
a question. If anyone as a comment please communicate...