Re: [cloudy_simulations] PROBLEM qheat did not converge grain pah-ab0805 in zone 422, error code = 14

Hi Patrizia,

Thanks for the report! I could reproduce the problem. The basic problem 

is that this simulation is spherical and also goes very deep, which 

causes the radiation field to become very diluted. This means that 

grains are only very infrequently hit by photons and they can cool to 

very low temperatures in between those hits. In fact a noticeable 

fraction of the grains cools below the lowest grain temperature that 

Cloudy treats (which is 1 mK). The stochastic heating code will then 

complain that it wants to lower the lower bound of the temperature 

distribution, but isn't allowed to do so since it already is at the 

minimum temperature that Cloudy can handle. I have relaxed the checks so 

that it doesn't complain any more in that particular case. The radiation 

from grains below 1 mK can be safely neglected -- they only emit very 

weakly at wavelengths beyond 1 meter. Such grains are also completely 

unrealistic by the way since quantum effects would take over at these 

very low temperatures (the stochastic heating code in Cloudy uses 

classical physics, which is a pretty decent approximation) .

If you pick up the new code here:

http://viewvc. nublado.org/ index.cgi/ branches/ c08_branch/ source/grains_ qheat.cpp? revision= 2696&root= cloudy

your sim should run OK (I assume you are using c08.00). However note 

that in these kind of simulations conditions get very extreme, which 

means that convergence problems can never be completely ruled out. If 

that happens, the code will revert to equilibrium treatment and still 

add in the emission from the grains under those assumptions. So there 

will not be an energy leak and the spectrum should still be a good 

approximation. Most likely the uncertainty in the optical properties of 

the grains is a far bigger source of error in the grain spectrum than 

this problem.