Incorrect HD cooling rate in the low density limit

[sglover]

The expression given in Galli & Palla (1998) for the low density HD cooling rate is incorrect: it includes cooling from the 0->1 and 1->2 collisional transitions, but in the low density limit this should actually be 0->1 and 0->2. (NB Daniele Galli has confirmed to me that this is a mistake in the original paper). The attached patch for calc_rates.src corrects this problem.

The revised rate gives roughly 25-30% less cooling at temperatures between 100 and 200 K.

[sglover]

Patch for calc_rates.src (against r1965)

[rpwagner]

Simon, thanks for bringing this up. Brian, I'm kicking this to you to verify and implement.

--Rick