| | :::That not what my stat.mech. text says ;-). Let's put it this way: the ''entropy'' formulation only applies to isolated systems. In fact, I think that the heat engine in the Clausius formulation should itself be taken as isolated. If we allow it to be (thermally or otherwise) coupled to anything else than a hot and a cold "temperature bath", I think we could make the heat flow up the temperature gradient, at the expense of course of something happening outside the engine. Of course, these thermal reservoirs are abstractions, etc. etc. Anyway, here as well, either we should take the system (engine) as isolated, or we should take it as the entire universe. Both cases are not going to help disprove evolution at any rate. [[User:PaulB|PaulB]] 12:24, 13 March 2007 (EDT) | | :::That not what my stat.mech. text says ;-). Let's put it this way: the ''entropy'' formulation only applies to isolated systems. In fact, I think that the heat engine in the Clausius formulation should itself be taken as isolated. If we allow it to be (thermally or otherwise) coupled to anything else than a hot and a cold "temperature bath", I think we could make the heat flow up the temperature gradient, at the expense of course of something happening outside the engine. Of course, these thermal reservoirs are abstractions, etc. etc. Anyway, here as well, either we should take the system (engine) as isolated, or we should take it as the entire universe. Both cases are not going to help disprove evolution at any rate. [[User:PaulB|PaulB]] 12:24, 13 March 2007 (EDT) |
| | + | ::::I think even the entropy formulation will work with open systems, so long as one takes care to quantify the entropy flow through the boundary. The requirement then becomes that entropy decrease within the boundary is no greater than the net entropy flow out of the system through the boundary. [[User:Tsumetai|Tsumetai]] 12:37, 13 March 2007 (EDT) |