Why does heat pass from hot to cold?

[sameeralord]

Hello everyone,

Heat passes from high to low, molecules pass from area of high concentration to low concentration, pressure everything. I was wondering why this was the case. I mean I can think of it like a force reaching terminal velocity, but what about heat does it have a force?

 

[Integral]

We cannot answer WHY this happens. We have OBSERVED this behavior, based on the observations we have defined quantities and derived mathematical relationships which provide very useful results. But all of the science of thermodynamics is based on the OBSERVATION of heat (a quantity we define) moving from hot to cold.

 

[rcgldr]

Temperature is related to speed of molecules. During collisions between molecules, the speeds tend to average out, increasing the speed of the slower molecules, increasing the speed of the faster molecules, so the kinetic energy and the related temperature of the molecules tends to even out over time, causing the heat to transfer from "hot" to "cold".

 

[nonequilibrium]

Integral, I have to disagree... We don't just observe and take it as a fact, we explain it with more fundamental knowledge/observation. That's the beauty of physics.

As Jeff Reid noted, it is actually a statistical fact. It's all about dynamic equilibrium. If you'll ever take Thermodynamics, you'll actually use methods of statistics to DEDUCE (in large systems) heat flows from hot to cold :) Jeff Reid already explained the basic principle behind the heat transfer, but the particle concentration is much the same: it's not that the concentrations become equal and then everything comes to a halt, no, the particles keep going from right to left and it's just that equal amounts fly left and right, not changing the macroscopic view of things. On microscopic scale, there is no such tendency to equalize concentrations, it's all just random behaviour.

 

[SpectraCat]

Integral, I have to disagree... We don't just observe and take it as a fact, we explain it with more fundamental knowledge/observation. That's the beauty of physics.

As Jeff Reid noted, it is actually a statistical fact. It's all about dynamic equilibrium. If you'll ever take Thermodynamics, you'll actually use methods of statistics to DEDUCE (in large systems) heat flows from hot to cold :) Jeff Reid already explained the basic principle behind the heat transfer, but the particle concentration is much the same: it's not that the concentrations become equal and then everything comes to a halt, no, the particles keep going from right to left and it's just that equal amounts fly left and right, not changing the macroscopic view of things. On microscopic scale, there is no such tendency to equalize concentrations, it's all just random behaviour.

This is something of a chicken and egg problem, but I tend to agree with integral. The concepts of heat and entropy are *defined* in terms of our observations of the physical world. Therefore it is no surprise that all of those concepts to which you refer "explain" why heat describes the flow of energy from high to low temperature reservoirs ... that is a tautological argument. I think integral's point is that, in order for anything you are saying to make sense, it must be consistent with the 2nd law of thermodynamics, and that is an empirically derived physical law. So, the OP's question can be construed as "why is the second law true", which is unanswerable.

 

[nonequilibrium]

The 2nd law is true because of statistical facts. And they follow directly out of mathematics.

 

[SpectraCat]

The 2nd law is true because of statistical facts. And they follow directly out of mathematics.

What are these statistical facts you mention? What *postulates* do you take to be true for your derivation? Most importantly, how do you know your derivation is correct?

 

[nonequilibrium]

Well, you're asking me to explain the whole of statistical thermodynamics. But for example, note that S may have been defined in terms of heat in the olden days, but the definition of entropy is now S = k lnW (actually it's even more fundamental, but in physics this'll do) with W nothing but the numbers of microstates corresponding to the microstate. To count the microstates, you use probability theory, so basically mathematics. The same probability theory tells you that in large systems (aka systems that we are occupied with) the chance of entropy decreasing (in the case of particles moving randomly on microscopic level; actually, interactions are allowed too, they just complicate things) is so small that you'd on average have to wait longer than the age of the universe for it to happen.

 

[SpectraCat]

Well, you're asking me to explain the whole of statistical thermodynamics. But for example, note that S may have been defined in terms of heat in the olden days, but the definition of entropy is now S = k lnW (actually it's even more fundamental, but in physics this'll do) with W nothing but the numbers of microstates corresponding to the microstate. To count the microstates, you use probability theory, so basically mathematics. The same probability theory tells you that in large systems (aka systems that we are occupied with) the chance of entropy decreasing (in the case of particles moving randomly on microscopic level; actually, interactions are allowed too, they just complicate things) is so small that you'd on average have to wait longer than the age of the universe for it to happen.

Ok, I agree with all of that, but you seem to be missing my point ... what I am saying is that all of what you have written is correct because it is consistent with experimental observations of the physical world, not the other way around. I am not familiar with any mathematically-based derivations of the second law from more abstract principles. In order for you to build up the "probability theory" that allows you to count microstates, you need to have some guiding principles that tell you how to rank the probabilities of the different microstates (e.g. higher energy microstates are less likely). In the end, there must be some subset of principles that are derived from observations of the world around us ... these are the physical laws that we take as postulates. While I agree with you that those principles have become more and more abstract as our understanding of physics has deepened, they are still in there somewhere in some form, and that is the point I think integral was trying to make ... it is certainly the one I was trying to make.

 

[DaveC426913]

If you put this into an analogy, it might help the OP understand without all the theory.

If you set all the billiard balls on a pool table, this is loosely analagous to a cold object comprised of slow-moving (cold) atoms.

Now you start knocking the cueball around (representing a hot object).

Ignoring friction, so that the cueball's energy is conserved, you can see that, eventually, all the balls will be moving slowly - there will be few or no fast-moving balls, and few if any stationary balls.

So, what has happened is that the hot object has shared its heat with the cold object, averaging out.

 

[epenguin]

In order for you to build up the "probability theory" that allows you to count microstates, you need to have some guiding principles that tell you how to rank the probabilities of the different microstates (e.g. higher energy microstates are less likely).

You don't really impose a principle like that, all you do essentially is count the microstates that have the same energy, subject to conservation of energy. For instance if you have a billion molecules, you could have all the energy in just one molecule and none in the others. You could have it all in the first, all in the second,... all in the billionth molecule. A billion ways to distribute it. But if you share it out more evenly there are vastly more than a billion ways of choosing how to share it out. Actually sharing it out completely equally is not the most probable, the most probable is a negative exponential called the Maxwell Boltzmann distribution.