I have questions(at the end) regarding an idea I thought of before.. Using the Kinetic Molecular Theory of Gases (KMT), we have air molecules all moving in random directions at high mean speed with it's heat content as the mean of it's kinetic energy at room temperature. The movement never stops, maybe except in absolute zero temperature.. -First and foremost, I'd like to make it clear that I'd like to extract this kinetic energy into something that could achieve work.. -Let's say we have a 'lucky tube'(this is purely an example..) For some extreme luck, everytime the air molecules collided with the tube's inner walls, it hits the right spot(a crack, a dirt, a scratch, etc..) in a way so it always get deflected to the rear end of the tube. -In such process, what used to be random motion molecules of air molecules are achieving order, they begin to move from one end to the other an orderly fashion, randomness is being transformed to order slowly. -In later stages of development, all the the molecules are now moving in one direction along the tube, front to back. In this manner, they still have the same level of kinetic energy right? And the same heat content. -Then we decide to place a turbine at the rear end of the tube. The turbine is attached or geared to an axle that drives a wheel of the automobile, thus, achieving work. -After the air has passed thru the turbine, which extracted mechanical energy from the air, the air stream will slow down. -Since all the air molecules are all moving in one direction(maybe in a twisting fashion after passing the turbine) and have slowed down, Its mean kinetic energy has decreased, therefore the heat content also, and finally, temperature drops. Finally, we have an engine that extracts the heat content of the air, does work and exhausts a stream of air that is colder than before. Basically, this is an indirect form of solar energy.. Sun heats ground, ground heats air and our engine extracts that heat into mechanical energy. Engine should work as long as air is in gaseous state. My questions are: -Given, the 'lucky tube' does exist, is it really possible to create unidirectional movement with KMT considering other aerodynamic factors as boundary layer, Reynolds Number, laminar flow, Bernoulli Effect, etc? -I know this is silly, but for the sake of argument, will it break Conservation of Energy?(not IMO, but a lot of people I mentioned this can't believe I won't be breaking any law) -Is the idea totally impossible to recreate? Are there other ways?