Recover energy from temperature

[MarkE]

This is idea is to recover energy from only one temperature not hot and cold. Look at left volumes when they are compress from X and Y forces. If Y forces don't compress I need to move for each volume the thickness of each volume but like Y decrease in the same time the tickness, I need to  move lower distance. Do you understand  ?
I added an image for show the reduction of thicness of each volume, more and more.

RC4 the second law of thermodynamics is a tyrannical monster that prevents such ideas from working.  Countless people have tried to cheat the second law and it has bested all of them.

[rc4]

Here, this is pressure (temperature) that compress volumes at left, what is the distance volumes moves ? It's not S*e*P, with P pressure, S surface of all volumes and t the thickness of a volume at start. It's Se'P with e' < e. Each time a volume move at right it lost thickness, e4<e3<e2<e1<e. What energy I need ? It's ( (S1-S)e+(S2-S1)e1+(S3-S2)e2+(S4-S3)e3+(S5-S4)e4 )* P. Sure surface is the same but thickness decrease with time. Do you understand this idea ?

Where pressure is constant at 0.1 bar don't help to recover energy, it's possible to have just usefull things like second image shown.

[MarkE]

rc4 what I see is two exponential horns wrapped around a closed path and connected back to back.  In the best possible case it can take a very small amount of work to drive fluid (gas) around the loop.  It will still take work, and you will have to drive that with an external energy source. 

[rc4]

MarkE, forget the last image (I deleted it), think with linear volumes please.  Do you agree with: ( (S1-S)e+(S2-S1)e1+(S3-S2)e2+(S4-S3)e3+(S5-S4)e4 )* P energy needed for compress volumes ?

Look at image below, you can see red lines are the true volume I move inside P.

It's possible to think with linear velocity for understand. If velocity is V for all walls, red surface don't move at V, they move at V' < V, because they are compressed with Y forces. In a time t, if velocity decrease, the distance decrease so I need to give less energy.

[MarkE]

rc4 even under the best of circumstances you have to do work to compress.  You will get less than that work back when you expand.  In order to execute a full circuit, you need energy from the outside.  Let's suppose that you use the body of the machine itself as your low temperature reservoir.   Just sitting in the environment brings the machine up to the ambient temperature and you lose your gradient.