MickLinux's Journal: PUBLIC DOMAIN -- direct-to-electricity heat engine

Okay, here's another idea: the direct-to-electricity heat engine.

Simply put, the efficiency of a real (not Carnot) heat engine is a function not only of the Carnot efficiency, but also of the number of undesirable accessible states, vs. the number of desirable accessible states.

So in that case, you want to use a gas with the fewest undesirable accessible states. Hydrogen? Well, not exactly. It's too reactive, for one thing. And there's a better gas, anyhow.

Free electrons. Yeah, electric current can double as your heat engine gas. As a bonus, electrons are what carry the heat. So double it up, and make your electric circuit the heat engine.

Those clearly will have fewer accessible states than a gas.

The heat engine compressor is composed of a flat layer of pins sticking out from a metallic plate; here's the schematic, as it looks along either the X-Z plane or the Y-Z plane (Z being vertical).

For one layer:

____/\_____/\_____/\_____/\_____/\_____

For multiple layers (more voltage):

____/\_____/\_____/\_____/\_____/\_____ (-) (Cool)

/\_____/\_____/\_____/\_____/\_____/\__

____/\_____/\_____/\_____/\_____/\_____

/\_____/\_____/\_____/\_____/\_____/\__ (+) (HOT)

Here's the process: you apply heat to the flat end (not to the pointy end). Heat transfer causes the electrons to boil to the far end, and the reducing radius magnifies the electric field. The electric field causesa electrons to jump from the surface, just as from the electron gun in your computer monitor. As they jump, they carry heat away.

Now, the voltage of this system will be small. But the smaller you can make the pins, and the more layers you can add, the higher the voltage and the greater the possible heat transfer. The greater the heat transfer, the greater the possible current. So you need to make these with chip production capabilities: condense, photograph and etch.

---------Addendum 12/17/03:

Jump over to

http://physics.berkeley.edu/research/zettl/projects/emission.html

and you'll see that there is an easy method for doing this. Get about 20 wafers for every volt you want, with gold evaporated onto one side, and a slightly thicker layer of insulating material added in a ring around the edge of the wafer. Then stack them within a vacuum. Seal. Now, the insulation acts to prevent the needles from contacting the gold backing of the next layer.
Attach a thicker layer of metal to the front (Gold surface) and back (nanotube surface); preferably, such that the front reaches around the edge and up to leave a strip of metal accesible from the back edge. That way, you can abut stacks of 20 of these one against the other, to get increased voltage.
    ______ ______
| ====== | ======
|______ |______

Now, heat the front side, cool the back side, and measure the voltage, current, and power output as a function of temperature.