Thinking about Asimov

[George S]

Asimov built his robots with positronic brains. The word positron was first used in 1933 (before that it was the antielectron theorized in 1928 and discovered in 1932); he was writing in 1939. Was he, perhaps, the first to use the word positronic? If only he'd thought of imaging instead of robots he'd have arrived at PET scans. Positron Emission Tomography.

I first encountered tomography back the tomographs were done on the x-ray table. It was a way to take three to ten sheets of x-ray film and expose them simultaneously with each sheet imaging a different vertical depth. Lots of radiation and only a few slices. Next was computer-assisted tomography (CAT). Then PET ... tomography of metabolic activity.

The 3D images are built in the computer and 2D projections onto a screen are made that can be rotated about.

Could it be made more real by printing on a 3D printer? The closer the slices the finer the 3D printer prints.

It builds from a beginning when there is simply no image ... and then a single point somewhere is the first to be printed (perhaps simultaneous with [or nearly so]) ... and as that first slice hardens the print-head moves up to the next slice ... now at the printhead liquid plastic is beginning to change state and solidify ... any given drop will solidify in time.

Perhaps the device that produces the PET image might be said to have a positronic brain in part.

 

[lpetrich]

That was just an irrelevant future-tech detail. No need to get literal-minded about it.

In his works, the positrons were working parts of positronic brains, just like electrons for electronics. Positrons in PET scans are for probing the brain by introducing radionuclides whose atoms go to some places but not to others.

It must be pointed out that there is no way that a positronic brain can work. It takes a LOT of energy in one place to make a positron, the energy an electron would get with a million-volt battery. But if a positron encounters an electron, it releases that energy again, usually into 2 or 3 gamma rays. So a positronic brain would quickly give itself a big dose of ionizing radiation.

As to making the positrons, there are two main ways:

• Electromagnetic pair production. This requires accelerating electrons or nuclei and shooting them at some target material. If they have enough energy, then some of that energy will go into making electron-positron pairs.
• Weak interactions: positronic beta decay. This is from proton-rich radionuclides, and the most convenient way of making them is to shoot protons at some target material.

Needless to say, neither process is very efficient. I once found somewhere a figure of about 10^(-3) for a few hundred MeV electron beam. That's over a hundred times the energy needed to make the positrons themselves. That's not even counting the energy efficiency of a particle accelerator, something that I've found hard to get good numbers for.


In fairness, there is an approach to positrons that many semiconductors have. Holes. Absences of electrons. Semiconductors are first made extremely pure, then "doped" with various substances that either add or subtract electrons as desired. Subtracting electrons gives holes. Their connection with positrons is with an early interpretation of positrons as holes in a sea of ordinary electrons with negative energy.

 

[repoman]

lpetrich, did you make the milankovitch related videos on youtube about the long term eccentricity and inclination precession of the planets?



I can't read the axes of the graph, because of the low resolution.
Also for the inclination videos.

 

[lpetrich]

For the eccentricity, the axes are
e*cos(ω)
e*sin(ω)
e = eccentricity
ω = ecliptic longitude of the perihelion, relative to present vernal equinox

For the inclination, the axes are
sin(i)*cos(Ω)
sin(i)*sin(Ω)
i = inclination, relative to present Earth orbit plane
Ω = ecliptic longitude of the ascending node, relative to present vernal equinox

 

[repoman]

So, e and sin(i) are the amplitudes of disturbance from an ideal circular and non-tilted orbit.

This "mathematical space" representation is confusing to me. I want to be able to see an exaggerated eccentricity animation of the orbit slowly precessing and getting more and less elliptical, basically have the spreadsheet by of e(t) and ω(t). For each of those times draw the unique ellipse it represents.

The video makes me think that the rotational direction of precession can reverse for short times, is that true?

I am confused why e and ω are combined into one graph. What does the graph represent?

 

[lpetrich]

Here's how one might picture them.

The inclination one is sort of the direction of the orbit's north pole.

The eccentricity one is sort of the offset of the orbit's center from the Sun.

 

[repoman]

Thanks,

I understand what the orbital elements are, anyway, it does not matter so much if that graph is hard to understand.