Jake Young is a 
You have got to check out this story from NPR:
The compass has been around since at least the 12th century, but scientists still don't know exactly how the Earth generates the magnetic field that keeps a compass needle pointing north.But geophysicist Dan Lathrop is trying to find out -- by building his own planet.
His latest effort at the University of Maryland towers over him, a massive stainless steel sphere that looks like a prop from some old science fiction movie. Later this year he plans to fill it with molten metal and set the whole 26-ton ball spinning. At top speed the equator will whirl by at 80 miles an hour.
"It was a little scary the first time we spun it up," he says.
If all goes well, the planet will generate its own magnetic field.
If you go to the NPR site you can see a video of the 12 foot wide ball that they intend to spin. In order to simulate the molten iron at the Earth's core, they will be using molten elemental sodium. This is all fine and good experimentally. Sodium has a charge and is capable of creating a magnetic field. It also melts at something like an achievable temperature.
However, considering the possibility that some could escape, sodium is also a problem. It reacts violently with water, making fires challenging to deal with. (Don't believe me. Check out this video of even a tiny amount being put in water. Now multiply that by 26 tons.) Incidentally, where do you even get that much sodium?
So this is going to be a cool experiment that will no doubt produce interesting results. On the other hand, I will be quite pleased to be nowhere near the University of Maryland when they actually try it.
Physical Science
Comments
Wouldn't mercury (the metal, not the planet) be safer? Probably easier to work with, too.
Posted by: Ahcuah | June 2, 2008 11:59 AM
I don't think that mercury would work because mercury doesn't allow for current flow. It has been a while since I had this in chemistry, but I think the electrons in molten sodium are free to move around like the electrons in iron. This isn't true with mercury. So the rotation of mercury wouldn't causes a current and hence would not generate a magnetic field.
I imagine you would have to heat mercury a lot before it would give up electrons whereas Na will give up that first at the drop of a hat.
Posted by: Jake Young | June 2, 2008 12:34 PM
Well, mercury is (A) ridiculously dense (13.5 times denser than water, while sodium is about the same density as water); (B) way more expensive than sodium (sodium in industrial quantities is only about 20 *cents* per pound, compared to roughly 8-10 *dollars* per pound for mercury), and (C) so toxic that a spill of this magnitude would get tagged as a major chemical disaster. All it all, sodium is a much better choice. The risk of fire won't be all that bad, either, as long as the whole area is kept dry and there is a good spill-containment basin. There is a lot of experience in handling liquid sodium, too: it has been used as coolant in applications like fast-breeder nuclear reactors.
Posted by: tceisele | June 2, 2008 12:41 PM
To get this experiment to work you need a low viscosity to get as much turbulence as possible as well as a low resistivity to get as tangled a magnetic field as possible. (The ratio of these two is also important.)
I don't number the numbers, but liquid sodium (as opposed to, say, salt water) is chosen for its much lower viscosity and resistivity.
Posted by: csrster | June 2, 2008 2:00 PM
tceisele,
Thanks. I'm sure the density (and subsequent weight) is an important consideration. And cost, too, is a good reason. Should have thought of that stuff myself.
Jake,
I don't understand your comment about mercury "doesn't allow for current flow". It's a metal. Current flow is what defines a metal. That's why mercury is (well, was) used for such things as thermostat switches. Is there some other property of mercury that you are thinking of?
Posted by: Ahcuah | June 2, 2008 3:37 PM
Actually forget what I said Ahcuah. I guess I had it in my head that Hg doesn't conduct electricity, but it clearly does.
Posted by: Jake Young | June 2, 2008 4:55 PM
In the video of the "burning sodium", note that the sodium itself is not burning; the sodium reacts with the water to form sodium hydroxide (a powerful base) and hydrogen gas. The reaction is exothermic, and the hydrogen gas is hot enough to burn when it comes in contact with air.
If water comes in contact with the sodium metal, it will dissociate, and the hydrogen will burn. Once the hydrogen is gone, no more burning, no more danger.
Posted by: Stephen Schaffter | June 2, 2008 6:01 PM
Once the hydrogen is gone, no more burning, no more danger.
Well, except for the rather caustic Sodium Hydroxide resulting from that reaction.
Posted by: Left_Wing_Fox | June 2, 2008 6:34 PM
Consider this Core information;
The design and creation of Earth's core was and is more than just a random creation of nature. From the beginning Earth's core was Intelligently Designed and purposed for a specific celestial function . and has now begin a timely predestined shift of it's mantle plates in the at first gradual return back to it's original function this has been referred to as a tectonic event . Such regressing changes are those which has off-sit increasing earthquakes, rising surface temperature, and increasing temperature changes in ocean and sea floors which will continue heating up until all such waters reach boiling degrees and began to evaporate . .
BUT THATS JUST THE BEGINNING !. The mantled core will return to such a heated state until the very surface and crust of Earths surface will peel away from the core as one would peal off the hull of an orange and cast it into space . The Core of this Planet will become once again a temporary supper-heated source similar to that which scientist knew had to exist, but didnt understand where it came from, or what it actually was that cause the accumulation of cosmic element unto the combustion of that big bang . This in a much much smaller since makes our solar system ground zero for a very special work which will soon take place . Earths Core Facility will not be working alone.
Im just an Old Soul passing through, with something important to say .
Posted by: Evelyn SoD | June 2, 2008 8:32 PM
The ability to carry a current is not the issue, I don't think -- you want a metal that has some degree of magnetic susceptibility. Sodium has a small susceptibility, while mercury is diamagnetic -- it actually reduces the magnetic field it is placed in.
I don't see how this experiment could work, however -- iron is *much* more magnetic than sodium (like, 50,000x more or something), so he's using a much less magnetic material in a much smaller sphere. I'd be surprised if he detects anything, although I'm just some guy reading an NPR article, and he's got a 10-foot ball of sodium in his lab!
Posted by: Len | June 2, 2008 10:54 PM
Len - the basic principle of using liquid sodium to create a "homogeneous dynamo" has already been demonstrated, which much smaller set-ups. This experiment is just on a larger scale than anything previously. See, for example: http://www.physics.wisc.edu/news/06s-forest/Schwarzschild06.pdf
Posted by: Dunc | June 3, 2008 6:04 AM
"...sodium is also a problem."
They plan to neutralize it with chlorine gas, and French fry potatoes over it using the residual heat. Can you bring the sodas and burgers?
You'll find that these people are the salt of the Earth.
Posted by: Ian | June 3, 2008 8:07 AM
I don't think magnetic susceptibility is relevant - the Sun manages to create a dynamo in a hydrogen/helium plasma.
Posted by: csrster | June 4, 2008 7:15 AM