Like most techno-geeks, I am fond of large gadgets.
When I was a wee lad, Los Alamos Scientific Laboratory (it was still a scientific laboratory back then) completed the Los Alamos Meson Physics Facility. Like the ill-fated Superconducting Supercollider, LAMPF was basically a giant hole in the ground into which the government poured money. Unlike SSC, LAMPF was actually completed, and at approximately the time and cost originally projected. The facility consisted of a proton accelerator installed in a trench almost a mile long cut, into the top of one of the volcanic mesas that characterize the Los Alamos landscape. This accelerator could accelerate protons to an energy of 800 MeV, about a hundred times the energy with which protons and neutrons are bound into atomic nuclei, an energy characteristic of a temperature of almost 10 trillion degrees Centigrade. In fact, what I remember most about the opening of LAMPF was a giant banner strung up on one of the local cliffs, with the happy message: "800 MEV!"
When protons hit the right target with this much energy, they produce copious numbers of mesons, unstable particles which consist of a quark and an antiquark bound together by the nuclear strong force. Some short-lived radiation is also produced, which is one reason for placing the accelerator in a hole in the ground. Much of our current understanding of mesons and of the atomic nucleus came from experiments like those performed at LAMPF.
LAMPF has since become LANSCE, the Los Alamos Neutron Scattering Science Center. The accelerator is now optimized to produce neutrons for various scientific experiments. As an astrophysicist, I am particularly interested in the use of LANSCE to study r-process nuclei.
Let me explain that last part. I am part of a team of physicists at LANL that was recently awarded some internal research money to look at the effects of the very first supernovae on the early Universe. Supernovae are exploding stars, which for a few weeks shine with the brightness of an entire galaxy. They are believed to be responsible for producing all the heavy elements in the Universe. The very first supernovae, which exploded just a few million years after the Big Bang, are thought to have had a profound effect on the formation of the first galaxies. However, these particular exploding stars did not contain any heavy metals, so they are thought to have behaved somewhat differently from today's supernovas. Just what the differences were is a problem in computational physics.
Today's supernovae are classified into two broad categories, Type I and Type II. Type I supernovae are characterized by the absence of any hydrogen that can be detected with astronomical spectrographs. They are believed to be white dwarf stars, in which the original supply of hydrogen and helium has been completely converted to carbon and oxygen. White dwarf stars are not massive enough to burn the carbon and oxygen further; they simply do not have enough gravitational potential energy to heat their cores to the necessary temperatures. Instead, these stars collapse into superdense cinders the size of a planet, and slowly cool into oblivion. Sometimes, though, the white dwarf has a companion star from which it is able to "steal" enough mass to bring its carbon to the ignition point. When this happens in such dense bodies, the entire supply of carbon and oxygen burns at once, yielding iron, which has the most stable nuclei of any element. The energy released is sufficient to tear the star into pieces, leaving an expanding cloud of superhot iron.
Type II supernovae are more mysterious. They are known to be very massive young stars in which much of the original hydrogen is unburned. It is presently believed that they explode when their superhot cores begin to emit large quantities of neutrinos, which are fundamental particles with very small masses (around a millionth the mass of an electron) that pass right through any normal substance. In fact, countless numbers of neutrinos from the Sun are passing harmlessly through your body as you read this. At night, these neutrinos have also passed through the entire mass of the Earth without effect.
The neutrinos from the superhot core of a massive star pass right out of the star, taking their energy with them. This results in a catastrophic collapse of the core of the star to nuclear densities. If the star is not too massive, the strong nuclear force then reverses the collapse, sending a shock wave into the outer layers of the star that produces the visible supernova.
Many details of this process are not fully understood. It is clear that the "bounce" at the end of the collapse is not enough, by itself, to produce the supernova. Some other mechanism must get enough energy out of the core and into the outer layers to produce the explosion. The neutrinos themselves are probably responsible; at the ultrahigh densities involved, enough are reabsorbed to reheat the outer parts of the core. But most models that have been calculated so far do not get enough neutrinos out of the core to do the job. Perhaps rapid convection in the core is responsible for letting enough neutrinos out. We just aren't certain yet.
We are pretty sure about the end result, though: The outer layers of the star are blasted into space, leaving behind a neutron star whose density is that of an atomic nucleus -- an incredible quadrillion times the density of ordinary water. The outer layers contain large quantities of carbon, oxygen, and other elements from earlier stages in the star's lifetime, plus traces of r-processed heavy elements.
In the fiery cauldon of a Type II supernova, much of the matter is converted to neutrons, then blasted out at a temperature of 5 billion degrees Centigrade or more. Under these conditions, neutrons and protons freely combine with nuclei and are as freely torn off again. This sets up a kind of equilibrium between the nuclei present, in which the more stable nuclei are favored. But since the material begins with a huge excess of neutrons, the nuclei formed tend to be unusually neutron-rich. The weak nuclear force allows the excess neutrons to slowly decay into protons -- slow in a relative sense. The most neutron-rich nuclei live a few tenths to thousandths of a second before one of their neutrons decays into a proton. This seems like a short time, but the entire core collapse and rebound is over in just a few seconds.
The formation of atomic nuclei in a neutron-rich environment is known as the rapid process, or r-process. The heavy nuclei so formed show a distinctive patter of abundances that distinguishes them from nuclei formed by slow neutron bombardment (s-process) or proton bombardment (p-process.) Most of the heavy elements found on the Earth today appear to have been formed through r-process in a Type II supernova in the distant past.
A full understanding of a Type II supernova requires knowledge of the properties of neutron-rich nuclei. LANSCE has one of the best facilities around for studying these nuclei. The researchers there are identifying the masses, lifetimes, and decay modes of such nuclei, and this data will greatly improve the accuracy of Type II supernova simulations.
Assuming, of course, that LANSCE is allowed to remain in business. The Los Alamos Monitor reports:
The Los Alamos Neutron Science Center is approaching a geriatric crisis, the Department of Energy Inspector General reported last week. "The ability of LANSCE to provide needed research capabilities in the future is uncertain," Inspector General Gregory H. Friedman wrote in his cover letter to Energy Secretary Spencer Abraham.
...The accumulated deficiencies have caused the overall reliability of the beam to decline to 77 percent, which the IG said is 8 percent below the national average for similar accelerators.
Danneskiold said the lab had made an agreement with the NNSA to operate LANSC at a 75 percent level for 2003-2004, actually exceeding the agreement. He said there is a $138 million plan to extend the facilities lifespan.
So on the basis of an 8% deficiency in operating efficiency, as measured against an arbitrary baseline not mentioned in the operating contract, DoE is considering shutting down LANSCE. Sadly, this is not unusual for Federal "watchdogs," who have a habit of criticizing federally funded research centers on the basis of wholly meaningless statistical measures.
For example, it was not that many years ago that Sandia was criticized by the Congressional Budget Office for using only 36% of the compute cycles on its latest supercomputer. The truth is that the supercomputer was so modern that the computational scientists at Sandia were still trying to find the best ways to program it. You can't do that effectively in batch mode, but you can't do much better than 36% utilization except in batch mode. The cycle usage was completely meaningless as a measure of the quality of the research Sandia was doing with its supercomputer.
Basic scientific knowledge has the attributes of a public good. As a result, public funding of scientific research ought to be easy to defend, particularly when the research supports national defense, as does much of the research at Los Alamos. This does not mean unqualified support for boondoggles like SSC; but productive research facilities like LANSCE deserve better than to be dropped because they are 8% less "efficient" than some arbitrary baseline, particularly when that baseline is not mentioned in the contract for running the facility.