This is an old revision of the document!
Please feel free to add (and answer!) questions. NOTE: There is far more to tell than we can fit into a tour… the visitors are starting to note “tour bloat”, and so am I. Perhaps the most popular questions in this FAQ could become a fact sheet to hand out.
(from Jessica Kolp) Each Wednesday at 3:30-5:00, there is a public tour at FRIB Trailer #6. For the safety of our visitors, the following rules apply for FRIB construction site tours:
(From Sam Austin's book, Up From Nothing) Total funding to the lab from 1962-2008 has been $664.5 million in 2014 dollars. There should be another ~$100 million in funding by the time the CCF is shut down. Since the FRIB project costs about $730 million, the funding for cyclotrons at MSU just about equals it.
Because most of our parts are metal, 3D printing isn't yet useful for many purposes. However, NSCL now has enough demand for plastic parts that we have one 3D printer working constantly!
The higher energy (200 MeV/u for uranium) afforded by the FRIB accelerator will:
From hundreds of gigabytes to several terabytes!
A couple I've come up with lately:
The more isotopes we discover and learn about, the better our overall picture of how the nucleus works. If you were a biologist and tried to understand living organisms in general, you wouldn't limit your research to only a few species! Note also - while those short-lived isotopes do not exist on Earth, they could be made in stellar processes and thus be key to understanding astrophysics.
(from Ashley Garrett) The highest exposure to employees is for operators (300 mrem/year) who perform maintenance of the cyclotrons and target areas at NSCL. Depending on the amount of work in the cyclotrons, some end the year at 100 mrem or less. The Nuclear Regulatory Commission has estimated the loss of life expectancy due to exposure and how it compares to other potential hazards:
(from David Morrissey) The answer depends on the mass of the incident ion — the lightest ions will go up to 400 meters, the heaviest ions from the CCF will go about 25 meters. Note: the cycstopper is expected to accept high particle rates (>10^8/s) and manage ions with half-lives less than 50 ms.
Strategies for responding:
(from Georg Bollen, 6/20/14 Greensheet) Providing even limited multi-user capability is desirable for NSCL and will be mandatory for FRIB. Such multi-user capability can be achieved by accessing rare isotopes that are produced but not used and presently lost on slits in the fragment separator. A promising approach is the use of a helium-jet transport system. Also see 3/17/16 Greensheet
Generally, they are trapped in the A1900 Fragment Separator and completely unused. However, (from Georg Bollen, 11/8/13 Greensheet) the Applied Isotope Program is exploring how to make currently unused isotopes from a beam accessible for applications such as bio-medicine, national security, energy science and material engineering.
Our research involves the fragmentation of heavy stable isotopes to lighter and unstable isotopes, so not directly. However: (from Zach Kohley, 6/13/14 Greensheet) The Coincident Fission Fragment Detector (CFFD) is a new device that is currently being constructed at the NSCL for measurements of fusion-fission and quasifission reactions… [and] will guide the future experiments at FRIB aimed at the production of new neutron-rich isotopes of [superheavy elements].
NSF commitment 2015: $23 million annual operating budget. CCF upgrade cost $20 million. MSU also funds us. Start-up costs?
Our research into nuclei helps explain such things as the origin/abundance of the elements, how stars explode, the nature of neutron stars, and tests of the standard model!
(from Peter Grivins) It depends on the nuclei being accelerated. In a typical high energy beam like Oxygen-16/18, lots of neutrons can be generated. I've seen levels of at least 7+ R/hr neutron flux at a time. That means in about 1 hour, you will exceed the legal occupational limit for trained radiation workers. I'm also ignoring quality factors. So, just add up the totals: mild radiation sickness typically starts at 50 R, while certain death (LD 100/14) starts at 600-1000 R.
“Low-conductivity water” is used to cool the cyclotrons. LCW reduces corrosion in copper equipment.
(from Kortney Cooper) The gas cell in the N4 vault runs at around 85-95 torr/115-125 millibar, and the cycstopper should run at similar pressures. (This is slightly more than 0.1 atmospheres)
Tin (Z=50) does, with 10 stable isotopes.
The average seems to be 10-20.
One way is through a Particle Identification (PID) plot, which graphs particles according to their change in energy passing through the detector (proportional to Z, element) and their time of flight (larger masses are slower, so proportional to A, gives isotope).
(Note: these answers can still be improved upon)
Some examples of recent experiments are on the NSCL website, or see below:
According to RISAC report on 12/8/06: “Shell structure of exotic nuclei with knockout reactions”, “Shell structure changes in exotic nuclei”, and “78-Ni lifetime”. We did pioneer the use of a superconducting cyclotron, including one for neutron therapy at Harper Hospital. We are also doing important work in the field of astrophysics, helping to explain observed astronomical phenomena by studying the nuclear processes involved. Research here is helping explain the r-process (formation of heavy elements in supernovae) and the nature of nuclear processes/explosions on neutron stars. In early 2007, we discovered the heaviest known isotope of silicon. NSCL is also the birthplace for many new technologies (such as superconducting cyclotrons, next-generation ion sources, and radiation-resistant magnets) that are spun off into other applications.
Scientists must send in an application proposing an experiment. An international committee reviews all proposals and awards beam time to those proposals of the highest scientific merit and greatest attainability. The average time alloted an experiment is 120 hours. In the weeks and months before the experiment, the research group and NSCL operations plan and prepare, so all the bugs are worked out before the assigned time period begins. Time not spent experimenting is time lost. When the time comes, the project runs 24 hours a day, 7 days a week. Each hour of experimentation is the reason NSF supports us, and when there’s an experiment running it costs $5000/hr to operate the NSCL. Experimenters do NOT pay to do research here. We want to maximize the amount of time the lab is available for science, about 4500 hours in most years! Experimenters spend months preparing so all the bugs are worked out before their assigned time begins.
The National Science Foundation funds our operating budget, so beam time is free to researchers who are approved by the international committee. MSU contributes as well, and of course many outside users are funded by grants from DOE, NASA, NSF, etc.
Our fragmentation process requires a fast beam and produces a fast beam of rare isotopes, but certain experiments can't be performed on a beam traveling half the speed of light! Many astrophysical events (such as stellar nuclear reactions) occur at a much lower energy, and so it is userful to replicate that. Reaccelerated beams also offer higher beam quality and access to many experimental techniques that were developed for low-energy beams.
(from Paul Mantica) In Paul Mantica's group, the main “chemistry” connection is our NMR measurements to determine the magnetic properties of nuclei far from stability. NMR is one of the main characterization tools used by chemists; typically H-1, C-13, and P-31 are probed to look at local fields at atomic sites to determine molecular structure. Most odd-mass, stable nuclei have well-known magnetic moments, from which the unknown perturbative fields are measured (the chemical shift) and molecular structure defined. At the NSCL, we use NMR with a well known “perturbative” field to measure the unknown magnetic moments of exotic nuclei. Here, the magnetic property of the nucleus provides insight into the complicated stucture of the nuclear wavefunction. The magnetic fields we use are much smaller than those in chemistry NMR (factor of 20 or more), but since we use radioactive atoms our sensitivity for detecting the resonance condition is enhanced by 14 orders of magnitude.
Dave Morrissey's groups also has a significant chemistry component. Their measurements using the gas cell are providing critically important data for the molecular formation in the gas phase. The NSCL gas cell thermalizes the fast beams for subsequent use in the LEBIT Penning Trap Mass Spectrometer. The process of stopping involves a lot of chemistry. Although many believed stopping in He buffer gas would lead directly to +1 ion formation, Dave's group has seen a complicated array of chemical reactions and different speciation depending on the chemical element involved. To date, they have studied S, As, Ge, Ga, Br, Ca among others.
See Medical applications of nuclear science research on the NSCL website.
(from unknown source) There is more than one type of exploding star, and each explosive scenario makes a different mix of elements and operates in a different way. The very nature of an explosion, however, gives the various scenarios some common properties. Some are faster than others, but all are fast when compared to the timescale for standard nuclear burning, such as takes place in the sun. Because the processes are fast, it becomes likely that the unstable product of one reaction will participate in another reaction before it returns to stability. In fact, large portions of key reaction sequences take place away from stability involving rare-isotopes. These processes are not yet fully understood in large part because the properties of many of these isotopes are unknown. The NSCL is one of only a few facilities in the world that can study many of the rare-isotopes important to key explosive burning processes, which produced much of the current elemental abundances, especially those of heavy elements (heavier than iron).
Cosmic rays are much more intense in orbit and in space than they are on the surface of the earth. Standard electronic components function differently under bombardment by these high-energy particles. Your computer memory may start functioning like a particle detector. What should be a zero becomes a one after a heavy ion streams through. Components to be used in space are tested at the NSCL under the impact of individual high-energy heavy ions to make sure that astronauts can depend on them. (there may be other examples)
(from Wikipedia) Niobium is not found free in nature, rather as part of minerals such as columbite and pyrochlore. Brazil and Canada are the largest producers of niobium, which costs about as much as silver per ounce. Niobium is mostly used to strengthen steel, but is notable because niobium-titanium alloy remains superconducting (critical temp: 9.2K) under high current and magnetic field. Thus, the superconducting electromagnets at NSCL are made possible. (additional note: the NbTi wire costs about 5 cents per foot)
(from Dallas Cole) Many of our beams are created using monoisotopic elements or elements that are naturally very high in isotopic abundance of the isotope we require, such as 238Uranium or 40Calcium. For these beams we purchase materials from common domestic chemical suppliers, such as Alfa Aesar in Massachusetts. Most of our enriched isotopes used to produce beams, such as 48Calcium or 78Krypton come from Russia through either US or Canadian distributors. The US has an inventory of these isotopes at Oak Ridge, but does not actively produce them anymore. The competitive commercial culture in Russia makes them substantially more economical in supplying our needs. (from Andreas Stolz) Up to Calcium (Z=20) beams are produced with all electrons stripped off (totally ionized) in the K1200. The heaviest beam of those would be Ca-48.
(from Peter Miller) There are three dees in our cyclotrons for a couple of reasons.
1. Mechanical fit– With the magnet pole tip design we used, there is space in the three valleys between the hills to install dees with enough space around them to hold the operating voltage. They fit there even though the gap between the upper pole and the lower pole of the magnet is small (about 2 inches). This is essential to get the magnetic field profile we need. (We would not be able to put dees in the pole tip gap, as was done traditionally.)
2. More accelerating gaps (particles gain more energy per turn)– The ions get an accelerating kick every time the cross the gap between the dee and the hill (2 gaps per dee). With 3 dees there are 6 accelerations on each turn. The cyclotron would work with one dee, but there would be 1/3 the energy gain per turn, so the ion would have to make three times as many turns to reach the final speed. If you could triple the voltage on that one dee you would compensate for this effect, but this is not practical. We run the dees at the highest voltage that the structure will hold reliably without much sparking.
3. Symmetrical forces on the particles – By using 3 dees spaced symmetrically around the circle we avoid pushing the ions into an off-centered orbit when they enter the cyclotron near its center. Such errors in the centering of the orbit can damage the beam by causing it to increase in cross sectional area, even to the point where it will hit the dees and many particles will be lost before they accelerate to full speed. With only one dee, the particles would get two kicks from it (60 degrees apart on its path) and would coast all they way around (300 degrees more) before repeating the process. You can imagine how this asymmetry in the pattern of forces would tend to push the orbit off center, toward the side of the cyclotron where the dee is.
In general, it is true that a dee does not have to be D-shaped to work. A D-shaped one has a width of 180 degrees. The charged particle does not experience any electic force while inside the dee; when it enters or leaves the dee it is accelerated by the electric field that is concentrated just outside the dee in the gap as it passes through it.
The condition for the particle to gain maximum energy per turn is: « The particle must cross the center line of the dee just as the voltage passes through zero. » If the particle arrives at the center line of the dee earlier or later than the zero crossing time, the energy it gains from crossing the two gaps will be less than the maximum possible.
It is necessary to have the correct time synchronization between the revolving particles in the cyclotron and the oscillating voltage on the dees that produces the acceleration. In our cyclotrons the dee voltages are phased individually to be 120 degrees shifted in time, in order to match the condition for maximum energy gain.
(from David Poe, 10/09) The “k” value comes from the equation E⁄A=k(〖q/A)〗^2, where E/A is in MeV/u, q is the charge state and A is the atomic mass. Labels such as k50 and k500 refer to the strongest magnetic fields we can run. For a cyclotron accelerating protons, q=A and thus q/A=1, so k is simply the energy of the protons in MeV. The k value thus represents the energy of a proton beam that could be accelerated by the same field. Note: K500 was the first single-turn extraction cyclotron, assuring a precise knowledge of the energy of particles on their exit.
See the wiki about cyclotrons: Advantages/Limitations of Cyclotrons. To summarize, the three pole tips in our cyclotrons can be used to accelerate the same nuclei again and again. This translates to a major savings in cost and space. Also, the design of the K1200/K500 compensates for relativistic effects, so they can accelerate nuclei up to a significant fraction of the speed of light (the other wiki asserts a limitation of about 1% light speed).
They can get to higher energies by adding length, while cyclotrons are limited by their set diameter. LINACs do not require magnetic fields to constrain the particle to a circular path (which results in synchrotron radiation and loss of energy). FRIB will produce rare-isotope beams of increased intensity and variety that will ensure world-competitive research past 2030.
There are many cyclotrons in the world, and that number is increasing as they are built for industrial/commercial uses, such as making rare isotopes and medical treatments. The largest is the 56-foot-wide TRIUMF in Vancouver, Canada. The K1200 right here at NSCL produces the second-highest energy (up to 200 MeV/u). Our closest competitors in rare isotope research are RIKEN in Wako, Japan, which uses three cyclotrons including a recently-completed (and new highest-energy) superconducting one, and Gesellschaft für Schwerionenforschung (GSI) in Darmstadt, Germany, which uses a linear accelerator and synchrotron. You may know other, larger accelerators, like those at Fermilab and CERN: those are called “synchrotrons” and while they can achieve much higher energies than our facility, they are used for high-energy physics (probing the nature of individual particles, rather than a nucleus).
Acceleration is dependent on the charge state of the injected nuclei. By allowing the Ion Source to produce ions with lower charge states and pre-accelerate them with the K500, we can strip MORE electrons off the faster beam nuclei and get more efficient acceleration with the K1200. The result: higher beam current AND energy.
(from Chris Magsig) The wire that we use is for beamline dipoles and quads. It is around .040“ to .080” in diameter which is somewhat smaller than the S800 or K1200 wire. The copper to superconductor ratio is around 5 to 1 making it fairly pliable. The wire has a formvar insulation to prevent shorts. We do have to be careful not to kink or abruptly bend the wire to prevent breaks in the superconducting filament. We use devices like air and magnetic particle clutches to maintain a consistent tension on the wire while winding. The winding table uses pulleys and springs to keep the wire from breaking. There is a digital readout that is linked to a linear resistor to monitor the predetermined wire tension. Quads are around 1.5 pounds and dipoles are around 8 pounds. The wire tension is the most critical factor of winding. If it is too much, the wire will break and if it is too low, the coil has loose wires and there is not enough room in the curing form. After winding we record resistance, dissipation, inductance, and a quality factor. We determine how good the coils are based on these values.
While the ion is confined within one cubic mm in the Penning trap (a combination of a strong magnetic field and a static electric quadrupole field), one can tune another RF field to find a resonance that will match the cyclotron frequency of the ion, the rate at which it “orbits” determined by q/mB. Knowing charge and field strength, find m. LEBIT's precision is 10 parts in a billion, which is equivalent to weighing a 100-ton steam engine and tell whether the engineer had left a one-dollar bill on board.
(from Raman Anantaraman) The minimum time has shortened significantly in recent years, down to about 12-14 hours. Having 24 hours alloted is preferable, however!
(from Andreas Stolz) For a fragmentation target it is good to use material with a low element number. You need nucleons for the fragmentation process, but at the same time you want to have a small electron density. Electrons in the target material would only slow down and blow up the beam, but not trigger the production of rare isotopes. Beryllium also has very good thermal properties. That means with the combination of low density, relatively high melting point, and a very good thermal conductivity the beryllium targets can be used with a high power deposition in the target without melting. Beryllium is actually the best material available under these considerations.
Actually, you can separate ionized U-235 and U-238 with a magnetic field, but there are more methods of enrichment that are much more efficient. While this technique was employed during WWII, it has essentially become obsolete.
These nuclei get stopped in the beamstopper or in the magnets.
Beam density is measured in current… that's a good description for a beam of moving positively-charged ions. Our ECR can produce varying amounts of ions, depending on the desired element. One example is a “primary beam” from the source of oxygen-16. The ECR can reliably produce 100 pnA (particle nano-amps) of this element, which equates to between 10 and 100 billion particles per second (a beam power of about 1.3 kW). Once the CCF has fragmented the primary beam and produced a filtered rare isotope beam, it represents a tiny fraction of the initial current. Beams made of isotopes near stability will contain millions of particles per second, while extremely unstable/rare isotopes could be produced at the rate of one per hour or even less!
Total travel time is less than one hundred microseconds, dominated by the time it takes to accelerate from the ion source through both cyclotrons. Total travel length is almost three miles.
Glad you asked!
(from Peter Miller) The radio frequency amplifiers that output the AC (frequency about 23 MHz) that is needed to accelerate the particles in the cyclotron, need 20 kV DC for the vacuum tube in the final stage. We are talking about 300 kW of power input (20 kV and 15 amperes) for each amplifier, one per dee. If you had a source of 600 V DC that would deliver the necessary power (would need 500 A for each amplifier), you would need to convert it to AC and step the voltage up to 20 KV to match the vacuum tube requirement using a transformer, then convert back to DC.
Our power supply takes the 13.2 kV three phase AC power from the electric power line and, by means of a transformer, produces the correct voltage, converts it to DC, and delivers that to the amplifier.
The size of the transformer needed is determined by the power level, in this case we need about 1 Mega-watt for the K1200. Our transformer, and the spare, is located behind a fence outside the High Bay between the building and the large white helium gas tank.
The dee voltage of 140 kV is generated in the resonant copper cavity surrounding each dee. The voltage on the dee is several times larger than the voltage output from the amplifier because of the effect of the resonance between the current in the cavity and the driving current from the amplifier. The voltage builds up over many cycles of the driving signal if the natural resonant frequency of the cavity has been adjusted to match the frequency of the driving signal.
The steel shell around the cyclotron is called the “yoke” and serves to propagate the magnetic field, making it as strong as possible.
(from Allyn McCartney) The liquid helium (LHe) flows to the superconducting magnets inside coaxial vacuum jacketed stainless steel pipes. There are multiple layers of insulation inside the vacuum space separating the inner and outer pipes. Often there is also a liquid nitrogen cooled copper shield between the 4.5 K helium pipe and the 300 K outside world. The outer most vacuum jacket can be either steel or stainless steel. I have several drawings that illustrate some of the various schemes we use to insulate the helium pipes. When we cool the stainless steel pipes they contract enough that we have to allow some movement when they go back and forth between warm and cold. We use flexible stainless steel hoses at times to allow for thermal contraction without pulling the welds apart.
Many of our beam line magnet connections do not use stainless steel for the helium lines. They use a special allow called Invar (also known generically as FeNi36 (64FeNi in the US), is a nickel steel alloy notable for its uniquely low coefficient of thermal expansion (CTE or ?))From Wikipedia.
Also we don't actually use liquid helium pumps, but push the cryogenics around the lab with pressure differentials. We run the LHe at around 11 psig and the helium compressor inlet low pressure side is just above 1.5psig.
(from Raman Anantaraman) The lab's annual liquid helium budget is over $300,000.
(from Paul Zeller, 9/13) Currently the facility is recorded at 235,963 square feet. The SRF high bay and the FRIB project will approximately double that number. I tell people we are about a quarter million square feet, and after FRIB is constructed we will be about half a million square feet.
(from Al Zeller) The steel came from Japan, the conductor from Italy and the coil bobbins from Colorado. Everything else was procured locally. The complete dipoles were built here in the S3 pit and the East High Bay.
Read up on Fiestaware and lantern mantles at the Oak Ridge website. However, the pertinent exposure data are:
Thanks to the NSCL, MSU's nuclear physics program is ranked number one in the nation (now ahead of MIT) by US News & World Report in its 2010 issue of America's Best Graduate Schools. NSCL currently educates about 10% of the nation's PhD students in nuclear science (physics and chemistry). We also educate and employ many undergraduates.
Dark green = poured concrete, permanent. Light green = stacked concrete blocks, movable. We're also getting in some large, shaped “Morris” blocks for the rebuilt walls in this summer's reconstruction.
The cyclotrons produce 3-5 Tesla over a fairly large area. The floating magnet is Neodymium, a powerful rare-earth magnet. Even so, as an N35-grade magnet (a classification of power for Nd magnets), you would probably measure 0.25 Tesla about 1/16” away from its surface. The Earth's magnetic field is 0.00005 Tesla; our cyclotrons are about 100,000 times stronger.
It is a YBCO superconductor, standing for Yttrium Barium Copper Oxide. This ceramic becomes superconducting at about 92K, which means that LN (77K) is sufficient to cool it. More info at the Wikipedia YBCO article. You can find places to buy YBCO superconductors.
(from Helmut Laumer) The present contract is $0.26/gal of liquid nitrogen and $9.4/gal of liquid helium. The production power cost for liquid nitrogen (which is available free from the air) is half the price; hence not cost effective to produce our own. The reliquefaction power cost for liquid helium is about $0.25/gallon, hence that is cost effective and we only buy gas to cover our losses. We get both from a chemical supplier.
The Meissner effect in superconductors like this black ceramic yttrium based superconductor acts to exclude magnetic fields from the material. Since the electrical resistance is zero, supercurrents are generated in the material to exclude the magnetic fields from a magnet brought near it. The currents which cancel the external field produce magnetic poles which mirror the poles of the permanent magnet, repelling them to provide the lift to levitate the magnet. The levitation process is quite remarkable. Since the levitating currents in the superconductor meet no resistance, they can adjust almost instantly to maintain the levitation. The suspended magnet can be moved, put into oscillation, or even spun rapidly and the levitation currents will adjust to keep it in suspension.
(from Terry Monahan) Someone slipped on some ice in the parking lot and was hurt. They missed one day of work. It occurred around 2001.