FERMI NATIONAL ACCELERATOR LABORATORY

 

 

 

 

 

 

 

MICHAEL COACHMAN

 

FERMILAB SUMMER INTERNSHIP IN SCIENCE & TECHNOLOGY

 

 

 

 

 

8 GeV TRANSFER LINE

MAGNET PRODUCTION AND DEVELOPMENT

 

 

 

 

 

 

 

Abstract: This paper will describe the importance of the 8 GeV transfer line project and the contributions I made.

 

 

The Magnet Maker

 

This summer at Fermilab I was pleased to be apart of the initial design and construction of the Recycler Ring. The Recycler Ring Project is one of the world's leading high energy physics experiments. The effect of the Recycler Ring on the Tevatron Collider operations is to radically change the means by which the antiproton and proton bunches are generated. The Recycler Ring will provide a long term repository for pre-cooled antiprotons, provide a reliable storage of antiprotons, and double the number of antiprotons for the next store.

 

These three objectives will save money, energy, time, and double the luminosity in the collider. By providing a long term repository for antiprotons the Accumulator Ring will run below 10 mA of antiproton current, whereas before it was running 200 mA. Using permanent magnets in the Recycler Ring will make it insensitive to lightning storms, power outages, computer glitches, and human error therefore reducing the antiproton storage failure rate of approximately 1 fault per week. Also antiprotons won't be discarded after use, but decelerated and re-integrated into the 8 GeV antiprotons in the Recycler, the total number of antiprotons available for the next store will be approximately doubled, doubling the peak luminosity of the beam.

 

Before the three thousand meter Recycler Ring can be built and placed in the Main Injector tunnel in the year 1998, several test must be done including the construction of a seven hundred and fifty meter 8 GeV proton transfer line from the Booster to the Main Injector using the same permanent magnet design. The permanent magnets for the line will be completed in December '96, installation into the Main Injector tunnel for use will be completed in spring '97. My job this summer has been to help build, perform test, analyze data, and draft procedures for testing and analyzing data from the magnets that will be used in this 8 GeV transfer line.

 

 

Magnets are constructed of stainless steel walls, packed with 4x6x1 and 4x3x1 inch strontium ferrite bricks, and aluminum spacers, all of which are inexpensive materials. The 8 GeV transfer line will consist of two types of magnets, Double Dipoles and Gradients.

 

The name Double Dipole comes from the fact that the bricks inside the magnet are dipoles and are glued together in pairs, the name Gradient comes from the fact that the inner pole piece assembly is made into an opposite sloping tunnel shape. There will be roughly 106 magnets in the 8 GeV transfer line: 41 Double Dipoles and 65 Gradients. In the Recycler Ring there will be roughly 344 Gradients: 216 will be 4.0 m long and 128 will be 2.67 m long.

 

I began the summer at the beginning of the magnet making process, magnetizing the bricks which fill the magnets. We buy our strontium ferrite bricks non-magnetized and we magnetize them using a 13.5 ton electromagnet ramped to 2000 amps at 4.0 volts. I learned to use a Sun Station running UNIX to control the ramping process.

 

Once we have shoved each brick into the electromagnet we then switch the Sun Stations program to test mode and we measure the magnetic field strength of each brick and store all the information using the UNIX program. Each magnet consists of 96 full bricks (4x6x1 bricks) and 176 half bricks (4x3x1 bricks), considering that bricks are magnetized in pairs and measured individually, 30 seconds per process, this part of magnet assembly is very time consuming. Although a device which can do both procedures at the same time is currently in production and should be completed by the beginning of August '96.

 

 

UNIX is nice for collecting data, but is limiting to say the least if you want to manipulate data. Therefore I learned how to transfer data from the Sun Station to a Macintosh in a Microsoft Excel file; my colleagues and I later drafted these procedures into an instructional document so that others that follow can also perform this task. Once in Microsoft Excel we analysis the brick data and construct stacking plans on how the bricks should be placed inside the magnet for a desired field strength.

 

After the stack plan has been followed and the magnet has been fully assembled my colleagues and I go back to the Sun Station and measure the magnetic field strength and flux inside the magnet. We use a flipcoil device connected to integrators that are run by the UNIX program. From the measurements we take we extract and/or add bricks to the removable end-brick positions on the ends of the magnet. Each brick contributes a different percentage to the overall strength of the magnet, a combination of sorts which sum to 4.6% per set of end-bricks. Our target magnetic flux is .7545907 V-s, we often start with a magnet that has a magnetic flux 3.68% too high and trim it down to a 0.05% error, although 0.20% error is acceptable.

 

Trying to trim a magnet to a target strength is not as direct a process as it may seem. Each brick has a different magnetic strength therefore each magnet has a different strength and each end-brick for trimming proposes in each end position contributes a differ percentage to the strength of the magnet. These combinations of strength variations, unlike the textbook world of school, re-iterated that physics is not an exact science, nor is the real world. So, with a lot of trial and error my colleague and I found the averages and standard deviations of each magnet and end-brick positions and combinations. From these experiments we formulated a magnet trimming procedure for future production trimming.

 

Some of the experiments we performed on the magnets include time dependency tests. These were done to find out at what rate does the magnetic bricks inside the magnets lose strength over time. We found that they lose strength in a logarithmic fashion. We performed several temperature dependency tests and have found that temperature changes fluctuate the bricks' magnetic strength in a random fashion. These test were done over a temperature range of 32o to 72o F; considering that the temperature in the tunnel varies from 55o to 95o F the data from these test is very informative. We also performed several freezing tests to see if freezing a magnet would change its chemistry and therefore its strength. We found that by freezing a magnet its weak crystal lattices would straighten into correct formation from the shock and actually increase the magnets strength in the range of 0.02% to 0.16%.

 

My opinion of the 8 GeV transfer line project and the use of permanent magnets is positive. We have the right people creating the line, everyone full of ideas, open to change, and completely knowledgeable of magnets and what it will take this line to run. Through magnets themselves are pretty tricky. I believe we can make them to the correct strength and flux, we can align them to accommodate the specs for the line within reason of the sextapole correction magnets, and I believe the line will hold the beam of protons. Although I am worried about future problems we may not know of yet.

 

The magnets are very fragile, the strontium ferrite bricks are brittle and shatter with very little force. This could pose problems because we won't know if a brick is shattered of fractured due to transportation or storage until the magnet is in the line. If the brick is only fractured then an initial test wouldn't show the problem, yet shattering is eminent.

 

 

Future problems may exist due to the energy the bricks are losing to the beam and each other. The magnetic bricks are constantly attracting and opposing each other inside a magnet producing a slow loss of energy in themselves; and when beam is introduced after storage even more energy will be lost to the beam, further weakening the bricks. This makes me wonder if the magnets that have been stored longer will be weaker. I also wonder if the loss of bonding energy will affect how easy it will be to shatter or fracture a brick. Finally, one must ask if the bricks will become more vulnerable to temperature changes in the tunnel due to moisture, longevity, or loss of energy.

 

In time I do believe that all problems that may come up with the 8 GeV transfer line will be solved. The Recycler Ring is of a much higher magnitude and will pose many higher order problems, some of them we may have solved form the 8 GeV transfer line and others we will have to come up with answers as we go along. In the end I know that the permanent magnet design will work and that I served an important part on the team that produced a new era of Collider operation at Fermilab.

 

Return to the main page.


Author:
Thomas Egan of Marist High School, Chicago, IL. This project was constructed as part of the Teachers Research Associate (TRAC) Program from the Fermi National Accelerator Laboratory in Batavia, IL. This project is also in conjunction with Aurora University, Aurora, IL.

Produced on: August 7, 1996