Active Brazing Joins Titanium-Graphite Beam Target for Subatomic Particle Detection

15/08/2012

Beam target performs in harsh environment

The beam target consists of a series of graphite segments lined by tubes filled with water for cooling. The particle beams strike the segments at the start of its 450-mile  journey to create neutrinos. Beam targets, used in a harsh environment including high temperatures, vibration, and radiation, must be strong and robust, and able to withstand significant mechanical stresses. Replacing beam targets when they degrade is expensive due to the time, manpower, and equipment costs associated with the aggressive radioactive environment.

Cory Crowley, mechanical engineer at Fermilab, explains that when the experiment began in 2005, scientists thought each target would last about a 1.5 years, a goal that was met with the first series of targets. However, after a few years, targets failed at an accelerated rate, lasting only about six months. “Since it takes a long time to build the targets, and it is difficult to source all the parts for the assembly, if it fails every six months and takes a month to replace, experiment time could be severely  compromised,” said Crowley.

Fermilab started to search for new target options as targets began to run out.  Researchers also considered retrofitting older targets to ensure that a target was available to keep the experiment running. In the original beam target design, which included many transitions, stainless steel tubing was joined to graphite by soldering using electron beam welding. This fabrication method led to some quality control issues and inconsistent results.

Failures were related to weld quality, the number of transitions used on the targets, and precision of assembly. Inspection of failed  targets showed that in most cases failure was due to cracks in transition welds. This is a serious problem, because the tube is used as a water line for cooling segments. If segments overheat or do not cool properly, the entire target can degrade or melt.

Active brazing improves joint strength

Fermilab engineers created a new target design consisting of a single tube  approximately 51 in. long, which uses three 180 degree bends rather than joints or transitions. In addition, the new design replaced stainless steel tubes with titanium tubes, which added strength and corrosion resistance. The same graphite segments used in the stainless design are used inside the tubes, each about 0.8 in. long. The small segments are stacked and brazed in place about 0.010 in. apart. The segments remain in place until depleted.

The challenge was to find a method of attaching the graphite to the titanium. Because Fermilab engineers had limited experience with possible titanium to graphite joining methods, they turned to Morgan Technical Ceramics Wesgo Metals (MTC Wesgo), located in Hayward, Calif. MTC Wesgo has a long history of experience in active metal brazing, a process that allows metal to be bonded directly to nonmetallic materials, which typically require a metallization layer. Active metal brazing allows direct wetting of the alloy to the substrate material, eliminating several steps in the joining process and creating an extremely strong bond seal.

MTC Wesgo developed more than 15 braze alloy compositions, which directly bond oxide and non-oxide ceramics and synthetics to metal or other materials, including graphite, diamond, and sapphire. Alloy compositions include those designed for use in very low temperature settings to very high temperature applications in the range of 500 to 1000°C. Alloys are set to meet specific service temperature conditions, as well as the requirements of all components to be joined. Figure 1 compares traditional brazing with active brazing.

[Brazing Process Steps]

Fig. 1 — Comparison of traditional brazing and active brazing

According to the company, the active alloy process provides a more robust joint, with high bond strength. Alloys are designed to withstand thermal cycling, and will not crack, break, or undergo fatigue failure. For the MINOS experiment beam target, MTC Wesgo ran samples using a variety of alloys, after which Fermilab engineers conducted tests to review the part’s mechanical properties. 

Visually, test joints had no braze drips, and the foil adhered well to the graphite and titanium and did not expand beyond its initial boundaries. A second mechanical strength test performed in a tension testing device showed no breakage in the bonding regions (Fig.2). A sample cross section of the bonding region was prepared, polished, and mounted for metallographic examination at 1000´, which revealed a strong interaction between the braze foil and titanium. The active metal brazing alloy had an interaction one-third of the way through the tube wall, which results in the braze material creating an extremely strong metallurgical bond.

[Pull test and fixture. Courtesy of FermiLab]

Fig. 2 — Pull test and fixture. Courtesy of FermiLab

A mechanical bending test was conducted to determine whether or not the strong interaction between two materials might create a brittleness that could lead to cracking.  Test results indicated that the brazing foil adhered to the tube like tape. A final thermal cycling test was conducted to look at the possibility that the graphite might crack during rapid temperature changes. A brazed sample was taken from ice water and placed into boiling water over a few seconds, and after 30 cycles, the brazed sample showed no deterioration.

Fermilab decided to proceed with this joining approach following the test program. Engineers visited MTC Wesgo’s Hayward site to work with the company to manufacture a target core. Assembly of parts and testing of the new design was carried out in a large vacuum furnace installed in a clean room. The team came up with a successful design after several iterations, design changes, and perfection of sample geometry to ensure all dimensions would work when the assembly was placed in the furnace (Fig. 3).

Fig. 3 — Beam target assembly and manufacturing

 Graphite segments and titanium tubes were prepared at Fermi and shipped to MTC Wesgo, who brazed and shipped the assembly back for inspection and testing. It was checked out in the inspection lab, where the team ensured it met the extremely tight tolerances required, that it was flat, and that spacing between the graphite segments was correct so they would fit properly in the assembly. 

The new target is being held as the emergency spare for the experiment, because the emergency retrofitting succeeded in temporarily stemming the tide of beam target failures and scientists opted to keep the old design in place until it fails. However, the new titanium graphite design is being considered as the prototype for a new long baseline neutrino experiment (LBNE) for which Fermilab is pursuing funding from DOE. This is a higher energy experiment that will require an even stronger and more robust target.

References
1. MINOS Neutrino Experiment Launched at Fermilab, March
4, 2005; http://www.fnal.gov/pub/presspass/press_releases/
minos_dedication_3-4-05.html.