On the leading edge

New cavity production process could reduce costs, provide high performance

One of the JLab-produced single-cell, single-crystal niobium cavities, of the low-loss design proposed for the International Linear Collider, awaits testing in the Vertical Test Area of the Test Lab.

The Accelerator Division's Institute for Superconducting Radiofrequency (SRF) Science & Technology is a world leader in SRF accelerator technology research and design. Now the newest idea out of the Institute may revolutionize the way accelerating cavities are produced -- making the manufacturing process faster and cheaper, while producing cavities that could potentially outperform any other niobium cavities ever tested.

Superconducting accelerator cavities, such as the ones used in JLab's accelerator to accelerate electrons, are usually made of niobium metal. Like salt crystals, niobium crystals can be grown in a variety of sizes. In accelerator scientist lingo, niobium metal composed of smaller crystals is called finegrain, while larger crystal material is called large-grain. Ordinarily, accelerator cavities are fabricated from sheets of fine-grain niobium. These sheets are forged from a large chunk of niobium called an ingot and then rolled. The rolling procedure crushes the grains so that each sheet contains grains that are more or less a uniform small size that can be easily stamped into a desired shape. The sheets are then pressed into parts, and the parts are welded together to make a complete cavity.

Gary Slack, a senior machinist in the CNC (Computer Numeric Controlled) Machine Shop, shows off the 7-cell, 12 GeV Upgrade cavity design (left) and the low-loss design proposed for the International Linear Collider. Click here for a high resolution photograph.

A JLab team came up with a way to simplify this process. Instead of working to make fine-grain niobium, the process renders pieces with readily available large-grain niobium. The team, led by Senior Staff Scientists Peter Kneisel and Ganapati Myneni, procured standard ingots of large-grain niobium through an industrial partnership with Tadeu Carneiro, spokesperson and manager of Reference Metals Co., Bridgeville, Penn. The JLab team obtained sheets of large-grain niobium by simply slicing them off an ingot using wire electrical discharge machining. "It's just like slicing up a sausage," Kneisel notes. Instead of rolling the metal to produce a fine-grain sheet, the team used a slice off of the ingot and applied the standard process of deep drawing to coax the large-grain material into parts of the desired shape. They then welded the parts together. "So the only difference between what we are doing now and typically what everybody else does, is that we have a different kind of material -- not fine-grain material, but large-grain," Kneisel adds.

The new fabrication process could simplify manufacturing while reducing cost and assembly time. For instance, eliminating the rolling procedure also eliminated related annealing steps, a softening process where the niobium is heated and then slowly cooled. "Intermediate annealing steps are necessary to remove stresses in the material and re-crystallize it. These manufacturing steps have the inherent risk of introducing unwanted impurities in the material," Kneisel explains. These impurities are removed by chemical etching, which entails using acids to strip away the surface layers of the metal. The chemical etching step was kept in the new fabrication process; early tests suggest however that much less etching is required.

A large-grain, single-cell niobium cavity in JLab’s 12 GeV Upgrade design. Click here for a high resolution photograph.

The team found they could remove two more steps from the process: an electropolishing step, where the cavity is cleaned by being suspended in acid and exposed to an electric current, and the baking procedure, where the cavities are baked for about two days in an industrial oven. In all, the new fabrication process could result in an estimated 35 percent savings in the cost of producing cavities and less time-consuming quality assurance procedures.

The Jefferson Lab team used the process to fabricate four single-cell cavities in two designs for testing. Two single-cell cavities were made with large-grain niobium in Jefferson Lab's own 12 GeV Upgrade design. The team also made use of one large grain of niobium to make two more singlecell cavities out of single niobium crystals. One of the single-crystal single- cell cavities was made in the same shape as the low-loss design proposed as an improvement to the baseline for the International Linear Collider (ILC), and the other was built in the 12 GeV Upgrade design.

Kneisel presented results of tests on the single-cell cavities in a poster session at the recent 2005 Particle Accelerator Conference in Knoxville, Tenn. The team found that all the cavities performed well in early tests; however, the cavities made with single-crystal niobium were superior performers.

A look inside two single-cell cavities shows the smoother, more reflective surface inside the cavity made with large-grain niobium (on the right) versus the cavity made with fine-grain niobium (on the left) through what has been the standard process. Click here for a high resolution photograph.

In particular, the ILC-style cavity, though a scaled-down version of what the ILC would require, reached a remarkable accelerating gradient (its ability to transfer energy into particles per unit of cavity length). In preliminary tests conducted at -271 degrees Celsius, the cavity's accelerating gradient exceeded both the ILC specification of 28 MV/m (MegaVolts per meter) and eventual goal of 35 MV/m. After adding a brief, low-temperature bake back into the fabrication process, the cavity achieved an accelerating gradient of 45 MV/m, which is roughly equal to Cornell's current world record of 46 MV/m, when measurement uncertainty and differing experimental conditions are taken into account.

The cavity in the 12 GeV Upgrade design performed even better. It reached a fundamental limit of superconducting niobium in its ability to store an accelerating field. Niobium loses its superconducting properties beyond this limit. Such field levels have never been achieved before in accelerating cavities. The team is now in the process of assembling a large-grain seven-cell cavity in the 12 GeV design, along with a fine-grain version for comparison. They expect to have the results of tests on the two complete cavities by this fall.

The team who built the cavities and conducted the tests includes Peter Kneisel, Ganapati Myneni, Gianluigi Ciovati, Jacek Sekutowicz (DESY), Larry Turlington, Robert Manus, Gary Slack, Steve Manning and Pete Kushnick. This work is being conducted in the Institute of SRF Science and Technology headed by Warren Funk.