Performance of the H[sup −] Ion Source Supporting 1-MW Beam Operations at SNS

Abstract

Sealed operation of a rf driven ion source for a compact neutron generator to be used for associated particle imaginga) Rev. Sci. Instrum. 81, 02B908 (2010); 10.1063/1.3266114 Development of a 27.12 MHz radio frequency driven ion source with 3 mTorr operation pressure for neutron generators Rev. Sci. Instrum. 76, 103302 (2005); Abstract. The Spallation Neutron Source (SNS) at Oak Ridge National Laboratory reached 1-MW of beam power in September 2009, and now routinely operates near 1-MW for the production of neutrons. This paper reviews the performance, operational issues, implemented and planned mitigations of the SNS Hion source to support such high power-level beams with high availability. Some results from R&D activities are also briefly described.

Recent performance and ignition tests of the pulsed SNS H− source for 1-MW neutron production

“…Increasing the antenna loop inductance reduced, but did not eliminate the problem during the next two runs [6].…”

“…After ramping up the plasma power to produce 1-MW beams in fall 2009, plasma outages became frequent towards the end of the last 4.5-week source service cycle of the 17-week run [6]. Apparently, the extended plasma operation reduced the plasma impurities to a level which significantly increased the breakdown voltage.…”

“…Accordingly, the required ~50 mA H -beams and the ~6% duty factor in SNS are significant challenges. These are met with the LBNL designed and built H -source [2], which continues to be improved [3][4][5][6][7][8]. Briefly, the cesium-enhanced, multicusp ion source has a 2.5 turn antenna inside the plasma chamber as seen in Fig.…”

Record performance of and extraction studies with the Spallation Neutron Source H− injector

“…Replacing the original stainless steel washer with a conical Mo converter brought the beam up to ~30 mA and beyond in 2008 [5]. Several additional improvements raised the beam to the desired ~40 mA [4,6] in 2009 and in 2010. In that year a method was developed to measure the RFQ input beam current [7], with the two 2010 measurements shown in Fig.…”

“…This was a serious issue in the early years when the subtle differences between RF system failures and antenna failures led to prolonged tuning in an effort to restore the RF system. This problem was reduced by a more robust RF installation on ground potential [6] and by routine analysis of the plasma light emission [23]. Likely, hidden excess porosity in strongly exposed locations, such as the leg bends, caused the porcelain to overheat and eventually to melt after prolonged exposure.…”