Improvements to the internal and external antenna H− ion sources at the Spallation Neutron Sourcea) R. F. Welton, V. G. Dudnikov, B. X. Han, S. N. Murray, T. R. Pennisi, C. Pillar, M. Santana, M. P. Stockli, and M. W. Turvey Citation: Review of Scientific Instruments 85, 02B135 (2014); doi: 10.1063/1.4858177 View online: http://dx.doi.org/10.1063/1.4858177 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/2?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Modeling of neutrals in the Linac4 H− ion source plasma: Hydrogen atom production density profile and Hα intensity by collisional radiative modela) Rev. Sci. Instrum. 85, 02B118 (2014); 10.1063/1.4833016 A new extraction system for the Linac4 H−ion sourcea) Rev. Sci. Instrum. 83, 02B710 (2012); 10.1063/1.3670344 The continued development of the Spallation Neutron Source external antenna H − ion sourcea) Rev. Sci. Instrum. 81, 02A727 (2010); 10.1063/1.3301601 General design of the International Fusion Materials Irradiation Facility deuteron injector: Source and beam linea) Rev. Sci. Instrum. 81, 02B301 (2010); 10.1063/1.3257998 Beam measurements on the H − source and low energy beam transport system for the Spallation Neutron Source (abstract) Rev. Sci. Instrum. 73, 1020 (2002); 10.1063/1.1433170
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 02B135 (2014)
Improvements to the internal and external antenna H− ion sources at the Spallation Neutron Sourcea) R. F. Welton,1,b) V. G. Dudnikov,2 B. X. Han,1 S. N. Murray,1 T. R. Pennisi,1 C. Pillar,1 M. Santana,1 M. P. Stockli,1 and M. W. Turvey3 1
Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee 37830-6471, USA Muons, Inc., 552 N. Batavia Avenue, Batavia, Illinois 60510, USA 3 Villanova University, 800E. Lancaster Ave, Villanova, Pennsylvania 19085, USA 2
(Presented 10 September 2013; received 27 September 2013; accepted 10 December 2013; published online 30 January 2014) The Spallation Neutron Source (SNS), a large scale neutron production facility, routinely operates with 30–40 mA peak current in the linac. Recent measurements have shown that our RF-driven internal antenna, Cs-enhanced, multi-cusp ion sources injects ∼55 mA of H− beam current (∼1 ms, 60 Hz) at 65-kV into a Radio Frequency Quadrupole (RFQ) accelerator through a closely coupled electrostatic Low-Energy Beam Transport system. Over the last several years a decrease in RFQ transmission and issues with internal antennas has stimulated source development at the SNS both for the internal and external antenna ion sources. This report discusses progress in improving internal antenna reliability, H− yield improvements which resulted from modifications to the outlet aperture assembly (applicable to both internal and external antenna sources) and studies made of the long standing problem of beam persistence with the external antenna source. The current status of the external antenna ion source will also be presented. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4858177] I. INTRODUCTION
The pulsed U.S. Spallation Neutron Source (SNS) is operating near ∼1 MW, which requires a beam of ∼38 mA of H− delivered to the LINAC. The recent milestone of delivering 1.2 MW of beam power to target has been achieved and plans are in place to reach 1.4 MW and later, possibly to 2–3 MW by increasing the beam energy from 1 to 1.3 GeV and the LINAC beam current to ∼60 mA. Presently, the SNS uses an RF-driven, Cs-enhanced, multicusp, baseline ion source originally designed at Lawrence Berkeley National Laboratory (LBNL) which features an internal porcelain enamel coated antenna1 as shown in Fig. 1. Plasma ignition is achieved by continuous application of ∼300 W of 13.56 MHz RF and H− production is achieved by the application of ∼55–65 kW of 2 MHz power pulsed at ∼1 ms, 60 Hz, to the internal antenna. The source is conditioned prior to cesiation for 3 h at 50 kW, 60 Hz, ∼0.88 ms. Cesiation is achieved by raising the temperature of the Cs collar containing a charge of ∼30 mg of Cs in the form of Cs2 CrO4 to 550 ◦ C for 12 min then reducing to nominal 170 ◦ C. Beam from the source is focused into a Radio Frequency Quadrupole (RFQ) accelerator by an electrostatic Low Energy Beam Transport (LEBT).2 Historically, ion source antenna failures have impacted SNS operation both directly by interrupting neutron production or indirectly by interrupting ion source testing efforts either on the Front End (FE) of the accelerator or on the ion source test stand. Recovering from such a failure typically requires about 8 hours for installing, conditioning and tuning a) Contributed paper, published as part of the Proceedings of the 15th
International Conference on Ion Sources, Chiba, Japan, September 2013. b) Electronic mail: [email protected]
0034-6748/2014/85(2)/02B135/3/$30.00
up a spare source. The failed source is then refurbished over about 3 man-days. This problem became particularly apparent in 2011 when a large number of antenna failures disrupted operations and we adopted a detailed antenna inspection, sorting and installation strategy which has largely mitigated the problem. The strategy has been described in detail earlier3, 4 and will be summarized here. Specifically, this report provides updated statistics of antennas used at the SNS since 2011 in order to better understand the effectiveness of the approach and provides guidance for other facilities using RF sources with internal antennas.4, 5 In 2012, in order to compensate for the RFQ accelerator’s reduced transmission2 the test stand was utilized to study performance variations of our set of baseline, internal antenna ion sources.3 A key outcome of this study was that it was found that increasing the aperture diameter of the Mo ionization surface and roughening the surface (see Fig. 1) also increased the H− output from the source. These findings are discussed in this report. In early 2013, after completing this study we resumed our external antenna testing campaign. This report discusses the results of six recent, 1–2 week runs on the test stand. Measured beam current, power efficiency, and beam persistence from the current configuration of the source are summarized.
II. INTERNAL ANTENNA RELIABILITY
After a rash of 12 antenna failures on the SNS FE in the first part of 2011 we imposed a strict process for selecting antennas which are installed in ion sources used on the SNS.3 Antennas were selected from a large inventory of porcelain coated Cu antennas produced by Cherokee Porcelain Enamel
85, 02B135-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
02B135-2
Welton et al.
Rev. Sci. Instrum. 85, 02B135 (2014)
FIG. 1. Schematic cross-section of the SNS internal and external antenna ion source. AlN plasma chamber shown as transparent.
Corporation.6 The procedure is as follows:
r All selected antennas must initially pass a salt water high pot test to 2 kV.
r Antenna coating thickness must be strictly within 0.5 < t < 0.7 mm and the smaller radius leg bend areas should be completely free of any tactilely perceptible surface defects registered using a dental pick. These are known as Tier II antennas. r The entire surface of a Tier II antenna (which will be exposed to plasma) is inspected and then sorted by the number of defects noted using the dental pick. The best 10% of these are cosmetically in near perfect condition and are called Tier I antennas. r Only Tier I antennas are used for neutron production at the SNS. r Tier II antennas are used on the test stand and during FE testing during maintenance periods. In May of 2011, the 2 MHz pulse length used for source conditioning was reduced from 1.2 ms (RF duty factor: 7.2%) to 1.0 ms (6%). In the next several months the SNS experienced 6 antenna failures, in part driven by aggressive testing. In August 2011, the conditioning pulse length was lowered to 0.88 ms (5.3%) and there was another failure within 2 months. In October of 2011, the above listed, refined selection criteria were implemented. Only 1 Tier I antenna has failed in the past 2 years, a result that is skewed by the fact that the 2 MHz power was lowered as the SNS power was lowered to 850 kW during about half of that time. Figure 2(a) shows the number of antenna failures on the SNS accelerator per year during neutron production and failures including all time periods (neutron production + maintenance periods) before and after the implementation of the above selection process in CY 11. Figure 2(b) shows the percentage of antenna found damaged after use on the SNS accelerator per year. No obvious correlation was found between antenna coating thickness and those which failed. This difference in the rate of antenna failure and damage after the implementation of the above selection criteria suggests that the majority of failures experienced by these antennas at the SNS and elsewhere over the last decade originated from mechanical surface de-
fects present prior to operation. Other contributing factors are likely gradual improvements in the manufacture process. Currently about 1 in 10 antennas received qualify for a Tier I status making the effective cost of each antenna ∼$5 k USD. However, these costs can be reduced by increasing the fraction of antennas accepted as Tier I, because a reduction in the failure rate of Tier II antenna has also been noted. III. IMPROVING IONIZATION EFFICIENCY
A study of baseline source variations conducted on the test stand in 2012 showed that increasing the inside diameter of the Mo ionization surface (see Fig. 1) from 7 to 8.4 mm and slightly roughening the surface yields more H− beam for a fixed RF power. This effect was explored on the test stand
FIG. 2. (a) Number of antenna failures on the SNS accelerator per year during neutron production and failures including all time periods. (b) Percentage of antenna found damaged after use on the SNS accelerator per year. Arrow indicates implementation of the selection criteria.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
02B135-3
Welton et al.
Rev. Sci. Instrum. 85, 02B135 (2014) TABLE I. Summary of the 2013 external antenna ion source experiments performed on the test stand. 2013 Run #
FIG. 3. Stable 12 day run (2013 run #5) on the test stand, showing archived H− beam current in mA and applied RF power.
in baseline source #5 and #6 for beam currents which could be sustained for days: beam current was increased in #5 from 38 mA to 45.5 mA and in source #6 from 39 mA to 46 mA for equal RF power ∼65 kW.3 This larger aperture ionization surface has now been installed in nearly all our baseline sources as well as our external antenna ion sources. Electron emission from the source was not changed significantly. IV. EXTERNAL ANTENNA SOURCE PROGRESS
The uncertainty and risk associated with increasing the beam current for future facility upgrades and having only a single source internal antenna manufacturer continues to motivate the development of the SNS external antenna ion source. After recently restarting the external antenna testing effort in 2013, six 1–2 week tests have been performed to date (labeled run-1 through run-6) on the test stand with focus on achieving and demonstrating consistent performance through minor changes to the source configuration, preparation, and operating procedures. Runs 2 and 3 will not be discussed due to assembly errors. The basic source configuration is shown in Fig. 1 and is described in detail in Ref. 2 along with the results of earlier source testing. A single source was used for the present set of experiments configured as described earlier specifically with the options of the RF plasma gun (200 V, 200 W); 400 G air-cooled multicusp magnet array; Mo ionization surface with φ ∼ 8 mm opening (secured with 2–4 bolts to the Cs collar). In runs 4 and 5 an auxiliary Cs-Bi dispenser containing 20 mg of Cs sealed with an indium plug was added inside the view port tube shown in Fig. 1 which could be ohmically heated by an external power supply. Although in both of these runs this dispenser was not intentionally activated, in run-4 the indium seal, which was located closer to the plasma then in run-5, was found melted (T > 150 ◦ C). In run-5 the seal was found intact (T < 150 ◦ C), as the dispenser had been placed further from the plasma wall. In both cases it was unlikely Cs from this dispenser was released into the plasma due to Cs-Bi activation temperatures T > 500 ◦ C and limited plasma heat flux. The electron dump feedthrough insulator upgrade described in Ref. 7 was added to the source for runs 5 and 6. All ceramic parts were vacuum baked at 200 ◦ C for many days prior to installation. The conditioning and cesiation procedure employed here were very similar to the baseline recipes described above.
1 4 5 6
RF power 34–50 kW ∼37 kW ∼39 kW ∼36 kW
H− (min/max)
Type efficiency
AVG beam decay
Duration (days)
35–36 mA ∼1 mA/kW ∼2 mA/day ∼9 days 49–55 mA ∼1.4 mA/kW ∼0.6 mA/day ∼9 days 47–49 mA ∼1.2 mA/kW ∼0.2 mA/day ∼12 days 50–56 mA ∼1.5 mA/kW ∼1 mA/day ∼6 days
Figure 3 shows an example plot of the archived H− beam current and RF power data taken during run #5 and Table I summarizes all the 2013 external antenna tests to date. These data show that the RF power efficiency is 1–1.5 mA/kW, better than baseline sources, but inconsistent beam persistence remains problematic with H− beam decay ranging from 0 to 8 mA/day for fixed RF power. It should be noted that the Cs-Bi injector was present for both runs which showed beam growth over the course of days as opposed to the more typical beam decay. In 2013, run 5 was the only run where full electron dump voltage (∼6 kV) could be maintained throughout the run. Voltage decline was likely due to beam loading of the electron dump electrode (see Fig. 1). Currently, LEBT heating is also problematic which causes discharges in the LEBT especially after cesiation. We hope the LEBT heating issues can be resolved by the use of a water cooled extract electrode. We are also hoping to address the electron dumping voltage issue electrically from the power supply side. Once these issues are addressed and consistent performance is achieved we plan to perform longer tests on the test stand to determine long-term reliability of the source and eventually deploy it on the SNS accelerator.
ACKNOWLEDGMENTS
This work was supported by SNS through UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 for the (U.S.) Department of Energy. 1 R.
F. Welton, S. N. Murray, M. A. Janney, and M. P. Stockli, Rev. Sci. Instrum. 73, 1008 (2002). 2 M. P. Stockli, B. X. Han, S. N. Murray, T. R. Pennisi, M. Santana, and R. Welton, “Recent performance of the SNS H− ion source and low-energy beam transport system,” Rev. Sci. Instrum. (these proceedings). 3 R. F. Welton, V. G. Dudnikov, K. R. Gawne, B. X. Han, S. N. Murray, T. R. Pennisi, R. T. Roseberry, M. Santana, M. P. Stockli, and M. W. Turvey, “H− radio frequency source development at the Spallation Neutron Source,” Rev. Sci. Instrum. 83, 02A725 (2012). 4 R. F. Welton, V. G. Dudnikov, B. X. Han, S. N. Murray, T. R. Pennisi, R. T. Roseberry, M. Santana, and M. P. Stockli, AIP Conf. Proc. 1515, 341–348 (2013). 5 A. Ueno, Y. Namekawa, S. Yamazaki, I. Koizumi, K. Ikegami, A. Takagi, and H. Oguri, AIP Conf. Proc. 1515, 331–340 (2013). 6 Cherokee Porcelain Enamel Corporation, 2717 Independence Lane, Knoxville, TN 37914, USA. 7 R. F. Welton, N. J. Desai, B. X. Han, E. A. Kenic, S. N. Murray, T. R. Pennisi, K. G. Potter, B. R. Lang, M. Santana, and M. P. Stockli, AIP Conf. Proc. 1390, 226–234 (2011).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 02B135 (2014)
Improvements to the internal and external antenna H− ion sources at the Spallation Neutron Sourcea) R. F. Welton,1,b) V. G. Dudnikov,2 B. X. Han,1 S. N. Murray,1 T. R. Pennisi,1 C. Pillar,1 M. Santana,1 M. P. Stockli,1 and M. W. Turvey3 1
Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee 37830-6471, USA Muons, Inc., 552 N. Batavia Avenue, Batavia, Illinois 60510, USA 3 Villanova University, 800E. Lancaster Ave, Villanova, Pennsylvania 19085, USA 2
(Presented 10 September 2013; received 27 September 2013; accepted 10 December 2013; published online 30 January 2014) The Spallation Neutron Source (SNS), a large scale neutron production facility, routinely operates with 30–40 mA peak current in the linac. Recent measurements have shown that our RF-driven internal antenna, Cs-enhanced, multi-cusp ion sources injects ∼55 mA of H− beam current (∼1 ms, 60 Hz) at 65-kV into a Radio Frequency Quadrupole (RFQ) accelerator through a closely coupled electrostatic Low-Energy Beam Transport system. Over the last several years a decrease in RFQ transmission and issues with internal antennas has stimulated source development at the SNS both for the internal and external antenna ion sources. This report discusses progress in improving internal antenna reliability, H− yield improvements which resulted from modifications to the outlet aperture assembly (applicable to both internal and external antenna sources) and studies made of the long standing problem of beam persistence with the external antenna source. The current status of the external antenna ion source will also be presented. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4858177] I. INTRODUCTION
The pulsed U.S. Spallation Neutron Source (SNS) is operating near ∼1 MW, which requires a beam of ∼38 mA of H− delivered to the LINAC. The recent milestone of delivering 1.2 MW of beam power to target has been achieved and plans are in place to reach 1.4 MW and later, possibly to 2–3 MW by increasing the beam energy from 1 to 1.3 GeV and the LINAC beam current to ∼60 mA. Presently, the SNS uses an RF-driven, Cs-enhanced, multicusp, baseline ion source originally designed at Lawrence Berkeley National Laboratory (LBNL) which features an internal porcelain enamel coated antenna1 as shown in Fig. 1. Plasma ignition is achieved by continuous application of ∼300 W of 13.56 MHz RF and H− production is achieved by the application of ∼55–65 kW of 2 MHz power pulsed at ∼1 ms, 60 Hz, to the internal antenna. The source is conditioned prior to cesiation for 3 h at 50 kW, 60 Hz, ∼0.88 ms. Cesiation is achieved by raising the temperature of the Cs collar containing a charge of ∼30 mg of Cs in the form of Cs2 CrO4 to 550 ◦ C for 12 min then reducing to nominal 170 ◦ C. Beam from the source is focused into a Radio Frequency Quadrupole (RFQ) accelerator by an electrostatic Low Energy Beam Transport (LEBT).2 Historically, ion source antenna failures have impacted SNS operation both directly by interrupting neutron production or indirectly by interrupting ion source testing efforts either on the Front End (FE) of the accelerator or on the ion source test stand. Recovering from such a failure typically requires about 8 hours for installing, conditioning and tuning a) Contributed paper, published as part of the Proceedings of the 15th
International Conference on Ion Sources, Chiba, Japan, September 2013. b) Electronic mail: [email protected]
0034-6748/2014/85(2)/02B135/3/$30.00
up a spare source. The failed source is then refurbished over about 3 man-days. This problem became particularly apparent in 2011 when a large number of antenna failures disrupted operations and we adopted a detailed antenna inspection, sorting and installation strategy which has largely mitigated the problem. The strategy has been described in detail earlier3, 4 and will be summarized here. Specifically, this report provides updated statistics of antennas used at the SNS since 2011 in order to better understand the effectiveness of the approach and provides guidance for other facilities using RF sources with internal antennas.4, 5 In 2012, in order to compensate for the RFQ accelerator’s reduced transmission2 the test stand was utilized to study performance variations of our set of baseline, internal antenna ion sources.3 A key outcome of this study was that it was found that increasing the aperture diameter of the Mo ionization surface and roughening the surface (see Fig. 1) also increased the H− output from the source. These findings are discussed in this report. In early 2013, after completing this study we resumed our external antenna testing campaign. This report discusses the results of six recent, 1–2 week runs on the test stand. Measured beam current, power efficiency, and beam persistence from the current configuration of the source are summarized.
II. INTERNAL ANTENNA RELIABILITY
After a rash of 12 antenna failures on the SNS FE in the first part of 2011 we imposed a strict process for selecting antennas which are installed in ion sources used on the SNS.3 Antennas were selected from a large inventory of porcelain coated Cu antennas produced by Cherokee Porcelain Enamel
85, 02B135-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
02B135-2
Welton et al.
Rev. Sci. Instrum. 85, 02B135 (2014)
FIG. 1. Schematic cross-section of the SNS internal and external antenna ion source. AlN plasma chamber shown as transparent.
Corporation.6 The procedure is as follows:
r All selected antennas must initially pass a salt water high pot test to 2 kV.
r Antenna coating thickness must be strictly within 0.5 < t < 0.7 mm and the smaller radius leg bend areas should be completely free of any tactilely perceptible surface defects registered using a dental pick. These are known as Tier II antennas. r The entire surface of a Tier II antenna (which will be exposed to plasma) is inspected and then sorted by the number of defects noted using the dental pick. The best 10% of these are cosmetically in near perfect condition and are called Tier I antennas. r Only Tier I antennas are used for neutron production at the SNS. r Tier II antennas are used on the test stand and during FE testing during maintenance periods. In May of 2011, the 2 MHz pulse length used for source conditioning was reduced from 1.2 ms (RF duty factor: 7.2%) to 1.0 ms (6%). In the next several months the SNS experienced 6 antenna failures, in part driven by aggressive testing. In August 2011, the conditioning pulse length was lowered to 0.88 ms (5.3%) and there was another failure within 2 months. In October of 2011, the above listed, refined selection criteria were implemented. Only 1 Tier I antenna has failed in the past 2 years, a result that is skewed by the fact that the 2 MHz power was lowered as the SNS power was lowered to 850 kW during about half of that time. Figure 2(a) shows the number of antenna failures on the SNS accelerator per year during neutron production and failures including all time periods (neutron production + maintenance periods) before and after the implementation of the above selection process in CY 11. Figure 2(b) shows the percentage of antenna found damaged after use on the SNS accelerator per year. No obvious correlation was found between antenna coating thickness and those which failed. This difference in the rate of antenna failure and damage after the implementation of the above selection criteria suggests that the majority of failures experienced by these antennas at the SNS and elsewhere over the last decade originated from mechanical surface de-
fects present prior to operation. Other contributing factors are likely gradual improvements in the manufacture process. Currently about 1 in 10 antennas received qualify for a Tier I status making the effective cost of each antenna ∼$5 k USD. However, these costs can be reduced by increasing the fraction of antennas accepted as Tier I, because a reduction in the failure rate of Tier II antenna has also been noted. III. IMPROVING IONIZATION EFFICIENCY
A study of baseline source variations conducted on the test stand in 2012 showed that increasing the inside diameter of the Mo ionization surface (see Fig. 1) from 7 to 8.4 mm and slightly roughening the surface yields more H− beam for a fixed RF power. This effect was explored on the test stand
FIG. 2. (a) Number of antenna failures on the SNS accelerator per year during neutron production and failures including all time periods. (b) Percentage of antenna found damaged after use on the SNS accelerator per year. Arrow indicates implementation of the selection criteria.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
02B135-3
Welton et al.
Rev. Sci. Instrum. 85, 02B135 (2014) TABLE I. Summary of the 2013 external antenna ion source experiments performed on the test stand. 2013 Run #
FIG. 3. Stable 12 day run (2013 run #5) on the test stand, showing archived H− beam current in mA and applied RF power.
in baseline source #5 and #6 for beam currents which could be sustained for days: beam current was increased in #5 from 38 mA to 45.5 mA and in source #6 from 39 mA to 46 mA for equal RF power ∼65 kW.3 This larger aperture ionization surface has now been installed in nearly all our baseline sources as well as our external antenna ion sources. Electron emission from the source was not changed significantly. IV. EXTERNAL ANTENNA SOURCE PROGRESS
The uncertainty and risk associated with increasing the beam current for future facility upgrades and having only a single source internal antenna manufacturer continues to motivate the development of the SNS external antenna ion source. After recently restarting the external antenna testing effort in 2013, six 1–2 week tests have been performed to date (labeled run-1 through run-6) on the test stand with focus on achieving and demonstrating consistent performance through minor changes to the source configuration, preparation, and operating procedures. Runs 2 and 3 will not be discussed due to assembly errors. The basic source configuration is shown in Fig. 1 and is described in detail in Ref. 2 along with the results of earlier source testing. A single source was used for the present set of experiments configured as described earlier specifically with the options of the RF plasma gun (200 V, 200 W); 400 G air-cooled multicusp magnet array; Mo ionization surface with φ ∼ 8 mm opening (secured with 2–4 bolts to the Cs collar). In runs 4 and 5 an auxiliary Cs-Bi dispenser containing 20 mg of Cs sealed with an indium plug was added inside the view port tube shown in Fig. 1 which could be ohmically heated by an external power supply. Although in both of these runs this dispenser was not intentionally activated, in run-4 the indium seal, which was located closer to the plasma then in run-5, was found melted (T > 150 ◦ C). In run-5 the seal was found intact (T < 150 ◦ C), as the dispenser had been placed further from the plasma wall. In both cases it was unlikely Cs from this dispenser was released into the plasma due to Cs-Bi activation temperatures T > 500 ◦ C and limited plasma heat flux. The electron dump feedthrough insulator upgrade described in Ref. 7 was added to the source for runs 5 and 6. All ceramic parts were vacuum baked at 200 ◦ C for many days prior to installation. The conditioning and cesiation procedure employed here were very similar to the baseline recipes described above.
1 4 5 6
RF power 34–50 kW ∼37 kW ∼39 kW ∼36 kW
H− (min/max)
Type efficiency
AVG beam decay
Duration (days)
35–36 mA ∼1 mA/kW ∼2 mA/day ∼9 days 49–55 mA ∼1.4 mA/kW ∼0.6 mA/day ∼9 days 47–49 mA ∼1.2 mA/kW ∼0.2 mA/day ∼12 days 50–56 mA ∼1.5 mA/kW ∼1 mA/day ∼6 days
Figure 3 shows an example plot of the archived H− beam current and RF power data taken during run #5 and Table I summarizes all the 2013 external antenna tests to date. These data show that the RF power efficiency is 1–1.5 mA/kW, better than baseline sources, but inconsistent beam persistence remains problematic with H− beam decay ranging from 0 to 8 mA/day for fixed RF power. It should be noted that the Cs-Bi injector was present for both runs which showed beam growth over the course of days as opposed to the more typical beam decay. In 2013, run 5 was the only run where full electron dump voltage (∼6 kV) could be maintained throughout the run. Voltage decline was likely due to beam loading of the electron dump electrode (see Fig. 1). Currently, LEBT heating is also problematic which causes discharges in the LEBT especially after cesiation. We hope the LEBT heating issues can be resolved by the use of a water cooled extract electrode. We are also hoping to address the electron dumping voltage issue electrically from the power supply side. Once these issues are addressed and consistent performance is achieved we plan to perform longer tests on the test stand to determine long-term reliability of the source and eventually deploy it on the SNS accelerator.
ACKNOWLEDGMENTS
This work was supported by SNS through UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 for the (U.S.) Department of Energy. 1 R.
F. Welton, S. N. Murray, M. A. Janney, and M. P. Stockli, Rev. Sci. Instrum. 73, 1008 (2002). 2 M. P. Stockli, B. X. Han, S. N. Murray, T. R. Pennisi, M. Santana, and R. Welton, “Recent performance of the SNS H− ion source and low-energy beam transport system,” Rev. Sci. Instrum. (these proceedings). 3 R. F. Welton, V. G. Dudnikov, K. R. Gawne, B. X. Han, S. N. Murray, T. R. Pennisi, R. T. Roseberry, M. Santana, M. P. Stockli, and M. W. Turvey, “H− radio frequency source development at the Spallation Neutron Source,” Rev. Sci. Instrum. 83, 02A725 (2012). 4 R. F. Welton, V. G. Dudnikov, B. X. Han, S. N. Murray, T. R. Pennisi, R. T. Roseberry, M. Santana, and M. P. Stockli, AIP Conf. Proc. 1515, 341–348 (2013). 5 A. Ueno, Y. Namekawa, S. Yamazaki, I. Koizumi, K. Ikegami, A. Takagi, and H. Oguri, AIP Conf. Proc. 1515, 331–340 (2013). 6 Cherokee Porcelain Enamel Corporation, 2717 Independence Lane, Knoxville, TN 37914, USA. 7 R. F. Welton, N. J. Desai, B. X. Han, E. A. Kenic, S. N. Murray, T. R. Pennisi, K. G. Potter, B. R. Lang, M. Santana, and M. P. Stockli, AIP Conf. Proc. 1390, 226–234 (2011).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 131.94.16.10 On: Tue, 23 Dec 2014 21:33:23
