Search

Spallation Neutron Source ramps to higher power levels that can be sustained with high availability. The goal is 1.4MWdespite a compromised radio frequency quadrupole (RFQ), which requires higher radio frequency power than design levels to approach the nominal beam transmission. Unfortunately at higher power the RFQ often loses its thermal stability, a problem apparently enhanced by beam losses and high influxes of hydrogen. Delivering as much H- beam as possible with the least amount of hydrogen led to plasma outages. The root cause is the dense 1-ms long ~55-kW 2-MHz plasma pulses reflecting ~90% of the continuous ~300 W, 13-MHz power, which was mitigated with a 4-ms filter for the reflected power signal and an outage resistant, slightly detuned 13-MHz match. Lowering the H2 gas also increased the H- beam current to ~55 mA and increased the RFQ transmission by ~7% (relative). [ABSTRACT FROM AUTHOR]

The Oak Ridge National Laboratory operates the Spallation Neutron Source, consisting of a H- ion source, a 1 GeV linac and an accumulator ring. The accumulated <1 μs-long, ~35 A beam pulses are extracted from the ring at 60 Hz and directed onto a liquid Hg target. Spalled neutrons are directed to ~20 world class instruments. Currently, the facility operates routinely with ~1.2 MW of average beam power, which soon will be raised to 1.4 MW. A future upgrade with a second target station calls for raising the power to 2.8 MW. This paper describes the status of two accelerator components expected to play important roles in achieving these goals: a recently acquired RFQ accelerator and the external antenna ion source. Currently, the RFQ is being conditioned in a newly constructed 2.5 MeV Integrated Test Facility (ITF) and the external antenna source is also being tested on a separate test stand. This paper presents the results of experiments and the testing of these systems. [ABSTRACT FROM AUTHOR]

The Spallation Neutron Source H- ion source is operated with a pulsed 2-MHz RF (50-60 kW) to produce the 1-ms long, ~50 mA H- beams at 60 Hz. A continuous low power (~300 W) 13.56-MHz RF plasma, which is initially ignited with a H2 pressure bump, serves as starter plasma for the pulsed high power 2-MHz RF discharges. To reduce the risk of plasma outages at lower H2 flow rates which is desired for improved performance of the following radio frequency quadrupole, the 13.56-MHz RF matching network was characterized over a broad range of its two tuning capacitors. The H-α line intensity of the 13.56-MHz RF plasma and the reflected power of the 13.56-MHz RF were mapped against the capacitor settings. Optimal tunes for the maximum H-α intensity are consistent with the optimal tunes for minimum reflected power. Low limits of the H2 flow rate not causing plasma outages were explored within the range of the map. A tune region that allows lower H2 flow rate has been identified, which differs from the optimal tune for global minimum reflected power that was mostly used in the past. [ABSTRACT FROM AUTHOR]

Recent measurements of the H- beam current show that SNS is injecting about 55 mA into the RFQ compared to ~45 mA in 2010. Since 2010, the H- beam exiting the RFQ dropped from ~40 mA to ~34 mA, which is sufficient for 1 MW of beam power. To minimize the impact of the RFQ degradation, the service cycle of the best performing source was extended to 6 weeks. The only degradation is fluctuations in the electron dump voltage towards the end of some service cycles, a problem that is being investigated. Very recently, the RFQ was retuned, which partly restored its transmission. In addition, the electrostatic low-energy beam transport system was reengineered to double its heat sinking and equipped with a thermocouple that monitors the temperature of the ground electrode between the two Einzel lenses. The recorded data show that emissions from the source at high voltage dominate the heat load. Emissions from the partly Cs-covered first lens cause the temperature to peak several hours after starting up. On rare occasions, the temperature can also peak due to corona discharges between the center ground electrode and one of the lenses. [ABSTRACT FROM AUTHOR]

Since 2009, the Spallation Neutron Source (SNS) has been producing neutrons with ion beam powers near 1 MW, which requires the extraction of ∼50 mA H- ions from the ion source with a ∼5% duty factor. The 50 mA are achieved after an initial dose of ∼3 mg of Cs and heating the Cs collar to ∼170 °C. The 50 mA normally persist for the entire 4-week source service cycles. Fundamental processes are reviewed to elucidate the persistence of the SNS H- beams without a steady feed of Cs and why the Cs collar temperature may have to be kept near 170 °C. [ABSTRACT FROM AUTHOR]

Existing RF ion sources for accelerators have specific efficiencies for H+ and H- ion generation ~3-5 mA/cm2 kW, where about 50 kW of RF power is typically needed for 50 mA beam current production. The Saddle Antenna (SA) surface plasma source (SPS) described here was developed to improve H- ion production efficiency, reliability, and availability. In SA RF ion source, the efficiency of positive ion generation in the plasma has been improved to 200 mA/cm2 kW. After cesiation, the current of negative ions to the collector was increased from 1 mA to 10 mA with RF power ~1.5 kW in the plasma (6 mm diameter emission aperture) and up to 30 mA with ~4 kW RF. Continuous wave (CW) operation of the SA SPS has been tested on the test stand. The general design of the CW SA SPS is based on the pulsed version. Some modifications were made to improve the cooling and cesiation stability. CW operation with negative ion extraction was tested with RF power up to ~1.2 kW in the plasma with production up to Ic = 7 mA. A stable long time generation of H- beam without degradation was demonstrated in RF discharge with AlN discharge chamber. [ABSTRACT FROM AUTHOR]

Extraction of negative ions from a saddle antenna radio-frequency surface plasma source is considered. Several versions of new plasma generators with different antennas and magnetic field configurations were tested in the small Oak Ridge National Laboratory Spallation Neutron Source Test Stand. The efficiency of positive ion generation in plasma has been improved to 200 mA/cm² kW from 2.5 mA/cm² kW. A small oven was developed for cesiation by cesium compounds and alloy decomposition. After cesiation, a current of negative ions to the collector was increased from 1 mA to 10 mA with 1.5 kW RF power in the plasma and longitudinal magnetic field B1 ~ 250 G. The specific efficiency of H- production was increased to 20 mA/cm² kW from 2.5 mA/cm² kW. [ABSTRACT FROM AUTHOR]

A prototype RF H- surface plasma source (SPS) with saddle (SA) RF antenna is developed which will provide better power efficiency for high pulsed and average current, higher brightness with longer lifetime and higher reliability. Several versions of new plasma generators with small AlN discharge chambers and different antennas and magnetic field configurations were tested in the plasma source test stand. A prototype SA SPS was installed in the Spallation Neutron Source (SNS) ion source test stand with a larger, normal-sized SNS AlN chamber that achieved unanalyzed peak currents of up to 67 mA with an apparent efficiency up to 1.6 mA/kW. Control experiments with H- beam produced by SNS SPS with internal and external antennas were conducted. A new version of the RF triggering plasma gun has been designed. A saddle antenna SPS with water cooling is fabricated for high duty factor testing. [ABSTRACT FROM AUTHOR]

The LANSCE accelerator is currently powered by a filament-driven, biased converter-type H - ion source that operates at 10%, the highest plasma duty factor for this type of source, using only ∼2.2 SCCM of H 2 . The ion source needs to be replaced every 4 weeks, which takes up to 4 days. The measured negative beam current of 12-16 mA falls below the desired 24 mA acceptance of the LANCSE accelerator. The SNS (Spallation Neutron Source) RF-driven, H - ion source injects ∼50 mA of H - beam into the SNS accelerator at 60 Hz with a 6% duty factor and an availability of >99.5% but requires ∼30 SCCM of H 2 . Up to 7 A h of H - have been produced during the 14-weeks-long source service cycles, which is unprecedented for small emittance, high-current, pulsed H - ion sources. The emittance of the SNS source is slightly smaller than the emittance of the LANSCE source. The SNS source also features unrivaled low Cs consumption and can be installed and started up in <12 h. LANSCE and SNS are working toward the use of SNS H - ion sources on the LANSCE accelerator because they could (a) fill the LANSCE accelerator to its capacity, (b) decrease the source replacement time by a factor of up to 7, and (c) increase source lifetime by a factor of about 4. This paper discusses some of the challenges that emerge when trying to match a different H - source into an existing injector with significantly different characteristics and operating regimes.

The Spallation Neutron Source H(-) ion source is operated with a pulsed 2-MHz RF (50-60 kW) to produce the 1-ms long, ∼50 mA H(-) beams at 60 Hz. A continuous low power (∼300 W) 13.56-MHz RF plasma, which is initially ignited with a H2 pressure bump, serves as starter plasma for the pulsed high power 2-MHz RF discharges. To reduce the risk of plasma outages at lower H2 flow rates which is desired for improved performance of the following radio frequency quadrupole, the 13.56-MHz RF matching network was characterized over a broad range of its two tuning capacitors. The H-α line intensity of the 13.56-MHz RF plasma and the reflected power of the 13.56-MHz RF were mapped against the capacitor settings. Optimal tunes for the maximum H-α intensity are consistent with the optimal tunes for minimum reflected power. Low limits of the H2 flow rate not causing plasma outages were explored within the range of the map. A tune region that allows lower H2 flow rate has been identified, which differs from the optimal tune for global minimum reflected power that was mostly used in the past.

Spallation Neutron Source ramps to higher power levels that can be sustained with high availability. The goal is 1.4 MW despite a compromised radio frequency quadrupole (RFQ), which requires higher radio frequency power than design levels to approach the nominal beam transmission. Unfortunately at higher power the RFQ often loses its thermal stability, a problem apparently enhanced by beam losses and high influxes of hydrogen. Delivering as much H(-) beam as possible with the least amount of hydrogen led to plasma outages. The root cause is the dense 1-ms long ∼55-kW 2-MHz plasma pulses reflecting ∼90% of the continuous ∼300 W, 13-MHz power, which was mitigated with a 4-ms filter for the reflected power signal and an outage resistant, slightly detuned 13-MHz match. Lowering the H2 gas also increased the H(-) beam current to ∼55 mA and increased the RFQ transmission by ∼7% (relative).

Recent measurements of the H(-) beam current show that SNS is injecting about 55 mA into the RFQ compared to ∼45 mA in 2010. Since 2010, the H(-) beam exiting the RFQ dropped from ∼40 mA to ∼34 mA, which is sufficient for 1 MW of beam power. To minimize the impact of the RFQ degradation, the service cycle of the best performing source was extended to 6 weeks. The only degradation is fluctuations in the electron dump voltage towards the end of some service cycles, a problem that is being investigated. Very recently, the RFQ was retuned, which partly restored its transmission. In addition, the electrostatic low-energy beam transport system was reengineered to double its heat sinking and equipped with a thermocouple that monitors the temperature of the ground electrode between the two Einzel lenses. The recorded data show that emissions from the source at high voltage dominate the heat load. Emissions from the partly Cs-covered first lens cause the temperature to peak several hours after starting up. On rare occasions, the temperature can also peak due to corona discharges between the center ground electrode and one of the lenses.