HIGH POWER BEAM DUMP AND TARGET / ACCELERATOR INTERFACE PROCEDURES * J. Galambos, W. Blokland, D. Brown, C. Peters, M. Plum, Spallation Neutron Source, ORNL, Oak Ridge, TN 37831, U.S.A. Abstract Satisfying operational procedures and limits for the beam target interface is a critical concern for high power operation at spallation neutron sources. At the Oak Ridge Spallation Neutron Source (SNS) a number of protective measures are instituted to ensure that the beam position, beam size and peak intensity are within acceptable limits at the target and high power Ring Injection Dump (RID). The high power beam dump typically handles up to 50-100 kw of beam power and its setup is complicated by the fact that there are two separate beam components simultaneously directed to the dump. The beam on target is typically in the 800-1000 kw average power level, delivered in sub- s 60 Hz pulses. Setup techniques using beam measurements to quantify the beam parameters at the target and dump will be described. However, not all the instrumentation used for the setup and initial qualification is available during high power operation. Additional techniques are used to monitor the beam during high power operation to ensure the setup conditions are maintained, and these are also described. INTRODUCTION Design, operation and maintenance of high power hadron beam targets and dumps present significant challenges [1]. As the beam power approaches the mega- Watt range, the target issues become especially severe. Typically, a high intensity secondary beam is desired favoring a small primary beam footprint, but heat removal constraints limit the minimum size of the beam. Accelerator designers prefer a peaked beam profile on target to minimize the amount of beam near the beam pipe, whereas target designers prefer as uniform a power distribution on target as possible. Pulsed beams introduce transient effects such as fatigue and cavitation, further complicating target survivability. Ultimately compromises in these various considerations are made in the design solution based on assumptions on the beam properties impingent on the target. Ensuring that the beam characteristics are within design limits is critical to ensure target integrity, and this task itself represents challenges. It is especially difficult given the harsh environment associated with high power targets and dumps. In this paper we describe the beam property requirements at the high power target and beam dump, the beam setup techniques used to qualify high power operation, and the live monitoring systems used to ensure that the beam setup is maintained during high power operations at the SNS [2,3]. We will also review the beam interface experience for the SNS Ring Injection Dump (RID), which receives up to 100 kw during operation, and with the SNS primary target. RING INJECTION DUMP Between 5-10% of the accelerated H - beam is not fully stripped in the Ring charge exchange injection process, and this waste beam is transported to the Ring Injection Dump (RID). For a 1 MW power delivered to the target, this represents up to ~100 kw delivered to this beam dump, which represents a significant challenge. Further complicating this system is the fact that there are two species of waste beam delivered to the RID: 1) beam that misses the charge exchange stripper foil (H - ) and 2) beam that hits the foil, but is only partially stripped (H 0 ). This injection process is described in Ref. [4]. The final approach to the RID is shown schematically in Fig.1, and includes 1) a region of ~9 m including beam instrumentation that is inside the tunnel (accessible), and 2) a long drift (~10.5 m) in a buried beam pipe, and 3) a vacuum window and target assembly that are accessible from above in a dedicated service building. Beam Instrumentation The beam instrumentation in the RID beam-line used for target interface related measurements includes: three Beam Position Monitors (BPMs), a wire scanner profile measurement, and two current measurement devices. The only beam instrumentation at the window/target area is a set of indirect halo temperature thermocouples mounted around the outer perimeter of the vacuum window assembly. There are no direct beam measurements at the window or dump assembly. Beam Constraints and Production Setup Primary concerns in the beam setup are preventing over-focusing of the waste beams on the vacuum window and dump, and ensuring that the waste beams are properly centered at the window and target. There is a long drift (~20 m) from the last focusing quadrupole to the beam dump, which provides a natural expansion of the waste beam sizes at the dump, which are estimated to be ~ 35 mm RMS. The final focusing quadrupole strength in this beam-line is limited, to prevent over-focusing of the beam sizes at the dump window. The waste beam sizes at the dump are not measured. * ORNL is managed by UT-Battelle, LLC, under contract DE-AC05-00OR22725 for the U.S. Dept. of Energy.
The primary effort during the production beam qualification is ensuring that both waste beams intersect the target and vacuum window near the center. This beam centering is done by extrapolating the beam position measured at the three BPMs in the dump line (see Fig. 1) to the window using a simple linear transport model. The extrapolation is straightforward for the H - waste beam, since the centering is performed with the primary stripper foil extracted, guaranteeing a 100% H - beam. For the H 0 waste beam, we simulate the H 0 trajectory by appropriate adjustment of the intervening dipole magnet fields, before it is completely stripped by a large/thick secondary foil. There is some uncertainty associated with this part of the setup, estimated to translate to a few mm in the beam position at the window. If the measured beam positions at the window are > 30 mm from the center, adjustments are made with steering elements, until this criterion is satisfied. Tunnel end ~ 10 m dri Dump Quad BPMs, Wire scanner, Current monitor Figure 1: Schematic layout of the Ring Injection Dump beam-line geometry. Several modifications to the injection dump have improved this setup process. A horizontal bending C- magnet that primarily affects the H - waste beam (the magnet is located at a point of maximal waste beam separation) was added, to provide an independent horizontal steering control. Initially, the beam position measurements in the dump line involved use of a wire scanner profile measurement. This was quite slow, and resulted in the waste beam setup being quite tedious. Two BPMs were added to bring the total to three in the drift region, and now the wire scanner is no longer used for beam centering. A real time phosphor screen imaging system for the dump vacuum window providing direct position and size measurement is under consideration, but this effort has not started. Post Power Ramp-up The above described setup process is typically done with a low intensity (few s) 1 Hz tuning beam. Not until the beam power is increased, do the dump window thermocouples begin to show a response to any impingent beam. The restrictions imposed on the beam centering described above do not necessarily result in a low beam loss tune in the dump line, but do ensure the waste beams start near the window center. The initial (immediately after production setup) waste beam coordinates are shown in Figure 2a, for several setup configurations in the 2010-2011 period. After the full beam power is reached, it is normal to perform tune up adjustments for reasons including beam loss reduction, or reducing foil flutter. This tuning involves steering that affects the waste beam trajectories. During tuning, the thermocouple levels at the injection dump window are restricted as indicated in Table 1. This process ensures that the waste beams start nearly centered, and any post ramp-up steering that results in the waste beams approaching the edge of the window will be limited from too close an approach by the near complete coverage of thermocouples. If either waste beam does approach the window edge, the halo thermocouple temperature response quickly rises, and appropriate limits are placed on these temperatures. Figure 2b shows the waste beam center measured positions after the end of the same production runs shown in Figure 2a. There is considerable deviation in the post production waste beam horizontal positions relative to the initial setup. This deviation is a result of the above mentioned tuning that occurs during production. Controls There are multiple layers of controls in place during operations to ensure that the design parameters of the beam dump are not violated, as summarized in Table 1. There are three levels of enforcement, should a monitored quantity exceed a limit: 1) alarm, 2) software trip, and 3) hardware trip. Alarm limits cause an audible alarm, but do not stop beam. These are useful for advance notice of a slow drift (seconds to minutes) towards an unsafe condition, allowing for operator correction. Alarm limits are typically implemented with a larger margin than any trip limit. Software trips inhibit the beam by stopping the beam permit in the timing system, but typically do not turn off equipment to ensure that the beam is off. These trip limits have an intermediate margin from the operational limit, and are usually programmed to run on a control system input-output computer (IOC), with
reaction times of ms. Finally the most severe and most reliable protective measure implements a hardware trip, which is typically enforced by a PLC circuit, sometimes accompanied by forced shutdown of equipment required to accelerate beam. a) b) Ini al Setup 80 H 60 H0 40 20 0 20 40 60 80 80 60 40 20 0 20 40 60 80 X (mm) Y (mm) Post Produc on 80 H 60 H0 40 20 0 20 40 60 80 80 60 40 20 0 20 40 60 80 X (mm) Figure 2: RID Waste beam centroids, a) before production start, and b) and after the end of a 1 month production run. Y (mm) Table 1: Summary of Injection Dump run time protective measures Limit Type Halo thermocouples < 55 C Software Halo thermocouples < 65 C Quadrupole / Dipoles < 10% from nominal Power < input limit Software + Beam current < 100 ma (instantaneous) Beam current < 15 ma (over 100 s) Beam current < 3 ma (over 1 ms) Beam current < 0.1 ma (over 1 s) Loss monitors + calculated using current monitor input BEAM ON TARGET and software Controlling the beam on target is also a critical concern [2,3]. Primary control requirements are shown in Table 2, which ensure limiting the peak intensity on the target to < 1.72x10 16 protons/m 2, limiting the peak time averaged current density (over 10 sec) < 0.17A/m 2, 90% of the power within the nominal spot size of 200 mm horizontal and 70 mm vertical, and keeping the beam centered on the target center. Beam Instrumentation Beam instrumentation in the beam-line approaching the target is discussed in [Plum 2005]. Instrumentation directly involved in the beam production setup and production monitoring includes 11 BPMs, 4 wire scanner beam profile measurement devices, one beam current monitor, 1 Harp profile monitor (9.52 m from the target), halo thermocouple instrumentation on at the beam window 2.14 m upstream from the target and a Target Imaging System (TIS) that utilizes a phosphor coating directly on the target [5,6]. Also, there are 33 BLMs in the transport line to the Target, as indicated in Figure 3. Beam Constraints and Production Setup The primary exercise in setting up the production beam on target is ensuring that the peak beam intensity, the beam size and the beam center on the target are acceptable. First, the beam size and profile on the target are determined indirectly. RMS beam sizes are measured at the 5 profile measurement locations directly (9.5 56 m) upstream of the target. An envelope beam model is used to match the measured beam sizes and extrapolate the beam size to the target. The beam profiles at the Harp are assumed to be the same on the target, with only the RMS sizes changed as per model extrapolation. This information, along with the measured beam current, provides the peak beam density, and beam sizes on the target. The beam center at the target is provided by extrapolation (using the same envelop model) of the measured positions at the BPMs through the lattice to the target, and is also available directly through the TIS. We used the BPM method for initial beam setup until early 2011, after which the TIS beam center measurement was adopted. The BPM extrapolation method differs from the TIS beam center measurement by about -8 mm horizontal and -2 mm vertical.
Figure 3: Layout of the beam-line approach to the Target. Table 2: Summary of Target run time protective measures. Limit Type Harp beam size within 10% of setup Alarm Harp peak intensity < 110% of setup Alarm TIS within 6 (4) mm center horizontal Alarm (vertical) Halo thermocouples < 205-350 o C Alarm ++ Halo thermocouples < 450 o C Quadrupoles within 3.5% nominal Dipoles within 0.5 A setup Last 4 dipole correctors within 0.5 A setup Power < prescribed limit depending on Software run conditions Vertical position offset < 4mm Software + Horizontal position offset < 6 mm Horizontal (vertical) beam size < 49 (17) Software cm Ring injection and extraction kicker Software * waveform monitors Loss monitors and software + This is also indirectly enforced with the waveform monitors ++ limit varies by thermocouple based on experience * Being converted to hardware The TIS also provides direct profile information, but profile interpretation is complicated by non-uniform degradation of phosphor emission with time, possible phosphor light induced from the intense radiation produced in the target and the complicated optical system used to gather the image. Because of these concerns the TIS is not presently used to control the beam size and intensity. We have compared the beam sizes on target from the extrapolation and TIS methods, and we typically see differences up to ~40%. But these differences are not always the same between setups, and they are not well understood. Controls methods employed during production operation to protect the target, and ensure the limits discussed above are respected, are summarized in Table 2. Production Stability of Beam Parameters After the initial beam characterization on target, the power is ramped up. To ensure that the beam position stays centered, we also monitor the Ring injection and extraction pulsed kicker magnet current waveforms to ensure they are within prescribed tolerances of the nominal levels. This is critical because it is possible for just one of the 13 extraction kickers to vary in amplitude or time, and cause the beam position to deviate beyond allowable limits, without causing the beam loss to increase above the trip limits. This situation was determined empirically. Until February 2011 beam centering on the target, after power ramp-up, was ensured by maintaining balanced halo thermocouple temperature measurements at the beam window just upstream of the target. Once the beam power reached a few hundred kw, operators would steer the beam to ensure left-right and up-down balanced halo monitor temperature readings. This process assumed symmetric halo profiles and that the center of the thermocouple array is the same as the center of the target. After Feb. 2011, the TIS became the primary diagnostic for beam centering, and asymmetric halo temperatures were accepted, but the thermocouple temperatures are limited to a value based on calculations for a safe amount
of beam impingent at the outer radius of the window. Figure 3a shows the history of the difference in the target window halo thermocouple differences (left-right and updown). After February 2011, there is an increase in the vertical asymmetry. Fig. 3b shows the beam center reported by the TIS, with a corresponding reduction in the beam position after February 2011. We note that the typical fluctuation of the beam position on the target over month long production runs is typically < 1-2 mm. a) b) Figure 2: a) Temperature asymmetry at the target beam window, and b) beam center measured by the TIS Figure 3: Long term beam size measurement at the Harp upstream of the target. Another quantity of interest is the beam size on target. As mentioned above, the nominal method for this determination is by propagating measured beam sizes downstream to the target. The only active upstream beam profile measurement during operations is the Harp 9.5 m from the target. Figure 3 shows long-term stability of the beam size measured at the Harp. The discontinuity at the end of 2010 is due to the adoption of new beam-line lattice optics between the Harp and the target (beam size on the target did not change). The beam size stability is typically < 1-2 mm over the course of a one-month production run. Also we monitor the integrated beam flux on the vacuum windows in the RID and target beam lines and for the targets, to compute radiation damage effects (e.g. DPA). This information is used for lifetime / replacement schedule purposes. SUMMARY Ensuring proper control of the beam parameters on high intensity beam dumps and targets is a critical concern. The methods employed at the SNS to ensure beam on target control and the typical beam on target stability are summarized. Some issues have arisen that were not anticipated during design. These include implications of operator tuning to optimize the operational setup after the production has been initially qualified. Also, multiple instrumentation methods can give differing indications of the beam parameters on target. REFERENCES [1] P. Hurh, et al., Targetry Challenges at Megawatt Proton Accelerator Facilities, THPFI083, Proceedings of the 4 th International Particle Accelerator Conference p. 3484 (2013); http://accelconf.web.cern.ch/accelconf/ipac2013/p apers/thpfi082.pdf. [2] M. Plum, M. Holding, T. McManamy, Beam Parameter Measurement and Control at the SNS Target, Proceedings of 2005 Particle Accelerator Conference, Knoxville, Tennessee, p. 3913 (2005); http://accelconf.web.cern.ch/accelconf/p05/papers/f pae074.pdf [3] S. Henderson et al., Exploration of Beam Fault Scenarios for the SNS, Proceedings of the 2003 Particle Accelerator Conference, p. 1573 (2003); http://accelconf.web.cern.ch/accelconf/p03/papers /TPPE012.PDF [4] M. Plum, SNS Injection and Extraction Systems Issues and Solutions, Proceedings of Hadron Beam 2008, Nashville, Tennessee, USA, p. 268-270, http://accelconf.web.cern.ch/accelconf/hb2008/pap ers/wgc04.pdf. [5] T. Shea et al., Status of Beam Imaging Developments for the SNS Target, Proceedings of DIPAC09, Basel, Switzerland, p. 38-41, (2009), http://accelconf.web.cern.ch/accelconf/d09/papers/m ooc04.pdf [6] W. Blokland et al., SNS Target Imaging System Software and Analysis, Proceedings of BIW10, Santa Fe, NM, US, p. 93-95, http://accelconf.web.cern.ch/accelconf/biw2010/pa pers/tupsm003.pdf