© 1983 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material
for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers
or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
2138 IEEE Transactions on Nuclear Science, Vol. NS-30, No. 4, August 1983
REMOTE HANDLING AND ACCELERATORS*
Mahlon T. Wilson, AT-DO, MS H811
LOS Alamos National Laboratory, Los Alamos, NM 87545
Summary LAMPF operated for over a year (1974) at reduced
intensities to permit hands-on maintenance because
The high-current levels of contemporary and pro- some of the main beamline components and much of the
posed accelerator facilities induce radiation levels shielding had not yet been installed. Another year
into components, requiring consideration be given to (1975) was required to complete equipment and shield-
maintenance techniques that reduce personnel exposure. ing installation. During that time, a second manipu-
Typical components involved include beamstops, tar- lative system was proposed that was more versatile
qets, collimators, windows, and instrumentation that for use in areas not requiring the complete shielded
intercepts the direct beam. Also included are beam handling MERRIMAC provided. This system, named
extraction, injection, splitting, and kicking regions, MONITOR.6g7 and its offspring have become the prime
as well as ourposeful spill areas where beam tails are manipulative systems; MERRIMAC has been scrapped, and
trimmed and neutral particles are deposited. Scat- the complete shielding requirement has been waived.
tered beam and secondary particles activate components ~MONITOR began in 1976 as a hydraulic-powered
al? along a beamline such as vacuum pipes, magnets, truck self-loader to which a switch-operated electric
and shielding. Maintenance techniques vary from manipulator (PaR 150) was attached. The boom of the
"hands-on" to TV-viewed operation using state-of-the- self-loader is capable of placing the small, human-
art servomanipulators. Bottom- or side-entry casks arm sized, manipulator near the beamline components
are used with thimble-type tarqet and diagnostic located as far as 8 m below the top of the shielding.
assemblies. Long-handled- tools -are operated from Viewing is with television. A prototype hydraulic
behind shadow shields. Swinging shield doors, master/slave manipulator was obtained from the Navy,
unstacking block, and horizontally rolling shield modified slightly, and mounted alongside the PaR.
roofs are all used to provide access. Common to all This arm was manufactured by the Remotion Company,
techniques is the need to make operations simple and weighs 8 kg, and responds to the position of the mas-
to provide a means of seeing and reaching the area. ter arm very rapidly.' The PaR 150 manipulator
later was replaced by the PaR 3000 salvaged from
Introduction MERRIMAC. This equipment was used on several occa-
sions including working with a beam exit window and
A survey paper was presented at this conference collimator in a lo4 R/h environment. Great increase
ir 1975 that discussed radiation levels, shielding, in manipulative speed was achieved through the acqui-
radiation damage, and remote handling at the then new sition of two TOS Model 229 bilateral servomanipula-
high-power accelerator facilities of LAMPF, SIN, t&-s of lo-kg capacity. These manipulators have been
TRIUMF, and FERMILAB.' The facility descriptions used as a pair of arms or singly, paired with either
given have remained fairly constant over the inter- the Remotion arm or one of the PaR manipulators.g~'O
vening 8 years; however, operational experience has LAMPF today has experienced the most remote han-
been gained, and some changes in remote-handling dling and the highest radiation levels of any accel-
equipment have occurred. This paper will discuss the erator. The main proton line has radiation levels of
additional information. 1000 R/h in the first target cell, 2000 R/h in the
second, 800 R/h at the biomed target cell and lo5 R/h
LAMPF in the beamstop area. All operations are performed
remotely using TV viewing. There is absolutely no
Design of the LAMPF Line-A equipment and facili- direct viewino because radiation levels from 10 to
ties was based on a philosophy of providing ready 25 R/h exist "at the edge of the open shield doors.
removal of equipment expected to require routine main- Shadow shields and distance are used to keep radiation
tenance (position monitors and targets), while provid- levels within limits where people must work on top of I
ing general access to all other components.' The the shielding.
position monitors and targets are mounted on shielding The remote-handling crew consists of 3 engineers,
stalks that are inserted into vacuum-tight enclosures 4 draftsmen, and approximately 15 technicians, under
that penetrate to the surface of the bulk shielding the direction of Donald L. Grisham. They can field
from the beamline. The drives, utilities, and vacuum three crews for continuous operation. A fourth crew
seals are, therefore, in an accessible location for builds tooling, and components to support the manipu-
hands-on-connection and service.3y4 Access to lating crews. When not operating, they are continu-
the beamline components is provided by horizontallv allv maintaininq and upqradinq the manipulative sys-
opening shield doors that weigh hundreds of tons: terns and building new systems, They are assembling
This is accomplished bv insertinq roller trucks under -
a third MONITOR to utilize the third TOS Model 229
the doors and'transferrinq the weight of the door on;o manipulator that they have acquired. They have a good
the trucks by hydraulically expanding the trucks, operating history because of their preventative main-
which are then oushed or Dulled with a pair of hori- tenance program. An example being a 5-month stint of
zontal cylinders.5 The 'original requirements that operating 24 h/day, 7 days/week (minus 3 days to rest
the radiation emanating from the beamline components people) with only 24 h of downtime to repair manipu-
be shielded and that all components removed from the lator problems. The first two TOS arms have clocked
beamline be transported in a cask resulted in the 210 000 h each in practice, tune-up, and operation
construction of a portable hot cell dubbed MERRIMAC. with about 6000 of this being actual working hours.
The shield doors, shielding configuration, and hot The LAMPF main beamline has been plagued with
cells of Area A were all designed to be compatible water and vacuum leaks resulting in an awesome assort-
with the MERRIMAC concept, which put rather tight ment of tasks that range from the replacement of five
constraints on the configuration of the shield doors. of the 17-ton target-cell triplets, through welding,
*Work supported by the IJS Department of Energy.
soldering, polishing flanges, to making up electrical diameter, 45-cm-high vacuum chamber by jacking the top
connectors. A time-consuming task has been the half of the cyclotron upward by 1.2 m. A service
removal of the close-in shielding that is fitted bridge is inserted into the opened vacuum chamber,
around the beamline components within the target with the inner end supported on the cyclotron center
cells. They have been reducing the number of pieces post and the outer end supported on wheels that ride
by welding them together into larger sections. on the cyclotron periphery. A drive system positions
the bridge at any radial location desired. Various
- specialized trolleys can be mounted on the service
bridge to accomplish inspection and maintenance
The main proton beamline, two target assemblies, functions.
and the beam dump are located within a vault that is Personnel access into the vacuum chamber requires
approximately 40 m long, 5 m wide, and 3 m high. rhe the remote placement of ~10 tons of 5-cm-thick lead
roof slabs are removable to provide access from above. shields over the vacuum chamber walls 'where neutral-
Convenient personnel access is provided. The target ized beam particles have impinged at the beam plane
assemblies are surrounded by close-in shielding and a level. It requires 220 h to install or remove these
side-entry cask is used to transport the target assem- lead shields remotely with the service bridgetrolley
blies to hot cells for maintenance. Maintenance was system. Four copper blocks located at places of known
intended to be accomplished by hand or with the use of high beam spill also are removed. This shielding and
shadow shields and long tongs. A manipulator system removing procedure reduces the general background
called MINIMAC" was assembled to handle unforeseen within the chamber to the lD- to 15-mR/h level.
events that involve radiation levels high enough to A goal of 300 uA by 1986 will require replac-
preclude the use of people. ing the 80 resonator panels whose drooping has been a
MINIMAC consists of one PaR Model 3000 switch- continuous problem. The present resonator panels are
controlled manipulator mounted on a trolly system that x0.8 m wide by 3 m long and weigh 275 kg each.
rurs on a 4- by 6-m frame. Suspended from the frame There are specialized trolleys for remotely replacing
are television cameras and a tool board. A 3-D tele- both upper and lower resonator panels of the current
vision camera is mounted on the manipulator. This design.
camera also may be used as a Z-D camera to reduce Beamline-component maintenance requires urstack-
operator eye fatigue when the nature of the manipula- ing the covering concrete blocks to expose the com-
tion does not require depth preception. A bent Quasar ponent. A platform, having a 5-cm-thick lead floor,
telescope allows an observer to peer over a shadow is placed over the opening and an operator on the
shield for additional viewing capability. platform performs manipulations 3 to 5 m below, using
Although MINIMAC was conceived for emergency use long-handled tools. Television is used to provide
only, exoerience showed it could be used for many viewing. Radiation levels of 10 to 15 R/h at the
operations that oriqinally were planned to be accom- beamline are attenuated to the lo-mR/h range at the
plished with shadow-shields and long tongs. MINIMAC operator by this system. The beamline vacuum connec-
has had approximately a month's usage each year since tions use double indium seals, which on occasion
1979, doinq modifications or repairs to the main beam- develop leaks. Repair requires four to six hours to
line production-target enclosures. Radiation levels unstack shielding, one hour to set up the platform and
in the lOOO- to 2000-R/h level were measured on com- actually change out the seals, and another four to six
ponents removed by MINIMAC. hours to replace the shielding. All beamline compo-
SIN has two engineers and four technicians, under nents other than the target chambers are maintained in
the direction of Eyke Wagner, that are responsible this fashion.
for the remote-handling effort. Five individuals are The target systems corsist of shielding thimbles,
capable of operating the manipulator. below which are suspended ladder targets that can con-
Wagner has developed a clever method of disposing sist of Be, Cu, H20, and C. Typically, these are
of activated components. He has designed concrete run until a water leak develops somewhere in the lad-
blocks whose outside dimensions are the same as SIN der, which is attached to the thimble with Swagelock
standard shielding blocks but contain a large void fittings. Routine replacement of a target ladder
into which an activated component is lowered remotely. requires about a half day. The thimble is withdrawn
A lid covers the component, and the remaining void may vertically into a bottom-entry cask that is hoisted
be filled with grout, completing the encapsulation. into a hot'cell for reolacement using master/slave
The activated component and its surrounding concrete manipulators. A bellows associated with the vertical-
thus serve as a shield block for the remainder of the motion portion of the target system has a lifetime of
life of SIN, by which time the component will have about a year. A full day is required to change this
decayed to manageable levels. bellows remotely.
Future plans include changing the thin target TRIUMF has acquired a Unimate programmable robot
system in 1984 and the thick target and beamstop in in cooperation with the University of British Colum-
1988. The beamline vault will be filled with shield- bia Engineering Department. The robot will be used
ing. Vertical shield plugs will provide access to the for welding operations within the vacuum chamber.
beamline components. The electrical and water ser- The remote-handling section at TRIUMF consists
vices for the components will rise through the shield- of one engineer, four designers, two electronic tech-
ing for manual maintenance. The beamline vacuum con- nicians, two mechanical technicians, ard one hot-cell
nections will consist of inflated bellows attached technician, under the direction of William Cameron.
between pairs of flanges similar to the system used on Of this group, two are familiar with hot cell opera-
their cyclotron vacuum chamber. A shielded area will tions, three with beamline remote handling, and five
be assembled from standard blocks and equipped with with service bridge-trolley handling.'3,14
lead-glass windows and master/slave manipulators to
allow performance of more on-site work on activated CERN
For over a dozen years, Roger A. Horne at CERN
TRIIJMF has been promoting the use of remote handling using
servomanipulators to reduce personnel dosage. At that
TRIUMF routinely produces a 450-MeV proton beam time he acquired a pair of Mascot servomanipulators
with a l!lO-PA intensity. Cyclotron maintenance is that were manufactured by Selenia in Italy." The
based upon the ability to gain access into the 18-m various CERN groups historically preferred to do
hands-on maintenance; however, because of growing FERMILAB
radiation levels and a stable work force, the dosages
of the maintenance workers is approaching the CERN The final beam-transport magnets and the produc-
limit of 1.5 R/yr. The decision to build LEP with no tion tarqets for the neutrino area at FERMILAB are
increase in personnel or relaxation of the dosage mounted on bedplates, which are picked up and carried
limit has increased interest considerably in doing bv a railroad svstem.17 Various train assemblies.
more work remotely. which are dozens-of meters long, exist to provide an
Horne has assembled a remote-handling vehicle, assortment of target and beam geometries. Trains may
named MANTIS after the insect it resembles, that has be driven between locations in an evacuatable target
become so in demand to handle modifications and tube for producing mesons, in shielded storage tun-
repairs that management has provided funds to build nels, and in a service building for maintenance or
additional upgraded systems and new manipulators. equipment change out. The service building contains
MANTIS consists of a self-propelled hydraulic bridge-mounted servomanipulators.'e One pair of
truck loader that carries the Mascot manipulator servomanipulators are the prototype Mark IV units
arms.16 In the traveling modes, it folds up into a developed by the Remote Control Division of the
compact package 1 m wide, 2.2 m high and 4.5 m long, Argonne National Laboratory. The second pair are
weighing 6 tons. The compact configuration is slightly modified CRL Model M manipulators, which look
required to enable it to thread among the equipment somewhat like the prototypes but with one arm having a
in the accelerator tunnels. A quarter-ton of spooled higher capacity. All viewing uses television. The
cable mounted on a trailer allows MANTIS to maneuver control room for remote operations is 130 m outside
80 m in its manipulations; another 250 m of cable the shielded area. Hot items requiring remote main-
connect the trailer to the master control station. tenance are moved to the service building.
Before beginning work, either MANTIS or an operator The train system provides a very fast way to
runs into the work area, unrolling a black rubber mat change out a large number of complicated and heavy
upon which a white line is drawn. MANTIS then auto- components. The just-removed train may be stored for
matically follows the white line in and out of the future use or modified. Since 1975, three trains have
area. This is a clever and quick way to deliver com- been scrapped, requiring ~500 h of remote-handling
ponents and pick up tools without the risk of collid- operations. The average hottest item is 30-50 R/h at
inq with shield blocks or accelerator apparatus. It 36 cm and remote-handling crews have worked on items
was observed that a white line painted on the floor as hot as 200 R/h. They prefer to do as much dis-
was auicklv eroded away because people tended to fol- assembly and reassembly Remotely, as time allows, to
low 'it when walking,- and equipment configurations reduce dosage, using local shielding of hot areas to
changed so frequently that no path remained the same permit some hands-on work. Early trains that were
for any length of time. hastily assembled proved to be difficult to work on
Upon arriving at the work area MANTIS unfolds remotely. The later trains had more attention qiven
itself and an operator can position the MASCOT manip- to remote maintainability, which has permitted moge to
ulators anywhere within an 8-m radius, including over be done at lower dosaqes. Jack Lindberg directs the
shielding walls. remote-handling effort-at FERMILAB.
The 13-yr-old Mascot servomanipulators have oper-
ated over 2500 h. They use glass-tube electronics, SLAC
and as only a half-dozen arms were manufactured, parts
must be custom constructed. Prototype work has been The SLAC switchyard was provided with long tool-
completed at CERN on a new generation of compact ser- maintenance capability. However, equipment reliabil-
vomanipulators using some of the concepts advanced by ity has been very good, and the activation is more a
Kentner Wilson's hydraulic arm.' A complete arm nuisance than a problem. SLAC beam powers of 200 to
will weigh 20 kg and can handle 25 kg. It is intended 300 kW are only half of those of a few years ago
to build six of these force-reflecting hydraulic ser- because of changing experimental emphasis. Targets
vomanipulators within the next year. The remote- and beam dumps are activated to the 15 R/h range on
handling community is watching this development with contact after a day of cooling. Additional coolinq is
interest because the power, speed, and simplicity of allowed to permit maintenance work to be done by hand.
hydraulic manipulators is appreciated; however, the Dieter Walz. having been responsible for the desion4 of
risk of spraying oil around a hot cell and the tradi- the components that absorb power, is the most know-
tional use of electrical motor power manipulators have ledgeable at SLAC about radiation levels and how they
restricted use within the reactor fuel community. design to accommodate them.
Under the direction of Roger Horne, the permanent
remote-handling team consists of an engineer, a tech- Manipulators
nician,, and a mechanic. Operations require two peo-
ple, one of which concentrates on the manipulations Manipulative equipment originated within the
while the other observes, plans, and helps make deci- National Laboratories in support of the needs of fis-
sions. After a half-day of work, the operator leaves sion reactor fuel development. These original designs
because of fatigue, the observer becomes the operator, were developed into commercial products by private
and the third person becomes the new observer. industry, and for the past generation PaR and CRL have
The remote-handling group prides themselves on been household acronyms within the remote-handling
being able to go anywhere and do any job quickly. To community. Essentially, every established hot-cell
do this may require a few months preparation to do a Laboratory in the country has through or over-the-wall
few weeks of work, esoecially if special tooling is master/slave manipulators manufactured by Central
required or access is particularly difficult. Most of Research Laboratories (now a division of Sargent
the areas of CERN were built without regard for ease Industries)." The heavier manipulations are han-
of remote maintenance. The remote-handling group dled by rate-controlled (switch-operated motors with
insists that a senior technician who really knows the no feedback) Programmed and Remote Systems (now
nuts and bolts be their only interface with the group GCA/PaR Svstems)" manipulators that are mounted .__- on _.
controlling the area where work is to be done, and telescopi;g tubes supported by trolleys riding on
that he be with them for guidance when they are doing bridges within the hot cell. A few other companies
the remote work. As the necessity of remote opera- have been involved intermittently with the production
tions increases, more groups are asking for guidance of manipulators, but the market is declining and
in equipment design to facilitate remote maintenance. development costs are so high that new-equipment
introductions are infrequent and usually are either 7. J. J. Burgerjon, E. L. Ekberg, D. C. Grisham, R.
funded by the first customer or are an entrepreneural A. Horne, R. E. Meyer, C. R. Flatau, K. B. Wil-
spin-off of a device developed at a Laboratory. An son, "A Solution for Remote Handling in Acceler-
example is the creation of TOS, TeleOperator Systems ator Installations," IEEE Trans. Nucl. Sci. 2,
Corp.,*' by Carl Flatau whose first product was No. 3, p. 1580 (June 1977).
produced on an order by LAMPF and is a descendent of
servomanipulators he developed at Brookhaven National 8. Kentner B. Wilson, "SERVOARM-A Water Hydraulic
Laboratory as a proposed solution to the maintenance Master-Slave Manipulator," Proc. 23 Conf. on
problems associated with increasing the intensity of Remote Systems Technology, American Nuclear
the AGS in the midsixties." Society, p. 233 (1975).
A very expensive and slowly acquired legacy of
experience within the remote-handlinq machining indus- 9. D. L. Grisham, E. L. Ekberg, J. E. Lambert, R.
try could well wither away from lack of adequate mar- E. Meyer, P. J. Stroik, and M. D. Wickham,
kets to support the skilled teams necessary to produce "MONITOR 1979," IEEE Trans. Nucl. Sci. 3,
these items. The sale of one or two servomanipulators No. 3, p. 3376 (June 1979).
a year by CRL or TOS is inadequate to keep these
products alive. 10. D. L. Grisham, E. L. Ekberg, J. E. Lambert, R.
"Why not send a robot in to do the work?" is an E. Meyer, P. J. Stroik, and M. 0. Wickham,
often asked question. There are 23000 industrial "MONITOR Update-1979," Proc 27 Conf. on Remote
robots at work in the United States today. Most are Systems Technology, American Nuclear Society,
quite primitive, with no capacity to see or feel or p. 268 (1979).
respond to their environment in any significant way.
Poor mechanical accuracy and repeatability, along with 11. Eyke Wagner and Albin Janett, "MINIMAC-The
inadequate control and cumbersome teaching, will delay Remote-Controlled Manipulator with Stereo Tele-
robot entry into the delicate and often unpredictable vision Viewing at the SIN Accelerator Facility,"
realm of accelerator remote handling. The most likely Proc. 26 Conf. on Remote Systems Technology,
scenario would be the adding of intelligence to a American Nuclear Society, p. 62 (1978).
specific operation, like Roger Horne has done with
MANTIS in havinq it follow a white line. 12. Eyke Wagner, SIN, CH-5234 Villigen, Switzerland,
Approximately 500 people involved in all aspects 011-41-56-99-3574.
of remote handlinq constitute the membership of the
Remote Systems Technology Division of the American 13. William Cameron, TRIUMF, 4004 Wesbrook Mall,
Nuclear Society. The yearly publication, "Proceedings Vancouver, B. C., Canada V6T 2A3, (604) 228-4711.
of the Conference on Remote Systems Technology," is a
compilation of the complete papers presented at their 14. Clive Mark, TRIUMF, 4004 Wesbrook Mall, Van-
conferences durinq the year. The dozen senior people couver, 6. C., Canada V6T 2A3, (604) 228-4711.
involved with -the -accelerator remote handling
described in this paoer are reaular oarticipants in 15. R. A. Horne and M. Ellefsplass, "Long-Range,
the activities of this Division: A buyer's guide is High-Speed Remote Handling at the CERN 26-GeV
published annually by the America Nuclear Society that Proton Synchrotron," Proc. 23 Conf. on Remote
lists suppliers that cater to the nuclear industry. Systems Technology, America1 Nuclear Society,
p. 247 (1975).
16. R. A. Horne, "MANTIS-A Compact Mobile Remote-
1. Mahlon T. Wilson, "Radioactivity, Shielding, Handling System for Accelerator Halls and Tun-
Radiation Damage, and Remote Handling," IEEE nels," Proc. 26 Conf. on Remote Systems Tech-
Trans. Nucl. Sci. 2, No. 3, p. 1042 (June 1975). nology, American Nuclear Society, p. 55 (1978).
2. Mahlon T. Wilson, "Remote Maintenance Concepts 17. J. Lindberg, J. Grimson, J. Simon, and D. Theri-
for The Los Alamos Meson Physics Facility," IEEE ot, "FERMILAB Target Areas and Target Train Sys-
Trans. Nucl. Sci. fi, No. 3, p. 588 (June 1969). tems," Proc. 23 Conf. on Remote Systems Tech-
nology, American Nuclear Society, p. 24 (1975).
3. Mahlon T. Wilson, L. L. Thorn, L. 0. Linquist,
0. L. Grisham, "The Evolution of the LAMPF High 18. J. Lindberg, J. Grimson, S. Mori, J. Simon, and
Power Pion Production Target Mechanisms," IEEE H. Stredde, "Operating Experience at FERMILAB,"
Trans. Nucl. Sci. 4, No. 3, p. 1574 (June 1977). Proc. 26 Conf. Remote Systems Technology, Ameri-
can Nuclear Society, p. 65 (1978).
4. L. Agnew, T. S. Baldwin, D. L. Grisham, R. C.
Holmberg, J. E. Lambert, L. 0. Lindquist, R. D. 19. Sargent Industries, Central Research Laboratories
Reiswig, and L. L. Thorn, "Graphite Targets for Division, Red Wing, MN 55066, (612) 388-3565.
Use in High Intensity Beams at LAMPF," IEEE
Trans. Nucl. Sci. 26, No. 3, p. 4143 (June 1979). 20. GCA/PaR Systems, 3460 Lexington Avenue No.,
St. Paul, MN 55112, (612) 484-7261.
5. Mahlon T. Wilson, "Kiloton Shield Doors at
LAMPF," Proc. 22 Conf. on Remote Systems Tech- 21. TOS, TeleOperator Systems Corp., 45 Knickerbocker
nology, American Nuclear Society, p. 43 (1974). Ave, Bohemia, NY 11716, (516) 567-8787.
6. R. A. Horne and E. L. Ekberg, "MONITOR-A Versa- 22. C. R. Flatau, "General Purpose Servo-Manipulator
tile Remote Handling System," Proc. 23 Conf. on for Remote Maintenance of Accelerators," IEEE
Remote Systems Technology, American Nuclear Trans. Nucl. Sci. 16, No. 3, p. 594 (June 1969).
Society, p. 242 (1975).