Generic control system design for the Cassette

Fusion Engineering and Design75–79(2005)553–557Generic control system design for the CassetteMultifunctional Mover and other ITERremote handling equipmentM.Irving a,b,∗,J.Palmer b,M.Siuko ca ENEA CR Brasimone,40032Camugnano,Bologna,Italyb EFDA-Close Support Unit-Garching,Boltzmannstrasse2,85748Garching,Germanyc Tampere University of Technology,P.O.Box589,33101Tampere,FinlandAvailable online26July2005AbstractA new suite of Remote Handling(RH)equipment,addressing the latest design of the divertor region,is currently being specified for ITER,with procurement of the Cassette Multifunctional Mover(CMM)scheduled for2005–2006.This presents a unique opportunity to address a vital component of the RH system–the control system–and this paper introduces a number of concepts based on replication of design and implementation for functionally similar RH systems.Correctly implemented,this should improve equipment availability and reliability,reduce overall support requirements,and minimise the risks of serious failure.©2005Published by Elsevier B.V.Keywords:ITER;Divertor;Remote Handling;Control systems1.IntroductionIn general,mechanical handling equipment for industry is specified from a principally functional per-spective.These functions concentrate on the mechani-cal aspects of the tasks to be carried out,but although RH equipment may appear to fall into the same cate-∗Corresponding author.Tel.:+498932994141/+390534801419; fax:+498932994224/+390534801225.E-mail addresses:mike.irving@brasimone.enea.it,mike.irving@(M.Irving).gory,there are other important factors which must be considered.Fundamental to this process is recoverabil-ity,since the handling equipment must be deployed in hostile areas where human intervention and repair is not possible,or is at best extremely undesirable.A failure which renders the RH equipment immobile inside the ITER vessel or ducts is considered catastrophic and could compromise the whole project.To minimise the risks of such an occurrence,in-depth studies of various aspects of the RH design are clearly essential,and this paper will focus on one aspect:design and construction of the control systems0920-3796/$–see front matter©2005Published by Elsevier B.V. doi:10.1016/j.fusengdes.2005.06.174554M.Irving et al./Fusion Engineering and Design75–79(2005)553–557and Human–Machine Interfaces(HMI’s)which form the interface between the operator and the remote equipment.2.Two types of RH equipmentFrom a functional and control perspective,ITER RH equipment can be placed into two main categories: Movers(slow,repetitive,large motions,heavy loads, programmed moves)Manipulators(fast,light,responsive,low payloads, versatile,master–slave)With few exceptions,the concepts for ITER RH equipment to date fall quite neatly into these two types. Divertor and blanket handling devices are clearly heavy component‘movers’.Motorised RH tooling,such as pipe cutters and welders,apart from size,is likely to be very similar in control requirements to the mover category.Manipulators,on the other hand,require dexterity and versatility,and are highly complex and specialised mechanisms requiring very tight real-time control. Whilst they can still benefit from some common guide-lines at the specification stage,they are generally less amenable than movers to standardisation.As an example of this,the existing ITER RH Diver-tor equipment prototyped at Brasimone:the Cassette Toroidal Mover(CTM),the Second and Radial Cas-sette Carriers(SCC and RCC),the Tractor Unit(TRC), the Central Cassette Carrier(CCC),and the Auxiliary Duct Vehicle can all be considered‘movers’.However, whilst all are similar in function,and all were able to carry out their specified tasks successfully,Irving [1],the control systems and HMI’s which they employ could not have been more different.3.Problems with this approachEven in this relatively small subset of RH equip-ment,it was found that each system required substantial time,both for operator familiarisation,and for the tech-nical staff to analyse and understand the system well enough to identify and resolve problems efficiently. In practise,it was found that a different member of the team became the local‘expert’on each item of equipment,since a deep technical appreciation of all the equipment was beyond the capabilities of any indi-vidual.Clearly,in an ITER context,such dependency on individuals increases the potential risks of incorrect decision-making should this individual be unavailable during critical periods,such as operational failure or maintenance.Recoverability of failed remote equipment is another primary consideration.In the past,design and manufacture of the controller has generally been the responsibility of the company producing the mechani-cal system,as would likely happen for normal materials handling equipment in industry.The result of this is that,in addition to wide varieties in system architec-ture and HMI’s,built-in facilities for on-line moni-toring of the equipment,and provision for effecting recovery varies enormously from one supplier to the next.Ultimately it is essential to acknowledge that,as well as handling normal operation of the equipment, the control system and its interface form a critical link between the operator and a potentially failed remote device.Clearly this is an area worthy of formal defini-tion,rather than a supplement to a mechanical system specification.4.An alternative approach:a generic architecture for the CMM systemThe fact that the majority of ITER RH equip-ment can be easily categorised indicates that common designs may well be possible,at least for the‘mover’type.One basic design can be utilised for many,pos-sibly all the ITER RH movers:the CMM,the Cassette Toroidal Mover,the transport cask,the transport cask to port docking system and many more.This is the approach suggested for the CMM.As outlined in Palmer[2],the CMM system will comprise of a total of justfive axes:lift,tilt,radial drive, and two end effector movements,see Fig.1.Due to the requirements of high thrust in small volumes,all but the radial drive will employ water–hydraulic actuators.Conventionally,this type of actuator is driven by a servovalve which controls theflow-rate according to a (small)currentflowing in its coil.This current is usu-ally generated by an analogue circuit which itself takes a±10V command,a motion control industry standard.M.Irving et al./Fusion Engineering and Design75–79(2005)553–557555Fig.1.The joints of the CMM.Fig.2.The CMM control system concept.556M.Irving et al./Fusion Engineering and Design75–79(2005)553–557In general,servovalves are very similar in size and per-formance.Electric servomotors develop a torque proportional to the current passing through them,provided by an electronic unit called a servoamplifier,which can be specified to take the same±10V command to pro-duce a commanded velocity.Small servomotors will require small servoamplifiers and vice versa,but the input requirements are usually identical.So despite the CMM actuator types being very diverse,the control requirements are still very similar.Prior to the1990s,the processing heart of the control system usually comprised of a Central Processor Unit (CPU),and a number of memory and analogue/digital input/output cards,with the software written specifi-cally for each application.In addition to being complex, bulky and difficult to program,the designers and pro-grammers had to integrate all the safety and diagnostic provision,requiring a range of experienced personnel throughout the development of the equipment.The modern approach employs a dedicated subsys-tem,known as a Motion Controller(MC),which has become the de facto standard in industry.This inte-grates all the processing,memory and input/output interfaces normally required onto a single board han-dling typically up to eight axes,which can be duplicated for more complex requirements.Less standard appli-cations can readily be accommodated by additional special purpose boards.The system as a whole incorpo-rates all the complex motion control software,includ-ing management of the servo loops at high refresh rates, and comprehensive safety provision,representing sev-eral man–years of dedicated effort.After10–15years, the MC industry is now very mature and active,and since the MC is a versatile device,the programmer has only to select the functionality required for the particu-lar application.This is the direction chosen by JET for its major mover type equipment,and the experiences have been very positive,Haist[3].In this context,the CMM is quite a simple device, and a block diagram of a provisional control system, based on a single MC,appears as Fig.2.5.General characteristics of RH control systemsEssential features of an RH controller mustinclude:Fig.3.JET Boom Controller,showing modularity.•Designed in recoverability:◦good access to all signals and power wires for monitoring and/or override◦use of dc brushed motors for easy basic recovery.M.Irving et al./Fusion Engineering and Design75–79(2005)553–557557◦software allowing supervisors to override inter-locks and access internal information easily •A common architecture for:◦reduced collective knowledge requirement◦lower dependency on individual‘experts’◦improved prospects of recovery of failed equip-ment◦reduced spares provision•The use of MC’s at the heart of the control system design.Desirable features include:•modularity of hardware and software(see Fig.3):◦to enable repairs by simple replacement of com-plete racks◦for upgrading of complete subsystems with low downtime◦to allow component level repairs to be carried out in the workshop•built-in:◦diagnostics for fault identification without test equipment or technical drawings◦provision for continuous monitoring of key parametersThese features have been incorporated into an initial requirements specification and outline design for the CMM control system,Irving[4].6.ConclusionsITER can avoid many of the problems already encountered in RH fusion environments by speci-fying thefinal control system separately from the mechanics.By defining such specifications generically where possible,features specific to RH applications can be collectively addressed,resulting in reduced risks,lowered costs and improved availability.As thefirst of a new suite of ITER RH equipment,the CMM is the ideal opportunity to put this concept into practise.References[1]M.Irving,et al.,Remote operational trials with the ITER FDRdivertor handling equipment,in:Proceedings of the22nd SOFT, Helsinki,September,2002.[2]J.Palmer et al.,Recent developments towards ITER2001Diver-tor Maintenance,This Symposium.[3]B.Haist,et al.,A rational approach to Remote Handling equip-ment control system design,in:Proceedings of the Ameri-can Nuclear Society Robotics and Remote System Conference, Washington,March,2001.[4]M.Irving,Specification and design of the Cassette Multifunc-tional Mover(CMM)control system and HMI for the ITER application,ENEA Report to EFDA MOV-CI-S-001,October 2003.。