《The integration, commissioning and initial operation of the SLACBaBar detector magnet & cryogenic plant.doc》由会员分享，可在线阅读，更多相关《The integration, commissioning and initial operation of the SLACBaBar detector magnet & cryogenic plant.doc（8页珍藏版）》请在三一办公上搜索。
1、The Integration, Commissioning and Initial Operation of the SLAC-BaBar Detector Magnet & Cryogenic PlantP. L. Anthony,1 W. Burgess,1 A. Candia,1 W.C. Craddock,1 W. Kaminskas,1 P. Fabbricatore,2 T.G. OConnor,3 C. Priano,2 and P. Valente41SLAC, Stanford, CA 94309, USA2INFN, Genova, Italy3LLNL, Livermo
2、re, CA 94551, USA4Ansaldo Energia, Genova, ItalyABSTRACTThe Stanford Linear Accelerator Center (SLAC) BaBar liquefier1 was commissioned in August 1997 and the 27 MJ BaBar detector solenoid magnet2 in April 1998. The magnet and its systems have been operating successfully since October 1998. Valuable
3、 experience has been gained, resulting in system hardware and procedural changes from the original specification. Commissioning and operation of the plant with its integrated magnet/liquefier PLC based control system is discussed. Cryogenic performance aspects, including thermo-syphon cooling, heat
4、load optimization, system cooldown, magnet fast discharge characteristics and liquefier operations, are detailed.INTRODUCTIONPerformance, reliability and efficiency are the critical aspects pertinent to any High Energy Physics cryogenic system. Emphasis may vary dependent on accelerator/detector siz
5、e, e.g. CERNs LHC or SLACs SCFF. However, they are all directly related to cost. Serious reductions in high energy physics funding in recent years, demand that prudence and sound judgement be exercised during all aspects of project design and construction to minimize cost, both constructional and op
6、erational. These criterions are often amplified by schedule constraints.The design, construction and installation of the SLAC/BaBar liquefier and superconducting solenoid magnet systems were subject to all of the above.LIQUEFIER OPERATIONBy July 1, 1999 the liquefier, which is a modified Linde TCF 2
7、00, had operated successfully for 7,000 hours. These include liquefier and magnet commissioning and two high energy physics runs. During this period, there has been no significant plant malfunction. Total lost physics time was 3 hours, which was attributable to operator error when he inadvertently c
8、licked on the “Disconnect Coldbox” button and stopped the plant, proving that PLC systems can not cater for everything. The only minor problem worth noting was with the I/P controller of a cryogenic valve pneumatic actuator, which went out of span and whose output could not be satisfactorily zeroed.
9、 Replacing the controller quickly solved the problem. Fortunately this did not happen during a magnet operational period.Judicious optimization of plant PLC parameters to satisfy all operating modes is nearly complete, allowing for fully automatic “Push Button” control in all cases. The operating mo
10、des are shown in Figure 1.During the first two coldbox/magnet cooldowns, cooling of the 4,000L LHe supply dewar was somewhat inefficient. This can be attributed to the fact that there is no vacuum insulation between the coldbox/dewar transfer line supply, return and vent channels, coupled with the s
11、hort circuit of the supply/vent lines at the dewar top. This is a tri-axial line, designed to minimize heat inleak to the magnet LHe supply during normal operation, with liquid withdrawn from the bottom of the dewar and all temperatures circa 4.5K. This situation was greatly improved during the thir
12、d cooldown by withdrawing gas from the dewar bottom via a second transfer line, which is normally used for transferring LHe to auxiliary dewars. Figure 2, shows the temperature/time history for the three cooldowns.SOLENOID MAGNET OPERATIONCooling of the coil conductor (Rutherford type cable imbedded
13、/co-extruded in a pure aluminum matrix) and its aluminum support cylinder, is intended to be by the thermo-siphon technique, i.e. convection cooling by liquid helium circulating from a header reservoir via aluminum cooling loops welded to the support cylinder. The system, including its pipings, valv
14、es and controls were designed accordingly, see Figure 3. The coldbox and 4,000L supply dewar can be cooled down and filled either before coil cooldown or simultaneously with the coil. Coil cooldown to 40K is done by forced flow of cold helium gas, via CV171/CV3 mixing with 290K gas via CV163/CV1, th
15、en to 7K via CV171/CV3 alone. Final cooldown to 4.5K and reservoir liquid helium fill are done via CV170 (coldbox liquid helium supply) and CV3. Reservoir LHe level is controlled via CV170. The gas returns via the reservoir and CV4, which is always open and (dependant on temperature) CV172 or CV173.
16、 CV2 and CV8, the thermo-siphon valve, are closed during the cooldown. When the coil/support cylinder temperatures have stabilized, CV2 and CV8 are opened and CV3 is closed, to establish thermo-siphon. However, establishing thermo-siphon was not always readily achievable. The fact that it can take u
17、p to an hour to confirm that thermo-siphon is not working, during which period the coil and support warm up by 0.2K, necessitating further temperature stabilization delays, decided us to operate with forced flow cooling. The LHe supply, coil pipings and valves are plumbed to accommodate both modes.
18、Fortunately the transfer heat loads via the co-axial, gas screened line3 (3 watt) and valve box pipings are low, minimizing the vapor fraction of the LHe supply.MAGNET FAST DISCHARGEThe coil is protected with the standard method of a resistor in parallel. If a quench is detected (50 mV unbalanced si
19、gnal between the two voltages in two layers), a breaker opens, closing the current in the coil and dump resistor. The peak voltage at the coil ends can be as high as 340V. However, as the center tap of the dump resistor is shorted to ground, then the maximum voltage to ground is 170 V.Fast discharge
20、 from the nominal current (4,600A) instigates a quench due the induced heating of the supporting cylinder (Quench Back). The coil temperature increases uniformly to 37K. In these conditions about 5 hours are needed to cool-down the coil again, fill the reservoir and be ready for re-charging. Typical
21、 current decay (from 4600 A) is shown in Figure 4. The voltage drops at the two layers are shown in Figure 5. The fast current decay at the beginning of the dump is due to the energy transferred from coil to cylinder.While commissioning the magnet, a problem developed with the coil/liquid helium res
22、ervoir pressure relief system. During fast discharge, the pressure increase is so rapid that the pneumatically actuated quench vent valve (V9) did not have time to fully open. As a consequence, on one occasion the bursting disc ruptured (6 bar), before the relief valve (5 bar) had time to open. The
23、problem was solved by disconnecting the command signal to V9 from the VME control system and hardwiring it directly to the breaker status signal. To increase its rate of opening, a 4 liter buffer volume was also added to the air supply immediately before the valves pneumatic actuator. If the breaker
24、 opens at currents less than 100A, V9 will not open. Irrespective of the magnet status, powered/unpowered, if the reservoir pressure exceeds 4 bar a software interlock also opensV9. In either case (breaker or pressure interlock), to limit helium loss, a timer closes V9 after seven seconds. These mod
25、ifications have been successfully tested.Current Lead HeatersDue to height restrictions in the experimental hall, the length of the current leads could not be optimized, resulting in condensation and ice. This has been effectively eliminated by the installation of four 0.5 inch diameter, 3 inch long
26、, 200 watt cartridge heaters mounted into each connecting bus bar. Originally powered at 110V these now operate at 48V for safety. Power is supplied via zero crossover SCRs, to reduce noise and controlled at 293K by programmable temperature controllers. Wiring the heaters via two 330K “Klixons” moun
27、ted directly on each bus bar provides over-temperature protection. Lead temperatures are also monitored, but not protected, by the PLC system with over/under temperature alarms set at 300K/280K respectively. PLC software over-temperature protection will be added during the next non-operational perio
28、d in October 1999.PLC CONTROL SYSTEMThe helium liquefier is controlled by an Allen Bradley, sixty slot, SLC500/04 PLC system programmed with RSLogix ladder logic. The ladder logic is displayed and modified on a dedicated Windows NT PC while it is actively controlling the liquefier. All PCs and SLC50
29、0 PLCs are linked by Allen Bradleys hardware and software Data Highway Plus (DH+) network. A separate SLC500 is used in our compressor room facility and is used for the Data Highway Plus I/O communication link with the liquefier PLC. The liquefier PLC controls virtually all compressor functions. Nat
30、ional Instruments BridgeView 2.0 program serves as the graphical user interface (GUI) to the PLC and runs on a dedicated 266 MHz dual Pentium II PC. BridgeView is essentially LabView with the ability to access tags from a PLC. The original BaBar solenoid control system as supplied by Ansaldo was a V
31、ME based system with an OS-9 operating system. This system, after being fully tested and commissioned, was replaced with RSLogix running on the liquefier SLC500 and a separate BridgeView program running on a dedicated 200 MHz Pentium II PC. VME control was changed because it was felt that two comple
32、tely different control systems was too difficult to operate, maintain and upgrade with changing operational and engineering staff over the coming decade. We also consider the RSLogix logic easier to understand and modify than the C programming in the VME system. The original magnet control system us
33、ed a combination HP 34401 A multimeter and HP 3852A data acquisition scanner module to record the analog signals. This hardware and signal monitoring was preserved in the new control system and is read into the magnet BridgeView system through a GPIB interface. BridgeView is also used to control the
34、 data acquisition process of the two HP units via the same GPIB bus. The magnet control system has ten hardwired interlocks that will initiate a fast discharge when tripped. These are also duplicated as software interlocks at lower settings that will cause the power supply to stop ramping up or to r
35、amp down. We are able to turn off the liquefier PC without affecting liquefier or magnet operation. However, we are unable to turn off the magnet PC due to magnet strain gage and certain magnet temperature software interlocks that come only through the GPIB bus and are not duplicated as direct PLC i
36、nputs. These software interlocks can be bypassed online in the ladder logic. We have experienced no mishaps in ladder logic modifications in online programming. With our magnet control system upgrade, the SLC500 control system now consists of 88 direct digital I/O and 114 analog I/O channels in addi
37、tion to the 80 analog signals coming through the GPIB board. The liquefier BridgeView program has 516 tags and the magnet BridgeView program has 162 tags. RSLogix scan time is 40 to 50 msec. BridgeView response time on both PC is in the two to ten second range. The standard X windows SLAC control pr
38、ogram (SCP) is used for ramping the power supply. A stand alone chassis is used for additional solenoid and bucking coil power supply monitoring. All analog and digital outputs and 24 VDC power are individually fused at the terminal blocks with fuses that are easily removable. This has proved to be
39、an absolutely invaluable feature during commissioning and during certain operations when absolute guarantee is required that a certain valve not change state or other action not occur. EPICS InterfaceThe Liquefier/Solenoid control system is also required to exchange information with the BaBar collab
40、oration Detector Control System4 (DCS). BaBar uses EPICS (Experimental Physics and Industrial Control Software) on VME CPU boards running the VxWorks real-time operating system for its DCS. This is now installed, performing well and providing operational support. The information exchanged between Ba
41、Bar on the Liquefier/Solenoid control system includes: Liquefier and Solenoid hardware interlock status, Solenoid and Bucking Coil power supply status (including current read back), operational state of pumps, compressors, and turbines that might affect BaBar data acquisition electronics (e.g. sourc
42、es of electronic noise and mechanical vibration), Solenoid temperature and BaBar status.A BaBar standard VME crate (21 slot, 6U, Motorola based CPU board running VxWorks and EPICS) is used to provide the interface between the Liquefier/Solenoid control system and BaBar status. BaBar standard interlo
43、ck modules are located in this crate to give BaBar direct readout of power supply interlocks. Communication between this crate and the Allen-Bradley SLC504 control system is implemented using an Allen-Bradley 6008-SV1R Scanner module in the VME crate and a 1747DCM in the SLC504 system. Module choice
44、 was based on the fact that EPICS already provided driver support for these. They also provide a bi-directional link between the BaBar Detector Control System and the Liquefier/Solenoid control system. Desired status information is passed between the control systems using the 8 input/output register
45、s of the 1747DCM.World Wide WebOne of the most beneficial diagnostic and operational aids was making all plant and magnet coil parameters accessible via the World Wide Web. System experts are able to monitor performance, “on-line” from home or office, as and when required 24 hours per day. Thus equi
46、pped, less experienced on-shift staff can be safely coached through any necessary intervention by telephone. This has greatly reduced on-site engineering manpower requirements. The disadvantage is that the whole world is able to monitor your progress, errors and problems. At the moment Web access pe
47、rmits monitoring only. However, interactive access and control, with “cast-iron” passwords of course, of some parameters is being contemplated.The web page can be found at: http:/cryocon2.slac.stanford.edu:8080/ACKNOWLEDGEMENTSWork supported by Department of Energy contract DE-AC03-76SF00515REFERENC
48、ES1.Burgess, W. et al. In: Proceedings of the Seventeenth International Cryogenic Engineering Conference, Bournemouth, UK, (1998) 149-1562.Fabbricatore, P. et al. In: IEEE Transactions on Magnets-12 (1996) 2210-2213.3.Blessing, H. et al. In: Advances in Cryogenic Engineering-27 (1982) 761-768.4.Abrahm, G. et al. In: paper titled The Detector Control System for BaBar presented at the Real-Time Conference RT99, Santa Fe, NM, June, 1999.我的大学爱情观1、什么是大学爱情：大学是一个相对宽松，时间自由，自己支配的环境，也正因为这样，培植爱情之花最肥沃的土地。大学生恋爱一直是大学校园的热门话题，恋爱和学业也就自然成为了大学生在校期间面对的两个主要问题。恋爱关系处理得好、正确，健康，可以成为学习和事业的催化剂，使人学习努力、成绩上升；恋爱关系处理
