| Miniproposals | ||||||||||
|
| Operators | |
| Session leader(s): | Ian Hutchinson |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Joe Daigle,Vinny Bertolino |
| Engineering Operator Run Comment |
| orthog controllers |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Entered: Jul 7 2004 03:49:11:350PM |
| Author: To Be Determined |
| Session Leader Summary 941130 SL Hutchinson, PO Wolfe, EO Daigle Finish of Installation of Orthogonal Controllers. Gradually making the transition to all shape controllers. The following were done: Exercise the orthogonal EF4 controller but took it back out because EF4 is mostly at full voltage demand all the time and so the orthogonal controller is not orthogonal but actully messes everything else up. Programmed some steps on RXL and ZXL (Shots 9 and 10) to demonstrate that they do the right thing. Tweaked the Ip controller during the ramp down to help CLEARIN to circularize the plasma after 1.0s. This worked only to a certain extent. Again we have problems with supply currents reaching limits (i.e. zero). Switched off all the coil current control in segment 2. (.1s on) by shot 28. Brought the full shape control on at the start of segment 2 (shot 32 on). Therefore, shots 32 and 33 were under full shape control. They seemed to work well. Most of the difficulties were associated with EF2 supplies railing during the first 0.5s or turning off on ramp down (which loses CLEARIN control). For settings see the PO log. Scorecard: Shot,Duration(s),Ipmax(MA) Shot,Duration(s),Ipmax 1 0.000 0.010 2 0.000 0.044 3 1.110 0.843 4 0.538 0.771 5 1.321 0.838 6 1.102 0.843 7 1.072 0.843 8 1.328 0.827 9 1.075 0.840 10 1.152 0.838 11 0.000 0.016 12 1.096 0.835 13 1.295 0.838 14 0.873 0.890 15 0.000 0.026 16 1.071 0.835 17 0.000 0.026 18 0.679 0.892 19 0.000 0.024 20 0.386 0.595 21 0.462 0.869 22 1.293 0.830 23 1.278 0.840 24 1.277 0.838 25 0.427 0.786 26 0.000 0.021 27 1.141 0.843 28 1.280 0.838 29 1.107 0.833 30 1.295 0.835 31 1.299 0.838 32 1.305 0.843 33 1.295 0.838 Finished summary for 941130 Total of 33 shots. 33 shots. 26 plasmas. 5 fizzles. 2 duds. Fizzle = gt 17 lt 50kA. |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:36:52:067PM |
| Author: To Be Determined |
| Physics Operator summary for 941130 Mini-proposal 065: Orthogonal controllers SL: Hutchinson PO: wolfe EO: Bertolino/Daigle The goal of today's run was to complete the installation and verification of the orthogonal controller set, and to explore their use to improve the SN discharge scenario. This was largely accomplished, although some additional tuning remains to be done to the discharge. All shots after #28 had the current control (IC_* wires) completely off except for IC_EF4 in segment 2, which begins at 0.1 sec. Only orthogonal shape controllers are used in those shots 28-33 (except EF4). The transition to a diverted discharge was moved earlier, to 0.34 sec on all shots after 28. Orthogonality of the RXL and ZXL controllers was tested by inserting 100 msec steps on each on shots 9 and 10. These looked quite good; the step amplitude was observed and the response time was consistent with the advertised rate on the Physics PID display. There was some crosstalk between the two at the mm level. There was also an interaction of ZXL with Ip, but this was small. Getting RXU and ZXU to work correctly without railing the EF2 supplies was a struggle. This was sort of accomplished by the end of the day. The present programming of RXU=-.04 and ZXU=+0.02 seems to give a reasonably good single null; A slightly more positive ZXU might be better, but may require too much negative current in EF2U with these controllers. There also seems to be some interaction between ZXU and CLEARIN. On shot 31 the gain was doubled on ZXU, which resulted in about a 1cm change in the value of ZXU, and a few mm increase in SSEP, the proximity of the secondary separatrix; however, the inner gap, which had been nicely controlled at about 1.5cm on shot 30 began to sag during the shot to about 1.2cm. The present setting of CLEARIN=-0.005 seems to correspond to about a 1.2 to 1.5cm inner gap, with some interaction with ZXU as noted above. An attempt to increase the gap by setting CLEARIN=0. was made on shots 12-14. Shot 14 had a serious vertical instability problem, so I set the CLEARIN back down to -.005. This seemed to be another case of an interaction with ZXU, which was being varied on those shots. The final set of shots had some significant moly injections. After the fact, it was noted that the earlier diversion was causing the x-point to transit the nose on the way to its final location, and one significant moly burst coincided with the time the x-point was on the nose. Also some later bursts may have been due to the strike point going to the V at the bottom of the inner divertor. It seems likely that the early evolution of RXL should be modified to move the x-point further out between 0.2 and 0.5 sec; at present it is actually set up to be slightly inboard of its final location during that interval. It might also be useful to raise the ZXL by 5 mm to get the strike off the bottom of the inner divertor. Some progress was made in understanding the problems with rampdown disruptions. Reducing the OH2 component of the Ip controller in the ramp-down phase, as had been done last June, seemed to help some. The reductions were done in several steps from shot 30 to 33, finally reducing the OH2 gains by about a factor of 2. It was noted that the EF2 supplies go to zero current at about 1.15 sec, which explains why CLEARIN is ineffective in the later rampdown. Also, it looks like EF3 is going to zero before the plasma current gets all the way down, which will screw up the RCUR control some. A vertical instability was observed during the ZCUR drop on shot 28; this was "cured" on shot 29 by slowing down the ZCUR drop to 200msec, but earlier shots had had no problem with a 100msec ramp. The problem seemed to be that the plasma was going diverted (nearly DN) almost immediately after the start of the ramp, and the elongation got to nearly 1.8. Problems: --------- Shot 4 was killed by overgassing when NL_04 dropped out during the pulse. There were also power supply problems on this shot (post-disruption) and when the engineers went out to investigate they smelled smoke from the cell. On further investigation, it was discovered that the smoke was coming from the motor of an exhaust blower in the cell. The smoke was also thick inside the TCI enclosure, which had led to the interferometer drop-out. We had more problems than yesterday with fizzles and dirty plasma following disruptions. The disruption on shot 14 put the RGA off the air. After the run it was discovered that the connector had been knocked off. This was replaced by Bob Childs and the RGA now seems ok. Recommendations: ---------------- Now that all current controllers in SEGMENT 2 are off (except for EF4) there are a bunch of newly freed wires. If density control were moved from wire 9 to wire 14, we could set the block of 9-12 to IP reference, allowing installation of additional shape controllers. This would be very convenient for future development, since we could switch in control on things like STRIKE, KAPPA, CLEAROUT, etc., as needed during the segment. We are not presently using the integral control on the NL_07 control on wire 9, but it has been used in the past (?); therefore it would be advisable to have Horne hook up an integrator to wire 14 if we make this switch. As Ian has been pointing out, the problem of asking for too much current from the EF2 supplies would be ameliorated if we could get more positive current from EF4 faster. This would require getting more voltage, perhaps by adding one or more additional TMX supplies in series with the present reverse polarity set. Scorecard: (see the logbook for blow-by-blow descriptions) ---------- Tests 1 (OH instrumentation test on shot 1) Plasmas 26 Fizzles 5 Duds 1 ------------------ Total 33 |
| Session Leader Comments |
| Physics Operator Comments | |||
| Nov 30 1994 08:21:06:870AM | Steve Wolfe | MP #065 - Otho. Controllers
SL-Hutch PO-Wolfe EO-Bertolino/Daigle Start from 941129025 PLan is to adjust the ZXU programming to try for nominal shape. Note: EF2U was railed on 941129025 yesterday, not on 24; hopefully nominal will be in between. Once RXU and ZXU are satisfactory, play with shape parameters to test orthogonality. ECDC on at 5AM, off at 7:30 | |
| Nov 30 1994 08:29:44:640AM | 941130001 | Steve Wolfe | Shot #941130001
OH only test (will be routine for first run of the week) Test ran ok (10kA Oh1 and Oh2's) |
| Nov 30 1994 08:41:51:170AM | 941130002 | Steve Wolfe | Shot #941130002
Set up from shot 941129025. Fill = 1.3e-5 Torr Fizzle. |
| Nov 30 1994 08:50:10:050AM | 941130002 | Steve Wolfe | Shot #941130002
Set up from shot 941129025. Fill = 1.3e-5 Torr Fizzle. |
| Nov 30 1994 08:57:35:950AM | 941130003 | Steve Wolfe | Shot #941130003
Raise fill to 1.4e-5 and try again Plasma - disrupts on rampdown at 600kA |
| Nov 30 1994 09:39:50:470AM | 941130004 | Steve Wolfe | Shot #941130004
Lower ZXU program from 0.1 to 0.06 Disrupts early. PF fault at 0.53 sec NL_04 dropped out and the density feedback overgassed the plasma. Smoke in the cell was found to be due to a motor in an exhaust fan. The TCI laser enclosure was filled with smoke. |
| Nov 30 1994 09:59:04:830AM | 941130005 | Steve Wolfe | Shot #941130005
Repeat of shot 4, hopefully without the smoke. Good plasma EF2s did not rail (max around 3500 amps each) ZXU error is now -30000 for most of shot |
| Nov 30 1994 10:12:42:030AM | 941130006 | Steve Wolfe | Shot #941130006
Raise P gain on ZXU to 2 at .35, 8 at 0.5sec (hybrid units) Disrupted during rampdown (1.1 sec, 640 kA) ZXU error down to -20000. |
| Nov 30 1994 10:22:33:920AM | 941130007 | Steve Wolfe | Shot #941130007
Increase gain for ZXU to 9.5, 0.0 during rampdown Program ZXU to stay at 0.06 until 1 sec. Similar. disrupts on ramp-down. ZXU now at -15000. |
| Nov 30 1994 10:51:57:150AM | 941130008 | Steve Wolfe | Shot #941130008
Argon this shot. Put in IC_EF4 contihh controller Had to fix a bug in XCONTIHH.PRO (variable NIP not defined) Ramped down ok. Less EF2U current and more EF4 current - maybe it did what it should RXU seems less well controlled early. EF4 was maxed out before 0.7 seconds. Full blown vertical instability at the very end, several cycles. |
| Nov 30 1994 11:11:47:670AM | 941130009 | Steve Wolfe | Shot #941130009
Trying to put back the old controller in EF4. READ_FROM is broken; USed load_wire Put a 1.5cm step in RXL from 0.7 to 0.8sec Back to the disruption on ramp-down. It seemed to get the step, gets to zero error in about 60msec, Physics P gain is 36. Got a 1.5cmm excursion, as advertised. |
| Nov 30 1994 11:32:24:920AM | 941130010 | Steve Wolfe | Shot #941130010
Raise gain on NL from 5 to 8. Take out RXL step , put in ZXL step up by 1cm from .75 to .85 Still disrupting on ramp-down. I need to try to deal with that. Got the ZXL step, along with a small perturbation on RXL The ZXL crosstalk from the RXL step was a few mm (EFIT) The RXL crosstalk from the ZXL step is about 2mm (EFIT) H-mode started earlier |
| Nov 30 1994 11:48:39:240AM | 941130011 | Steve Wolfe | Shot #941130011
Take out ZXL step. Change clearin to stay at 0 (instead of -.005). Extend till 0.9sec Dud.? Br looks a little positive, but it did earlier. |
| Nov 30 1994 12:00:50:070PM | 941130012 | Steve Wolfe | Shot #941130012
Try again, I can't see anything wrong. Take out ZXL step. Change clearin to stay at 0 (instead of -.005). Extend till 0.9sec Shot ran and disrupted on rampdown. Seemed to get a little more clearin? |
| Nov 30 1994 12:17:01:840PM | 941130013 | Steve Wolfe | Shot #941130013
Move step in ZCUR and ZXL to 0.5 sec. Shot ran. EF2U maxed out for the whole flattop. |
| Nov 30 1994 12:37:41:350PM | 941130014 | Steve Wolfe | Shot #941130014
Change programming on ZXU to +0.04 (from 0.06) Disruption near end of flattop (full current) Vertical instability - it went up. ZXU error still around -15000 to -20000 EF2U still maxed out. I don't understand this. |
| Nov 30 1994 12:48:15:200PM | 941130015 | Steve Wolfe | Shot #941130015
Change CLEARIN back to -.005 fizzle - probably just dirt from the disruption |
| Nov 30 1994 01:01:09:290PM | 941130016 | Steve Wolfe | Shot #941130016
Try again with CLEARIN at -.005 Ran but shaky. Disrupted at start of ramp-down. This time the EF2s did not rail? Multiple injections |
| Nov 30 1994 01:13:16:410PM | 941130017 | Steve Wolfe | Shot #941130017
Increase RXU gain from .64 to 1.0 Fizzle - probably dirt. |
| Nov 30 1994 01:28:49:240PM | 941130018 | Steve Wolfe | Shot #941130018
Try again with Increase RXU gain from .64 to 1.0 Disruption in flattop. EF2U railed again. |
| Nov 30 1994 01:36:24:570PM | 941130019 | Steve Wolfe | Shot #941130019
Drop ZXU program to 0.02 (from 0.04) Fizzle |
| Nov 30 1994 01:46:24:930PM | 941130020 | Steve Wolfe | Shot #941130020
try again with ZXU at +0.02 Short and ugly. |
| Nov 30 1994 02:09:38:430PM | 941130021 | Steve Wolfe | Shot #941130021
try again with ZXU at +0.02 Lasted longer still disrupted. |
| Nov 30 1994 02:21:37:380PM | 941130022 | Steve Wolfe | Shot #941130022
try again with ZXU at +0.02 This one ran. |
| Nov 30 1994 02:37:27:740PM | 941130023 | Steve Wolfe | Shot #941130023
Moved RCUR in between 200 and 300msec Changed clearin to turn on at 100msec Went all the way through Seemed to work about as expected, except the EF2U almost rails |
| Nov 30 1994 02:49:41:220PM | 941130024 | Steve Wolfe | Shot #941130024
Moved control on ZXL to 0.1 sec That worked almost all the way through |
| Nov 30 1994 03:15:37:590PM | 941130025 | Steve Wolfe | Shot #941130025
Turn control on IC_EF1L and IC_EF2L off (moved back to < 100msec) Renamed wire 10 to IC_OH2_off Change ZXL program to start at +0.03, ramp down to -0.01 at 0.4 sec Disrupted at 400msec. |
| Nov 30 1994 03:26:07:630PM | 941130026 | Steve Wolfe | Shot #941130026
Turn control down by a factor of 4 on IC_EF1U and IC_EF2U off Fizzle |
| Nov 30 1994 03:40:08:040PM | 941130027 | Steve Wolfe | Shot #941130027
Try again with control down by a factor of 4 on IC_EF1U and IC_EF2U relative to 25. Disrupted on rampdown. Big thrash in vertical position at .38 to .5 EF2L rails from .33 to .48 Multiple injections |
| Nov 30 1994 03:53:39:430PM | 941130028 | Steve Wolfe | Shot #941130028
Turn the EF2U amd EF1U control off completely Made it through all the way. EF2s did not rail. Still have a vertical thrash. |
| Nov 30 1994 04:08:58:960PM | 941130029 | Steve Wolfe | Shot #941130029
Program IC_EF4 to 2000A earlier Move start of ZCUR ramp from 0.4 to 0.3 Disrupted on rampdown |
| Nov 30 1994 04:17:23:910PM | 941130030 | Steve Wolfe | Shot #941130030
Put in a controller for Ip with 10% less on OH2s during rampdown. Made it all the way through. |
| Nov 30 1994 04:38:00:190PM | 941130031 | Steve Wolfe | Shot #941130031
Extend integrator on slow ZCUR from 1.0 sec to 1.8sec Drop the OH2s in the IP rampdown controller by another 10% Increase factor on ZXU and change controller time constant to get a factor of 2 in gain. Ran ok - thrash at .52 sec ZXU error down by factor of 2 Brief Hmode |
| Nov 30 1994 04:49:54:510PM | 941130032 | Steve Wolfe | Shot #941130032
Moved the ZXU and RXU control to turn on at .1 sec Went through, H-mode is better. Note - EF2s go to zero at 1.15 sec, which kills clearin control on rampdown |
| Nov 30 1994 05:24:35:280PM | 941130033 | Steve Wolfe | Shot #941130033
Drop the OH2 on the IP rampdown controller another 20%. Raise density from .6 to .8e19. Runs ok, no better on rampdown but no worse. Starts dithering around .6sec and drops into H-mode at .85. Note moly burst early coincides with x-point passing the nose. Later bursts may be striking near the crotch of the inner divertor. Inner gap starts at 1.5cm, but sags to 1.2 by 1 second. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 08:26:26:500AM | Test | Ok | oh coax test |
| 2 | 08:39:35:280AM | Plasma | Ok | no faults |
| 3 | 08:51:12:690AM | Plasma | Ok | no faults |
| 4 | 09:07:48:480AM | Plasma | Ok | disruption induced comm faults, blower motor in cell smoking- removed fr |
| 5 | 09:40:57:990AM | Plasma | Ok | no faults |
| 6 | 10:03:05:240AM | Plasma | Ok | no faults |
| 7 | 10:17:28:740AM | Plasma | Ok | no faults |
| 8 | 10:39:18:250AM | Plasma | Ok | no faults |
| 9 | 11:03:19:990AM | Plasma | Ok | no faults |
| 10 | 11:22:03:450AM | Plasma | Ok | no faults |
| 11 | 11:37:45:050AM | Plasma | Ok | no faults |
| 12 | 11:51:12:930AM | Plasma | Ok | no faults |
| 13 | 12:06:49:820PM | Plasma | Ok | no faults |
| 14 | 12:19:50:410PM | Plasma | Ok | no faults |
| 15 | 12:40:08:830PM | Plasma | Ok | No Faults. |
| 16 | 12:54:16:880PM | Plasma | Ok | No Faults. |
| 17 | 01:06:31:490PM | Plasma | Ok | No Faults. |
| 18 | 01:16:46:110PM | Plasma | Ok | No faults |
| 19 | 01:31:51:160PM | Plasma | Ok | No faults |
| 20 | 01:41:07:350PM | Plasma | Ok | No faults |
| 21 | 01:53:16:840PM | Plasma | Ok | No Faults. |
| 22 | 02:11:26:810PM | Plasma | Ok | No faults |
| 23 | 02:27:27:100PM | Plasma | Ok | No faults. |
| 24 | 02:42:00:600PM | Plasma | Ok | OH2 L comm fault |
| 25 | 03:02:58:200PM | Plasma | Ok | No Faults. |
| 26 | 03:19:14:500PM | Plasma | Ok | No Faults. |
| 27 | 03:32:46:500PM | Plasma | Ok | Comm fault OH1 EF1U |
| 28 | 03:44:32:030PM | Plasma | Ok | No faults |
| 29 | 03:58:50:100PM | Plasma | Ok | No Faults. |
| 30 | 04:12:39:770PM | Plasma | Ok | No Faults |
| 31 | 04:28:47:170PM | Plasma | Ok | |
| 32 | 04:42:23:250PM | Plasma | Ok | No faults. |
| 33 | 04:57:27:480PM | Plasma | Ok | No faults |