| Miniproposals | ||||||||||
|
| Operators | |
| Session leader(s): | Steve Golovato |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Frank Silva,Vinny Bertolino |
| Engineering Operator Run Comment |
| 7.9t wolfe |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Entered: Jul 7 2004 03:49:15:040PM |
| Author: To Be Determined |
Run summary - 960131 SL: Golovato PO: Wolfe EO: Silva/Bertolino Run 960131 - He3 minority heating at 8 T with bori=onized walls. Run Plans: Start with 1MA shot similar to 960117003 (but with better RF tuning) Density scan at 1MA (one higher, one lower). Concentration scan at 1MA (one higher, one lower) Choose best condition, repeat shots for profile. Raise current to 1.2MA at best condition, repeat shots for profiles. Go back to 1MA and try PEP modes (Li pellets) This was a successsful run although we only got 15 shots because of slow cooling at 8T and didn't get to the PEP mode shots. We re-established the conditons of the last boronized 8T run (960117) on shot 3 and added position feedback on shot 4. Shots 5 ands 6 were at lower and higher density and both dithered. The higher density shot on 960117 went into H-mode at 2.5MW while we had only 2MW today. The lower density may be near the low density limit and requires more power to transition into H-mode. We settled on the intermediate density of nl04(programmed)=1.e20 and did a concentration scan. We did a higher concentration for shots 7 and 8, getting dithering and off-axis electron heating on shot 8. This is predicted by theory from direct electron absorption from the mode converted Bernstein wave at high He3 fraction. It was also seen on shot 960117012. We then went down in concentration for shots 9 and 10. Shot 11 had strike point sweeping in the divertor to avoid a hot spot on the outer diveror which is visible in the TV view. It didn't help noticably. This seemed to be the best condition so shots 12-14 were repeats to get a hirex Ti profile. Shot 15 was a 1.2MA shot at the same conditions. The H-modes were shorter with less density rise but the higher current led to higher stored energy, 136 kj compared to 125 on shot 11. Under the best conditions, the H-modes last less than 200ms and tend to be ended by some kind of mini-injection which is visible on the 2pi bolometer and rf loading, but not the moly monitor. The H-factors are about 1.2 assuming 100% absorption of the rf power. If the heating efficiency is less for He3 minority heating because of low single-pass absorption, the H-factor would be higher. For these shots, an efficiency of 60% would raise the H-factor to a more respectable 1.8. It would be good to use change in slope of the diamagnetic stored energy signal to get a handle on the power absorbed. This should be done during the next 8T run and requires a no-plasma baseline shot. The next 8T run should also investigate PEP-mode (finally!) and attempt higher rf power since so many shots today were dithering. For information on engineering problems and shot statistics, see the physics operator summary. For shot-to-shot comments, see the logbook under the session_leader topic. |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:36:55:957PM |
| Author: To Be Determined |
| Physics Operator Summary for 960131 MP# 117 D(He3) Heating at 7.9T SL: Glovato PO: wolfe EO: Silva/Bertolino Run plan: ---------- This run is a continuation of 960117 and used the same equlibrium programming. 1.Run 1MA shots similar to 960117003 (but with better RF tuning) 2.Concentration scan at 1MA (one higher, one lower) 3.Density scan at 1MA (one higher, one lower). 4.Choose best condition, repeat shots for profile. 5.Raise current to 1.2MA at best condition, repeat shots for profiles. 6.Go back to 1MA and try PEP modes (Li pellets) Engineering setup: ------------------ No Discharge Cleaning at all (no overnight, no pre-run) Fill A-side (really K-side) to 20psi D2 Fill B-top with 5psi D2 Fill B-side lower with 4psi Argon B-side upper filled with He3 (as at present) Fiducial setup as on 960130002: A-side and B-top enabled. Do not enable B-side lower (Ar) Begin experimental program with setup from 960117003 EXCEPT TF current set to 224KA TF on PLC control Enable A-side,B-top,B-side upper (He3), B-side lower (Ar) Results: ------- This was a pretty good run, except that the number of shots was low because of long cooling times. We completed the first four items on the list and got one good discharge at 1.2 MA. No pellet shots were attempted today. We ended up with 15 shots, including the fiducial on shot 1. Operationally, the machine performed well, except for the long cooling times, over 30minutes on average. The RF performed well all day at 2.0 MW, and on the last shot D-port got up to 900kW, which may indicate its arcing problem isn't as bad as previously thought. One problem with 8Tesla runs is that one tends to be very conservative about making changes once things are working, because the consequences of losing a shot are such a high fraction of the run. Following a successful fiducial, we loaded a revised 960117003, with changes to the RCUR offset and programming along the lines recommended by Ian yesterday. It disrupted at .5 sec and I wimped out and went back to the 960117 programming for shot 3, which went fine. The rest of the day used this programming with only minor changes. We then tried a density "scan" on shots 4,5,6 (NL_04 1.0,0.8,1.3e20), attempting to keep the He3 concentration constant by adjusting the He3 pulse width roughly in proportion. Actually the concentration varied during these, and the density we got didn't quite match what we asked for, but we declared 1.0 to be the target point anyway. It is worth noting that the low density point had an injection early in the RF pulse, which may have prevented it from going into a good H-mode. This point would be worth revisiting in a later run. We then did a concentration scan on #8,9,10, with the lowest concentration being selected. This had a long ELM-free period terminated (possibly) by an injection. Shot 11 was like shot 10 except that a zigzag sweep was programmed in ZXL to try to reduce a hot spot, which has been apparent on the video, identified as being at J-port outer divertor. This may have helped a little. This shot was selected for repeating for profiles. There was a very long delay (over 1:20) between 11 and 12 as the cooling had to be suspended while the LN2 tank was filled. Shot 12, nominally a repeat, was an ELMy H-mode after the E-port transmitter came on late. Shot 13 was closer, but dropped out of H-mode sooner. Shot 14 was similar. Shot 15 was at 1.2MA and ran successfully and got somewhat higher stored energy (136kJ). The EF3 supply apparently encountered a current limit at 10.8kA, although it was supposed to be set for 15kA. This resulted in a 7mm RCUR excursion and probably contributed to the slow current rise to 1.2MA. Vinnie will check into this, and hopefully fix it, tomorrow. See the Logbook under PHYSICS_OPERATOR for shot by shot records. Other problems, lessons learned: -------------------------------- Apparently some shots, including 960130040, have a segment 2 named "Fiducial Plasma, Lower X-pt" even though they aren't fiducials; this one is a 1.1MA case. Having noted this, I also got JAS to fix the bug that caused PCS to display titlebars with incorrect segment names on the ALL_P and ALL_V screens. There was no ECE on the fiducial shot 1 because a valve was closed on the ECE beamline. This was opened on shot 2. Should be checked before first shot on the ECE PLC screen. A number of spectroscopy batch jobs for the first 3 shots on the ZENITH queue were held up because of a number cruncher left running at interactive priority. PHYSOP can quite easily handle all the scopes normally used by the Physics Operator. I ran with CMODX9 logged in to PHYSOP and had good response with a total of six scopes, a traverser, PCS, an IDL session, LOGBOOK, and a few other processes. The only thing I ran on an AXP (Barolo) was the EFIT movie. It seems that logging into an already-burdened AXP from the xterm CMODX9 is not a good use of resources. Scorecard: ---------- Shots 15 Plasmas 15 Duds 0 Fizzles 0 Tests 0 We are now up to 54 straight plasmas (not counting intentional tests) this week. There were two disruptions: Shots 2 and 7; neither produced a warning, although 7 was close. Shot 7 had a brick (large impurity injection), while 2 had bad position and eventually went vertical. |
| Session Leader Comments | |||
| Jan 31 1996 02:09:10:740PM | Steve Golovato | run plans - 960131
Start with 1MA shot similar to 960117003 (but with better RF tuning) Density scan at 1MA (one higher, one lower). Concentration scan at 1MA (one higher, one lower) Choose best condition, repeat shots for profile. Raise current to 1.2MA at best condition, repeat shots for profiles. Go back to 1MA and try PEP modes (Li pellets) | |
| Jan 31 1996 08:52:29:400AM | 960131001 | Steve Golovato | shot 960131001 - fiducial.
e-port only, 1.1 MW. good h-mode. |
| Jan 31 1996 09:05:49:160AM | 960131002 | Steve Golovato | shot 960131002 - reload 960117003, 8T, 50ms He3 puff, nl04=1.e20.
disrupted at 460 ms. |
| Jan 31 1996 09:31:09:400AM | 960131003 | Steve Golovato | shot 960131003 - move He3 puff earlier, now .6-.65s at 90v, .65-1.1s
at 30v. load rf pulses form shot 960130040, rf starts at about .68s. plasma, h-mode. d-port ran, e-port had some trips. 129kj, hfac=1.2 |
| Jan 31 1996 10:04:34:100AM | 960131004 | Steve Golovato | shot 960131004 - repeat, feedback on, e-port tuned a little.
plasma. h-mode started later. 125kj. |
| Jan 31 1996 10:26:51:870AM | 960131005 | Steve Golovato | shot 960131005 - nl04=.8, He3 puff 40ms (same ratio, lower density)
plasma, rf ran, no h-mode. dithering. brick at 790ms. |
| Jan 31 1996 11:17:13:990AM | 960131006 | Steve Golovato | shot 960131006 - nl04=1.3, He3 puff 65ms (same ratio, lower density)
plasma, dithering, may need more power to get h-mode at this density. |
| Jan 31 1996 11:40:01:510AM | 960131007 | Steve Golovato | shot 960131007 - nl04=1., He3 puff 70ms, concentration scan.
last three shots had Prad~20% in lmode. plasma, disruption at 800ms, brick hit plasma. |
| Jan 31 1996 12:14:53:310PM | 960131008 | Steve Golovato | shot 960131008 - repeat.
plasma. no h-mode. dithering. off-axis electron heating seen on ece. |
| Jan 31 1996 01:06:39:740PM | 960131009 | Steve Golovato | shot 960131009 - He3 puff to 30ms.
plasma, h-mode ends at some kind of injection, like shot 4. |
| Jan 31 1996 01:21:04:330PM | 960131010 | Steve Golovato | shot 960131010 - He3 puff to 20ms.
plasma, h-mode again, again lost at emmision from somewhere. |
| Jan 31 1996 01:52:58:050PM | 960131011 | Steve Golovato | shot 960131011 - repeat with strike point sweeping to avoid emmision.
plasma, similar to last shot. went into h-mode sooner, stayed in a little longer, 180ms, before blip (no pun intended) on 2pi which leads to end of h-mode as in previous shots. there is a clear hot spot on the outer divertor which may lead to our injection problems. this appears to be best condition, we will do 2 more shots here to get Ti profile. |
| Jan 31 1996 03:26:45:740PM | 960131012 | Steve Golovato | shot 960131012 - repeat for Ti profile.
plasma, h-mode when e-port came on late. brick at .67sec. |
| Jan 31 1996 03:56:24:640PM | 960131013 | Steve Golovato | shot 960131013 - repeat for Ti profile.
plasma. h-mode lost after mini-injection seen on 2pi, not on moly monitor. |
| Jan 31 1996 04:09:51:820PM | 960131014 | Steve Golovato | shot 960131014 - last shot for Ti profile.
plasma. much like last shot. |
| Jan 31 1996 04:48:16:360PM | 960131015 | Steve Golovato | shot 960131015 - go to 1.2 MA, all other conditions the same.
plasma. 1.2 MA at 850ms. elmy h-modes. |
| Physics Operator Comments | |||
| Jan 31 1996 06:24:01:590AM | Steve Wolfe | Run plan for Wed. 960131
MP 117 D(He3) heating at 7.9T SL: Golovato PO: wolfe Engineering setup: ------------------ No Discharge Cleaning at all (no overnight, no pre-run) Fill A-side (really K-side) to 20psi D2 Fill B-top with 5psi D2 Fill B-side lower with 4psi Argon B-side upper filled with He3 (as at present) Fiducial setup as on 960130002: A-side and B-top enabled. Do not enable B-side lower (Ar) Begin experimental program with setup from 960117003 EXCEPT TF current set to 224KA TF on PLC control Enable A-side,B-top,B-side upper (He3), B-side lower (Ar) Note: This run will be at high field (7.9Tesla). The TF scanner must be operational on all shots, and cooling should be such as to keep the TF core temperature from ratcheting up. Run plan: ---------- Run 1MA shots similar to 960117003 (but with better RF tuning) Concentration scan at 1MA (one higher, one lower) Density scan at 1MA (one higher, one lower). Choose best condition, repeat shots for profile. Raise current to 1.2MA at best condition, repeat shots for profiles. Go back to 1MA and try PEP modes (Li pellets) | |
| Jan 31 1996 06:51:32:290AM | Steve Wolfe | Load 960130002 for fiducial. Load at 31-Jan-1996 06:07:15.00
Coldstart done 31-JAN-1996 06:09:18.66 LOAD from 960117003 into PCS. DO NOT LOAD_INTO_MODEL until after fiducial. | |
| Jan 31 1996 07:49:17:350AM | Steve Wolfe | Ian's recommendation from yesterday is to use RCUR offset=3000
for segment 2. IMPORT_SEG from 960130040 into seg 3 and stare at it. For some reason this segment has title "Fiducial Plasma, Lower X-pt" although in fact it is a 1.1MA shot! There is probably a bug in PCS about when the title gets stored? Change it now, since I don't want to be confused. Note that the title bars on the All_p's screen seem to reflect the original titles when PCS got brought up: Seg 2 says "Fiducial Plasma, Lower X-pt" and segment 3 says "Diverted Plasma, Upper X-pt". Send Mail to JAS. The original trace from Granetz run has an initial RCUR of .655 and moves out to 0.666133 with some funny jogs. Ian's trace from #40 (a 1.1MA shot) is also at .655 before segment switch, jumps to 0.644880 at .1sec, then moves out to 0.660000. As advertised, he turns on a third line with offset=3000 at 0.1 sec (at segment switch). He obviously evolved this from his end-of-shot sequence, because he still has the offset=8000 row there, turning on at -2.5 sec and off at 0.1sec. Note that the Granetz seg 1 has RCUR at 0.1sec of 0.653180 with an offset of 10000. TO make the change, I need to make a 5mm inward shift at .1sec, then increase that to 1cm at .5sec, when we get to 1MA. I think I'll see how the Granetz prescription goes first. | |
| Jan 31 1996 08:05:54:540AM | Steve Wolfe | Changed my mind - go ahead and make the change: cut in the RCUR trace
from shot 960130040 and change the offset from 8000 to 3000. None of this will get loaded until after a successful fiducial. | |
| Jan 31 1996 08:30:59:860AM | 960131001 | Steve Wolfe | Shot 960131001 Fiducial setup (from 960130002) - Successful
|
| Jan 31 1996 09:03:08:860AM | 960131002 | Steve Wolfe | Shot 960131002
Load from 960117003, with changes to RCUR as noted. (offset to 3000) Note that Molymonitor analysis did not run on #1,but data is there. Disrupted early at 0.5sec |
| Jan 31 1996 09:19:07:290AM | 960131003 | Steve Wolfe | Shot 960131003
MOve He3 puff to 0.6 extend thru 1.1sec. Reload RCUR wire from 960117003. Maybe that little jog was a good thing? Note- all the spectroscopy batch jobs seem to be hung up on Zenith. Killed YWANG's IDL job. Shot ran ok. Disrupts at 1.47. RF gets 2MW. H-mode at .78. 127kJ. |
| Jan 31 1996 09:56:41:430AM | 960131004 | Steve Wolfe | Shot 960131004
Decrease RFload feedback weight to -1500. Feedback will be on this shot. Ran to 1.45sec. H-mode at .9, dithering earlier. RF 2MW from .72, 125kJ. |
| Jan 31 1996 10:30:49:820AM | 960131005 | Steve Wolfe | Shot 960131005
NL04 to 0.8e20 He3 (PG2) initial puff to 40msec (from 50) Steve decided to do the density "scan" first Plasma - Dithering, no real H-mode, NL04 goes to 1.2. RF stays on. 85kJ. INjection at 780ms. |
| Jan 31 1996 11:07:49:630AM | 960131006 | Steve Wolfe | Shot 960131006
NL04 to 1.3e20 He3 (PG2) initial puff to 65msec Put in a rampdown jog on RCUR at 1.3sec, change offset(1.3) to 3000. Plasma - dithering, no H-mode. RF stays on. No obvious improvement in rampdown. |
| Jan 31 1996 11:38:20:870AM | 960131007 | Steve Wolfe | Shot 960131007
NL04 back to 1.0e20 He3 (PG2) initial puff to 70 msec; start conc. scan Disruption at .77sec -- Brick! It came from the outer divertor; the video caught it in the act. |
| Jan 31 1996 12:06:15:450PM | 960131008 | Steve Wolfe | Shot 960131008
Repeat: NL04 back to 1.0e20 He3 (PG2) initial puff to 70 msec; start conc. scan Plasma ramps all the way down. RF stays on. Dithering the whole RF pulse. 80kJ. |
| Jan 31 1996 01:02:52:560PM | 960131009 | Steve Wolfe | Shot 960131009
Repeat: NL04 back to 1.0e20 He3 (PG2) initial puff to 30 msec (low conc. point) Plasma - ELM-free H-mode .74 to .90, then drops out. RF stays on. 120kJ Maybe a little injection drops it out of H-mode? What I've been calling dithers appear to be a sinusoidal oscillation on H-alpha, around 250Hz. |
| Jan 31 1996 01:13:43:500PM | 960131010 | Steve Wolfe | Shot 960131010
Repeat: NL04 back to 1.0e20 He3 (PG2) initial puff to 20 msec (lower conc. point) Plasma - Multiple ELM-free H-modes. Also some injections? RF stays on 125kJ. |
| Jan 31 1996 01:50:45:030PM | 960131011 | Steve Wolfe | Shot 960131011
Put in a +/- .005 zigzag on RXL to try to reduce hotspot on outer divertor Plasma - Similar to previous case. Still se hot spot on video, I'm not sure if its moving. Steve thinks this best case. Do profile here (2 repeats). |
| Jan 31 1996 03:13:16:810PM | 960131012 | Steve Wolfe | Shot 960131012
Repeat. Long delay (1:2) for LN2 tank refilling before shot 12. Plasma - Multiple injections, good ELMy H-mode after .9 when RF goes to full power. 120kJ |
| Jan 31 1996 03:50:43:370PM | 960131013 | Steve Wolfe | Shot 960131013
Repeat. Plasma - More like #11 but injection(?) terminates H-mode earlier |
| Jan 31 1996 04:09:22:590PM | 960131014 | Steve Wolfe | Shot 960131014
Another repeat. Plasma - Multiple H-modes, pretty similar to 11 and 13. 116kJ |
| Jan 31 1996 04:49:18:110PM | 960131015 | Steve Wolfe | Shot 960131015
Go to 1.2MA at .65sec. Plasma runs all the way through. Didn't get to 1.2 until around .8sec. Similar H-modes to the lower currents. 136kJ. |
| Jan 31 1996 05:16:00:000PM | 960131015 | Steve Wolfe | Shot 960131015
Go to 1.2MA at .65sec. Plasma runs all the way through. Didn't get to 1.2 until around .8sec. Similar H-modes to the lower currents. 136kJ. The RCUR error has a large excursion (around 7mm) from .6 to .85, during the ramp up to 1.2MA. It looks to me as if EF3 is on a current rail at that time, as the voltage does not follow its demand; this is peculiar as the current is at 10.8kA and the limit was supposed to be at 15kA. Vinnie agrees that it looks like a current rail at 10.8, and suggests that the PLC may not have been changed from shot 1, which had a 12kA limit; this would imply "only" a 10-15% calibration error instead of 40-50%. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 08:29:03:610AM | Plasma | Ok | fiducial |
| 2 | 08:48:16:220AM | Plasma | Ok | |
| 3 | 09:10:42:720AM | Plasma | Ok | |
| 4 | 09:44:02:850AM | Plasma | Ok | |
| 5 | 10:16:40:190AM | Plasma | Ok | |
| 6 | 11:01:21:690AM | Plasma | Ok | |
| 7 | 11:25:00:370AM | Plasma | Ok | oh2u bypass comm fault |
| 8 | 11:57:36:300AM | Plasma | Ok | |
| 9 | 12:44:05:600PM | Plasma | Ok | |
| 10 | 01:05:19:260PM | Plasma | Ok | |
| 11 | 01:35:38:600PM | Plasma | Ok | |
| 12 | 03:04:51:590PM | Plasma | Ok | |
| 13 | 03:40:43:110PM | Plasma | Ok | |
| 14 | 03:59:42:550PM | Plasma | Ok | |
| 15 | 04:40:31:470PM | Plasma | Ok | |