| Miniproposals | ||||||||||
| ||||||||||
|
| Operators | |
| Session leader(s): | Jim Terry,Jim Irby |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Bill Cochran |
| Engineering Operator Run Comment |
| MP#283 Long Pulse, MP#286 Divertor/Long Pulse. Irby/Terry/Wolfe/Cochran |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Entered: Jul 7 2004 03:49:28:210PM |
| Author: To Be Determined |
| Session Leader Summary for run 1010802 --- MP283 Long Pulse SL: Irby/Terry PO: Wolfe The first objective of this run was to extend the discharge length of the machine to beyond 3 seconds. We extended the pulse length in 0.2 s increments until on shot 21 we had a 3.2 s discharge (runaway, low density, into rampdown). Most of the problems associated with achieving this length discharge are gone over in great detail in the Physop summary. Note that from the engineering standpoint the supplies and support systems performed as expected. I2t, and TC measurements indicated we were operating in a safe parameter space on all systems except for the TF interphase transformers for which more cooling will be added before we attempt another long pulse run. If the power room ambient temperature had been below 100F on this day we probably would not have had a problem with these transformers. The backporch indicated no real change in TF joint temperature because of the long pulse operation. The TF scanner indicated resistances as high as 4 uOhms (as compared to 1.87 uOhms from the backporch). The OH coax resistances varied considerably from shot-to-shot, but were generally below 2 uOhms with peak shots as high as 2.6 uOhm. From Jim Terry: This summary covers that part of the run plan that deals with MP #286a ("Divertor heating during long-pulse, low-density operation"). The most important aspect of that MP concerns divertor heating during the long pulses which have low plasma density and medium levels of auxiliary power. The most important measurements are the IR camera measurements of divertor surface temperatures and measurement of any increasing impurity levels that might be due to the divertor heat loads. The major results form this run are: 1) Current flattop was extended to 2.88 s 2) We could not control the strike point for these long pulses. 3) A low confinement mode is observed after about 1.5 s in these low density, ohmic plasmas. See shots 007,009,010,011. 4) The wall effects are small even for pulses as long as 2.9 s. 5) The particle confinement got significantly worse with time on shot 022 shot 001 - Ip flattop to 1.5s shot 002 - Ip flattop to 1.7s shot 003 - Ip flattop to 1.9s, strong Ha and 2Pi-bolo oscillation (@~500 Hz) (dithering?) shot 004 - Ip flattop to 2.1s, 0.4 MW RF fro 0.6 to 2.1 s - caused two H-mode periods shot 005 - Ip flattop to 2.1s, 0.3 MW RF from 0.6 to 2.1 s - caused EDA H-mode after 1.38s We can't hold strike pt. shot 006 - Ip flattop to 2.13s, no RF and no H-mode, losing strike pt control at 1.4s MoXXXII 127 A signal const at 3.0 units, FeXXII - 132 A signal const at 0.5 units, OVI 150 A signal 0.5 units rising to 1.1 units, FVI - 139 A signal const at 0.2 units. H_alpha rise after 1.4 s. This is due to stronger emission on inner nose. Almost double-null at that time???? Note this is also the time on shot 005 when the plasma went into H-mode, even though the RF power was on starting at 0.6 s. shot 007 - Ip flattop to 2.275s, two Mo injections Turning STIKEPSI off, going to XPT control H_alpha rise after about 1.45 s. This is due to stronger emission on inner nose. At this time the sawteeth disappear, and Te and density decrease, indicating poorer confinement. The impurity levels (Mo, Fe and soft x-ray signal) are also lower during this mode. shot 008 - Ip flattop to 0.9s, disrupts - not due to an injection shot 009 - Ip flattop to 2.3 s, multiple Mo injections, H_alpha rise from 1.5 s to 2.1 s. MoXXXII 127 A signal drops after 2.1 s to 2.0 units (from 4.5 units). IR shows relatively faint toroidal band with surface temperature that increases with time. Low confinement mode seen again, starting at 1.5 s. Could this have to do with current penetration? shot 010 - Ip flattop to 2.3 s, RF on at ~0.5 MW, causing dithers between ELMfree and EDA H-mode (starts at 1.86 s). Mo and Fe up strongly with RF, as well as with H-mode. Mo XXXII to 17 units pre-H-Mode, 20 units during H-Mode. Low confinement mode is not seen. shot 011 - Ip flattop to 2.3 s, RF on at 1.0 MW for 1.2 s, causing dithers between ELMfree and L-mode, Mo and Fe up strongly with H-mode. Mo XXXII to 32 units during H-Mode Low confinement mode is not seen. shot 012 - Ip flattop to 2.41 s, disrupted, RF on at ~1.2 MW for 1.8 s, causing dithers between ELMfree and L-mode, Mo and Fe up strongly with H-mode. Mo XXXII to 31 units during H-Mode. shot 013 - dud shot 014 - Ip flattop to 2.37 s, disrupted, RF on at ~1.2 MW for 1.5 s, causing dithers between ELMfree, EDA, and L-mode shot 015 - fizzle shot 016 - Ip flattop to 2.41 s, disrupted, RF on at between 1.2 and 2.5 MW for 1.5 s, causing dithers between ELMfree, EDA, and L-mode shot 017 - fizzle shot 018 - Ip flattop to 2.5 s, disrupted, RF on at between 1.2 and 2.6 MW for 1.9 s, causing dithers between ELMfree, EDA, and L-mode shot 019 - Ip flattop to 2.5 s, no RF, very low density, NL_04=2.1e19 and slightly decreasing, no density feedback after 1 s. Lots of hard x-rays. The density time history says that the walls are still pumping and that the inventory put into the plasma FROM THE WALLS is not very large. IR camera shows hot spots reaches >800 Deg C (at which point the camera saturates), the symmetric band temperature reaches 500 deg C. shot 020 - fizzle shot 021 - Ip flattop to 2.7 s, no RF, very low density, NL_04=1.56e19 and essentially constant, no density feedback after 1 s. (This is essentially a repeat of 019, albeit with even lower density). Lots of hard x-rays. shot 022 - Ip flattop to 2.88 s, disrupted, NL_04=1.53e19, but with two "Pitcher" puffs, one at 1.0 s the next at 2.0 s. The response to the puffs is very interesting. Even though the pre-puff density was the same at each time, the gas "take-up was more efficient for the first puff (NL04 to 5.1e19 instead of to 4.1e19 for the second puff. The gas that wasn't taken be the plasma showed up in the main chamber and divertor, with the midplane pressure increasing to 0.26 mTorr on the 1st puff but to 0.071 mTorr on the 2nd. This occurred on a slowly increasing midplane pressure "background". The Halpha signal has the same trend as the midplane pressure. All of this implies that the particle confinement got worse with time. Note also that the tight correlation that we normally see between density and midplane pressure is broken in these low density, long pulse cases. |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:37:14:600PM |
| Author: To Be Determined |
| Physics Operator Summary for Thurs Aug 2, 2001 MP#283 Long Pulse MP#286 Divertor/Long Pulse SL: Irby,Terry PO: Wolfe Engineering Setup ----------------- Overnight ECDC in Helium at 5e-5Torr ****NOTE: that's Helium, not D2**** ending two hours before the run. Run to begin at 9:00AM and end at 7:00PM Gas setup: fill B-Top with 6 psi D2 Hybrid enabled (PG4) fill B-side lower with 1 psi Ar Hybrid enabled (PG1) *fill B-side upper with 10 psi He Hybrid DISABLED (PG2) (for ECDC) fill C-side with 30 psi D2 Hybrid enabled (PG3) *fill J-Bottom with 20 psi N2 Hybrid DISABLED(PG5) *NOTE: That's HELIUM in B-side upper and NITROGEN in J-bottom Enable the following gate valves, assuming no vacuum problems: ECE, DNB, VUV Power systems as on 1010727001 for first shot, except tf invert time moved to 3.65sec (for back porch) and fizzle delays set to 200msec on TF and 100msec on the OH coils. The power systems settings will be varied throughout the day. RF Setup: All transmitters in RUN-ON by 8:00AM Run Plan --------- PCS Start from previously prepared shot in USER10:[WOLFE.MDS.SHOTS.HYBRID.PCS]pcs_1010802000 which should be similar to 1010727001, except: End of shot is in segment#3, turns on at 1.4sec TF reduced to 5T, with back porch waveform All gains, controllers, extended to >4sec (seg3) Clear predictors in seg3 moved to 5sec Chopper voltage to 100V Ref 9-12 set to ip in segs 2,3; strkpsi on wire 12 (no gain). The idea is to drag the segment#3 traces to reposition the end of flattop and rampdown times shot-to-shot. We will enable STRKPSI control as early as possible, in preparation for sweeping later in the run. RF will ultimately want to run one transmitter at a time for a constant power up to 1.5MW through the flattop. Early shots may be used for tuning, but RF-induced disruptions should not be allowed to delay progress in extending the pulse length. RF is required for the second part of the experiment (MP286a). During the first half of the run, RF can/should tune up and raise the power to a level that does NOT result in H-mode transitions. Because the tf backporch is an important diagnostic for the magnet heating, and we only get it if the fizzle detector does not terminate the shot, we will move the fizzle detector off time earlier in rampdown than usual. We normally turn off the f.d. at 0.2sec into rampdown; for these shots we will turn it off at 0.1sec into rampdown, so that for any shot that makes it that far, we will get a backporch. This means that shots that disrupt in rampdown will run the PF supplies to their rails until the PF invert time, which is 500msec after end-of-flattop. We may re-evaluate this strategy during the run. For the first day of this experiment, the main goal is to extend the pulse length. If plasma problems are encountered, we will attempt to extend the tf flattop pulse at least to 3.0sec, assuming engineering criteria are satisfied, by disabling the fizzle detector to the TF. Evaluation of taup* from gas puffs and perhaps flappers is the next priority. We will try to get as far as doing the divertor sweeping part of the second MP (#286a). Most likely, the N2 puffing will be deferred to the second day. Note that there is a PCS limitation which prevents us from having BOLOFEED and STRKPSI control active at the same time. This could perhaps be circumvented by removing IC_EF4 control and installing BOLOFEED on wire 15. DIVBOLO1 (input_85 to the hybrid) will be used as the feedback source for BOLOFEED. This signal is from the core bolometer array channel which views just inside the separatrix. The following sequence corresponds to SUCCESSFUL shots. Each increment in pulse length must be accomplished before proceeding, except that disruptions late in flattop while tf and oh fizzle delays are in place may be acceptable. Shot#1: baseline shot. No RF Shot#2: extend flattop to 1.7sec (seg3) Program strkpsi to match shot1 evolution, no gain extend PF invert times to 2.2, fizzle off time to 1.8 tf invert time to 3.85 RF may tune at low power Note that the programmed shot has nl_04 at 7e19 (in L-mode). It will be programmed as flat for the duration of each pulse (with the exception of the wall-loading-expt/Pitcher puff shots, i.e. shots 9-12 Shot#3: tune parameters if needed to reduce voltage, PF excursions extend flattop to 1.9sec turn on strkpsi gain, reduce rxl,zxl gains extend PF invert times to 2.4, fizzle off time to 2.0 tf invert time to 4.05 RF may tune at low power try turning off IC_EF4 gain in seg3 Shot#4: extend flattop to 2.1sec extend PF invert times to 2.6, fizzle off to 2.2 tf invert time to 4.25 RF may tune at low power if shot runs ok with IC_EF4 gain off, then replace it with bolofeed on seg3, wire 15, no gain Shot#5: extend flattop to 2.3sec extend PF invert times to 2.8, fizzle off to 2.4 tf invert time to 4.45 Increase OH1,OH2L,OH2U current limits to -30kA Remove delays on tf, oh* fizzle detector *note that the delays should be removed when we increase the limits RF may tune at low power Shot#6: extend flattop to 2.5sec, extend tf waveform if necessary, begin programming strkpsi waveform extend PF invert times to 3.0, fizzle off to 2.6 tf invert time to 4.65 RF may begin increasing power, if we avoid H-modes and injections, up to 1.5 MW or until ELMfree H-modes occur. The RF should be on from t=0.6s until the end of the flat-top. Shot#7: extend flattop to 2.7sec (if feasible) extend PF invert times to 3.2, fizzle off to 2.8 extend TF invert time to 4.85 Shot#8: extend flattop to 3.0sec (if feasible) extend PF invert times to 3.5, fizzle off to 3.1 extend TF invert time to 5.15 Shot#9-10: for longest pulse achieved (or backed off a little) program short gas puffs (pitcher puffs) for taup* RF may need to back off for these puffs Shot#11-12:program flapper pulses for wall pumping measurement RF may need to back off during flapping Shot#13-17: Gas/nl4 program back to flat Begin strikepoint sweeping, 0.5sec sweeps. Depending on results, RF may increase power. Shot#18-25: If necessary, turn off STRKPSI control, increase gains on RXL,ZXL back to shot#1 values in seg3, reset ref9-12 to "10", and begin bolofeed control for N2 puffing. Alternatively, if removing IC_EF4 worked, return STRKPSI program to nominal and run with bolofeed on wire 15. Program a time history similar to the no-puff history but successively increase the requested level shot-by-shot until a clear increase in edge radiation due to the N_2 is observed. Note the fractional increase over the no-puff level. ------------------------------------------------------- Results: ------- Overnight ECDC in helium ran as requested. Began by loading the prepared shot, with a few last minute tweaks Things started off well, but by step 4 we began to run into the EF1U current rail. The surface voltage was around 1.1V, whether or not we turned on the RF. The H-mode threshold was incredibly low, less than 500kW; however, the H-modes were plagued by H/L transitions, and the voltage was not significantly reduced. It also became apparent that we were not going to be able to sweep the strikepoint late in the shot, because ef1L was also close to railing, and using strkpsi control to even keep the strike below the nose caused the xpoint to move down, and until I cranked up the RXL gain, onto the inner nose. On shot#7 we abandoned strkpsi control. Replacing IC_EF4 control with a preprogrammed V_EF4=200V, going to zero in rampdown, was reasonably successful, with ef2byef4 gain reduced from 2 to 1 to avoid instability. This was left in place for most of the run, although if the strkpsi experiment were not to be feasible we could have used bolofeed on wire#9 with Ref 9-12 set to "10". That is probably the way to do it tomorrow, if we want to do the N2 puffs. The early shots had increases in halpha and gas pressure around the 1.4sec segment switch. This was probably related to change in strikepoint and feedback, but we never really figured out what was going on. Shots#7 and 9 have interesting examples of this effect. In #9 the jump happened at 1.5sec, well after the seg switch. It may have been associated with a change in the divertor or the core MHD; there was a dramatic effect on sawteeth. We tried lots of combinations of RF waveforms. We got up to ~1.3MW continuous through the flattop (.6 to 2.5 or so) and with the the timing stacked so we got up to 2.5MW for 1sec, and 1.2MW for ~.5sec before and after. The problem was most likely that the H/D got too high during the early shots (up to .13) and we couldn't get it back down with He or D2 ECDC between shots. On shot#19 we stopped the RF, both because they had CAMAC problems on their monitor signals, perhaps because of the power room temperature, and because it wasn't helping. We started a shot with 2e19 density program and turned off the gas in seg3. We got a runaway shot that ran into rampdown (2.85) and showed lots of heating in the divertor, especially when the strike rose to the outer nose. Shot 21 was an even worse runaway shot, but it made it all the way through rampdown, with eof at 2.7sec and plasma to 3.2sec. If you don't mind the runaways, this is the best shot of the day to look at. The density got down to 1.6e19. Shot 22 was programmed for eof at 3.0sec with density feedback back on at 3e19, and two gas puffs at 1.0sec and 2.0 sec. This shot disrupted at 2.88sec still in "flattop", having lost vertical control when the EF1's railed. There were still runaways, but much lower. The response to the puffs is interesting and unexpected. The later puff was much less effective than the first in raising the density, though it was a larger contribution to the h-alpha and midplane pressure. There was no sign of wall pumping. Earlier in the run we decided to begin increasing the tf pulse length past the plasma, since tf cooling was not limiting and the magnet was behaving well. We exceeded the original target for the tf flattop, running shots 19,21 and 22 to 3.2sec EOF, with a full backporch. For these shots the tf fizzle delay was raised to 1.5sec, and the fizzle off time was left at 2.8 to ensure getting the long flattop and back porch. We had three shots on which the PF breaker pulled out with a time overcurrent indication, 2 on OH2U and 1 on OH1. Apparently the setting on these relays was incorrect because of a miscommunication about the gain of the CT feeding the relay. A change was made to the oh2u setting, and similar changes need to be made to the oh1 and oh2l before the next long pulse run. When this is done, we expect the relay to fault before the I^2t circuits, but they should be ok for the present experiment. Aside from these trips, there were no heating or other problems with any of the supplies (or coils), and the shot cycle time was limited primarily by the time to set up the next shot, rather than the cooling. Things remaining to be done include the flapper shots, some more pitcher puffs, and the bolofeed with N2. The strikepoint sweep back down the plate looks completely infeasible, unless a miracle occurs with the RF and the voltage drops a lot. The only long shot that did not rail the EF1U was #21, with a loop voltage around .6V, but that was seriouslu running away. Scorecard --------- Tests 0 Duds 1 Fizzles 3 Plasmas 18 ------- -- Shots 22 |
| Session Leader Comments | |||
| Aug 2 2001 09:16:44:403AM | 1010802001 | Jim Irby | Good shot TF MB79 turn at 1.75 uOhms |
| Aug 2 2001 09:51:19:983AM | 1010802002 | Jim Irby | TF MB79 still about 1.75 uOhms at 3.8 s oh1: 0.5150 uohm oh2u: 1.1474 uohm oh2l: 1.6550 uohm No new time dependant changes in OH resistances |
| Aug 2 2001 10:25:03:597AM | 1010802003 | Jim Irby | TF taps okay at 4 s oh1: 0.5063 uohm oh2u: 1.1962 uohm oh2l: 1.6509 uohm 80 RPM delta in Alternator speed ---- deltaE 198 MJ |
| Aug 2 2001 10:38:48:187AM | 1010802004 | Jim Irby | No backporch for the shot --- disrupted oh1: 0.5023 uohm oh2u: 1.1248 uohm oh2l: 1.6007 uohm DeltaE alternator of 212 MJ TF resistance temp from -176 to -157 C or dT of +19C |
| Aug 2 2001 11:11:12:920AM | 1010802005 | Jim Irby | No TF ramp again oh1: 0.4450 uohm oh2u: 1.0212 uohm oh2l: 1.1934 uohm 220 MJ out of alternator |
| Aug 2 2001 11:48:49:727AM | 1010802006 | Jim Irby | No backporch 228 MJ dE oh1: 0.5218 uohm oh2u: 0.8418 uohm oh2l: 1.2125 uohm |
| Aug 2 2001 12:29:07:537PM | 1010802007 | Jim Irby | No backporch but only 2.28 s of plasma oh1: 0.5317 uohm oh2u: 1.0383 uohm oh2l: 1.0868 uohm dE of 224 MJ |
| Aug 2 2001 12:50:00:157PM | 1010802008 | Jim Irby | Lost plasma at 0.89 s oh1: 0.5541 uohm oh2u: 1.1457 uohm oh2l: 1.1926 uohm |
| Aug 2 2001 01:09:05:473PM | 1010802009 | Jim Irby | Plasma into rampdown Got backporch, but still can't measure joint resistances because of pickup from crowbared OH. oh1: 0.6139 uohm oh2u: 1.0225 uohm oh2l: 1.0924 uohm dE 229 MJ |
| Aug 2 2001 01:39:45:110PM | 1010802010 | Jim Irby | Plasma for 2.61 s Delta Ts from coil resistance measurements: TF ~20C OH2U ~30C OH1 ~30C OH2L ~30C oh1: 0.5199 uohm oh2u: 1.0666 uohm oh2l: 1.5529 uohm dE 230 MJ |
| Aug 2 2001 02:44:38:157PM | 1010802011 | Jim Irby | Plasma for 2.78 s Finally got backporch: TF MB79 at 1.85 uOhms at 4.2 s oh1: 0.4829 uohm oh2u: 0.7747 uohm oh2l: 1.3929 uohm 229 MJ |
| Aug 2 2001 03:05:31:683PM | 1010802012 | Jim Irby | lost plasma at 2.4 s, no backporch oh1: 0.6230 uohm oh2u: 0.9024 uohm oh2l: 1.7087 uohm 241 MJ |
| Aug 2 2001 03:32:18:907PM | 1010802013 | Jim Irby | Dud oh1: 0.5594 uohm oh2u: 1.3582 uohm oh2l: 1.1159 uohm |
| Aug 2 2001 04:19:19:230PM | 1010802014 | Jim Irby | Plasma for 2.37 s TF resistance indicates a 23C increase in temp oh1: 1.4311 uohm oh2u: 1.7637 uohm oh2l: 1.7459 uohm duds are not good |
| Aug 2 2001 04:26:23:273PM | 1010802015 | Jim Irby | TF MB79 at 1.4 uohm, 142 kA, 1.4 s oh1: 0.7242 uohm oh2u: 1.3334 uohm oh2l: 1.1026 uohm |
| Aug 2 2001 04:52:40:640PM | 1010802016 | Jim Irby | Plasma to 2.41 s TF resistance temp up 25 C oh1: 1.3815 uohm oh2u: 1.6764 uohm oh2l: 1.7873 uohm |
| Aug 2 2001 05:10:02:373PM | 1010802017 | Jim Irby | fizzle oh1: 0.6823 uohm oh2u: 1.3781 uohm oh2l: 1.0800 uohm |
| Aug 2 2001 05:36:16:720PM | 1010802018 | Jim Irby | Plasma for 2.54 s, no backporch yet oh1: 1.4374 uohm oh2u: 1.6101 uohm oh2l: 1.6973 uohm |
| Aug 2 2001 06:17:46:363PM | 1010802019 | Jim Irby | Plasma for 2.83 s TF resistance delta T of 29C M79 at about 1.77 uOhms at 5 s oh1: 0.6747 uohm oh2u: 1.3598 uohm oh2l: 1.0380 uohm dE of 262 MJ IR guys report hot spots in divertor (signal saturated above 800C) |
| Aug 2 2001 06:15:48:953PM | 1010802020 | Jim Irby | fizzle oh1: 0.6071 uohm oh2u: 1.4576 uohm oh2l: 1.0123 uohm |
| Aug 2 2001 06:35:10:647PM | 1010802021 | Jim Irby | Plasma for 3.18 s TF resistance change of 29C MB79 at 1.8 uOhms oh1: 1.2764 uohm oh2u: 1.6448 uohm oh2l: 1.6537 uohm dE 266 MG |
| Aug 2 2001 06:55:40:920PM | 1010802022 | Jim Irby | Plasma for 2.86 s oh1: 0.6880 uohm oh2u: 1.0114 uohm oh2l: 1.5862 uohm |
| Physics Operator Comments | |||
| Jul 27 2001 02:12:35:030PM | Steve Wolfe | Beginning to set up PCS for the long pulse operations. Load From from 1010727001 Import seg3 from seg2; I'm going to use seg3 for the last half drop tf to 139000 (5T) in both seg2 and 3. in seg 3, kill points before .5sec, insert knot at 1.499 turn seg3 on at 1.4sec copy strkpsi wire into wire 12 in seg2 and 3; set 9-12 to IP ref. strkpsi gain at zero, need to look at error sig. in seg 3, add a knot at t=-2 to all P wires similar for the V wires. Extend all gains in seg3 to 4.0sec Add a back porch to the tf in seg3, initially 3.7:4.0sec Save in pcs$wolfe:pcs_1010802001 | |
| Jul 27 2001 03:23:07:080PM | Steve Wolfe | Checking the controllers, I find some wires with zeroed controllers at 3.0sec like zcur3. eliminate those in seg 3. Also, most of the predictors have a clear at 3.0sec. Move these to 5.0sec Also, extend the integrator gates to 4.0sec Now resave it | |
| Jul 27 2001 03:48:54:273PM | Steve Wolfe | Copy seg3 to seg4, and move the rampdown time to 3.5sec. Modify non-zero traces in all_v's. save it this way. I can turn seg 4 on and use this for hybrid tests (*not during a run though) | |
| Jul 28 2001 01:07:27:470PM | Steve Wolfe | Test of hybrid for long pulse operation. 7/28/01 Load from pcs$wolfe:pcs_1010802000 turn seg 3 off and seg4 (full length pulse) on Load into Model Do a coldstart - logs ok take a test shot #80 with freq=500. POUT,VOUT, MOUT look ok Reload from 1010727028 (last shot yesterday) take test shot#81, back at 100Hz dig rate Looks ok. | |
| Aug 2 2001 08:01:53:667AM | Steve Wolfe | Setting up PCS for long pulse run. Load from pcs$wolfe 1010802000 seg1: BR0 to .00175, IC_EF4 to -1305 (as seemed to work yesterday) pg4 left at 27msec, as on 1010727001 seg2: STRKPSI waveform to 0 (was -.005) seg3: "" "" Load at 2-Aug-2001 07:37:12.00 coldstart - all matrices respond ok. | |
| Aug 2 2001 08:22:10:110AM | Steve Wolfe | Note that strkpsi predictor is set for ZSOUT=-.54, which is 2cm below Ian's default point. Therefore, I can't quite use the IanFlux scope directly to compare. I'll leave the predictor this way, because I believe we will want to put the strike point lower rather than higher for best IR camera views and interpretation, before we start sweeping. The starting shot under xpoint control is expected to have the strike near -.52 in flattop, i.e. near Ian's default location. I expect this will come out something like +.005 relative to the loaded predictor. | |
| Aug 2 2001 09:26:05:280AM | 1010802001 | Steve Wolfe | shot#1: plasma, some runaways early. full length bouncer, pressure is ok, Bz,Br a little negative. So far, so good. Next shot: increment length by .2sec EF4 back to -1300. strkpsi from .001 to .004 in seg2 and 3, no gain vpg3 to zero |
| Aug 2 2001 09:35:47:813AM | 1010802002 | Steve Wolfe | shot#2: increment length by .2sec EF4 back to -1300. strkpsi from .001 to .004 in seg2 and 3, no gain vpg3 to zero plasma full length. 2.2sec no runaways injection at 1.2sec starts some dithers and a small density increase. breakdown an tad late, no bounce seg 2 physaout error is steady at 1280a-m in seg2 but takes a jump in seg 3. Must have misprogrammed. next shot: go to shot#3 plan setup. |
| Aug 2 2001 10:02:18:837AM | 1010802003 | Steve Wolfe | shot#3: increment length by .2sec go to shot#3 plan setup, except leave strkpsi off, turn off IC_EF4 in seg3, program 100V on VEF4 plasma, disrupted in late rampdown 2.33. ef1U hit limit at end of flattop. breakdown is still late. hash on halpha from 1sec on. power supplies may have hit a timer limit. RF is on from .7 to 1.9 at .25 to .3MW. EFIT seems to stop early. It's set for 101 points! |
| Aug 2 2001 10:41:56:050AM | 1010802004 | Steve Wolfe | shot#4: increment length by .2sec drop zxu from .04 to .02 in seg 2 and 3 Turn on seg3 strkpsi control gain=5 from -2 to 2.1 zxl gain from 8->2 in same range, then back to 8. plasma, disrupted at 2.1sec, lots of thrashing, got hmodes. Seem to have some thrashing, probably the strkpsi control. EFIT still seems to be broken. ef2by ef4 goes crazy at the h-modes. Looks like the xpoint bangs the inner nose. There is a VDE related to tuning ZXL back on in rampdown. |
| Aug 2 2001 11:11:27:653AM | 1010802005 | Steve Wolfe | shot#5: increment length by .2sec, EOF at 2.3sec leave zxl gain at 2 in rampdown. drop ef2 by ef4 gain from 2->1, and raise VEF4 to 150V. move initial strkpsi value up to .00575 to reduce jump at seg switch increase rxl gain from 4->8 in seg 3 until 2.3sec EOF disrupts near eof at 2.1. Got another h-mode. When you want them the threshold is higher. It seems to be happening at seg switch, so it must be the shape change with strkpsi. I forgot to extend strkpsi gain, it kills it when it turns off. next shot: repeat this length, try to fix it. |
| Aug 2 2001 11:45:25:487AM | 1010802006 | Steve Wolfe | shot#6: EOF at 2.3sec strkpsi gain down to 2, off at 2.3 move rxu from -.028->-.01 in seg 2 and 3 Makes it to 2.246 and disrupts again We're getting upper null at the end and EF1U is railing! Strike point sweep isn't going to work. Try going back to xpoint control. |
| Aug 2 2001 12:26:40:190PM | 1010802007 | Steve Wolfe | shot#7: strkpsi gain to zero, RXL back to 4, ZXL to 8 drop the density program to 5e19 increase oh current limits to -30kA raise V_ef4 from 150->200 disrupts at 2.28! Still getting a rise in Halpha at segment switch. ???????? There is an injection that preceeds the halpha rise and follows seg switch. next shot, drop zxl to be constant at -.01 no other changes |
| Aug 2 2001 12:50:25:720PM | 1010802008 | Steve Wolfe | shot#8: drop zxl to be constant at -.01 no other changes disrupts early at .89. Nothing was changed at that time (seg2) It's a vde for no obvious reason. Oscillations on EF2 and OH2 before chopper rails. next shot: nl04 to 6e19 in seg2 and 3 no other changes |
| Aug 2 2001 01:10:22:633PM | 1010802009 | Steve Wolfe | shot#9: drop zxl to be constant at -.01 nl04 to 6e19 in seg2 and 3 no other changes plasma ramps down until 2.7sec pf breaker pulled. oh2's and ef1's both on their current rails. something is causing an increase in halpha, midplane pressure, gas at 1.5sec. It isn't at seg switch. sawteeth either disappear or get small and fast; could be m=1 continuous. next shot: raise nl4 to 7e19 and let RF turn on |
| Aug 2 2001 01:46:20:907PM | 1010802010 | Steve Wolfe | shot#10:raise nl4 to 7e19 and let RF turn on plasma, multiple h-modes from 1.2sec gets to 2.6sec breaker pulls. oh1 rails at 28kA (set for 30) oh2u rails at 28.9 " oh2l rails at 27.4 " ef1U,L rail at 5kA ef2's go to zero. next shot: raise the RF |
| Aug 2 2001 02:49:24:433PM | 1010802011 | Steve Wolfe | shot#11:raise the RF engineers have changed a setting on the time-overcurrent relay. multiple h-modes, breaker stayed in. plasma to 2.77sec, ramped all the way down despite railing oh2's and ef1's. ef1 is on the rail from 2sec on, the others rail in rampdown. next shot:increment to 2.5sec raise oh2L current limit by 2kA, to get closer to 30kA. raise nl4 to 8e19, pg3 gas puff at 20V |
| Aug 2 2001 03:05:36:900PM | 1010802012 | Steve Wolfe | shot#12:increment to 2.5sec raise oh2L current limit by 2kA, to get closer to 30kA. raise nl4 to 8e19, pg3 gas puff at 20V disrupted at 2.4sec, in flattop multiple h-modes. very late breakdown, 8.3msec; almost lost that one. 1.2MW of RF with a few trips. ef1's both rail in flattop, then we lose it vertically voltage is still a little over 1volt. next shot. Turn off density feeback and leave pg3 puff on during h-mode. |
| Aug 2 2001 03:21:26:540PM | 1010802013 | Steve Wolfe | shot#13: Turn off density feeback and leave pg3 puff on during h-mode. BR0 to .002 dud! next shot: ecdc for 10min in helium drop the pg4 to 24msec |
| Aug 2 2001 04:25:05:850PM | 1010802015 | Steve Wolfe | shot#15: ecdc for 10min in helium stack the RF to get 2.5MW for 1sec of the pulse. extend tf to 3sec another dud. NO gas puff. had a plc problem. |
| Aug 2 2001 04:46:15:720PM | 1010802016 | Steve Wolfe | shot#16: ecdc for 10min in deuterium and try again B-side upper now has D2 stack the RF to get 2.5MW for 1sec of the pulse. extend tf to 3sec disrupts at 2.4sec again. got multiple H-modes, got up to 2.5MW but no increase in temp or drop in loop voltage H/D still high? |
| Aug 2 2001 05:07:06:167PM | 1010802017 | Steve Wolfe | shot#17: ecdc for 10min in deuterium again no beam, no argon. B-side upper now has D2 stack the RF to get 2.5MW for 1sec of the pulse. tf at 3sec. If this doesn't work, we'll try dropping the density to minimum (no gas) fizzle. pressure is up. |
| Aug 2 2001 05:27:19:767PM | 1010802018 | Steve Wolfe | shot#18: No ECDC pg4 down to 22msec, ef4 to -1310 no beam, no argon. stack the RF to get 2.5MW for 1sec of the pulse. tf at 3sec. If this doesn't work, we'll try dropping the density to minimum (no gas) plasma to 2.54sec, breaker pulled again. again with the multiple h-modes. breakdown still a little late, bz still a little high next shot: Turn off RF (they have a camac problem anyway) drop density to 2e13, no pg3 puff |
| Aug 2 2001 05:50:33:577PM | 1010802019 | Steve Wolfe | shot#19: Turn off RF (they have a camac problem anyway) drop density to 2e13, no pg3 puff tf eof to 3.2, invert time to 5.35 dnb will fire again, the H/D was not affected. plasma to 2.85 sec, into rampdown (with runaways) ef1U rails at 2.3, ef1L doesn't quite rail . Vsurf is ~.95 next shot: extend current eof to 2.7sec |
| Aug 2 2001 06:10:31:723PM | 1010802020 | Steve Wolfe | shot#20: extend current eof to 2.7sec turn gas off everywhere after 1sec another fizzle. bounced. |
| Aug 2 2001 06:27:46:337PM | 1010802021 | Steve Wolfe | shot#21: back to 24msec on pg4 extend current eof to 2.7sec turn gas off everywhere after 1sec plasma to 3.2sec, all the way down xray snow whites out the tv screen. nl4 is at 1.6e19. next shot: nl4 feedback to 3e19 pitcher puffs. |
| Aug 2 2001 07:00:14:353PM | 1010802022 | Steve Wolfe | shot#22: add 1msec to pg4 nl4 feedback to 3e19 pitcher puffs at 1.0 and 2.0sec plasma disrupts at 2.88sec first and second pitcher puffs are different density response, different decay runaways much lower, but voltage is back to ~1V, and the ef1's railed again. breaker pulled, another time overcurrent on OH1 this time. It happened after the disruption. That's the last shot for today. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 09:03:54:690AM | Plasma | Ok | |
| 2 | 09:26:48:030AM | Plasma | Ok | |
| 3 | 09:51:48:690AM | Plasma | Bad | TF and EF1 hit the power supplies backup timers |
| 4 | 10:23:44:797AM | Plasma | Ok | |
| 5 | 11:00:40:883AM | Plasma | Ok | |
| 6 | 11:34:52:467AM | Plasma | Ok | |
| 7 | 12:04:04:397PM | Plasma | Ok | |
| 8 | 12:29:24:883PM | Plasma | Ok | |
| 9 | 12:52:33:693PM | Plasma | Bad | PF Fault, Breaker opened at end of shot |
| 10 | 01:29:22:047PM | Plasma | Bad | PF fault, Breaker opened from oh2u timed overcurrent relay |
| 11 | 02:29:06:613PM | Plasma | Ok | PF Breaker stayed in. |
| 12 | 02:52:47:610PM | Plasma | Ok | |
| 13 | 03:17:03:330PM | Plasma | Ok | |
| 14 | 03:40:48:687PM | Test | Bad | |
| 17 | 05:01:48:890PM | Plasma | Ok | |
| 18 | 05:20:21:613PM | Plasma | Bad | OH1 type 2 flt at end of shot |
| 19 | 05:42:30:953PM | Plasma | Ok | |
| 20 | 06:08:31:033PM | Plasma | Ok | |
| 21 | 06:22:08:530PM | Plasma | Ok | |
| 22 | 06:44:45:633PM | Plasma | Bad | PF Fault, Breaker opened at end of shot |