Miniproposals

Miniproposal: 300 Date Filed: 6/14/2002 Title: TF Ramps First Author: Lee Session Leader: Catherine Fiore (shots 1-21)

Miniproposal: 344b Date Filed: 7/1/2003 Title: 100% Bootstrap Fraction Plasmas First Author: Peter Politzer Session Leader: Unknown (shots 22-40) Session Leader: Peter Politzer (shots 22-40)

Engineering Operator Run Comment

Run part 1 MP 352, 303, and part of 300 -- ITB control. SL:C Fiore PO: S

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:33:430PM

Author: Catherine Fiore

1030702 Part A: The purpose of this experiment is two-fold: find the effect on the ITB if the field is ramped up beyond the point where and ITB is established after an ITB is formed. Secondly, it is to provide an ITB target with sufficient density to have the FRCECE diagnostic at or near cutoff in order to study probable density fluctuations. First, set up an EDA H-mode ITB plasma with 1.5-2 MW off axis ICRf at 4.5 T. Once the ITB timing is known, add a toroidal field ramp, initially to 4.6 T to study the effect of the ITB. If the ITB is not extinguished, increase the ramp to 4.7 T and determine where the ITB cutoff is, if indeed it is extinguised by rising field. Two shots are needed without ITB (slightly higher target field) to provide a non-ITB baseline for the FRCECE ---------------- We achieved a number of satisfactory ITBs with at least some ITBs on all shots between 8 and 18 with the exception of shot 17. TF ramp experiments started on shot 12. There is a suggestion that rising field may clamp the ITB when it reaches slightly above 4.6 T, but H to L back transitions may interfere with solid interpretation here. ---------------------------------------------------------------------------- 1030702001 No power. Data entry problem on the C-Coil set up. 1030702002. Good plasma but no H-mode. RF power was only 1 MW. Dalsa data was not good. 1030702003 H-mode started but was not steady. Disruption. Dalsa diagnostic seems to have recovered from problems it had on shot 2. 1030702004 H-mode still ratty: RF faulting. 1030702005 H-mode started then the plasma disrupted. Injection of some sort. 1030702006 H-mode started then the plasma disrupted again. 1030702007 No plasma. 1030702008 H-mode, with some rf dropouts. Looks like an ITB was forming before an injection and rf dropout. We will try to get a clean version of this. 1030702009--H-mode, still with rf dropouts. ITB formed at 1 s and lasted until the big rf dropout at 1.3 s. We will try again to clean it up. 1030702010 Two nice H-modes which drop out with the rf dropouts. ITB forms in both. We will back off slightly on the rf power to try and make a clean pulse. 1030702011 Long H-mode with weak ITB plagued with small drop outs and EN modes. We will start trying the Tf ramp experiments on the next shot. 1030702012 ITB starts, but a big rf drop out ended it when the toroidal field reached 4.6 T, making interpretation difficult. We will repeat. 1030702013 Two ITBs, the 2nd coinciding with the toroidal field ramp. An H-mode back transition again makes interpretation difficult. We will try again. 1030702014 One ITB, still with a back transition problem during the toroidal field ramp. We are going to adjust the target density. 1030702015 H-mode started later because of the increase in the target density. One ITB which formed as the field was rampingup, did not seem to care that the field was increasing, thus giving us a higher field point for making an ITB than we have previously achieved with the resonance on the high field side. We will try ramping the field up faster and slightly further. 1030702016 The late H-mode start makes interpreting the apparent beginning of an ITB hard to interpret. It is possible that the ITB rise was clamped due to the Tf increase, but we need more time for the ITB to fully develop in order to be positive about this. We will back off slightly on the target density. 1030702018 ITB forms, with the density remaining peaked through the back transition at 1.05 s. It looks like the peaking clamps at 1.2s when the toroidal field reaches 4.6 T, but a back transition quickly destroys the ITB. We are next doing 2 no rf gas puffing shots as requested by the FRCECE group. 1030702019 Reference shot 1 for FRCECE. No rf, no ITB 1030702020 No plasma 1030702021 FRCECE reference shot #2

Entered: Jul 7 2004 03:49:33:443PM

Author: Peter Politzer

Subject: summary_1030702B From: Pete Politzer Date: Wed, 02 Jul 2003 16:12:06 -0700 Newsgroups: psfc.cmod.bboard Session Summary run 1030702b July 2, 2003 (afternoon) Miniproposal 344B - 100% Bootstrap Fraction Plasmas shots 1030702022-40 session leader (nominal): Pete Politzer (with Rip Perkins) physics operator (and all-around wizard): Steve Wolfe Objective: The overall objective of this experiment is to study the behavior of fully noninductive plasmas (without induction), and to determine the characteristics of the self-consistent current and pressure profiles found with near 100% bootstrap current. The initial step for today's run was to establish a noninductive plasma, by switching control of the OH and PF coils from current programming to voltage control. We planned to extend the pulse up to 5 s in order to demonstrate that a stationary state without induction is possible, and to begin to explore parametric variation of that state. Results: As predicted by some, the afternoon was spent working on the voltage control settings. The initial conditions were an Ohmic plasma, at about 600 kA, with B= 5.3 T. Initially, the RF was turned on at 0.6 s and voltage control was started at 0.7 s. The best case for this timing was 026, which had about 0.9 MW of ICRF power, and only derivative gain on the voltage control. The time for transition to voltage control was moved to 0.45 s, the derivative gain was adjusted, and a second transmitter arrived, leading to shot 029. This had about 2.1 MW into the plasma, sustaining approximately constant central electron density and temperature to 1.2 s. In general, the second transmitter tripped out when the current dropped to 300 kA, establishing an operational low-current limit. Usually disruption occurred at about 200 kA, shortly thereafter. For most of the afternoon, two transmitters were working. The power delivered increased steadily during the day, to 2.8 MW in shot 040. Adding proportional gain increased the current decay rate, but did manage to hold the control parameter at the set level. The current profile was peaking significantly as Ip decayed, ad indicated by a rapidly rising internal inductance. As it appeared that this reflected the peaked resistivity profile, the toroidal field was reduced to try to broaded the Te profile. The best case at 5.0 T was shot 038 (probably the best shot of the day). In shot 039, the field was reduced further, to 4.8 T. In shot 040, the gains were reset to the values for 029, and the density was slightly increased. This one is also worth a further look. We need to understand better what the control system is trying to do. It may be necessary to find (or develop) a better proxy for the plasma surface flux. Also, more heating power is clearly needed (compare 026 and 029). We reached beta_p = 1.5 transiently with 2.8 MW, so the bootstrap current was not sufficient to take over from the inductive current. Something in the neighborhood of 2.5 is needed.

Physics Operator Summaries

Entered: Jul 7 2004 04:37:18:520PM

Author: Steve Wolfe

Physics Operator Summary for Wed. July 2, 2003 This was a two part, 10 hour extended run: Part 1: ITB studies (MP#352,303, and part of 300) SL: Fiore EO: Byford PO: wolfe Part 2: 100% bootstrap plasmas MP#344B SL: Politzer,Perkins (by remote access and telephone from GA) EO: Dekow PO: wolfe FUTURE PHYSOPs NOTE: Part 1 is in seg2 and part 2 is in seg3 in PCS. The part 1 part of the day comprised shots#2-21 (see below for what happened to shot#1). Part 2 was shots #22-40. --------------------------------------------------------------------- Part 1 started with a target like 1021010004, 800kA 4.5T HFS resonance ITB. Total power of 2MW was called for. MP#352 was piggybacking, looking at fluctuations on ECE as it went through cutoff, around 1.8e20; two ohmic shots with density raised to 1.8e20 by puffing were taken to provide background for this experiment. Aside from that the whole run was done with essentially the same RF. On shot 12 we began ramping the field back up from 4.5T to see if the ITB would be maintained or clamped. Several of these shots were suggestive of just this effect, but we seemed to be unable to avoid H-L back transitions just at the crucial time, so the analysis of these shots will be difficult. Early in the run we experienced a number of disruptions whose cause was not immediately obvious. It turned out that EF2U was receiving one or more spurious glitches on the hybrid control signal, which were causing it to slam full on, resulting in loss of position control and subsequent bashing on the outboard limiter/antennas and then disrupt. No such glitches appear on the hybrid output digitized in the interface room. Examples are shot #3, 5, and 6. Looking back it also turns out that shots 1030701007 and 1030701008 also disrupted due to the same effect. The glitches have amplitude of -12.8V, while saturated output from the hybrid appears to be limited to 11.1V or so. The glitches appear to be being generated in the AFOL receiver at the TMX rack. Suspicion was drawn to the J-port transmitters (isn't it always?) and J was turned off for shots #7-13, and the glitches did not recur. On shot 14, J3 was fired into the dummy load just before rampdown, and the glitch re-appeared and caused another disruption. Firing J4 into dummy load during rampdown did not cause any glitches on shots 15 and 16; On shot 17-18 J3 was put back into dummy load but the interference did not reoccur. The interference appears to be intermittant but it seems likely that there is an RF leak in the J3 line between the transmitter and the switch/dummy load. J was left off for the rest of the day. This problem needs to be investigated and fixed before next week's RF runs. The startups early in the run were very bouncy and many had strong runaway problems; in most cases there was a severe bounce and the current would actually drop significantly before taking off. This behavior has been present for over a week now. I found that the bounces were completely eliminated by changing the segment 1 RCUR offset from 3600 to 5000a-m; this was implemented in shot#10. Previous attempts to improve the startup by raising the pg4 pre-puff (to 30msec) and changing ic_ef4u from -1305 to -1290 were ineffective, although they did no obvious harm. I left ic_ef4u at -1290, but dropped the pre-puff later in the day as indicated by slowing of the current raise. Some shots, especially later in the day, had some runaways, but the current rises after #10 were all fast and monotonic; the runaways seemed curable by raising the pre-puff again. I strongly suggest using a startup like that on shot#10 in the next runs. On shot#1, we had an INIT failure on a TMX J221 and had to run through a no-power cycle. This turned out to be because we had loaded an power systems setup from 2002 which of course did not have the C-coils. The widget came up with the string "HYBRID" in the time fields for start, invert, etc. for the c-coils, and since we weren't using the c-coils today, nothing was filled in to replace these, which were then loaded as values to the J221. Of course, this caused and invalid data error at INIT time. When time values were loaded into the fields, the next shot went fine. This is a bug in the Engineering Power Systems Control Panel software. It needs to be able to deal properly with missing values in these fields when loading old shots. Until it gets fixed, physops and engops need to be aware that those fields must not contain non-numeric values. ------------------------------------------------------------------- Part 2 of this run was an attempt to develop a high bootstrap fraction (up to 100%) plasma. The approach was to start from a low current, low density H-mode target and change from current control to psi0 control with derivative gain, hopefully resulting in a clamped loop voltage. The psi0 predictor was loaded on wire 10, which is the one with the high gain (tau_d=10msec) derivative stage. Some progress was made, but the performance of the Dpsi0 control was not as effective as had been hoped, and we did not get close to a self-sustained steady-state current. More power might have helped, but we didn't want to lose shots to the J3/EF2U interference problem. I had previously set up a shot in PCS with the initial target shot in segment 3. This was basically 1030530024 with everything extended to 5 sec (being very optimistic) and the switchover from current to derivative psi0 control at 0.7sec. RF was asked to start at .6sec and run until 2sec or the fizzle detector turned them off. TF invert was set to 2.0sec, and f.d. to 1.9. PF supplies are normally inverted at 2sec. The idea was to extend the pulse length as necessary. It didn't become necessary as all the discharges in this part of the run decayed in less than 2sec, or else disrupted. As we were getting ready to begin this part, I realized that the psi0 predictor was giving very large values, of the order of 10^5. Looking at the code, I found that the flux value is being divided by calfac, which is defined as 2R mu0/a =8e-6, which is used in other places to convert from flux differences to amp-m, but is wrong for this quantity. Assuming this is just a factor though it shouldn't make a difference, except in trying to figure out what the flux really is. Shots #22-24 were mistakenly taken with segment 3 on but set to async, and therefore were not really switching to psi0 control. I think these were basically free-running on segment 1 pre-program. It was clear that nothing dramatic was happening at 0.7sec, when the changeover from Ip control was supposed to be happening, but we were getting long slow current decays, so I thought maybe it was working. Since the psi0 error was not approaching a constant, I tried raising the D gain to maximum, adding in wire 11, and finally turning on P gain with a target value of 50000, which was the value in the phys_aout scope at the time of the gain switch (another mistake). Meanwhile the RF was getting turned on and improved, so the changes we observed in these shots were the result of the RF, not of the gain changes I was making. I finally noticed that seg 3 was async before shot 25, when I looked at the PID outputs and found no evidence that the seg3 gains were being applied. I then just set it back to sync and let it run, with all the accumulated gain changes, including the P gain. This led to a disruption of course. For the next shot (26) I turned off the Pgain and dropped the D gain back to 2.0. This sort of worked, but the switch happened right at oh1 crossover, so oh1 had zero current througout the shot and the clearin was badly controlled. From shot#27 I moved the switch from Ip to psi0 control earlier to .45sec, to try to prevent crossover. After some more adjustments this was accomplished at shot#29, which has oh1 current nearly flat at -1900A, and eventually got vloop nearly zero. However, the psi0 was not close to being held constant for most of the shot. The next several shots were spent trying to improve the feedback control. Much higher derivative gain was unstable. The first attempt at going back to P gain repeated the mistake I had made earlier: I drew the waveform as a positive value, +130000 Ian-units=1.04Wb, which was what I was seeing at the appropriate time in the aout scope with the target at 0. Of course, this scope is the error signal in the sense of "what you asked for - what you got", so the right target was -130000. Not surprisingly, what I got on shot 31 was a nasty disruption. Shot #32 used -130000, as it should and ended up clamping the psi0 signal with an error around -30000ians=-.24Wb, as expected driving the oh1 back toward the precharge by a few kA as the current fell. However, while use of P gain did a better job of holding the psi0 signal constant, it seemed to cause the current to drop too rapidly, reaching 200kA by ~.7 or .8sec. All of these shots had strongly rising li as the current fell, indicating the profiles were becoming more peaked, rather than less as would be the case for a bootstrap dominated plasma. The RF managed to stay on pretty well until the current reached ~300kA, at which point E-port would fault out. D did better down to lower current. On shots 34-39 we tried dropping the tf to see the effect of heating off-axis. We first went to 5T and then to 4.8, but the peaked profiles persisted. Shot 39 went back to all D gain, and seemed better, very similar to 29. Shot 40 was again all D gain, back at 5.4T . It was supposed to be higher target density, but I didn't change the program early enough, so it ended up about the same density as earlier shots. Good shots (or those worth further study) included 29, 38, 39, 40. The highest values of betap were ~1.5, which is respectable but we really wanted 2.5. Values of li got as high as 2.5, which is not a good thing, at least for what we wanted. Many of the low-current disruptions were vertical instabilities at n/ncrit>1.4 due to the high li. Scorecard ---------- No Power Shots 1 Duds 0 Fizzles 2 Plasmas 37 ------ -- Shots 40

Session Leader Comments
Jul 2 2003 07:30:52:013AM		Catherine Fiore	The purpose of this experiment is two-fold: find the effect on the ITB if the field is ramped up beyond the point where and ITB is established after an ITB is formed. Secondly, it is to provide an ITB target with sufficient density to have the FRCECE diagnostic at or near cutoff in order to study probable density fluctuations. First, set up an EDA H-mode ITB plasma with 1.5-2 MW off axis ICRf at 4.5 T. Once the ITB timing is known, add a toroidal field ramp, initially to 4.6 T to study the effect of the ITB. If the ITB is not extinguished, increase the ramp to 4.7 T and determine where the ITB cutoff is, if indeed it is extinguised by rising field. Two shots are needed without ITB (slightly higher target field) to provide a non-ITB baseline for the FRCECE.
Jul 2 2003 08:03:23:457AM	1030702001	Catherine Fiore	No power. I'll leave Steve to explain it.
Jul 2 2003 08:25:34:863AM	1030702002	Catherine Fiore	Good plasma but no H-mode. RF power was only 1 MW.
Jul 2 2003 08:34:30:160AM	1030702003	Catherine Fiore	H-mode started but was not steady. Disruption. Dalsa diagnostic seems to have recovered from problems it had on shot 2.
Jul 2 2003 08:44:35:720AM	1030702004	Catherine Fiore	H-mode still ratty: RF faulting.
Jul 2 2003 09:02:43:577AM	1030702005	Catherine Fiore	H-mode started then the plasma disrupted. Injection of some sort.
Jul 2 2003 09:07:40:540AM	1030702006	Catherine Fiore	H-mode started then the plasma disrupted again.
Jul 2 2003 09:27:08:537AM	1030702007	Catherine Fiore	No plasma.
Jul 2 2003 10:19:20:383AM	1030702008	Catherine Fiore	H-mode, with some rf dropouts. Looks like an ITB was forming before an injection and rf dropout. We will try to get a clean version of this.
Jul 2 2003 10:19:53:523AM	1030702009	Catherine Fiore	H-mode, still with rf dropouts. ITB formed at 1 s and lasted until the big rf dropout at 1.3 s. We will try again to clean it up.
Jul 2 2003 10:34:33:187AM	1030702010	Catherine Fiore	Two nice H-modes which drop out with the rf dropouts. ITB forms in both. We will back off slightly on the rf power to try and make a clean pulse.
Jul 2 2003 10:52:02:620AM	1030702011	Catherine Fiore	Long H-mode with weak ITB plagued with small drop outs and EN modes. We will start trying the Tf ramp experiments on the next shot.
Jul 2 2003 11:00:20:943AM	1030702012	Catherine Fiore	ITB starts, but a big rf drop out ended it when the toroidal field reached 4.6 T, making interpretation difficult. We will repeat.
Jul 2 2003 11:16:37:313AM	1030702013	Catherine Fiore	Two ITBs, the 2nd coinciding with the toroidal field ramp. An H-mode back transition again makes interpretation difficult. We will try again.
Jul 2 2003 11:32:22:463AM	1030702014	Catherine Fiore	One ITB, still with a back transition problem during the toroidal field ramp. We are going to adjust the target density.
Jul 2 2003 11:43:28:387AM	1030702015	Catherine Fiore	H-mode started later because of the increase in the target density. One ITB which formed as the field was ramping up, did not seem to care that the field was increasing, thus giving us a higher field point for making an ITB than we have previously achieved with the resonance on the high field side. We will try ramping the field up faster and slightly further.
Jul 2 2003 11:54:44:090AM	1030702016	Catherine Fiore	The late H-mode start makes interpreting the apparent beginning of an ITB hard to interpret. It is possible that the ITB rise was clamped due to the Tf increase, but we need more time for the ITB to fully develop in order to be positive about this. We will back off slightly on the target density.
Jul 2 2003 12:03:23:243PM	1030702017	Catherine Fiore	Frequent H to L back transitions prevented good ITB development on this shot.
Jul 2 2003 12:30:13:037PM	1030702018	Catherine Fiore	ITB forms, with the density remaining peaked through the back transition at 1.05 s. It looks like the peaking clamps at 1.2s when the toroidal field reaches 4.6 T, but a back transition quickly destroys the ITB. We are next doing 2 no rf gas puffing shots as requested by the FRCECE group.
Jul 2 2003 12:38:03:563PM	1030702019	Catherine Fiore	Reference shot 1 for FRCECE. No rf, no ITB
Jul 2 2003 12:47:54:000PM	1030702020	Catherine Fiore	No plasma
Jul 2 2003 01:00:34:770PM	1030702021	Catherine Fiore	FRCECE reference shot #2
Jul 2 2003 05:41:01:930PM	1030702029	Peter Politzer	look closely at this shot. voltage gets to and holds at zero. good rf pulse.

Physics Operator Comments
Jul 1 2003 05:32:26:567PM		Steve Wolfe	Setting up PCS for Wed. First part is ITB Studies: MP#352, 303, and part of 300 SL: Fiore PO: Wolfe Engineering setup ----------------- overnight ECDC in D2, stopping 1-2 hours prior to the run. Run begins at 8:00 and this portion ends at 14:00. Power system setup as on 1021010004 Gas setup - fill B-top with 6 psi of D2 Hybrid enabled fill B-side lower with 1 psi of Ar Hybrid enabled fill B-side upper with 10 psi of D2 Hybrid Disabled (for ECDC) fill C-side with 30 psi of D2 Hybrid enabled The following gate valves should be enabled, assuming no vacuum problems: ECE, VUV, DNB RF setup: in RUN ON by 7:00AM J, D and E , pulse at 0.6s to 1.6s, 1.5 to 2 Mw. ------------------------------------------------------------------------------- Run plan: First RF plasma shot, load from 1021010004 (0.8 MA, 4.5 T). The purpose of this experiment is two-fold: find the effect on the ITB if the field is ramped up beyond the point where and ITB is established after an ITB is formed. Secondly, it is to provide an ITB target with sufficient density to have the FRCECE diagnostic at or near cutoff in order to study probable density fluctuations. First, set up an EDA H-mode ITB plasma with 1.5-2 MW off axis ICRf at 4.5 T. Once the ITB timing is known, add a toroidal field ramp, initially to 4.6 T to study the effect of the ITB. If the ITB is not extinguished, increase the ramp to 4.7 T and determine where the ITB cutoff is, if indeed it is extinguised by rising field. Two shots are needed without ITB (slightly higher target field) to provide a non-ITB baseline for the FRCECE Depending on results, either increase the current or the RF power or both to try to improve ITB performance. If compatible with the ITB experiment, begin developing the constant psi0 control for the second half of the day during rampdowns.
Jul 1 2003 05:57:53:450PM		Steve Wolfe	Import Seg2 from 1021010004 (0.8MA, 4.5T ITB) Fixup V_10 and bolofeed wire9, as indicated in ack widget. Waveforms look pretty vanilla, except the pg3 waveform in seg2 is all -6V to 0V, i.e. no maint. gas. Leave it for now, we'll see how it runs. Put a psi0 predictor and controller on wire10, which has a tau_d of 10, with no gain. Leave waveform at 0. so I'll see the value on the phys_aout scope (give or take a sign). If it doesn't interfere with the experiment, I'll test this out in rampdown, to get a head start on part 2. Do a build_all. I expected some timestamp problems, but didn't get any. Either jas has fixed it or they only show up in load. Save this shot as /home/wolfe/pcs_trees/pcs_1030702100, just in case I screw it up setting up the part 2 shot.
Jul 1 2003 05:58:51:590PM		Steve Wolfe	Now start setting up the target for the bootstrap run. Engineering setup for Part 2 of Wed. 7/2/2003 MP#344B (or A, if B isn't available) 100% Bootstrap Plasmas SL: Politzer, Perkins PO: Wolfe Power System as on 1030530024 Except: All invert times moved to 2.0sec Fizzle Detector off time to 1.9sec Gas setup - fill B-top with 6 psi of D2 Hybrid enabled fill B-side lower with 1 psi of Ar Hybrid enabled fill B-side upper with 10 psi of D2 Hybrid Disabled (for ECDC) fill C-side with 30 psi of D2 Hybrid enabled The following gate valves should be enabled, assuming no vacuum problems: ECE, VUV, DNB ------------------------------------------------------------------------- RF for first shot: pulse from 0.6 to 2.0sec, 3MW Using whatever combination of transmitters seems best to the RF Group. PCS Setup --------- Load seg2 from 1030530024 and extend all shape,density,TF, etc. traces to 5.0sec (just in case we get there). Make sure all predictor and controller switches are moved late. Put psi0 control on a suitable (non-Ip referenced) wire, and once the shot is established, turn it on at 0.55sec, using derivative gain and perhaps P gain with a flat drawn waveform corresponding to the value of psi at that time in the target shot. Run Plan -------- Step 1: Setup shot like 1030530024 (600kA, target density nl04=6e19 m^{-3}) with all shape parameters and tf extended to 2sec. Turn on 3-4MW RF at 0.6sec and switch from Ip to psi0 (derivative) control at .7sec. If necessary, adjust timing to avoid OH1 crossover. Tune up psi0 gains and adjust RF tuning trying for H-mode plasma with Vsurf=0, RF staying on, and gaps and shape approximately constant. Get MSE data at initial and late times. Note evolution of sawteeth on ECE (or Soft Xrays). Monitor radiated power, strike point temperatures and antenna thermocouples. Engineers should monitor I2t signals, magnet heating, etc. as on the long pulse runs 1010802 and 1010803. [5-10 shots] Step 2: Extend pulse length in 0.5sec increments (or more as indicated) if and when plasma and RF are surviving that long. Try to remember to extend timebases on key diagnostics. If we go beyond 3sec, delay Thomson Scattering to insure getting profiles at the end. Keep moving the DNB pulse later as the pulse length is extended. Monitor TS profiles and VB array for signs of density peaking. Continue to monitor heating of walls, antennas, magnets. As current drops, look for indications of fast ion loss. We are looking for something that looks like a steady-state with finite current and zero loop voltage. See if MSE can tell whether angles are still changing at late times. [5 shots] Step 3: If it looks like there will be time, optimize by varying target density +/-25% and/or RF power 2-4MW; if density is falling with current we may want to puff into the H-mode to maintain slowing down time short enough to avoid tail ion loss. [5 shots] Step 4: Document best plasma. Assuming we have something that looks like an interesting (steady or near-steady) final state, verify that we have MSE data at the end, repeat the early state MSE, and fill in intermediate points as possible. [2-4 shots] Follow-up scans (next time, or if all works wonderfully well the first time): Step 5. Power scan. See what happens at higher power, vary current at which RF is turned on or increased. Step 6. Density scan. Extend scan started in step 3, at 4 MW ICRF. Step 7. ITB control. If an ITB occurs, try to control peaking by varying B during the shot or varying power. Otherwise, we will likely want to repeat the experiments starting from an ITB discharge to take advantage of the higher bootstrap current associated with grad n relative to grad T.
Jul 1 2003 06:47:46:137PM		Steve Wolfe	Import into seg3 (I know I said 2 in the run plan) from seg2 of 1030530024 Change it's name to "0.6MA, SN bootstrap target" Extend all P traces (except Ip) to 5sec; use values around 1sec in cases where the waveform was varying over nominal flattop. Copy the psi0 wire from seg2 into seg3; change the time of the clear predictor from 3 to 10sec, and extend the waveform to 5sec. extend v traces to 5sec, change tf from -157kA to -153kA, set oh voltages to 0. extend gain traces except for ip, which I'll zero at .7; turn on psi0 at D=5.0 at 0.7sec Check tht the controllers stay on. Turn off seg2, on seg3, build_all Turn off seg3 on seg2, build_all Save as /home/wolfe/pcs_trees/pcs_1030702200
Jul 2 2003 06:26:11:633AM		Steve Wolfe	Setting up PCS for Wed. July 2 Load_from /home/wolfe/pcs_trees/ shot=1030702200 Load/clean Load at 2-Jul-2003 06:25:06.00 Open tree /home/wolfe/pcs_scratch -1 Open tree done
Jul 2 2003 07:23:18:830AM		Steve Wolfe	The present startup has IC_EF4U = -1305 PG4 = 26msec BR0=0.000 (which seems to be giving -.001T) RCUR offset= 3600. ; the predictor thinks rcur error=-1000 to -2000A-m before 25msec. The initial V_EF3 points are -.001 2650 .005 2600 .012 2700 .026 2200 .020 2250 .026 2200 .046 2300 I want to get rid of the hards, but I'm not sure what to change, so we'll leave it alone for the first shot anyway.
Jul 2 2003 07:51:01:373AM		Steve Wolfe	The RGA is still showing a 28 peak, indicating a small leak; it seems to be getting worse with cooling. We may need to deal with this before the day is over.
Jul 2 2003 08:03:52:567AM	1030702001	Steve Wolfe	shot#1: No power shot. Had an INIT failure on TMX camac because the C-Coil settings got left as "HYBRID" in the time fields. Changed to nominal (?) settings for next shot.
Jul 2 2003 08:15:40:677AM	1030702002	Steve Wolfe	shot#2: plasma full length. No H-mode, only got D ab=nd a ratty J for 1MW startup was still bouncy, but no hards. Next shot: increase pg4 from 26 to 30msec
Jul 2 2003 08:33:15:330AM	1030702003	Steve Wolfe	shot#3: increase pg4 from 26 to 30msec; hopefully E (and the plasma) will go this time. DALSA seems to have no signal on shot 2. plasma with H-modes disrupted at 1.3sec Had some hards early. Injection at 1.32sec killed it.
Jul 2 2003 08:44:18:033AM	1030702004	Steve Wolfe	shot#4: plasma h-modes, disrupted in rampdown 1.68. startup maybe better, but still has blips on halpha One early small burst of hards dumps at .12 sec rf still faulting (E and J, D is ok). next shot: repeat, hope for better RF
Jul 2 2003 08:57:53:710AM	1030702005	Steve Wolfe	shot#5: repeat, hope for better RF plasma, early disruption at .72sec All the RF's were on. Hards are back. next shot: ic_ef4u from -1305 to -1290
Jul 2 2003 09:11:57:030AM	1030702006	Steve Wolfe	shot#6: ic_ef4u from -1305 to -1290 plasma disrupts at .95sec. No hards this time Next shot increase outer gap.
Jul 2 2003 09:36:15:350AM	1030702007	Steve Wolfe	shot#7: increase outer gap. rcur negative by 3mm fizzle. Got to 50kA at 13msec. Looking at shot 6 harder, it looks like the hybrid signal in the ef2u supply has junk on the hybrid demand signal that isn't on the hybrid signal in the interface room. However, the jumps are not coincident with RF faults. Try turning off J for next shot. If that doesn't help, maybe try swapping the hybrid acquisition and live outputs?
Jul 2 2003 09:43:01:483AM	1030702007	Steve Wolfe	shot#8: Try turning off J. If that doesn't help, maybe try swapping the hybrid acquisition and live outputs? This will be a no-power shot: data acq cycle is hung up
Jul 2 2003 10:03:39:950AM	1030702008	Steve Wolfe	shot#8 (second take): Take out the gap change (3mm rcur) that was put in for shot 8 before, since this is a supply or hybrid problem. Try turning off J. If that doesn't help, maybe try swapping the hybrid acquisition and live outputs? plasma full length. Looks like the beginnings of an ITB, after some RF faulting. bouncy startup almost didn't make it, but no hards. No sign of the glitches on the EF2U supply signal, which doesn't prove much, since we had clean shots earlier; however, all the disruptions show the effect. next shot: repeat
Jul 2 2003 10:17:59:927AM	1030702009	Steve Wolfe	shot#9: repeat plasma full length. ITB forms after an initial H-L that seems to be radiation-related. Later RF faulting may not be hurting, but they don't help. Hards are back, big bounce in startup. next shot: seg1 rcur offset from 3600 to 5000.
Jul 2 2003 10:34:37:953AM	1030702010	Steve Wolfe	shot#10: seg1 rcur offset from 3600 to 5000. plasma full length. Startup looks better, no turnaround or even hesitation. No hards. looks like we're getting ITB's but not keeping them. next shot: drop rf power 200kW (100 from each) to try to get them to run through
Jul 2 2003 10:41:25:277AM	1030702011	Steve Wolfe	shot#11: drop rf power 200kW (100 from each) to try to get them to run through plasma full length - the circumstantial evidence against J in the EF2U interference is building up. ITB not as strong, still faulting on E but not on D startup still looks good, no hards. next shot: field ramp
Jul 2 2003 10:57:25:723AM	1030702012	Steve Wolfe	shot#12:field ramp from -124kA at 1.1 to -126.8 at 1.3 to -129.6 at 1.5 plasma full length, just a hint of hesitation in early current rise. got the field ramp next shot: repeat
Jul 2 2003 11:15:09:767AM	1030702013	Steve Wolfe	shot#13: repeat plasma full length. two rf dropouts, just at the wrong time. next shot: RF will fire j3 into dummy load during rampdown to test ef2 effect. drop pg4 from 30 to 28msec otherwise repeat
Jul 2 2003 11:28:17:957AM	1030702014	Steve Wolfe	shot#14: pg4 from 30 to 28msec J3 into dummy load in rampdown plasma disrupts at 1.48sec, the ef2U glitch has returned. Still 2 dropouts, .98 and 1.2, which really screws up the ITB data. nest shot: raise target by 1e19
Jul 2 2003 11:40:37:867AM	1030702015	Steve Wolfe	shot#15: raise target by from .85 to .95e20 J4 into dummy load in rampdown this time plasma full length next shot: increase tf ramp
Jul 2 2003 11:50:17:053AM	1030702016	Steve Wolfe	shot#16: increase tf ramp to get to 135.2kA at 1.5sec J4 still in dummy load plasma full length; startup is getting slower. next shot: pg4 from 28 to 25msec J3 back into dummy load nl4 to 9e19
Jul 2 2003 12:03:20:933PM	1030702017	Steve Wolfe	shot#17: pg4 from 28 to 25msec J3 back into dummy load from 1.6 to 1.8 nl4 to 9e19 (from 9.5) plasma full length, no interference on ef2U demand this time. H-mode is earlier, 0.9sec, but multiple H-L transitions. next shot: down to 8e19
Jul 2 2003 12:21:40:260PM	1030702018	Steve Wolfe	shot#17: down to 8e19. delay for false I2t reading to get cleared plasma, full length. looks like an ITB next shot: Turn off RF, take comparison shots for Lynn and Phillips.
Jul 2 2003 12:40:40:897PM	1030702019	Steve Wolfe	shot#19: program nl4 to 1.7e20 starting at .8 to 1.5 plasma disrupted at 1.74, not from ef2 but just from the density density was ramping up pretty linearly. next shot: No Bt ramp, get density of 1.8e20 by 1.1sec
Jul 2 2003 12:45:10:137PM	1030702020	Steve Wolfe	shot#20: No Bt ramp, get density of 1.8e20 by 1.1sec fizzle next shot: drop the prefill (should have done it last time).
Jul 2 2003 01:02:33:977PM	1030702021	Steve Wolfe	shot#21: drop pg4 to 22 from 25msec plasma disrupted at 1.7sec, they got their density. That concludes part 1. Now we go to the bootstrap run.
Jul 2 2003 01:29:09:660PM	1030702022	Steve Wolfe	shot#22: Beginning bootstrap run. Turn off seg 2, turn on seg1 psi0 gain turns on at .7, RF on at .6 plasma ran down to 200kA at 1.5sec, no RF (my error) deriv(psi0) is definitely not zero, it seems to continue ramping until very late. Probably not enough deriv gain. 0.7 sec is after crossover, but it may not be with RF, so don't turn on P gain yet.
Jul 2 2003 01:53:01:470PM	1030702023	Steve Wolfe	shot#23: Raise D gain from 5 to 9. According to ihh "physics gain" the D gain is only 0.9, which is way too small. Try using a second wire with different tau_d On wire 11 now as well; that claims a factor of 10 smaller D gain, as expected. If it's right it won't help much. Oh, I've still got a voltage on the V_oh1 trace! kill it, and the oh2 and ef2 voltage as well. The fizzle detector really did take out shot 22, I'll need to adjust it but first let's try to get the control going. plasma to 1.6sec, f.d. killed it again. current acyually rises after 0.7sec, then decays. Got rf, no H-mode? big injection at 0.125, almost lost it then. Vsurf never goes below .3V, and it's limited on the inside. This isn't working with D gain. I'm going to try holding psi0 with P gain as well, have to draw a value of +50000 in ihh units.
Jul 2 2003 02:12:16:283PM	1030702024	Steve Wolfe	shot#24: Draw psi0 = 50000, flat , and P gain at 3.0 until 3 sec. I don't see why it's still ramping the oh plasma, only to 1.3 sec this time. It drops to .5MA, holds flat, then ramps down after 1sec Somehow, segment 3 has been on as Async instead of sync. I have no idea what that is going to do, possibly nothing? We may have been running on seg 1 the whole time. I see no evidence of signals changing on the pid out at the right times next shot: reload it, with it on sync this time.
Jul 2 2003 02:19:54:920PM	1030702025	Steve Wolfe	shot#25: reload it, with seg 3 on sync this time. plasma disrupted at .77, curent ramped up, probably the P gain to the wrong flux. Yes, it was at 140000 ians, and I told it to go to 50000 in no time. next shot: back to all D gain
Jul 2 2003 02:38:52:777PM	1030702026	Steve Wolfe	shot#26: back to all D gain, drop wire 10 D from 9 to 2, turn off 11 plasma, runs down to 96kA at 1.6sec, Vsurf drops to .2V Inner gap is growing to 2.5cm. It's going the right way, but it looks like oh1 is right at crossover. next shot: move turn on of vloop control from .7 to .45 sec.
Jul 2 2003 02:51:25:843PM	1030702027	Steve Wolfe	shot #27: Change control swich time from .7 to .45 sec to avoid crossover. plasma down to 140kA at 1.6sec Still makes it into crossover, voltage still not zero RF at about 1.5MW, stays on better. next shot: raise D10 from 1 to 3, also zero the feedforward at .45.
Jul 2 2003 03:15:58:913PM	1030702028	Steve Wolfe	shot #28: raise D10 from 2 to 6, also zero the feedforward at .45. plasma disrupted at .53 sec Definitely went unstable. next shot: back off on the D10 gain
Jul 2 2003 03:35:07:550PM	1030702029	Steve Wolfe	shot #29: Drop D10 gain from 6 to 3 try for more rf power (2MW total). plasma tio 1,4sec. OH1 is flat at -1900A This one dies from a vertical instability, n/nc=1.4 li is still going up. Got 2.1MW until it faults I need to change cutip to keep the EFIT's going below 200kA next shot: increase D from 3 to 4
Jul 2 2003 03:43:20:920PM	1030702030	Steve Wolfe	shot #30: increase D from 3 to 4 plasma to 1.3sec, Voltages look very hashy. lost the RF earlier, got to 2.7MW but then ran max faults. next shot: back off D, turn on P at gain of 1
Jul 2 2003 03:59:35:603PM	1030702031	Steve Wolfe	shot #31: D back to 3, turn on P at gain of 1, target=130000 disrupted at .57. I think the target shoould have been -130000 next shot: change target to -130000, drop P gain from 1 to .2
Jul 2 2003 04:17:29:943PM	1030702032	Steve Wolfe	shot #32: change target to -130000, drop P gain from 1 to .2 plasma .92 sec, got down to 200kA, voltage gets to be very close to zero. psi0 error gets to -40000 ians~ 0.3Wb. I can increase that P gain some. next shot: P gain from .2 to .4
Jul 2 2003 04:30:59:560PM	1030702033	Steve Wolfe	shot #33: P gain from .2 to .4 disrupted at .75 at 300kA. psi0 is flat at error of -30000. Try moving heating off axis. Drop tf to 5.0T
Jul 2 2003 04:41:56:323PM	1030702034	Steve Wolfe	shot #34:Try moving heating off axis. Drop tf to 5.0T plasma disrupts at .75sec next shot: RF on at 0.4 sec
Jul 2 2003 04:56:51:287PM	1030702035	Steve Wolfe	shot #35: RF on at 0.4 sec, no other changes. plasma to .75, rf has multiple trips, get multiple H-modes. injection at .14, next shot: pg4 to 24msec, sjw is retuing the RF
Jul 2 2003 05:05:09:313PM	1030702036	Steve Wolfe	shot #36: pg4 to 24msec, sjw is retuning the RF plasma to .78, rf stayed on next shot, drop P gain
Jul 2 2003 05:22:26:217PM	1030702037	Steve Wolfe	shot #37: drop P gain from .4 to .1 plasma to 1sec, 2.8MW and steady until about 300kA final is ~200kA. next shot:pg4 to 28msec, tf to 4.8T
Jul 2 2003 06:17:38:110PM	1030702038	Steve Wolfe	shot #38:pg4 to 28msec, tf to 4.8T (135kA) plasma to 1.02sec, to slightly lower current. Operator error: I changed the field in seg1 instead of seg3. next shot: zero P gain, so it looks like 29, and drop tf to 4.8
Jul 2 2003 05:49:17:363PM	1030702039	Steve Wolfe	shot #39: zero P gain, so it looks like 29, and drop tf to 4.8 plasma to 1sec, down to 150kA. li during rf is ~1.5 event at .8 sec drops stored energy, current decay next shot: raise density, and back to 5.3T
Jul 2 2003 06:11:34:197PM	1030702040	Steve Wolfe	shot #40: raise density to 7.5e19, and back to 5.3T plasma to 1.3 sec, to 200kA. target density didn't go up? I didn't change the program before .4, so we were already in H-mode and at 1.1e20 before then. This looks quite a bit like 29.

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	07:57:16:163AM	Plasma	Ok	C
2	08:10:04:840AM	Plasma	Ok
3	08:23:34:470AM	Plasma	Ok
4	08:37:19:710AM	Plasma	Ok
5	08:50:01:197AM	Plasma	Ok
6	09:02:56:577AM	Plasma	Ok
7	09:15:49:550AM	Plasma	Ok
8	09:55:06:423AM	Plasma	Ok
9	10:08:20:067AM	Plasma	Ok
10	10:21:28:823AM	Plasma	Ok
11	10:35:32:877AM	Plasma	Ok
12	10:49:56:650AM	Plasma	Ok
13	11:05:07:203AM	Plasma	Ok
14	11:18:04:863AM	Plasma	Ok
15	11:30:54:243AM	Plasma	Ok
16	11:43:30:247AM	Plasma	Ok
17	11:56:35:530AM	Plasma	Ok
18	12:16:20:580PM	Plasma	Ok
19	12:29:21:990PM	Plasma	Ok
20	12:42:20:540PM	Plasma	Ok
21	12:55:25:747PM	Plasma	Ok
22	01:17:40:000PM	Plasma	Ok
23	01:42:34:930PM	Plasma	Ok
24	01:57:35:040PM	Plasma	Ok
25	02:12:29:860PM	Plasma	Ok
26	02:27:26:007PM	Plasma	Ok
27	02:43:24:813PM	Plasma	Ok
28	03:11:32:327PM	Plasma	Ok
29	03:24:11:550PM	Plasma	Ok
30	03:38:10:180PM	Plasma	Ok
31	03:53:33:017PM	Plasma	Ok
32	04:06:09:363PM	Plasma	Ok
33	04:20:00:400PM	Plasma	Ok
34	04:34:31:580PM	Plasma	Ok
35	04:47:19:083PM	Plasma	Ok
36	05:00:08:883PM	Plasma	Ok
37	05:12:47:637PM	Plasma	Ok
38	05:25:58:767PM	Plasma	Ok
39	05:38:50:210PM	Plasma	Ok
40	05:54:12:700PM	Plasma	Ok