TRIDENT direct-drive LDRD shots
(Psychedelic Snakes),
May 2003

Contents

• Objectives
• Schedule and statistics
• Sample materials
• Targets
• Drive conditions
• Diagnostics
• Alignment
• Shot summary
• Notable events
• Experimental observations
• Evaluation
• Acknowledgements

Objectives

Primary objectives:

1. To field the "GRIM" interferometer on direct-drive materials experiments simultaneously with VISAR and sample recovery. 100% successful.
2. To obtain GRIM and VISAR data on samples of NiAl, particularly bicrystals. 100% successful.

Opportunistic:

1. To characterise the spatial uniformity of directly-driven loads using the Fresnel zone plate. 100% successful.
2. To obtain useful material response data on materials used in precursor experiments, e.g. Si. 100% successful.

Success ratings are based on what we expected to achieve during the session, not on all that could possibly be done.

Schedule and statistics

• Plan refers to TRIDENT schedule ~Apr 2003.
• We were asked ~13 May whether we could slip our start date by a week or two to help with TRIDENT staff schedules. This was actually very helpful (thanks to Tom Ortiz and Nathan Okamoto providing us with alignment and timing beams during the week of 19 May) as it allowed us to install and align the GRIM at leisure.
• The schedule was only remotely workable because we were able to leave the VISAR equipment in place from the previous series which finished before.
• In order to help accommodate the harmonic generation experiments in the optics setup lab, light was diverted from the front-end using an etalon. This tended to make a dip during the drive pulse, but was not deemed significant from the point of view of dynamic loading. (Element-to-element variations were typically of a similar order.)
• As the following series was in the north target area, we were able to dismantle and tidy our equipment at leisure during the following week.

Shot allocations
	Plan	Actual	Comments
Setup	~6-10	-	Actual: counted among data shots below.
NiAl single crystals	~8-12	3	Obtained all desired data.
NiAl bicrystals	2	6	Additional samples delivered from ASU.
Si crystals	(setup)	13
Cu	(setup)	1
RuAl	0	2	Scoping shots.
Total	~20	25

Firing duration and rate
	Expected	Actual	Comments
Experiment time	4.0	4.0
Shots/day	8.0	6.25	Preshot setup was generally more involved than expected.

• Experiment time ignores extended setup time allowed during the week before the live shots.
• Setup time is subtracted from experiment time in calculation of shot rate.

Sample materials

Targets

Figure 1: Sample holder

Samples were typically 5-8 mm in size, and the flyer holder was not suitable to hold them in place directly. Washers were punched from 250 µm Cu foil, the outer diameter being a close fit within the flyer holder and the inner diameter roughly 5 mm. Samples were fastened across the hole using thin strips of insulating tape to hold their edges, and an additional washer punched out of Cu or plastic shim stock if it was felt necessary. The sample/washer assembly was then held against one face of the flyer holder by packing the holder with plastic rings.

Figure 2: Mounted sample (NiAl bicrystal B6 post-shot; scale: cm)

Drive conditions

The flyer holder obscured the view from the east alignment telescope, which is the usual way of checking alignment before a direct drive shot. Alignment was set approximately before the first shot of the day by replacing the target holder with a 2" diameter disc with a hole a couple of mm diameter in the center. This was not quite representative of real samples, and the center of the drive was sometimes appreciably off-center in terms of shock arrival (slight lag at one edge) and surface displacement (off-center dome shape).

Diagnostics

The optical layout used was based on the standard VISAR layout, adjusted slightly to allow GRIM access through the west port. There was some conflict in setting up the diagnostics before each shot, as the adjustments required for the GRIM necessitated blocking the VISAR beams for a while.

Figure 3: Experimental layout (GRIM shown in simplified form).

Figure 4: Setup inside target chamber (view from S and above).

Line VISAR

The camera slits were set to 300 µm.

Fringe constant

Timing relative to drive beam

The timing offsets for the west streak camera had been established accurately by Tom Tierney for the preceding Be melt series, following the previous procedure of passing an element of the drive pulse through the VISAR optics to the camera, and adjusting until the element appeared close to the center of the camera. The east streak camera was brought up during the experiments, and there was no opportunity to time it in using the same procedure. The delay was established approximately by detecting the peak in the probe laser pulse, and then refined for the 50 ns sweep period by adjusting it slightly to record the arrival of a shock wave during one of the experiments. This proved to be an adequate way to establish timing, though the detection of a drive beam element is preferable in general.

VISAR streak camera delays
sweep period	W delay	E delay
(ns)	(ns)	(ns)
50	2762	2777 (est)
20	2811
10	2824.5
5	2834.5

Delays are for element 7 to be nominally at center of record. The Q-switch channel delay was 2450 ns.

Spatial scale

Fiducial markers

GRIM

The GRIM is a Mach-Zehnder type interferometer in which the sample is located on one arm and is imaged to the recording camera. Light reflected from the sample interferes with light from a reference arm, with a deliberate misalignment introduced (as with the line-VISAR) so that a pattern of fringes is produced. The path length and lenses were duplicated in the reference arm to allow straight fringes to be set up. The fringe pattern was recorded in 2D using a framing camera. A spatial carrier frequency algorithm was used to convert perturbations in the fringes to spatial displacements; for this algorithm to be successful, straight fringes of high contrast were needed, preferably with a small spatial scale. The GRIM design included balance controls to correct for targets of different reflectivity. A Joulemeter was used to estimate the energy in each arm, and to check the power level before exposing the camera.

Several cameras were considered for recording the GRIM images. We eventually managed to borrow a Hadland Imacon 200 system capturing 8 frames. This camera was driven internally with a 200 MHz clock: the minimum (nominal) gate time was 5 ns, as was the unit in which frame delay times could be varied. The frame start time was triggered on the next clock pulse after the trigger signal was received, so there was jitter of in principle 0 to 5 ns later than the specified frame time. 5 ns gate time would be too long for samples moving at several hundred m/s: the GRIM fringes would be excessively blurred. We used short illumination pulses to freeze the fringe pattern on a shorter time scale. A HeNe beam was used for static alignment; a second set of images was collected from each shot using the HeNe light after an internal ~150 ms.

The spatial scale was established by imaging a Ronchi ruling of 1000 lines per in. The optical system rotated the image of the sample by 90^o. In the sample plane, the horizontal and vertical scales were 1.8 and 1.5 µm/pixel, respectively. The image dimensions were 1022x1289 pixels, so the field-of-view was 1.8 x 1.9 mm (horizontal x vertical).

TRIDENT's "holography" (H) beam was used to generate a train of 200 ps pulses at 527 nm wavelength (green). A single pulse was rattled from a mirror and amplified. The pulses were synchronized to the main oscillator for the drive beam; the pulse train was amplified using an envelope whose timing was adjusted optically with reference to the drive beam. The system gave pulses at intervals of 6.6 ns; a value chosen because the laser was previously set up this way. It would be possible to increase the separation up to ~100 ns with some trouble. At 6.6 ns, the pulse spacing matched the inter-frame time of the Imacon quite well - as far as we know we never skipped an image because of mistimed pulses. A gated optical imager (GOI) could be used for pulses of shorter separation.

With the camera gain set low, the full H beam energy (~1 mJ in the pulse train) was needed to obtain a record, equivalent to ~5 µJ at the camera. With higher gain an order of magnitude less light was needed. There is scope to make the optical design more efficient if needed in future.

Timings were established and monitored by capturing trigger signals and photodiode output (from the drive and H beams) on an oscilloscope.

The GRIM was triggered from the user DG535 unit in the target area, with the signal fed to another DG535 which triggered the camera itself and the monitoring oscilloscope. The user DG535 was set to 2765 ns and the second DG535 to 0. With these settings, the drive pulse occurred on the oscilloscope at approx 200 ns, measured using a photodiode. The camera frame times were then set using the Imacon's control computer.

A triggering problem was encountered in the Imacon and the LeCroy monitoring oscilloscope. The Imacon periodically failed to produce its "Trigger Out" and "Monitor" outputs, although the images were successfully acquired. Additionally, the LeCroy scope sometimes failed to trigger on its DG535 trigger. Consequentially, many shots had either incomplete or non-existent scope records. However, most of these shots had "setup shot" records (without the drive beam).

We had been concerned that stray light from the drive beam might swamp the GRIM, as both were operated at the same wavelength. We tried to avoid exposing GRIM frames around the time of the drive pulse; in some shots this was difficult to achieve. In the event, we experienced little or no problem with crosstalk. It is likely that the relatively massive target holder helped in this respect.

Alignment

The system was set up so that the diagnostic surface of the sample was nominally at target chamber center, as defined by alignment telescopes to the east of and below the target, as had been aligned precisely in the preceding Be melt experiments. (The design of the sample holder made it impossible to use the alignment cameras on each experiment, so the effective center of the experiment drifted during the course of the series.) The nominal axis of the sample was defined by the initial orientation of the line VISAR. The GRIM beams were then aligned approximately to a sample in the field of view of the VISAR, and the GRIM was adjusted to give satisfactory fringes. At this point, the VISAR was realigned to the target center as defined by the GRIM. Successively smaller spatial features (letters on a resolution chart) were then imaged using both diagnostics making minor adjustments, to establish a common center.

Ideally, the sample holder would have been aligned so that the translation axes were perfectly parallel or orthogonal to the VISAR axis. Because of the finite precision of alignment, and the repositioning of the VISAR following GRIM alignment, this was done to a finite accuracy. Further, the components of the sample holder (particularly the Cu washers) were made to "eyeball" tolerance, and the samples were thus imperfectly centered in the holder. As a result, adjustments such as tilting or translation caused the sample to move out of focus or away from the center of the field of view of one or both diagnostics. Multiple adjustments were needed to establish an adequate working alignment for some shots, and the effective center of the experiment moved around during the series.

The illumination line of the VISAR was visible on preshot GRIM pictures (except for unmarked Si), allowing the relative alignment to be verified before most shots.

Alignment of the bicrystal boundary was a challenge in the previous LDRD series with VISAR only (Guinness), and was expected to be a major problem this time. In fact the modified VISAR optical system (including brighter probe laser) and presence of the GRIM made alignment significantly easier. We managed to align on scratches or ink marks. Scratches or other surface features were more convenient to check focusing. Ink marks could cause problems because they could obscure significant areas of the GRIM frames; however, it was possible to remove the ink marks in situ after alignment using a cotton bud soaked in ethanol. Ink marks are thus recommended as convenient without being complicated to lay down or damaging to the surface of the sample.

With the optical setup used in this series, the VISAR had a relatively high magnification and the GRIM a relatively large field of view, of which a subset was recorded with the framing camera. This was convenient as it allowed some flexibility in alignment, compensating for deficiencies in the sample holder and positioning system, though it was wasteful in GRIM illumination.

Shot record

• Target stack listed from drive side.
• Drive energy measured by calorimetry, as usual where the beam enters the target area.
• Aluminised Si: Al layer ~1000 Å
• Sample recovered unless stated.

shot	target	VISAR					GRIM	Drive		Comments
		W sweep	W delay	E sweep	E delay	laser delay	delay	duration	energy
		(ns)	(ns)	(ns)	(ns)	(ns)	(ns)	(ns)	(J)
27 May
15872	Si (100) 32 µm/Al	50	2765			2460	125, 140, 145, 150, ..., 170	2.5	5.6	GRIM exposures 5 ns except for first (10 ns); gain 2 for first 3 frames then 5. VISAR signal low (slit too narrow / misaligned) but saw shock then reflectivity dropped (Al layer separated?); GRIM: saw fringes before, poss after, then disappeared at late time. Sample destroyed.
15873	Al/Si (100) 32 µm	50	2765			2460	125, 140, 145, 150, ..., 170	2.5	10	Lumpy drive, reasonable VISAR; squirly GRIM in later frames. Sample destroyed.
15874	NiAl (100) #11 303 µm	50	2807			2505	170, 185, 190, 195, ..., 215	2.5	8	VISAR jump; GRIM little effect
15875	Al/Si (100) 32 µm	50	2765			2460	125, 140, 145, 150, ..., 170	2.5	10	Beautiful VISAR, GRIM squirly after breakout (missing frame closest). Sample destroyed.
15876	Si (100) 32 µm	50	2765			2460	125, 140, 8e6, 145, ..., 165	2.5	22	Pretty nice VISAR, GRIM squirly after breakout. Sample destroyed.
15877	Si (100) 655 µm	50	2835			2530	190, 210, 8e6, 215, ..., 235	2.5	21	Beautiful VISAR, GRIM showed slight bowing after breakout. Sample recovered in pieces.
15878	Si (100) 655 µm	50	2835			2530	190, 210, 8e6, 215, ..., 235	2.5	74	Good VISAR and GRIM; GRIM showed dark "seeds". Sample destroyed.
28 May
15880	NiAl (100) #10 307 µm	50	2807			2505	170, 190, 220, 195, 200, ..., 215	2.5	34	Good VISAR, not much difference between GRIM frames.
15881	NiAl (100) #3 257 µm | (110) #6 253 µm	50	2797	50	2827	2495	150, 170, 200, 175, 180, ..., 195	2.5	41	Good VISAR (E late); not much difference between GRIM frames.
15882	Si (100) 396 µm	50	2797	50	2812	2495	150, 170, 200, 175, 180, ..., 195	2.5	31	Sample scribed with "Z" on drive side. Good VISAR (E in time); not much difference between GRIM frames; no obvious effect from Z. Sample destroyed.
15883	NiAl bicrystal set 2a #1 224 µm	20	2843	50	2809	2492	140, 155, 185, 160, 165, ..., 180	2.5	16	good VISARs, good GRIM
15884	NiAl bicrystal set 2a #2 245 µm	20	2843	50	2809	2492	140, 155, 185, 160, 165, ..., 180	2.5	17	good VISARs and GRIM; GRIM fov relatively obscured by alignment blobs
15885	Cu foil 233 ± 4 µm	20	2870	50	2836	2519	170, 185, 215, 190, 195, ..., 210	2.5	46	W VISAR: OK but early in frame; E missed. GRIM looks good.
29 May
15887	Si (100) 396 µm	20	2822	50	2856	2505	150, 170, 200, 175, 180, ..., 195	2.5	86	VISARs and GRIM good, few measles late in time nr alignment scratch.
15888	NiAl bicrystal B5 260 µm	20	2856	50	2822	2505	150, 170, 200, 175, 180, ..., 195	2.5	24	Wiped ink marks off pre-shot. W VISAR late, E fine (saw spatial variation); GRIM good.
15889	NiAl bicrystal B6 288 µm	20	2851	50	2812	2500	145, 165, 195, 170, 175, ..., 190	2.5	182	Good VISAR and GRIM; sample disappeared but shield survived(!)
15890	NiAl bicrystal set 3 #4 (?) 210 µm	20	2841	50	2807	2490	145, 155, 215, 165, 170, ..., 205	2.5	24	Sample size only slightly greater than hole in Cu holding ring. Good VISAR and GRIM; sample recovered.
15891	Si (111) 401 ± 2 µm	20	2846	50	2817	2500	150, 170, 200, 175, 180, ..., 195	2.5	26	Good VISAR and GRIM (W late in frame); inverse measles at late time. Edges of sample recovered in small pieces.
15892	NiAl bicrystal set 3 #1 (?) 210 µm	20	2841	50	2807	2490	145, 155, 215, 165, 170, ..., 205	2.5	20	Sample a push-fit in Cu holding ring. Good VISAR and GRIM; sample recovered.
30 May
15894	Si (111) 401 ± 2 µm	20	2851	50	2817	2500	150, 170, 200, 175, 180, ..., 195	2.5	85	Good VISARs and GRIM; black measles.
15895	Si (111) 401 ± 2 µm	20	2851	50	2817	2500	150, 170, 210, 175, 180, 185, 195, 205	2.5	14	Good VISARs and GRIM.
15896	Si (111) 401 ± 2 µm	20	2851	50	2817	2500	150, 170, 210, 175, 180, 185, 195, 205	2.5	55	Probe laser multi-moded; W VISAR may be usable. GRIM good.
15897	RuAl 188 µm	50	2782	50	2822	2490	150, 170, 210, 175, 180, 185, 195, 205	2.5	15	Jump visible on W VISAR, ~25 ns transit. GRIM good.
15898	RuAl 217 µm	20	2836	50	2802	2490	150, 170, 210, 175, 180, 185, 195, 205	2.5	189	VISARs and GRIM good; elastic precursor and shock; sample recovered apart from region in centre.
15899	Si (111) 401 ± 2 µm / LiF 2 mm	20	2851	50	2817	2500	150, 170, 210, 175, 180, 185, 195, 205	2.5	15	Fringes between Si and window. Good VISARs and GRIM. Sample recovered totally intact.

Notable events

1. 27 May: FedEx deliver bicrystal samples from ASU; packaging crushed in transit so samples mixed together; some doubt over identification of samples in set 3.
2. 29 May: Leak in turbo pump: chamber evacuation was slower than usual.

Experimental observations

1. In (100) Si, black specks ~100 µm across appreared in the GRIM tens of nanoseconds after shock breakout, initially as a reflectivity change but subsequently becoming a change in relief. In (111) Si, the speckles appeared lighter than the bulk for some energies. The formation and evolution varied with drive energy.
2. Drive energy was relatively difficult to control at 2.5 ns (compared with 1 ns); variations of order tens of percent though pulse shapes were fairly good.
3. The second RuAl sample probably had a hole near one edge of the GRIM frame (the material was reportedly porous); a bright region was visible supposedly tens of ns after the end of the drive pulse. We suggest this is evidence of a low power tail following the nominal pulse.
4. Imacon was inconvenient for setting GRIM contrast as the rate at which profiles could be obtained was many seconds. Suggested improvements: Photometrics camera used for TIM TA-46 using beam split from near Imacon; improved Imacon driving software e.g. Labview interface.
5. Rebuilt VISAR probe laser was much more stable and brighter than previously.
6. The alignment system and target holder were not ideal and could usefully be redesigned for future experiments.
7. There were inefficiencies in the order of alignments and tests before each shot, probably losing us a few shots over the week. A procedure could usefully be developed to improve efficiency, based on our practical experience from this series.
8. The contrast in the GRIM fringes was better at longer spatial wavelength, indicating a source of spatial blur somewhere in the recording system.
9. Grain boundaries were generally visible in both VISAR and GRIM (clearer).
10. The sample surface remained flat and reflective enough for GRIM fringes, even following shock breakout on the highest energy shots.
11. The GRIM was tolerant to adjustments in beam pointing necessary to cope with wander in the effective center of the experiment as the series progressed.
12. It would be desirable in future shots to include temporal fiducials on all the diagnostics from a common master signal, synchronized to the drive beam.

Evaluation

1. "To integrate the GRIM diagnostic with TRIDENT direct-drive VISAR and sample recovery experiments."

   The GRIM and VISAR were integrated with no major difficulties. The optical paths were physically separate, the GRIM looking off-normal. This made alignment somewhat awkward, but future improvements in the design of the target holder and positioning assembly should make life easier. Sample recovery by edge clamping worked as before. We also managed to obtain satisfactory GRIM and VISAR signals from Si through a LiF window (indicating that the sample surface was adequately reflective); the presence of the window reduces the amount of 2D flow induced by deceleration, compared with edge clamping.

2. "To obtain simultaneous GRIM and VISAR data on NiAl bicrystals."

   Simultaneous GRIM and VISAR records were obtained from 6 bicrystals, with apparently adequate alignment on the boundary. The samples were recovered, though there had been insufficient time for the degree of preshot characterization desired. These shots included a pair of samples of (probably) identical orientations, shocked from opposite sides, and thus providing potentially the first direct test of the energy localization concept which is the key process underlying this project. A third sample was retained unshocked for a posteriori characterization.

Secondary objectives:

1. "To characterise the spatial uniformity of directly-driven loads using the Fresnel zone plate."

   We obtained VISAR and GRIM data showing simultaneous shock arrival to the precision of the diagnostics (best: VISAR, ~50 ps in 50 ns) except for shots where the drive beam was noticeably misaligned; initial velocity was uniform to within the limit possible by eyeballing fringes i.e. probably ~100 m/s; Small-scale (~100 µm) variations of displacement were apparent with amplitudes ~0.1 µm after ~10 ns, presumably caused by the speckle pattern in the drive beam. Large-scale ~2 mm target bowing of amplitude ~10 µm appeared after several tens of ns, presumably from the spatial envelope of the drive beam or because the samples were held by their edges. It would be useful to combine measurements of this type with beam profiling in the future.

2. "To obtain useful material response data on materials used in precursor experiments, e.g. Si."

   Clean shock/release VISAR records were obtained for Si (100) and (111), Cu, and RuAl. Corresponding GRIM records were obtained, though the RuAl surface may have been too rough for the analysis scheme to handle well. Interesting spatial variations were observed in Si at relatively late times.

Acknowledgements

Distribution