Overview of Project Orion

Transcription

1 Overview of Project Orion Nicholas W. Hopps, Thomas H. Bett, Nicholas Cann, Colin N. Danson, Stuart J. Duffield, David A. Egan, Stephen P. Elsmere, Mark T. Girling, Ewan J. Harvey, David I. Hillier, David J. Hoarty, Paul M. R. Jinks, Michael J. Norman, Stefan J. F. Parker, Paul A. Treadwell, David N. Winter. AWE, Aldermaston, Berkshire, United Kingdom. RG7 4PR ABSTRACT Project Orion will provide a facility for performing high energy density plasma physics experiments at AWE. The laser consists of ten, nanosecond beam lines delivering a total of 5kJ with 0.1-5ns temporally shaped pulses and two short pulse beam lines, each producing 500J in 0.5ps with intensity > 10^21 W/cm^2. The performance of the Orion laser is reported as the first phase of commissioning (one short and one long pulse beam) concludes. Target shots with all beam lines will begin in Keywords: High power lasers, Nd:glass lasers, OPCPA, high energy density physics. 1. INTRODUCTION Construction has recently been completed for the Orion laser facility at AWE in the UK. The first two beam lines have been commissioned with shots onto targets. Orion, which is to be used for high energy density plasma physics experiments, consists of ten long pulse beams, and two short pulse beams, with the following baseline performance: the long pulse beams each deliver 500J at 351nm in a user-variable pulse length from 100ps to 5ns; the short pulse beams each deliver 500J around 1054nm with a pulse length adjustable between 0.5ps and 20ps. The experimental strategy for Orion in the early years of operation is to acquire data to benchmark and underwrite the material modelling capability at AWE. The case for Orion was made on the basis that short pulse laser interactions allow states of matter at extremes of temperature and density to be accessed. The other argument for Orion was that the UK needed to have a large laser facility, in some way complementary to the NIF in the US, which could act both as a staging platform to NIF and allow interactions on a quid pro quo basis. Figure 1 shows a schematic of the facility layout. The long pulse beams consist of two stacks of five beam lines, supported on each side of a large space frame structure. Each stack is seeded with a pre-amplifier module (PAM), which amplifies pulses up to the 100mJ level and spatially shapes the beam. Each PAM also has smoothing by spatial dispersion functionality in two dimensions (2D-SSD). The PAMs are seeded with pulses with a peak power of roughly 50mW, using a pulse generation system housed beneath the laser hall, known as Optical Pulse Generation 1 (OPG1). The long pulse beam architecture comprises four 200mm aperture disk amplifiers (LG770 Nd:phosphate glass), which are four-passed using angular multiplexing. The beams are separated and redirected near the pinhole plane of a long spatial filter. The output at the fundamental wavelength is about 750J, with a beam diameter of 300mm, which is then frequency tripled in the target hall. Residual fundamental and doubled light is removed from the beam by a sequence of five dichroic mirrors. These mirrors also facilitate path length requirements and act to rotate the beams such that they arrive on target in two opposing cones of five beams with polarisation in the p state relative to a plane normal to the cones mutual axis. Focusing is via f/4 lenses and there is the capability to quickly insert phase plates and debris shields. The short pulse beam line components are supported on a second space frame, one beam line per side. These are seeded with pre-amplified pulses from Optical Pulse Generator 2 (OPG2). Within OPG2, a single commercial mode-locked Ti:sapphire laser, operating with about 12nm of bandwidth (FWHM) around 1054nm, is split into two beams and directed into two Offner triplet stretcher systems. The output from these is 6ns in duration and is directed into the pre- High Power Lasers for Fusion Research, edited by Abdul A. S. Awwal, A. Mike Dunne, Hiroshi Azechi, Brian E. Kruschwitz, Proc. of SPIE Vol. 7916, 79160C 2011 SPIE CCC code: X/11/$18 doi: / Proc. of SPIE Vol C-1

2 amplifier stages, which are based on the OPCPA principle. The pulses are amplified in three stages to roughly 50mJ before being image-relayed upstairs into the laser hall and the short pulse beam lines. The first stage in each short pulse laser chain is a four-passed rod amplifier system, which uses two rods, phosphate and silicate glass, to maintain bandwidth of the pulse. From here, the beam line design uses a mostly single-passed architecture, increasing in aperture though 100mm, 150mm and 200mm disk amplifiers. There are also Faraday isolators at 100mm and 150mm aperture to isolate the laser gain from potential target back-reflections and back-scatter. The beams are ultimately expanded to 600mm diameter before injection into each compressor vessel. They are then transported under vacuum into the target chamber for focusing with f/3 off-axis paraboloidal mirrors. A deformable mirror, at the 100mm aperture region of the beam line, is used to maintain good focusing properties to access the highest possible intensity. 10 Long Pulse Beams Long Pulse Preamplifiers Target Area Laser Hall 1.5m thick concrete shield wall 2 Short Pulse Beams Compressors (13m grating separation) Long & Short Pulse Generation & Pulsed Power System (10MJ) on ground floor Figure 1. Schematic of the Orion facility An extensive suite of target diagnostics is to be made available, which are either fixed in location or designed to be fielded via a Ten Inch Manipulator (TIM) design, to facilitate better interoperability with other laser facilities. 2. LONG PULSE DESIGN The original source of light is a commercial, distributed feedback (DFB) fiber laser delivering up to 1W of CW power at nm. This beam is chopped into a 1kHz chain of ~ ns pulses using an in-fiber acousto-optic modulator and then split between two outputs by the chopper unit. The two chains of pulses are sent to the two, independent, integratedoptic (IO) modulators. These modulators sculpt the 0.1-5ns shaped pulses and are each driven by an arbitrary waveform generator (AWG), plus a square pulse generator to reduce the pulse rise/fall time to less than 100ps. The shaped pulses are sent to the PAMs along 2 long transport fibers. The system is entirely fiber coupled which minimizes maintenance/realignment and makes it inherently safer than an equivalent free space system. The system architecture is summarized in Figure 2. When the two-way split and the inherent system losses are accounted for, the output of the system is roughly 50mW (e.g. 50pJ for a 1ns square pulse). This is more than adequate to ensure good signal to noise ratio from the PAM. Proc. of SPIE Vol C-2

3 DFB Laser Beam dump Chopper unit IO Modulator 1 AWG 1 Pulse gen. 1 AWG 2 Pulse gen. 2 IO Modulator 2 PAM 1 PAM 2 Figure 2. OPG1 system schematic The PAMs house a regenerative amplifier, based on a flash lamp pumped Nd:YLF rod, a spatial shaping stage, a 2D- SSD system based on bulk medium phase modulators, a four-passed 32mm aperture rod amplifier and a diagnostics package. These systems are mounted on a both sides of a vertically mounted optical table with dimensions 5.5m by 1.5m. A schematic of both sides is shown in figure 3. The regenerative amplifier is a self-imaged 4f cavity which utilizes user-introduced polarization losses in the cavity to limit the fluence to safe levels for shorter pulses, while still providing sufficient energy extraction from the rod to achieve good shot-to-shot energy stability. The beam shaping consists of over-filling a serrated aperture. The residual dome from the Gaussian beam is used to compensate the radial gain non-uniformity in the 32mm aperture rod amplifier. The SSD system uses phase modulators at 2.45GHz (PM1 in figure 3) and 10.4GHz (PM2) to impart spectral sidebands on the beam, which are dispersed in orthogonal directions using Littrow gratings. Further gratings are required to precompensate the resultant temporal shear across the aperture of the beam. The rod amplifier system contrives four passes of the rod using polarization switching with a quarter waveplate. A permanent magnet Faraday rotator (FR7) switches the beam in and out of the rod amplifier system. A second Faraday rotator (FR8) inside the four-pass cavity is used to compensate any thermally induced birefringence in the rod and minimize the pre-pulse seen after two passes of the rod. The rod amplifier system is angularly multiplexed so that the on-axis path can be blocked at the focal plane of its relay telescopes, so as to inhibit oscillation. Figure 3. PAM system schematic Proc. of SPIE Vol C-3

4 Each PAM is capable of delivering multiple Joules, however, in the current system design they are only required to yield up to 200mJ. The output from each PAM is split either five or ten ways (there is an option to seed either all ten beam lines with a single front-end, or have independent seeding delay and pulse shape of each stack of five beams lines). Each long pulse beam line requires about 15mJ of seed energy to achieve its required 750J output at 1053nm. The PAM output is split using polarizer / waveplate combinations and are image relayed into the long, transport spatial filter (TSF), where they are injected near the focal plane. The beam is expanded in the TSF to 170mm before injection into the four 200mm aperture disk amplifiers. Each amplifier houses three Nd:glass slabs (LG770) and delivers a small-signal gain of up to 2.4. In practice the gain used is about 2.1, so as to control the potential for parasitic lasing in the cavity. The beam is angularly multiplexed such that different passes of the amplifiers can be spatially separated near the pinhole plane of the TSF. After two passes of the four amplifiers, the beam re-enters the TSF and is directed, near the pinhole plane, into a reverser system, where it is re-collimated at 40mm diameter and returned to the TSF. At this point, the energy is about 4J. The reverser also houses a Pockels cell to isolate the gain from passes 1 and 2 and passes 3 and 4. The final two passes of the disk amplifiers brings the energy to about 750J. This time it passes straight through the TSF, being expanded to 300mm, and is directed into the target hall. Other beamlines OPG1 (ground floor) Main Beamline PAM IO Amplitude modulators (0.1 5ns) DFB Fibre CW 200mm disk amplifiers x 4 Reverser Transport Spatial Filter 4-pass disk amplifier stage f/4 Final Focus Optics Figure 4. Long pulse beam line schematic Harmonic conversion crystals Frequency tripling is effected by a pair of KDP crystals. The first is a type I doubler with thickness of 14mm to achieve the optimal 2:1 energy ratio of doubled and fundamental light at the nominal operating point. The second is a type II mixing crystal, of 12mm thickness. The result is a third harmonic beam with a polarization the same as the input fundamental beam. Five dichroic transport mirrors each discriminate between residual wavelengths and the 3ω beam with an efficiency of ~100:1, giving a very pure 3ω beam at the target. As mentioned above, their geometry is such that the beams is orientated onto target in two opposing cones of five beams. The beam lines themselves do not oppose each other though the target chamber; the two cones of beams are mirror images of each other about the target plane (i.e. a plane at the apex of each cone, orthogonal to their mutual axis). The axis of each individual beam is also rotated by the geometry of the transport mirrors such that the polarization of each beam is in the p state, relative to the target plane, to better couple laser radiation into typical target plasmas. As can be seen from figure 5, the result is a complex layout in the Target Hall. Proc. of SPIE Vol C-4

5 Figure 5. Layout of beam transport optics in the target hall, showing the transformation of the rectilinear beam line layout to a conical geometry. Final focusing of the beams is achieved using a final optics assembly (LPFOA) which consists of (in this order) an optional kinoform phase plate, a vacuum window, a 1.2m focal length lens and a debris shield. 3. LONG PULSE BEAM LINE PERFORMANCE RESULTS Initial commissioning of the first long pulse beam line has taken place. To date up to 600J at 1ω has been fired. Commissioning at 3ω has been limited to about 100J (with about 300J of fundamental light). Figure 6 shows the spatial properties of the beam (near- and far-field). Figure 6. Near- and far-field profiles (uncorrected and corrected) of a long pulse beam line at 1ω. The uncorrected far-field profile corresponds to 5λ of astigmatism. The far-field performance is limited by the system aberration, which is dominated by the heat loading in the disk amplifiers. For example, after five consecutive shots, roughly 5λ of (mostly) astigmatism has been measured using a wavefront sensor in the 1ω output diagnostics package. A static wavefront corrector, a deformed mirror in the reverser Proc. of SPIE Vol C-5

6 section, was optimized to remove the astigmatism from the middle shot of a working day, thus reducing the greatest astigmatism seen on any given shot. The performance of the 3ω beam has so far been limited to about 100J (a function of the relevant project milestones). Figure 7 shows near- and far-field images. The 3ω spot has a 90% of the energy within a diameter of 50µm at the target. A pair of photodiodes were used to measure pre-pulse contrast. This measurement shows that pre-pulses are not present above the measurement s dynamic range of Figure 7a. Long pulse 3ω near- and far-field profiles. The far-field profile is from the same shot as the corrected far-field in figure SHORT PULSE DESIGN The short pulses are originally generated by a commercial Ti:sapphire oscillator delivering pulses with a bandwidth of about 12nm around 1054nm. Immediately these pulses are split to seed each of the two short pulse beam lines. A common oscillator is used to realize the best timing jitter between the two short pulse beams. Each side of the split, the pulses are sent to two independent Offner triplet stretcher systems. These are four-passed systems with an effective grating separation of 3.25m per pass, with a 1480mm -1 diffraction grating at 47.9 angle of incidence. This compensates the compressor with a 13m grating separation and the same grating specification. Two independent stretchers are used so that each short pulse beam can operate with a different pulse duration, specified to be between 0.5 and 20ps (FWHM). The resultant chirp is 300 ps nm -1. The stretcher imparts a hard spectral clip at 18nm, so the stretched pulse duration is 6ns; there is significant power (about 20% of peak) at these pulse extremities. The first stage of amplification takes place within the OPG2 subsystem. An optical parametric chirped pulse amplifier (OPCPA) is used to conserve pulse bandwidth. The OPCPA systems for both beam lines are pumped by a single, commercially sourced, pump laser. This is an Nd:YAG system operating at the second harmonic, which was specified to have very flat pulses (within 10%) both spatially and temporally. The pump pulse is of 6ns duration and so amplifies the full bandwidth from the stretcher. The amplification occurs in three stages, the final two of which exhibit strong pump depletion. Near-degenerate, type I phase matching in LBO is used throughout. The more intense regions of the pulse (spatially and temporally) are driven to the point of back-conversion to achieve optimal output energy stability. Also, the signal pulse mimics the spatial and temporal properties of the pump pulse, that is, it becomes a top-hat, square pulse. Each OPCPA stage can produce about 150mJ, which is much greater than that required to meet the short pulse baseline performance, which results in greater operational flexibility. Figure 8 summarizes the OPCPA performance. Proc. of SPIE Vol C-6

7 I Figure 8a. OPCPA near- and far-field output OPA Energy Stability 300 Number of Shots (100 total) OPA Pump Normalised Energy Figure 8b. OPCPA output energy stability The OPCPA output is expanded and then apodized at 16mm diameter such that only the cleanest, central region of the beam is injected into the main short pulse laser chain. This begins with a four-passed, mixed glass, 32mm aperture rod amplifier subsystem. This uses phosphate and silicate glass in order to maintain as much bandwidth as possible before injection into the disk amplifiers. Other than the use of two rods, the design is very similar to the rod amplifiers on the long pulse PAMs. This system produces about 2J for baseline performance with 9nm (FWHM) bandwidth. The beam is then expanded to 86mm diameter before it is injected into a 100mm aperture Faraday rotator. This has a two-fold function: it allows a double pass of the first disk amplifier and acts as an optical isolator for back-reflections. In the double passed disk amplifier the pulse experiences a small signal gain of about 6 per pass. This is somewhat lower than the gain for a monochromatic pulse. The beam is then expanded to 140mm diameter, where it passes through a 150mm aperture disk amplifer, with a gain of about 4. The next component is another Faraday rotator, with the sole function of isolation against back reflections. The beam is then expanded to 180mm and single passes three 200mm amplifiers, of the same design as those on the long pulse beams. The beam line is shown schematically in figure 9. Proc. of SPIE Vol C-7

8 Rod amplifier stage Disk amplifier stage Adaptive Optic Faraday Isolator disk amplifiers 2nd beamline Broadband Preamplifier (OPA) Pulse stretcher Pulse generation OPG2 (Ground Floor) Spatial filter Spatial filter Compressed pulse beam Diffraction transport (in vacuum) grating Spatial filter Main Beamline Target chamber f/3 Off-axis parabola Compressor (vacuum) Figure 9. Schematic of a short pulse beam line. The result is pulses with about 700J and 5nm bandwith. These are expanded to 600mm diameter and image-relayed into the compressor vacuum vessels, which contain single pass compressor systems, using the same grating configuration as the stretcher, and a 13m grating separation. The gratings are gold-coated and have 940mm aperture in the horizontal (dispersive) plane. There is a slight spectral truncation in the near-field on the second grating. Long vacuum transport systems send the beam from the compressor vessels to the target chamber, where they are focused with f/3 paraboloidal mirrors. 4. SHORT PULSE BEAM LINE PERFORMANCE RESULTS One short pulse beam line has been operated at up to 400J prior to insertion into the compressor vessel and 200J with compressed pulses. The performance data are shown in figure 10. The far-field image shown is at the output of the final amplifiers, without applying any wavefront correction, after three consecutive shots. A comparable image with wavefront correction is also shown. An image taken after the compressor vessel is shown. Initially, this appeared slightly different and required further work to optimize aberrations though the compressor and in the compressed pulse diagnostics station. Figure 10a. Short pulse beam output near field profile. The laterally dispersed beam through the compressor clips the second grating. Proc. of SPIE Vol C-8

9 Figure 10 b, c, d. Short pulse beam line far-field images uncorrected, corrected and post-compressor, respectively. The output of the amplifier stage shows that 4nm of bandwidth is supported by the beam line. There is further scope to increase the silicate rod contribution, so this should increase slightly in future. Figure 11 shows the pulse spectrum at the amplifier output and figure 12 shows an autocorrelation trace, suggesting a sub-picosecond pulse duration. There is further optimization of the pulse compression to be undertaken. 4nm Figure 11. Spectrum supported by the short pulse beam line, measured at the amplifier output Autocorrelation intensity Autocorrelation delay Figure 12. Autocorrelation trace measured in the compressed pulse diagnostics station. Proc. of SPIE Vol C-9

10 5. PLASMA PHYSICS PROGRAMME Experiments on HELEN (Orion s predecessor facility) have shown that useful opacity data can be obtained from extremely hot, dense plasma heated by short pulse lasers. Plasma temperatures achieved were in excess of 600eV (more than 6 million degrees centigrade), in plasmas with densities around solid aluminium (1-3g/cc). On ORION plasmas will be created at higher density than those on HELEN allowing high temperature opacity experiments to be carried out at conditions closer to local thermodynamic equilibrium than possible on HELEN. The data from the HELEN experiments have advanced the understanding of the heating mechanisms in short pulse experiments to the point where we can now realistically access possible experiments with Orion. The very high temperatures in short pulse laser- solid target interactions are due mainly to resistive heating by currents circulating in the target. In essence the laser can be thought of as an extremely fast current switch and the plasma experiments as measuring the after effects of this enormous current (the current density is in excess of Am -2 ). On Orion, temperatures are expected to be in excess of 1keV, producing thermal pressures of several Gigabars, and ponderomotive pressures in excess of this, during the laser pulse. The shocks produced by the rapid heating of the target rapidly propagate outward through material surrounding the laser spot. After an initial rapid decay, a pressure of several hundred Megabars is maintained for durations on the order of 100ps; long enough for ultra-fast diagnostics to be used to measure the material properties at higher pressure than previously achieved in laboratory experiments. Orion will also be used to make material measurements at lower temperature but at higher density than achieved previously. The combination of long pulse compression and short pulse heating will be used to study a number of dense plasma effects which influence high temperature equation of state and opacity such as pressure ionization, Stark broadening and ion-ion coupling. The second short pulse laser will be used to diagnose these high density plasmas using high energy X-ray backlighting and Thomson scattering. Orion will also be used to carry out experiments in hohlraums to measure the properties of lower density plasma heated by X-radiation. Experiments are planned that use X-ray backlighting for a variety of experiments. For example, to sample population distributions in high principal quantum number states to aid in the equation of state development with the CASSANDRA atomic physics code, using grating spectrometers built as part of the Orion project. It will also be possible to investigate the lower temperature constitutive properties of materials such as strength. Strength experiments using Laue diffraction techniques that are currently being carried out on US facilities are also planned for the early years of Orion operation. In short, the Orion experiments will be a mix of hohlraum based experiments using techniques tried and tested over several decades of experiments on HELEN and US lasers, extensions of short pulse HELEN experiments to more extreme conditions, and novel experiments accessing new plasma regimes; an example of which are the shock experiments at hundreds of Megabars. 6. CONCLUSIONS The initial commissioning of one short pulse and one long pulse beam line of the Orion facility is now complete. A substantial fraction of the full, baseline energy performance has been demonstrated for both types of beam line. Further work is required to optimize the performance characteristics in detail. This will be done in parallel with commissioning the other nine long pulse and one short pulse beam line during The first experiments to look at materials properties in the high temperature, high density regime will commence in the Spring of Proc. of SPIE Vol C-10