Experiments to investigate the effects of radiative cooling on plasma jet collimation
C. D. Gregory, A. Diziere, H. Aoki, M. Besio, S. Bouquet, E. Falize, T. Ide, B. Loupias, C. Michaut, T. Morita, S. A. Pikuz Jr., A. Ravasio, Y. Kuramtisu, Y. Sakawa, H. Takabe, H. Tanji, N. C. Woolsey, M. Koenig
aa r X i v : . [ phy s i c s . p l a s m - ph ] N ov Astrophysics and Space ScienceDOI 10.1007/s ••••• - ••• - •••• - • Experiments to investigate the effects of radiative coolingon plasma jet collimation
C. D. Gregory • A. Diziere • H. Aoki • H. Tanji • T. Ide • M. Besio • S. Bouquet • E. Falize • B. Loupias • C. Michaut • T. Morita • S. A. Pikuz Jr. • A. Ravasio • Y. Kuramitsu • Y. Sakawa • H. Takabe • N. C. Woolsey • M. Koenig c (cid:13) Springer-Verlag ••••
C. D. GregoryLaboratoire pour l’Utilisation des Lasers Intenses, UMR7605,CNRS - CEA - Universit´e, Paris VI - Ecole Polytechnique, 91128Palaiseau Cedex, FranceA. DiziereLaboratoire pour l’Utilisation des Lasers Intenses, UMR7605,CNRS - CEA - Universit´e, Paris VI - Ecole Polytechnique, 91128Palaiseau Cedex, FranceH. AokiInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanH. TanjiInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanT. IdeInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanM. BesioDipartmento di Fisica “G. Occhialini” University of Milano-Bicocca, Milan, ItalyS. BouquetCEA/DIF/DPTA BP 12, 91680 Bruyers-le-Chatel, FranceE. FalizeCEA/DIF/DPTA BP 12, 91680 Bruyers-le-Chatel, FranceB. LoupiasLaboratoire pour l’Utilisation des Lasers Intenses, UMR7605,CNRS - CEA - Universit´e, Paris VI - Ecole Polytechnique, 91128Palaiseau Cedex, FranceC. MichautLUTH, Observatoire de Paris, CNRS, Universit´e Paris Diderot,Place Jules Janssen, 92190, Meudon, FranceT. MoritaInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanS. A. Pikuz Jr.Joint Institute for High Temperatures of RAS, Izhorskaya 13/19,Moscow, 125412, RussiaA. RavasioLaboratoire pour l’Utilisation des Lasers Intenses, UMR7605,CNRS - CEA - Universit´e, Paris VI - Ecole Polytechnique, 91128Palaiseau Cedex, France
Abstract
Preliminary experiments have been per-formed to investigate the effects of radiative coolingon plasma jets. Thin (3 µ m - 5 µ m) conical shells wereirradiated with an intense laser, driving jets with veloc-ities >
100 km s − . Through the use of different targetsmaterials - aluminium, copper and gold - the degree ofradiative losses was altered, and their importance forjet collimation investigated. A number of temporally-resoved optical diagnostics was used, providing infor-mation about the jet evolution. Gold jets were seento be narrower than those from copper targets, whilealuminium targets produced the least collimated flows. Keywords
Young stellar objects; Jets and outflows;Laboratory astrophysics
Y. KuramitsuInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanY. SakawaInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanH. TakabeInstitute of Laser Engineering, Osaka University, Suita, Osaka,565-0871, JapanN. C. WoolseyDepartment of Physics, University of York, Heslington, YO105DD, UKM. KoenigLaboratoire pour l’Utilisation des Lasers Intenses, UMR7605,CNRS - CEA - Universit´e, Paris VI - Ecole Polytechnique, 91128Palaiseau Cedex, France Current address: Department of Physics, University of York,Heslington, YO10 5DD, UK LUTH, Observatoire de Paris, CNRS, Universit´e Paris Diderot,Place Jules Janssen, 92190, Meudon, France
Jets from young stellar objects (YSOs) are associatedwith the accretion phase of stellar evolution, which lastsfor around the first 10 years of a young star’s life,see Reipurth and Bally (2001) for a review. The jetsare seen propagating away from the star at speeds ofthe order of 500 km s − , with lengths of up to 1 pc,and with aspect ratios (jet length / jet width) of 10or more. The flows are often seen to terminate in re-gions of optical emission, and contain a series of brightknots - these regions of emission are known as Herbig-Haro (HH) objects. The relative proximity of thesesystems means that observations of radiation from HHobjects have allowed a large amount of high-quality ob-servational data to be collected. These objects havedynamic time-scales of the order of a decade, and it isthus possible to observe the jet as it evolves. Thesedata, along with theoretical and computational mod-elling, have lead to significant improvements in the un-derstanding of YSO jets. Despite this, questions stillexist surrounding the physics of both the jet launchingand propagation, and so the possibility of performingwell-designed laboratory simulations that contribute tocurrent level of understanding therefore has potentiallylarge benefits. This experiment represents part of anongoing campaign (Gregory et al. 2008; Loupias et al.2007; Gregory et al. 2008b; Waugh et al. 2009) to aid inthe understanding of aspects of jet propagation physics,and in particular the high degree of collimation oververy large distances. There are three main physicalprocesses which are thought to be important: 1/ iner-tial confinement of the flow by the ambient medium,2/ radiative cooling of the jet, causing a drop in theinternal thermal pressure and a collapse on axis, and3/ magnetic fields which act to restrict the radial ex-pansion of the charged particles. A number of exper-iments have taken place in recent years to investigatejet propagation. These have demonstrated the prop-agation of a jet moving into an ambient medium, ei-ther stationary (Nicola¨ı et al. 2008; Foster et al. 2005;Ampleford et al. 2005) or in the form of a plasma cross-wind (Lebedev et al. 2004), and an increase in jet colli-mation due to radiative cooling (Shigemori et al. 2000;Lebedev et al. 2002; Purvis et al. 2010). The work pre-sented here is aimed at studying the effects of radia-tive losses on the collimation of plasma jets. Conicaltargets are irradiated with an intense laser, resultingin jets of high velocity. By choosing different targetsmaterials - here gold, copper and aluminium - the im-portance of radiative losses can be altered, since higheratomic number materials radiate more efficiently. Thisradiation cools the jets on axis, and lowers the thermal pressure driving the radial expansion. As a result, moreradiative jets are expected to have larger aspect ratioand to be more collimated. Results are given from twoexperiemental campaigns, the first at the GEKKO XII,12 beam, 10 kJ, Nd:glass laser system at the Institutefor Laser Engineering, Japan. The second experimenttook place at the PICO2000 facility at the Ecole Poly-technique, France, which houses a 100 J, picosecond,Nd:glass laser and a 1 kJ, nanosecond, Nd:glass laser. The primary experimental configuration is shown inFigure 1. Five beams of the GEKKO XII laser irra-diated the apex of the conical shell, delivering around500 J of laser light in a 500 ps pulse, at a wavelength of351 nm. The focal spot size was 600 µ m. The plasmafrom the rear face of the target is focused on axis dueto the conical geometry, and forms a jet. A second,mJ, 527 nm, 20 ns laser pulse probed the system inthe direction perpendicular to the plasma flow. Thedelay between the arrival of the probe beam and thedrive lasers was varied up to a maximum of 70 ns. Amodified Normarski interferometer analysed the beam,and was detected with a gated optical imager (GOI)with a temporal resolution of 250 ps. In addition, theself emission from the jet at a wavelength of 450nm ± µ m thick gold, 5 µ m thick copper, or 3 µ m thickaluminium. All the targets had a full opening angle of140 ◦ and a diameter of around 1.2 mm. emission Probe and self emission axis b) End−on view for PICO2000probe and X−rays, JetLaser X−raysJet o a) Top view Fig. 1
The experimental set-up. Panel a) shows the orien-tation of the optical probe and self-emission diagnostics usedin both experiments, as well as the x-ray radiography usedonly in the PICO200 experiment. The angular separationbetween the optical and x-ray axes used in the PICO2000experiment is shown in panel b). This configuration allowedboth diagnostics to be used simultaneously. In this view thejet is propagating out of the page.
A second experiment took place at the PICO2000facility. In this case only copper targets were used, andwere irradiated with ∼
400 J, 1 ns, 532 nm laser pulses, adiative cooling and jet collimation 3 focused through hybrid phase plates to give a 400 µ mfocal spot. A mJ, 527 nm, 8 ns laser was again used toprobe in the direction perpendicular to the jet propa-gation axis, and was detected with a GOI with a timeresolution of 120 ps. Here, no interferometer was used,and the diagnostic was used in a shadowgraphy config-uration. A streak camera, with a time window of 50ns, recorded the self-emission at 450nm ±
10 nm trans-verse to the jet propgation direction. An additionalx-ray radiography diagnostic was implemented for thisexperiment. Laser pulses of 60 J, 30 ps at 1064 nm,and a focal spot of 50 µ m, irradiated a Ti foil placed 20mm away from the concial shell. The resulting Ti-k α emission at 4.75 keV was used to radiograph the jetsat varying times during their evolution, using imagingplate as the detector. The axis of the x-ray and opticaldiagnostics were separated by 22 degrees, allowing bothdiagnostics to be used simultaneaously. Figure 2 shows the results of the experiment at theGEKKO XII facility for targets of gold, copper, and alu-minium. In all images the laser is incident from the left,and the jet propagates from left to right. The originalhorizontal postion of the target is indicated by a dashedvertical line. The columns show, respectively, interfer-ograms taken with a probe delay of 50 ns, the streakedself-emission over the first 28 ns of the jet evolution,the imaged self-emission after 50 ns, and the imagedself-emission after 70 ns. In each of the self-emissionimages, a dark region is seen close to the original tar-get position. This is due to absorption of the emittedlight by plasma generated from the target mount. Inthe case of the gold jet, the interferograms and the self-emission images show a jet with aspect ratio (jet lengthto jet width) of around 2.5, and the flow appears to beconverging onto the axis even at distance far from theoriginal target postion. The initial velocity of the jetcan be inferred from the streaked self-emission diag-nostic, and is 130 km s − . The results for the coppertargets indicate a jet moving with a velocity of 120 kms − . The reason that this jet has a lower velocity thanfor gold is likely due to fluctations in the laser energyon target and difficulties with the laser beam alignmentprocess. In this case the jet diverges as it moves awayfrom the target surface, as seen in the self-emission im-ages, which show a flow opening angle of ∼ ◦ . In thecase of the aluminium cones, the plasma flowing fromthe target surface moves initially with a much highervelocity of 470 km s − . Due to this the leading edge of the expansion is out of the view window for the in-terferogram, and for the self-emission image after 70ns.Figure 3 shows data taken during the PICO2000 ex-periment. Panel a) shows a shadowgraphic image ofa cone target taken after 20 ns. In panel b), a corre-sponding image is shown for a simple plane foil of 5 µ mthick copper. The flow from the conical foil is narrowerand has propagated a greater distance from the originaltarget position. These data indicate the importance ofthe target geometry for the formation of jets. Panelc) shows a mass density profile for a jet after 12 ns,inferred form the x-ray radiography. a) b) D e n s it y [ g / cc ] c) 1 mm µ m 1 mm500 0.350.00 Fig. 3
Shadowgraphs for a) a concial copper target, andb) a plane copper target, taken with a probe delay of 20ns during the PICO2000 experiement. Panel c) shows adensity profile after 12 ns, inferred from x-ray radiography.In this x-ray image, the jet propagates from bottom to top,and the orginal position of the target is at the bottom edgeof the image.
The data presented in Figures 2 and 3 show that forboth campaigns, the experimental configuration wassuccessful in generating plasma jets. Panels a) andb) of Figures 3 indicated the importance of the coni-cal geometry of the target: the cone target producesa significantly more collimated outflow than a simple e)a) b) c) d)f) g) h)i) j) l)k) 2 mm2 mm2 mm2 mm2 mm2 mm 2 mm 2 mm2 mm 2 mm2 mm2 mm
Fig. 2
The data from the optical diagnostics used in the GEKKO XII experiment. The top, middle and bottom rowsrepresent data for gold, copper, and aluminium targets respectively. The colums, from left to right, show data from theinterferometry with a probe delay of 50 ns, the streaked self-emission over the first 28 ns of the jet evolution, the imagedself-emission after 50 ns, and the imaged self-emission after 70 ns. The dashed line indicates the intial target location onthe horizontal axis. See text for further details. plane target. The optical and x-ray probes are com-plimentary - the x-rays are able to diagnose the densepart of the jet close to the target surface, but the moretenous regions of the flow around the jet head are belowthe sensitivity of the diagnostic. In contrast, the opti-cal probe is unable to penetrate the high-density coreof the jet, but is well suited to imaging relatively lowplasma densities. The capability of fielding these twodiagnostics on a single shot is therefore beneficial, andallows a more complete picture of the dynamics of thejet to be built.The results shown in Figure 2 suggest that the jetcollimation is increased for higher-atomic number tar-gets. Gold jets are narrower than those of copper, andthe least collimated flows are from the aluminium tar-gets. Heavier materials radiate more efficiently, andthe results are consistent with computer simulations(Mizuta, Yamada & Takabe 2002), which have shownthat increased radiation can lead to a drop in the ther-mal pressure on the axis, and a radial collapse of lab-oratory jets increasing the collimation. Although the initial data are consitent with the the idea that moreradiative jets are better collimated, firm conclusionscannot yet be reached. The reason for the large ra-dius of the aluminium flows may be because these coneswere too thin, and were exploded by the laser. In addi-tions, the velocity of the aluminium jets is much higherthan for copper or gold. This means that at the sameexperimental delay the aluminium jets have reached alater stage in their evolution, and that simple compar-ison of images taken at the same delay is not suitablefor meaningful conclusions to be drawn. Future ex-periments will adjust the target design to conserve aconstant areal mass density for all materials - by in-creasing the thickness of the lighter materials - in anattempt to create jets with similar expansion speeds.In addition, the time at which images are recorded willbe normalised to the jet velocity for each material, ef-fectively comparing jets of the same length, in order torecord a snapshot of the jets at the same point in theirevolution. adiative cooling and jet collimation 5
The time-resolved self-emission diagnostics showsthat the aluminium jets move around 3.5 times fasterthan the copper and gold jets. It is therefore expectedthat the aluminium jets are 3 . Thin conical targets of aluminium, copper and goldhave been irradiated with an intense laser. The result-ing flows have been studied with a variety of opticaldiagnostics and x-ray radiography, and jets are seen toform. The preliminary findings from these experimentssupport the conclusion that jets that radiate more effi-ciently are better collimated, due to the subsequent de-crease in the internal thermal pressure and a collapse onaxis. Future experiments will investigate a reasonableway to compare jets of different materials, since giventhe dissparate propagation velocities a simple compar-ison at some fixed time delay is not appropriate. Inaddition an ambient medium will be introduced intothe jet propagation region to simulate the interstellarmedium, and to increase the astrophysical relevence ofthis work.
Acknowledgements
The authors wish to thank thetarget preparation group at the Rutherford-AppletonLaboratory and Gabriel Schaumann of the TechnicalUniversity of Darmstadt for fabricating the cone targes.CDG acknowledges financial support from R´egion Ile-de-France and RTRA.
References