34
VIS MCAO at the VLT AOF! S. Esposito, G. Agapito, M. Bonaglia, L. Busoni, INAF-Arcetri T. Fusco, B. Neichel, LAM VLT AO-Community day 20-21 Sept., 2016
VISMCAOattheVLTAOF!
S.Esposito,G.Agapito,M.Bonaglia,L.Busoni,INAF-ArcetriT.Fusco,B.Neichel,LAM
VLTAO-Communityday20-21Sept.,2016
Sep2015:AOFfactsandconsidera7onsØ AdeformablesecondarymirrorDSMwith1170actuators,conjugatedto~ground
Ø Fourlaserguidestars20weach,drivinga40x40SHSforGLAO(4WFSsintotal).
GLAOforNIRandVIS,narrowfieldmodeforVIS
SCAOwithNGSandLGS(ERIS)
Addingpost-focalDMsischangingthegame=>increasecorrectedFoVbeyondlimita[onofnaturalangularanisoplana[sm.
~20cmactuatorspacingprojectedonM1
Considerusingthe21cmspacing(and420Wlaser)topushcorrec[ontoshorterwavelengths.
AVisibleLGSM
CAOsystemwith20
arcsec
FoV(2DMs)
Correc7oninthevisible……650nmimagesfromForerunneratLBT
…visibleisdoable….
-welltuned500modesreconstructors-NCPAcompensatedat6and3lambda/D-indomepsfswithandwithout0.8seeingturbulenceshowingStrehlof50%.
similarimagesfromSPHERE…
A) Largevisibledetectorsarecheap(comparetoNIR),anddetectorquality
ismuchbeder(darkcurrent,cosme[c,warmandsimple)B) B)Skybackgroundissmall(1000to10000[mesdarkerthanK),
differencewithspaceissmalltoo.C)“low-noise”(<1e-RON)large(4kx4k)andfast(10frame/seconds)detectorsalreadyexist!(e.g.Gachetal.)
C1:PostprocessingTip-Tiltcorrec[on(100%skycoverage,imagesarere-centeredpost-facto)C2:Higherorderpost-processing(e.g.mul[-framedeconvolu[on)inordertorecoversomeofthepar[alcorrec[onoftheAO
=>C1,2mayrelaxsignificantlytheconstraintsontheAOsystem!
Advantagesofvisibleobserva7onsw.r.t.NIR
VISMCAO4VLT:talksummaryAfirstassessmentof#ofLGSsand#ofDMsconsideringAOFavailabiltyi.e.4LGSsandDSM(1170acts)
E2Esimula[onexploringrestrictedparameterspaceforLGSsandDMs(assumeVLTenvironmentalparameters)
compareVLTandHSTPSFsat650nm:SNRinRbandfilter
ConclusionsaboutVISMCAOperformance(andFoV)
A5(4)LGSs&2post-focalDMsoptomechanicalsketch.
Basiclimi7ngfactorsforVISMCAOsystem
totalwferror:wewanttoachieve~30%SRat650nm(e.g30%ofenergyina2lambda/Dpatch),requires~1rad^2overallAOerror.At650nm1radeq.to~100nm
(1)generalizedanisoplana[sm(&fipng)error(#DMs,#Acts)(2)tomographicerror(#ofLGS,thetaofLGS)
Tomograficreconstruc[onandDMscorrec[onefficiency
plotdoesnotincludefipngerror(dact/ro)^5/3
CHAPITRE 8. QUELLE OA POUR EAGLE?
Fig. 8.3 – En haut : influence du nombre de DMs sur l’EE moyenne en fonction de la tailledu champ scientifique. En bas : illustration du residu de correction (en niveau de gris) enfonction de l’altitude et de la frequence spatiale pour deux ou trois DMs.
Le principal avantage de cette methode est que l’on peut alors envisager de travailleren boucle fermee, ou plutot en “pseudo boucle fermee”. En e↵et, dans le cas de la SMCAO,chaque ASO associe aux etoiles lasers sera place derriere le DM associe a son sous-champen plus du M4. Compare a une configuration boucle ouverte, cela permet de reduire ladynamique necessaire sur les ASOs1. De plus, chaque DM post-focal est vu par un ASOce qui permet d’en controler les formes et de reduire l’impact des non-linearites.
1Il faut tout de meme noter que les ASOs voient une correction “onde plane” des DMs, alors que l’onde
incidente est spherique. Autrement dit, chaque ASO subit l’e↵et de cone ce qui necessite une dynamique
de mesure importante.
Memoire de these Benoıt NEICHEL 244
tel-0
0366
529,
ver
sion
1 -
8 M
ar 2
009
CHAPITRE 8. QUELLE OA POUR EAGLE?
Fig. 8.3 – En haut : influence du nombre de DMs sur l’EE moyenne en fonction de la tailledu champ scientifique. En bas : illustration du residu de correction (en niveau de gris) enfonction de l’altitude et de la frequence spatiale pour deux ou trois DMs.
Le principal avantage de cette methode est que l’on peut alors envisager de travailleren boucle fermee, ou plutot en “pseudo boucle fermee”. En e↵et, dans le cas de la SMCAO,chaque ASO associe aux etoiles lasers sera place derriere le DM associe a son sous-champen plus du M4. Compare a une configuration boucle ouverte, cela permet de reduire ladynamique necessaire sur les ASOs1. De plus, chaque DM post-focal est vu par un ASOce qui permet d’en controler les formes et de reduire l’impact des non-linearites.
1Il faut tout de meme noter que les ASOs voient une correction “onde plane” des DMs, alors que l’onde
incidente est spherique. Autrement dit, chaque ASO subit l’e↵et de cone ce qui necessite une dynamique
de mesure importante.
Memoire de these Benoıt NEICHEL 244
tel-0
0366
529,
ver
sion
1 -
8 M
ar 2
009
theta>17”,2DMsalwayslimitedby#ofDMs
theta<45”,3DMsalwayslimitedby#ofLGS
plotshowingresidualrmsforWFreconstruc[onandDMsplacementsop[mizedforscienceFoV~eq.toLGSasterism
tomogenaniso
datafromT.Fusco,LAM
70nm
Basicerrorbudget
1)tomo+gen_anisoplana[s=~90nm2)DMfipngerror~50nm(DSM)3)LGSphotonnoise~504)TTresidualNGS~30nm(fromE2Emodaldecomposi[on)
VLTAOFenvironmentalparametersassumed.-ExpectedfluxesforESOAOF(~100phot/ms/sub,40x40sub)-seeing=0.73arcsec-L0=25m
3DMs(2postfocal&5LGS)sigma_tot~125nn….movetoE2Esimula[ons
E2Esimula[onsmainparameters• Atmosphere:
– Seeing(@zenith)0.66”– L025m– Cn2ERISprofile(10layers)– zenithangle30°– inputwfstd.dev.1042nm
• NGS:– WFS:2x2SH– GSon-axiswithR=12,19
• LGS:– WFS:4,5&940x40SH– asterism:
• 4LGS@FoV(150fot/sub/frame)• 5LGS:4@Fov+1on-axis(150fot/sub/frame)
• 9LGS:4@FoV+4@FoV/2+1on-axis(75fot/sub/frame)
• DM1(ASM1172acts):– Height0m– 945KLmodes
• DM2(ALPAO241acts)– Height5000m– 252KLmodes
• DM3(ALPAO241acts)– Height10500m– 252KLmodes
4+1LGS~3+1LGS,maysavecentralLGS,(TBC)
Asterisms:
E2Eresults@650nmAnima[on!(1)brightLGS(x10);(2)brightNGS(mag12);(3)centeredNGS
Resultscomparison
courtesy:M.LeLouarnCourtesy:T.Fusco
MLeLouarn&T.Fuscosimula[onresultsinagreementwithArcetriones.
(Briefly)Effectsof:(1)offaxisNGS,(2)fluxofNGS,(3)fluxofLGS,(4)3+1~4+1(whenLGSisconcerned)
3+1~4+1(whenLGSisconcerned...)
FWHM
EE50%
SR
BrightLGS(x10)BrightNGS(R=12)On-axisNGS
performanceof3+1LGSsisslightlyworsethan4+1but3+1isdoablewithpresent4LGSofAOF
BrightNGSoffaxis
EE50%
FWHM
SR
FaintNGSonaxis:results@650nmEE
SR
Rmag19thgivesslightlyreducedSRsovertheFoV,Hbandsourceshouldimproveresults
Differencesarelimitedtofirst10-20mas
EEprofiles
FWHM
SR
System SC: tip tilt reference star
16
FoV for
30% R<15 R<16 R<17 R<18 R<19
sky cov.@GP [as] [as] [as] [as] [as]
GSCNGP 86 67 53 41 29NOMADNGP 81 64 51 40 31
GSCSGP 84 66 51 40 30NOMADSGP 80 63 50 41 33
Average 83 65 52 41 31
3x104randomdirec[onsin5°radiuscirclearoundGalac[cPoles.GSCandNOMADcatalogues
usingpostprocessingwithfastreadoutVISCCDstoreduceneedsforNGSstar
realLGSfluxes:results@650nmFWHM
SR
EE50%
FWHMSR
BrightLGS(x10)VSTruefluxLGSBrightNGS(R=12)On-axisNGS Casewiththelargestvaria[onfound
Opto-mechanics2postfocalDMs4+1(3+1),8+1LGS1ormoreNGS(VISorNIR)
MCAOrelayarrangement
DM10km
DM5km
K-mirror(removeseleva[on:M2&LGSsarefixed)
OAP2
OAP1
LGSdichroic(foldsNGSlightbelowMCAObench) NGSfocus
(mustbemechanicallyderotatedtofollowsky)
~500mm
VLTlight
NGSandLGSfootprintsonDMs
[email protected]=44mm
9xLGSfootprint:Max22.5’’off-axis
DM@10kmDiammax=44mm
NGSfootprint:Max22.5’’off-axis
IRNGSWFSfocus
9xLGSWFSs
DM5km
MCAObench&LGS9xWFSsarrangement
Blue=NGS+LGSlightpathYellow=LGSlightpath
DM10km
InstrumentVISfocus
LGStrombone
NGS/LGSdichroic
NGSpartadachedtomechanicalderotatortotracksky
9xLGSpickoff~1000mm
LGSTrombone
LGS2xbeamexpander
9xLGSWFSbelowMCAOrelay
~1000m
m
8xpickoffmirrors(inred)for11.25’’and22.5’’off-axisLGSs(10mmdiameter)On-axisLGSistransmided
from45x45arscecto90x90
possibili[esforadoublesizeFoVlike90x90withsameperformances:1)op[ona:MCAOsystemwitha4-5postfocalDMsand~10LGS2)op[onb:a2x245x45arcseca)2postfocalDMsb)5WFS/LGS(4+1)[led,atotalof9LGS3)op[onc:….
3DMs:DSM+2postfocalDMs,5LGS(4+1)
AlreadydiscussedinthepastbyR.Ragazzoni&B.NeichelfigurefromB.Neichel,LAM
HSTandVLTVIScomparison
JWSTinVIS
“JWSTwillbediffrac0onlimitedat2μm,definedashavingaStrehlra0o>0.80.JWSTwillachievethisimagequalityusingusingperiodicwave-frontsensingandcontroloftheprimarymirror.Theobservatoryandpoin0ng-controlsystemaredesignedtolimitimagemo0ontolessthan7milliarcsecondsduringobserva0ons.”
80%SRat2um=>142nmrms,thisgivesa15%SR,assumingno[p[ltresidualerror.
hdps://jwst.stsci.edu/instrumenta[on/telescope-and-poin[ng/image-quality-and-psfs
HST&VLTVISMCAOwearereferringtoACS/WFCinthefollowingdiscussion
VLTMCAOVISvsHST/WFCObjectdetectedfluxforR=25:• VLT=20.0ph/s• HST=2.0ph/sBackgroundflux:• VLT=846ph/asec2/s• HST=35ph/asec2/sPlateScale:• VLT=10mas• HST/ACS-WFC=50masRON:• VLT=2e-/pixel• HST/ACS-WFC=8e-/pixel
Throughputinsimula[onsHSTACS-WFCF625W-77~0.4,[548-707]VLT(VIMOSRfilter)~0.3,[558-730]
SNR = γ source γ source +γbckgρ2 +σ ron
2 ∗npix
Results@650nm
RAWContrast EncircledEnergy%
(1)realLGSflux,(2)NGS12mag(3)NGSonaxis
curvesofthesamecolorrepresentprofilesindifferentposi[onoftheFoValwaysgivenasdiameter
SourceR=25@650nm,brightNGS
Profilesinphotons
EEinphotons
SNR←y-Linear→y-Log
RealLGSfluxBrightNGS(R=12)On-axisNGS
SourceR=30@650nm,brightNGS
Profilesinphotons
EEinphotons
SNR←y-Linear→y-Log
RealLGSfluxBrightNGS(R=12)On-axisNGS
SourceR=30@650nm,faintNGS
Profilesinphotons
EEinphotons
SNR←y-Linear→y-Log
RealLGSfluxBrightandFaintNGSOn-axisNGS
Resumingcomparison:dt=1000ssourceRmag
SNRHSTACS/WFC SNRVLTVISMCAO SNRra7o
Exp.7me
10mas 50mas 10mas 50mas
25 13 30 49 55 3.8-1.8 14.4-3.4
30 0.28 0.56 1.6 0.77 5.7-1.9 32.5-1.7
Ø 30magRstaratSNR=5Ø VLT~10^4s(2.7h)over20masdisk² HST~8e4s(22h)over100masdisk² HST~3.2e5s(88h)over20masdisk
correctedFoVis30x30or45x45withminorlossesinperformance(2postfocalDMs,5LGS)
ConclusionsIØ errorbudgetsandE2Esimula[onsshowsthat
AOFHWASM+4LGSplus2small(50mm)postfocalDMsprovidesagoodMCAOsystemforVIS(650nm),<SR>30%for30”diameter.
Ø 2postfocalDMs,4+1LGS(30”or45”)and5WFSs40x40subapertures.Afirstcompact1mx1mx1marrangementforoptomechisoutlined.
Ø ona30magsource(in1000s)VLTachievesbederSNRthanHST.Upto67meson10masradandupto1.9on50masrad.VLTdetects30magatSNR=5in2.7hagainstHST22h.
2yearspost-docposi[onavailableatLAMtoworkonWideFieldAOVLT3
ConclusionsII
TheVISMCAOatVLT(doesbenefitalotfromAOFexis[ngHW):(1) itprovidesgoodperformancecomparedtoHST,
detectsmag30in~3h(HST~20h)at5xtheHSTspa[alresolu[on(10mas).
(2) itisdesignedusingcommercialDMsandparts,seemsoflimitedcomplexityandcost.
