Home

FLAMES User Manual

1. 3 X Y and PSSN can be found in the binary FLAMES FIBER table FU FPS gt Retractor number E UE E WE NO SL PA E E 49000 107 25 If p so 77 os F113 134 147 165 183 203 E 3000 ER a x lt 2 2000 d E lt gt 7 1000 E o Lilli d H E E E E EJ E HE 3 E E O Soo 1000 1500 2000 x pixel Figure 25 Top panel Geometric layout of IFU fibers including the X Y and Position of the sub slits for an individual IFU The IFU orientation long short axis on the plate is also shown Bottom panel Schematic layout of the IFU spectral format not to scale blue solid line
2. tweak time refers to the time for which the field was configured 6 5 2 HDU3 FLAMES FIBRE Table This table is a static table in the sense that for all files of a given period it should be the same This table contains the association between the fibre buttons and retractors and the slit position and the transmission of the fibres as measured in the laboratory before shipping to Chile The table is changed only when some major problem or change occurs e g that a fibre subsystem needs to be replaced with a spare Several of the data contained are useful for engineering purposes the reader is invited to concentrate on field 2 FPS and field 8 RP which give the position in the slit and the position number of the button on OzPoz respectively One example of this table can be retrieved via the web from FLAMES User Manual VLT MAN ESO 13700 2994 73 http www eso org observing dfo quality GIRAFFE txt fibre dat Schematic examples for the raw data format of the different fibre types are given in Ap pendix 7 1 Description of the table columns Col 1 Slit Slit name from fibre type and plate number Col 2 FPS Progressive fibre position number in the slit Col 3 SSN Subslit number Col 4 PSSN Fibre Position in the subslit Col 5 Retractor Serial Number of the retractor Col 6 BN Serial Number of the button used in the retractor Col 7 FBN Serial Number of the fibre used for the button Col 8 RP Retractor position on the
3. 85 Figure 33 Transmission of the GIRAFFE Low Resolution filters 03 08 Wavelength is in nm FLAMES User Manual VLT MAN ESO 13700 2994 86 7 3 FLAMES calibration times Table 14 lists the calibrations times used for calibrating MEDUSA observations Flat field times are exposure time per frame Typically 3 frames are taken more in the case of blue settings IFU exposure times are twice the MEDUSA values For ARGUS the attached screen flat field exposure times are twice the MEDUSA values ARGUS calibrations using the robot are typically 4 times longer than the MEDUSA ones as the robot first needs to illuminate the sky fibers and then the ARGUS array The quality of the ARGUS robot flats is not as good as for MEDUSA as the illumination of the array is not so uniform For this reason screenflats are preferred for ARGUS although in the blue the integration times required are very long 7 4 Comparison between old and new HR gratings In early October 2003 the high resolution grating on GIRAFFE was changed leading to an increase in sensitivity at the loss of spectral resolutions in some setups A comparison between old and new gratings is shown Table 15 below FLAMES User Manual VLT MAN ESO 13700 2994 Mode Acenter ETC name t arc robot t FF robot t FF screen nm sec sec sec MEDUSA 379 0 HRO1 850 630 5384 0 MEDUSA 395 8 HR02 280 630 577 0 MEDUSA 412 4 HRO03 270 630 566 0 M
4. CCD This usually induces limitations in the maximum attainable S N ratio the measured vs expected S N ratios depart more and more and the measured S N ratio tends to a maximum asymptotic value For FLAMES S N ratios have been obtained in a single exposure up to S N of 400 The departure from the photon noise in this regime was very high We consider this value as the limiting single exposure S N value 3 7 2 Enhanced Dark Current after a FIERA Start up When the CCD Control System FIERA has to be restarted e g due to a unrecoverable error or a general failure of the CCD the level of the dark current will be higher than the value measured in the running system approximately an extra 5 ADU in an hour long GIRAFFE exposure following the shutdown and an extra 0 5 ADU RMS noise It is important to check the performance of the detectors by taking e g a dark exposures of a few minutes in binned mode An interval of 2 3 hours is normally sufficient to return to optimal performance of the CCD 3 8 GIRAFFE Features and Problems 3 8 1 Low counts in blue attached screen flatfields It is almost impossible to get sufficient flux with a reasonable exposure time for the bluest setups for attached screen flats In the bluest three HR settings there are typically only a few hundred ADU detected in a 20 minute exposure time in the bluest part of the CCD Hence accurate flatfielding for the HR1 HR2 HR3 and LR1 settings for ARGUS ONLY is not presently possib
5. FLAMES User Manual VLT MAN ESO 13700 2994 17 Table 4 Full Corrector transmission as function of wavelength and radial distance of the object from center in arcminutes It includes as well pupil decentering effects for a MEDUSA aperture Wave Distance from optical axis arcmin ma 0 2 4 6 8 10 11 12 365 0 789 0 788 0 788 0 784 0 776 0 754 0 735 0 692 405 0 894 0 890 0 888 0 888 0 881 0 842 0 848 0 802 486 0 919 0 912 0 906 0 902 0 903 0 900 0 890 0 847 586 0 914 0 905 0 897 0 890 0 889 0 895 0 887 0 856 656 0 898 0 887 0 879 0 871 0 869 0 879 0 872 0 842 800 0 880 0 869 0 858 0 849 0 846 0 854 0 854 0 826 1014 0 843 0 830 0 819 0 809 0 805 0 811 0 816 0 792 two for IFUs and one for ARGUS The different fibres have different diameters and lengths and are organized in different slit systems each feeding the spectrographs Finally the light reaches the last two components the UVES RED and the GIRAFFE spectrographs where it is dispersed and detected The next sections describe the FLAMES subsystems as one follows the optical path going from the telescope to the instrument detectors It is intended to guide the user in the selection of the optimal instrument configuration for the observing programme The functionalities of the different subunits are explained and reference is made
6. 1 Object ID MANDATORY 2 Right ascension in hh mm ss ss J2000 MANDATORY FLAMES User Manual VLT MAN ESO 13700 2994 53 3 4 5 6 T 8 LABEL M67 central field UTDATE 2001 04 23 CENTRE 08 51 22 82 11 50 09 4 EQUINOX J2000 0 67_00005 08 50 15 98 11 33 58 6 P 8 16 1 B V 0 614 67_00006 08 50 26 13 11 33 59 2 P 9 80 2 B V 0 034 67_00007 08 51 01 87 11 34 01 5 P S 3 B V 0 008 67_00012 08 52 04 43 11 34 10 2 P 90d 4 B V 0 379 67_00017 08 50 52 99 11 34 16 3 P 7 28 5 B V 0 008 67_00020 08 50 22 52 11 34 20 8 P 553 6 B V 0 018 67_00022 08 50 33 83 11 34 21 6 P 4 66 7 B V 0 209 67_00023 08 50 42 31 11 34 22 0 P 8 69 8 B V 0 013 67_00025 08 51 31 60 11 34 24 2 P 7 48 9 B V 0 064 Figure 19 Example of an input file for FPOSS Declination in dd mm ss ss J2000 MANDATORY Object type MANDATORY G VLT Guide Star Magnitudes ideally between R 9 and R 11 F FACB astrometric fiducial star Field Acquisition Coherent Bundle Due to the limited dynamical range of the technical CCD these stars must not have a difference in R magnitude exceeding 3 magnitudes The absolute magnitude may be from R 8 15 P Program target same as M M MEDUSA target I IFU target J IFU SKY target A ARGUS sky target U UVES target S Sky generic can be allocated by any fibre type Priority 9 to 1 from 9 highest to lower respectively MANDATORY Target M
7. 2003 1 The filter number 2 The central wavelength in nm 3 Resolving power R in MEDUSA mode for old and new HR gratings 4 Average efficiency for old and new HR gratings in percent FLAMES User Manual VLT MAN ESO 13700 2994 000 89
8. ADU 3 1 e rms 65000 ADU 30 sec 1 5 e Pix h up to 10 gt 0 99995 in disp dir Pix 40 50 2098 3008 lt 60um peak to peak FLAMES User Manual VLT MAN ESO 13700 2994 36 E y o Y m Lab curve o e Standard stars E E gQ o PS K 8 b G s eee Sup wb d oO e EI e n 3 D N f O o 600 800 1000 Wavelength nm Figure 13 The ratio of the Quantum Efficiency of Zeus new MIT CCD Nigel Old MIT CCD The curve is based upon both measurements in the lab and also standard star mea surements A factor 2 increase in response at 900 nm is apparent tions From 500 700 nm the QE of both CCDs is the same to within 5 per cent Redder than this the QE of Zeus relative to Nigel rises rapidly being a factor of 2 at 900 nm The saturation level of the replacement CCD is now 65 000 compared to some 40 000 with Nigel Because Zeus is a thicker CCD than Nigel the fringing is also much reduced although the cosmic ray count rate increased Fig 14 shows part of a FLAMES UVES extracted flatfield at 860 nm taken with both CCDs in which this reduced fringing is evident A summary of the properties of the red arm scientific CCDs is given in Tab 10 The detailed QE curves can be found in the UVES database available through the ETC The detector in the red camera consists of a mosaic of one EEV CCD 44 82 and one MIT LL CCID 20 4kx2k CCD this to optimize the detector response as a function of wavelengt
9. FF exposure time has to be taken into account when computing overheads Note that in an attached NasA flatfield in the bluest three HR settings there are typically only a few hundred ADU detected in a 20 minute exposure time in the bluest part of the CCD Hence accurate flatfielding for the HR1 HR2 HR3 and LR1 settings for ARGUS ONLY is not presently possible within service mode These four settings are therefore only offered in VISITOR MODE IN ARGUS to allow for more time to take the attached flatfields in the daytime Even then the number of derived counts obtained at the far blue end may only be around a thousand in a few hours For Medusa and IFU the robot flats may be used so all setups are accepted in Service or Visitor mode e Daytime Calibrations 0 minutes Bias frames flatfield and Th Ar calibration lamp exposures are taken only during the day with the same instrument and detector setup as the science exposures Standard calibrations are carried out automatically by the Observatory No overheads need to be accounted for FLAMES User Manual VLT MAN ESO 13700 2994 60 5 The Calibration of FLAMES Data 5 1 General Concept Given the possibility of using two spectrographs in many setups the possibility of obtaining suitable calibrations has been a constant concern for FLAMES The operation concept relies on the fact that all necessary calibrations can be taken during the day and they have an accuracy level to guarantee that a sky
10. FITS file middle of the chip in y They each affect a single column longer dimension of the CCD and are almost parallel to the echelle orders They would appear as broadish emission lines in the bluer part of the extracted spectrum of a faint object Nigel Old MIT CCD In the MIT LL chip red side of the CCD mosaic there is a trap in the column X 1609 which might show up as a slight depression over 130 pixels in the extracted spectrum of one order Zeus New MIT CCD In the MIT LL chip red side of the CCD mosaic there is a bad column at 1254 weakly visible in the bias images FLAMES User Manual VLT MAN ESO 13700 2994 43 4 Preparing the Observations 4 1 Introduction Before the actual execution of observations several steps have to be taken The preparation of an observing programme is split in two parts Phase I and Phase II In Phase I i e the application for VLT observing time the emphasis is put on the scientific justification and on the technical feasibility of the proposed observations For the specific case of FLAMES the proponents must clearly show that they have or will have the proper target list including astrometry prior to Phase II In Phase II the successful applicants have to prepare their detailed observing plan including the instrument setups using via the Phase II Preparation P2PP tool Prior to Phase II however it is fundamental that the applicants have prepared the proper files containing
11. Target Setup File column 6 FLAMES User Manual VLT MAN ESO 13700 2994 72 Col 14 Comments User comments from Target Setup File column 8 The table FITS header contains additional information from the configuration process The following compiles the most important keywords Keyword Example Value Comment Information from Target Setup File een FILENAME w_Cen_ COMMED8 025151 Configuration file name CENRA 201 700124999997 13 26 48 03 Field centre mean RA Degrees CENDEC 47 5219444444436 47 31 20 0 Field centre mean Dec Degrees CENEQNX 2000 Equinox of Field Centre FK5 Julian ALLOCGUI 4 Number of allocated FACB stars ALLOCOBJ 122 Number of allocated objects ALLOCSKY 18 Number of allocated sky positions Information from Configuration Process ee ACTMJD 52808 9479166665 Actual MJD of tweak time ACTUTC 2003 06 18T22 45 00 Actual UTC of tweak time ATMPRES 745 7 Atmospheric pressure millibars ATMRHUM 8 Atmospheric relative humidity percent ATMTEMP 13 4 Atmospheric temperature celsius PLATE 1 Identifier of the used positioner plate FACBWLEN 730 FACB wavelength nm GIRAWLEN 679 7 GIRAFFE wavelength nm UVESWLEN 580 UVES wavelength nm A dE VARGUS Was USE 0 SF sss 5 Ss stra sSe TSS A AS O AS SAR ARGSUSED T Flag indicating if ARGUS used ARGSCALE 1 1 ARGUS Scale ARGPOSAN 90 Position Angle of ARGUS on sky Degrees ARGANGLE 0 Orientation of ARGUS Degrees
12. a system of thermally compensating bars that move one lens inside the camera Note however that the collimator changes cannot be compensated with this system and that these are instead compensated through the slit focus movement 3 5 4 GIRAFFE scientific CCD Carreras At the end of P80 the old blue optimized GIRAFFE CCD Bruce was replaced by a thicker red optimized chip Carreras The quantum efficiency of this new chip is significantly higher for wavelengths larger than 700nm e g a gain of factor of 2 at 880nm Between 400 and 700nm the quantum efficiency essentially remains the same as before In the extreme blue lt lt 400 nm a quantum efficiency loss up to 25 was measured in the lab but could not be confirmed on sky yet Figure 10 shows the quantum efficiency ratio between Carreras and Bruce Thanks to the thickness of this new CCD 40m fringing is strongly reduced 6 times smaller peak to valley amplitude at 880 nm see Fig 11 The overall geometry of the Carreras num FLAMES User Manual VLT MAN ESO 13700 2994 30 80000 LR8 new LR8 old 70000 60000 50000 40000 30000 20000 2805 500 1000 1500 2000 2500 3000 3500 4000 4500 Pixel Figure 11 Extracted flat fields for the LR08 setup centered at 881 7 nm We compare the new chip Carreras blue curve and the old one Bruce black curve The two data sets were obtained with the same exposure times and with the same flat
13. and left on a porch located just outside the plate The disposition of the fibres on the plate s is similar in that MEDUSA UVES FACB IFU and IFU sky retractors are disposed in the same way on the two plates every even numbered retractor is a MEDUSA one e Trolley Main structure holding the plates The trolley can perform two main move ments it can approach or retract the Nasmyth adaptor rotator to engage the plate or disengage it Furthermore it can rotate the structure holding the plates in order to exchange the plates between the adaptor rotator and the positioner robot e DA system and gripper robot This unit is at the very core of the whole system It grips and releases the magnetic buttons at the positions reached via the R 0 polar robot The gripper requires a back illumination system that means some light shining through the fibres from the spectrograph towards the plate A video system records this back illumination light and performs an image analysis for two purposes first to reach the required high position accuracy of the optical center of the fibre button and second to detect if the magnetic button is properly picked by the gripper and properly released on the plate The polar coordinates of a placed fibre are stored in an internal permanent memory NVRAM and are used for the next positioning e OzPoz is equipped with a calibration box which moves with the gripper This calibration box hosts an optical s
14. disabled The large majority of the failures are due to dust on the plate and or on the button itself and are readily solved within 1 2 days by cleaning the plate and button However it might happen that a button remains disabled for a longer time due to a more severe problem If the configuration file provided by the user makes use of a broken fibre the problematic fibre is simply ignored by the system which places all available fibers leaving the broken fibres parked Thus if during phase IT preparation a given target has been allocated to the problematic fibre no data will be produced for this target There is an additional detail Since FLAMES has two observing plates available the fact that a given fibre is disabled in one plate does not mean that the fibre with the same fiber number is also disabled in the other plate For instance if object XYZ 1 has been allocated to fibre Medusa 116 and this fibre is broken in plate 1 but not in plate 2 then a spectrum of XYZ 1 will be produced when the user s OB is observed with plate 2 whereas with plate 1 no data will be collected Since the amount of broken fibres is small 1 2 per plate most of the objects will have data produced by both plates However if the user has a have to observe target he or she must pay attention to the list of broken fibres and not allocate this particular and precious target to any of the fibres in the list of broken fibers Note that since FPOSS doesn t kno
15. each plate They are identical to the object IFUs with the exception that only 1 fibre takes the light from the central microlens The fibres of each IFU are organized in a special way on the microlens array to guarantee the maximum of contiguity between fibre exit and fibre input The output of the twenty IFU fibres plus the Sky IFU fibre constitute an IFU subslit with therefore twenty one fibres In total the IFU slit is composed of fifteen subslits In addition five subslits contain in addition a calibration fibre fed by the GIRAFFE simultaneous calibration box The center to center distance between the fibres in the subslits is only 1 47 times the fibre diameter core which implies that the contamination between adjacent fibres is rather high about 10 in any case a even higher level of contamination is always present at the fibre entrance level in normal observing conditions 3 4 5 IFU Orientation Fibres are kept in constant tension by the springs in the retractors and the buttons are free to float in the gripper before being placed When a button is placed the fibres are always oriented along the line which joins the position of placement and the retractor parking position as seen in the FPOSS mimic This means that the IFUs are always oriented within a few degrees with the long side perpendicular to the conjuction fibre retractor that is the long FLAMES User Manual VLT MAN ESO 13700 2994 26 Table 7 Summary of GIRAFF
16. field lamp The fringing is strongly reduced on the new chip 6 times smaller The higher efficiency of Carreras is also well visible a factor of about 2 at this wavelength ber of pixels in and X and Y directions over prescan regions is the same as Bruce s Carreras is equipped with a continuous flow cryostat which provides a high mechanical sta bility The liquid nitrogen tank is exchanged every 14 days and a seal between the CCD and the enclosure ensures the thermal insulation of the spectrograph Carreras is controlled by a FIERA controller We have opted not to offer any possibility of changing the CCD readout characteristics i e windowing variable read out speed and binning are not possible for the user Note that on chip binning would undersample some regions of the spectra also in MEDUSA mode A compromise has been adopted for the read out speed the chosen setup works at a read speed of 225 kPixel sec which implies a reading time of 43 seconds and excellent read out noise 4 3 e Pixel at 225 kPix sec reading speed and low gain The CCD working temper ature is 120 C to minimize some of the blemishes and to ensure negligible dark current le Pixel hour The cosmetic quality of Carreras is excellent The electronic glow that was well visible on the FLAMES User Manual VLT MAN ESO 13700 2994 31 Table 8 Measured properties of the GIRAFFE scientific CCD Carreras Quantum efficiency 56 Q 350nm 98
17. is not much room for maneuver Usually if some observations require repetition the plate will need to be reconfigured the field be re acquired and re executed after another OB has been executed 6 2 Pointing and Guiding FLAMES is not equipped with any auxiliary slit viewer or imaging system in addition to the 4 FACBs therefore the whole system relies on the relative accuracy between the targets the VLT guide star and the 4 FACBs After the telescope has been pointed the images from the 4 FACBs are recorded on the techni cal CCD and the centroids in each of the quadrants are computed and offsets calculated The fiducial stars in the FACBs are therefore the ONLY sources linking the geometry of the plate to the sky The target VLT guide star and fiducial FACBs coordinates must be in the same reference system i e their coordinates must be computed from the same astrometric solution with a relative accuracy better than 0 3 arcseconds to avoid wasting telescope time To help the users in their observations preparation ESO has performed a pre FLAMES stellar survey using the Wide Field Imager WEI at the ESO MPI 2 2 m telescope http www eso org science eis In order to guarantee a correct centering and offset calculation FACB reference stars should not have close visual companions within 3 arcseconds Telescope Guide Stars should have magnitude R between 9 and 11 FACB stars should be brighter than R 15 and be of comparable ma
18. mode the retractor positions of fibres adjacent to the simultaneous calibrations are as follows Medusa plate 1 retractor positions 24 44 84 104 144 164 194 224 264 Medusa plate 2 retractor positions 24 44 84 106 130 170 204 224 264 3 8 3 In focus Ghosts and Scattered Light A 3 scattered light level is observed in the reddest 300 pixels of the spectrum the feature is rather sharp and most likely caused by a white light ghost i e a reflection inside the collimator it affects all the 2048 pixels along the slit direction 3 8 4 CCD Defects e The old detector Bruce until P80 The EEV chip Bruce has very few cosmetic defects The most noticeable is a hot column which affect all the pixels of row 420 redder than pixel 1270 This column does not affect the same spectrum in all setups due to slight shifts between the different setups The most affected spectra are spectrum number 24 in MEDUSA mode which corresponds to button 58 but a movement of 1 spectrum depending on the chosen setup and on the long term spectrograph spectrograph stability is possible Other defects can be generally eliminated by flat fielding In the upper red corner i e at fibres with high button numbers a rectangle of about 350 x 200 pixels shows a higher level of counts in long exposures up to about 60 electrons in one hour This feature is generally referred as the glow e The new detector Carreras after P80
19. reflection on the grating the second order spectrum is re imaged by the camera on the CCDs These ghosts appear as in focus echelle orders with a steeper inclination and approximately twice the order separa tion than the primary spectra The effect is relevant with the cross disperser 4 at the far FLAMES User Manual VLT MAN ESO 13700 2994 42 end of the spectrum central wavelength 860 nm where the efficiency of the CCD decreases and the importance of reflections is higher On the MIT chip upper CCD in the mosaic reflections from the EEV lower CCD of the mosaic are seen For reference see the flat field data available on the UVES web page The relative intensity of the ghosts to the primary echelle orders depends on the shape of the target spectrum With flat fields and at a central wavelength of 860 nm the intensity of the ghost orders is lt 1 of the real orders on the EEV chip and between 1 and 7 on the MIT chip 3 9 6 CCD Cosmetic Defects The CCDs of UVES are of excellent cosmetic quality The number of hot or less sensitive pixels is limited lt 0 1 and has little effect on the quality of the data because of the good sampling The defects which depending on the signal to noise of the spectrum might be visible in the extracted data are listed below In the EEV CCD blue side of the CCD mosaic there are four trails of hot pixels which appear in long exposures X coordinates 3896 3963 4052 and 4140 in an unbinned
20. subtraction to better than 2 is possible To achieve this goal a high photometric fibre stability is required and an overall high instrument setup repeatability and stability All calibrations are carried out using the calibration unit of the Fibre Positioner see next section and Section 2 3 With the exception of the attached Nasmyth FF calibrations the observer is not supposed to prepare any calibration OBs calibrations will be provided by the Observatory following the FLAMES Calibration Plan 3 5 2 Positioner Calibration Unit The positioner calibration unit has been conceived to provide the user with the following performance e Fibre to fibre relative flux illumination flat field relative illumination better than 0 3 This ensures that accurate fibre to fibre relative transmission can be derived using the flat spectra acquired with the positioner e Integration time per button less than ten seconds For a flat field level of at least 8000 e7 pixel and for a Th Ar level of at least twenty lines per GIRAFFE setup This is valid for the majority of the setups In the BLUE setups calibrations require longer integration time and the targeted flux cannot always be obtained This performance has been obtained using a mixture of hardware and operations The fibres to be calibrated are first positioned following a spiral pattern on the plate In order to guarantee the same flat field relative illumination it is necessary to minimize l
21. the MEDUSA values see 5 2 Configuration time is excluded FLAMES User Manual VLT MAN ESO 13700 2994 88 Fil TA Order Order AA nm AA nm bi R AvEf AvE 4 nm old new old new old new old new 1 379 0 15 15 16 7 16 7 22 500 22 500 29 0 43 5 2 395 8 14 14 19 5 19 5 19600 19600 26 7 50 3 3 412 4 14 14 16 8 16 8 24 800 24 800 30 3 36 9 4 429 7 13 13 20 5 20 5 20 350 20 350 36 4 50 7 5 447 1 13 12 17 6 24 7 26 000 18470 29 2 48 3 6 465 6 12 12 22 1 222 20 350 20 350 41 8 60 9 7 484 5 12 11 19 0 27 4 26700 18529 32 1 45 6 8 504 8 11 11 24 6 24 6 20000 20000 44 2 57 1 9 525 8 11 10 21 3 31 0 25900 17750 38 4 39 0 10 548 8 10 10 28 0 28 0 19800 19800 44 4 58 4 11 572 8 10 10 24 3 24 3 24 200 24 200 43 4 41 6 12 599 3 9 9 32 5 32 5 18 700 18700 38 2 58 4 13 627 3 9 9 28 6 28 5 22 500 22 500 47 6 51 9 14 651 5 9 8 24 3 39 4 28 800 17740 28 7 54 2 15 679 7 8 8 35 9 35 9 19300 19 300 44 5 61 3 16 710 5 8 8 31 3 31 3 23 900 23 900 42 0 44 1 17 737 0 8 7 26 5 45 9 30200 17 425 25 6 64 8 18 769 1 7 7 42 0 42 0 18 400 18 400 46 2 65 4 19 805 3 T7 6 36 9 59 0 22200 13867 42 0 50 0 20 836 6 E 6 31 4 56 0 28 600 16036 29 8 70 6 21 875 7 6 6 51 7 51 7 16 200 16 200 42 9 62 6 22 920 5 6 5 45 9 74 9 19000 11642 50 7 LT Table 15 Comparison between the GIRAFFE HR grating used pre and post October 10
22. the spectral format for the different FLAMES fibre types on the raw images FLAMES User Manual VLT MAN ESO 13700 2994 76 7 1 1 GIRAFFE MEDUSA MEDUSA FES gt E T E i 4000 3000 Di x a 2000 D 1000 O O 500 1000 1500 x pixel Figure 24 Schematic layout of the MEDUSA spectral format blue solid lines object fibres red dots calibration fibres The direction of the increasing fibre number in the slit FPS and increasing wavelength A are indicated FLAMES User Manual VLT MAN ESO 13700 2994 7 1 2 GIRAFFE IFU IFU configuration shown for PA 0 deg East PSSN 1 20 4 20 11 10 41 Plate 3 18 19 12 9 2 3 Y gt North 2 17 116 13 8 5 4 1 15 14 7 6 1 2 3 4 5 6 X Notes 1 Position Angle PA 315 deg ORIENT in binary OzPoz table PA North East 2 For IFUs with SKY fibers the PSSN numbers should be increased by 1 TT IFU Fiber m Retractor
23. therefore limiting the dead time between observations to less than 15 minutes e A link to the UVES spectrograph RED arm via eight single object fibres per plate e A high and intermediate resolution optical spectrograph GIRAFFE with its own fibre systems in three possible configurations MEDUSA IFU ARGUS e A coordinating observing software system that allows simultaneous UVES and GI RAFFE observations The operation of FLAMES requires that the observer has her his own target coordinate list with a relative astrometric accuracy better than 0 3 arcsec rms at the time of the Phase 2 proposal preparation The minimum object separation is 11 arcsec which is limited entirely by the size of the magnetic buttons The Fibre Positioner is able to position the fibres with an accuracy better than 0 1 arcsec peak to peak In addition to the targets the user must also provide coordinates for one VLT guide star and four fiducial stars in the same astrometric solution as the targets The VLT guide star is used to first point the telescope and to close the active optics loop while the four fiducials are used to correct this pointing for further small offsets in coordinates due to corrections of the field geometry Ideally it should have an R magnitude of between 9 11 2 1 UVES FIBRE mode UVES is the high resolution spectrograph of the VLT UT2 It has been designed for working in slit mode only but was modified to add a fibre mode
24. to their measured performance Efficiencies e g in the form of transmission curves of the main instrument components including the CCDs are available in the FLAMES database accessible through the GIRAFFE and UVES Exposure Time Calculators see 4 7 1 3 2 Corrector The optical corrector is a doublet of BK7 equivalent lenses of 900 mm diameter In order to maintain a good transmission over a large wavelength range the lenses have been coated with a single layer of Mg The function of the corrector is to give an excellent image quality over the whole 25 arcmin FLAMES field of view and to provide a pupil located at the center of curvature of the focal plate The corrector is mounted with a cross support onto the Nasmyth adaptor rotator The support also hosts the three attaching points for the Positioner fibre plates When the whole optical train is taken into account including telescope optics and vignetting the effective transmission of the corrector depends on the observing wavelength and on the distance of the object to the field center as expressed in Table 4 The corrector and therefore the FLAMES plates are positioned in the optical path AFTER FLAMES User Manual VLT MAN ESO 13700 2994 18 the VLT guide probe This implies that the guide probe will vignet the field of view It is therefore very important to select carefully the VLT guide star VLT guide star should have an R magnitude between R 9 and R 11 for optimal pe
25. 17 07 07 04 Sect 1 4 Combined cfg A comment 2 25 11 04 Sect 1 2 1 3 2 2 4 5 1 4 1 6 Update for P75 1 7 2 8 1 4 3 5 1 1 2 11 03 05 some Figures quality improved 79 01 09 06 none Version for Period 79 Phase I and II 79 1 25 12 06 Sect 29 added Version for Period 79 Phase II 80 27 02 07 none Version for Period 80 Phase I and II 81 27 02 07 none Version for Period 81 Phase I and II 82 08 03 08 none Version for Period 82 Phase I and II 82 1 05 06 08 Sect 3 5 4 updated Version for Period 82 Phase I and II Figure 10 new Table 8 updated Table 14 updated Hypertext links updated 82 2 25 06 08 Fig 11 new 83 08 09 08 none Version for Period 83 Phase I and II 83 1 23 11 08 Minor typo changes 84 0 25 02 09 All Version for Period 84 Phase I Visitor Mode only for HR1 3 and LR1 in Argus Figs added to show why sky subtraction can be important Comment about UVES SimCal Refs added for RV accuracy Sky subtraction 84 1 18 06 09 All Figs changed added for FACBs reconstructed IFU Argus images and IFU throughput Other minor typo changes 85 0 28 08 09 All MIT UVES CCD upgrade Calibration plan ammended Guide star magnitudes changed 85 1 18 12 09 Sect 4 3 2 table 13 table 11 Minor modifications FLAMES User Manual VLT MAN ESO 13700 2994 This page was intentionally left almost blank FLAMES User Manual VLT MAN ESO 13700 2994 Contents 1 Introduction 1 1 On the Contents of the FL
26. 2994 68 6 FLAMES Observing Operations This chapter explains in some detail the operation of FLAMES A sound understanding of the rather peculiar FLAMES operations procedure is crucial for good planning of visitor AND service mode observations We summarize again a number of particulars which need to be taken into consideration for FLAMES operations 1 The relative distance between objects is going to change during an observation therefore in general long observations should be avoided 2 Observations of the same objects at different wavelengths or multiple observations of the same objects should be executed only after re positioning of the fibres 3 The Fibre Positioner configures while observing this implies that two OBs are running at the same time Also the positioner needs to know the mean time of the next observation while the current one is still running In practice it gives some rigidity to the whole operation scheme 4 When used in combined mode FLAMES produces UVES and GIRAFFE frames 5 Each FLAMES sub system Positioner GIRAFFE UVES has its own Observing Soft ware OS The complete system is coordinated by the FLAMES Super OS which is the only OS allowed to talk to the Telescope Control Software TCS The frames produced by the spectrographs are complemented with information coming from the TCS 6 1 During the Night Observations with FLAMES are carried out at the Console of UT2 located in the VLT Control Bu
27. 4 reference stars and link the telescope to the targets Flatfield FF Spectrum obtained from light source with a flat i e without spectral fea tures energy distribution e g a tungsten lamp The registered signal provides infor mation about the response of the detector allowing a determination of the variation in sensitivity from pixel to pixel the echelle order shape the presence of bad columns on the detector etc Free Setting A setting of the instrument defined by the observer generally with a different wavelength readout or binning than any of the standard settings No free settings are available with FLAMES Grating The main light dispersing elements of UVES and GIRAFFE are echelle gratings Guide star A point source used for accurate tracking and active control of the telescope mirrors ideally with magnitude between R 9 and R 11 Maintenance Technical procedures developed to control and maintain the quality of tele scope instrument and detector Observation Block OB A logical unit of exposures needed to obtain a coherent set of data Encompasses all relevant information for a successful data acquisition on a target It consists of target information a set of templates parameter files for the templates conditions requirements and comments concerning the specified observations It represents the entity the short term scheduler deals with Constructing Observation Blocks is part of the Phase II Proposal Preparat
28. 8 404 9 19 5 19600 31300 3 63 0 14 H412 4 HRO3 403 3 412 4 420 1 16 8 24800 39000 4 59 1 13 H429 7 HRO4 418 8 429 7 439 2 20 5 20350 32500 5 55 1 12 H447 1A HROSA 434 0 447 1 458 7 24 7 18470 29481 5 63 9 13 H447 1B HROSB 437 6 447 1 455 2 17 6 26000 41500 6 59 1 12 H465 6 HRO6 453 8 465 6 475 9 22 2 20350 32500 7 54 5 11 H484 5A HRO7A 470 0 484 5 497 4 27 4 18529 29632 7 63 9 12 H484 5B HRO7B 474 2 484 5 493 2 19 0 26700 42700 8 58 5 11 H504 8 HRO8 491 7 504 8 516 3 24 6 20000 32000 9 53 3 10 H525 8A HROQA 509 5 525 8 540 4 31 0 17750 28372 9 63 2 11 H525 8B HRO9B 514 3 525 8 535 6 21 3 25900 41400 10 57 3 10 H548 8 HR10 533 9 548 8 561 9 28 0 19800 31600 11 62 0 10 H572 8 HR11 559 7 572 8 584 0 24 3 24200 38700 12 55 6 9 H599 3 HR12 582 1 599 3 614 6 32 5 18700 29900 13 60 3 9 H627 3 HR13 612 0 627 3 640 5 28 5 22500 36000 14 52 6 8 H651 5A HR14A 630 8 651 5 670 1 39 4 17740 28334 14 65 1 9 H651 5B HR14B 638 3 651 5 662 6 24 3 28800 46000 15 8 H665 0 HR15N 647 665 0 679 32 17000 28 000 15 56 4 8 H679 7 HR15 660 7 679 7 696 5 35 9 19300 30800 16 61 1 8 H710 5 HR16 693 7 710 5 725 0 31 3 23900 38000 17 51 7 7 H737 0A HR17A 712 9 737 0 758 7 45 9 17425 27869 17 65 9
29. 8 H737 0B HR17B 722 5 737 0 749 0 26 5 30200 48300 18 55 4 7 H769 1 HR18 746 8 769 1 788 9 42 0 18 400 29400 19 46 9 6 H805 3A HR19A 774 5 805 3 833 5 59 0 13867 22175 19 60 1 7 H805 3B HR19B 785 6 805 3 822 5 36 9 22200 35500 20 49 6 6 H836 6A HR20A 807 3 836 6 863 2 56 0 16036 25511 20 64 9 7 H836 6B HR20B 819 5 836 6 850 9 31 4 28600 45500 21 53 2 6 H875 7 HR21 848 4 875 7 900 1 51 7 16 200 25900 22 43 7 5 H920 5A HR22A 881 6 920 5 956 5 74 9 11642 18628 22 57 9 6 H920 5B HR22B 896 0 920 5 941 9 45 9 19000 30400 Table 11 All 31 high resolution setups of GIRAFFE with 316 lines mm and 63 5 blaze grating These setups are valid for observations taken after October 10th 2003 when a new HR grating was installed A comparison between the old and new HR gratings is given in Sect 7 4 The B settings always have lower efficiency than the A settings In particular at Acent the efficiency ratios are approximately as follows H447 1B H447 1A 0 6 H484 5B H484 5A 0 7 H525 8B H525 8A 0 7 H651 5B H651 5A 0 5 H737 0B H737 0A 0 3 H805 3B H805 3A 0 7 H836 6B H836 6A 0 25 H920 5B H920 5A 0 7 The H665 0 setting covers both Ha and Li6707 Previous to P74 settings such as H447 1A H484 5A etc were called H447 1 H484 5 without the A suffix Note that for ARGUS normal programmes that HRO1 HRO2 and H
30. A 400 nm 87 500 nm 86 600 nm 90 700 nm 81 800 nm 54 900 nm Thickness 40 ym Number of pixels 2048 x 4096 Pixel size 15 wm Gain low 2 35e7 ADU Readout noise 4 3e7 rms Saturation low gain 60000 ADU Full frame readout 225 kPix sec 43 sec Dark current levels 120 C 1 e Pix h Cosmic hit event rate Non linearity Fringing amplitude 880 nm CTE Readout direction Prescan Overscan areas 3 14 0 18 event min cm lt 0 5 up to 5 gt 0 99999 in disp direction Pix 1 50 and 2098 2148 top right corner of Bruce images has disappeared No bad columns rows are present and only one hot pixel affecting some of its neighbours was detected A summary of the GIRAFFE Carreras CCD characteristics for the readout mode 225kHz 1x1 low which is the only one presently offered is given in Table 8 3 5 5 Spectral Format and Efficiency In GIRAFFE the spectra are parallel in dispersion along the long side of the detector i e in readout direction while on the short side is parallel to the slit Spectra are curved with the central part closer to each other than the edges The lines of constant wavelength describe arcs of low curvature with respect to the CCD pixels This implies that the wavelength coverage is slightly shifted by a few ngstroms between the fibres at the edge and the fibres at the center of the slit which are shifted to the red An example of a T
31. AMES User Mamual 1 2 Information available outside this Manual 2 Capabilities of the Facility UVES FIBRE mode 5 4440 4485 be ORO SERRE R ARE EEE OOS GIRAPPE s ec e eh a Oe a AS SE A A FLAMES Observing Modes 2 00 pee eee eee ene Limitations and Caveats 2 ca A 8 hee KE RE EK eee Da Bee Oe ee ee FLAMES within the VLT Observatory 02 00002 eae High resolution spectrographs at ESO La Silla FLAMES Sample Observations and Calibrations Ed IY s su ce ek tet Ree SEGRE EERE DSS Se OEE SES CN AG DI 2 4 2 5 2 6 f 28 29 MI den oe eee E mee Eee A E eoa AR e E E 2 10 Abbreviations and Acronyms 3 FLAMES Characteristics and Sub Systems 3 1 3 2 3 3 3 4 3 9 3 6 Opto mechanical Layout lt 4 22 26 eb ed eee Ee eae ee es SOMOS y paa A A A ROE BEEN ADA wo Pie Positioner TOPS ois eue ENEE A RAY A A 3 3 1 Positioner Performance Characteristics Buttons and Fibre SysteMS o 0 002 eee eee 3 4 1 3 4 2 3 4 3 3 4 4 3 4 5 3 4 6 ag es ARA UVES Ewe ee ee eee le eed Ge bee wb eS a MEDUSA TORES es ee Gok eR A wR Be ee e IEU EE EEN IFU Orientation ss o s A Aert RRR AE NEE E EE E E ARGUS DIES carl desa de o et eA A EE Ee AE EE GIRAFPE cra ar ARE A A 3 5 1 3 5 2 3 5 3 3 5 4 3 9 9 3 5 6 3 5 1 CU WM lt 4 4 2s om Bea Rae ra AA Filters and the Filter Wheel Dioptri
32. C so AA eas HA SRE ES SEER Ee ERE SEES Simultaneous Calibrations 2 4 e e ri eae eb aa ee ew ee ee es vil 38 38 38 38 38 39 39 39 40 40 40 40 41 41 42 43 43 44 45 45 47 48 52 53 54 54 56 56 57 57 57 57 58 58 FLAMES User Manual VLT MAN ESO 13700 2994 viii Oo Lonasio Cala da ea A BE CSR we ES 61 Sol GIRAFFE Leonesa Unib cirios e ew oS 61 90 The UVES Callbration Umit ceca ae ei hae ee aa Coka Ee RA 61 5 7 Fibre to Fibre Transmission and Sky Subtraction issues 62 DS ape ial CalprationS lt gt idas rad OES EA DEER EE ERE E GH 63 S81 Detector Plata e eo lenceria OS 24 Se EE oH x 63 5 8 2 Use of Telluric Standard Stars to correct for Fringing or atmospheric Lines 63 5 9 FLAMES Science Calibration Plan 0 00 000 eee nee 66 6 FLAMES Observing Operations 68 Gil D ring the Mie o se esc a sice a saian e a aa EM ERE ES 68 G2 Pomtne and Guiding s lt s X ssori da Mee enG Sane E E eR ES 69 6 3 ARGUS fast observations ooo a a a e 70 6 4 Evaluation of the Results Offline Data Analysis 70 6 5 FLAMES Raw Data Structite 4 ssc cocosii peod biete GeO 71 Od HDU OzPoztable lt lt lt span we aoe Ow Oe EH RES 71 65 2 HDU3 FLAMES FIBRE Table 2 0 46 2 oe ee ww ke ee ea 72 7 Appendix 75 Til FLAMES Raw Data Spectral Format oe es 6 ee a ew se er Ba 75 TLI GIRAFFE MEDUSA s ja EEN AAA RA 76 Gil GIRAFFE IF U coca RR ee Ked a aa e T7 TL3 GIRAPRE lt ARGUS
33. E spectrograph characteristics Type Echelle Order selection Filters Collimator beam 180 mm Collimator aperture F 5 HR Echelle 204 x 408 mm 316 lines mm 63 4 blaze angle LR Grating 156 x 204mm 600 lines mm 38 blaze angle Camera Focal Length 360 mm mean Detector 2048 x 4096 15 ym EEV CCD Scale 0 19 arcsec pixel Slit height 76 8 mm side of the IFU 3 arcseconds is perpendicular to the fibre the short 2 arcseconds is oriented along the fibre retractor direction See Fig 25 Fig 26 shows which IFU fibre is which in the reconstructed image produced by the ESO pipeline 3 4 6 ARGUS Fibres The ARGUS system is a fixed array of 14x22 microlenses similar to the IFUs located in the center of Plate 2 ARGUS is further equipped with a focal enlarger system allowing to switch between a scale of 0 52 microlens 1 1 to a finer scale of 0 3 microlens 1 1 67 In addition to the object fibres fifteen ARGUS Sky fibres are available on the plate they are built identical to the IFU Sky fibres i e with only the central fibre present The ARGUS array is best flatfielded using attached screen flatfields Fig 9 shows the ARGUS array geometry and Fig 23 reconstructed ARGUS images produced by the ESO pipeline 3 5 GIRAFFE GIRAFFE is a fully dioptric spectrograph with a beamsize of 180mm and is able to support the 76 8mm longslits fed by the different GIRAFFE fibre systems It has been conceived to minimiz
34. EDUSA 429 7 HRO4 95 470 189 0 MEDUSA 447 14 HRO5A 455 420 144 0 MEDUSA 447 1B HRO5B 785 630 288 0 MEDUSA 465 6 HRO06 280 315 98 0 MEDUSA 484 54 HRO7A 425 250 50 0 MEDUSA 484 5B HRO7B 965 250 140 0 MEDUSA 504 8 HROS 160 380 30 0 MEDUSA 525 8A HRO9A 710 240 60 0 MEDUSA 525 8B HRO9B 710 315 80 0 MEDUSA 048 8 HR10 220 190 25 0 MEDUSA 572 8 HR11 345 285 30 0 MEDUSA 599 3 HR12 235 190 14 0 MEDUSA 627 3 HR13 220 190 18 0 MEDUSA 651 54 HR14A 815 140 10 0 MEDUSA 651 5B HR14B 815 240 22 0 MEDUSA 665 0 HRI5N 125 95 8 6 MEDUSA 679 7 HR15 125 125 8 3 MEDUSA 710 5 HR16 160 125 7 9 MEDUSA 737 0A HR17A 110 60 7 9 MEDUSA 737 0B HR17B 110 315 16 7 MEDUSA 769 1 HR18 140 60 5 0 MEDUSA 805 34 HRIOA 140 30 3 3 MEDUSA 805 3B HR19B 300 60 4 6 MEDUSA 836 64 HR20A 160 30 2 4 MEDUSA 836 6B HR20B 160 110 11 8 MEDUSA 875 7 HR21 125 50 2 1 MEDUSA 920 5 HR22A 110 60 2 7 MEDUSA 920 5 HR22B 160 95 4 5 MEDUSA 385 7 LRO1 330 630 300 0 MEDUSA 427 2 LRO2 125 110 57 0 MEDUSA 479 7 LRO3 125 60 40 0 MEDUSA 043 1 LRO4 170 60 12 0 MEDUSA 614 2 LRO5 125 60 7 5 MEDUSA 682 2 LRO6 110 60 5 0 MEDUSA 773 4 LRO7 30 30 2 1 MEDUSA 881 7 LRO8 amp 45 16 1 5 UVES 6FIB 520 80 30 80 UVES 7 1 8FIB 580 60 20 40 UVES 8FIB 860 60 20 40 87 Table 14 Integration times for ThAr arcs and W flats for both Robot calibrations and attached Screen flats IFU calibs robot amp screen and ARGUS screen FF times are twice the MEDUSA values ARGUS robot calibration times are 4 times
35. EUROPEAN SOUTHERN OBSERVATORY Organisation Europ ene pour des Recherches Astronomiques dans H misph re Austral Europ ische Organisation ftir astronomische Forschung in der s dlichen Hemisphare ESO European Southern Observatory Karl Schwarzschild Str 2 D 85748 Garching bei Miinchen Very Large Telescope Paranal Science Operations FLAMES User Manual Doc No VLT MAN ESO 13700 2994 Issue 85 0 Date 18 12 2009 A Kaufer D Naef C Melo J Smoker C Martayan is AA EE EE Date Signature A Kaufer PROVE HE Date Signature C Dumas tl A IRA Date Signature FLAMES User Manual VLT MAN ESO 13700 2994 This page was intentionally left blank FLAMES User Manual VLT MAN ESO 13700 2994 Change Record 111 FLAMES User Manual VLT MAN ESO 13700 2994 iv Issue Rev Date Section Parag affected Reason Initiation Documents Remarks 0 5 15 09 01 all draft 0 6 Apr 02 all draft 0 7 Sep 02 all CfP P71 appendix added 1 0 21 03 03 all First Release for P71 1 1 21 07 03 all Updates and Corrections Results ARGUS commissioning Release for P72 1 12 24 10 03 some HR update IFU geometry calibration times added 1 13 15 01 04 some Minor changes for P73 1 14 01 03 04 Table 3 1 New HR settings added 1 15 04 03 04 Table 3 1 New HR setting added 1 16 18 06 04 Sects 1 4 1 5 2 2 3 1 3 3 1 Update for P74 Sects 4 4 5 1 1 6 2 7 3 Figs 2 6 Tables 3 1 7 1 7 3 1
36. Each FACB bundle has an effective diameter of 2 4 An example of a TCCD image with the four stars in the FACB bundles is shown in Figure 4 Positioning Software This is based on a well tested and complex code developed initially for the 2dF system at Anglo Australian Observatory AAO This so called delta task allows crossing of the fibres in their final positions and determines the button movements sequence from a given fibre configuration to the next solving a traveling salesman problem It is worth noticing that to reach a new complex configuration more than one move per fibre may be needed A more detailed explanation can be found in the FPOSS manual 3 3 1 Positioner Performance Characteristics The main performance characteristics of the Fibre Positioner can be summarized as follows 1 2 N DO Ooh A Q 3 4 Mechanical Positioning accuracy better than 0 08 arcsec Positioning time 6 seconds move The whole configuring time also depends on the number of moves necessary to re configure the plate that is the number of moves per fibre needed to re position one button as well as by the number of attempts per button needed to achieve the required positional accuracy Possibility to configure the next observation while observing Plate Exchange time less than 180 seconds w o considering field acquisition Calibration unit equipped with Th Ar Ne and FF lamps Performing fine centering of the plate on the sky
37. For a given spectral format the spectral resolution is very uniform along the chip Measure ments show a variation of the resolving power along the chip of less than 4 rms It is important to recall that the efficiency can change substantially within the same setup by almost a factor two if the wavelength of interest is close edge of the order The ETC gives a very reasonable representation of the final spectral format 3 5 6 GIRAFFE Setup Stability and Repeatability GIRAFFE was built to be mechanically very stable the slit exchange mechanics were designed to ensure high setup repeatability the basic requirement being that the calibrations made during the day would well reproduce the night observations This has been achieved as confirmed form tests in Garching and Paranal The flexures due to temperature variations measured in GIRAFFE in long term tests amount to 0 4 pixels K along and perpendicular to dispersion which translates into typical shifts of 0 2 pixel in the 12 hours night day interval on the detector The number given above includes the repeatability of the setup since the FLAMES User Manual VLT MAN ESO 13700 2994 33 tests were done by swapping setups between the tests Note that GIRAFFE has no thermal compensation in the axis perpendicular to the dispersion so that occasional large changes in the temperature can cause bigger shifts than those noted above In these cases the calibrations need to be repeated as the extra
38. HR3 and LR1 need to be observed in visitor mode only This only affects Argus all Medusa and IFU settings will continue to be observed in Service or Visitor mode e The atmospheric effects depend on the wavelength of observation The VLT pointing and guiding is made for a given wavelength While it is possible in P2PP to specify two different observing wavelengths for the UVES and GIRAFFE fibres the VLT will point and guide only to the GIRAFFE wavelength in this combined mode This implies that the pointing will be correct but if the airmass is changing drastically during the observations and the UVES and GIRAFFE wavelengths differ dramatically then the UVES A is disregarded and the UVES fibres may loose a considerable amount of light For the same reason it is important that if the observer wishes to observe the same objects at different wavelengths two OBs and therefore two different fibre positionings are made repositioning the fibres for the correct wavelength We finally note that in a combined observation if the UVES part is more important then the GIRAFFE configuration wavelength can be chosen to be close to the UVES wavelength e g H572 8 in the case of using UVES 580 nm Of course this would lead to entrance losses in the GIRAFFE part if the GIRAFFE observation wavelength is different from the above e In general long e g longer than 60 minutes for most declinations see 4 4 and or repeated observations of the same objects ar
39. RO3 are available in visitor mode only due to the difficulties of obtaining sufficient counts in the attached screen flatfields For Medusa IFU all settings are available in Service or Visitor Mode 1 Filter number 2 Exit angle 3 Wavelengths at the edges and center of the CCD in nm 4 Resolving power R for MEDUSA IFU ARGUS FLAMES User Manual VLT MAN ESO 13700 2994 47 Table 12 All 8 low resolution setups of GIRAFFE with 600 lines mm and 34 0 blaze grating Note that for ARGUS normal programmes that LRO1 is available in Visitor Mode only due to the difficulties of obtaining sufficient counts in the attached screen flatfields For Medusa IFU all settings are available in Service or Visitor Mode Fil 7 Ord p2pp ETC Astart Acentor Aenda AA Resolving Power e H name nm nm nm nm MED IFU amp ARG 1 32 4 5 L385 7 LRO1 362 0 385 7 408 1 46 1 8000 12800 2 27 9 4 L427 2 LR02 396 4 427 2 456 7 60 3 6000 10200 A 32 2 4 L479 7 LRO3 450 1 479 7 507 8 57 8 7500 12000 4 26 3 3 L543 1 LRO4 501 5 543 1 583 1 81 6 6000 9600 5 30 6 3 L614 2 LRO5 574 1 614 2 652 4 78 3 7400 11800 6 34 9 3 L682 2 LRO6 643 8 682 2 718 4 74 6 8600 13700 7 24 7 2 L773 4 LRO7 710 2 773 4 834 3 124 0 5600 8900 8 29 0 2 L881 7 LRO8 820 6 881 7 940 0 119 0 6500 10400 4 3 2 UVES Standard Setting
40. S FLAMES P2PP htm1l Tutorial If the observations have been prepared carefully using FPOSS to define the FLAMES modes and targets to fibre assignments as saved in the Target Setup File and the ETC to define the required instrument setups and exposure times the use of the P2PP tool almost trivial one acquisition template and one or several observing templates have to be combined in one OB 4 8 1 Acquisition Templates There are 4 acquisition templates for FLAMES The first three correspond to the three in strument modes UVES COMBINED and GIRAFFE The fourth template is available in visitor mode only and is the fast acquisition in the ARGUS mode of GIRAFFE only In the first three templates the observer has to fill in the same 2 parameters only a Name of the Target Setup File created by FPOSS to be associated to the template via a file selector box b Observing wavelength from a pull down menu with all FLAMES standard setups In case of combined observations both the GIRAFFE and UVES observing wavelengths have to be indicated Finally the fourth template is for the fast acquisition mode of ARGUS where no FPOSS setup file is needed See the template reference manual for details FLAMES User Manual VLT MAN ESO 13700 2994 58 4 8 2 Observing Templates There are 5 observing templates one for UVES one for GIRAFFE one for COMBINED and two for ARGUS In the first three cases only a few parameters are required to be
41. The EEV chip Carreras has very clean cosmetics It has no known bad column or row The number of dark pixels i e pixels with flux values less than 50 of the local mean is 62 There is only one hot pixel located at 185 3040 This hot pixel affects the counts recorded for some of its neighbours The exact number of affected surrounding FLAMES User Manual VLT MAN ESO 13700 2994 40 pixels depends on the CCD temperature typically about 10 pixels at the operating temperature 3 9 UVES Features and Problems We note that in early July 2009 the RED MIT CCD of UVES Nigel was replaced by Zeus that has a higher throughput redwards of 700 nm and less fringing See Sect 3 6 2 s 3 9 1 UVES SimCal lamp too bright for long exposures When observing with the FLAMES UVES 7 1 setting the strength of the simultaneous ThAr calibration lamp is set for an exposure time of 15 minutes Unlike the SimCal lamp in GIRAFFE there is no filter available in UVES to reduce the flux for long exposures This leads to saturation in some lines in long integration times 3 9 2 Fibre Overlap in the 520 nm Setup Below 500 nm the order separation becomes too small to accommodate all 8 UVES fibres without overlap Therefore the decker of the UVES slit masks the UVES fibres 8 and 9 corresponding to buttons 103 and 235 respectively if the 520 nm setting is selected If these two fibres have been assigned to targets the light will be lost Hence only 6 fi
42. agnitude Target program ID integer number Comments Fields not labeled as MANDATORY are in fact optional for FPOSS For possible additional information contained in the Comment field see the FPOSS manual 4 5 1 Run FPOSS to Prepare the Target Setup Files The FPOSS manual provides a full explanation of its use here we give just a brief summary of the general concepts The use of the FPOSS follows the following simple steps 1 Loading of the input file with Guide Fiducial and Target stars 2 Selecting the VLT Guide star 3 Selecting the FLAMES observing mode 4 5 6 7 Assignment of the fibres to target objects Assignment of the fibres to sky positions if needed Checking configuration over hour angle range Saving of the fibre assignment to the Target Setup File FLAMES User Manual VLT MAN ESO 13700 2994 54 Once the saving is performed a Target Setup File is generated containing all the information necessary for the observation with the exception of the spectrograph setup This file is of utmost importance The files contains a checksum and must not be edited If the file is edited the following process P2PP will not proceed and will not be able to create OBs Only Target Setup Files created by the FPOSS are accepted by the system 4 6 Broken and low transmission fibres Although FLAMES fibres are mechanically very stable the gripper might occasionally fail to move them leading to the faulty fibre to be
43. amp variability This is achieved by sweeping the gripper above the fibres several times the gripper movement is so accurate to guarantee the same illumination time fibre to a level of a fraction of percent These flats can be therefore used for both flat fielding and fibre to fibre transmission mea surement purposes In the case of Th Ar exposures the gripper moves from one fibre to the next and dwells on the fibre for a given amount of time then goes to the next For ARGUS calibrations both flats amp arcs the robot first illuminates the sky fibers by sweeping in front of them with its lamp on Then it illuminates the ARGUS array which is located at the center of plate 2 see Fig 9 by scanning over it with the lamp on As a consequence the total times exposure time is twice the IFU one half for sky fibres and half for the array FLAMES User Manual VLT MAN ESO 13700 2994 61 5 3 Nasmyth Screen To perform a very accurate sky subtraction it may be useful to acquire Nasmyth screen flat fields or so called attached flats These flats are obtained by illuminating the closed Nasmyth shutter with halogen lamps after the observations The fibres are not moved at all from their observing positions during this type of calibration They are maintained with the same geometry and torsion property This ensures the minimum difference between observing and calibration conditions but on the other hand especially for wavelengths blue
44. ample Input Fil o sc o bee ee be ers a EE ee RE Bs 53 Raw GIRAFFE IFU images of the solar spectrum on plate 1 top and plate 2 bottom taken in May 2009 Variations in the IFU responses on each plate are clear although the absolute level depends on the solar illumination so this figure should not be used to compare the two plates 0 0 55 Image taken at L881 7 nm showing many sky lines 64 Extracted spectra at LST lt soo so bo crisis ras 65 Argus reconstructed image with Argus position angle in the acquisition set to 45 degrees Top panel Original pointing Bottom panel Telescope moved by 1 0 arcseconds North and 1 0 arcseconds East i e the object moves 1 0 arcseconds South and 1 0 arcseconds West on ARGUS 74 GIRAFFE MEDUSA Spectral Format 76 GIRAFFE IFU Spectral Format o lt 6 4 er det etd ed 05444 084 KH 77 Reconstructed image of 15 IFU units produced by the pipeline 78 GIRAFFE ARGUS Spectral Format aoaaa 79 UVES FIBRE Special Format s o oe a roau dre Oe Se e E 80 GIRAPPE Filters HE DEDO A ros i soio e AER OR E ROR Re RD OS 81 GIRAFFE Filters HR DESTA Zoor tankio a a eR er 82 GIRAFFE Filters HR 13 18 NEE NEEN RRA ES 83 GIRAFFE Filters HR 19 22 LR 01 02 e EEN RRR Ew GE 84 GIRAFPE Filters LR 03 08 6 04 644 0a doa EN EE ORE a a REE 85 FLAMES User Manual VLT MAN ESO 13700 2994 1 1 Introduction 1 1 On the Contents of the FLAMES User Manual The curren
45. ape with time as the observed field moves across the sky Figure 16 illustrates the amplitude of this effect for an object located 9 arcmin away from the field center as a function of the hour angle of the observations and the target declination FLAMES User Manual VLT MAN ESO 13700 2994 49 Paranal P 743 mbar T 11 5 C FLAMES field 25 arcmin E A EA SE a pT ab T o A i pk f Ki g S ace lt ae A IA ON E E 6 5 4 3 2 1 0 1 2 3 4 55 6 HA hours Figure 16 Distance between the field center and an object located 9 arcmin away from it as a function of hour angle and declination of the field The dashed horizontal lines indicate the loci of constant 2 2 5 and 3 airmasses from bottom to top respectively Computed for a wavelength of 400 nm FLAMES User Manual VLT MAN ESO 13700 2994 50 Paranal P 743 mbar T 11 5 C FLAMES field 25 arcmin A O O E O ae ee aere av Asis 12 5aremin arcsec HA hours Figure 17 Same as previous figure but for an object located just at the edge of the field 12 5 arcmin from the center Also for a wavelength of 400 nm FLAMES User Manual VLT MAN ESO 13700 2994 ol Distance arcsec a E E A airmass Figure 18 Relative displacement between the central wavelength A 400nm and nearby wavelengths covered by typical GIRAFFE gratings Each curve co
46. aware of its amplitude in different observing conditions in order to correctly plan the observations A analytical formula that takes into account all the parameters affecting the amount of the differential atmospheric refraction has been given by Filippenko 1982 PASP 94 715 Ac cording to his calculation the refraction index n of the atmosphere at a certain wavelength A is given by 29498 1 255 4 P 1 1 049 0 0157T x 10 6P n A 1 107 x 164 328 Lal Es 146 Ac 41 172 720 883 1 0 003661T where P is the atmospheric pressure in mm Hg typically 557 25 for Paranal and T is the tem perature in C typically 11 5 for Paranal Once the refraction index is properly determined the displacement of the observed astronomical object with respect to its position without the atmosphere is dr X n A tan z where z is the zenith distance of the object The problem when observing a large field of view is that an object at the field corner will have a zenith distance z different from that at the field center z hence its observed position will be displaced with respect to the real one by a different quantity The size of this difference is proportional to tan z tan z and therefore varies non linearly with the zenith distance of the whole field Since the telescope is guiding with respect to the center of the field this effect causes distortions at the field edges that change sh
47. bres are available in the 520 nm setting 3 9 3 Fibre to fibre Contamination Given the limited separation between the UVES fibres a small degree of contamination exists between one fibre and its neighbour on the slit This can be seen easily in Figure 15 showing a trace of three UVES orders in direction perpendicular to the dispersion three groups of 8 fibres are seen and the reader can notice that the flux level in the interfibre is higher than the interorder light and that the flux at the base of one fibre is slightly overlapping with the neighbouring one The contamination can be divided into two main components 1 Diffuse light this light depends primarily on the total amount of light injected in the spectrograph it follows the echelle intensity curve and is estimated at the level of 0 2 pixel of the adjacent fibre overall intensity This implies that if 8 stars of similar intensity are observed their overall contribution to diffuse light will be at the 1 level of a single source however diffuse light has no spectral features and it appears as a continuous source 2 Fibre to fibre direct contamination the wings of two adjacent fibres slightly overlap and this gives a direct contamination including spectral features of one fibre to the next This value is however very low and it increases from 0 13 to 0 5 going from an integration over 5 to 7 pixels One fibre has a PSF FWHM of 4 5 pixels Gaussian PSF approximation b
48. bserving modes MEDUSA IFU ARGUS In addition it allows a simultaneous acquisition of UVES and GIRAFFE observations with the specific observing modes listed in Table 3 It is important to note that during a combined observation the exposure times for UVES and GIRAFFE do not need to be the same but the longest exposure time will determine the overall length of the observation Table 3 Summary of the various single and combined modes of FLAMES Spectrograph Mode Single Modes UVES a 8 target fibres 580 nm or 860 nm setups UVES b 7 target fibres 1 Simul Calib Fibre 580 nm setup only UVES c 6 target fibres 520 nm setup GIRAFFE MEDUSA GIRAFFE IFU GIRAFFE ARGUS Combined Modes UVES GIRAFFE UVES a b or c MEDUSA UVES GIRAFFE UVES a b or c IFU UVES GIRAFFE UVES a b or c ARGUS 2 4 Limitations and Caveats FLAMES is a complex instrument because of the different modes available and the multi object capability In order to operate it efficiently a number of limitations had to be imposed e The observer is responsible for the accuracy of the input catalogue A relative accuracy of better than 0 3 arcseconds rms is required to limit the losses FLAMES User Manual VLT MAN ESO 13700 2994 8 due to fibre object mismatch VLT guide star and fiducial stars must have coordinates in the same reference system as the objects No cr
49. by another MIT Zeus Figure 13 shows the Quantum Efficiency ratio of Zeus to Nigel plotted as a function of wave length both from measurements in the lab and also from preliminary standard star observa FLAMES User Manual VLT MAN ESO 13700 2994 Table 9 UVES echelle and cross disperser gratings Echelle gratings g mm Resolving Spatial Blaze Blaze Power resolution angle Eff RED 31 6 2 100 000 0 09 75 04 63 Cross disperser gratings g mm Wav range Average Wav of Peak Blaze nm Eff Eff am Eff CD 3 600 420 680 gt 60 520 68 CD 4 312 660 1100 gt 70 770 80 35 Table 10 Measured properties of UVES RED scientific CCDs Sting EEV and Zeus MIT Zeus is the replacement for the old MIT CCD Nigel and has higher QE above 700 nm and reduced fringing See Figs 13 and 14 EEV MIT LL Quantum efficiency Number of pixels Pixel size Gain low Readout noise Saturation low gain Full frame readout 225 kPix sec Dark current levels 120 C Fringing amplitude 850 nm CTE Readout direction Prescan Overscan areas Flatness 89 450nm 89 600 nm 2048 x 4096 15 um 1 47e7 ADU 3 4 e7 rms 65000 ADU 30 sec 0 5 e Pix h up to 40 gt 0 99995 in disp dir Pix 1 50 and 2098 2148 lt 30m peak to peak 84 800 nm 64 900 nm 2048 x 4096 15 ym low 1 41e
50. carioca A kae A E a 79 FLA UVES PIBRE escocesas eniad AA A 80 7 2 Characteristics of GIRAFFE Filters 81 Go FLAMES calibration times 2 44 06 sa cca eR A racao asasta 86 7 4 Comparison between old and new HR gratings 86 List of Figures O oo N1 on RAUN R Schematic view of an Integral Field Unit MEDUSA entrance losses due to fibre object decentering View of the Fibre Positioner and GIRAFFE on the Nasmyth A Platform 16 TCCD image of the Fiducial Stars acid we asrida saes keh 19 Histogram of Fibre Transmission at 600 nm aa a 00004 21 UVES Fibre Bundles and Slit Geometry 22 Schematic view of Microlenses EE EE a 23 MEDUSA Fibre Bundles and Slit Geometry 25 ARGUS Microlens Array and Slit Geometry 27 Quantum Efficiency ratio New chip Old chip 29 Fringing Old and New chip comparison e cessc os owna coe a e 30 FLAMES User Manual VLT MAN ESO 13700 2994 ix 12 13 14 15 16 17 18 19 20 21 K i 23 24 25 26 a 28 29 30 Al 32 33 GIRAFFE Spectral Format and Slit Curvature 006 32 QE Zeus II 36 Fingme tor Zeus and Nigel ru 8 ans AAA A ERA 37 UVES Fibre to Fibre Contamination 41 Atmospheric Dispersion Effects at 9 arcmin from center 49 Atmospheric Dispersion Effects at 12 5 arcmin from center 50 Chromatic Atmospheric Dispersion Effects 00 51 FPOSS S
51. cm wide Because grating masters of this size cannot be ruled a new process was developed in which a replica is made of two precisely aligned masters The result is called a monolithic mosaic and has a resolving power on the order of 2000000 and a stable Line Spread Function The groove density and hence the order length was selected such that the order length at 990 nm is equal to the CCD length Further information on the echelle and cross disperser gratings can be found in Table 9 The cross disperser unit is a grating turret with two gratings mounted back to back Se lection of the grating is done by rotation of the unit the angle of the grating is automatically set according to the required wavelength of the central echelle order The properties of the red cross disperser gratings 3 and 4 can be found in Table 9 The Camera is dioptric with an external focus to facilitate detector exchange Focus is set manually and then maintained automatically by thermal expansion rods in the camera support structure The red camera has unvignetted entrance apertures of 230 mm focal length of 500 mm and fields of 87 mm diameter Its image quality is 20 ym on axis to 30 um in the corners diameter of circle containing 80 of the energy The transmission curves can be found in the UVES database available through the instrument ETC 3 6 2 Scientific CCD Mosaic STING ZEUS In early July 2009 the MIT UVES red CCD Nigel was replaced
52. ction of fibre data are very sensitive to such shifts between science and calibration frames The grating position is monitored daily and the results are available here http www eso org observing dfo quality GIRAFFE reports HEALTH trend_report_STABILITY_HC html 3 5 7 GIRAFFE Calibration Units In addition to the calibrations performed through the Positioner calibration Unit illuminating all the fibres sequentially cf 3 3 GIRAFFE is equipped with two calibration units Simultaneous Calibration Unit SCU In order to limit the use of time consuming night calibrations in each GIRAFFE mode 5 fibres are devoted to the acquisition of simultaneous Th Ar spectra illuminated by the SCU during the science integration The unit is equipped with a tunable neutral density filter which allows good Th Ar exposure levels for integration times between 2 and 120 minutes If not deselected in the observing template these 5 SCU spectra are acquired automatically Note that for faint objects the SimCal spectra can contaminate neighbouring spectra especially for settings redward of 650 nm see Sect 3 8 2 Users can choose to take a 60 s exposure with the lamp ON then the long integration science target with it OFF then a 60 s exposure with it ON again in order to eliminate the possibility of contamination Longslit Calibration Unit One of the GIRAFFE slits is equipped with a longslit unit fed by a calibration system with an integrating sp
53. d 860 nm settings which have an aperture of 1 arcsecond each and are separated by 1 47 arcseconds The red camera is dioptric no central obstruction and provides an external focal plane for easy detector interfacing and upgrading during the lifetime of the instrument together with a large field good image quality and high optical transmission In the red arm a mosaic of two 4096 x 2048 pixels CCDs is offered separated by about 1 mm loss of one order in the gap The direction of the spectral dispersion echelle orders is along the larger dimension of the CCDs The instrument spectral formats wavelength coverage etc are always computed for these fixed CCD window settings The Arm Selector unit has four positions Free direct feed to the red arm Mirror 1 to feed the blue arm Dichroic 1 and Dichroic 2 to feed both arms In fibre mode the backside of Mirror 1 is used to feed the red arm of UVES with the light from the eight FLAMES fibres The working position of this unit is determined automatically by the instrument software once the instrument observing mode is selected 3 6 1 The RED Spectrograph Arm The red mirror collimator consists of two off axis parabolas and two flat mirrors It is of the white pupil type with two 200 mm pupils one for the echelle and one at the cross disperser camera which results in a moderate size of the optical components and a simplified design The red echelle grating is 840mm long and 210
54. d at a time e g it is not possible to observe simultaneously with MEDUSA IFU FLAMES User Manual VLT MAN ESO 13700 2994 45 In summary all the possible modes of FLAMES include 1 UVES 8 fibres to targets in 580 or 860 nm setup 1 arcsec diameter each 2 UVES 6 fibres to targets in 520 nm setup 1 arcsec diameter each 3 UVES7 1 7 fibres to targets 1 simultaneous calibration fibre illuminated with a Th Ar lamp only in the 580 nm setup 4 GIRAFFE MEDUSA 131 fibres to targets 1 2 arcsec diameter each 5 simultaneous calibration fibres illuminated with a Th Ar lamp 5 GIRAFFE IFU 15 movable rectangular Integral Field Units 2x3 arcsec each made of an array of 20 fibres 15 sky fibre units 6 GIRAFFE ARGUS Single fixed Integral Field Unit consisting of 14x22 microlenses with scale of either 0 52 or 0 3 arcsec each 7 Any combined simultaneous UVES or UVES7 1 plus GIRAFFE mode two simulta neous GIRAFFE modes are not possible However in order to insure a manageable calibration database only a limited amount of setup combinations such as CCD setting are offered It is also important to note that in the COMBINED modes the guiding can be performed only at one given wavelength even if the two spectrographs have different central wavelength settings This could result in some efficiency losses for the cases when the UVES and GI RAFFE central wavelength are far apart and or when the chromatic atmospheric
55. e dispersion The instrument does not include a remotely controlled focus adjustment since the cameras automatically compensate for temperature variations of the complete optical train from slit to CCD The image quality over the entire spectral range is better than 20 30 ym over the full CCD 80 of the energy This implies that in practice no noticeable variations are expected between different exposures The measured resolving power in Fibre mode is FLAMES User Manual VLT MAN ESO 13700 2994 38 R 47000 The efficiency of the RED arm of UVES in fibre mode is about 40 lower than UVES in slit mode for observations of a single point source It is rather difficult however to accurately predict the real differences in flux collection be tween the two modes because they will be function of the seeing and of how accurate fibre centering has been performed 3 7 FLAMES Features and Problems This section lists some of the features and problems common to FLAMES Several of these topics are very important for efficient observations with FLAMES 3 7 1 Maximum reachable S N ratio Fibre systems when coupled to spectrographs suffer from small photometric instabilities whose relevance depends on many factors including the fibre type the fibre system design the spectrograph design basically the full path from fibre entrance to detector This instability shows up as time variable fringing additional to the fringing produced by the
56. e Spectrograph gn lt ics sro te swa OE BEG EOE BEA Ee GIRAFFE scientific CCD Carreras o o ca roepa rre EN E Spectral Format and Edicieney e oos sor edsane se aabt eae Ps GIRAFFE Setup Stability and Repeatability GIRAFFE Calibration Units UVES FIBRE mode s e EE Ee resina 3 6 1 3 6 2 3 6 3 The RED Spectrograph ATM A EE e eaae E ALE Scientific CCD Mosaic STING ZEUS aaau aaa aaa Spectral Resolution and Overall Efficiency vi N e NI NOt Cu D 10 11 11 12 12 15 FLAMES User Manual VLT MAN ESO 13700 2994 3 7 3 8 3 9 FLAMES Features and Problems 0606 eee eee eee es 3 71 Maximum reachable S N Table ooo 3 7 2 Enhanced Dark Current after a FIERA Start up GIRAFFE Features and Problems e 3 8 1 Low counts in blue attached screen flatfields 3 8 2 Contamination from Simultaneous Th Ar Calibrations 3 8 3 In focus Ghosts and Scattered Light 384A CCD ie pacs muce wok k pug eR ae ee EOE BRE e Sos UVES Features amd Problems 2 sg re Ae ENN PEER ee a Be eS ek ER OS 3 9 1 UVES SimCal lamp too bright for long exposures 392 Fibre Overlap m the 620mm Setup lt 222 4245 4 se he ees 3 9 3 Fibre to fibre Contamination gt o lt a cea ssa careo tarea iu 3 94 Spectral Gaps in th RED ar nas ima diota di r tiis 3 9 5 Optical Ghosts in th
57. e better split in several observations made with different plates Since the geometry of the field will slightly change with time it is anyway recommended to reposition the fibres after each observation FLAMES User Manual VLT MAN ESO 13700 2994 1 T T T T T T B _ DD arcsec A Eer L See A 0 7 arcsec J GC B E _ 0 8 arcsec 4 tay 0 8 H Bo SC pay dvaresec _ Ze aR eee Ze SC N L TESS Pi ye 1 2 arcsec J 8 ECH SS NN 1 4 arcsec Sie L L J Se K doc SCH e p 0 6 aks Sla 3 E J fr on Ke a a T S Se e Ba AER E TA SA ES z RES Se ee E o A ES SS e 2 GE SR SH SE 0 4 E al Sk E No Se an W NN DS E Wee ee SS ol ae en SC St A CON 3 es ie 0 2 E _ 0 I il iT 1 I 1 0 0 2 0 4 0 6 Object Fibre centre displacement Figure 2 MEDUSA entrance losses as a function of seeing and object fibre decentering This plot shows how much flux can be lost due to bad astrometry The reader should evaluate the impact of the astrometric errors in their full statistical sense FLAMES User Manual VLT MAN ESO 13700 2994 10 e The positioning time is about 10 seconds fibre or 20 minutes for MEDUSA This implies that OBs shorter than 20 minutes will suffer considerable deadtime before the next observation is started In these cases the duty cycle is very bad another instrument is perhaps more suited for the observations Note that due to the implementation of the UVES slit the positio
58. e expected execution time of the OB as provided by the user and computes the mid time of the new observation The coordinates are transformed into FLAMES User Manual VLT MAN ESO 13700 2994 69 plate R positions the back illumination is switched on about 30 milliseconds fibre and the buttons are placed Once the configuration of the plate is completed it receives an unique identifier say plate 1 and a validity time stamp is generated When the OB is re run on BOB obs it looks for the status of plate 1 Finding it at the configuration position i e at the robot and knowing that no other observation is running on the telescope side it sets the telescope rotator back to the home zero position disengages the current plate rotates the tumbler and engages the new plate At this point the center field coordinates are sent to the telescope and to the rotator The VLT field acquisition can start search for guide star closing of the active optics loop and field stabilization The four FACBs check that the fiducial or reference stars are indeed in the right position or of needed telescope offsets are computed and applied The acquisition template is now finished and the observing template can start While running the observing template on BOB obs the next OB can be selected and run on BOB config which after some sanity checks starts the configuration of the next field Clearly this cycle is very critical and once started there
59. e far red Spectra aooaa a 3 9 6 CCD Cosmetic Defects 2 aoa ee bee RE a a a a 4 Preparing the Observations 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 5 The 5 1 G2 5 3 5 4 POCONO A sore e A eee ee AE DES ee es Se EE e we ES FLAMES Modes and basic Choices 0000 eee eee GIRAFFE and UVES Standard Settings 244 a6 40 ee oe A EE OO 4 3 1 GIRAFFE Standard Settings 2 0000020 432 UVES Standard Settings esc ke eA Se eR Se ee REE Differential Atmospheric Effects 2 20 0 0 cee ee Preparing the Target Input Files 0 0 0 eee eee ees 4 5 1 Run FPOSS to Prepare the Target Setup Files Broken and low transmission fibres oc 2b ewe Re ewe ee eS Introducing the Observation Blocks lt lt uo ses eek ee wk we Rex 4 7 1 GIRAFFE and UVES Exposure Time Calculators 4 7 2 Choice of the Sample Target o e s ec se dac canre raa kuron 4 7 3 Choice of Instrument Configuration and Spectral Format 4 7 4 Exposure Time and predicted Counts and S N Ratios PPP tOO k e eck oO ae ee ek AROE E A RUE POR ee Be koa ASL Acquisition Templates s 44a pi po we ORR ee OL RH OL ROR SH aoe Observing Templates s y or YK ee E EEN AA 4 8 3 Computing Time Overheads for your Programme Calibration of FLAMES Data General Concept lt lt oz cier E e bee ha dhe EE RES Ow Positioner Calibration Unite 64 4 a dak web G eee KS SOE SHE Nasmyi h ORC
60. e fibre end to be placed on the plate in front of the stellar target Second they are the support of the optical interface between the telescope beam and the fibre A magnet glued at the base of the button enables the fibres to be placed on the curved plate The diameter of the stainless steel button is 6mm 10 3 arcsec but in order to avoid collisions between the buttons the minimum allowed separation of two buttons is set to 11 arcsec The magnet is a tablet of 4mm diameter and 1 2mm high The magnetic attachment force is around 235 gr Single microlenses for MEDUSA and UVES fibres and arrays of microlenses for ARGUS and the IFUs are used to reduce the F 15 telescope beam to around F 5 into the fibres These lenses image the pupil of the telescope onto the fibre entranced surfaces This system helps to scramble the photometric variations produced by oscillations of the star on the fibre by guiding errors In the case of MEDUSA and UVES fibres the optics are rod lenses with their image focal plane on the flat surface The lens has been cut to 45 degrees to fold the telescope beam into the fibre parallel to the plate In the case of ARGUS and IF Us all arrays of microlenses are glued to a single total reflection 45 degrees prism 3 4 2 UVES Fibres Each of the positioner plates hosts eight 54 metre long fibres which bring the light to the UVES spectrograph on the Nasmyth platform B The UVES fibre concept is shown in figure 6 there are two bund
61. e maintenance and night calibrations special requirements were introduced to reduce setup shifts and to obtain accurate re positioning In this way one is able to use calibrations obtained in the afternoon Five fibres can be used to obtain simultaneous calibration spectra and to monitor the instrument drifts After passing from the slit unit the light is sent through the order sorting filters to the collimator The collimated beam is dispersed by one of the two high HR or low LR resolution echelle gratings After passing through the collimator again an intermediate spectrum is produced Finally the F 2 camera produces the image on the 2kx4k CCD see Figure 2 8 A summary of the most relevant GIRAFFE characteristics is given in Table 7 The GIRAFFE spectrograph obtained its name from the first concept for the instrument in which it was positioned vertically on the Nasmyth platform FLAMES User Manual VLT MAN ESO 13700 2994 27 ARGUS input 300 um 0 52 arcsec Spatial sampling 0 52 arcsec 4 2 mm 0 3 arcsec 6 6 mm pp Diagonal 7 8 mm Array of 22 x 14 channels ARGUS ARRANGEMENT Eg S RS BS ES BS ES ES x y E PA 0 BE Spectrograph ES slit ES BS e hal x N PA 0 oO 20 fibres of the Sub slit N 1 20 fibres of the Sub slit N 2 20 fibres of the Sub slit N 3 Sr INTEGRATION IN OZPOZ Sky button Tumbler Argus head o Figure 9 Geometry of the ARGUS microlens arra
62. ed to objects and 15 IFUs dedicated to sky measurements the latter contain the central fibre only A schematic representation of an IFU in its button is given in Figure 1 e 1 ARGUS slit The large integral unit ARGUS consists of a rectangular array of 22 by 14 microlenses It is fixed at the center of positioner plate 2 Two scales are available one with a sampling of 0 52 arcsec microlens scale 1 1 and a total aperture of 12 by 7 arcseconds and one with a sampling of 0 3 arcsec microlens and a total coverage of 6 6 by 4 2 arcseconds scale 1 1 67 15 ARGUS single sky fibres are also available These can be positioned within the 25 arcmin field cf Figure 9 The ARGUS long axis is along the N S direction for a position angle of 0 degrees with the PA entered in FPOSS being measured North East See Fig 23 Users should pay attention to the small list of broken fibres see Sec 4 6 GIRAFFE is operated with 39 fixed setups 31 high resolution 8 low resolution whose characteristics are given in Table 11 and Table 12 FLAMES User Manual VLT MAN ESO 13700 2994 7 For performance estimates based on measured transmission curves and performances the user is referred to the Exposure Time Calculator http www eso org observing etc A summary of the GIRAFFE characteristics is given in Table 1 including estimated best performance S N ratios 2 3 FLAMES Observing Modes The FLAMES observing software OS coordinates the various o
63. ee Sect 5 4 We also refer the reader to values of radial velocity accuracy claimed by a number of authors and shown in Table 2 that should be consulted for more information a The number of allocatable buttons is 132 but only 131 spectra are fully covered on the detector Table 1 FLAMES characteristics and observing capabilities The wavelength coverage Cover is given is nanometers The S N ratio is given per wavelength pixel as in the ETC and it refers to the mean S N ratio in the setups LR4 543 1 nm and HR10 548 8 nm using as inputs a G2 star for point like and elliptical galaxy for extended sources Additional assumptions include 1 hour exposure dark time 0 8 arcsecond seeing airmass 1 2 and a fibre perfectly centered on the object FLAMES User Manual VLT MAN ESO 13700 2994 5 Table 2 References for claimed radial velocity accuracy derived with FLAMES The resultant velocity error is affected by factors such as the grating setup used whether or not SimCal was activated the time period over which the observations were taken and reduction techniques amongst others The references contain much more specific information than it is possible to show in the table and should be consulted for more details Reference Mode Science Magnitudes RV err Battaglia et al MNRAS 383 183 2008 FL GIR Sculptor V 18 1 6 kms Bouchy et al A amp A 431 1105 2005 FL UVE OGLE survey V 15 7 51 ms Koc
64. effects of the atmosphere are relevant during the exposure see 4 4 4 3 GIRAFFE and UVES Standard Settings To facilitate the preparation of Observation Blocks 4 7 standard settings have been defined that allow the observer to select a pre defined instrument setting for which all param eters are fixed at optimal values and only the exposure time and number of observations are left to be decided The observers can only use these standard settings The automatic data processing pipelines are available for these standard settings only The FLAMES standard settings are given in Chapter 2 and repeated here for the sake of completeness 4 3 1 GIRAFFE Standard Settings GIRAFFE standard settings are given for the high and low resolution modes in Tables 11 and 12 respectively Each setting has a unique FITS keyword INS EXP MODE which is the same as the p2pp name given in the tables In these tables in addition to the central wavelengths of the settings their coverage and resolving power is given as measured with Th Ar lines Since the coverage varies slightly from fibre to fibre cf 8 3 5 5 the coverage given in the tables is the coverage interval common to all fibres FLAMES User Manual VLT MAN ESO 13700 2994 46 Filt 6 Ord p2pp ETC ABar ASen Aa AA RMED IFU amp ARC 1 61 1 15 H379 0 HRO1 370 0 379 0 386 7 16 7 22500 36000 2 58 2 14 H395 8 HR02 385 4 395
65. en in Table 5 The fibres of each plate are arranged in one subslit see Figure 6 The nine fibre centers are separated by 1 7 times the fibre core diameter implying that there is some degree of contam ination between adjacent fibres This contamination can be largely reduced by extracting the spectra on the central six or seven pixels Diffuse light is present and since this depends on FLAMES User Manual VLT MAN ESO 13700 2994 23 Focale length in air 3 5 mm o 2 6 m Flat surface gluing of the fibre 4 6 mm Focal plane Index F5 Focal length 3 5 mm in air 2 1 mm Focal plane Index BK7 Focal length 1 74 mm in air Figure 7 Scheme of the different microlens concepts FLAMES User Manual VLT MAN ESO 13700 2994 24 Table 5 UVES fibre transmission The values given here include all losses focal ratio degra dation optics and coupling For wavelengths redder than 600 nm the transmission is constant Variations of a few tens percents between different fibres have been measured Fibre 370nm 400nm 450nm 600 nm UVES 0 36 0 41 0 52 0 61 the overall light injected into the spectrograph the observer should be careful not to expose objects of too different spectral type and or luminosity if absolute spectroscopy is desired The simultaneous calibration fibre is the last one on the slit Figure 6 The two subslit centers are separated by 500 microns see Fig
66. ent not yet recorded in the current version of the user manual consult the FLAMES web page at http www eso org sci facilities paranal instruments flames e For technical information on the instrument not related to an observing programme con tact optics and mechanics Hans Dekker hdekker eso org the electronics systems Walter Nees wnees eso org the CCD detector systems Roland Reiss rreiss eso org the instrument software Peter Biereichel pbiereic eso org fibres and fibre system Gerardo Avila gavila eso org FLAMES User Manual VLT MAN ESO 13700 2994 3 2 Capabilities of the Facility FLAMES is the multi object intermediate and high resolution fibre facility of the VLT Mounted at the Nasmyth A platform of UT2 it offers a rather large corrected field of view 25 arcmin diameter In 2008 FLAMES completed 6 years of operations at the VLT and a conference was held to celebrate this The presentations are available at http www eso org sci facilities paranal instruments flames doc FLAMES 6th Anniversary FLAMES 6th Anniversary html and give a flavour of the science to date performed with the instrument FLAMES consists of several components e An optical Corrector providing excellent image quality and tele centricity over the full field of view of 25 arcmin diameter e A Fibre Positioner hosting two plates While one plate is observing the other one is positioning the fibres for the subsequent observations
67. ere the same Much reduced fringing is obvious b Extraction of a single order of the same fast rotating star with Zeus red and Nigel black Again the fringing in Zeus is much reduced data reduction The cosmetic quality of the scientific CCDs is very good The CCD cryostat is attached to the dioptric camera with the last optical element acting as the dewar entrance window The CCDs are operated at a temperature of 153 K A liquid nitrogen tank ensures continuous operation without manual intervention for 2 weeks The shutter is located between the cryostat window and the camera It is actuated by solenoids with an open close time of 50ms The illumination of the detectors is homogeneous within 50 ms but a minimum exposure time of 0 5 sec is recommended The reader is referred to the CCD webpages of the ESO Optical Detector Team for additional general information on the CCDs and the FIERA CCD Control System http www eso org odt 3 6 3 Spectral Resolution and Overall Efficiency In contrast to slit mode standalone UVES in fibre mode FLAMES UVES the user cannot set the resolving power by choosing the slit width and the resolving power is determined by the projection of the fibre apertures on the CCD The only variable factors which may affect the resolving power are the image quality of the optics including the focus and the alignment CCD effects chip tilt diffusion of photoelectrons charge transfer as well as the echell
68. es doc 11 Quality Control of VLT FLAMES GIRAFFE data Hanuschik et al 2004 SPIE con ference http www eso org sci facilities paranal instruments flames doc 2 9 Glossary Acquisition Accurate positioning of the telescope in order to center the target on the spectrograph slit FLAMES User Manual VLT MAN ESO 13700 2994 13 BIAS frame Read out of the CCD detector of a zero seconds integration time exposure with shutter closed The registered number of electrons per pixel has to be subtracted from a science exposure because these were not created by photons from the source Calibration Procedures to remove the instrumental signature from the scientific data e g subtract BIAS frames and divide by the flatfield Camera GIRAFFE and UVES have dioptric cameras imaging the dispersed parallel beams on the respective CCD detectors Charge Coupled Device CCD Electronic 2D array detector converting photons into electrons Cross disperser grating An echelle spectrograph contains two dispersive elements One is the echelle grating the other one is called the cross disperser grating UVES hosts two cross dispersers each with two different gratings The cross disperser grating determines the distance between the echelle orders Decker Reflecting and movable blades placed in front of the slit and determining its length FACB Fiducial Acquisition Coherent Bundles These are 4 bundles of coherent fibres to take images of
69. es large orders also in the blue The erating can turn on a turntable turret with very high accuracy 0 05 pixel and repeatability 0 05 pixel rms The whole 370 950 nm spectral range is covered by twenty four setups in 10 grating orders The original HR grating was replaced in October 2003 which has lead to an average efficiency gain of 46 per cent and loss of resolution of 15 per cent The low resolution grating has 600 lines mm and a blaze of 34 degrees The whole spectral range is covered with eight setups in 4 orders The grating size is 150 x 200 mm which implies that some vignetting 8 is present due to the geometry of the beam 180 mm In the future the aim is to provide a grating with the same characteristics but a larger size The transmission curve of the low resolution grating is also given in the Appendix FLAMES User Manual VLT MAN ESO 13700 2994 29 eg Gei e ul o ul QE Carreras new CCD QE Bruce old CCD kuch o 600 0 700 0 800 0 900 0 1000 0 Wavelength nm 800 400 0 500 0 Figure 10 Quantum Efficiency ratio between Carreras and Bruce This curve is based on lab measurements made at ESO Garching Sky tests are ongoing After being dispersed the light passes again through the collimator forms a real image at an intermediate focal plane and is finally imaged by a rather complex F 2 fully dioptric camera The camera has seven elements is thermally compensated for focus displacement through
70. from a Th Ar lamp or from a Ne lamp into a fibre In this way FF Th Ar and Ne calibrations can be obtained for GIRAFFE and for UVES The procedure to acquire these calibrations is to first position the fibres to be calibrated on the plate in a given pattern typically along a spiral pattern and then to sweep with the gripper over the buttons illuminating one fibre after the other one by one For FF calibration the procedure is to sweep continuously over the buttons illuminating them several times while for the Th Ar calibrations the gripper stops over each fibre for a number of seconds specified by the user e Field Acquisition Coherent Bundles FACBs Four magnetic buttons are equipped with a system of 19 coherent fibres each This bundle of fibres is used to obtain images of fiducial or reference stars one per bundle The four images are recorded on an ESO technical CCD TCCD the image centroids are computed and the proper offsets are calculated to center the fiducial stars into the bundles These 4 fiducial stars FLAMES User Manual VLT MAN ESO 13700 2994 20 represent the link between the sky and the plate coordinates therefore it is absolutely necessary that they are chosen carefully They must be sufficiently isolated in the same coordinate system as the target stars and of visual magnitude brighter than R 15 Given the limited dynamical range of the Technical CCD the FACB stars should be within a range of 3 magnitudes
71. fter pipeline processing the pipeline products are also forwarded to the offline workstation from where they can be accessed and inspected by the astronomer Standard data reduction tools like MIDAS IRAF or IDL are available for this purpose The rather complex data structure of FLAMES raw frames is described in the next section FLAMES User Manual VLT MAN ESO 13700 2994 Ti 6 5 FLAMES Raw Data Structure Given the complexity of the instrument the data must carry all the relevant information about the objects and the instrument configurations through the whole data flow process In the case of FLAMES two binary tables associated to the detector image are of utmost importance because they contain all the information to associate the spectra to the objects Both GIRAFFE and UVES FIBRE mode FITS data have the same data and header structure and include 3 FITS HDUs Header Units HDU1 image the image header contains in addition to the primary FITS keywords addi tional keywords for the status of telescope positioner spectrograph detector templates etc during the exposure HDU2 OzPoz_table associates the objects to the fibre buttons This is basically the same table as was provided in the Target Setup File for FPOSS plus additional information from OzPoz such as the R and 0 position of each button on the plate and the corre sponding positioning errors The table header contains keywords related to the fibre positioning proces
72. given a Setup grating and central wavelength from pull down menu b Observing time for each exposure Note that UVES and GIRAFFE may have different exposure times c Number of exposures d Switch for the simultaneous calibration GIRAFFE only Set to OFF in case negative effects on your observations are expected Of course in case of COMBINED observations this information needs to be given for UVES and GIRAFFE separately For the final two ARGUS cases the OB must include the infor mation above plus three additional parameters e Number of offsets f List of offsets in Right Ascension g List of offsets in Declination 4 8 3 Computing Time Overheads for your Programme With the Exposure Time Calculator the user obtains an estimate of the observing time needed to reach the desired S N depending on the object magnitude and observing configura tion To compute the total observing time required for the programme one needs to add the time for all actions required to carry on the scientific observation When applying for service or visitor mode observations the computation of the overheads is required and has to be included in the application The following estimates of the overheads must be used and are also the basis for the automatic calculation of execution times within the P2PP tool used for the final definition of the OBs in service and visitor mode e Target Acquisition 13 minutes except ARGUS fast Visitor M
73. gnitude within 3 magnitudes This latter requirement was imposed by the small dynamics of the FLAMES technical CCD 0 4095 Recently a new technical CCD has been installed The dynamics of this new chip is enhanced 0 65 535 but it is still recommended to use FACB stars with similar magnitudes Recent images will help to minimize errors due to neglected proper motion in the targets guide fiducial stars FLAMES User Manual VLT MAN ESO 13700 2994 70 The tracking of the telescope is corrected for errors of low frequency lt 1 Hz by an autoguiding facility The autoguider makes use of a guide star observed by the guide probe in the adaptor rotator that is moved into the telescope beam The guide star is selected by the observer in the input file to the FPOSS and needs to be in the same coordinate system as the targets and of the fiducial FACBs stars Once the telescope acquisition and active optics correction is executed some small shifts may still be present between the telescope and the target coordinates When the four FACBs start working the offsets of the 4 stars are computed and the operator may apply them to center fiducial stars on the FACBs After the centering is considering satisfactory the observing template can start and the science integration proceeds 6 3 ARGUS fast observations In Visitor mode only it is possible to move from field to field and take observations of different science targets at different wavele
74. h and to reduce fringing at far red wavelengths The gap between the two CCDs is 0 96 mm This gap and the non perfect alignment of the two chips require a separate extraction of the spectra of the two chips The CCD control system the ESO standard system FIERA reads the mosaic as a single image with 100 artificial pixels between the two sensitive areas The file has to be split before applying a standard echelle reduction package Windowing of the CCDs is not allowed neither is CCD binning in UVES FIBRE mode Only ONE read out mode of the CCDs is offered in visitor and service mode Low gain fast read out 225 kpix sec 1x1 binning The characteristics of these modes are given in Table 10 The linearity of the CCDs is measured to be better than 1 over the range from 200 e to the saturation limit in 1x1 binning The CCD parameters are periodically re measured at the observatory The updated values are entered in the instrument database and are recorded in the FITS headers for later use in the FLAMES User Manual VLT MAN ESO 13700 2994 37 Extracted counts Normalised counts Nigel 2 5 flatfield Zeus flatfield Nigel norm flux Zeus norm flux 0 1000 2000 3000 4000 0 1000 2000 3000 4000 0 Pixel Pixel Figure 14 a Extraction of a single order FLAMES UVES flatfield for Zeus new CCD red line and Nigel old CCD black line with the counts multiplied by 2 5 the integration times w
75. h Ar wavelength calibration spectrum is given in Figure 12 blue is towards the left The setups have been fixed to guarantee instrument operability By turning the grating dif ferent combinations can be obtained for a given central wavelength and grating The resolving power and coverage are both function of the grating angle FLAMES User Manual VLT MAN ESO 13700 2994 32 Figure 12 GIRAFFE CCD image of a Th Ar calibration lamp The fibres in the center of the slit have lines moved towards the blue left i e fibres corresponding to the center of the CCD will have a slightly redder wavelength coverage than the ones at the CCD edges For both gratings the spectral orders are quite long with respect to the detector In order to guarantee the whole coverage the grating s need to be rotated which causes the differences in resolution and spectral coverage between the different setups A higher angle corresponds to a higher resolution and to a smaller wavelength coverage The setups have also been selected to give a 10 overlap between consecutive setups This enables an easy connection between the different spectral slices of the same object The higher resolving power of the IFU and ARGUS modes compared to the MEDUSA mode is solely due to the smaller size of the fibres which projects to 2 4 2 6 pixels instead of 3 8 4 2 pixels of MEDUSA The spectral coverage for a given setup is the same in the MEDUSA IFU and ARGUS modes
76. h et al AJ 134 566 2007 FL GIR Leol dSph V 18 21 24 kms Loeillet et al A amp A 479 865 2008 FL GIR CoRoT fields V 11 16 30 ms Maxted et al MNRAS 385 2210 2008 FL GIR LM binaries I 14 19 0 6 kms Pancino et al ApJ 661 L155 2007 FL GIR w Cen B 13 15 0 2 0 8 kms Sommariva et al A amp A 493 947 2009 FL GIR M4 V 12 18 lt 0 3kms Tolstoy et al ApJ 617 119 2004 FL GIR Sculptor dSph V 19 5 2 0 kms Udalski et al A amp A 482 299 2008 FL UVE OGLE survey I 14 3 46 ms available With an aperture on the sky of 1 arcsec the fibres project onto five UVES pixels in the dispersion direction giving a resolving power of 47000 In addition to the eight fibres per plate an extra fibre fed by a separate calibration unit is available This fibre is used for simultaneous wavelength calibration in order to obtain very accurate radial velocities Only seven fibres can be devoted to astronomical objects when this simultaneous calibration fibre is used Note that this simultaneous calibration fibre mode is only available in the 580 nm setup For faint objects one or more fibres can be devoted to the sky When used in Fibre mode only the three standard UVES RED setups are offered with central wavelength of 520 580 and 860 nm respectively see the UVES user manual for details http www eso org sci facilities paranal instruments uves doc 2 2 GIRAFFE GIRAFFE is a medium high
77. he detailed technical information on the instrument e g transmission curves of the GIRAFFE filters can be found in the Appendices Chapter 7 The FLAMES Templates Reference Guide 1 contains detailed instructions for the use FLAMES User Manual VLT MAN ESO 13700 2994 2 of the observing and calibration templates The FPOSS manual 2 illustrates the use of the positioner software for the allocations of the fibres to the objects are given as separate documents although they should be considered by the user as PART of the present manual 1 2 Information available outside this Manual If you cannot find a specific piece of information in the FLAMES User Manual or in case you have remaining questions please check http www eso org sci observing or more specifically e For information on the instrument performance Phase I and Phase II proposal prepa ration please contact the User Support Division usd help eso org e For Phase II preparation of Service Mode Observation Blocks OBs follow the instruc tions given in the FLAMES specific P2PP page http www eso org observing p2pp FLAMES FLAMES P2PP html e For questions directly related to your granted observing run in Visitor Mode please contact Paranal Science Operations pso eso org and flames eso org Visitor mode specific information on FLAMES is found at http www eso org sci facilities paranal instruments flames visitor html e For updates on the instrum
78. here The slit is illuminated by an F 5 beam to simulate the F 5 fibre exit This slit unit is mainly used for engineering tests and is equipped with one FF one Ne and one Th Ar lamp 3 6 UVES FIBRE mode In fibre mode part of the pre slit area of UVES see UVES Manual is substituted by a fibre projector which transforms the fibre focal aperture F 3 at the fibre exit into a 25 mm parallel beam In front of the fibre projector a shutter allows the light from the fibres to reach the mode selection mirror where the fibre mode is selected After the light passes through a re imaging F 10 lens the regular red UVES slit and the UVES RED arm is used in fibre mode UVES is exhaustively described in the UVES User Manual only a very short description is given here see http www eso org sci facilities paranal instruments uves The UVES RED arm AA 420 1100 nm is a white pupil type design With a beam of 200 mm the off axis parabolic collimator illuminates the echelle grating of 214 x 840 x 125 mm with a large blaze angle 76 The echelle is used in quasi Littrow mode i e the angle of incidence and the angle of diffraction are equal but in a different plane which maximizes efficiency FLAMES User Manual VLT MAN ESO 13700 2994 34 The grating cross dispersers provide an order separation larger than 10 arcsec at any wave length in the spectral range 420 1100 nm This separation allows to host the 8 UVES fibres for the 580 nm an
79. ields 1 1 1 pix to pix sensitivity variations fibre to fibre transmission fibre localisation fibre PSF modelling blaze correction Slit Flatfields 3 1 7 pix to pix sensitivity variations Attached Fibre Flatfields n or high precision flatfielding Wavelength 1 1 1 dispersion solution resolving power Sim Fibre Order Definition 1 1 1 order and background def Sim Fibre Format Check 1 1 1 dispersion guess solution Bias 5 1 1 master biases bias chars Dark 3 1 30 master darks dark current cosmic rate FLAMES GIRAFFE Science Data Calibration Plan per instrument setting plate fibre mode resolution and central wavelength Calibration num freq purpose 1 days Robot Flatfields 3 1 1 pix to pix sensitivity variations fibre to fibre transmission fibre spectra localisation Attached Flatfields n or high prec flatfielding Medusa Attached Flatfields 3 1 1 high precision flatfielding Argus or IFU Wavelength 1 1 1 dispersion solution resolving power slit geometry Bias 5 1 1 master biases bias char Dark 3 1 30 master darks dark curr CRs IFU Flux Standard 1 1 7 response corr flux calib Attached Flats 3 1 1 rel trans IFU Sky fibres ARGUS Flux Standard E 1 7 response correction flux calib Attached Flats 3 1 1 rel trans ARGUS Sky fibres o r on request only corresponding OBs to be provided by user n number to be defined by user FLAMES User Manual VLT MAN ESO 13700
80. ilding just below the Paranal summit From there all telescopes and instruments are remotely controlled The telescope and instrument operator carries out the observations and checks that they perform correctly the main responsibility of the visiting astronomers is the selection of the OBs based on the sky conditions and on the results of the first observations The GIRAFFE and UVES FIBRE mode raw data are saved in the FLAMES workstation After the data has been transferred to the Archive workstation copies of the files are received on the astronomer s offline workstation and on the pipeline workstation where the automatic data reduction is running The pipeline products are eventually forwarded to the astronomer s offline workstation Note that the UVES FIBRE pipeline has been available from April 2003 the GIRAFFE pipeline from April 2004 Given the necessity to run target assignation at the Fibre Positioner together with target observation at the telescope two BOBs Broker for Observation Blocks are running simulta neously Although the two BOBs are perfectly symmetric and exchangeable for the sake of simplicity we will call BOB obs the one observing and BOB config the one configuring The Target Setup File generated by the FPOSS associating the fibres to the object coordi nates is linked to the OB through P2PP The OB is read by BOB config and the acquisition template is executed The positioner SW knows the actual time th
81. ilities paranal instruments flames doc 2 FPOSS User Manual VLT INS MAN AUS 13271 0079 http www eso org sci facilities paranal instruments flames doc 3 FLAMES Calibration Plan VLT PLA ESO 13700 3248 http www eso org sci facilities paranal instruments flames doc 4 UVES User Manual VLT MAN ESO 13200 1825 http www eso org sci facilities paranal instruments uves doc 5 P2PP Users Manual VLT MAN ESO 19200 1644 http www eso org observing p2pp P2PP tool html Manual 6 Mechanical features for the OzPoz positioner for the VLT Gillingham et al 2000 SPIE conference http www eso org sci facilities paranal instruments flames doc 7 Installation and commissioning of FLAMES the VLT Multifibre Facility Pasquini et al 2002 The Messenger 110 1 http www eso org sci publications messenger 8 Installation and first results of FLAMES the VLT multifibre facility Pasquini et al 2002 SPIE conference http www eso org sci facilities paranal instruments flames doc 9 The Data Reduction Software for GIRAFFE the VLT medium resolution multi object fiber fed spectrograph Blecha et al 2002 SPIE conference http www eso org sci facilities paranal instruments flames doc 10 Toward accurate radial velocities with the fibre fed GIRAFFE multi object VLT spectro graph Royer et al 2002 SPIE conference http www eso org sci facilities paranal instruments flam
82. ion files for all Standard Settings of the instrument Template A set of instructions for the performance of a standard operation on an instru ment typically an instrument and detector setups The templates represent specially devised sequences for all instrument operations and calibrations Template Signature File This file is a description of a Template and its parameters It contains information about the type and allowed ranges of the parameters some of the parameters have to be set by the observer Wavelength calibration Spectrum obtained from a reference emission line lamp e g Th Ar The wavelengths of the many emission lines are accurately known and are used to transform pixel space into wavelength space FLAMES User Manual VLT MAN ESO 13700 2994 15 2 10 Abbreviations and Acronyms AT BOB CAL CCD CD ESO ETC FLAMES FPOSS FRD IFU OB OS P2PP RTD STD SM TSF UVES VLT VM Acquisition Template Broker for Observation Blocks Calibration exposure Charge Coupled Device Cross disperser European Southern Observatory Exposure Time Calculator Fibre Large Array Multi Element Spectrograph Fibre Positioner Observing Support Software Focal Ratio Degradation deployable Integral Field Unit Observation Block Observation Software Phase II Proposal Preparation Real Time Display Standard star Service Mode Template Signature File Ultraviolet and Visual Echelle Spectrograph Very Large telescope Vi
83. ion Process Order Separation Filters In GIRAFFE the wavelength range covered in each setup is defined by using filters as predisperser inserted in the beam they reject all the light outside the defined bandpass which instead is dispersed by the echelle grating FLAMES User Manual VLT MAN ESO 13700 2994 14 Phase II Proposal Preparation P2PP During this phase the successful applicant whose Phase I proposal has been accepted based on the scientific rationale and technical feasi bility prepares the Observation Blocks to carry out the observing programme Focal Plates The Fibre Positioner can host up to 4 plates these are metallic spherical surfaces where the fibre buttons are positioned for the observations Only two plates are only currently in use Pre slit area UVES optical elements located in front of the spectrograph slits Spectrograph arm UVES consists of two separate spectrographs one optimized for the blue blue arm and one for the red wavelength region red arm Only the red arm is connected to the fibres Spectrograph slit Two parallel reflecting metal blades with an adjustable separation slit width form the entrance slit of the spectrograph The image of the astronomical source produced by the telescope is focused on the slit plane Standard Setting A pre defined setting of the instrument facilitating the preparation of the observations The Observatory keeps an updated database of the relevant calibrat
84. l instrument for observations in service mode carried out by the observatory staff i e in absence of the applicant All the information necessary to the execution of the observations has to be provided to ESO in the form of Observation Blocks prepared through the P2PP tool following the instruc tions sent to the applicants The Observatory staff will combine the execution of different programmes in the same night optimizing the time sequence seeing and moon requirements Observations carried out with the applicant present at the telescope are referred to as visitor mode observations In this mode the astronomer prepares or finalizes the OBs at the Obser vatory in advance of his her nights He she decides about the sequence of observations during the night but their execution is however still performed by the telescope and instrument operator To facilitate the preparation of Phase I and Phase II proposals in addition to the information provided in this User Manual ESO has developed sophisticated Exposure Time Calculators ETC one for GIRAFFE and one for UVES Fibre see 4 7 1 The ETC permits one to estimate the signal to noise ratio for a given configuration and exposure time taking into account specific atmospheric conditions and determines the spectral format resulting from the selected instrument setup The Observation Blocks OB are prepared using another ESO provided software tool called P2PP see http www eso org
85. le in UVES fibre mode with FLAMES When used in slit mode the RED arm of UVES is about 2 times more efficient than the FLAMES fibre link The multiplex advantage of using the FLAMES fibre link with respect to the slit mode can therefore be 3 4 depending if one or more fibres are dedicated to record the sky This has to be considered just as a rough number because the precise value will depend on the seeing and on resolving power adopted for UVES in slit mode GIRAFFE in particular in IFU mode can approach the typical resolution used with UVES and its use can be considered as a valid alternative to UVES slit when several sources are present in the field and a very large wavelength coverage is not required UVES is equipped with a Iodine cell for accurate radial velocity measurements While this system is likely more accurate than the multi fibre system of FLAMES it does not offer multiplex capabilities The UVES iodine cell cannot be used in combination with the fibres because it is located in the focal plane of Nasmyth B i e before the Fibre link to FLAMES See the UVES webpage for details http www eso org sci facilities paranal instruments uves e VIMOS at UTS has a smaller field of view than FLAMES a square of 14 x 14 arcmin utes but a higher multiplex gain up to 400 mini slits punched in to mask The major difference is the spectral format and a lower resolution R 4500 for a 0 5 arcsec wide slit VIMOS also has a I
86. le within service mode normal programmes These four settings are therefore only offered in visitor mode in Argus to allow for more time to take the attached flatfields in the daytime Even then the number of derived counts obtained at the far blue end FLAMES User Manual VLT MAN ESO 13700 2994 39 may only be around a thousand in a few hours Note that for Medusa and IFU settings the robot not screen flats are adequate in the blue so all settings are offered in service mode for these fibre systems 3 8 2 Contamination from Simultaneous Th Ar Calibrations Although GIRAFFE has very low level of scattered light the 5 simultaneous fibres in particu lar in the reddest setups may show very strong Argon lines These lines cannot be suppressed by any filter and give visible ghosts at the level of several ADUs over a large part of the CCD area It appears as a diffuse increase in the background 10 20 electrons with an increase up to 40 60 electrons numbers are indicative very close to the strong lines These ghosts may be very bad for those users interested in faint objects low S N ratio observations since they increase substantially the background Since the spectrograph is quite stable users who are observing faint objects and who are not interested in accurate radial velocity determinations should switch the simultaneous calibration OFF This can be done by filling in the appropriate field in the FLAMES observing templates For Medusa
87. les one per plate each with FLAMES User Manual VLT MAN ESO 13700 2994 22 UVES Fibre link overall view Connector 8 buttons per in the positioner sub slit on telescope side Simultaneous al Calibration Unit 1 2m hem 1 Fibre per 10 m long fibre 33 5 m long fibers 9 fibres per slit To UVES Anchored point lube of 4 mm Figure 6 Scheme showing the buttons fibre bundles and the geometry of the UVES slit eight buttons Every button hosts one fibre From the UVES simultaneous calibration box one additional 5 meter fibre reaches the UVES fibre slit Each of the two UVES fibre slits one per plate consists of 9 fibres although only eight fibres can be used simultaneously They have a core of 120 microns diameter and a cladding of 144 microns Each fibre is protected by a Polyamide jacket of 180 micron diameter The fibre to fibre separation center to center is 1 7 times the fibre core 1 7 arcseconds The UVES system works at the optimal F 3 focal ratio to minimize the focal ratio degradation FRD and therefore the transmission losses The exit is also at F 3 each fibre has an aperture of 1 arcsecond on the sky A microlens in front of the fibre converts the F 15 focal ratio of the VLT Nasmyth focus to F 3 behind the microlens the light is reflected towards the side of the button where the entrance of the fibre is located see Figure 6 The overall transmission of the UVES fibre system is giv
88. librations 5 5 1 GIRAFFE Longslit Unit One of the slits of GIRAFFE is equipped with a longslit unit which is used for engineering purposes calibrations obtained with this unit are of no interest for the observer 5 6 The UVES Calibration Unit The UVES calibration unit is a mechanical structure mounted on the Nasmyth rotator flange which in the case of UVES is kept fixed during observations It hosts continuum lamps which in combination with various filters are used for flatfield calibration and one Th Ar lamp for wavelength calibration The lamps are mounted on an integrating sphere and relay optics simulate the F 15 telescope beam The light from the lamps is fed into the instrument beam by 45 mirrors mounted on a slide FLAMES User Manual VLT MAN ESO 13700 2994 62 The flatfield spectra provide a good correction for the blaze function of the echelle They correct the pixel to pixel variation in CCD sensitivity as a function of the wavelength of the light In the red part of the spectra A gt 650 nm narrow fringes with peak to valley amplitudes up to 30 are present on the EEV CCD of the mosaic On the MIT LL CCD the fringes are less sharp and of smaller amplitude In the fibre mode of UVES long slit spectra are acquired with a slit longer than the extent of the fibre slit to ensure that even in case of small shifts between the observation and the calibrations the detector area covered by the fibres is covered by the long slit s
89. mat wavelength and order number as function of position on the detector and the expected S N for the specified target atmospheric conditions as a function of exposure time The ETC can also be used to access the efficiency curves of the various optical components as measured in the laboratory in advance of the installation While using the FLAMES ETC the user has to keep in mind two fundamental points 1 Some of the transmission factors are mean values for instance Table 4 shows how the corrector transmission varies with the distance from the field center The ETC assumes a distance of 8 arcminutes In the same way fibre to fibre transmission variations are present at the 5 10 level The ETC values are also mean values 2 With an aperture of only 1 2 and 1 0 arcseconds on the sky MEDUSA and UVES respectively the photon collecting efficiency will strongly depend on the accuracy of the astrometry The ETC is set to a default value of 0 3 arcseconds for the average object fibre displacement An option allows the user to specify the object fibre displacement to evaluate the effects of bad astrometry on the photon collecting efficiency 4 7 2 Choice of the Sample Target For the input flux distribution to the ETC four options can be selected 1 A blackbody energy distribution at a given temperature 2 a power law distribution 3 a template spec trum stellar spectra from spectral type O5 to M2 nebular spectra galaxy spectra o
90. n of the increasing fibre number in the slit FPS and increasing wavelength A are indicated Redder echelle orders are to the left FLAMES User Manual VLT MAN ESO 13700 2994 7 2 Characteristics of GIRAFFE Filters ww ol HR 01 HR 02 HR 03 HR 04 HR 05 HR 06 81 Figure 29 Transmission of the GIRAFFE High Resolution filters 01 06 Wavelength is in nm FLAMES User Manual VLT MAN ESO 13700 2994 82 HR 07 HR 08 HR 09 a HR 10 D 4 HR 11 HR 12 a a a A us se Ss Se rs Ss ms e ra a se Figure 30 Transmission of the GIRAFFE High Resolution filters 07 12 Wavelength is in nm FLAMES User Manual VLT MAN ESO 13700 2994 Figure 31 nm HR 13 HR 14 HR 15 HR 16 HR 17 HR 18 Transmission of the GIRAFFE High Resolution filters 13 18 Wavelength is in 83 FLAMES User Manual VLT MAN ESO 13700 2994 HR 19 HR 20 HR 21 HR 22 LR 01 LR 02 84 Figure 32 Transmission of the GIRAFFE High Resolution filters 19 22 and Low Resolution Filters 01 02 Wavelength is in nm FLAMES User Manual VLT MAN ESO 13700 2994 LR 03 LR 04 LR OS LR 06 LR 07 LR 08
91. ngths without reconfiguring plate 2 at the centre of which ARGUS is located These observations rely on the VLT guide star and hence do not use the FACBs These observations are performed using the FLAMES giraf_acq_argfast acquisiton template The 15 ARGUS sky single fibres are placed in a circle with radius defined by the user Use of the template saves time as swapping the plates back and forth is not necessary However swapping is avoided only if a The ARGUS sky fibres are at the same radius and or not used b The plate scale is the same for the two observations Note that the FLAMES giraf_acq_argfast template can be used with any of the following ob servation templates e FLAMES giraf_obs_argoff science target with offsets e FLAMES giraf_obs_argstd standard star observation with offsets e FLAMES giraf_obs_exp normal Giraffe science template 6 4 Evaluation of the Results Offline Data Analysis At the end of each integration the CCD frames are read out by the FIERA controller and transferred to the instrument workstation and subsequently to the archive During the readout the frames are displayed automatically on two Real Time Display RTD panels one for GIRAFFE one for UVES for first inspection using the standard RTD tools More detailed analysis of the new exposures or previous exposures has to be carried out on the astronomer s offline workstation where copies of the raw files are available within a few seconds A
92. ning of the UVES fibres for the next observation cannot be done simultaneously with the current observations but only after that the UVES shutter has been closed Repositioning of UVES fibres requires 90 seconds in total e The UVES simultaneous wavelength Th Ar lamp can accept exposure times in the range of 3 60 minutes for the 580 nm setup giving an acceptable level of exposure Shorter or longer exposure times however will result in under and over exposed Th Ar reference spectra respectively These will not be accepted in Service Mode In the 860 nm setup long exposures would provide heavily saturated Ar lines producing strong persistent remnants Neither the 860 or the 520nm setups are offered with the simultaneous calibration option e All observations must be prepared with the FPOSS preparatory tool See 4 5 and 4 5 1 and the FPOSS manual The Target Setup Files created by this tool must not be modified by the user This will cause the P2PP process to fail e The limited size of the MEDUSA and UVES fibres together with the lack of information on the object fibre displacement makes it impossible to compute the amount of flux lost therefore no absolute spectro photometry can be obtained with these fibre systems Unlike other multi object ESO instruments FLAMES does not have pre imaging capabilities to prepare target selection This implies that astrometric lists must be prepared by the observer Experience with other similar in
93. ntegral Field Unit This Unit is larger than the GIRAFFE ARGUS it may be as large as 60 x60 with a resolving power of 300 or as large as 30 30 with a resolving power of 2000 See the VIMOS webpage for details http www eso org sci facilities paranal instruments vimos e Xshooter at UT2 is designed to cover the spectral range from 300 2480 nm at medium resolution See the ESO website for the current status of this instrument http www eso org sci facilities paranal instruments xshooter 2 6 High resolution spectrographs at ESO La Silla High dispersion echelle format spectrographs available at ESO La Silla include FEROS at the MPG ESO 2 2m telescope R 48 000 and HARPS R 110 000 at the ESO 3 6m tele scope Information on these instruments can be found at http www 1s eso org lasilla sciops 2 7 FLAMES Sample Observations and Calibrations A large number of scientific observations of a variety of targets and their associated calibrations have been obtained during the FLAMES Commissioning and Science Verification runs They have been made publicly available at http www eso org sci activities vltcomm flames and FLAMES User Manual VLT MAN ESO 13700 2994 12 http www eso org sci activities vltsv flamessv FLAMES calibrations are available from the ESO archive at http archive eso org 2 8 Bibliography 1 FLAMES Templates Reference Guide VLT INS MAN ITA 13750 0009 http www eso org sci fac
94. ntial Atmospheric Effects An important problem that cannot be neglected when performing multi object spectroscopy in a large field is the differential refraction of the atmosphere This is a differential effect in the sense that the atmospheric refraction index and hence the direction of propagation of the light from a given star changes with both zenith distance and wavelength The consequences for astronomical observations are therefore two fold e An achromatic effect since the refraction index at a given wavelength changes non linearly with the zenith distance This is very important when observing large fields because stars in different position within the field can have significantly different zenith distances and therefore do not move in a coherent way across the sky making it impos sible to guide on the whole field e A chromatic effect because the refraction index changes with wavelength As a result the red and the blue part of the spectrum do not hit the fibre at the same position and therefore part of the stellar spectrum can fall outside the fibre entrance This effect is important only when observing over a wide spectral range especially in the blue it will therefore be more relevant for the GIRAFFE low resolution setups and for UVES The achromatic effect cannot be compensated since it is differential across the field and strongly dependent on the actual zenith distance It is therefore extremely important to be
95. observing p2pp It is worthwhile recalling that for FLAMES the OB preparation is very simple while most of the effort is required to provide objects fiducial stars and VLT guide stars in the same astrometric coordinate system with a relative rms accuracy better than 0 3 arcsecs A UVES fibre since April 2003 and a GIRAFFE from April 2004 data reduction pipeline runs at the observatory They enable automatic extraction and wavelength calibration of most settings in order to check of the quality of the observations resolution signal to noise ratio in the extracted spectra The science data are calibrated with calibration exposures from a calibration database which is updated when required Note that at the time of writing there is no sky subtraction available in the GIRAFFE pipeline 4 2 FLAMES Modes and basic Choices After the detailed description of the GIRAFFE and UVES spectrographs and their subsystems and functions Chapter 3 we provide an overview of the different FLAMES observing modes FLAMES is equipped with two spectrographs GIRAFFE and the RED arm of UVES UVES can be either used with all 8 fibres acquiring source or sky photons or by using seven fibres on sources and one fibre to record simultaneously a Th Ar arclamp spectrum UVES 7 1 GIRAFFE can operate in either MEDUSA IFU or ARGUS mode Simultaneous observations with UVES can be carried out with any of the GIRAFFE modes However only one GIRAFFE mode can be use
96. ode only which is 8 minutes The target acquisition includes the configuration of UVES fibres the homing of the telescope rotator to zero degrees the swapping of the plates and the acquisition of the field 7 minutes The telescope preset acquisition of the guide star and start of the active optics account for an additional 6 minutes e GIRAFFE and UVES Instrument setup 1 minute A new instrument setup takes 1 minutes for GIRAFFE and UVES FLAMES User Manual VLT MAN ESO 13700 2994 59 e GIRAFFE and UVES CCD readout 1 minute The readout time for the CCD mosaic in the UVES red arm and for GIRAFFE CCD is 1 minute each In combined mode all CCDs can be read in parallel e Plate Configuration 0 20 minutes Plate configurations take 20 minutes at most MEDUSA mode This does not translate into overheads if the running exposure on the other plate is at least 20 minutes long Plate configuration overheads are to be taken into account only when the exposure time on one plate is shorter than 20 minutes e Nighttime Screenflat Calibration 7 minutes If attached screen FF calibrations are requested at nighttime they will need on average 7 extra minutes Note that the attached FF exposure time strongly depends on the wavelength It is almost impossible to get sufficient flux with a decent exposure time for the bluest setups The in use exposure times for attached screen FF can be found in Table 14 t FF screen The exact attached
97. on its RED arm Each positioner plate has eight fibres connected to the red arm of UVES In 520 nm mode only 6 of these are FLAMES User Manual VLT MAN ESO 13700 2994 4 Spectro Mode N Objects Aperture IT H Cover UVES RED 8 with sky 1 0 47000 200 UVES7 RED 7 with sky 1 0 47000 200 1 Simul Calib GIRAF HR MEDUSA 131 with sky 1 2 190007 22 A 12 GIRAF LR MEDUSA 131 with sky 1 2 70001 A 9 5 GIRAF HR IFU 15 15 sky 2x3 300001 A 22 1 12 GIRAF LR IFU 15 15 sky 2x3 110001 A 9 5 GIRAF HR ARGUS 1 11 5x7 3 300004 1 22 A 12 or 6 6x4 2 GIRAF LR ARGUS 1 11 5x7 3 11000 A 9 5 or 6 6x4 2 Spectro Mode V S N 10 V S N 30 pix RV accuracy UVES RED 17 5 15 5 0 18 300 ms UVES7 RED 17 5 15 5 0 18 30 ms GIRAF HR MEDUSA 19 3 17 4 0 19 150 ms GIRAF LR MEDUSA 19 9 18 0 0 19 300 ms GIRAF HR IFU 17 9 15 9 0 19 150 ms GIRAF LR IFU 18 54 16 5 0 19 300 ms GIRAF HR ARGUS 17 9 15 9 0 19 150 ms GIRAF LR ARGUS 18 54 16 5 0 19 300 ms t The resolving powers R given here are only average values for details see Tables 11 and 12 which contain a description of all the GIRAFFE setups t Magnitudes for IFU and ARGUS modes are given for extended objects in surface bright ness magnitudes arcsecond 4 Radial velocity accuracy is estimated for a slowly rotating solar like star over several days S
98. oss check of the coordinates supplied by the user is performed by ESO The quality of the astrometry remains fully the observer s responsibility Common errors include using a mix of astrometric systems not correcting for stellar proper mo tions and assuming that bright stars always have accurate coordinates they don t Figure 2 shows the amount of flux lost in a MEDUSA fibre as a function of seeing and fibre to object decentering in fraction of arcseconds it is evident as bad coordinates may spoil completely the predicted performance The reader should consider the full implications of the statistical meaning of the astrometric accuracy if it is too bad some of the objects will not get light at all This factor is even more important for the UVES fibres which are 1 0 arcseconds in diameter as opposed to 1 2 arcseconds for MEDUSA fibres e Given the high number of possible configurations the spectral format is fixed for both GIRAFFE and UVES no CCD binning only one CCD readout speed no tuning of the wavelength no change in resolving power are possible e Since the day time calibration procedure is rather long up to several minutes setup especially in the bluest setups only a limited number of setups may be allowed per night both in service and visitor modes e Screen flats in the blue are very time consuming and may not provide sufficient counts to flatfield the data well Hence starting from P84 ARGUS settings HR1 HR2
99. ot to create light contamination in the spectrograph Thanks to this system while one is observing with a set of fibres on one plate any set of fibres on the other plate can be prepared by the positioner for the next observations 3 5 2 Filters and the Filter Wheel After the slit an interferometric filter selects the light according to the chosen wavelength and resolution In addition to excellent transmission and image quality these filters must fulfill very stringent requirements on the bandpass edges and blocking over the whole CCD sensitive bandpass in order to avoid pollution from adjacent spectral orders This is very critical especially in the blue where the orders are rather short in wavelength The transmission curves for all filters can be found in the Appendix Chapter 7 The thirty GIRAFFE filters are mounted on a filter wheel with four layers each with twelve filter positions A filter is selected by selecting the one of the 4 layers A D and one of the rotational filter positions 1 12 3 5 3 Dioptric Spectrograph After the light is passed through the filters it is reflected towards the collimator which works in double pass i e the light passes through it again after being dispersed The configuration angle is six degrees The main dispersers are two commercial echelles The high dispersion grating is a protected silver coated 200 x 400 mm 63 6 degree echelle R2 with a high groove density 316 lines mm which ensur
100. other red settings Fig 22 shows extracted spectra which emphasise the strength of the sky lines Before submitting a proposal PIs should consider downloading previous GIRAFFE spectra from the ESO archive to see how badly their spectra will be contaminated by sky lines We note that the current version of the GIRAFFE pipeline does not perform sky subtraction PIs should consider referring to the following articles amongst others on how to remove sky features in FLAMES data e Battaglia et al 2008 MNRAS 383 183 Contains a detailed description of how sky lines can removed from FLAMES GIRAFFE spectra e Koch et al 2007 AJ 134 566 An estimate in the final accuracy of sky subtraction of 3 per cent is given for Leo spectra e Koch et al 2006 AJ 131 895 An estimate in the final accuracy of sky subtraction of 2 per cent is given for Carina spectra FLAMES User Manual VLT MAN ESO 13700 2994 63 5 8 Special Calibrations 5 8 1 Detector Flats Detector flats from direct illumination of the CCD through the camera only are taken at regular intervals by the maintenance staff to monitor the CCD performance They and all other calibrations are available from the ESO VLT Science Archive at http archive eso org 5 8 2 Use of Telluric Standard Stars to correct for Fringing or atmospheric Lines Stars with featureless spectra typically white dwarfs or fast rotating hot stars can be used to provide a good template to correct fo
101. pectra Pixel to pixel variations can be reduced in this way 5 7 Fibre to Fibre Transmission and Sky Subtraction issues When dealing with fibre spectra proper sky subtraction is a concern For some science cases this may be the limiting factor on the quality of the reduced data In the present scheme i e without nod and shuffle technique it is critical to characterize the fibre to fibre relative transmission with excellent accuracy This task is not always trivial because fibres may develop photometric instabilities which depend on their history and on the way they are routed In FLAMES we have ensured that in normal operating conditions the fibre system is constant to better than 1 stability At this point the most critical issue is to find a way to uniformly illuminate the fibres This task is done by the positioner Other steps to obtain a correct sky subtraction involve a the knowledge of the transmission of the corrector which is given in Table 4 as a function of wavelength and position on the field of view and b a good spatial distribution of the fibres which can account for sky variations in the field of view It is also important that enough fibres are allocated to the recording of the sky Fig 21 shows an example of a GIRAFFE image of exposure of 2750 s taken with Carreras at a wavelength of L881 7 Aside from the many cosmic rays present the attention of the reader is drawn to the many sky emission lines present at this and
102. plate This number corresponds to the fibre number used e g in FPOSS All even numbers are MEDUSA fibres Col 9 17 wave Fibre Transmission values as measured in the lab Each column is a dif ferent wavelength Col 18 X x position of fibre in the reconstructed image matrix Col 19 Y y position of fibre in the reconstructed image matrix Col 20 FPD Fibre Position on the Detector For all setups except ARGUS this is the same as FPS For ARGUS it is reversed Added April 2004 Note that the ARGUS image reconstruction using the x and y columns for the table will give the image in the standard North East orientation on sky If the ARGUS position angle was O ARGPOSAN 0 N is along the X axis and E along the Y axis A non zero PA is shown in Figure 23 The position angle is counted in the standard sense i e N to E FLAMES User Manual VLT MAN ESO 13700 2994 74 ARGUS PA 45 degrees original position de E Reconstructed image ARGUS PA 45 degrees Tel moved 1 N and 1 E E Reconstructed image Figure 23 Argus reconstructed image with Argus position angle in the acquisition set to 45 degrees Top panel Original pointing Bottom panel Telescope moved by 1 0 arcseconds North and 1 0 arcseconds East i e the object moves 1 0 arcseconds South and 1 0 arcseconds West on ARGUS FLAMES User Manual VLT MAN ESO 13700 2994 75 7 Appendix 7 1 FLAMES Raw Data Spectral Format The following figures give a schematic view of
103. position of each object FLAMES User Manual VLT MAN ESO 13700 2994 92 during the exposure knowing the field coordinates and the time of the observation As shown in Figure 16 extremely rapid variations of airmass causes the position of an object close to the field edge to change by up to 2 arcsec in one hour exposure For this reason it is extremely important to carefully plan the duration of each single exposure in order to minimize the flux losses due to the fact that objects far away from the field center may move away from their fibre in the course of long exposures Observers should make sure that their observations are confined to the flat part of the curves shown in Figure 16 For instance fields at declinations below 30 can be observed continu ously for 3 4 hours before and after their culmination On the contrary fields at declination between 0 and 30 can be observed for no longer than 1 hour and as close as possible to zero hour angle In order to allow the Fibre Positioner to calculate the mean position of each object during the exposure it is necessary to provide an expected total execution time of the complete observation defined in the observing block OB 4 5 Preparing the Target Input Files All the information regarding the targets are usually defined using the Observation Support Software OSS a number of software tools intended to assist the user in this process For FLAMES the OSS consists of FPOSS
104. r a quasar spectrum 4 a single line at a wavelength width and flux level to be specified In all cases but point 4 the object magnitude in a given broad band filter has to be entered For extended sources the magnitudes are given per square arcsec In addition to the target it is necessary to enter the sky conditions phase of the moon and FWHM of seeing disc FLAMES User Manual VLT MAN ESO 13700 2994 57 4 7 3 Choice of Instrument Configuration and Spectral Format Based on the properties of the optical components of the spectrograph the format of the echelle spectrum covered by the detector depends solely on the selected central wavelength The instrument templates with standard settings can be selected from the pull down menu The corresponding spectral formats are given in Tables 11 amp 12 for Giraffe and 13 for UVES The final entry is the exposure time 4 7 4 Exposure Time and predicted Counts and S N Ratios The output of the ETC is a table listing the pixel size in wavelength the computed efficiency the total expected electrons for the object and the sky the maximum pixel intensity to monitor saturation the predicted S N ratio per extracted pixel in dispersion direction the central wavelength and the wavelength bin size 4 8 P2PP tool The Phase 2 PreParation P2PP tool allows the observer to construct OBs An online tutorial for the creation of FLAMES OBs is available at http www eso org observing p2pp FLAME
105. r fringing as an alternative to the use of internal flatfield lamps These spectra can also be used to identify and estimate the depth of atmospheric H20 and Oz absorption lines In case these are required we suggest that the users insert some of these objects among their targets FLAMES User Manual VLT MAN ESO 13700 2994 64 GAUTIER DUHR Figure 21 This image shows how especially in the red that there are many sky lines Removing them can be critical to obtaining good science output The exposure was taken using GIRAFFE at L881 7 nm for 2750 s FLAMES User Manual VLT MAN ESO 13700 2994 65 Ay a in klen WW l Wh 850 855 860 Wavelength angstroms A Juan R Lo N 850 875 3900 Wavelength angstroms Figure 22 Extracted spectra of the image in Fig 21 showing a number of bright sky emission lines FLAMES User Manual VLT MAN ESO 13700 2994 66 5 9 FLAMES Science Calibration Plan Table provides a summary of the FLAMES Science Calibration Plan as defined in the FLAMES Calibration Plan 3 We note that specphot standards are generally selected by the nighttime astronomer from the list available at http www eso org sci observing tools standards spectra FLAMES User Manual VLT MAN ESO 13700 2994 67 FLAMES UVES Science Data Calibration Plan per instrument setting plate fibre mode and central wavelength Calibration num freq purpose 1 days Robot Flatf
106. resolution spectrograph R 7500 45000 for the entire visible range 370 950nm It is equipped with two gratings high and low resolution and uses order sorting filters to select the required spectral range Each object can be only observed in one or a fraction of a single echelle order at once GIRAFFE is equipped with a 2kx4k EEV CCD 15 um pixels with a scale of 0 19 arcsec pixel The fibre system feeding GIRAFFE consists of the following components e 2 MEDUSA fibre slits one per positioner plate Up to 131 different objects including sky fibres are accessible in MEDUSA single fibre mode each with an aperture of 1 2 arcsec on the sky 5 additional fibres allow simultaneous calibration of every exposure e 2 IFU slits one per positioner plate Each deployable Integral Field Unit IFU consists FLAMES User Manual VLT MAN ESO 13700 2994 6 20 Square Microlenses Microlens Array positioned in the button Folding Prism H 6mm E Magnet 20 Fibres Figure 1 Schematic representation of a deployable Integral Field Unit IFU in its button The signal from the rectangular microlens system 0 52 arcsecond squared per microlens is brought to the Giraffe spectrograph through 20 fibres The fibres of one IFU form one subslit of the IFU slit of an array of 20 square microlenses of 0 52 arcsec side each giving a total almost rectangular aperture of 3 x 2arcsec For each plate there are 15 IFUs dedicat
107. rformance Although fainter objects to R 13 may work experience has shown that due to uncertainties in the magnitude and non ideal observing conditions e g cirrus or poor seeing the Active Optics loop may fail to close If this occurs then another guide star would need to be chosen that would likely vignet the fibres on the plate Finally it is also very important that the guide star is sufficiently isolated to avoid confusion in its choice Note that FLAMES does not have an atmospheric dispersion corrector 3 3 Fibre Positioner OzPoz The Fibre Positioner OzPoz is at the core of the FLAMES facility OzPoz is a rather large and complex system equipped with four plates two of which are currently in use see Figure 3 The Positioner can be subdivided into the following subsystems e Plates Two metallic dishes on which the magnetic buttons holding the fibres are at tached Each of the plates has a hole in the center In one plate plate 2 this hole hosts ARGUS Each plate has a curvature of 3950 mm to match the curvature of the corrector focal plane The corrector also places the telescope exit pupil at the center of curvature of the plate so fibres receive the full telescope beam regardless of their position on the plate e Retractors Mechanical systems maintaining the fibres in constant tension Each fibre is equipped with one retractor The retractors are the same for all fibres When parked the fibres are deposited
108. rresponds to a different declination and indicates the size of the relative motion between the center of the field and the object as function of hour angles The distance of 9arcmin has been chosen as the radius enclosing about one half of the field area For comparison Figure 17 shows the same effect for a star located at the edge of the field i e at 12 5 arcmin from the field center The effect is obviously non linear with the distance from the field center becoming rapidly worse towards the edges Figures 16 and 17 refer to a central wavelength of 400nm Due to the dependence of the refraction index upon wavelength the effect would be significantly smaller in the red than in the blue The central A of the observations is given as input to the acquisition template in order to allow the telescope to guide on the same wavelength However different regions of the spectrum will be displaced with respect to the central one and for large displacements they may fall outside the fibre entrance Figure 18 illustrates the displacement between a central wavelength of 400nm and four other wavelengths in a typical GIRAFFE spectrum covering a range of 60nm Two bluer wavelengths 370 and 385nm show positive displacement with respect to the central one i e the offset with respect to their theoretical position is larger while two redder wavelengths 415 and 430 nm show negative displacements The FLAMES Fibre Positioner is designed to calculate the mean
109. s UVES RED standard settings Mode Cross Below Min Central Max NMaxFib Gap Disp slit filter Wav Wav nm Wav nm RED CD 3 SHP700 414 520 621 6 1 RED CD 3 SHP700 476 580 684 8 RED CD 4 OG590 660 860 1060 8 10 Table 13 The 3 UVES red standard settings are listed below The two CCDs in the red camera are separated by approximately 0 96 mm this results in a gap in the wavelength coverage approximately centered on the central wavelength The start and end points of the spectral ranges reported in the table are generally conservative as they do not include the echelle orders which are outside the sensitive area of the CCD by more than 50 of their length No major changes to the UVES settings occurred after the upgrade of the UVES RED CCDs in July 2009 although there was a slight change in the angle between the two parts of the mosaic The standard settings for UVES are listed in Table 13 They are chosen such that together they cover the optical wavelength domain from 420 1100 nm The wavelength coverage is computed for the 4kx4k CCD mosaic of the UVES RED arm The below slit filters are used to suppress the second order of the CD gratings or undesired light from entering the spectrograph The wavelength coverage is incomplete above 993nm due to the absence of overlap between adjacent orders FLAMES User Manual VLT MAN ESO 13700 2994 48 4 4 Differe
110. s e g the time and wavelength for which the field was configure HDU3 FLAMES FIBRE Table contains the fibre description association between fibre buttons and position in the subslit and slit measured laboratory fibre transmission at different wavelengths fibre bundle For the ARGUS fibre bundle the X and Y position of the individual fibre in the reconstructed image matrix is given 6 5 1 HDU2 OzPoz_table The OzPoz binary table will be different for every frame this table associates the objects to the fibre buttons The basic information for this table is taken from the Target Setup File association object to fibre and object characteristics This information is complemented by OzPoz with all information related to the positioning of the fibres The table is structured as Col 1 Object Identification from Target Setup File column 1 Col 2 RA Right Ascension from Target Setup File column 2 Col 3 DEC Declination from Target Setup File column 3 Col 4 R Button R position on plate microns Col 5 R_Error Error in R microns Col 6 Theta Button 0 position on plate radians Col 7 Theta_Error Error in 0 microns Col 8 Type Object type MEDUSA IFU etc Col 9 Button OzPoz button number Col 10 Priority Object Priority from Target Setup File column 5 Col 11 Orient Button Orientation Col 12 In_Tol Tor F if positioned or not within tolerance 40 microns 0 08 Col 13 Magnitude Target Magnitude from
111. s object fibres red dots calibration fibres green dots sky fibres The direction of the increasing fibre number in the slit FPS and increasing wavelength are indicated as well as the retractor number for each IFU FLAMES User Manual VLT MAN ESO 13700 2994 78 A E i 219 237 253 165 183 203 a i o 113 131 D i 059 077 095 007 025 043 Figure 26 Reconstructed image of 15 IFU units produced by the pipeline FLAMES User Manual VLT MAN ESO 13700 2994 79 7 1 3 GIRAFFE ARGUS Note that SSN FPS and PSSN increase from RIGHT TO LEFT FPS in the current version of the fibre table PSSN ARGUS SSN 15 14 13 12 1110 9 8 76 5 4 3 2 1 1000 2000 x pixels Figure 27 Schematic layout of the ARGUS spectral format blue solid lines object fibres red dots calibration fibres green dots sky fibres The direction of the increasing fibre number in the slit FPS and increasing wavelength A are indicated Note that the directions of FPS SSN and PSSN are inverted w r t MEDUSA and IFU FLAMES User Manual VLT MAN ESO 13700 2994 80 7 1 4 UVES FIBRE UVES lt FPS soot rror T 3 3000 y pixel N O O O 1000 MIT REDU EEV REDL H 0 LJ F LL O 1000 2000 3000 4000 x pixel Figure 28 Schematic layout of the UVES FIBRE spectral format for one order blue solid lines object fibres red dots calibration fibre The directio
112. ses focal ratio degradation optics and coupling For wavelengths redder than 600 nm the transmission is constant Variations of a few tens percent between different fibres are measured FLAMES User Manual VLT MAN ESO 13700 2994 25 MEDUSA FIBRE LINK OVERALL VIEW Simultaneous S f calibration box Type A 5 bundles focal plate In the positioner Hytrel jacket Hytrel jacket L 7 5 m 1 Fibre per button Polyurethane Anchored point jacket 12 buttons per sub slits 13 fibres per slit In the positioner 12m 28 SS Type B 8 bundles focal plate 1 Fibre per A shrinkable tube button Polyurethane jacket L 6 m 3 mm 9 buttons per sub slits Anchored point tube o 4mm Number of buttons 132 buttons for star and sky Figure 8 Scheme showing the buttons fibre bundles and the geometry of the MEDUSA slit beam of the VLT to an F 7 focal ratio Due to the focal ration degradation FRD in the fibres the effective focal ratio at the fibre exit is F 5 The movable or deployable IFUs are a unique characteristics of GIRAFFE These devices can be placed all over the FLAMES field of view with the exception of the very center of the plate Underneath the microlenses a totally reflecting LLF1 prism sends the light to the fibres Each IFU contains twenty associated single fibres and each plate hosts fifteen IFUs In addition 15 Sky IFUs are present on
113. sitor Mode FLAMES User Manual VLT MAN ESO 13700 2994 16 3 FLAMES Characteristics and Sub Systems 3 1 Opto mechanical Layout Figure 3 is a view of two of the main components of the FLAMES facility the Fibre Positioner and GIRAFFE as seen from the telescope centerpiece on the telescope platform Figure 3 The Fibre Positioner and GIRAFFE as seen during the GIRAFFE integration on the Nasmyth platform The picture was taken from the telescope The positioner is looking towards the Nasmyth Focus where the corrector is placed and on the lower left the positioner electronics cabinet is seen GIRAFFE is opened and the optomechanical components are visible The instrument consists of five main parts The first part is the corrector which is mounted on the rotator The second part is the fibre positioner which allocates the fibres on the two plates mated to the Nasmyth adaptor rotator The positioner also hosts the calibration lamps used to obtain flat field and wavelength calibration spectra Furthermore it is equipped with a secondary astrometric and guiding system FACBs which consists of four imaging fibre bundles correcting small mismatches between the VLT and the observer coordinate system These first two components are common to all FLAMES configurations The light is collected through fibres equipped with microlenses into different fibre systems two for UVES one per plate and five for the GIRAFFE spectrograph two for MEDUSA
114. struments shows that most observation failures are due to improper target preparation Also given the relatively large field atmospheric effects e g differential refraction and its variations see 4 4 may be relevant and the reader is asked to consider them carefully when preparing the observations 2 5 FLAMES within the VLT Observatory A detailed overview of the different instruments on the VLT is given on the ESO homepage under VLT Instrumentation http www eso org sci facilities paranal instruments In the choice of the best instrument for a given observing programme the following possibilities should be considered VLT instruments that can perform spectroscopy in the UV Visual Red regions 300 1100 nm e FORS2 at UTI is an instrument operated at Cassegrain and has MOS capabilities and masks where up to 200 slitlets can be inserted The highest resolution possible is 6000 although only with certain setups See the FORS webpage for details at http www eso org sci facilities paranal instruments fors FLAMES User Manual VLT MAN ESO 13700 2994 11 e UVES at UT2 is the instrument which is closest to FLAMES in terms of spectral resolution In slit mode the resolving power of UVES can be up to 120000 The UVES red arm is also part of FLAMES but its blue arm 300 500nm is not connected to FLAMES When used in slit mode with a dichroic blue and red spectra can be recorded simultaneously This option is not availab
115. t version of the FLAMES user manual is available online as a retrievable pdf file at the FLAMES ESO home page http www eso org sci facilities paranal instruments flames doc Prior to the observing proposal application and or phase 2 announcements the User Manual is usually updated any significant changes are announced on the FLAMES web pages If you do not have access to the WWW a printed copy can be requested from ESO s Visiting Astronomers Section e mail visas eso org in Garching Germany Paper copies of a new version of the FLAMES User Manual are printed out only after a major revision of the document Chapter 2 is addressed to users who are not familiar with the FLAMES facility and who are interested in a quick overview of its capabilities as in the case of similar VLT and La Silla instruments This should enable a potential user to select the best instrument for a given observing program It also includes information on how to access FITS files of reference FLAMES spectra and a glossary of terms used in the Manual Chapter 3 provides the description of the instrument the instrument layout 3 1 its main components Corrector 3 2 Fibre Positioner 3 3 Fibre System 3 4 the properties of GIRAFFE and UVES 3 5 3 6 spectrographs including their resolving power and overall efficiency In addition it contains the requirements to be kept in mind while planning the observations or reducing the data It can be cons
116. the Fibre Positioner OSS i e a software pack age that takes an input file with the target coordinate list and allows the user to define automatically and or interactively the way the FLAMES fibre have to be allocated to the targets For this reason FPOSS is basically the core of the preparation of the FLAMES observations The subsequent step the definition of the observing sequence and exposure times with P2PP is then relatively straightforward The FPOSS software can be downloaded from the following page http www eso org observing p2pp 0SS FPOSS FPOSS tool html The FPOSS user manual can be retrieved at http www eso org sci facilities paranal instruments flames doc The target information flow starts with the creation of the target input file The latter is fed to FPOSS which then generates a Target Setup File containing the target guide star observing mode fiducial stars fibre and guide probe assignation This Target Setup File is associated to an OB via P2PP The content of the Target Setup File plus additional information is added as a FITS binary table to the final spectral images Since ESO has no means to check the correctness of the input file the astronomer must be very careful an error in such a file will propagate through the whole data flow without being detected The Target Input file is an ASCII file containing the following columns see Figure 19 for an example and the FPOSS User Manual for details
117. the target list and have already processed them with the FPOSS tool Together with the general Phase I and Phase II documentation http www eso org sci observing proposals the information contained in this chapter and in Chapter 5 provides guidelines for the Phase I and Phase II preparation process for FLAMES observations In Chapter 6 information is given for astronomers who come to Paranal to observe with FLAMES The preparation process can be summarized as follows Phase I e Scientific justification e Choice of instrument and mode Estimate of exposure time to reach the required S N ratio at the desired resolution Selection of the targets check availability of accurate coordinates Estimate of telescope and instrument overheads e Determination of scheduling constraints e g visibility time critical observations e Overview of observation plan e g target list calibration needs Phase II only for successful applicants bold represent tasks specific for FLAMES e Preparation of the target input files e Preparation of the positioner Target Setup Files with FPOSS Users should pay particular attention to the list of broken fibres see below Sec 4 6 e Identification of detailed instrument setups e Preparation of required Observation Blocks FLAMES User Manual VLT MAN ESO 13700 2994 44 e Recalculation of exposure time if new version of ETC has been released Due to its design and concept FLAMES is an idea
118. ulted by users who want to prepare an Observing Proposal Phase I but should definitely be read by those who have been granted observing time and have to prepare their observations Phase II In particular the description of the Atmospheric Effects affecting FLAMES observations and their consequences on planning and optimizing the observations is of fundamental importance In particular for MEDUSA mode the relative astrometry should be better than 0 3 arcseconds for all targets Chapter 4 presents the basic information needed to prepare an observing programme the various observing modes 4 2 the standard wavelength settings 4 3 and a description of the Exposure Time Calculator 4 7 1 This chapter explains how to prepare a target input file and how to generate a positioner allocation file It assumes that the reader is familiar with the fibre assignment software FPOSS and with the FLAMES templates The FPOSS manual and template descriptions are provided as separate documents and released before Phase II Chapter 5 deals with the calibration strategy wavelength flat fielding relative and abso lute calibrations of the data obtained in standard operation It also outlines the calibration techniques for high velocity accuracy and demanding sky subtraction Chapter 6 provides in formation for the visiting astronomers who come to the Paranal Observatory to use FLAMES A description of the raw data format is presented in 6 5 T
119. umination so this figure should not be used to compare the two plates FLAMES User Manual VLT MAN ESO 13700 2994 56 schedulable entity which means that the execution of an OB is normally not interrupted as soon as the target has been acquired An OB is executed only once when identical observation sequences are required e g repeated observations using the same instrument setting but different targets a series of OBs must be constructed The detailed definition of FLAMES Observation Blocks and Templates are given in the FLAMES Template Reference Guide 1 Usually one OB consists of two separate entities the acquisition template and the observation template s It is important to recall that although not mandatory it is much more convenient to repeat the fibre positioning when the same objects are observed through different setups and or very long integrations are required in order to minimize the effects of the atmosphere see 4 4 P2PP will read the information regarding the targets from the Target Setup File Note that only the files produced by FPOSS are accepted by P2PP 4 7 1 GIRAFFE and UVES Exposure Time Calculators The GIRAFFE and UVES FIBRE mode Exposure Time and Spectral Format Calcu lator ETC is accessible through the ESO WEB page at http www eso org observing etc The ETC models the instrument and detector in their different configurations It can be used to compute the detailed spectral for
120. ure 6 therefore spectra obtained with different plates are shifted by 40 pixels in the spectral direction on the CCD 3 43 MEDUSA Fibres Each plate also hosts 132 MEDUSA fibres Each button includes a single fibre and its con struction is similar to that of UVES In the case of MEDUSA the used focal ratio is F 5 The core of each MEDUSA fibre is 230 microns which corresponds to an aperture on the sky of 1 2 arcsecs They have a cladding of 253 microns and a protection buffer of 280 microns The MEDUSA fibres are 13 meters long and their typical overall transmission is given in Table 6 They are organized in a slit composed of several subslits The MEDUSA subslits are of two types a eight subslits hosting nine object fibres and b five subslits hosting thirteen twelve object one simultaneous calibration fibres This fibre slit follows the curvature of focal plane of GIRAFFE The center to center distance of the MEDUSA fibres is of 2 26 times the fibre core diameter this ensures a fibre to fibre contamination below 0 5 3 4 4 IFU Fibres Each Integral Field Unit IFU button is composed of twenty microlenses arranged in a rect angular shape see Figure 1 1 The microlenses are 0 52 squares They convert the F 15 Fibre 370nm 400nm 450nm 600 nm MEDUSA 0 47 0 52 0 55 0 61 ARGUS 0 52 0 58 0 62 0 70 IFUs 0 49 0 55 0 58 0 66 Table 6 GIRAFFE Fibre Transmission The values given here include all los
121. ut note that the PSF is not Gaussian FLAMES User Manual VLT MAN ESO 13700 2994 41 6000 4000 Pixel value 2000 1200 1300 1400 1500 164 Position PIXEL Figure 15 This figure shows a trace perpendicular to the dispersion of an UVES FIBRE frame containing three orders Note the flux overlap between contiguous fibres the UVES Fibre data reduction software is designed to deblend the contributions This implies that some care should be taken in not placing objects with large differences in magnitudes and possibly of very different nature e g emission and absorption line objects in adjacent positions on the detector Note that the UVES Fibre Data Reduction Software DRS has been developed to take into account and eliminate this fibre to fibre contamination 3 9 4 Spectral Gaps in the RED The CCD detector in the red arm see Section 2 4 consists of a mosaic of two chips separated by a gap of approximately 0 96 mm This results in the loss of one echelle order in the recorded spectrum around the central wavelength selected by the observer At 580 nm the gap spans around 5nm and at 860nm the gap is 10nm The dimension of the gap at any central wavelength can be predicted with high accuracy lt 0 5 nm using the instrument ETC 3 9 5 Optical Ghosts in the far red Spectra Spectra imaged on the CCD mosaic in the red arm are partly reflected back to the cross disperser grating through the camera lenses After a further
122. w with which plate your configuration will be observed it will allow you to allocate broken fibres It is up to the user to manually correct the configu ration deleting allocations and re allocating the target to another fibre by hand as described in the FPOSS manual making sure that highly important targets are not allocated to any of the broken fibres Fig 20 shows IFU sky spectra taken on plate 1 and and on plate 2 Differences in the relative responses of the IFUs are clearly present within each plate although the absolute value in this case just depended on the sky brightness 4 7 Introducing the Observation Blocks An Observation Block OB is a logical unit specifying the telescope instrument and detector parameters and the actions needed to obtain a single observation It is the smallest 2The updated list of broken fibres is available at http www eso org sci facilities paranal instruments flames visitor html Fibres and http www eso org observing p2pp FLAMES FLAMES P2PP html Fibres FLAMES User Manual VLT MAN ESO 13700 2994 55 STAT Wd Ww Cuts 1 WN Pixel Values lut 11 LU MAAA WA UAA AA AAN N Cuts 1 D Jee E rae e Value 17405 0 int Close Figure 20 Raw GIRAFFE IFU images of the solar spectrum on plate 1 top and plate 2 bottom taken in May 2009 Variations in the IFU responses on each plate are clear although the absolute level depends on the solar ill
123. wards of 420 nm it is very night time consuming For this reason starting in P84 the observatory only offers ARGUS HR1 HR2 HR3 and LR1 observations in visitor mode where more time will be available during the day on a best effort basis to obtain more screen flatfields Medusa and IFU systems remain available for all settings in service or visitor mode as they can use flatfields taken using the robot 5 4 Simultaneous Calibrations GIRAFFE is equipped with five simultaneous calibration fibres per slit Unless explicitly avoided by the user in the observing template every spectrum contains five simultaneous arc spectra evenly located along the CCD 2k width These spectra can be used to track the wavelength solution for all the fibres Tests on solar spectra during GIRAFFE integration in Garching reached high accuracy over a few days and tests on stars during commissioning showed that an accuracy of 70ms can be obtained on a time basis of a few hours for slowly rotating cool stars This was reached in 15 minute exposures for objects brighter than 14 3 magnitude in the HRO9 setup cf The ESO Messenger 110 1 More accurate and detailed long term RV studies are presently being carried out on old open clusters For FLAMES UVES the radial velocity instrumental error is about 30 50ms when using the 7 1 mode which includes a fibre dedicated to simultaneous calibration 580 nm only Astronomy amp Astrophysics 421 L13 5 5 Longslit Ca
124. with FACBs Minimum button separation 11 arcseconds button diameter 10 arcsecs Buttons and Fibre Systems FLAMES is equipped with different types of fibres for UVES and for the different modes of GIRAFFE At the output of the fibre system individual fibres are arranged in different subslit systems depending on the fibre type Each GIRAFFE mode has five fibres per slit devoted to simultaneous wavelength calibration in addition to the fibres coming from the Positioner These fibres provide five calibration spectra for each observation acquired with GIRAFFE The UVES system has a similar simultaneous calibration capability in that case one of the eight fibres is reserved for calibration In the evaluation of the instrument performance it has to be considered that among such a large number of fibres some dispersion exists in the fibre transmission Fibre transmission within the GIRAFFE F 5 and UVES F 10 apertures have been measured for every single fibre and the distribution of the transmission is given in Figure 5 for the different fibre types FLAMES User Manual VLT MAN ESO 13700 2994 21 FLAMES Fibre Transmission PJ OT Frequency bh D 40 45 50 55 60 BS ZU 75 EN Transmission Figure 5 Distribution of the transmission of the FLAMES fibres at 600 nm each fibre has been measured in laboratory 3 4 1 Magnetic Buttons The Magnetic Buttons have two purposes first they are the mechanical means which allows th
125. y The arrows e gt indicate the orientation of the subslits in the array and the direction of the increasing number of the fibre s position in the subslit PSSN as given in the static Fibre Binary Table cf Section 6 5 2 The x and y coordinate system refers to the X and Y columns of the Fibre Binary Table with which the ARGUS image array is reconstructed from the fibre s position in the ARGUS slit FPS For a ARGUS position angle of PA 0 the North East orientation on sky is indicated for the reconstructed image too This is the long axis of ARGUS See also Fig 23 FLAMES User Manual VLT MAN ESO 13700 2994 28 The different GIRAFFE sub units are described in more detail in the following sections 3 5 1 Slit Unit The slit unit contains six slots five are occupied by the GIRAFFE fibres and one is occupied by the long slit which can be illuminated by an internal calibration unit The slit unit is the most complex mechanical subsystem of GIRAFFE because it needs a very high stability and reproducibility Two movements allow to exchange the fibre slit and to set the fibres in focus In addition the slit unit is equipped with a number of back illumination LEDs These LEDs are powered and controlled by the Fibre Positioner They are used to allow the gripper camera to view the fibre output when positioning Clearly since they are lit during the GIRAFFE exposures special care was taken in keeping them very well light tight in order n
126. ystem which directs the light either from a tungsten lamp or FLAMES User Manual VLT MAN ESO 13700 2994 19 File FP TCCD RID Acquisition Offsets Offsets Template RA 0 014 Apply Continue DEC IE ABORT THETA 0 004 SEND ROTATOR OFFSET Validity E Offsetting m Loop frequency Ip images Guiding 8 Counter 0 Start guiding LAST APPLIED OFFSET Last RA 10 000 Last Dec 0 000 flopanAcquisition Qwflames FACB Offsets and FWHM Fibre 17 85 149 213 On Plate m m m m Image Ok m m m m Used RA o 0025 0 091 0 11 0 17 Dec 0 16 0 34 0 22 0 017 FWHM 0 75 0 52 0 61 0 61 FACB CCD Control ONLINE SR Object fpteca Ty tive span IT gt enn EEE ZIE 820 Y 755 5 Value 2019 Integration Time 4 4j s 17 19 29 400 8 99 41 25 50 Equinox 2000 P Use Graphic Plate At Rotator Plat el IE Min 1775 Max 5873 Bitpix 28 3 Save Image Low fi 811 High 3000 Auto Set Cut Levels FLAMES UVES Scae ZlzlSl2lt Camera fptecd Remaining Time Remaining Time H Z z 8 n ER 3537 10193 a Figure 4 Image of four fiducial stars as seen in the Technical CCD through the 4 Fiducial Acquisition Bundles FACBs Each bundle is composed of 19 ordered fibres which image the stars on 4 parts of the TCCD Each FACB has an effective diameter of 2 4 arcseconds

FLAMES User Manual

Contents

Download Pdf Manuals

Related Search

Related Contents