• Non ci sono risultati.

The end-to-end simulator for the E-ELT HIRES high resolution spectrograph

N/A
N/A
Protected

Academic year: 2021

Condividi "The end-to-end simulator for the E-ELT HIRES high resolution spectrograph"

Copied!
15
0
0

Testo completo

(1)

2017 Publication Year

2021-02-25T15:49:12Z Acceptance in OA@INAF

The end-to-end simulator for the E-ELT HIRES high resolution spectrograph Title

Genoni, Matteo; LANDONI, Marco; Riva, M.; PARIANI, Giorgio; MASON, Elena; et al. Authors 10.1117/12.2271953 DOI http://hdl.handle.net/20.500.12386/30619 Handle PROCEEDINGS OF SPIE Series 10329 Number

(2)

PROCEEDINGS OF SPIE

SPIEDigitalLibrary.org/conference-proceedings-of-spie

The end-to-end simulator for the

E-ELT HIRES high resolution

spectrograph

M. Genoni, M. Landoni, M. Riva, G. Pariani, E. Mason,

et al.

M. Genoni, M. Landoni, M. Riva, G. Pariani, E. Mason, P. Di Marcantonio,

K. Disseau, I. Di Varano, O. Gonzalez, P. Huke, H. Korhonen, Gianluca Li

Causi, "The end-to-end simulator for the E-ELT HIRES high resolution

spectrograph," Proc. SPIE 10329, Optical Measurement Systems for

Industrial Inspection X, 103290Z (26 June 2017); doi: 10.1117/12.2271953

Event: SPIE Optical Metrology, 2017, Munich, Germany

(3)

The end-to-end simulator for the E-ELT HIRES high resolution

spectrograph

M. Genoni*

ab

, M. Landoni

b

, M. Riva

b

, G. Pariani

b

, E. Mason

c

, P. Di Marcantonio

c

, K. Disseau

d

,

I. DiVarano

e

, O. Gonzalez

f

, P. Huke

d

, H. Korhonen

g

, Gianluca Li Causi

h

a

Univeristà degli Studi dell’ Insubria, Dipartimento di Scienza ed Alta Tecnologia, via Valleggio 11,

I-22100 Como, Italy;

b

Istituto Nazionale di Astrofisica – Osservatorio Astronomico di Brera-Merate, via E. Bianchi 46,

I-23807 Merate (LC), Italy

c

Istituto Nazionale di Astrofisica – Osservatorio Astronomico di Trieste, via Tiepolo 22, I-34131

Trieste

d

Georg-August-Univ. Göttingen, Institute for Astrophysics, Friedrich-Hundplatz 1, 37077 Göttingen

e

Leibniz-Institut für Astrophysik (AIP) , An der Sternwarte 16, 14482 Potsdam, Germania

Germany

f

UK Astronomy Technology Center (ATC) - Royal Observatory, Edinburgh, Blackford Hill –

Edinburgh

Niels Bohr Institute – Dark Cosmology Center Juliane Maries Vej 30, 2100 København Ø

h

Istituto Nazionale di Astrofisica – Istituto di Astrofisica e Planetologia Spaziali, via Fosso del

Cavaliere 100, Roma

ABSTRACT

We present the design, architecture and results of the End-to-End simulator model of the high resolution spectrograph HIRES for the European Extremely Large Telescope (E-ELT). This system can be used as a tool to characterize the spectrograph both by engineers and scientists. The model allows to simulate the behavior of photons starting from the scientific object (modeled bearing in mind the main science drivers) to the detector, considering also calibration light sources, and allowing to perform evaluation of the different parameters of the spectrograph design. In this paper, we will detail the architecture of the simulator and the computational model which are strongly characterized by modularity and flexibility that will be crucial in the next generation astronomical observation projects like E-ELT due to of the high complexity and long-time design and development. Finally, we present synthetic images obtained with the current version of the End-to-End simulator based on the E-ELT HIRES requirements (especially high radial velocity accuracy). Once ingested in the Data reduction Software (DRS), they will allow to verify that the instrument design can achieve the radial velocity accuracy needed by the HIRES science cases. Keywords: End-to-End model – End-to-End modular modeling – End-to-End flexible modeling – Spectrograph parametric model – High resolution spectrograph – E-ELT telescope.

INTRODUCTION: MOTIVATION AND MAIN GOALS

In the last decades astronomical observation projects have become more and more complex both in term of technologies, instrumentations, operative modes and procedures. This arises both from the infrastructure required by large aperture telescope and even more heterogeneous and complex instrumentation.

*matteo.genoni@brera.inaf.it;

g

Optical Measurement Systems for Industrial Inspection X, edited by Peter Lehmann, Wolfgang Osten, Armando Albertazzi Gonçalves Jr., Proc. of SPIE Vol. 10329, 103290Z · © 2017 SPIE

CCC code: 0277-786X/17/$18 · doi: 10.1117/12.2271953 Proc. of SPIE Vol. 10329 103290Z-1

(4)

In order to overcome and mitigate this situation it became necessary to develop End-to-End instrument models2,3, i.e.

simulators which allow physical modeling of the whole system from the light source to the reduced data.

The aim of our work is to design and develop the physical model of the different modules, which compose the entire End-to-End system, directly during the project design phase. This approach will benefit the entire project since it allows to:

• generate synthetic detector images starting from starlight or calibration light that closely represent the expected ones from the commissioned instrumentation, that will serve for detailed scientific verifications of the proposed observations;

• exploit the generated detector images for the development of the data reduction software in parallel to the instrument design, or alternatively, for test and verification purposes of existing data reduction pipelines, exploring the possibility of reusing existing pipeline code;

• keep the whole system and its different modules efficiently under control during every phase of the project (both design and assembly-integration phases) allowing to test directly the chosen design from simulated frames to data reduction and analyses.

• exploit reliable tools (the different operative modes of the simulator) at a system engineering level to evaluate the effect of the main instrument parameters directly on the final performance.

• evaluate the effects on the final performance of the different instrument architectures and technologies;

• achieve a deep understanding of the design as well as design optimization and improvements exploiting the capability of early identification of system level problems;

• gather an early performance verification purposes ensuring design requirements are met; • give reliable inputs for calibration procedures;

• improve the effectively achievable scientific goals, through the observation performance optimization during simulation, long before it is carried out in practice;

• plan efficiently the observation proposals and programs that will be run on the actual instrument;

Fundamental general aspects which characterize the simulator are its modularity and flexibility. They will be extremely important in the next generation telescopes and their related instruments like the E-ELT (European Extremely Large Telescope) because of the high complexity and long-time design and realization. Moreover these aspects will help an efficient evaluation of the effects of new technologies (e.g. instrumentations) and procedure (e .g. data reduction software). Modularity and flexibility will be of relevant importance in the design and building of an End-to-End simulator, which foresees the integration of different software packages that are aimed to allow a successful cooperation of a wide range of users: from project manager and engineers to instrument scientist and astronomers.

MAIN CAPABILITIES

In order to design and develop the End-to-End instrument simulator the following main tasks have been identified: • to evaluate the spectral energy distribution of different targets like star, QSO or galaxies (these are of course

related to the specific observation program); • to account for the effects of the atmosphere;

• to perform a simulation of the complete optical train of the instrument; in particular for all the possible operative modes (e.g. different resolutions) of the instrument and all possible operation phases;

• to take into account both optical aberrations and diffraction effects;

Proc. of SPIE Vol. 10329 103290Z-2

(5)

Nasmyth focal surface Pnmary mrror (1.47) Quaternary mYla (W) FMI mina mkror(M5) Toróny minor (M3)

Gaaty..+vananl local surface

• to take into account both mechanical and thermal effects;

• to perform high accuracy image reconstruction to be provided to data reduction pipeline; • to include a simulation of the detector properties and of the impact of cosmic rays.

END-TO-END MODEL STUDY CASE: THE E-ELT HIRES

We present in this section the specific study case we have focused on for the design and development of the simulator, which is the High Resolution Spectrograph at the European Extremely Large Telescope (this is for the moment called E-ELT HIRES in order to avoid wrong association with the HIRES spectrograph at Keck telescope). We give an overview of the E-ELT design and of the spectrograph proposed architecture concept.

E-ELT HIRES overview

The huge photon collecting power of the 39 m primary mirror diameter E-ELT (see Figure 1 and 2) coupled with a High Resolution Spectrograph (E-ELT HIRES) will allow the European high resolution community to make fundamental discoveries in a wide range of astrophysical areas, outlined by the Science Team of the E-ELT HIRES Initiative4,5,6:

• The study of Exo-planetary atmospheres and the detection of signatures of life on rocky exo-planets. • The chemical composition, atmospheres, structures and oscillations of stars.

• The spectroscopic study of the galaxies evolution as well as the three dimensional IGM reconstruction at high redshift.

• Fundamental constants (such as the fine-structure constant α and the proton-to-electron mass ratio µ) variation and the related cosmology.

The E-ELT will be the largest telescope to observe in visible and infra-red light; the baseline of the optical design (see

Fig. 1) is a five mirror solution1: aspherical (almost paraboloid) primary mirror M1, a convex secondary mirror M2 with

4 m diameter, concave tertiary mirror M3 with 3.75 m diameter, and two flat mirrors (called M4 and M5). These two latter mirrors have the purpose to feed two Nasmyth focal stations and for adaptive optics; below each Nasmyth platform a Gravity Invariant focal station, fed by a steerable and removable mirror (M6), will be located (see Fig. 2). In addition the M6 mirror and a Coudè-train relay optics will allow to feed a Coudè focal station, which will be specialized to host instruments requiring very high long term stability in terms of thermal and mechanical perturbations. The telescope structure (rendering shown in Fig. 3) will be alt-azimuth type.

Figure 1. E-ELT optical layout: Nasmyth focus (left) and Gravity invariant focus (right), taken from [1].

Proc. of SPIE Vol. 10329 103290Z-3

(6)

Figure 2. E-ELT structure, rendering, taken from [1].

The E-ELT spectrograph preliminary architecture concept, proposed by the E-ELT HIRES Initiative working group and at the moment in the phase A of the project, is highly modular. It foresees different independent fiber-fed echelle cross-dispersed spectrographs optimized for different wavelength bands of the whole spectral coverage of the instrument 370 nm to 2400 nm (see Figure 3 for details).

The different spectrometers can be divided according to their specific function into two units: the pslit unit, a re-imaging system which collects the light from the fiber optics and feeds the spectrometer unit, which has the usual purpose of separate the light into its constitutive wavelengths and then refocus them onto the detector surface. A fore-optics system in combination with a lens-let array is used to couple the telescope focal plane and the fiber fore-optics6, which

are vertically re-arranged to feed the pre-slit unit.

Proc. of SPIE Vol. 10329 103290Z-4

(7)

IRES i UHR -end ) -IFU -end )nt -end her t -ends Special R fibers (L < ? ? ?m) BB fibers (L <20 m) Calibration Unit Telecom fibers 0L <200 m) fiber to fiber interface ,._ l J Í \ Collimator and Echeo lle ( l Ó I { I °-^ k.. J tll

t

\

1 Collimator and ° Echelle J

\

Dichroics J m \ l Colllmator and Echelle Dichroics L J Collimator and Echelle (LL, isp. (Camera) K IR spectrograph 1 Figur The proposed slit arrangem • High subt • Ultra each • MO on sk In order to b capabilities a UHR modes feeding the sp the selected a is looking at allowing high effects are als In addition to the E-ELT HI particular to t at the momen re 3. E-ELT HI d architecture ents: h Resolution ( tending a size a-High Resolu h one subtendi S Medium Re ky. better accomp are foreseen fo there will be pectrograph m architecture is a slightly diff h resolving po so exploited in o the Resolvin IRES Initiativ the exoplanets nt with 10 m c IRES prelimina

also allow dif (HR) mode: R ≥ 0.87” on sk ution (UHR) ing a size ≥ 0. esolution (MR

plish the spec or both the hig e specific sub modules with s, at the mome ferent part of t ower performa n both the pre ng power and ve working gro s science) and class telescope ary architecture fferent main o R = 100000, 2 ky; mode: R = 1 75” on sky; R) mode: R ≥ cific scientific gh resolution b-modes (opti different fiber ent, the field d

the object. Mo ance keeping t

-slit and spect wavelength c oup involve th d gather spectr e like the VLT concept. observing mod 2 apertures (w 50000, 2 aper 10000, 10 fibe c goals, high mode and ult imized for sp rs systems. Th dicing (see for oreover in thi the componen trometer units coverage requi he accuracy o ra with signal T. des, dependin with one objec rtures (with o ers each looki

h radial veloc tra-high resolu pecific science he technique u r details refer s way the spe nts size within s of each modu irements, two n radial veloc -to-noise ratio g on resolving t and one sky one object and

ing at a differ city accuracy, ution mode. T e cases) whic used to feed th ence7), in whi ectrometers en n manufacturin ule. other relevan city measurem o ≥ 100 for fai g power and d y bundle of fib d one sky bun rent object sub

high through This means th ch will be im he spectrograp ich each fiber ntrance slit wi ng capabilities nt design drive ment of < 10 cm

int object that

different fiber bers) each one ndle of fibers) btending 0.86”

hput and IFU hat for HR and mplemented by ph entrance in r of the bundle dth is reduced s. Anamorphic ers derived by m/s (related in t is prohibitive -e ) ” U d y n e d c y n e

Proc. of SPIE Vol. 10329 103290Z-5

(8)

Calibratio C .. Science module

3imulato

1

r path &

pheric ule scope dule Image si ist &2n interfacE ulator order

es

Lig . Dal rimeter unit Detect modul Iht path to & Logic

pat-The required and interface the definition schematic ov Figur data orang The u In the follow interfaces. Science Obje This module science case, give the num is with the At capabilities d s. We underli n of interfaces erview). re 4. General s & logic path th ge, while in blu units which com

wing a brief de ect Module simulates the by loading a mber of photon tmospheric M

END-described abov

ned once mor s among diffe

schematic diagr hrough the diffe

ue other units mpose the Instru

efinition and e theoretical sp an input spect ns; according t Module.

TO-END M

ve are translat re that modula erent modules

ram of the End erent units are simulated by th ument Module description o pectral energy tral template, to the target o

MODEL AR

ted in differen arity and flexi and among d d-to-End simula shown. Instrum he End-to-End are grouped by of the main m y distribution which is resc object differen

RCHITEC

nt modules, ea ibility will be different modu ator modules, w ment related uni

model but ind y the grey box.

modules is giv (SED) for a s caled on the b nt noises are t Instrum

TURE

ach one with s

continuously ule units or el

with the related its of the End-to dependent from

en, focusing o

selected kind basis of the ob

taken into acc

ment modul specific units-y taken into co lements (see F interfaces. The o-End simulato E-ELT HIRES on their funct of target obje bject magnitu count. The mo e functionalities onsideration in Figure 4 for a

e light path and or are colored in S Spectrograph tionalities and ect, or specific ude in order to odule interface s n a d n h. d c o e

Proc. of SPIE Vol. 10329 103290Z-6

(9)

aw e I r

in

ir

R 68 60 60a0 WaveleyyN'AngstN BW 6020 940 6960 Ta4umOiiub U!/tlebmvb 6980 Atmospheric It generates t telluric spect observatory s This module transmission ADC residua Simulation M Figure blue th Telescope M The aim of th system correc alternative op account disto conditions aim parameter fro The second o The output in Instrument M This module PSF and inst optical fiber sources at a g Unit and Fib proceeding fo calibration so path along th different cond positions on t as the echelle c Module the sky emiss trum, see Fig site by the ado then combine profile (see a als if ADCs a Module. e 5. Example o he telluric spect odule his module is ctions, the im perative ways ortion and opt med to obtain om the telesco one, instead, re nterfaces (for b

Module has the purpo trument throu

according to given resolutio ber Link Unit for details), th ources and stra he optical train

ditions is eval the instrumen e grating blaze

ion lines spec gure 5) based option of the es the sky spec an example in are present). T

f part of simula trum. Both are

s to predict th mage at the te s. The first o tical quality a n the telescop ope are not fu elies on fully both implemen ose to genera ghput. In add the specific c on for flat-fiel t models the he Calibration ategy the pape

n taking into luated in order nt image plane e function and ctrum, at a gi d on the sky ESO tool Mo ctrum with the n Fig. 5) and The output int

ated spectrum i computed at a r he telescope p elescope focal ne is a simpl as the spatial, pe PSF at the ully defined an computation ntation) are w ate geometrica dition it mode case) and it p ld and/or wave link between n Unit produc er from Huke account disto r to compute t e. The transm d cross dispers

iven high reso y condition an olecfit8. At the e science one considering terfaces are w n output from t resolution 3 tim point spread fu l plane and th lified telescop , wavelength level of teles nd allows to p of PSF from t with Instrumen al images of t els also the lin

roduces the s elength calibr n the telescop ces the SED o

at al. in this p ortion, the eff

the whole poin missivity and r

ser diffraction

olution, and th nd standard a e moment a m and filters the the Atmosphe with Instrumen

the Atmospheri mes higher than

unction (PSF) he telescope t pe version th and time var scope focal pl perform simul the E-ELT as nt Module and the instrumen nk between t spectral energ ration. The mo e and instrum of the selecte proceeding). T fects of disper nt spread func eflectivity of n efficiency ar he atmospher atmospheric p mean airmass e resultant spe ere Dispersion nt Module, P ic Module; in r the main obser

) with and wi throughput. T hat simulate th

riation of the lane. This app lation also und illustrated in d Image Simul t focal plane, elescope and y distribution odule is organ ment (see pap ed calibration The Spectrogra rsive element ction (PSF) fo the different re modeled fo ric transmissio parameters o

value has bee ectrum with th n at given ze Polarimeter Un

red a G2V stella rving mode reso

ithout adaptiv The module ca he optical pat telescope PS proach could der simplified the paper of B lation Module to evaluate t instrument (o n of the select nized in units: per of Disseu sources (see aph Unit mod s and the opt or different wa optical compo r the instrume on profile (the f the specific en considered he atmospheric enith angle (or nit and Image

ar model and in olution. ve optics (AO) an run in two th taking into SF in differen be used when d assumptions Basden 20149 e. the instrumen optics train or ted calibration the Front End u et al. in this for details o dels the optica ical quality in avelengths and onents as wel ent throughpu e c d. c r e n ) o o nt n s. 9. nt r n d s f al n d ll ut

Proc. of SPIE Vol. 10329 103290Z-7

(10)

calculation. The Polarimeter11 Unit aim to simulate the induced telescope polarization effects and the final odd and even

flux at the polarimetry arm output. It is foreseen the simulation of the physical operative conditions in terms of ambient, mechanical and thermal effects. The module interfaces with Atmospheric Module, Telescope Module, the Image Simulation Module and Detector module.

Image Simulation Module

It simulates the image on the detector surface by generating low resolution spectra or high resolution echellograms according to the user required observing mode; synthetic frames are obtained by convolving the ideal image of a specific object or entrance slit at a specific wavelength (optical fiber at a specific wavelength in the case of fiber fed instruments) with the resulting PSF of the whole instrument. The module is split into two units: one, the First Order Unit, takes into account first order diffraction and optical blurring effects in the PSF to be used in the convolution with the ideal image; the other, called Second Order Unit, estimates second order effects on the PSF halo due to pupil obscurations and

non-uniform illumination (as also deeply described in Li Causi et al 201413). This module also combined the

efficiencies/throughput computed by other modules to simulate the number of photons hitting each pixel of the detector. Interfaces are with all other modules. Optical defocusing effects due to both flatness deviation of detector surface and beam divergence across the thickness of detector substrate (especially for thick full-depleted CCDs) are also taken into account by this module.

Detector Module

This module simulates the real final detector outputs, modeling the photon noise (due to the quantum nature of light) and all the detector physical effects: quantum efficiency, charge diffusion, charge transfer efficiency (CTE), read-out noise, conversion into ADU, detector bias, dark current, non-linearity, defects like bad pixel/column and contaminants like cosmic rays; it is interfaced with Image Simulation Module and DRS pipeline (which is not part of the End-to-End simulator).

Control program of the whole simulator architecture

The whole End-to-End simulator architecture is planned to be controlled and managed by a high level program, external

to the different modules listed above. This high level program integrates all the modules by setting the input data,

parameters and passing data between them; moreover it will also call the different modules at appropriate times, managing the data flow and it will determine the times at which a module must be run again to account for a change in its input data or parameters. At the current version most of the functions and operations that coordinate the different modules are implemented in MATLAB. A possible different implementation could be considered (e.g. C++) on the basis of the optimization of the performance in modularity and flexibility managing as well as the maximization of the End-to-End simulator usability by the wide range of users for which it is developed.

PRELIMINARY RESULTS: INSTRUMENT AND IMAGE SIMULATION MODULES

Instrument Module: Spectrograph Unit

The purpose of the Instrument Module is to simulate the physical aspects of the different optical components of the instrument, taking into account aberrations, distortion and diffraction effects as well as the physical operative conditions of the instrument in term of ambient, mechanical and thermal effects. Here we present the results concerning the spectrograph which can be modeled using two different alternative versions: the Parametric Version and Ray-Tracing Version. The former is based on a physical parametric model built with the physical equations and relations which characterize the optical elements, while the latter is a ray-tracing version built with commercial optical design software (e.g. ZEMAX). Dispersion and diffraction effects as well as optical component efficiency have been already taken into account, while ambient, mechanical and thermal effects will be introduced in the future versions. A detailed description

and verification of the paraxial parametric model can be found in a previous proceeding14. Below we list the main

parameters related to the spectrograph physical model, while in figure 6 we show a schematic representation of them.

Inputs Parameters form other modules: telescope primary mirror diameter, angular aperture diameter of the object

image on the telescope focal plane (which is actually related to the seeing condition) and pixel size.

Proc. of SPIE Vol. 10329 103290Z-8

(11)

lnput Parameters from other Modules: cope primary mirror diameter

Angular aperture diameter of the obiect image on the telescope focal plane Pixel size

Top Level Requirements & Input performance data: Resolving power

waveiengrn Dana Optical Efficiency

Spectral and spatial sampling values Inter -order separation

1

Instrument Module:

E -ELT High Resolution Spectrogr

Input variable parameters: Optical fibers number & working F -ratio Main collimator F- ratios

Echelle grating blaze angle & groove density Cross disperser working angle & line density Various possible anamorphic effects

The relevant output parameters for the instrument design: Echelle grating numbers

Spectrograph camera optics working F- ratios

Number of detectors required the collect the whole echellogram

Output for the final generation of synthetic echellogram: the coordinates of the projected resolution elements on the detector surface - echellogram format

Image Simulation Module

Input variable parameters: number of optical fibers and their working F-ratio; main collimator F-ratios (which could be

different in the x, spectral, and y, spatial, directions), echelle grating blaze angle, groove density and tile size (normally high resolution performances require a mosaic made by more than one echelle grating), cross disperser working angle and groove density, various different possible anamorphic effects generated by the spectrograph optical components.

Top level requirements and input performance data: resolving power, wavelength coverage and band division, spectral

and spatial sampling values, inter-order separation (separation of the different diffracted orders projected onto the detector surface).

Thecritical output parameters for the instrument design, mainly related to technical complexity and cost issues which

characterize the proposed architecture concept and/or involved or assessed technologies, are: echelle grating numbers to form the mosaic, spectrograph camera size and optics working F-ratios, number of detectors required the collect the whole echellogram.

Figure 6. Schematic of the Instrument Module for the E-ELT high resolution spectrograph main input and output (for both the instrument design at a system engineering level and for the final generation of synthetic echellogram) parameters.

The other relevant outputs which come out from this model are the coordinates of the projected resolution elements on the detector surface (in particular the coordinates the projected image of the central point of the spectrograph entrance slit is calculated in the paraxial parametric model); these are passed as input to the Image Simulation Module for the generation of synthetic diffracted spectra. An example of the echellogram format retrieved by the Instrument Module is shown in figure 7.

Proc. of SPIE Vol. 10329 103290Z-9

(12)

Figur the f and o Image Simul The aim is echellograms each single p (spectral reso Module (as w instrument ef image or spec imposing tha specific wave the different given. Inputs param geometry and object and sk plane, taking Input variabl case of high wavelength im of magnitude re 7. Echellogr fibers per each o

object in the Hi lation Modul to model the s according to pixel. Synthet olution elemen well as the in fficiency will ctral resolutio at the integral elength and fr object images meters form o d PSF param ky photons an into account a le parameters resolution spe mage barycen e the spectral r

ram format exam order are plotte igh Resolution m

le

e image on t the user requ tic frames are nt) with the re nformation of be used to co on element. In

l calculation r rom the sky. T s or spectral r

other modules

eters (parame nd efficiency o also the detec

s: number of p

ectrograph aim nter reconstruc resolution elem

mple from the ed, including the

mode (i.e. the p

the detector uired simulatio

e obtained by sulting PSF o

the ideal ima ompute the nu n practice this result is equa The accuracy o resolution elem s: object or w eters which d of the whole i tor quantum e pixels and pix med to reach ction accuracy ment barycent

Instrument Mo e dark fibers w plot could be rep

surface by g on mode; the o y convolving of the whole in age position o umber of pho will be done al to the num of the output f ments. As bef wavelength ima describe and m instrument ch efficiency). xel size. Top l

10 cm/s of ra y will be the p ter reconstruct

odule Paraxial which are exploi presentative of enerating low outputs forma the ideal im nstrument, wh on the detecto otons incident by integratin mber of photo format can be fore a list of t age coordinat model the poi hain (from the

level requirem adial velocity primary perfo tion accuracy Parametric Mo ited to separate f the High Resol

w resolution t is a matrix w mage of a spe hich is taken a or plane). In a t on the detec g the convolv ons received f e checked by e the main para te positions, si int spread fun e telescope thr ments and inpu precision and ormance of th

should be bet

odel computatio

the fibers dedi lution IFU mod

spectra or hi which represen ecific object o as input from t addition the i ctor surface pe ved signal in e from the spec evaluating the ameters and a ize of the ide nction), numb rough the inst ut performanc d wavelength is module; to tter than ∼ 1nm

on; note that al cated to the sky de). igh resolution nts the flux on or wavelength the Instrumen nformation o er each objec each pixel and cific object or e barycenter o schematic are al image, PSF ber of science trument image ce data: for the accuracy, the give an order m. ll y n n h nt f t d r f e F e e e e r

Proc. of SPIE Vol. 10329 103290Z-10

(13)

.035 0.03 .025 0.02 .015 0.01 .005 Distribution coi 0 0 50 VIM nor :k v 'w n Ö < Block thread

/

/Block Threa- d Thre:

lock Block Block

Pixe Block Thread Thread T imparison: convolution ] 100 linear coordinate x

i and Super- Lorentziai

onvolution ìuper-Lorentzian M =3i

150 : [urn]

Block Block Bloc

Bloc d X Thread thread thread th Pixel Fractii ad Thread

'read thread tread

thread on X We have dev computing CU integration al computation scanning of i since the sim particular, as few nm on th clear that the obtained star parameter M that represent The computa the schema of Figur Spre In particular, thread compu thread on a si Figur integ Pix el X veloped our o UDA by NVI lgorithm and, accuracy and ideal function mulations of explained in he detector an complexity in rting from the

= 30 (see refe t the PSF of th ation of each S f Figure 9 re 8. Left panel ad Function tha

for each pixe ute the value o ingle GPU ker

re 9. Schemat gral on each det

Pixe own integral c IDIA for the e , among othe cost. These p and PSF for synthetic spe the previous nd the number n time is a key e ideal illumin erence15 for de he spectrograp SRE image is

l: fiber exit illum at represent the el a Computa of surface inte rnel allowing tic of CUDA B tector pixel. el Y computation p evaluation of rs, foresees tw parameters are

the point con ctra gathered sections, each r of SRE (inc y factor in the nation profile etails, Equatio ph itself (right performed on mination profile PSF due to opt ation CUDA B egral in that po to decrease th Block and Wo procedure usin single point c wo different e the single pi nvolution eval d by HIRES h resolution e cluding sampl e design of ou e that can be on and left pa t panel of Figu

with

n a grid of 10x e assuming a Su tical aberrations Block is assig oint. This kin he time of com rker Thread ar ng an innovat convolution va variable para ixel partition luation. The a involve a hu element (SRE) ing anti-aliasi ur simulator. T safely model anel of Figure ure 8). x10 pixels sam uper-Lorentzian s of the spectro

gned and, for d of approach mputation up t

rchitecture for

tive approach alue. The proc ameters explo

for the integr adoption of pa ge amount o ) must be sim ing) is of the The image of led as a Supe 8) and a Gaus mpled in step n function. Righ graph itself. each one of h, allows to ru to 0.001 us for the computatio ed based on h cedure is base ited to direct ral calculation arallel comput f computation mulated with a order of 7 x single resolut er-Lorentzian ssian Point Sp of 0.1 pixels ht panel: Gauss the 100 subpi un in parallel 1 r each SRE. on of convoluti heavy paralle ed on standard tly control the n and the pitch ting is crucial nal power. In an accuracy o 109. It is thus tion element is function with pread Function as depicted in sian Point ixel, a worker 10.000 worker

ion and surface

l d e h l, n f s s h n n r r e

Proc. of SPIE Vol. 10329 103290Z-11

(14)

Since the pro cloud comput the echellogr allowing to fu on Amazon S simulated a G event PSFs fa With our app computation u

Figur order Righ

In this paper for the high r modularity an like E-ELT (E Moreover the different softw manager and We have des addition the p

blem is, for d

ting using the

ram with a dif urther parallel S3 storage ser G2V star spec alling on the tw proached based using a 5 g2x re 10. Syntheti r the free spectr ht inset: zoom o we have prese resolution spe nd flexibility. European Extr ey are of rele ware package engineers to i scribed the m preliminary re efinition, para e Elastic Comp fferent compu lization of the rvice. A samp ctrum with sim

wo spectrogra d on hybrid co .large instance

ic echellogram ral range increa n spectral order

CONCLUS

ented the deve ectrograph at t These aspec remely Large evant importan es and which i instrument sci main functiona esults regardin allelizable and puting 2 of A uter (namely e computation ple result that multaneous F aph apertures) ombination of es on AWS. of the E-ELT H ased by the 10% r

SIONS AN

elopment strat the E-ELT (E ts will be fun Telescope) b nce in the des is aimed to all ientist and astr alities and inte ng two module

d SREs are ind Amazon Web S “instance”, eq . Results are t we obtained w Fabry-Perot sp ) gathered by t f CUDA and HIRES Z-band %. In this case t

ND FUTURE

tegy and archi E-ELT HIRES ndamental in t ecause of the sign and build low a success ronomers. erfaces of the es have been d dependent of e Services (AW quipped with then stored in with our appr pectrum and p the E-ELT HI cloud comput

arm. The spect the dark fibers b

E DEVELO

itecture of the S). The simula the next gene high complex ding of a sim sful cooperatio e different mo described; in p each other, we WS). In details, NVDIA CUD simple binary roach is repor polarimetric o IRES Z-band s ting, we obtain

trum has been g between the tw

OPMENTS

e current versi

ator architectu eration astrono xity and long-t mulator which on of a wide r odules which particular the e exploited the , we compute DA card) runn y files automa rted in Figure observation (o spectrograph a ned the image

generated consi wo aperture are n

S

on of End-to-ure is characte omical observ time design an foresees the range of users compose the definition of t e capability o each order o ning on cloud atically upload 10, where we object odd and

arm. e in 0.5 day o

idering per each not illuminated End simulator erized by high vation projects nd realization integration o s: from projec e simulator. In the parameters f f d d e d f h d. r h s n. f t n s

Proc. of SPIE Vol. 10329 103290Z-12

(15)

and the implementation of Spectrograph Unit of the Instrument Module have been illustrated (referring to a previous paper for further details), showing an example of echellogram format simulation. For what concern the Image Simulation Module the key parameters for the pixel flux calculation have been defined and the innovative approach based on heavy parallel computing CUDA by NVIDIA and cloud computing architecture is described. In the future developments of the simulator the contribution of all the other modules will be integrated in order to run a complete end-to-end simulation.

REFERENCES

[1] European Southern Observatory, [The E-ELT construction proposal], (2011).

[2] Goodwin, M., Smedley, S., Barnes, S., Farrel, T., Barden, S., “Data simulator for the HERMES instrument”, “Proc. SPIE”, 7735, (2010).

[3] Jarno, A., Bacon, R., Ferruit, P., and Pécontal-Rousset, A., “Numerical simulation of the VLT/MUSE instrument” Proc. SPIE 7017, (2008).

[4] Maiolino, R., Haehnelt, M., Murphy, M., et al., “A Community Science Case for E-ELT HIRES”, arXiv:1310.3163, (2013).

[5] Origlia, L., Oliva, E., Maiolino, R., et al “SIMPLE: a high-resolution near-infrared spectrometer for the E-ELT”, “Proc. SPIE”, 7735, (2010).

[6] Zerbi, F. M., Bouchy, F., Fynbo, J., et al. “HIRES: The High Resolution Spectrograph for E-ELT” “Proc. SPIE” 9147, 914723-1, (2014).

[7] Parry, I., Bunker, A., Dean, A., et al. “CIRPASS: description, performance and astronomical results”, “Proc. SPIE”, 5492, 1135, (2004).

[8] Smette, A., Sana, H., Noll, S., et al. 2015, Astronomy & Astrophysics, 576, A77 (2015).

[9] Basden, A., Morris, S., Morris, T. and Myers, R. "AO modelling for wide-field E-ELT instrumentation using Monte-Carlo simulation" “Proc. SPIE”, 9148, (2014).

[10] Disseu, K, Huke, P., et al, “Fiber emulator for the performance simulation of the fiber-slit spectrograph HIRES”, this proceeding 10330-8 (2017).

[11] Di Varano, I., Strassmeier, K. G, et al. “Optical and mechanical architecture for the E-ELT HIRES polarimeter”, this proceeding 10330-12 (2017).

[12] Huke, P., Origlia, L., Riva, M., et al., “Phase A: Calibration Concepts for HIRES”, this proceeding 10329-91 (2017).

[13] Li Causi, G., et al, “PSF modeling by spikes simulations and wings measurements for the MOONS multi fiber spectrograph” Proc. of SPIE”, 9908, (2014)

[14] Genoni, M., Riva, M., et al., “Optical parametric evaluation model for a broadband high resolution spectrograph at E-ELT (E-ELT HIRES)” “Proc. SPIE” 9911, (2016).

[15] Shealy, D. L. and Ho nagle J. A., “Beam shaping profiles and propagation” “SPIE Conf. Laser Beam Shaping VI”, 5876, (2005).

[16] Schmalzl, E., Meisner, J., Venema, L., et al., “An end-to-end instrument model for the proposed E-ELT instrument METIS” Proc. SPIE” 8449, 84491P-1 (2012).

Proc. of SPIE Vol. 10329 103290Z-13

Riferimenti

Documenti correlati

The actual status of the work is focused on the experimental characterization of the energy selector system with laser-driven proton beams in order to study how the

To this purpose, we have adopted the Galerkin Finite Element Method for the numerical simulation of the Drift-Diffusion model in the D case, and we have proposed two novel methods

Another study was per- formed on micro milling by using a new approach for predicting cutting forces with an orthogo- nal cutting finite element model by considering the

By combining point cloud data sets of the same object generated with both technologies, or with the same technology but registered in different moment and/or

Al contrario, alcuni e pochi interessanti personaggi di Ansar al Islam potrebbero essere nell’area Nord: se questo fosse l’obiettivo non esplicito della richiesta, allora

Si trattava della prima rilevante innovazione nelle politiche socio-assistenziali, che di fatto introduceva interventi ispirati al mo- dello delle politiche attive del lavoro,

L’orientamento generale è comunque restrittivo: l’attinenza di una norma regionale alla materia del commercio non basta ad escluderne l’illegittimità quando introduca “nuovi

Da esse, viene dedotto che tra Draconte e Solone gli efeti giudicassero, oltre che sul φόνος μὴ ἐκ προνοίας e sul φόνος ἀκούσιος (cui è riferita,