HYPER Industrial Feasibility Study Final Presentation Precision Star Tracker Activity 3, WP 3100

Transcription

1 HYPER Industrial Feasibility Study Final Presentation Precision Star Tracker Activity 3, WP 3100 ESTEC, Noordwijk The Netherlands 6 -th March 2003

2 Agenda Introduction 1 PST Requirements 2 PST CCD Characteristics 3 PST System Trade-off 4 PST Baseline Configuration 5 PST Optics 6 PST internal baffle 7 PST Accuracy 8 Guide Star Catalogue 9 Conclusions 2 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

3 Introduction 1/2 PST purpose To allow the measurement of the Lense-Thirring effect This measurement is performed as relative measurement between the Precision Star Tracker (PST) giving angles between a guide star, fixed in inertial space and an atomic gyroscope direction, which has an extremely high short time sensitivity for rotation rates (angular rates). The PST is directed to far-distant guide stars, which are not affected by the Lense-Thirring effect. They represent a reference for the measurement and for the motions of the satellite and its control. The second measurement is performed with an Atomic Sagnac Unit (ASU), which measures the rotations of freely falling atoms relative to a series of laser beams, whose orientation is rigidly linked to the PST boresight 3 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

4 Introduction 2/2 Today GA has the capability to provide Star Sensors for a wide variety of mission requirements and applications, ranging from high accuracy pointing of scientific instruments and platform, to medium FOV sensors with AAD capability The required accuracy of the HYPER PST is about 1000 times more severe than most accurate GA star trackers (ISO, XMM). This makes this study very challenging 4 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

5 PST Requirements 1/4 The PST requirements, level 1 from HYP-2-05, are the following: Req.# Requirement Value (3) R1-PST-01 PST internal errors in the frequency range between 3.5*10-5 Hz and 5 Hz. R1-PST-02 External measurement errors (star, aberration, etc) in the frequency range between 3.5*10-5 Hz and 5 Hz. R1-PST-03 PST internal errors in the frequency range below 3.5*10-5 Hz R1-PST-04 External measurement errors (star, aberration, etc) in the frequency range below 3.5*10-5 Hz. R1-PST-05 Timing/Jitter < 1.2 * 10-8 rad (< 2.48 * 10-3 arcsec) < rad (<0.25 * 10-3 arcsec) < 1.2 * 10-9 rad (< 0.25 * 10-3 arcsec) < 1.2 * 10-9 rad (< 0.25 * 10-3 arcsec) 1 ms 5 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

6 PST Requirements 2/4 Level 2 from HYP-2-05 R1-PST-01 R2-PST Optical Distortion Residual Error < rad (< 0.02 * 10-3 arcsec) R2-PST Focal Length Variation with Temperature < rad (< 0.02 * 10-3 arcsec) R2-PST Focal Length Variation with Star Colour < rad (< 0.02 * 10-3 arcsec) R2-PST Photo Response Non-Uniformity Effect on Star Signal and Straylight < 4.1 * 10-9 rad (< 0.85 * 10-3 arcsec) R2-PST Dark Current Non-Uniformity < 2.9 * 10-9 rad (< 0.6 * 10-3 arcsec) R2-PST Centroiding Algorithm Error < 8.2 * 10-9 rad (< 1.7 * 10-3 arcsec) R2-PST Arithmetic Round-Off < 1.4 * 10-9 rad (< 0.29 * 10-3 arcsec) R2-PST Noise Equivalent Angle < 6.8 * 10-9 rad (<1.4 * 10-3 arcsec) 6 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

7 PST Requirements 3/4 Level 2 from HYP-2-05 R1-PST-02 R2-PST Relativistic Aberration < 1.2 * 10-9 rad (< 0.25 * 10-3 arcsec) R1-PST-03 No PST internal low frequency errors have been identified R1-PST-04 R1-PST-05 R2-PST Star Proper Motion < 10-5 rad (< 2.06 arcsec) R2-PST Star Parallax Error < 10-5 rad (< 2.06 arcsec) R2-PST Star Catalogue Error < 10-5 rad (< 2.06 arcsec) No 2nd level errors have been identified 7 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

8 PST Requirements 4/4 Other Requirements PST optics to fit within the following dimension: 387x387x700 mm (boresight) PST Update rate 10 Hz PST Optical entrance 190 mm PST Guide star catalogue To have at least 1 guide star always available 8 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

9 PST CCD Characteristics 1/1 To obtain typical value the following characteristics of the ATMEL TH7890 (used by GA ASTR) have been taken into account. Its main characteristics are: Full Well Capacity dark current 15 pa/cm 2 Quantum Efficiency 2*10 5 electrons (17 micron pixel size) See figure 9 6 -th March 2003, ESTEC, Noordwijk, The Netherlands

10 PST System Trade-off 1/3 To identify the baseline PST configuration the following guidelines have been followed: IFOV REDUCTION. In order to reduce the contribution of all errors that can be characterised in terms of fraction of pixels such as Centroid and NEA: this can be obtained by a longer focal length. INCREASE OF CCD FWC (Pixel size). In order to avoid CCD saturation CONSIDER LARGE TRACKING MATRIXES. In order to match large PSF produced by high F number (considered odd rows/columns matrixes from 3x3 up to 25x25 pixels) and reduce Centroid error in terms of fraction of pixel Use of GA simulator to evaluate PST performance, especially in terms of Centroid error and NEA th March 2003, ESTEC, Noordwijk, The Netherlands

11 PST System Trade-off 2/3 y' Used Centroid algorithm: C N w i1 N i1 ' i * Col Col i i z' C N w i1 N i1 ' i * Row Row i i w ' i i 1,2,...N i 9 *(1 m) w *(1 m) i N value is tied to the star spot size Increasing N means to decrease the centroid error the ratio pixel size / spot size is reduced: a smaller sampling period is obtained and then the rounding effect introduced by the physical pixel dimension is reduced Increasing N gives rise to greater sensitivity to CCD non-uniformities and noise more pixels and then more error contributions (1 for each pixel) new outer pixels having a higher weight in the barycentre computation d min 2.44* #* the centroid computation cannot be considered an average computation because the denominator of is not proportional to the total number of pixels but to the fraction of the star spot energy collected by the tracking matrix f th March 2003, ESTEC, Noordwijk, The Netherlands

12 PST System Trade-off 3/3 To identify the best N value the following simulation steps have been performed: Increase focal length to reduce the IFOV Check if the pixel size is able to contain all star signal, otherwise increase it Consider N=N 0 = min and evaluate performance in terms of Centroid error and NEA Find optimum N value: increase N value and evaluate if best performance in terms of Centroid error and of NEA degradation have been obtained, then decide if to continue increasing N or not Check if Centroid error and NEA are within the requirement Repeat all from the beginning d 2.44* #* stop when increasing focal length does not produce significant improvement in performance (too low signal to noise ratio) f th March 2003, ESTEC, Noordwijk, The Netherlands

13 PST Baseline Configuration 1/1 The identified baseline PST configuration is the following: Optical configuration: Ritchey-Chretien telescope Focal length 36 m F number 190 FOV ±25 arcsec CCD number of pixels 1024x1024 CCD pixel size 13 micron IFOV arcsec Integration time 100 ms, jitter < 1 ms magnitude range 2 V 4 Centroid algorithm Based on a 17x17 pixels tracking window th March 2003, ESTEC, Noordwijk, The Netherlands

14 PST Optics 1/5 The principal aim of the optics study has been to find a configuration with the smallest number of elements, reducing as small as possible the criticality of position errors of the optical elements PST OPTICS LAYOUT Compact 4 mirrors (2 parabolas + 2 flat) configuration Flat and parallel plate as closure window, secondary mirror holder and support of flat mirror M4 coating Minimised secondary mirror magnification (18 x) 36 m Effective Focal Length enough focal plane relief to accommodate the detector half cone 25 arc seconds FOV the FOV is limited to avoid mechanical overlap of mirrors M2/M4 and to avoid interference of the output beam with M1/M th March 2003, ESTEC, Noordwijk, The Netherlands

15 PST Optics 2/ th March 2003, ESTEC, Noordwijk, The Netherlands

16 PST Optics 3/ th March 2003, ESTEC, Noordwijk, The Netherlands

17 PST Optics 4/ th March 2003, ESTEC, Noordwijk, The Netherlands

18 PST Optics 5/ th March 2003, ESTEC, Noordwijk, The Netherlands

19 PST Internal Baffle 1/ th March 2003, ESTEC, Noordwijk, The Netherlands

20 PST Internal Baffle 2/2 Irradiance on focal plane, as computed by ASAP 7.1 assuming two diffusion models (far field point source at the extreme FOV) Chemglaze Z302 paint and Lambertian 5% reflectance Results No ghosts from mechanics Flat low level straylight th March 2003, ESTEC, Noordwijk, The Netherlands

21 PST Accuracy 1/4 Requirement number Requirement 3, rad (3, arcsec) R2-PST-01-01: optical distortion (0.02*10-3 ) R2-PST-01-02: focal length variation with temperature (0.02*10-3 ) R2-PST-01-03: focal length variation with colour (0.02*10-3 ) R2-PST-01-04: PRNU on star and on straylight 4.1*10-9 (0.85*10-3 ) R2-PST-01-06: CCD DSNU 2.9*10-9 (0.6*10-3 ) R2-PST-01-07: Centroid error 8.2*10-9 (1.7*10-3 ) R2-PST-02-01: relativistic aberration 1.2*10-9 (0.25*10-3 ) R2-PST-01-08: Arithmetic round off 1.4*10-9 (0.3*10-3 ) R2-PST-01-09: NEA 6.8*10-9 (1.4*10-3 ) R2-PST-04-01: Star proper motion 10-5 (2) R2-PST-04-02: Star parallax error 10-5 (2) R2-PST-04-03: Star Catalogue error 10-5 Estimated value 3, rad (3, arcsec) < (<0.02*10-3 ) < (<0.02*10-3 ) < (<0.02*10-3 ) 3.4*10-9 (0.74*10-3 ) 2.9*10-10 (0.06*10-3 ) 8.2*10-9 (1.7*10-3 ) < 1.2*10-9 (< 0.25*10-3 ) < (< 0.02*10-3 ) 6.8*10-9 (1.4*10-3 ) < 10-5 (< 2) < 10-5 (< 2) <10-5 (<2) (*) It depends on how the star moves within the pixel (depends on S/C attitude control system) (**) It depends on how the star moves within the CCD (depends on S/C attitude control system) (2) Frequency contrib. (**) (**) + 1 orbit (**) (*) (*) (*) 1 orbit 10 Hz 10 Hz 1 year 1 year 1 year th March 2003, ESTEC, Noordwijk, The Netherlands

22 PST Accuracy 2/4 Centroid error curves for baseline configuration th March 2003, ESTEC, Noordwijk, The Netherlands

23 PST Accuracy 3/4 NEA curves for baseline configuration th March 2003, ESTEC, Noordwijk, The Netherlands

24 PST Accuracy 4/4 Pixel edge effects induced by CCD non-uniformities Fig. 1 PRNU = 0% and DSNU =0 % Fig. 2 PRNU = 1% and DSNU =10 % Fig.. 1 Fig th March 2003, ESTEC, Noordwijk, The Netherlands

25 Guide Star Catalogue 1/8 The guide stars selection has been performed in accordance with the following criteria: Declination: to 10.5 degrees. In fact the guide star will be in a direction close to the normal to the orbit plane. The maximum angular distance from the normal is represented by the directions forming an angle of 10 deg. to the Earth limb: ±(30-10) deg. for 1000 Km orbit altitude, inclination 99.5 deg. Anti-Sun direction deg. EARTH LIMB GUIDE STARS Orbit Plane (inclination 99.5 deg.) deg. Anti-Sun direction File guidestar.ppt th March 2003, ESTEC, Noordwijk, The Netherlands

26 Guide Star Catalogue 2/8 Selection criteria (continued) Right ascension: the maximum allowable angular separation between two consecutive guide stars is 30 deg Brightest star: V +2, to avoid CCD saturation Faintest star: V such that the required sky coverage is guaranteed. Be non-variable, non-binary and no double Have an absolute proper motion known (3) better than 8 arcsec/year (from R2-PST-04-01) Have a Right ascension known (Star catalogue error, 3 ) better than 10-5 rad (i.e. 2 arcsec), (from R2-PST-04-02) Have a Declination known (Star catalogue error, 3 ) better than 10-5 rad (i.e. 2 arcsec), (from R2-PST-04-03) Have a parallax error known (3 ) better than 10-5 rad (i.e. 2 arcsec), (from R2-PST-04-02) th March 2003, ESTEC, Noordwijk, The Netherlands

27 Guide Star Catalogue 3/8 Hipparcos catalogue To perform the catalogue creation, the following fields of the Hipparcos catalogue have been taken into account: H01: Hipparcos Catalogue (HIP) identifier H02: Proximity flag H05: V magnitude H06: Coarse variability flag H08: Right ascension (), degrees (ICRS, Epoch=J ) H09: Declination (), degrees (ICRS, Epoch=J ) H10: Reference flag for astrometric parameters of double and multiple systems H11: Trigonometric parallax, [milliarcsec] H12: Right ascension proper motion * =.* cos(), ICRS [milliarcsec/year] H13: Declination proper motion, ICRS [milliarcsec/year] th March 2003, ESTEC, Noordwijk, The Netherlands

28 Guide Star Catalogue 4/8 Hipparcos catalogue (continued) H14: Standard error in Right ascension, [milliarcsec] H15: Standard error in Declination,, [milliarcsec] H16: Standard error of the trigonometric Parallax, [milliarcsec] H17: Standard error in Right ascension proper motion, * = cos(), [milliarcsec/year] H18: Standard error in Declination proper motion,, [milliarcsec/year] H37: Colour index, B-V H76: Spectral type th March 2003, ESTEC, Noordwijk, The Netherlands

29 Guide Star Catalogue 5/8 Final results Considering 2 V 4 a guide star catalogue has been generated: Total number of stars: <= Hip. entry no. <= <= HIP identifier <= <= V <= <= m_pst <= <= R.A. [deg.] <= <= DEC. [deg.] <= <= B-V <= <= Teff [K] <= 9733 Maximum (R.A.) = [deg.] (Req. <= 30 deg.) th March 2003, ESTEC, Noordwijk, The Netherlands

30 Guide Star Catalogue 6/8 Final results (continued) entry n. HIP Ra [deg.] Dec [deg.] V m PST B-V spectr. class K2III K0III G8V K2III K1III-IV G8III K0IV F6V th March 2003, ESTEC, Noordwijk, The Netherlands

31 Guide Star Catalogue 7/8 Final results (continued) Guide star co-ordinates th March 2003, ESTEC, Noordwijk, The Netherlands

32 Guide Star Catalogue 8/8 Final results (continued) Guide star catalogue histograms Number of catalogue stars vs. Teff and instrumental magnitude th March 2003, ESTEC, Noordwijk, The Netherlands

33 PST Conclusions 1/1 This study has shown the feasibility of HYPER PST within the requirements: The CCD selection is not critical The Optics design is effective, reliable and simple The baffle system shows very good straylight performance The star centre determination algorithm (centroid) is derived from well known GA algorithms The sky coverage is guaranteed in each period of the year th March 2003, ESTEC, Noordwijk, The Netherlands