The Swarm Vector Field Magnetometer (VFM): instrument commissioning & performance assessment José M. G. Merayo

Transcription

1 instrument commissioning & performance assessment José M. G. Merayo DTU Space, Technical University of Denmark Division Measurement & Instrumentation Systems

2 overview Fluxgate principle Amorphous magnetic material Very low noise <10 pt RMS Extremely high stability Coils create an homogeneous vector field inside the sphere Isotropic and extremely stable mechanical support for the coils The coils define uniquely the magnetic axes of the sensor Electronics Unit - EU 100x100x60 mm 3, 750 g Sensor Unit - SU Ø 82 mm, 280 g 2 DTU Space, Technical University of Denmark

3 operation R f Exc. Drive. f Amp. 2 f The voltage in the pick-up coil is e(t) = d /dt, i.e: e(t) = n A B dµ r (t) /dt By detecting the 2 nd harmonic in the synchronous demodulator, we extract information about the (DC) magnetic field B. In the feedback configuration, a (DC) current is applied to the detector coil via the resistor R f by an integrator so the sensor is in null field. 3 DTU Space, Technical University of Denmark

4 configuration The VFM sensor is mounted on the very stable Optical Bench together with the Star Trackers The VFM electronics unit is mounted on the body of the S/C The Optical Bench is located at a distance (~4m) from the body of the S/C to reduce the magnetic disturbances (remanent, induced, EMC,.) The Optical Bench is thermally isolated in order to minimize thermal gradients Z X Y XYZ is the sensor coordinate system 4 DTU Space, Technical University of Denmark

5 physical components The VFM magnetometer measurements E at 50 Hz are converted into the physical units by a linear transformation B( t) R P S E( t) b b: offset vector S: sensitivity diagonal matrix P: non-orthogonality skew matrix Z X Y R: rotation unitary matrix E(t): raw measurements in the sensor XYZ frame B(t): magnetic field components in the NEC frame The magnetic field vector data product is derived from the VFM GPS: time reference STR: inertial reference ASM: nt reference 5 DTU Space, Technical University of Denmark

6 B S [ T] B Z [ T] B Y [ T] B X [ T] The Swarm Vector Field Magnetometer (VFM): Availability and sample rate Since boom deployment, the VFM has been on constantly on all three satellites. 50 Swarm A, Boom deployment t [s] Fixed sample rate is 50Hz Down sampling in L1b processor No TM(5,4) => 100% validity All rates observed until now have been correctly tracked 6 DTU Space, Technical University of Denmark

7 B [nt] B/ t [nt/20ms] The Swarm Vector Field Magnetometer (VFM): Timing The VFM response features a linear phase, i.e. constant time delay. Variation of B and its constant trend, gives <0.7ms stability µ= 1.27nT/20ms t= 31pT RMS /µ=0.7ms 1.6 The VFM is internally time stamping its measurements and it tracks the system PPS VFM timing absolute accuracy measured pre-flight to 0,2ms sample # VFM timing accuracy in-flight is verified to 0,7ms level except from a potential bias relative to the onboard PPS The mean drift of the internal time is -60µs/PPS, with a variation of less than 2µs/week 7 DTU Space, Technical University of Denmark

8 Thermal Stability The VFM experiences very small temperature orbital variations as observed in-flight. Pre-flight the VFM has been characterized (over a very large temperature span); sensitivity 30ppm/K for the sensor 5ppm/K for the electronics Very low offset dependency Zero angular dependency The VFM experiences long term temperature variations: - Sensor: [-30ºC,-10ºC] =20ºC - Electronics: [+15ºC,+40ºC] =25ºC 8 DTU Space, Technical University of Denmark

9 Model parameters Even without corrections of the thermal susceptibility, the modelled contributions of the sensitivity and offset are very small. Contributions (modelled) of Sensitivity & Offset (in-flight) Temperature effect Span S [nt] O [nt] Sensor Orbital variation 0.05 C 0,1 0,025 Sensor Temperature delay 10 min. 0,03 0,0025 Sensor Daily variation (*) 2 C 3 (*) 1 (*) Sensor Monthly variation (*) 10 C 15 (*) 0,5 (*) Electronics Orbital variation 1.5 C 0,4 0,1 Electronics daily variation (*) 3 C 0.8 (*) 0,2 (*) Electronics monthly variation (*) 10 C 2,3 (*) 0,6 (*) (*) verifiable by in-flight scalar calibration 9 DTU Space, Technical University of Denmark

10 Level 1b parameters On ground the VFM has been calibrated and characterized - Absolutely wrt absolute scalar magnetometer (Proton magnetometer) 0.15nT RMS - Thermal drift in ±50ºC 0.20nT RMS - Time synchronization 0.2 ms In-flight, the parameters of the VFM have been estimated by comparing to the ASM magnetometer - Offsets: < 1nT - Scale factors: < 20 ppm/year - Non-orthogonalities: < 2 However this may be contaminated due to the residuals ASM/VFM - Estimated long term drift ~1nT/year 10 DTU Space, Technical University of Denmark

11 Noise Field variation as measured by VFM The VFM measurement noise has been established prior launch to 10 pt/ 1Hz The varying field when orbiting Earth makes direct noise performance impossible. However, by plotting the measurement noise at orbit positions known to have little natural variation, the instrument noise is verified in full accordance with the preflight measurements Field fluctuations during a one minute quiet period The RMS on the eclipse side (over one minute ) is < 31 pt RMS for all axis 11 DTU Space, Technical University of Denmark

12 Noise (2) The VFM measurement noise observes ~0.32Hz and harmonics 1Hz and harmonics VFM raw data: includes MTQ, S/C,. Power spectrum density of the instrument noise is in full accordance with the pre-flight measurements: 11 pt/ 1Hz Evolution of PSD 12 DTU Space, Technical University of Denmark

13 B [nt] B [ T] The Swarm Vector Field Magnetometer (VFM): Absolute accuracy Although the absolute accuracy of the VFM is a system task requiring a fully calibrated processing chain, an indication of the true performance may be indicated simply be subtracting the field measured by one Swarm satellite from that of another Magnetic model based on VFM measurements Model based on VFM measurements (pre-flight cal) Doy= Bs sph Bs VFM t [hour] Obviously this will require that the field gradient between the two spacecraft is subtracted, and will thus work best when they are close together. A preliminary comparison to an magnetic model shows the overall absolute accuracy = 10nT RMS ( <50 ) [º] 13 DTU Space, Technical University of Denmark

14 status Phases after launch LEOP (~2days) Commissioning (~4 months) 1 st operation (1/2 year) Full operation (mission lifetime) Function VFM Data Range VFM Data Rate VFM Data Availability VFM Data Continuity VFM Time Integrity VFM Compression VFM Housekeeping Performance VFM Noise VFM science temperatures VFM parameter (b, s, v) VFM parameter (drift) :not planned : reduced or not performed : performed : planned 14 DTU Space, Technical University of Denmark

15 Conclusion Nominal operations and full performance verified for all Swarm VFMs Availability: 100 % (On since boom deployment) Noise verified to 10 pt/ 1Hz ( < 35 pt RMS ) for all axes ICDB (based on VFM calibration 2009): Pre-flight parameters Small adjustment of scale factor Timing consistent with pre-flight (0.2 ms absolute) System level: ASM/VFM residuals, limits the vector data product at the moment 15 DTU Space, Technical University of Denmark

16 Residual [nt] ASM/VFM residuals On-ground, the comparison of the VFM scalar values computed for all directions wrt. Absolute Scalar Magnetometer in constant background field of 50 µt gave: 0.15 nt RMS (~1ppm) But In-flight, the same comparison gives few nt On the night side, it is possible to reduce the residuals On the day side, they are modulated with LT, and maximum at LT= # 16 DTU Space, Technical University of Denmark Swarm A: 1.2 nt RMS Swarm B: 0.9 nt RMS Swarm C: 0.8 nt RMS

17 ASM/VFM residuals (2) It appears that there is an external disturbance to the VFM, which causes the effect on the day side Thermo-electrical currents induced in the vicinity of the Optical Bench bracket under the VFM sensor Eclipse Eclipse 17 DTU Space, Technical University of Denmark

18 ASM/VFM residuals (3) A model that takes into account the geometry of the Optical bench wrt. Sun reduces the residuals by a factor of 3 However, the maneuvers are not well reduced Swarm A Swarm B - Swarm A: 0.5 nt RMS - Swarm B: 0.4 nt RMS Swarm C - Swarm C: 0.4 nt RMS 18 DTU Space, Technical University of Denmark