Disentangling the magnetic force noise contribution in LISA Pathfinder M. Nofrarias Institut de Ciències de l’Espai (IEEC-CSIC) Motivation X LISA Symposium - M Nofrarias Gainesville - 20/05/14 2 Magnetic Diagnostics in LPF • Magnetic diagnostics in LPF: - 4 Fluxgate magnetometers 2 Coils X LISA Symposium - M Nofrarias Gainesville - 20/05/14 3 Magnetic Diagnostics in LPF • Magnetic diagnostics in LPF: - 4 Fluxgate magnetometers 2 Coils X LISA Symposium - M Nofrarias (1) Estimation test mass magnetic parameters Gainesville - 20/05/14 - Injected signals - Estimation force noise contribution 3 Magnetic Diagnostics in LPF • Magnetic diagnostics in LPF: - 4 Fluxgate magnetometers 2 Coils (1) Estimation test mass magnetic parameters - Injected signals - Estimation force noise contribution (2) Magnetic field interpolation - Characterisation of magnetic env. - Contribution of mag. sources X LISA Symposium - M Nofrarias Gainesville - 20/05/14 3 Magnetic analysis #1: coils experiment • Coils attached to ISH to apply a force/torque on the TM • The response must allow a precise characterization of the the moment and susceptibility of the TMs X LISA Symposium - M Nofrarias Gainesville - 20/05/14 4 Magnetic analysis #1: coils experiment • Coils attached to ISH to apply a force/torque on the TM • The response must allow a precise characterization of the the moment and susceptibility of the TMs Measured X LISA Symposium - M Nofrarias Gainesville - 20/05/14 4 Magnetic analysis #1: coils experiment • Coils attached to ISH to apply a force/torque on the TM • The response must allow a precise characterization of the the moment and susceptibility of the TMs Measured Unknown X LISA Symposium - M Nofrarias Gainesville - 20/05/14 4 Magnetic analysis #1: coils experiment • Coils attached to ISH to apply a force/torque on the TM • The response must allow a precise characterization of the the moment and susceptibility of the TMs Measured Unknown X LISA Symposium - M Nofrarias Gainesville - 20/05/14 4 STOC Simulation #4 • STOC Simulations are exercises to train the team for LPF operations - different DA teams/sites mission-like schedule telecommand/telemetry as in operations • STOC end-to-end Simulation #4 (16-23 Nov 2013) included magnetic characterisation X LISA Symposium - M Nofrarias Gainesville - 20/05/14 5 STOC Simulation #4 • STOC Simulations are exercises to train the team for LPF operations - different DA teams/sites mission-like schedule telecommand/telemetry as in operations • STOC end-to-end Simulation #4 (16-23 Nov 2013) included magnetic characterisation X LISA Symposium - M Nofrarias Gainesville - 20/05/14 5 STOC Simulation #4 • STOC Simulations are exercises to train the team for LPF operations - different DA teams/sites mission-like schedule telecommand/telemetry as in operations Signal characteristics: • STOC end-to-end Simulation #4 (16-23 Nov 2013) included magnetic characterisation X LISA Symposium - M Nofrarias - Gainesville - 20/05/14 Four frequencies: [1, 3, 5, 7 ] mHz Fixed amplitude 1mA Applying ~5uT (interplanetary ~10nT) Duration 19h 5 Magnetic analysis pipeline X LISA Symposium - M Nofrarias Gainesville - 20/05/14 6 Magnetic analysis pipeline Download Preprocess Acceleration estimate compute mag. field heterodyne solve linear system X LISA Symposium - M Nofrarias Gainesville - 20/05/14 6 Analysis pipeline - displacement • Analysis based on telemetry parameters - applied voltage on coils, TM displacement magnetic experiment too close to CMS discharge (planning error) LTPDA 2.8.dev (R2012a) 2014−05−18 21:53:44.894 UTC ltpda: 637f026 iplot LTPDA 2.8.dev (R2012a) 2014−05−19 17:51:48.281 UTC ltpda: 637f026 −6 iplot 3 2 V1 V2 x 10 o1 o12 0 2 −2 Displacement [m] Voltage [V] 1 0 −1 −4 −6 −8 −10 −12 −2 −14 −3 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Origin: 2013−11−18 08:00:01.000 − Time [s] X LISA Symposium - M Nofrarias 5 5.5 −16 0.5 4 x 10 Gainesville - 20/05/14 1 1.5 2 2.5 3 3.5 4 4.5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5 5.5 4 x 10 7 Analysis pipeline - displacement • Analysis based on telemetry parameters - applied voltage on coils, TM displacement magnetic experiment too close to CMS discharge (planning error) LTPDA 2.8.dev (R2012a) 2014−05−18 21:53:44.894 UTC ltpda: 637f026 iplot LTPDA 2.8.dev (R2012a) 2014−05−19 17:51:48.281 UTC ltpda: 637f026 −6 iplot 3 2 V1 V2 x 10 o1 o12 0 2 −2 0 −1 −4 2 −8 −14 1.5 2 2.5 3 3.5 4 4.5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5 5.5 0 −10 −2 1 o1 o12 1 −12 −3 0.5 x 10 −6 −16 0.5 4 x 10 −1 Displacement [m] Displacement [m] Voltage [V] 1 LTPDA 2.8.dev (R2012a) 2014−05−19 17:51:49.228 UTC ltpda: 637f026 −8 iplot −2 −3 −4 1 −5 1.5 2 2.5 3 3.5 4 4.5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5 5.5 4 x 10 −6 −7 −8 0.5 X LISA Symposium - M Nofrarias Gainesville - 20/05/14 1 1.5 2 2.5 3 3.5 4 4.5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5 5.5 4 x 10 7 Analysis pipeline - force and torque • Translation into force and torques - requires system parameters (assumed to be known) LTPDA 2.8.dev (R2012a) 2014−05−18 19:10:43.430 UTC ltpda: 637f026 −9 iplot 1.29 LTPDA 2.8.dev (R2012a) 2014−05−18 19:58:13.898 UTC ltpda: 637f026 −12 iplot x 10 16 Force x 10 −Torque 14 1.285 −2 TM Torque [kg m rad s ] 1.275 10 8 2 TM Force [kg m s−2] 12 1.28 1.27 1.265 6 4 2 1.26 1.255 3 0 −2 3.5 4 4.5 5 Origin: 2013−11−18 08:00:01.000 − Time [s] X LISA Symposium - M Nofrarias 5.5 4 x 10 −4 3 Gainesville - 20/05/14 3.5 4 4.5 5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5.5 4 x 10 8 Analysis pipeline - force and torque • Translation into force and torques - requires system parameters (assumed to be known) LTPDA 2.8.dev 2.8.dev (R2012a) (R2012a) LTPDA 2014−05−18 19:10:44.524 19:10:43.430 UTC UTC 2014−05−18 ltpda: 637f026 637f026 ltpda: −9 −9 iplot iplot 1.29 LTPDA 2.8.dev (R2012a) 2014−05−18 19:58:15.410 19:58:13.898 UTC ltpda: 637f026 −12 iplot x 10 16 Force Force lowpass(Force) x 10 14 1.285 −Torque −Torque lowpass( −Torque) −2 TM Torque [kg m rad s ] 1.275 10 8 2 TM Force [kg m s−2] 12 1.28 1.27 1.265 6 4 2 1.26 1.255 3 0 −2 3.5 4 4.5 5 Origin: 2013−11−18 08:00:01.000 − Time [s] X LISA Symposium - M Nofrarias 5.5 4 x 10 −4 3 Gainesville - 20/05/14 3.5 4 4.5 5 Origin: 2013−11−18 08:00:01.000 − Time [s] 5.5 4 x 10 8 Analysis pipeline - force and torque • Translation into force and torques. Spectra: - Force contribution at f0 and 2xf0 Torque contribution at f0 High frequency IFO noise LTPDA 2.8.dev (R2012a) 2014−05−18 19:10:46.165 UTC ltpda: 637f026 iplot LTPDA 2.8.dev (R2012a) 2014−05−18 19:58:17.142 UTC ltpda: 637f026 iplot −10 −10 10 PSD(Force) PSD( −Torque) TM Torque [kg m2 rad s−2 Hz−1/2] TM Force [kg m s−2 Hz−1/2] 10 −11 10 −12 10 −13 10 −14 10 −11 10 −12 10 −13 10 −14 −4 10 −3 10 −2 10 Frequency [Hz] X LISA Symposium - M Nofrarias −1 10 0 10 10 −4 10 Gainesville - 20/05/14 −3 10 −2 10 Frequency [Hz] −1 10 0 10 9 Analysis pipeline - parameters • To determine the parameters: - heterodyne at the injected frequency to get the coupling coefficients solve the linear system - LTPDA 2.8.dev (R2012a) 2014−05−19 01:32:59.168 UTC ltpda: 637f026 −11 iplot 1.5 x 10 LTPDA 2.8.dev (R2012a) 2014−05−19 01:32:59.168 UTC ltpda: 637f026 −12 iplot Force Reconstructed x 10 8 1 Force Reconstructed 6 Force [kg m s−2] Force [kg m s−2] 4 0.5 0 2 0 −2 −4 −0.5 −6 −8 3.76 −1 −1.5 3 3.5 4 4.5 5 Origin: 2013−11−18 08:00:01.000 − Time [s] X LISA Symposium - M Nofrarias 3.78 3.8 3.82 3.84 3.86 3.88 3.9 3.92 3.94 Origin: 2013−11−18 08:00:01.000 − Time [s] 3.96 4 x 10 5.5 4 x 10 Gainesville - 20/05/14 10 Analysis pipeline - reconstructed force LTPDA 2.8.dev (R2012a) 2014−05−16 09:54:53.382 UTC ltpda: 637f026 iplot −9 10 TM Force −10 10 −11 −12 10 ! m kg √ Hzs2 " 10 −13 10 −14 10 −15 10 −5 10 X LISA Symposium - M Nofrarias −4 10 −3 10 −2 10 Fr equency [ Hz] Gainesville - 20/05/14 −1 10 0 10 11 Analysis pipeline - reconstructed force LTPDA 2.8.dev (R2012a) 2014−05−16 09:54:54.881 09:54:53.382 UTC ltpda: 637f026 iplot −9 10 TM Force TM Force component 0 −10 10 −11 −12 10 ! m kg √ Hzs2 " 10 −13 10 −14 10 −15 10 −5 10 X LISA Symposium - M Nofrarias −4 10 −3 10 −2 10 Fr equency [ Hz] Gainesville - 20/05/14 −1 10 0 10 11 Analysis pipeline - reconstructed force LTPDA2.8.dev 2.8.dev(R2012a) (R2012a) LTPDA 2014−05−1609:54:56.563 09:54:54.881 UTC 09:54:53.382UTC 2014−05−16 ltpda:637f026 637f026 ltpda: iplot iplot −9 −9 10 10 TM Force TMForce Force TM component component 0 0 −10 −10 2 10 10 0 component −11 −11 m m kg kg √ √ Hzs Hzs22 "" 10 10 −12 −12 !! 10 10 −13 10−13 10 −14 10−14 10 −15 10−15 −5 10 10−5 10 X LISA Symposium - M Nofrarias −4 10−4 10 −3 −2 −2 10−3 10 10 10 Fr equency equency [[ Hz] Fr Hz] Gainesville - 20/05/14 −1 −1 10 10 0 0 10 10 11 Analysis pipeline - reconstructed force LTPDA2.8.dev 2.8.dev(R2012a) (R2012a) LTPDA 2014−05−1609:54:56.563 09:54:54.881 UTC 09:54:53.382UTC 2014−05−16 ltpda:637f026 637f026 ltpda: iplot iplot −9 −9 10 10 TM Force TMForce Force TM component component 0 0 −10 −10 2 10 10 0 component −11 −11 m m kg kg √ √ Hzs Hzs22 "" 10 10 −12 −12 !! 10 10 −13 10−13 10 −14 10−14 10 −15 10−15 −5 10 10−5 10 −4 10−4 10 −3 −2 −2 10−3 10 10 10 Fr equency equency [[ Hz] Fr Hz] −1 −1 10 10 0 0 10 10 Last step in the pipeline is noise projection (requires magnetic field) X LISA Symposium - M Nofrarias Gainesville - 20/05/14 11 Magnetic analysis #2: magnetic field extrapolation Magnetometers are fluxgate, so magnetically active - Magnetic field needs to be interpolated to TM position Neural networks found to solve the problem - need to be ‘trained’, so relying on previous knowledge 1 σ1=9.2% σ2=1.45% 0.5 0 −100 M Diaz-Aguilo, A Lobo, E Garcia-Berro. Exp. Astronomy 30 2011 Pdf (Bz) Gainesville - 20/05/14 σ1=7.91% σ2=1.98% TM1 TM2 σ1=4.42% σ2=3.67% TM1 TM2 0.5 0 100 Interpolation error (%) 0.5 0 −100 0 100 Interpolation error (%) 1 0 −100 X LISA Symposium - M Nofrarias 1 TM1 TM2 y • ~ 30 magnetic active units surrounding the TMs create a complex mag. environment Pdf (B ) - 0 100 Interpolation error (%) 1 Pdf (|B|) • can’t be near the TMs Pdf (Bx) • σ1=4.28% σ2=0.83% TM1 TM2 0.5 0 −100 0 100 Interpolation error (%) 12 Exploring the LPF magnetic map X LISA Symposium - M Nofrarias Gainesville - 20/05/14 13 Exploring the LPF magnetic map X LISA Symposium - M Nofrarias Gainesville - 20/05/14 13 Exploring the LPF magnetic map X LISA Symposium - M Nofrarias Gainesville - 20/05/14 13 Exploring the LPF magnetic map X LISA Symposium - M Nofrarias Gainesville - 20/05/14 13 FEEP to Cold Gas thrusters FEEPs Cold Gas (measured sources) X LISA Symposium - M Nofrarias Gainesville - 20/05/14 14 Sources contribution - magnetometers Nominal contribution of sources to mag. measurement 40 20 0 0 10 20 Sources Magnetometer #3 80 60 30 40 20 0 0 X LISA Symposium - M Nofrarias 10 Sources 20 30 ULU CGLV MY CGLV MXPY CGLV PXPY 80 60 40 20 0 0 10 20 Sources Magnetometer #4 80 30 CGLV MY CGLV MXPY CGLV PXPY 60 Magnetometer #2 60 PM d1 PM d2 CGLV MY CGLV MXPY CGLV PXPY 80 Mag. Field Contribution [%] Magnetometer #1 Mag. Field Contribution [%] Dominated by cold gas latch valves (CGLV) CGLV MY CGLV MXPY CGLV PXPY Mag. Field Contribution [%] Mag. Field Contribution [%] - CCU • 40 20 0 0 Gainesville - 20/05/14 10 Sources 20 30 15 Sources contribution - test masses Test Mass #1 Test Mass #2 35 30 25 20 15 TM contribution dominated by Cold Gas LV (~70%) • Cold Gas Thrusters (and other units) are actively tested during operations • Could we turn the problem into a source detection/subtraction ? 35 30 25 20 15 10 10 5 5 0 0 • 40 Mag. Field Contribution [%] Mag. Field Contribution [%] 40 ULU CGLV MY CGLV MXPY CGLV PXPY 45 CCU CGLV MY CGLV MXPY CGLV PXPY CCU 45 50 PM d1 PM d2 50 10 20 X LISA Symposium - M Nofrarias 30 0 0 10 Gainesville - 20/05/14 20 30 16 Magnetic sources location exercise • The LPF spacecraft went through a Thermal Balance/Thermal Vacuum test (Oct 2011) - including magnetometers in the payload X LISA Symposium - M Nofrarias Gainesville - 20/05/14 17 OSTT campaign real data X LISA Symposium - M Nofrarias Gainesville - 20/05/14 18 OSTT campaign real data X LISA Symposium - M Nofrarias Gainesville - 20/05/14 18 OSTT campaign real data X LISA Symposium - M Nofrarias Gainesville - 20/05/14 18 OSTT campaign real data X−component 0.03 0.02 0 −0.01 −0.02 −0.03 0.04 0.01 0 −0.01 −0.02 0.02 0 −0.02 −0.04 −0.03 −0.06 −0.05 −0.04 −0.08 4.325 4.33 4.335 4.34 4.345 Origin: 2011−10−22 08:54:09.938 − Time [s] X LISA Symposium - M Nofrarias 4.35 5 x 10 −0.05 4.32 4.325 4.33 4.335 4.34 4.345 Origin: 2011−10−22 08:54:09.938 − Time [s] Gainesville - 20/05/14 Mag. #1 Mag. #2 Mag. #3 Mag. #4 0.06 −0.04 −0.06 4.32 Z−component 0.08 Mag. #1 Mag. #2 Mag. #3 Mag. #4 0.03 Magnetic Field [uT] 0.01 LTPDA 2.8.dev (R2012a) 2014−05−17 13:50:41.515 UTC ltpda: 637f026 iplot Y−component 0.04 Mag. #1 Mag. #2 Mag. #3 Mag. #4 0.02 Magnetic Field [uT] LTPDA 2.8.dev (R2012a) 2014−05−17 13:50:39.353 UTC ltpda: 637f026 iplot Magnetic Field [uT] LTPDA 2.8.dev (R2012a) 2014−05−17 13:50:37.669 UTC ltpda: 637f026 iplot 4.35 5 x 10 −0.1 4.32 4.325 4.33 4.335 4.34 4.345 Origin: 2011−10−22 08:54:09.938 − Time [s] 18 4.35 5 x 10 Magnetic MUSIC • LPF magnetometers can be treated as a sensor array - take advantage of geometry to locate sources X LISA Symposium - M Nofrarias Gainesville - 20/05/14 19 Magnetic MUSIC • LPF magnetometers can be treated as a sensor array - • take advantage of geometry to locate sources MUltiple SIgnal Classification (MUSIC) method is based on subspace properties of the covariance matrix to: - Schmidt, R.O, IEEE Vol. AP 34 (March 1986) locate sources determine number of sources in received signal X LISA Symposium - M Nofrarias Gainesville - 20/05/14 19 Magnetic MUSIC • LPF magnetometers can be treated as a sensor array - • take advantage of geometry to locate sources MUltiple SIgnal Classification (MUSIC) method is based on subspace properties of the covariance matrix to: - Schmidt, R.O, IEEE Vol. AP 34 (March 1986) locate sources determine number of sources in received signal Music algorithm: X LISA Symposium - M Nofrarias Gainesville - 20/05/14 19 Magnetic MUSIC - real data X LISA Symposium - M Nofrarias Gainesville - 20/05/14 20 Magnetic MUSIC - real data X LISA Symposium - M Nofrarias Gainesville - 20/05/14 20 Magnetic MUSIC - real data Star Tracker Electronics Sun Sensor X LISA Symposium - M Nofrarias Gainesville - 20/05/14 20 Conclusions • Force noise due to magnetic coupling of the test mass is the main contribution coming from direct forces at low frequency • We have developed a data analysis pipeline for the inflight experiments with coils. Tested in STOC Simulation • Magnetic field needs to be interpolated to the TM position. Work to improve better understand the structure and main contributors of the magnetic field • We are developing tools to help locate magnetic sources introducing a magnetic signal in the system X LISA Symposium - M Nofrarias Gainesville - 20/05/14 21 Thank you