QD0 stabilization L. Brunetti 1, N. Allemandou 1, J.-P. Baud 1, G. Balik 1, G. Deleglise 1, A. Jeremie 1, S. Vilalte 1 B. Caron 2, C. Hernandez 2, (LAViSta Team) 1 : LAPP-IN2P3-CNRS, Université de Savoie, Annecy, France & 2 : SYMME-POLYTECH Annecy-Chambéry, Université de Savoie, Annecy, France

2 CLIC Challenge vs QD0 stabilization Final focus CLIC R&D: Sub-nanometer beams Developments of Lavista team are dedicated to the final focus CLIC CERN QD0 Stabilization 0,2 nm 0,1Hz

3 Introduction  IP Beam based feedback : already developed in collaboration with CERN since 2010  Mechanical stabilization has to be achieved CLIC CERN QD0 Stabilization3 Final focus : beam stabilization strategy  At the IP (mechanical + beam feedback), we aim at 0,2nm at 0,1Hz

4 IP feedback Beam trajectory control : simulation under Placet CLIC CERN QD0 Stabilization - Caron B et al, 2012, “Vibration control of the beam of the future linear collider”, Control Engineering Practice. - G. Balik et al, 2012, “ Integrated simulation of ground motion mitigation, techniques for the future compact linear collider (CLIC) “, Nuclear Instruments and Methods in Physics Research Luminosity vs control ON or OFF and vs model of seismic motion (deal under Placet)

CLIC CERN QD0 Stabilization5 Prototype : Active control demonstration Results : control with commercial sensors (2 geophones and 2 accelerometers)  0,6 nm Balik et al, “Active control of a subnanometer isolator“, JIMMSS.  Main limitation : SENSOR (simulation and experiment).

6 Transfer on a real scale CLIC CERN QD0 Stabilization6 Demonstration active table to QD0 active control ? One active foot Several active feet  Structure : QD0 Magnet  Sensors  Actuators  Integration: control, data processing, real time… Mecatronics challenge Schedule : CLIC development phase

Mechanical structure Simulation studies of QD0 :  Initial study performed by the team of M. Modena CERN : mechanics aspects…  LAVISTA : control of a elementary prototype  FEM : dynamics of the quadrupole given its modal response  State-space model and control strategy  Test : on a real-scale prototype – “dummy QD0 magnet” CLIC CERN QD0 Stabilization7

8 Finite elements simulations Free system 224 Hz 312 Hz 450 Hz 555 Hz Antonio: 208 Hz Antonio: 295 Hz Antonio: 416 Hz Antonio: 511 Hz (i) Simulation studies of QD0 : Free system Simulation studies of QD0 : Constrained system 243 Hz Antonio: 190 Hz 286 Hz Antonio: 260 Hz 332 Hz Antonio: 310 Hz 388 Hz Antonio: 366 Hz

QD0 activities (ii) State space model and control strategy :  Targets : several aspects have to be defined - Location and number of active feet - Type of active feet - Degrees of freedom - Type of control (SISO, MIMO) - To adjust the specifications of actuators and sensors - Conditioning, real time processing…  After normalizing, the system can be expressed in the form of a state space model  Integration in a control loop using Simulink with the whole simulation (sensor, actuator, ADC, DAC, Data processing…. And seismic motion model and its coherence) : first simulations done (results are interesting) CLIC CERN QD0 Stabilization9 ForceDisplacement  Another model : the input is the voltage applied to the PZT actuators and the output is QD0 displacement : current stage

CLIC CERN QD0 Stabilization10  Design and modelling of the large prototype (recruitment of a student in progress)  Modal analysis of the short QD0 to confirm the simulation results (without coils). Prospectives  Objectives of the prototype : -Dynamics behavior (eigenfrequencies, damping…) -Size -Geometry -Mass  Most elementary as possible for machining, assembling, cost, delays…  SD0? (iii) Dummy QD0 magnet prototype :

CLIC CERN QD0 Stabilization11 Actuator Industrial solution : PZT actuator  Two challenging ways : internal development or industrial partnership… Manufacturer identified Past “similar” developments for vibration control dedicated to machining Specifications will be the result of the whole simulation (prototype, sensor…) Collaboration with another French laboratory for the powering part. No commercial solution for dynamics, resolution, load, stiffness…  Funding plan is needed A French proposal (to ANR agency) has been recently submitted, 1 st selection at beginning of March.  Has to be tested on the dummy magnet prototype Example of an large actuator Small size PZT actuator

CLIC CERN QD0 Stabilization12 Sensor Already the limitation for the «demonstration » active table: Sensors noise Sensors transfer function  No commercial solution, so different challenging options: - Laboratory developments - Commercial investigations - PACMAN program (3.2)…  Internal development at LAPP

CLIC CERN QD0 Stabilization13 Measurements Comparison with Guralp and Wilcoxon:  Efficient  Approach validated → patent  Next stage : miniaturization Demonstration sensor

CLIC CERN QD0 Stabilization14 Measurements Comparison with Guralp and Wilcoxon: Miniaturized sensor Accelerometer LAPP sensor Geophone 6TGeophone 3ESP  Still efficient  Control :

CLIC CERN QD0 Stabilization15 Sensor Development in progress:  “Medium” bandwidth : assembling and tests in progress  Validation of the machining and some of the sensor properties  “High” bandwidth: equivalent to accelerometers  Less challenging  Could be interesting for control  “Large” bandwidth: development in progress in term of machining and method  Method tests will be done the next month.  Most challenging – measure very low frequencies and on a large bandwidth - Very strategically for control  French AAP (call for project) with SYMME for material properties and process  Sensitivity to magnetic field  Tests on an experimental site : first at CERN (stabilization group), then ATF2, CLIC module or CLEX… Future:  A new mechanical engineer has joined the team.

CLIC CERN QD0 Stabilization16 Sensor Status:  A first innovative version has demonstrated the feasibility : performances are about the same as Guralp 6T.  2 miniaturized versions with our own electronics : still almost the same performance, but small adjustments have to be made (tuning, drift...)  Promising results in control  “More industrial” versions are in progress whose one in low frequencies  Adapted to control  Cost  Size Advantages:

CLIC CERN QD0 Stabilization17 Conclusion vs QD0 stabilization Structure:  Simulations in progress (f.e., state space model, prototype CAD...) Actuator:  Partnerships identified and the strategy chosen  Funds, to be found Sensor:  Efficient and innovative sensor is being developed  Several versions are manufactured to be able to build one for very low frequencies and efficient on a large bandwidth  Test on a site is foreseen

ATF2 : 2013 Measurement campaign Collaboration LAPP - CERN  FD relative displacement measurements  Feedforward ATF2 GM System team: A. Jeremie, K.Artoos, C.Charrondière, J.Pfingstner, D.Schulte, R.Tomas-Garcia, B.Bolzon (past experience) LCWS TOKYO QD0 Stabilization18

FD : Final Doublet Shintake monitor : beam size measurement ATF2 layout LCWS TOKYO QD0 Stabilization 19 Final Doublet relative displacement measurements :

ATF2 Objectives : ATF2 – Strategy of stabilization  Steady and repetitive beam with a radius of 37 nm at the focus point.  Past : relative motion between shintake monitor and final doublets of 6 nm 0,1 Hz in the vertical axis (i.e. B. Bolzon results). 20 Sol Shintake monitor Last focusing magnets Beam Interference fringes Previous magnets Relative motion beetwen Shintake monitor and last focusing magnets 4m Efficient coherence of ground motion: measured on site  Very rigid and independent supports i.e. B.Bolzon thesis LCWS TOKYO QD0 Stabilization

Shintake monitor has more components and is in operation: QF1FF has been replaced by a heavier magnet with better field quality and larger radius: LCWS TOKYO QD0 Stabilization QD0 QF1 ATF2 - Why making new measurements?

With ground motion, relative motion at 1 Hz of Shintake to [QD0; QF1] : 2008 by B. BOLZONToleranceMeasurement [SM-QD0]Measurement [SM-QF1] Vertical7 nm (for QD0) 20 nm (for QF1) 4.8 nm6.3 nm Perpendicular to the beam ~ 500 nm30.7 nm30.6 nm Parallel to the beam~ 10,000 nm36.5 nm27.1 nm 2013 by A. JEREMIEToleranceMeasurement [SM-QD0]Measurement [SM-QF1] Vertical7 nm (for QD0) 20 nm (for QF1) 4.8 nm30 nm Parallel to the beam~ 10,000 nm25 nm290 nm LCWS TOKYO QD0 Stabilization  New QF1 : relative motion of Shintake monitor to new QF1 > Tolerance  Outside tolerance for 2% effect on beam!