+ All Categories
Home > Documents > Modeling VRALA, the Next-Generation Actuator for High ...C. Del Vecchio1 G. Agapito1 G. Tomassi2 E....

Modeling VRALA, the Next-Generation Actuator for High ...C. Del Vecchio1 G. Agapito1 G. Tomassi2 E....

Date post: 25-Jan-2021
Category:
Upload: others
View: 1 times
Download: 0 times
Share this document with a friend
50
Vrala Del Vecchio, Agapito, Tomassi, de Santis Background The AO Principle The Design Drivers Statics The Approach The Model Dynamics The Governing Equations The Open-Loop Response The Closed-Loop Response Summary Modeling VRALA, the Next-Generation Actuator for High-Density, Tick Secondary Mirrors for Astronomy Comsol for Adaptive Optics C. Del Vecchio 1 G. Agapito 1 G. Tomassi 2 E. de Santis 2 1 INAF-OAA Florence, Italy 2 University of Cassino – DAEMI Cassino, Italy 2010 Comsol Conference Paris, Nov 19 2010 Presented at the COMSOL Conference 2010 Paris
Transcript
  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Modeling VRALA, the Next-GenerationActuator for High-Density, Tick Secondary

    Mirrors for AstronomyComsol for Adaptive Optics

    C. Del Vecchio1 G. Agapito1 G. Tomassi2

    E. de Santis2

    1INAF-OAA Florence, Italy2University of Cassino – DAEMI Cassino, Italy

    2010 Comsol Conference Paris, Nov 19 2010

    Presented at the COMSOL Conference 2010 Paris

    http://www.comsol.com/conf_cd_2011_eu

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Compensating the Atmospheric TurbulenceThe Control System Concept

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Adaptive Optics on board the TelescopeSystem Overview

    [Riccardi et al., 2004]

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Actuating the DM & Sensing the DisplacementsThe LBT Voice-Coil Actuator

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Basic Requirements of High Order DM’sThe Specs are very Severe

    rms force (turbulence correction) .363 Nmax force (static) .36 Nmax force (dynamic) 1.27 Nstroke (usable) ±100 µmstroke (mechanical) ±150 µmbandwidth 1 kHztypical inter-actuator spacing 25 mmtypical actuator length ≤ 60 mmtypical mover mass ≤ 10× 10−3 kgDC resistance 2 to 2.5 Ω

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    DM Stiffness vs. DM Thickness & Act SpacingThe Plate Stiffness is Strongly Non-Linear

    The plate stiffnessKflex ∝ t3 × (1/d)4

    t = thickness d = dimension

    What ifthe inter-actuator spacing is slightly reducedthe thickness is slightly increased

    HIGHER ORDER DM d = 30→ 25 mm (16%)ELT PANELS t = 1.6→ 2 mm (20%)

    } 2×Kflex

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Design Criterion: Avoid Thermal PollutionThe (usual) Basic Question and the (enhanced) Answer

    reduce the local seeing⇓

    reduce any local heating⇓

    given the force, reduce the power⇓

    maximize the efficiency(the force-to-power ratio)

    WHILE

    ♣ respecting the geometry♣ minimizing the emi

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic CoreVariable Reluctance LM: F =

    ZS−1

    2(H · B) n + (n · H) BT dS

    [Del Vecchio et al., 2010]

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic CoreVariable Reluctance LM: F =

    ZS−1

    2(H · B) n + (n · H) BT dS

    [Del Vecchio et al., 2010]

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic CoreVariable Reluctance LM: F =

    ZS−1

    2(H · B) n + (n · H) BT dS

    [Del Vecchio et al., 2010]

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Full Coil ApproximationThe Filling Factor ϕ Dramatically Reduces the DOF’s

    StaticRf =

    ϕN2 R

    J2f = ϕJ2 = ϕ

    (I

    Aw

    )2Ff = 1ϕF

    V f =√ϕ

    N VTransientJf = N IAw

    V indf = NV ind

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Numerical Optimization IA Single Matlab Script to Fully Calculate the Magnetic Response

    geometry define the (very simple!) basic components inthe r -z plane

    meshing get the elements (typically 10000) and embedthem in the azimuthal currents applicationmode

    physics definethe physical properties of the chosenmaterials (via tables or plots provided bythe manufacturers), including the airthe input external current density (with theproper correction factor)

    solution solve the non linear system (of typically 20000equations) for Aϕ

    post-proc. compute the magnetic force via the Maxwellstress tensor (multiplying by ϕ)

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Numerical Optimization IIMaterials & Geometry: �max ≈ 7 N×W−1

    mat 64% of 17× 17 combinations � ≥ 6 N×W−1geom � = �

    (hmov , rstato ,W ,H,R

    )7→ � = �

    (W ,H,R

    )⇓

    �� ��Iron losses≤ 1.6% of the DC power, via frequency analysis

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Numerical Optimization IIIPrototypes: the Magnetostatics results are experimentally confirmed

    value Pure Fe FerriteR1 [mm] 1.5 1.5R2 [mm] 6 4.5R3 [mm] 11 9R4 [mm] 12.5 10.75h [mm] 12 5turns 400 85

    force [N] 1.95 0.71voltage [V] 1 0.75current [A] 0.2 0.8power [W] 0.2 0.6� [N ×W−1] 9.75 1.18

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Analytical OptimizationThe Comsol Results Match the Analytical Ones

    rstati ↪→ R1R −W/2 ↪→ R2R + W/2 ↪→ R3

    rstato ↪→ R4H ↪→ h

    � = K(R24 − R23 + R22 − R21)(R3 − R2)

    R3 + R2

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Setting Up the ModelThe Full Coil is Implemented Multi-Physically

    Aφ︸︷︷︸m/s

    ⊗ r , z︸︷︷︸ALE

    ⊗F = (M + m0)z̈︸ ︷︷ ︸ODE

    ⊗ (Vext − Vind )/R︸ ︷︷ ︸coupling eq

    where Vext = IextR Iext =√ϕ AAw

    V f =∫

    A

    (−ef + Jfρ)2πrA

    dA

    where∫

    A

    ef 2πrA

    dA is V indf

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic InertiaEnergy Balance

    P(t) = Vi(t) = PFe + PCu︸ ︷︷ ︸Pheat

    +Pmagn + Pkinet

    Pheat � P ⇒ P≈{

    Pmagn for t ≤ .8Pmagn + Pkinet for t > .8�� ��The e/m inertia is a big issue for the control system

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic InertiaEnergy Balance

    P(t) = Vi(t) = PFe + PCu︸ ︷︷ ︸Pheat

    +Pmagn + Pkinet

    Pheat � P ⇒ P≈{

    Pmagn for t ≤ .8Pmagn + Pkinet for t > .8�� ��The e/m inertia is a big issue for the control system

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Electromagnetic InertiaEnergy Balance

    P(t) = Vi(t) = PFe + PCu︸ ︷︷ ︸Pheat

    +Pmagn + Pkinet

    The stored magnetic energy

    Pmagn =∫

    V

    (∫ B0

    H dB

    )dV

    defined via Comsol functionsafter Matlab numerical integration

    Pheat � P ⇒ P≈{

    Pmagn for t ≤ .8Pmagn + Pkinet for t > .8�� ��The e/m inertia is a big issue for the control system

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Outline

    1 BackgroundThe AO PrincipleThe Design Drivers

    2 StaticsThe ApproachThe Model

    3 DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Real-Time-Updating LQR IThe Block Diagram

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Real-Time-Updating LQR IIDesign Principles

    the state x is defined as x(k) = [ I(k) z̈(k) ż(k) z(k) ]′(sampling @ 50 kHz at each pk )

    I = current signalz = position feedback (z ⇒ ż ⇒ z̈)

    at each k step, the system matrices update determinesA(p) and B(p) in x(k + 1) = A(p)x(k) + B(p)u(k)

    A(p) 7→B(p) 7→ LQR(p,λ) 7→ F (k) (control matrix)

    2 limits: |Ii | ≤ Imax &∣∣∣∣dIidt

    ∣∣∣∣ ≤ dImaxIi 7→ coil current splitter 7→

    {I ti if Ii > 0Ibi if Ii < 0�� ��λ is the forgetting factor, according to RLS theory

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Real-Time-Updating LQR IIIThe 2 Transfer Functions

    The plant is the combination of 2 TF’s

    1 F = f (z, I), non linear, but linearized @ p = (z̄, Ī) byf (z, I) ≈ k(p) + kz(p)δz + kI(p)δI

    k(p) = f (p)kz(p) =

    ∂f (z,I)∂z

    ∣∣∣(p)

    kI(p) =∂f (z,I)

    ∂I

    ∣∣∣(p)

    2 I(s)Ii (s)

    = 1ωs+1 , a first order low-pass filter

    ω is the time constantI(s) is the Laplace transform of the coil currentIi (s) is the Laplace transform of the input current

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Closed-Loop ResultsThe Step Response

    z0 = 0; δ = 1 to 5 µm

    �� ��ts for δ = 4 and 5 µm exceed by 10% the goal

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Closed-Loop ResultsThe Step Response

    z0 = −100 to 100 µm; δ = 1 µm

    �� ��ts ≤ 1 ms

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Closed-Loop ResultsThe Step Response

    z0 = −100 to 100 µm; δ = 2 µm

    �� ��ts ≤ 1 ms

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    The Closed-Loop ResultsThe Step Response

    The maximum average powercomputed over the entire time domain

    ranges from 1.3 to 10.7 mW

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Lessons Learned & Future Work

    A challenging projectThe high-order, long-stroke, very large deformable mirrorsof the next generation telescopes require very large forces

    and unprecedented actuator densities. The simple and veryeffective magnetic circuit of VRALA is well-suited to

    accomplishing this goal.

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Lessons Learned & Future Work

    The actuator can accomplish the demandingspecifications with

    � = 7 N×W−1 → low power dissipationts = .71 ms for δ = 1 µm→ high speed

    Φ ≥ 25 mm→ small separationsThe Comsol results are (statically) verified by

    two very simple, preliminary prototypesan analytical optimization

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Lessons Learned & Future Work

    Further pushing the technology boundariesVRALA, the last chapter of the short but rich history of theAO technology has established many achievements. The

    encouraging results indicate the near future developments.

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Lessons Learned & Future Work

    Complete prototype (provided with a feedback capacitive sensor)possible construction issuesclosed loop responsepower dissipation

    Further computationspossible alternative control system designsclosed loop frequency responserefined multiphysics (HT+NS+SM)

    cooling system designmagneto-mechanics as a function of T

    3D modelingwhole system simulation (mutual effects)effects of tolerances/errors

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Some Final Explanations

    �� ��Variable Reluctance Adaptive mirror Linear Actuator

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Some Final Explanations

    Vråla To roar(Swedish) (English)

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    BackgroundThe AO Principle

    The Design Drivers

    StaticsThe Approach

    The Model

    DynamicsThe GoverningEquations

    The Open-LoopResponse

    The Closed-LoopResponse

    Summary

    Some Final Explanations

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    Appendix

    For Further Reading I

    Del Vecchio, C., Marignetti, F., Agapito, G., Tomassi, G.,and Riccardi, A. (2010).Vrala: Designing and prototyping a novel,high-efficiency actuator for large adaptive mirrors.In Ellerbroek, B. L., Hart, M., Hubin, N., and Wizinowich,P. L., editors, Adaptive Optics Systems, volume 7036 ofProc. SPIE. SPIE.

  • Vrala

    Del Vecchio,Agapito,Tomassi,de Santis

    Appendix

    For Further Reading II

    Riccardi, A., Brusa, G., Xompero, M., Zanotti, D., DelVecchio, C., Salinari, P., Ranfagni, P., Gallieni, D., Biasi,R., Andrighettoni, M., Miller, S., and Mantegazza, P.(2004).The adaptive secondary mirrors for the Large BinocularTelescope: a progress report.In Bonaccini Calia, D., Ellerbroek, B. L., and Ragazzoni,R., editors, Advancements in Adaptive Optics, volume5490 of Proc. SPIE, pages 1564–1571. SPIE.

    BackgroundThe AO PrincipleThe Design Drivers

    StaticsThe ApproachThe Model

    DynamicsThe Governing EquationsThe Open-Loop ResponseThe Closed-Loop Response

    SummaryAppendix


Recommended