1RIA, Valencia March 30th, 2012
Francisco PradaInstituto de Física Teórica (UAM/CSIC)
Instituto de Astrofísica de Andalucía (CSIC)
BigBOSSBigBOSS Focal Plane Assembly Focal Plane Assembly
[DOE Review of BigBOSS, Berkeley, Dec. 6-7, 2011]
2
ContentsContents
• BigBOSS Spanish Participation Group (BSPG)• WBS & Tasks assigned• Focal plate requirements, status, interfaces and R&D scope• Positioner requirements, status, R&D scope
3
BigBOSSBigBOSS in Spain in Spain
Spanish Participation Group:
• Instituto de Astrofísica de Andalucía, IAA-CSIC Contact: F. Prada (BSPG Spokesperson)• Instituto de Astrofísica de Canarias, IAC Contact: C. Allende-Prieto• Institut de Ciencies del Cosmos, ICC-UB Contact: J. Miralda-Escudé• Instituto de Física Teórica, IFT UAM/CSIC Conctact: A. González-Arroyo• Observatori Astronomic de la Universitat de Valencia, OA-UV Contact: V. Martínez
4
• IAA-CSIC BigBOSS Team:
F. Prada, P.I. & Instrument ScientistM. Azzaro, Project ManagerS. Becerril, System & Mechanical EngineerJ. Sánchez, Electronics EngineerA.D. Montero-Dorta, PostdocIn collaboration with the Spanish Company AVS.Other collaborations:CEI UAM+CSIC, EPS Madrid (electronics)MultiDark Consolider Project
• UCB/LBNL Team:R. Besuner, J. Edelstein, P. Jelinsky, P. Perry, M. Sholl, J.Silber, C. Schenk, Z. Zhou.
• USTC Team:C. Zhai, H. Hu
Contributors to the Focal Plane activitiesContributors to the Focal Plane activities
5
Reference Design
Focal Plane System
6
Fibers(pointing down)
7
Focal Plane WBS & TasksFocal Plane WBS & Tasks• Focal plate.
• Requirements and error budget• Topology• CAD modeling• FEM analysis (with LBNL)
• Structural• Thermal Transient
• Search for proper materials AlSi machineability tests• Thermal study (with LBNL)• Focal Plate tolerancing (with LBNL)
• Theta-theta fiber positioner.• Requirements and error budget• Concept design• Motor tests• Electronics concept• Positioning software
• Develop Interfaces• Focal Plate/Actuator interfaces• Interfaces for fiducial fibers• Interfaces for guiders• Repeatable, semi-kinematic mounts• Adapter interfaces
8
Focal Plate RequirementsFocal Plate Requirements
!"#"$%&'$"()*$"+"#,-'./$',0"'1/2%&'3&%,"
!"#$%&'()'*+,"*,(&- .///0(1%'2*3%&4 0%5*6(4*178*#%,%& /9:.#;1*,%',<8+=4%-- >//##
;(-8?(4%&'28,+< >@##'(4',<%'A(+*1'B"&)*+%
C*D8"-'()'+"&E*,"&%'()')(+*1'-"&)*+% F@9.#
4%5*+%&'6$$/$-'/#'7'8*$"29/#
A86"&%'%&&(&'()')(+*1'21*,% GHI/J#'K6(*1L'@/J#MN<%&#*1'*4D'6&*E8,*?(4*1'-*6'()')(+*121*,%
GH@/J#
4%5*+%&'6$$/$-'/#'&%,"$%&'8*$"29/#':;<=
O*,%&*1'2(-8?(4846'*++"&*+P'()'<(1%- GHQ/J#
4%5*+%&'>*&,'6$$/$-0(1%'*46"1*&',(1%&*4+% GH/9/@R
>"+?"$%,)$"'@"()*$"+"#,-B"&E8E*1',%#2%&*,"&%'&*46% S@/RT',('UV/RTW2%&*?46',%#2%&*,"&%'&*46% S>/RT',('I/RT
0(1%'+<*#)%& /9>',('/9I##X("4,'Y*46%'%D6%'&*D8"- Z4D%&+",
Focal Plane Patrol discs layout
Focal Plate/Actuator Interfacefor IAA/AVS, USTC & LBNL
A#,"$.%2"'@"()*$"+"#,-["8D%&-\']D"+8*1']$%&-\'=84%?+'#("4?46-\'%,+
9
Focal Plate topologyFocal Plate topology
Because of the curvature of the plate, the pitch of the positioners cannot be uniform and themaximum difference will be of the order of 72 !. Also, because of the step motors, the fibercan move on a grid of nearly 210 million positions within the patrol disc of each positioner.
The theta-theta concept, combined withstepper motors, gives a discrete set ofpositions available to the fiber on the patroldisc. The worst-case resolution available isabout 4 ! (requirement is < 5 !).
The deformationof different partsof the border ofthe focal plate.The curvatureradius is muchsmaller than thereal one (750mm) in order tomake the effectmore visible.
" 975mm
10
Aluminum, steel and Al/Si alloy are considered. Tests are ongoing in order to ascertain the best-structurally and thermally behaving material. Al/Si was tested also for machining, because it is aless known alloy, with the advantage that its expansion coefficient is very close to that of steeland is stiffer than normal Aluminum.A bulk of Al/Si was set under test in order to find the best machining method. Milling of themetal is found to give the best results both for flat surfaces and forming holes (drilling and wirecutting were also tested). No problems are found with small wall thickness. Tapping the alloywith small threads does not present special difficulties. No overheating is observed.
Drilled (left) versus milled (right) holes Milled holes finish detail
Focal Plate material testsFocal Plate material tests
11
The most critical gravity deformation is in the Z direction. Analyses performed in collaborationwith LBNL show that, with an Al/Si plate 100 mm thick and clamped boundaries, the maximumZ displacement can be kept to less than 4!m (requirement is <~15!m).
Al/Si Focal Plate Gravitational DeflectionAl/Si Focal Plate Gravitational Deflection
Thermal DistortionThermal Distortion
12
B>66C14 um peak deflection
DCE4AFE412 um peak deflection
• Thermal gradient deflection through focal plate:Constant heat flux (q=1155 w/m2) at z=0 for 52sec
• Axi-symmetric finite element model of focal plate and adapter
Fiber Side
FixedBarrel I/F
Heat Flux Side
Preliminary results suggest that aluminum may be acceptable, despite its relatively high CTE dueto its high thermal diffusivity and rapid equilibration of through-the-thickness temperaturegradients.
Fiber Fiber PositionerPositioner Retention Retention
• Development and testing of interface hardware forclamping fiber positioners into focal plate
• Must retain positioner with long-term stability in arbitraryorientations with varying environmental conditions
• Principle is to use a tapered screw in a retained insert orflexural ring to clamp multiple adjacent positioners
• Other possibilities will be explored (e.g. actuator screweddirectly onto the front of the plate)
14
Focal Plate R&D ScopeFocal Plate R&D Scope
During the R&D phase the most critical issues of the designshall be faced up in order to produce a safe and consistentpreliminary design.
Most of them will be mitigated by the production and testingof a sub-scale plate. Indeed, this will give us importantinformation about the long-term dimensional stability of thematerial and the errors produced during the manufacturing.The thermal/structural FEM models will also be checkedthrough the sub-scale plate.
Thus, both the interface with the fiber positioner and theerror budget tolerances of the Focal Plate will berevised/confirmed. Efficient metrology and manufacturingprocedures for the further production of the Focal Plate willalso be established.
All these inputs will result in a realistic cost estimate of theFocal Plate component.
15
Actuator RequirementsActuator Requirements
!"#"$%&'$"()*$"+"#,- 4%5*+%&'6$$/$-'/#'&%,"$%&'8*$"29/#':;<=A8$%&'+(&%'D8*#%,%& >@/J# O*,%&*1'2(-8?(4846'2&%+8-8(4'()'*+,"*,(& G.J#C%2(-8?(4846'?#% Q.- O*,%&'2(-8?(4846'*++"&*+P'()'*+,"*,(& GQ/J#
A"11'+(E%&*6%'()'A(+*1'B"&)*+%>I9^V##'D8*#%,&%';*,&(1
78-+;(-8?(4%&'-,%2'84<(#(6%4%8,P G@J#
@4D'+P+1%'*$-(1",%'%&&(& G.J#4/)#9#G'%#8'HI"$'$/)9#G$"()*$"+"#,-
;*,&(1'78-+'[&*4"1*&8,P G@J#
X("4?46',&*_%+,(&P'()'A8$%&A%&&"1%'84,(',<%'`+,"*,(&
A&(#'$*+='-8D%',(')&(4,'-8D%K(&'E8+%E%&-*M
a&&(&'2&(D"+%D'$%,b%%4'-"++%--8E%',&*+=846+(&&%+?(4-
G^J#
X*58#*1'$%4D846'&*D8"-'()',<%]$%&
./## N&*+=846'+("&-%'%&&(& G.J#
C("?46'()']$%&',<&("6<',<%*+,"*,(&
c>9.##'+1%*&'D8*#%,&%&("?46')&(#'%4D',('%4D
4%5*+%&'>*&,'6$$/$-A#,"$.%2"'J*,0',0"'1/2%&'3&%," >/6V'(E%&'@.##'1%46,< ;(-8?(4%&'*4D')%&&"1%'<(1D%&',(1%&*4+% GH/9>.R
4%5*+%&'6$$/$-'/#'7'8*$"29/# >"+?"$%,)$"'@"()*$"+"#,-A8$%&'2(-8?(4%&'%&&(&'K(E%&'&*46%()'#(?(4M
GH@/J#B"&E8E*1',%#2%&*,"&%'&*46% S@/RT',('UV/RT
;&%+8-8(4'(E%&'&*46%'()'#(?(4 GH.J# W2%&*?46',%#2%&*,"&%'&*46% S>/RT',('I/RTB2<%&8+*1'7%2*&,"&% GH.J#X("4?46'2&%+8-8(4'A8$%&'A%&&"1%b9&9,9'`+,"*,(&
GH.J# K,0"$-
X("4?46'2&%+8-8(4'`+,"*,(&b9&9,'A(+*1';1*,%
GH.J# O8)%?#% @./\///'+P+1%-
IAA/AVS Actuator ConceptIAA/AVS Actuator Concept
16
• Two rotational DoFs (theta-theta concept)• Backlash cancelled through standard torsion springs
(successfully tested)• Gearing ratios: 1/512 for ROT2 and 1/1024 for ROT1• Two stops limit the range of movement for ROT1 (1
turn) and ROT2 (half turn)• Both rotations are provided with a semi-closed loop
device which sense the movement of the fiber in realtime. This is crucial for avoiding incidental collisions.
• Overall features: 10mm diameter, "200mm length, 75gweight
• Final design of the prototype close to be completed.
17Francisco Prada Talk B2.4 Focal Plane
SimulationSimulation ofof thethe FiberFiber PositioningPositioning Software Software
18
The Faulhaber motor and gearbox, preloaded with asteel spring was tested with a 20X microscope.
The 6 mm Faulhaber motor with the 1024:1 gearbox.
Results are scaled to the Patrol Disc radius (worst-case) of 7 mm.• Steady position shift in disable mode # 5 % of step (0.1 !)• Step homogeneity (gearbox output) = ± 47% of step (± 1 !)• Repeatability = ± 0.8 !• Residual backlash (with preload) = 0.9 !• Short run error # 5 !• Longest run error (whole turn run = 20480 steps) = 15 steps or 32 !• Average absolute positioning error < 65 !• Power Consumption < 80 mW (not measurable in sleep mode)• Life test passed (equivalent to 5 years operation)
PositionerPositioner motor tests motor tests
The Faulhaber motor+gearbox is totally suitable for the BigBOSS positioner.
19
The electronics must be constrained within theactuator envelope (10 mm). The current concept isbased on the SIDE electronics, but with a Zigbeemodulus for wireless communications and a Halleffect indexing in semi-close loop for security ofmovement (very important in case of power loss ormechanical failure or actuator collisions).
PositionerPositioner electronics electronics
Electronics board can be kept within 40x8x3.5mm,Including Hall effect indexing for both motors andtemperature sensors
Electronics & Motor Assembly Hall effect & Thermo sensor
In short, the software can properly deal with objects in common areas (shared by 2 or even 3 positioners). This isnot possible with a random assignment. A summary of the advantages is:• ~2% improvement as compared to a random assignment.• ~4.5% for slight field rotations (+/- 2°, BigBOSS) and 5-6 % for unlimited rotations or small shifts (TBC).• Fiber collisions are solved optimally.• Dynamic collisions (during fiber positioning) are minimized and avoided. For BigBOSS, several hundreds of thousands targets are gained, or, equivalently, ~350 sq deg (~50 tiles).
20
Assignment & anti-collision softwareAssignment & anti-collision software
See Morales et al. 2011
Draining Algorithm
21
ActuatorActuator R&DR&D ScopeScope
Some of the main actuator uncertainties will be mitigatedby the production and testing of a 12mm-pitchdemonstrator prototype. Thus, the actuator interface andlocking mechanism will be tested. The fulfillment of themechanical requirements and the error budget estimatewill also be revised or confirmed.
Further issues (actuator collisions, integration of the FiberFerrule into the actuators, electronics, control software,etc) will be mitigated by a further stage with theproduction and testing of a sub-system of 19 units.
Finally, appropriate cost minimization and estimate forserial production of the Actuators will be done. Simulation of the 19 actuator sub-
system mounted on the Focal Plate
22
Summary:Summary:Focal Plane R&D Work Plan 2012Focal Plane R&D Work Plan 2012
• Sub-scale plate prototype production and testing by end of 2012.• Actuator prototype produced by early Summer 2012.
Testing LCDM with BOSS/MultiDark
Nuza et al. 2012
FirstFirst BAO BAO resultsresults fromfrom BOSS! BOSS!
arXiv:submit/0446072From: David Schlegel <[email protected]>Date: Thu, 29 Mar 2012 13:11:54 EST (2799kb,D)
Title: The clustering of galaxies in the SDSS-IIIBaryon Oscillation Spectroscopic Survey: BaryonAcoustic Oscillations in the Data Release 9Spectroscopic Galaxy Sample
FirstFirst BAO BAO resultsresults fromfrom BOSS! BOSS!