ISTITUTO NAZIONALE DI FISICA NUCLEARE Preventivo per

Transcript

ISTITUTO NAZIONALE DI FISICA
NUCLEARE
Preventivo per l'anno 2005
Codice
Esperimento
AURIGA
Rapp. Naz.: M. Cerdonio
Rappresentante nazionale:
Struttura di appartenenza:
Posizione nell'I.N.F.N.:
Gruppo
2
M. Cerdonio
PD
INFORMAZIONI GENERALI
Rivelazione onde gravitazionali
Linea di ricerca
L.N.L. Legnaro
Laboratorio ove
si raccolgono i dati
AURIGA
Sigla dello
esperimento
assegnata
dal laboratorio
Acceleratore usato
Fascio
(sigla e
caratteristiche)
Onde gravitazionali
Processo fisico
studiato
Apparato
strumentale
utilizzato
massa risonante a temperatura
ultracriogenica
FE, FI, PD, LNL, TN
Sezioni partecipanti
all'esperimento
Istituto Fotonica e Nanotecnologie (IFN)
Istituzioni esterne Trento,
all'Ente partecipanti Istituto Nazionale di Ottica Applicata (INOA)
Firenze
5 anni
Durata esperimento
Mod EC. 1
(a cura del responsabile nazionale)
ISTITUTO NAZIONALE DI FISICA NUCLEARE
Struttura
TN
Codice
Esperimento
AURIGA
Resp. loc.: Giovanni PRODI
Gruppo
2
PREVENTIVO LOCALE DI SPESA PER L'ANNO 2005
In KEuro
IMPORTI
VOCI
DI
SPESA
DESCRIZIONE DELLA SPESA
Parziali
Totale Compet.
SJ
missioni a Legnaro
collaborazione con ROG e VIRGO
22,0
10,0
collaborazioni analisi dati con gruppi stranieri (IGEC, LIGO, GEO,
VIRGO−Francia)
16,0
4,0
collaborazione sviluppo amplificatori SQUID
17,0
6,0
6,0
4,0
4,0
elio liquido
manutenzione strumenti e attrezzature
sensori di vibrazione e microfoni
materiali
lavorazioni meccaniche
Ore CPU
Spazio Disco
32,0
24,0
4,0
conferenza Amaldi 6, giappone (1 partecipazione)
Consorzio
di cui SJ
Cassette
37,0
Altro
sistema acquisizione 8 canali con condizionamento segnale per sensori
vibrazione (scheda di acquisizione + pc + accessori)
8,0
3,0
controllore di temperatura per refrigeratore a diluizione
sviluppo e realizzazione di 1 K pot per il funzionamento ultracriogenico di
AURIGA
11,0
15,0
5,0
sviluppo e realizzazione degli ancoraggi termici fra il refrigeratore a diluizione
e le sospensioni criogeniche di AURIGA per il funzionamento ultracriogenico
Totale
20,0
124,0
di cui SJ
0,0
Sono previsti interventi e/o impiantistica che ricadono sotto la disciplina della legge Merloni ?
Breve descrizione dell'intervento:
Mod EC./EN. 2
(a cura del responsabile locale)
A cura della
Comm.ne
Scientifica
Nazionale
Struttura
TN
Codice
Esperimento
AURIGA
Resp. loc.: Giovanni Prodi
Gruppo
2
ALLEGATO MODELLO EC2
Mod EC./EN. 2a Pagina 1
Struttura
TN
Codice
Esperimento
AURIGA
Resp. loc.: Giovanni Prodi
Gruppo
2
Codice
Esperimento
AURIGA
NUCLEARE
Gruppo
2
PREVENTIVO GLOBALE DI SPESA PER L'ANNO 2005
In KEuro
A CARICO DELL' I.N.F.N.
Materiale
di
consumo
Struttura
Missioni
interne
FE
FI
LNL
PD
TN
10,0
3,0
4,0
30,0
32,0
4,0
4,0
6,5
29,0
24,0
5,0
60,0
49,0
37,0
TOTALI 79,0
67,5
151,0
Missioni
estere
SJ
SJ
SJ
Trasporti
e
facchinaggi
SJ
Spese
di
calcolo
Affitti
e
Materiale Costruzione
manutenzioneinventariabile apparati
SJ
SJ
1,0
13,0
14,0
SJ
SJ
2,0
1,0
27,5
55,5
11,0
95,0
20,0
35,0
97,0
115,0
35,0
A
carico
di altri
Enti
TOTALE
Compet.
SJ
17,0
13,0
111,0
258,5
124,0
0,0
0,0
0,0
35,0 3,0
15,0
523,5 35,0 18,0
NB. La colonna A carico di altri enti deve essere compilata obbligatoriamente
Mod EC./EN. 4
NUCLEARE
Codice
Esperimento
AURIGA
A) ATTIVITA' SVOLTA FINO A GIUGNO 2004
Il rivelatore AURIGA è entrato in funzione nel Dicembre 2003 (si veda
www.auriga.lnl.infn.it > status) con sensibilità e banda in accordo con
quelle attese (100 Hz per S_hh < 5 E−21 /sqrt(Hz) con barra a 4.5 K), ma
lo spettro di rumore non è soddisfacente per la presenza di alcune righe
non stazionarie nella banda di sensibilità. Durante questi mesi il
rivelatore è stato calibrato ed è stato mantenuto in funzione con finalità
principalmente diagnostiche. In particolare, si sono studiati i fenomeni di
up−conversion di eccitazioni a bassa frequenza (10−200 Hz) in rumore nella
banda di sensibilità.
La presenza delle righe spurie ha richiesto un notevole impegno sia per
estendere le funzionalità del sistema di analisi dati sia per la determinazione della
qualità dei dati. In particolare, dopo il porting del sistema di acquisizione ormai
obsoleto da Linux RedHat 6.1. su piattafoprma dualPIII a Linux RedHat 9 su
piattaforma PIV e l'upgrade di tutte le librerie utilizzate dai sistemi di analisi ed
acquisizione si e' messa a punto, nelle condizioni attuali, la caratterizzazione del
rumore intrinseco del rivelatore (termico, back−action e bianco) e della funzione di
trasferimento del segnale gravitazionale (calibrazione).
E` iniziato làllestimento di stadi aggiuntivi di sospensione meccanica esterna
del rivelatore al fine di diminuire làmpiezza delle risonanze meccaniche
nellìntervallo 10−200 Hz e rendere trascurabili i fenomeni di up−conversion.
E` iniziata làttività di studio di metodologie per la ricerca di segnali
impulsivi mediante una rete di rivelatori con sensibilità (antenna pattern,
banda, etc.) diversa, anche con metodi di una analisi coerente. Per quest'ultimo
metodo si sono iniziate simulazioni in particolare per AURIGA−VIRGO e per il
network IGEC.
E' stato allestito un sistema di controllo dei processi di acquisizione e
implementato un sistema di allarmi.
Nel corso dei primi 6 mesi dell'anno, in vista della prossima prova
completa della trasduzione optomeccanica su una barra criogenica, si
sono effettuate misure fino alle temperature di 4K del fattore di
qualità meccanico del trasduttore ottico. Le misure si sono articolate
nella caratterizzazione prima del solo corpo risonante (dopo averlo
modificato per renderlo compatible con l'operazione criogenica) e poi
del trasduttore completo: esso consiste nel corpo risonante e nella
parte rigida che supporta il resto delle ottiche necessarie e ne
permette l'allineamento. Per la prova completa della trasduzione optomeccanica
su una barra criogenica e' stata allestita i) una camera anecoica
attorno al tavolo ottico che supporta il laser stabilizzato e ii) un terzo sistema di
acquisizione dati .
Lo sviluppo di amplificatori SQUID a doppio stadio accoppiati a risonatori
a bassissime perdite ha finora dimostrato la termicità del rumore del
risonatore LC fino a 60 mK ed il comportamento in temperatura del rumore
additivo di due tipologie di sensori. Si prosegue per misurare il rumore di
back action a quelle temperature.
E' iniziato lo studio dell’utilizzo di espansioni “wavelets” per la classificazione ed
identificazione dei segnali del rivelatore, in particolare (mediante metodi
montecarlo) di falsi allarmi e falsi dismissal utilizzando wavelet packets con alberi
binari.
B) ATTIVITA' PREVISTA PER L'ANNO 2005
Gruppo
2
Una volta ottenuto un livello soddisfacente di prestazioni di rumore del
rivelatore, le priorità dellàttività 2005 del gruppo saranno:
i) mantenere in misura il rivelatore;
ii) produrre e validare i dati raccolti;
iii) ricercare segnali impulsivi mediante osservazioni congiunte con altri
rivelatori;
iv) dimostrare il funzionamento criogenico di un sistema di trasduzione
ottico installato su una barra. In particolare saranno studiate le
prestazioni di rumore complessive in funzione della temperatura.
Altri obiettivi dellàttività 2005 saranno:
− realizzazione dei componenti necessari per il funzionamento
ultracriogenico di AURIGA, in particolare una 1 K pot che produca disturbi
vibrazionali trascurabili e ancoraggi termici fra il refrigeratore a
diluizione e gli stadi di sospensione criogenica dellàntenna tali da
preservare làttenuazione dei disturbi vibrazionali. Questi sono
componeneti essenziali per raggiungere gli obiettivi finali di sensibilità
del run di AURIGA;
− affinamento dell’analisi dati, diagnostica e debugging di AURIGA: si cerchera’ di
caratterizzare nel modo piu’ robusto il rumore, classificare gli eventi, produrre
Frames per l’analisi dati coerente con altri rivelatori sia interferometrici che
risonanti (ricerca di eventi impulsivi, sorgenti periodiche e misure del fondo
stocastico).
− prova completa della trasduzione optomeccanica su una barra criogenica: le parti
necessarie (fibra ottica che trasporta il fascio stabilizzato dal banco fino alla
sezione baricentrale della barra + banchino con ottiche di adattamento su
supporti motorizzati compatibili con l'ambiente criogenico e fotodiodo
di rivelazione + trasduttore optomeccanico risonante) verranno montate
su una barra identica a quella in Auriga raffreddata alle temperature criogeniche,
compensando attivamente per i disallineamenti introdotti dalla contrazione
termica; verranno poi caratterizzate le sorgenti di rumore e la sensibilità alle onde
gravitazionali del sistema formato dalla barra criogenica dotata di trasduzione
optomeccanica; diagnostica e debugging utilizzeranno gli stessi tools di analisi
dati implementati per AURIGA.
− proseguimento dello sviluppo di sistemi di amplificazione a doppio SQUID
a temperature ultracriogeniche con diverse tipologie di sensori per
abbassare il livello di rumore ad 1 kHz.
C) FINANZIAMENTI GLOBALI AVUTI NEGLI ANNI PRECEDENTI
Anno
finanziario
Missioni
interne
1989
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
TOTALE
Mod EC. 5
Missioni
estere
Materiale di Trasporti e
consumo facchinaggi
190,5
49,0
51,6
40,2
48,0
51,6
56,2
54,7
48,0
62,5
59,5
104,8
25,8
23,2
12,9
7,7
15,4
20,6
23,8
23,5
26,0
33,5
402,3
95,5
102,2
103,2
124,4
150,8
136,8
100,7
98,5
135,0
122,0
11,8
711,8
317,2
1571,4
11,8
Spese di
calcolo
7,7
7,7
In kEuro
Affitti e
Materiale
Costruzione
manutenzione inventariabile
apparati
2,5
TOTALE
7,7
14,4
14,4
17,0
13,9
12,9
5,5
5,5
10,5
726,1
62,4
159,0
58,3
30,4
109,2
83,1
84,7
74,5
35,0
31,5
1350,5
72,3
7,7
180,7
61,9
30,4
20,6
103,3
92,5
122,5
52,5
2796,2
305,0
351,4
409,7
286,8
374,4
331,2
380,1
342,5
386,5
309,5
104,3
1454,2
2094,9
6273,3
(a cura del rappresentante nazionale)
Codice
Esperimento
AURIGA
NUCLEARE
Gruppo
2
PREVISIONE DI SPESA
Piano finanziario globale di spesa
In KEuro
ANNI
Missioni Missioni
FINANZIARI interne estere
79,0
80,0
81,0
82,0
83,0
2005
2006
2007
2008
2009
TOTALI
67,5
50,0
60,0
50,0
65,0
405,0 292,5
Mod EC./EN. 6
Spese
Materiale
Affitti e
Trasporti e
di
di
manutenzione inventariabile apparati
facchinaggi
calcolo
consumo
150,0
97,0
14,0
151,0
50,0
40,0
14,5
150,0
45,0
40,0
15,0
150,0
40,0
30,0
15,5
150,0
10,0
10,0
16,0
150,0
751,0
0,0
0,0
75,0
217,0
295,0
TOTALE
Compet.
558,5
384,5
391,0
367,5
334,0
2035,5
Struttura
TN
Codice
Esperimento
AURIGA
Resp. loc.: Giovanni PRODI
Gruppo
2
COMPOSIZIONE DEL GRUPPO DI RICERCA
N
1
2
3
4
5
6
7
8
RICERCATORE
Cognome e Nome
Qualifica
Dipendenti
Incarichi
Affer.
al
gruppo
.
Art.
23
Ruolo
Ricerca Assoc
BAGGIO Lucio
BONALDI Michele
FALFERI Paolo
MION Alessandro
POGGI Silvia
PRODI Giovanni Andrea
VINANTE Andrea
VITALE Stefano
AsRic
Ric.
Ric.
Dott.
Dott.
P.A.
AsRic
P.O.
2
2
2
2
2
2
2
2
%
100
40
50
100
100
70
80
15
N
TECNOLOGI
Cognome e Nome
1 MEZZENA Renato
Qualifica
Incarichi %
Ass.
Ruolo Art. 23
Tecnol.
Univ.
40
Dipendenti
Numero totale dei Tecnologi
Tecnologi Full Time Equivalent
N
TECNICI
Cognome e Nome
Qualifica
Incarichi
Dipendenti
Collab.
Ruolo Art. 15
tecnica
1 GENNARA Pierino
2 GOTTARDI Fabrizio
3 SALOMON Claudio
Numero totale dei ricercatori
Ricercatori Full Time Equivalent
Univ.
Univ.
Univ.
Annotazioni:
mesi−uomo
Osservazioni del direttore della struttura in merito alla
disponibilità di personale e attrezzature
Mod EC./EN. 7
%
Assoc.
tecnica
8 Numero totale dei Tecnici
5.55 Tecnici Full Time Equivalent
SERVIZI TECNICI
Denominazione
1
0.4
10
10
10
3
0.3
Codice
Esperimento
AURIGA
NUCLEARE
Gruppo
2
MILESTONES PROPOSTE PER IL 2005
Data
completamento
Descrizione
30/06/2005
test completo a 4.5 K del trasduttore optomeccanico su una barra identica ad AURIGA
31/12/2005
10 mesi di acquisizione dati a sensibilta' impulsiva di almeno 0.1 mK e con banda utile di
almeno 50 Hz
30/10/2005
realizzazione componenti critici per funzionamento ultracriogenico di AURIGA: 1 K pot con
rumorosita' trascurabile e ancoraggi termici verso le sospensioni criogeniche
31/12/2005
risultati osservativi di ricerca congiunta di eventi impulsivi
31/07/2005
amplificazione a doppio SQUID a circa 15 h−bar a 1KHz
Mod EC./EN. 8
NUCLEARE
Codice
Esperimento
Gruppo
DUAL−RD
2
Rapp. Naz.: Massimo Cerdonio
Massimo Cerdonio
PD
Onde gravitazionali
Linea di ricerca
Fi, Na, L.N.L., Trento
Laboratorio ove
Sigla dello
esperimento
assegnata
dal laboratorio
Acceleratore usato
Fascio
(sigla e
caratteristiche)
Processo fisico
studiato
Prestazioni di rivelatori acustici di onde
gravitazionali a doppia massa sensibile
Apparato
strumentale
utilizzato
Fi, LNL, Na, Pd, Gruppo Collegato di Trento,
all'esperimento
Sezione di Trento Istituto Fotonica e
Istituzioni esterne Nanotecnologie (ITC−CNR)
all'Ente partecipanti Dipartimenti di Fisica di Padova e Trento
3 anni
Durata esperimento
Mod EC. 1
Struttura
TN
Codice
Esperimento
DUAL−RD
Resp. loc.: Michele Bonaldi
Gruppo
2
In KEuro
IMPORTI
VOCI
DI
SPESA
Parziali
Totale Compet.
SJ
Collegamento con le altre sezioni
di cui SJ
5,0
5,0
Collegamento con Glasgow per sviluppo test mass
Collegamento con Giessen per sviluppo SQUID
2,0
3,0
Materiali
Elio liquido
3,0
2,0
5,0
5,0
Consorzio
Ore CPU
Spazio Disco
Cassette
Altro
Generatore di funzioni ad alta tensione
Pompa rotativa a secco
Vasche ultrasuoni pulizia condensatori
18,0
2,0
5,0
Lavorazioni meccaniche camera test capacità
5,0
25,0
5,0
Totale
45,0
di cui SJ
0,0
Mod EC./EN. 2
A cura della
Comm.ne
Scientifica
Nazionale
Struttura
TN
Codice
Esperimento
DUAL−RD
Gruppo
2
Struttura
TN
Codice
Esperimento
DUAL−RD
Gruppo
2
Codice
Esperimento
Gruppo
DUAL−RD
2
NUCLEARE
In KEuro
Struttura
Missioni
interne
FI
LNL
NA
PD
TN
11,0
1,0
5,0
3,0
5,0
TOTALI 25,0
Materiale
di
consumo
Missioni
estere
SJ
SJ
SJ
Trasporti
e
facchinaggi
SJ
Spese
di
calcolo
Affitti
e
Materiale Costruzione TOTALE
Compet.
SJ
SJ
SJ
SJ
17,5
7,0
5,0
5,0
26,5
5,0
10,0
5,0
5,0
15,0
23,0
25,0
10,0
51,5
80,5
SJ
10,0
9,0
5,0
62,0
6,0
40,0
45,0
45,0
30,0
0,0
0,0
30,0
12,0
31,0
198,0
72,0
Mod EC./EN. 4
A
carico
di altri
Enti
Codice
Esperimento
Gruppo
DUAL−RD
2
NUCLEARE
• studio di un amplificatore meccanico con guadagno pari a 10 dotato di trasduzione optomeccanica. Modellizzazione FEM tramite il
software commerciale ANSYS del rumore termico di un oscillatore meccanico lontano dalla risonanza, prima e dopo il suo accoppiamento
a un amplificatore meccanico.
• progettazione di un oscillatore meccanico dotato di amplificatore meccanico a larga banda tale da poter misurare l'effetto
dell'amplificatore sul rumore termico per mezzo di trasduzione optomeccanica ai LNL.
Sezione di Firenze − suspension development
• Study and preliminary dimensioning of a passive suspension for a dummy payload.
• Set up of the FEM tool (ANSYS) for modelling the passive suspension: definition of the parts that can be parametrized.
• Operation of (already realized) standard prototypes of inertial sensors and their characterization.
Sezione di Firenze − optic development
• Improvement of the frequency stabilization of the laser, with a goal of 109 Hz^2/Hz in the range 1−5 kHz.
• Extending the amplitude stabilization of the laser in the range 1−5 kHz.
• Implementation of wide reading area optical cavities.
Sezione di Napoli
• Definition and preliminary implementation of the necessary changes on the Napoli suspension chain to perform the tests for the R−D.
• Design and preliminary tests of the digital control system hardware and software architecture to be used for the first tests of the
suspended mass control.
• Characterization of the inertial sensors prototypes used in the active control of the suspension, to evaluate their sensitivity and dynamics
in connection with the scientific requirements and the architecture of the control system.
• Definition and characterization of the conditioning electronics used to interface the signals from sensors to the digital control system.
Sezione di Padova − Laboratori Legnaro
• study of wideband mechanical amplifiers: leverage, velocity transformer.
• Construction of a vacuum chamber for the room temperature thermal noise measurement of the mechanical oscillator equipped with
broadband mechanical amplifier: passive thermal insulation and mechanical suspension.
• study of impedance matching networks optimized for the capacitive readout.
Gruppo Collegato di Trento
• Thermal noise modelling of mechanical amplifier in order to evaluate and control its role in the detector.
• Study of the noise of electro−mechanical devices coupled to the detector: define requirements on the SQUID−based readout so to
guarantee at the detector the same energy sensitivity as in bench tests of the SQUID amplifier.
Anno
finanziario
Missioni
interne
Missioni
estere
consumo facchinaggi
Spese di
calcolo
In kEuro
Affitti e
Materiale
Costruzione
apparati
TOTALE
2004
TOTALE
Mod EC. 5
DUAL R&D
feasibility study for a
Wideband Resonant Gravitational Wave Detector
- Research Proposal June 2004
Contents
1 Introduction
3
2 Scientific relevance
2.1 Expected sources in the high frequency range . . . . . . . . . . . . . . .
2.2 The European scene . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
4
4
3 Dual detector theory
3.1 One dimensional model: Standard Quantum Limit
duction . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Full modal expansion and selective readout . . . . .
3.3 Expected sensitivity at the SQL . . . . . . . . . . .
6
and back-action re. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
6
7
8
4 R&D goals and methods
10
4.1 Detector design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2 Readout system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.3 Test mass development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5 Research status
13
5.1 Mechanical amplification . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
6 References
16
7 Qualifications of the proponents
18
7.1 Functional chart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
7.2 Participants (from 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
7.3 Selected publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1
8 Research program
8.1 2005 . . . . . .
8.2 2006 . . . . . .
8.3 2007 . . . . . .
8.4 Instrumentation
. . . . . . .
. . . . . . .
. . . . . . .
available for
. . . . . . . .
. . . . . . . .
. . . . . . . .
the research
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
24
24
25
26
28
9 Detailed funding request (in Italiano)
29
10 Other fundings
32
2
1
Introduction
Substantial progress has been made over the last 40 years in preparing instruments and
methods to search for gravitational waves from the universe. Theoretical astrophysics
and numerical relativity focus on candidate sources, whose signal will be efficiently
extracted by the dominating noise with analysis methods developed ad-hoc. Resonant
mass ”bar” detectors, the first historically to come to continuous operation, have been
improved by 4 orders of magnitude in energy sensitivity, so that they can detect energy
changes of a 2300 kg bar as little as a few hundred of quanta of vibration at about
1kHz with a bandwidth of 100 Hz [1]. The 5 bar detectors distributed worldwide have
operated for a few years as a network, giving for the first time significant upper limits
to the yearly rate of violent Gravitational Waves (GW) events in the Galaxy [2, 3]. A
first generation of long baseline interferometric detectors is now operating or coming
into operation [4, 5, 6] and, complemented by the upgraded bars, may well give a first
detection in the next few years.
However it is commonly accepted in the community that, to enter the ”observatory
phase” and open up a new GW astronomy, a substantial improvement in detector sensitivities should be achieved. Traditional acoustic detectors are seriously limited in terms
of bandwidth (typically to about 10% of the resonant frequency) due to the usage of
the resonant transducer, which is needed to reduce the effect of the noise of the final
amplifier. Bandwidth enhancements can be obtained by the use of a two-modes (mechanical + electrical) resonant readout [1], but a different approach is needed to fully
exploit the potential sensitivity of resonant detectors and make them complementary to
the advanced versions of interferometric detectors. We are actively investigating a novel
detection scheme, the ”dual” resonator system [7, 8, 9], which can provide both high
sensitivity and wide bandwidth.
Given the interest of the novel ”dual” resonator GW detector, we propose here to perform a complete study of the critical issues:
• Detector design: limit sensitivity, noise control, underground operation
• Readout system: geometry and noise to the quantum limit
• Test mass development: choice of material and fabrication method
The R&D we propose here would have a duration of 3 years and its result will be the
demonstration at prototype level of the relevant technologies needed for the readout
realization. At the same time the groups involved in this proposal will also complete
the modelling of dual detector configurations, the optimization of their expected performances and the establishment of the main technological requirements needed for their
realization. Then at the end of this project it will be possible to establish the guidelines for the detailed design of a dual detectors with optimal sensitivity and very wide
bandwidth.
3
2
2.1
Scientific relevance
Expected sources in the high frequency range
We support the common idea that a complete observatory, sensitive in a wide frequency
range, will permit to follow the star system evolution during a long time interval, improving the signal reconstruction and the detection probabilities. For this reason collaborations are ongoing to set up the multi-window observation concept, i.e. the linking
between different GW detectors (acoustic detectors, interferometers) and all the other
astronomical telescopes and particle detectors. In particular the Dual detector covers
the high frequency part (1 - 7 kHz) of the GW spectrum, and could contribute to the
detection of the following GW sources:
Binary systems of coalescing neutron stars or black holes. The system inspiral
will produce a chirping signal from the lowest frequency of the detector network up
to about 1 kHz. The merger event is still not well understood, but the ringdown
will be in the range of 1 - 10 kHz for reasonable remnant black hole masses and
spins. Because the signal for the chirp will be quite strong, these events should be
visible well outside the Virgo cluster. Thus, the large volume of space allows for a
reasonable event rate in the detector.
Perturbed relativistic stars. Newly born Neutron Stars undergo bars instabilities
or oscillate in their quasi-normal modes (such as f-mode and r-mode), and emit
gravitational waves in the kHz range. The properties of these signals depend
crucially on the dynamical properties of the stars and therefore on their structure,
composition and equation of state. In particular insight could be made on the
precise equation of state of matter at supernuclear density.
Low mass x-rays binaries. LMXBs have a constant supply of energy from the accretion. The GW frequency will be twice the spin frequency of the neutron star. The
signal will be modified due to Doppler shifting from the neutron star in its orbit
as well as possible changes in the neutron star spin due to the accretion. Thus,
the signal is ”dirty” and may be difficult to extract, but the number of known
potential sources is about 100, and the position and frequency are reasonably well
known.
Supernovae and GRBs as impulsive sources. This rare events spread the energy
transported by the GW in a large frequency band and their detection is optimized
by a very broad band network of detectors with different detection methods and
frequency ranges.
2.2
The European scene
Currently, several gravitational wave detectors are active in Europe. There are three
cryogenic resonant bar detectors (AURIGA, EXPLORER and NAUTILUS) that are
managed by INFN in Italy and at CERN. The GEO interferometric detector is operating close to Hannover (Germany-UK collaboration) and the French-Italian VIRGO
collaboration is quickly bringing the interferometer located close to Pisa (Italy) to an
4
operational level. Almost all these detectors have been designed to be sensitive in a well
limited frequency range and this is limiting their capability to fully characterize the GW
signal and to cover all possible kind of sources. The VIRGO detector has been designed
to show the largest frequency band of sensitivity, but in any case it is optimized for the
low frequency range (10-1000 Hz).
Figure 1: Sensitivity of the proposed EGO observatory compared with some current
and proposed detectors. In the high frequency range the Dual detector physical limit is
competitive and complementary with the foreseen limits of next generation of Interferometric detectors. (Excerpt from the EGO Observatory proposal [10].)
Several European organizations have manifested interest in a future, wide band GW
observatory. A common proposal document ”European Gravitational waves advanced
Observatory design” has been produced [10], and testifies the whole community agreement toward the design of an European observatory. This new infrastructure will be
competitive with respect to the detectors that in other countries are under consideration
(LIGO2 and LIGO3 in USA, LCGT in Japan). In fact the observatory, formed by several
detectors dedicated to specific frequency ranges and operating as a single instrument,
will permit to cover a large fraction of the possible gravitational wave sources, ranging
from few Hz up to ten kHz (Fig. 1). In this framework the DUAL detector represents a
promising detector covering the high frequency range of the GW spectrum, and one of
the specific goals of the proposal is to exploit the bandwidth complementarity between
Dual and Interferometric detectors.
The technology needed to realize such large band GW observatory is partially under
study with several R&D activitiesin the different countries. The European Community
supports in part this research under the ILIAS consortium (STREGA activity). We
participate to this project with an R&D on thermal noise reduction in GW detectors.
5
3
Dual detector theory
A dual detector is formed by two nested massive bodies whose quadrupolar modes (ie the
modes sensitive to the GW signal) resonate at different frequencies. In ref. [7] the two
resonators are two spheres, an inner full one and a hollow outer one, but a dual cylinder
configuration can also be evaluated [9] (Fig. 2). In a dual detector the signal is read in
the gap between the two masses as their differential deformations: the centers of mass
of the two bodies coincide and remain mutually at rest while the masses resonate, thus
providing for the rest frame of the measurement. The sensitivity of the dual detector is
predicted to be of interest in the frequency range between the first quadrupolar modes
of the two masses. This can be as broad as a few kHz in the kHz range.
Figure 2: Different Dual detector configurations.
3.1
One dimensional model: Standard Quantum Limit and
back-action reduction
A basic dual detector can be represented as the simple one dimensional system shown in
Fig. 3(a) formed by two different and independent mechanical oscillators whose positions
are x1 and x2 ; a force Fe driving the two oscillators is evaluated by measuring the
difference xd = x1 − x2 . In the frequency region between the resonant frequencies of
the two oscillators, Fe drives the slow resonator above its resonance νs and the fast one
below its resonance νf . The responses of the two are thus out of phase by π radians
and therefore the differential motion xd results in a signal enhancement with respect
to the response of the single oscillators [Fig. 3(b)], as shown by the transfer function
HFe = xd /Fe [Fig. 3(c)].
In addition such a scheme leads to a reduction of the back-action noise within the
frequency range of sensitivity of the detector. In fact the back-action noise force drives
coherently with opposite direction the two oscillators: the system response is nearly
exactly in phase and the consequent differential displacement is highly depressed at a
frequency f ∗∗ , as shown in the transfer function Hba = xd /Fba [Fig. 3(c)]. The features
of the back-action reduction are discussed in more details in ref. [8].
6
Figure 3: (a) One-dimensional “dual” detector: the same force Fe is measured
by the relative displacement xd of two resonators. (b) Transfer functions of the
slow resonator (continuous line) and of the fast resonator (dashed line). (c) Dual
detector transfer functions: signal HFe = xd /Fe (continuous line), back-action
Hba = xd /Fba (dashed line). (d) Wideband optimized noise.
We measure xd with a real displacement amplifier, described as an ideal amplifier with
an additive displacement noise generator and force noise generator that back-acts on
the system under measurement. Let SXX (ω) = Sxx and SF F (ω) = Sf f be respectively
the frequency-independent power spectral densities of these noise generators. The total
displacement noise due to the amplifier is then Sxx + |Hba (ω)|2 Sf f and the noise power
spectrum on the measurement of Fe is (see for instance ref. [11]):
SFe (ω) = Sxx + |Hba (ω)|2 Sf f /|HFe (ω)|2 .
(1)
If we take as reference an operation at the Standard Quantum Limit (SQL), we may
consider Sxx Sf f = ~2 , and the noise figure can be optimized by adjusting the ratio
Sxx /Sf f . In a wide bandwidth detection strategy Sxx and Sf f must be balanced to give
the lowest noise within the bandwidth. In doing so we profit by the subtraction effect
in the back-action noise transfer function Hba , and obtain a dip at the frequency f ∗∗ ,
as shown in Fig. 3(d). We finally notice that, to fully exploit the back-action reduction
features, f ∗∗ should be placed, by a proper choice of the system parameters, amid the
oscillator frequencies.
3.2
Full modal expansion and selective readout
In the case of a three-dimensional body, the dynamics of elastic deformations is given
as the superposition of the dynamics of an almost infinite number of normal modes of
7
vibration [12]. An obvious way to preserve the convenient features for signal and back
action noise outlined above, is to bring the real system to be as close as possible to
the idealized two modes system. In fact when only the first quadrupolar mode can be
considered for each body, the response to a GW of such a system can again be described
by the simple one-dimensional model. This can be accomplished by a selective readout,
capable of rejecting a large number of normal modes on the basis of their symmetry.
This geometrically based mode selection senses the surface position of a body on specific
regions, so that the related deformations can be combined with a weight/sign properly
chosen to optimize the total response to normal modes of quadrupolar symmetry.
As an example we can apply this concepts to a dual cylinder detector [Fig. 4(a)], a
convenient geometry where the advantages of the proposed scheme can be fully exploited.
In the dual cylinder we average the differential displacement in 4 distinct areas x1..4
[Fig. 4(b)] and combine them to obtain Xd = x1 + x3 − x2 − x4 . The detector displays
its maximal sensitivity when the GW propagates along the z axis, the symmetry axis
of the system.
Such a strategy to select specific vibrational modes and to reject a class of unwanted
modes is conceptually different from the strategy now employed in GW acoustic detectors. Resonant bar detectors [2, 3] and spherical detectors [13, 14, 15] reconstruct the
amplitude of the normal modes excited by the GWs by the use of resonant displacement
readouts coupled to the modes. With a proper choice of the readout surfaces, the resonant scheme is not sensitive to the thermal noise of out of resonance modes and gives
an efficient frequency based mode selection. However this feature necessarily limits the
detector bandwidth, due to the thermal noise contribution of the resonant transducer,
as mentioned above. By contrast the geometrically based mode selection selects gravitationally sensitive normal modes by means of their geometrical characteristics, and
shows a significant rejection of non quadrupolar modes without affecting the detector
bandwidth. For this reason it can be effectively applied to a “dual” wideband detector.
3.3
Expected sensitivity at the SQL
In close analogy with Eq. (1), the detector sensitivity to GW is obtained by its mechanical transfer functions Hba (ω) and Hgw (ω) as:
Shh (ω) = Sxx + |Hba (ω)|2 Sf f /|Hgw (ω)|2 .
(2)
As discussed above, Sxx and Sf f are the frequency-independent power spectral densities
that define the amplifier noise performances. For operation at the SQL we consider
Sxx Sf f = ~2 and optimize the noise figure by adjusting the ratio Sxx /Sf f .
To show the limits of our design, we evaluate the sensitivity at the SQL of some practical
configurations of detector material and geometry. As usual a low dissipation material is
required to reduce the effect of the thermal noise. Molybdenum represents an interesting
choice, as it shows high cross-section for GWs and its mechanical dissipation was investigated at low temperature giving Q/T > 2×108 K−1 for acoustic normal modes [16]. In
Fig. 5 is shown the SQL sensitivity of a Mo detector with dimensions within the present
technological production capabilities. In Fig. 5 we also show the SQL of a detector made
of SiC, a ceramic material currently used to produce large mirrors or structures [17],
with mechanical and thermal properties of interest here but not yet characterized in
8
Figure 4: (a) The two concentric cylinders, Ce , Ci are made of materials of
density ρe , ρi and have the same height. The inner cylinder may also have null
internal radius. The relative distance between the two bodies is measured in 4
regions (in black), each of area ST , for the whole cylinder height. (b) Section of
the detector showing the signal enhancement obtained when a GW signal drives
the external cylinder above resonance and the internal cylinder below resonance.
The difference Xd = x1 + x3 − x2 − x4 , proportional to the GW strength, is not
dependent on a number of non GW sensitive modes.
Figure 5: Predicted spectral strain SQL sensitivities of two different dual detector
configurations. The predicted SQL sensitivities of two advanced interferometric
detectors are also shown (dotted lines): LCGT, Advanced LIGO, and a narrow
band design of Advanced LIGO. Continuous line: Mo dual detector, inner cylinder
radius 0.25 m, outer cylinder internal-external radius 0.26 - 0.47 m, height 2.35 m,
weight 4.8 + 11.6 tons, fundamental quadrupolar modes 5189 Hz and 1012 Hz,
amplifier noise Sxx = 10 × 10−46 m2 /Hz, Q/T > 2 × 108 K−1 . Dashed line: SiC
detector, inner cylinder radius 0.82 m, outer cylinder internal-external radius 0.83 1.44 m, height 3 m, weight 20.5 - 41.7 tons, Sxx = 6×10−46 m2 /Hz, Q/T > 2×108
K−1 .
9
terms of low temperature mechanical dissipation. More aggressive configurations could
be designed if we can drop the assumption of flat noise spectral densities and optimize
at each frequency the ratio between displacement and force noise.
4
4.1
R&D goals and methods
Detector design
The full detector design must address, with a multidisciplinary approach, the challenging
problem of the noise control and the cooling down to ultracryogenic temperatures of test
masses in the range of 50-100 tons. The detector cryostat positioning on ground, the
basement and the prefiltering stage must be defined, as well as the inner suspension
filtering chain.
Seismic noise control. Low frequency (100 mHz - some 100 Hz) seismic vibrations
can induce acoustic emission in the passive insulation chain components, with the
possibility of giving rise to up-conversion phenomena. Moreover any low frequency
differential motion between the two masses constituting Dual must be limited to
allow the proper working of the readout. The actual requirements on the low
frequency motion on the DUAL test masses will be determined at the beginning
of the R&D; preliminary estimations give an upper limit of the order of 10−10
m. The performances of low frequency pre-insulation stages will be evaluated
in the case of heavy masses (50-100 tons). The sensor, the passive insulation
and the active control will be defined during the project. The low temperature
performances of the selected sensor will be measured in the LNL facility.
Underground operation. It is expected that noise induced by cosmic rays will be
relevant at the sensitivities predicted for Dual. Therefore it is important to assess
the compatibility and requirements for underground operation, also in relation to
the environmental noise expected therein.
4.2
Readout system
The non-resonant readouts for a ”dual” detector are evolutions both conceptually and
in technology of the resonant readouts used in the bar detectors. In order to achieve the
wide frequency interval of high sensitivity, order of many kHz, the differential deformation between the facing surfaces of the nested bodies needs to be measured by readouts
that are not in mechanical resonance. Another relevant requirement for the transducer
system is to sense the deformation of the resonant masses on a wide surface, in order
to be less sensitive to the resonant modes of higher frequency, which do not carry any
gravitational signal. In this way the thermal noise of the detector is minimized, while
preserving the sensitivity to the signal. A further progress in this noise reduction can
be achieved by a readout scheme which is geometrically selective to the fundamental
quadrupolar modes, as that shown in Figure 4. This selectivity also allows to clean the
bandwidth from the spurious modes not sensitive to gravitational waves [9].
The dual cylinder sensitivity curves are optimized with
√ a quantum limited readout with
−23
displacement sensitivity of the order of 3 × 10
m/ Hz. This figure is impressive and
10
indeed has not been achieved so far, and we stress that a wide-band readout cannot
profit from the displacement amplification at resonance of a conventional transducers.
Up to now, √
the lowest displacement noise we have achieved experimentally is about
−20
5 × 10 m/ Hz in the kHz range, demonstrated by the capacitive readout based on
SQUID amplifiers [18]. Concerning the optical readout based on Fabry-Perot cavities
that we are developing,
√ sensitivity at room temperature
√ the demonstrated displacement
−20
−18
is around 10 m/ Hz [19], while 3 × 10 m/ Hz are planned for the cryogenic
readout system designed for the AURIGA detector. The properties of these transducer
systems are very different and both of them show peculiar advantages and could be
profitably implemented in different configurations of dual detectors. We discuss now
the advances over the present technology which must be achieved to allow their use as
readout in a dual detector.
Transduction efficiency. The efficiency of the signal conversion by the transducer, is
proportional to the bias field and the square root of the capacitance in the case
of the capacitive transducer, and correspondingly to the square root of the light
power injected in the cavity and to the finesse of the cavity in the case of the
optical transducer.
For the capacitive transducer we aim at increasing by at least an order of magnitude (to achieve at least 100MV/m [21]) the static bias electric field between the
plates.
For the optical transducer, we will increase the light power by a factor of 1000, up
to about 1W, and the finesse up to 106 . Moreover, it will be necessary to improve
the active frequency stabilization (to better than 10−9 Hz 2 /Hz) and to implement
a different readout configuration, where one measures the difference in the length
of two close cavities both settled on the detector. Conceiving a method for tuning
the cavity without introducing extra noise is a challenging task.
The geometry - extension and symmetry - of the readout surface. The capability of implementing a quantum limited readout requires to extend the area
where the displacement is averaged over. In fact, transducers with a small surface
’see’ many acoustic modes of the detector, up to high frequencies. Each mode has
its own thermal noise and is excited by the back action noise. It is thus useful to
extend the sensitive region, averaging out high order modes. In this way, we can
approach a configuration with only two effective modes, i.e. the two lowest order
modes, sensitive to the gravitational wave and necessary for wide-band selective
detection [8, 9]. Moreover the reduction of thermal noise makes easier to meet the
requirements on the operating temperature and on the mechanical losses of the
detector. A second major step is the development of a specific readout geometric
design that is sensitive to the deformations that have quadrupolar symmetry, thus
reducing the response of those with different symmetries that are not sensitive to
the gravitational radiation.
For the capacitive transducer a conceptual configuration that realizes this readout
selectivity has already been proposed [9]. In this case the selectivity allows to clean
the bandwidth from spurious modes and makes possible an effective back-action
reduction. As for the optical readout, we have proposed a new cavity (Folded
11
Fabry-Perot [22]) which allows to extend the effective waist size. Such cavities has
not yet been built and its implementation and test is one of the purposes of this
research project.
Noise to the quantum limit. Standard quantum limited performances of the detector can be reached only if the readout system itself is at the quantum limit. As
for the capacitive transducer, one of the crucial improvements with respect to the
state-of-the-art technology regards the energy resolution of the SQUID amplifier.
Up to now the best performance achieved in a setup that can be implemented in
a bar detector corresponds to about 200 hbar [23], but an energy resolution of
about 10 quanta in SQUID amplifiers without input load have been recently obtained by the proponents of this project [24] and energy resolutions approaching
the quantum limit have been obtained, in different experimental configurations
and operation frequencies, by other groups [25, 26, 27]. Therefore, the main limit
to the realization of a practical quantum limited amplifier doesn’t seem to be the
production of the SQUID itself but the preservation of its noise performances even
with the resonant and spatially extended load represented by the detector transducer. Medium-term improvements should come from the reduction of the 1/f
noise contribution and from the suppression of the ”hot electron effect” [28] down
to ultracryogenic temperatures by means of corrections of the design of a SQUID
chip already tested. The final step towards the quantum limited SQUID amplifier requires the implementation of a SQUID with new design and new impedance
matching stage.
As for the optical transducer, we have already achieved a shot-noise limited displacement sensitivity, i.e., a quantum-limited displacement noise, for a laser power
of some mW. Such result must be extended to higher power, requiring a careful
choice of the materials and optical devices and properly built detectors. Concerning the force noise, we remark that the mechanical effect of quantum noise in the
radiation pressure has never been observed. Indeed, in present experiments such
kind of fluctuations are always overwhelmed by thermal (or technical) noise. A few
groups are working on table-top experiments in order to demonstrate mechanical
quantum effects of the light, and reaching the force noise quantum limit is the
goal of all the proposed advanced GW detectors. However, a strong experimental
effort is necessary in this direction. Major progresses are necessary also regard the
cryogenic operation.
The needs for a mechanical amplifier The improvements described in the previous
√
section could allow to reach, in the near future, sensitivities of a few 10−22 m/√Hz
for both capacitive and optical readouts. Then to achieve the needed 10−23 m/ Hz
displacement sensitivity range, it may be necessary to develop an alternative and
non-resonant device to amplify the differential deformation of the massive bodies
[29]. In particular a device with gain 10 on a bandwidth of about 5 kHz in the
range 0-10 kHz should be used as a first amplification stage.
Mechanical amplifiers based on the elastic deformation of monolithic devices compliant mechanisms - are well known for their applications in mechanical engineering [30]. Their application to GW detectors seems promising but the contributed noise needs to be investigated thoroughly. We notice that a mechanical
12
amplification will also work as a mechanical impedance matching stage, since it
affects the balance force-displacement and in particular the back-action forces due
to the readout. This feature could be also helpful to fit the detector mechanical
impedance to the noise impedance of the amplifier, in order to obtain the so called
”noise matching” condition and to optimize the signal to noise of the system. In
the case of optical readout, mechanical amplifiers would also allow to reduce the
laser power used in the detector cavity, thus helping the cryogenic operation.
4.3
Test mass development
√
To reach interesting sensitivities (better than 10−23 / Hz) the use of test masses in the
range 50 - 100 tons is required. The material must have high cross section for the GW
and at the same time should have low mechanical losses as well as large enough thermal
conductivity to allow cooling to cryogenic temperatures. The thermal noise contribution
is in fact evaluated as negligible in respect to the amplifier Standard Quantum Limit
at ratios Q/T ≥ 5 × 108 K−1 . The thermal and mechanical properties of the selected
material will set the constrains to the detector cryostat design.
Large masses production methods. Possible candidate materials are the same currently considered for the production of the interferometers test masses, that is
Sapphire, SiC and Si. Hydroxide-catalysis bonding is the most promising method
for materials which are interesting for their low mechanical losses at low temperature, but other techniques will also be investigated, as hot hydrostatic pressing
and fiber infiltration (for C/SiC only).
Material properties. Measurements of the low temperature properties of material
samples will be performed in the Legnaro ultracryogenic test facility. Quality
factor and possibly thermal noise will be measured at low temperatures. Rough
estimates on thermal conductivity and thermal capacity will be obtained by the
sample base temperature. On the basis of the material properties the cooling and
thermal link requirements will be evaluated.
Inter mass suspension. Following the requirements on the low frequency motion of
the dual test masses, an inter masses suspension will be studied that allows the
use of a mechanical amplifier.
5
Research status
Some of the proposed topics are currently under theoretical investigation study by a
small number of researchers.
5.1
Mechanical amplification
Our design goal is a mechanical leverage able to perform at least a gain factor of 10
on the displacement in a wide bandwidth. The proposed leverage features a rhombic
structure with localized rigid hinges (see Fig. 6): the variation on its length D represents
13
Figure 6: Leverage prototype: the variations on its length D represent the
displacement signal and are converted in variations on the distance L. The
displacement L can be measured by an optical readout, and two mirror
holders are placed on to the amplifier device to make the optical cavity.
the displacement signal and are converted in variations on the distance L. The amplifier
gain factor depends on the frequency of the displacement signal and was evaluated
within a modelling software framework (Pro/Engineer - Pro/Mechanica). In our model
the static gain is essentially maintained up to the frequency of the internal resonance
modes of the device. At higher frequencies the system stiffness is reduced and the
leverage gain is consequently spoiled off. The displacement gain factor is α ' 10 and
the structure is free from internal resonance up to about 1.5 kHz. We are aware that the
localized rigid hinges could be a big source of thermoelastic noise if used in a sensitive
detector and then further evaluation of this noise component will be necessary.
In order to establish the amplifier thermal noise contribution in a resonant detector, we
designed a system made of our displacement amplifier nested with an hollow mechanical
test oscillator (see Fig. 7). The input variable is the test oscillator length D, that should
be measured as an amplified distance at the leverage output L. We plan to measure the
relevant lengths D and L by the use of a displacement optical readout. For this sake
two mirror holders are placed on to the amplifier device to make an optical cavity. The
Pound-Drever signal of the Fabry-Perot cavity inside the amplifier will then be compared
with the one of a reference frequency stabilized Fabry-Perot cavity. In the same way
two mirrors can be placed on the test resonator and its length D can be measured by
14
Figure 7: The overall assembly of the testing oscillator (dark gray) with the
inner mechanical amplifier (gray). Two laser beams are shown and drive
two orthogonal optical cavities which allow the measurement of the relevant
quantities: the test resonator length D variations and the corresponding
variations in the length L of the mechanical amplifier. In principle the two
cavities can be used to measure in the same time the two distance to obtain
the system transfer function, but for the moment we only plan to separately
measure the mechanical thermal noise on these quantities.
a second FP optical cavity. To reduce the vibrational noise the test resonator will be
suspended by a mechanical isolator inside a vacuum chamber. The full system will be
housed in a thermally stabilized box.
15
6
References
[1] http://www.auriga.lnl.infn.it
[2] Z.A. Allen, P. Astone, L. Baggio, D. Busby, M. Bassan, D.G. Blair, M. Bonaldi,
P. Bonifazi, P. Carelli, M. Cerdonio, E. Coccia, L. Conti, C. Cosmelli, Visconti
VC, S. D’Antonio, V. Fafone, P. Falferi, P. Fortini, S. Frasca, W.O. Hamilton,
I.S. Heng, E.N. Ivanov, W.W Johnson, M. Kingham, C.R. Locke, A. Marini, V.
Martinucci, E. Mauceli, M. P. McHugh, R. Mezzena, Y. Minenkov, I. Modena,
G. Modestino, A. Moleti, A. Ortolan, G. V. Pallottino, G. Pizzella, G. A. Prodi,
E. Rocco, F. Ronga, F. Salemi, G. Santostasi, L. Taffarello, R. Terenzi, M. E.
Tobar, G. Vedovato, A. Vinante, M. Visco, S. Vitale, L. Votano, J.P. Zendri
Phys. Rev. Lett. 85, 5046 (2000).
[3] P. Astone, D. Babusci, L. Baggio, M. Bassan, D. G. Blair, M. Bonaldi, P. Bonifazi, D. Busby, P. Carelli, M. Cerdonio, E. Coccia, L. Conti, C. Cosmelli, S.
D’Antonio, V. Fafone, P. Falferi, P. Fortini, S. Frasca, G. Giordano, W. O.
Hamilton, I. S. Heng, E. N. Ivanov, W. W. Johnson, A. Marini, E. Mauceli, M.
P. McHugh, R. Mezzena, Y. Minenkov, I. Modena, G. Modestino, A. Moleti, A.
Ortolan, G. V. Pallottino, G. Pizzella, G. A. Prodi, L. Quintieri, A. Rocchi, E.
Rocco, F. Ronga, F. Salemi, G. Santostasi, L. Taffarello, R. Terenzi, M. E. Tobar, G. Torrioli, G. Vedovato, A. Vinante, M. Visco, S. Vitale, and J. P. Zendri
Phys. Rev. D 68, 022001 (2003).
[4] D. Sigg, Class. Quantum Grav. 19, 1429 (2002).
[5] F. Acernese et al., Class. Quantum Grav. 19 1421 (2002).
[6] See for up to date results: http://www.ligo.org.
[7] M. Cerdonio, L. Conti, J.A. Lobo, A. Ortolan, L. Taffarello, and J. P. Zendri,
Phys. Rev. Lett. 87 031101 (2001).
[8] T. Briant, M. Cerdonio, L. Conti, A. Heidmann, A. Lobo, M. Pinard,
Phys. Rev. D 67, 102005 (2003).
[9] Michele Bonaldi, Massimo Cerdonio, Livia Conti, Michel Pinard, Giovanni A.
Prodi, Luca Taffarello, and Jean Pierre Zendri,
Phys. Rev. D 68 102004 (2003)
[10] http://virgo-bwulf.pg.infn.it/ punturo/FP6/
[11] V.B. Braginsky, F.Ya. Khalili and P.S. Volikov, Physics Letters A 287, 31 (2001).
[12] A.E.H. Love, A treatise on the mathematical theory of elasticity (Dover, New
York, 1944).
[13] C.Z. Zhou and P.F. Michelson, Phys. Rev. D 51, 2517 (1995).
16
[14] S.M. Merkowitz and W. W. Johnson, Phys. Rev. D 56, 7513 (1997).
[15] J.A. Lobo, Phys. Rev. D 52, 591 (1995).
[16] W. Duffy, Jr., J. Appl. Phys. 72, 5628 (1992).
[17] B. Harnisch et al., ESA bulletin 95 (1998).
[18] J.P. Zendri et al., to appear on the proc. of the XXXVIIIth Rencontres de
Moriond - Gravitational Waves and Experimental Gravity, Les Arcs 1800, March
2003 (Edition Frontiers, Paris)
[19] L. Conti, M. De Rosa and F.Marin, J. Opt. Soc. Am. B 20, 462 (2003).
[20] L. Conti, M. De Rosa, F. Marin, L. Taffarello,M. Cerdonio,
J. Appl. Phys. 93, 3589 (2003).
[21] S. Kobayashi, IEEE Trans. on Diel. Elec. Ins. 4, 841 (1997).
[22] F. Marin, L. Conti and M. De Rosa, Phys. Lett. A 309, 15 (2003).
[23] A. Vinante, R. Mezzena, G. A. Prodi, S. Vitale, M. Cerdonio, P. Falferi, and M.
Bonaldi, Appl. Phys. Lett. 79, 2597 (2001)
[24] A. Vinante, poster presented at the 5th Amaldi Conference on Gravitational
Waves, Tirrenia (Italy), July 6-11 2003.
[25] D. D. Awschalom, J. R. Rozen, M. B. Ketchen, W. J. Gallagher, A. W. Kleinsasser, R. L. Sandstrom and B. Bumble, Appl. Phys. Lett. 53, 2108 (1988)
[26] P. Carelli, M. G. Castellano, G. Torrioli and R. Leoni, Appl. Phys. Lett. 72, 115
(1998)
[27] M. Mück, J. B. Kycia and J. Clarke, Appl. Phys. Lett. 78, 967 (2001).
[28] F. C. Wellstood, C. Urbina, and J. Clarke, Phys. Rev. B 49, 5942 (1994).
[29] H.J. Paik, G.M. Harry, T. Stevenson, in Proceedings of the Seven Marcel Grossmann Meeting on General Relativity, Stanford, 1994, R.T. Jantzen, G. Mac
Kreiser, R. Ruffini eds., (World Scientific, Singapore, 1996 ) p. 1483.
[30] S. Kota, Smart Materials Bulletin, p.7, March 2001; J.F.Tressler et al., IEEE
Trans. on Ultrasonics, Ferroelectrics and Frequency Controls, 45 1363 (1998); J.
Zhang et al., Ultrasonics 37, 387 (1999).
17
7
Qualifications of the proponents
The INFN sections participating to this R&D project are experienced in the design and
developments of acoustic and interferometric Gravitational Wave detectors.
For what concerns the Florence Section activity, as usual when both the components
(Urbino and Florence Universities) are involved in INFN experiments, the sharing of
responsibilities and activities is settled in order not to have duplication or dispersion of
resources but on the contrary an optimization of the investments by INFN on both the
sites. In this particular case of DUAL R&D experiment, the HW related to inertial (or
viscous) sensors/dampers will be fully developed in the laboratory in Florence, whilst
the start-up of the FEM analysis and subsequent tool development will be pursued in
Urbino.
A main part of the experimental activity will be hosted by the Laboratori Nazionali di
Legnaro, in the laboratories of the AURIGA experiments. In particular the experimental
activity of the personnel by the INFN Padova section will be hosted by the LNL, which
contributes all the technical services. For this reason the participation of the LNL to
the experiment goes well beyond the 0.2 FTE of the following table.
18
Working group
FTE
Sezione di Firenze
Experience
2.5
Fi/Urb suspension group
Development of inertial sensor of displacement and related control loops of the VIRGO superattenuators and
measurement of the seismic noise in the area of the interferometer.
Fi optic development group
Metrology tools for optical measurements. Development
of optical readout systems.
Laboratori Nazionali di Legnaro
0.2
Provide the infrastructures and the laboratory necessary
to the resonant cryogenic detectors AURIGA
Sezione di Napoli
2.1
Development of inertial sensor of displacement and related control loops of the VIRGO superattenuators and
measurement of the seismic noise in the area of the interferometer.
Sezione di Padova
2.4
Design, construction and operation of the resonant cryogenic detector AURIGA. Development of capacitive and
optical readout.
1.6
Design, construction and operation of the resonant cryogenic detector AURIGA. Development of SQUID based
amplification and matching chains for capacitive readout
systems.
TOTAL
8.8
19
7.1
Functional chart
20
7.2
Participants (from 2005)
Sezione di Firenze
Maurizio De Rosa - ric. INOA (75 %)
Giovanni Giusfredi - primo ric. INOA (30%)
Massimo Inguscio - prof. ord. (20 %)
Francesco Marin - prof. ass. (Resp. Locale, 30 %)
Massimo Mazzoni - R.U. Univ.(40 %)
Bruna Perniola - tecnologa Univ. (40 %)
Flavio Vetrano - prof. ord. (20 %)
Contact persons:
Francesco Marin ([email protected])
Massimo Mazzoni ([email protected] )
Laboratori Nazionali
di Legnaro
Antonello Ortolan - ricercatore (20%)
([email protected] )
Sezione di Napoli
Rosario De Rosa - ricercatore Univ.(Resp. Locale, 50%)
Fabrizio Barone - prof. ass. (30%)
Leopoldo Milano - prof. ord. (30%)
Silvio Pardi - dottorando (100%)
Contact person:
Rosario De Rosa ([email protected] )
Sezione di Padova
Michele Bignotto - dottorando (100%)
Massimo Cerdonio - prof. ord. (Resp. Nazionale, 40%)
Livia Conti - ass. ricerca Univ.(Resp. Locale, 30% )
Luca Taffarello - tecnologo INFN (20%)
Jean-Pierre Zendri - ricercatore INFN(50%)
Contact persons:
Massimo Cerdonio ([email protected])
Livia Conti ([email protected] )
Gruppo Collegato
di Trento
Michele Bonaldi - ricercatore ITC-IFN (Resp. Locale,
50%)
Paolo Falferi - ricercatore ITC-IFN (30%)
Renato Mezzena - tecnologo Univ.(30%)
Giovanni Andrea Prodi - prof. ass. (30%)
Andrea Vinante - ass. ricerca Univ.(20%)
Contact person:
Michele Bonaldi ([email protected])
21
7.3
Selected publications
• Wideband Dual Sphere Detector of Gravitational Waves
M. Cerdonio, L. Conti, J.A. Lobo, A. Ortolan, L. Taffarello, and J. P. Zendri,
Phys. Rev. Lett. 87 031101 (2001).
• Thermal and back-action noises in dual-sphere gravitational-wave detectors
T. Briant, M. Cerdonio, L. Conti, A. Heidmann, A. Lobo, M. Pinard,
Phys. Rev. D 67, 102005 (2003).
• Selective readout and back-action reduction for wideband acoustic gravitational wave detectors
Michele Bonaldi, Massimo Cerdonio, Livia Conti, Michel Pinard, Giovanni A. Prodi, Luca Taffarello,
and Jean Pierre Zendri,
Phys. Rev. D 68 102004 (2003)
• A folded Fabry-Perot cavity for optical sensing in gravitational wave detectors
F. Marin, L. Conti and M. De Rosa
Phys. Lett. A 309 15-23 (2003)
• Room temperature gravitational wave bar detector with optomechanical readout
L. Conti, M. De Rosa, F. Marin, L.Taffarello, M. Cerdonio
J. of Appl. Phys. 93 3589 (2003)
• Experimental measurement of the dynamic photothermal effect in Fabry-Perot cavities for
gravitational wave detectors,
M. De Rosa, L. Conti, M. Cerdonio, M. Pinard, F. Marin,
Phys. Rev. Lett. 89 (2002) 237402
• Thermoelastic effects at low temperatures and quantum limits in displacement measurements,
M. Cerdonio, L. Conti, A. Heidmann, M. Pinard,
Phys. Rev. D 63 082003 (2001)
• Low-amplitude noise laser for AURIGA detector optical readout
L. Conti, M. De Rosa, F. Marin,
Appl. Opt. 39, 5732 (2000)
• High spectral purity laser system for the AURIGA detector optical readout
L. Conti, M. De Rosa, F. Marin,
J. Opt. Soc. Am. B 20, 462 (2003)
• Dc superconducting quantum interference device amplifier for gravitational wave detectors with
a true noise temperature of 16 µK
A. Vinante, R. Mezzena, G. A. Prodi, S. Vitale, M. Cerdonio, P. Falferi, and M. Bonaldi,
Appl. Phys. Lett. 79, 2597 (2001)
• Sensitivity enhancement of Quantum Design dc Superconducting Interference Devices in twostage configuration
R. Mezzena, A. Vinante, P. Falferi, M. Bonaldi, G.A. Prodi, S. Vitale and M. Simmonds
Rev. Sci. Instrum. 72, 3694 (2001)
• Noise sources and dissipation mechanisms of a 120 ~ SQUID amplifier
P. Falferi, M. Bonaldi, A. Cavalleri, M. Cerdonio, A. Vinante, R. Mezzena, Ke-xi Xu, G.A. Prodi,
and S. Vitale,
Appl. Phys. Lett. 82, 931 (2003)
22
• A real-time procedure for noise uncoupling in laser interferometry
F. Barone, R. De Rosa, A. Eleuteri, L. Milano, R. Tagliaferri, K. Qipiani,
IEEE Transactions on Nuclear Sciences 49 411-416 (2001).
• A procedure for noise uncoupling in laser Interferometry
F. Barone, E. Calloni, R. De Rosa, A. Eleuteri, L. Milano, K. Qipiani,
Classical and Quantum Gravity 19 1529-1536 (2002).
• The Virgo Suspensions
The Virgo Collaboration,
• The Inertial Damping of the VIRGO superattenuator and the residual motion of the mirror
• Last stage control and mechanical transfer function measurement of the VIRGO suspensions
Review of Scientific Instruments 73 2143-2149 (2002).
23
8
8.1
Research program
2005
Sezione di Firenze - suspension development
• Study and preliminary dimensioning of a passive suspension for a dummy
payload.
• Set up of the FEM tool (ANSYS) for modelling the passive suspension: definition of the parts that can be parametrized.
• Operation of (already realized) standard prototypes of inertial sensors and
their characterization.
Sezione di Firenze - optic development
• Improvement of the frequency stabilization of the laser, with a goal of 10−9
Hz 2 /Hz in the range 1-5 kHz.
• Extending the amplitude stabilization of the laser in the range 1-5 kHz.
• Implementation of wide reading area optical cavities.
Sezione di Napoli
• Definition and preliminary implementation of the necessary changes on the
Napoli suspension chain to perform the tests for the R&D.
• Design and preliminary tests of the digital control system hardware and software architecture to be used for the first tests of the suspended mass control.
• Characterization of the inertial sensors prototypes used in the active control
of the suspension, to evaluate their sensitivity and dynamics in connection
with the scientific requirements and the architecture of the control system.
• Definition and characterization of the conditioning electronics used to interface the signals from sensors to the digital control system.
Sezione di Padova - Laboratori Legnaro
• study of wideband mechanical amplifiers: leverage, velocity transformer.
• Construction of a vacuum chamber for the room temperature thermal noise
measurement of the mechanical oscillator equipped with broadband mechanical amplifier: passive thermal insulation and mechanical suspension.
• study of impedance matching networks optimized for the capacitive readout.
• Thermal noise modelling of mechanical amplifier in order to evaluate and
control its role in the detector.
• Study of the noise of electro-mechanical devices coupled to the detector: define requirements on the SQUID-based readout so to guarantee at the detector
the same energy sensitivity as in bench tests of the SQUID amplifier.
24
8.2
2006
• Detailed development (on the basis of the knowledge gained on the detector structure) of the first stages of the passive suspension and of the
ground/basement/inverted pendulum system.
• Study of the suspension wiring
• Sensor conformation to the requirements. Realization of prototypes of ’linear’
sensors. First tests at cryogenic temperatures
• Project of the control loop for both kind of sensors
• Improvement of the simulation tool
• Improvement of the tilt sensor
• Table-top test of a Folded Fabry-Perot (FFP): alignment, finesse measurement.
• Optimization of a photodetection bench for high power laser
Sezione di Napoli
• Study of the wiring between the suspension stage, suspended mass and payload.
• Integration of the inertial sensors in the suspension stage and suspended mass.
• Cryogenic test of the narrow band inertial sensors.
• Study and development of control system strategies and architectures for
optimizing the suspension control in connection with the performances of the
wide and narrow band sensors and actuators.
• Room temperature measurement of the mechanical quality factor of the mechanical amplifier and of the effect of its coupling to mechanical oscillator.
Measurements of the thermal noise of the composed system.
• realization of a active thermal stabilization for the thermal noise measurement
of a mechanical oscillator with mechanical amplifier.
• Room and cryogenic temperature measurements of the mechanical quality
factor of candidate materials for the Dual detector (SiC, Si).
• definition of the requirements for underground laboratories to host the Dual
detector.
• Design and realization of a mechanical amplifier with a capacitive readout.
• Definition of the requirements on the detector test masses.
25
• Definition of a mechanical amplifier tailored for capacitive readout.
• Definition of a procedure for surface preparing and conditioning that allows
to increase the bias field up to 100MV/m for capacitive readout.
• Improvement of bandwidth and dynamic range of the two-stage SQUID amplifier in order to match the Dual requirements. Measurements of the noise
temperature in the 2-10 kHz range and definition of the constructive properties of new SQUID chips.
• Implementation of wideband noise matching chains: electro-mechanical matching and/or development of SQUID amplifiers with specific input impedance.
8.3
2007
• realization of inertial sensor prototypes working at low temperatures and
designed for the stabilization of the inertial platform in the range between
0.1Hz and several hundreds of Hz by means of hierarchical control.
• realization of a tilt sensor to be characterized on controllable platforms.
• Final design of the passive suspension for masses in the range 10-100 ton with
sismic reduction at the suspension point down to 10−10 m with hierarchical
control.
• Study of viscous dampers.
• Final setup of the FEM modelling tool.
• Implementation of frequency and amplitude stabilization for a high power
laser.
• Vacuum test of FFP, with high power laser.
• Conceiving and testing Q-keeping methods for fixing the mirrors.
Sezione di Napoli
• Implementation of the final inertial sensor prototype, compatible with the
cryogenic environment, for the complete control loop of the inertial platform,
in the band 0.1 - 100 Hz.
• Study of the effectiveness of a tilt-meter as a part of the control system
transducers.
• Final design of a suspension chain able to sustain a payload of 103 - 104 Kg
with a seismic isolation based on both passive and active damping.
• Study of fabrication/assembling procedures to realize a Dual detector from
smaller dimension samples while preserving a high mechanical quality factor
at cryogenic temperatures.
26
• Measurements of the mechanical quality factor at cryogenic temperatures of
mechanical resonators constituted by smaller assembled samples
• Measurements of the mechanical quality factor at cryogenic temperatures of
a mechanical resonator equipped with mechanical amplifier
• Measurements on a full wideband capacitive readout: mechanical amplifier,
wideband matching network and SQUID.
• Estimation of cooling and thermal link requirements for DUAL test masses
and technologies selection.
• Study of the inter masses suspension will be studied that allows the use of a
mechanical amplifier.
• Realization of the wideband matching network and SQUID for the capacitive
readout.
• Selective detection schemes for measurement of small displacements: full calculation of thermal and backaction noise at the detector output.
27
8.4
Instrumentation available for the research
The following instruments and systems are already available and will be used for the
activities foreseen in the first year of the research program:
Firenze Two optical tables
Nd-YAG laser (100mW) with ’noise eater’ and external phase modulator
Two mechanically isolated Zerodur optical cavities
with thermally-stabilized vacuum chambers
Laser beam shape analyzer
Laminar flow bench
Oscilloscope with FFT
Padova optical table in a sound-proof environment
Nd-YAG laser (50mW) with ’noise eater’ and external
phase modulator
Mech. isolated Zerodur optical cavity in thermallystabilized vacuum chamber
Oscilloscope with FFT, spectrum analyzer, general
instrumentation (signal generators, low-noise amplifiers, lockin)
Cryogenic facility with mechanical isolation
Clean room, class 5000
Movable clean room, class 5000
Data acquisition system with 5kHz sampling rate
Napoli Multi Stage Pendular Suspension
NIM crate for Accelerometer Electronic
Oscilloscope
Low Frequency Spectrum Analyser
VME Crate for Control and Acquisition System
CPU for VME + OS
Workstation for development and simulations
Trento Commercial SQUID system
Dilution refrigerator
PC based Digital Signal Analyzer
Workstation for development and simulations
28
50%
30%
30%
80%
80%
50%
20%
20%
20%
30%
50%
50%
50%
20%
50%
50%
100%
30%
100%
100%
50%
50%
50%
50%
100%
9
Detailed funding request
2005
Working group
Totale
Giustificazione
39 ke
Missioni: 6 ke interne, 0 ke estero
Consumo: 12.5 ke licenza annuale ANSYS, 3 ke accelerometri, 3ke cablaggi vari
Costruzione apparati: 7 ke disegni progetti
Inventariabile: 4.5 ke workstation, 3 ke generatore di
funzioni
23 ke
Consumo: 8 ke (specchi, ottiche, montaggi ottici, materiale da vuoto, elettronica, fibre ottiche)
Costruzione apparati:
Inventariabile:
10 ke (oscillatore, amplificatori a
guadagno e banda variabile)
6 ke
Consumo: 5 ke
Sezione di Napoli
40 ke
Consumo: 10 ke (cablaggi, componenti meccanici ed
elettronici).
Costruzione apparati: 10 ke (modifiche della Sospensione di Napoli per i test per R&D)
Inventariabile: 15 ke (acquisto di ADC e DAC VME
boards per interfacc. con controllo digitale).
Sezione di Padova
45 ke
Missioni: 3 ke interne, 5 ke estero (coll. ESTEC e Glasgow per test masses devel., Parigi optical readout)
Consumo: 5 ke materiali
Costruzione apparati: 4 ke banco ottico, 5 ke lavorazioni
meccaniche
Inventariabile: 20 ke laser, 3 ke pompa ionica
45 ke
Missioni: 5 ke interne, 5 ke estero (coll. Glasgow per
test masses development, Giessen SQUID devel.)
Consumo: 5 ke materiali ed elio liquido
Costruzione apparati: 5 ke lavorazioni meccaniche
Inventariabile: 18 ke generatore di funzioni alta tensione,
2 ke pompa rotativa pulita, 5k e vasche pulizia
Sezione di Firenze
TOTALE
198 ke
29
2006
Working group
Totale
Giustificazione
60 ke
Consumo: 12.5 ke licenza annuale ANSYS, 5.5 ke
Costruzione apparati: 20 ke modelli in scala, 10 ke accelerometri
Inventariabile:
71 ke
Missioni: 5 ke interne, 4 ke estero (Kastler-Brossel)
Costruzione apparati: 10 ke costruzione camera da vuoto
per FFP
Inventariabile: 40 ke (laser di potenza (Nd:YAG, 1 o 2
W), pompe da vuoto)
6 ke
Consumo: 5 ke
Sezione di Napoli
64 ke
Missioni: 8 ke interne, 4 ke estero (coll. Caltech e
TAMA per sviluppo accelerometri)
Consumo: 20 ke (Componenti elettronici per interfacciamento, componenti meccanici).
Costruzione apparati: 17 ke realizzazione protipi di accelerometri
Inventariabile: 15 ke (realizzazione del primo prototipo
di sistema digitale di controllo VME).
Sezione di Padova
50 ke
Missioni: 5 ke interne, 5 ke estero (coll. ESTEC e Glasgow test masses development, Parigi optical readout)
Consumo: 10 ke materiali, 5 ke elio liquido, 5 ke lavorazioni meccaniche e trattamenti
Costruzione apparati: 20 ke ottiche (isolatori, lenti,
specchi, montaggi, cav riferimento)
Inventariabile:
48 ke
Consumo: 5 ke SQUID, 5 ke materiali ed elio liquido
Costruzione apparati: 10 ke camera da vuoto per test
alti campi elettrici
Inventariabile: 15 ke saldatrice ad ultrasuoni, 5 ke
cappa a flusso
Sezione di Firenze
TOTALE
299 ke
30
2007
Working group
Totale
Giustificazione
62 ke
Consumo: 10 ke
Costruzione apparati: 20 ke disegni, 20 ke accelerometri
Inventariabile:
25 ke
Missioni: 8 ke interne, 2 ke estero (Kastler-Brossel)
Inventariabile: 5 ke modulatore elettro-ottico per laser
di potenza
6 ke
Consumo: 5 ke
Sezione di Napoli
62 ke
Missioni: 10 ke interne, 2 ke estero (coll. Caltech e
TAMA per sviluppo accelerometri)
Consumo: 10 ke (Componentistica meccanica ed elettronica per il sistema finale).
Costruzione apparati: 30 ke realizzazione accelerometri
Inventariabile: 10 ke (completamento sistema di controllo).
Sezione di Padova
35 ke
test masses development, Parigi optical readout)
Consumo: 10 ke materiali, 10 ke elio liquido, 5 ke lavorazioni meccaniche e trattamenti
Inventariabile:
30 ke
Consumo: 5 ke SQUID custom , 5 ke elio liquido
Inventariabile: 10 ke elettronica II SQUID
Sezione di Firenze
TOTALE
220 ke
31
10
Other fundings
Approved projects
Institution
Call type / title
Main topic
Duration
Funding
EGO
R&D on GW research
Readouts for DUAL
3 years
1 research assistant
EU
Integrated Infrastructure
Initiative / STREGA
(ILIAS)
Thermal noise reduction in GW detectors
readouts
5 years
3 years post-doc
90 ke consumables
Submitted projects
Institution
EU
Call type / title
FP6 Design Study /
EGO Observatory
Main topic
Test mass
properties.
32
material
Duration
Funding
4 years
4 years post-doc
3 years technical assistant
280 ke consumables
Codice
Esperimento
Gruppo
DUAL−RD
2
NUCLEARE
PREVISIONE DI SPESA
In KEuro
ANNI
Missioni Missioni
2005
2006
2007
TOTALI
Mod EC./EN. 6
25,0
32,0
39,0
10,0
20,0
16,0
96,0
46,0
Spese
Materiale
Affitti e
Trasporti e
di
di
facchinaggi
calcolo
consumo
31,0
80,5
51,5
87,0
75,0
85,0
70,0
25,0
70,0
206,5
0,0
0,0
0,0
180,5
188,0
TOTALE
Compet.
198,0
299,0
220,0
717,0
Struttura
TN
Codice
Esperimento
DUAL−RD
Gruppo
2
N
1
2
3
4
RICERCATORE
Cognome e Nome
Qualifica
Dipendenti
Incarichi
Affer.
al
. gruppo
%
2
2
2
2
50
30
30
20
N
Ruolo Art. 23 RicercaAssoc
BONALDI Michele
FALFERI Paolo
VINANTE Andrea
Ric.
Ric.
P.A.
AsRic
Qualifica
Incarichi %
Ass.
Ruolo Art. 23
Tecnol.
Univ.
30
Dipendenti
TECNICI
Cognome e Nome
1
0.3
Qualifica
Incarichi
Dipendenti
Collab.
Ruolo Art. 15
tecnica
Annotazioni:
mesi−uomo
Mod EC./EN. 7
%
Assoc.
tecnica
SERVIZI TECNICI
Denominazione
Cognome e Nome
1 MEZZENA Renato
N
TECNOLOGI
0
0
NUCLEARE
Codice
Esperimento
Gruppo
DUAL−RD
2
Data
completamento
Descrizione
30/09/2005
Realizzazione di un amplificatore meccanico a banda larga e caratterizzazione della sua funzione di trasferimento.
30/06/2005
Modellizzazione del rumore termico di un amplificatore meccanico a banda larga.
30/06/2005
Caratterizzazione dei sensori inerziali a banda stretta da utilizzare per il controllo.
30/11/2005
Stabilizzazione di frequenza e ampiezza di un laser Nd:YAG nella zona 1−5 kHz
30/06/2005
Caratterizzazione completa nella banda di interesse dei sensori inerziali
a banda larga già disponibili (Virgo−like).
30/11/2005
Realizzazione dei primi modelli FEM di sospensione passiva interfacciata allo
stadio di preisolamento terreno/supporto con individuazione dei parametri di interesse.
30/11/2005
Realizzazione dell'elettronica di condizionamento dei segnali da usare per il controllo della sospensione.
Mod EC./EN. 8
Codice
Esperimento
Gruppo
LISA−RD
2
Rapp. Naz.: Stefano VITALE
NUCLEARE
Stefano VITALE
TN
Ricerca di onde gravitazionali su satellite
Linea di ricerca
Trento
Laboratorio ove
Sigla dello
esperimento
assegnata
dal laboratorio
Acceleratore usato
Fascio
(sigla e
caratteristiche)
Emissione di onde gravitazionali
Processo fisico
studiato
Interferometro laser su satelliti in orbita eliocentrica
Apparato
strumentale
utilizzato
Trento, Roma II, Firenze − Urbino, Perugia, Napoli
all'esperimento
Max Planck Institut, Onera (Parigi), NASA, ESA.
Istituzioni esterne
all'Ente partecipanti
molti anni
Durata esperimento
Mod EC. 1
Struttura
TN
Codice
Esperimento
LISA−RD
Resp. loc.: Rita Dolesi
Gruppo
2
In KEuro
IMPORTI
VOCI
DI
SPESA
Parziali
Totale Compet.
SJ
Partecipazione a riunioni della collaborazione INFN LISA PF
di cui SJ
10,0
10,0
Partecipazione a congressi e a riunioni del TEAM internazionale di LTP e
LISA
20,0
Materiali da costruzione, componenti elettronici e da vuoto
12,0
20,0
12,0
Consorzio
Ore CPU
Spazio Disco
Cassette
Altro
Up−grading sistemi di pompaggio facility pendolo 1 massa Strumentazione 35,0
elettronica varia (oscillators,stab. var. power suppl., oscilloscope Tecktronix
200 MHz, Computer+DAC). Upgrading read−out ot
Prototipi di parti di sensore capacitivo.Prototipo di sensore capacitivo
rappresentativo del Flight Model per LTP.Supporto alla costruzione della
facility a LNGS. Upgrading della facility per il test
35,0
120,0
120,0
Totale
197,0
di cui SJ
0,0
Mod EC./EN. 2
A cura della
Comm.ne
Scientifica
Nazionale
Struttura
TN
Codice
Esperimento
LISA−RD
Gruppo
2
Struttura
TN
Codice
Esperimento
LISA−RD
Gruppo
2
Codice
Esperimento
Gruppo
LISA−RD
2
NUCLEARE
In KEuro
Materiale
di
consumo
Trasporti
e
facchinaggi
Spese
di
calcolo
Affitti
e
manutenzioneinventariabile apparati
Struttura
Missioni
interne
FI
NA
PG
RM2
TN
8,0
10,0
9,0
8,0
10,0
16,0
8,0
11,0
7,0
20,0
4,0
3,0
32,0
4,0
12,0
21,0
3,0
13,0
70,0
35,0
16,0
120,0
TOTALI 45,0
62,0
55,0
142,0
209,0
Missioni
estere
SJ
SJ
SJ
SJ
SJ
SJ
SJ
SJ
43,0
30,0
31,0
31,0
A carico
di altri
Enti
TOTALE
Compet.
SJ
92,0
54,0
65,0
105,0
197,0
31,0
0,0
0,0
0,0
0,0
2600,0
513,0 31,0 2600,0
Mod EC./EN. 4
Codice
Esperimento
Gruppo
LISA−RD
2
NUCLEARE
L
Anno
finanziario
Missioni
interne
2000
2001
2002
2003
2004
TOTALE
Mod EC. 5
Missioni
estere
consumo facchinaggi
Spese di
calcolo
In kEuro
Affitti e
Materiale
Costruzione
apparati
TOTALE
1,5
4,0
16,5
30,0
5,0
6,5
6,5
33,5
3,0
7,0
11,5
26,0
23,7
20,0
5,0
71,0
140,5
25,0
19,0
73,0
179,0
23,7
54,5
41,5
178,5
409,0
52,0
51,5
47,5
260,2
296,0
707,2
LISA PF project:
Reshaping of the workplan: optical readout system, and
advanced on ground test facility at LNGS
May 20, 2004
1
2
2.1
2.2
2.3
3
4
5
6
INTRODUCTION
THE ON GROUND TEST FACILITY
Status of the on ground testing
Test facilities upgrades
LNGS
OPTICAL READOUT SYSTEM
WORKPLAN/MILESTONES
MEMBERS OF THE COLLABORATION AND FINANCIAL SUPPORT
REFERENCES
2
2
3
12
20
24
40
42
42
1
1
INTRODUCTION
LISA (Laser Interferometer Space Antenna) [1] will be the first high sensitivity space-borne
gravitational wave detector. LISA, a joint ESA-NASA endeavor, consists of a constellation of 3
spacecraft in heliocentric orbits. Orbits are adjusted such that the three spacecraft maintain an
equilateral triangle formation with a 5 106 km sidelength. Each spacecraft contains a pair of testmasses nominally in pure geodesic motion, with no mechanical contact to the spacecraft itself. Each
test-mass serves as the end-mirror of a single arm interferometer, with the other end-mirror in one of
the other two spacecraft. Two semi-independent two-arm interferometers are formed by taking the
difference of the signals from the 3 independent arms. LISA sensitivity goal is a strain power spectral
density of 4×10-21 1/√Hz at around 3 mHz. The useful bandwidth is between 0.1 mHz and 0.1 Hz.
LISA falls in a special category among gravitational wave detectors, as it has “guaranteed” sources in
a set of galactic binary systems for which we can calculate the expected signal, a very bright one also.
LISA will also look for a wide range of sources, including super-massive black hole binaries (very
bright even at z>3), the capture of stellar objects by a super-massive black hole, and primordial
gravitational radiation backgrounds from the Big Bang.
The LISA sensitivity is limited at low frequencies by the ability to set the test-masses in perfect free
fall. No acceleration in excess of 3×10-15 ms-2/√Hz can be tolerated at the frequency of 0.1 mHz. In
order to achieve this level of performance, a key point is to keep the spacecraft as stationary as
possible around the test-masses, at least along the two directions of the laser beams. This is achieved
by a “drag-free” control loop where a set of micro-thrusters are driven by the signal generated by a
precision capacitive displacement sensor, and force the spacecraft to follow the test-mass. The
ensemble of the test-mass and the capacitive sensor is referred to here as the inertial sensor. This
technique cannot be fully tested on the ground. The reasons for this are many, the leading one being
the difficulty to measure very tiny forces against a background of the 1 g terrestrial gravitational
acceleration.
Because of this difficulty, ESA has decided to implement a test-flight known as the LISA Pathfinder.
NASA will also participate in the mission. The ESA test instrument is known as the LISA Test-flight
Package (LTP). The basic idea behind the LTP is to squeeze one LISA arm from 5 106 km to a few
centimeters and place it on board a single spacecraft (S/C). Thus the key elements are two nominally
free-falling test-masses and a laser interferometer whose purpose is to read the relative acceleration
between the test-masses.
The Trento group provides the Principal Investigator of the LTP experiment. They also have the
leadership in the development of the inertial sensor for the LTP, which has been carried out thus far
with ESA funding while the flight model will be provided by ASI.
This proposal concerns two basic items: an on ground test facility, and a redundant optical readout
system of the position of the test masses in LTP. In what follows, the two items will be described.
2
THE ON GROUND TEST FACILITY
For the last 3 years INFN has funded an intense campaign of development of prototype sensors, in
preparation for the flight model, as well as the development of ground based methods to test, to the
best possible levels, the free-fall condition. The key instrument of this testing effort has proven to be a
2
torsion pendulum bench. We have recently demonstrated a performance that, for many disturbances, is
within a factor 100 from LISA and a factor 10 from the LTP goal.
Our method as developed up to now, though very powerful, can realize this “free-fall” condition for
one degree of freedom at a time. In addition the laboratory conditions in term of ground tilt, mass
motion and temperature oscillations are currently limiting factors of our current pendulum
performance and would represent even more serious problems for the next generations of higher
sensibility torsion pendulum and of multi-degree of freedom suspension system.
The on-ground testing for LISA and LTP would then greatly benefit from moving the next generation
of test benches to the environment of Laboratori Nazionali del Gran Sasso.
Our proposal is then to install and run the next generation of ground testing facilities for the LISA testmass geodesic performance at the LNGS.
2.1
Status of the on ground testing
The key instrument of this testing effort has been a torsion pendulum bench.
It has allowed demonstration of capacitive sensing performance within LISA specifications and, as a
consequence, provided experimental confirmation to the predictions from the applied noise and
performance models.
We also successfully employed the technique to make sensitive, low frequency, single degree of
freedom torque noise measurements, based on the high sensitivity of the torsion pendulum, which
consists of a lightweight LISA test mass (a 40 mm cube) suspended by a thin fiber[9]. As shown in
Figure 1 the torque noise floor of the apparatus was measured to be less than 20 fN m Hz between
0.3 and 10 mHz, with a minimum of 4 fN m Hz at 3 mHz. This torque noise can be converted into
an upper limit for random differential force noise exerted by the LISA gravitational sensor prototype
on the suspended test mass. The “armlength” used to make this conversion depends on the specific
disturbance mechanism. For back-action effects related to the readout and actuation circuitry, the
armlength is half the separation between adjacent sensing electrodes (roughly 10 mm) and the upper
limit can be set at 1.5 pm s 2 Hz in the frequency range 0.3 to 10 mHz, when referred to a bulk
LISA test mass of the same size. For random inelastic molecular impacts, the torque noise levels can
be converted into a minimum acceleration noise below 200 fm s 2 Hz at 3 mHz. This last limit
corresponds to roughly a factor 70 over the LISA flight goal, and a factor 7 over the LTP flight test
goal.
3
Figure 1 Torque and acceleration noise upper limits for LISA, compared to the instrument limit (dominated at low
frequency by thermal noise and at high frequency by the readout noise), compared with the LISA goal. The red curve data
were calculated from the blue curve by subtracting the torque induced by floor tilt, through the measured tilt-twist coupling
coefficient and pendulum transfer function (see later in this document for details). The sharp peak at 0.5 mHz is an artifact
of a test mass charge measurement performed during the 101 hour run, analyzed here by cutting it into three 120 ks
windows. The acceleration levels, compared with the LISA goal, are evaluated assuming an armlength a = 20 mm to
convert torque into force noise, and a 1.3 kg LISA cubic test mass.
The torsion pendulum high torque sensitivity, which translates into a resolution near 0.1 fN m over 1
hour measurement, can also be exploited to perform, with a modulation technique, characterization of
individual important disturbance sources of acceleration noise. It is possible to single out these
interactions, studying their governing parameters and coupling mechanisms, by performing coherent
experiments, where the source under consideration is modulated, and the torque exerted on the test
mass is measured by coherent demodulation of the pendulum twist angle. These experiments, besides
characterizing these specific inertial sensor disturbances at the level of the LISA specs, constitute
significant ground testing of the same source characterization procedures that will be employed in
flight during the LTP mission[7]. In Figure 2 and Figure 3 we report on the measurement of two such
effects. Figure 2 reports the evaluation of the negative, additive “spring constant” associated the
voltages needed for the sensor readout, which is a significant validation of the electrostatic modeling
of the real sensor. Figure 3 sketches the results of the measurement of the rotational combination of
stray DC bias on the sensor electrodes, an effect which is expected to be significant at low frequencies
for LISA, in its mixing with low frequency electrostatic noise (for example induced by the random
cosmic rays arrival).
By means of this technique, measurements of the sensor electrostatic stiffness at the 5% level and
detection and compensation of stray dc electrostatic biases at the millivolt level have been successfully
performed [9][10][8].
4
Figure 2 Plot showing the 1fm, 3fm, and 5fm (5, 15, and 25 mHz) torque components in the electrostatic rotational stiffness
measurement, as a function of the relative sensor - test mass rotation angle φsens, measured by the gravitational sensor itself.
The torque is induced by alternatively switching on and off the readout bias source at 1fm , with the test mass at different
relative rotation angles with respect to the sensor housing. The torque were reconstructed by demodulating the deflection
angles measured by the autocollimator, and applying the torsion pendulum transfer function. The roughly 5:3:1 ratio of the
slopes corresponds to the 1/f square wave spectral content. The sub-fN m resolution for the 1fm data is smaller than the
point markers.
Figure 3: Plot of DC bias measurement torques for applied bias amplitude V∆ = 3 V at fm = 5 mHz, as a function of the
applied DC compensation voltage VC. The measured torque amplitude is proportional to a residual DC imbalance, which,
for spatially uniform stray DC biases, is ∆φ + 4 VC.
Upgrading of the existing facility and the creation of new test benches, with better sensitivities and
additional functionalities, will allow us to further characterize disturbance effects and to verify the
performance of key features of the inertial sensor.
There is also a strong interest in extending the LISA measurement band below 0.1 mHz, possibly
down to .01 mHz 0. This possibility is hampered by the lack of knowledge of the sensor at these very
5
low frequencies, and extensing the pendulum measurements to these very low frequencies would also
be very valuable.
Noise sources that have been identified but not yet investigated include various thermal gradients
effects (related to the radiometric effect, radiation pressure and temperature dependent outgassing
effects) and additive residual couplings to the test mass relative displacement.
Important key functionalities whose performance should be investigated include the test mass charge
measurement and control system based on UV and the so called “low frequency suspension”. Both for
LISA and for the LTP, all relative test mass-spacecraft degrees of freedom must be controlled by
either the drag-free loop or the “low frequency suspension. The role of the latter is to counteract dc
forces at very low frequencies and, within the measurement band, to stabilize the test masses. Proper
control laws have been identified, and the baseline actuation strategy is to provide actuation forces by
means of amplitude-modulated square pulse train voltages applied to the electrodes. In order to
guarantee the performance of these systems, it is important to verify, for example, that the low
frequency suspension actuation laws does not introduce excess stray forces in the measurement band
and the actual frequency dependent gain of the control laws. Stray effects of the “low frequency
suspension” are expected to be created by several sources of “cross-talk” among different degrees of
freedom, like cross-talk of actuation force/torque along other DOF into a force along the sensitive axis,
cross-talk of displacement/rotation of other DOF into the sensitive axis displacement readout that is
used for drag-free and/or electrostatic suspension and non diagonal terms of parasitic stiffness matrix.
The creation of new test benches is also expected to significantly improve our testing ability of the
overall inertial sensor performance by better addressing these issues, both in terms of sensitivity and in
representativeness of flight experiments.
Before describing the planned upgrades, and with the aim of introducing the discussion concerning the
opportunities offered by the LNGS, we briefly describe the torsion pendulum facility currently
operating in Trento, with particular reference to the impact of environmental noise coupling on its
performance.
2.1.1 Torsion pendulum currently operating in Trento: apparatus description
A schematic of the torsion pendulum facility is shown in Figure 4a. A vacuum vessel accommodates a
prototype sensor surrounding the test mass and its 6-channel capacitive-inductive readout electronics.
The chamber is mounted on a platform whose inclination can be adjusted, while the whole facility sits
on a concrete slab partially isolated from laboratory floor. As shown in Figure 4b, the torsion
pendulum is composed of a hollow gold-coated Ti cube, with s = 40 mm sides and 2 mm wall
thickness, and a supporting Al bar, on which a stopper plate and optical mirror for independent readout
are mounted. The test mass is electrically isolated by a ceramic spacer, while the rest of the pendulum
is grounded through the torsion fiber, a Au-coated W wire nominally 25 µm thick and 1 m long. The
pendulum free torsional period is T0 = 515.1 s, with an energy decay time τ0 ≈ 1.35 ·105 s,
corresponding to a quality factor Q ≈1650.
The torsion pendulum hangs from a magnetic eddy current damper upper stage consisting of a W fiber,
with radius r ≈ 50 µm and length l ≈15 cm, supporting an Al disk surrounded by toroidal rare earth
magnets. The magnetic damper reduces the swing mode energy decay time to ≈ 70 s, without affecting
the torsional mode performance because of the cylindrical symmetry of its design. This double
suspension ensures that the main torsion fiber suspension point hangs essentially vertical, while giving
6
a negligible contribution to the torsional mode; it is rotationally much stiffer than the main fiber (the
spring constant scales Γ ∝ l r-4 for round fibers).
Figure 4 (a) Sketch of the experimental apparatus. (b) The test mass, its support, and the stopper plate that prevents the test
mass from hitting the sensor electrodes. The pendulum has a moment of inertia I = 338 ± 5 g cm2 and weighs 101.4 g. (c)
Schematic top view of the sensing electrodes; relevant dimensions are Rφ = 10.25 mm, and d = 2 mm.
The capacitive sensor, a Mo-Shapal prototype [6](Figure 5b), can be centered, based on the sensor 6channel capacitive-inductive readout, around the suspended test mass using a 5 degree of freedom
micromanipulator, while the fiber suspension point can be raised in z and rotated around z axis. The
displacement sensor angular sensitivity, ≈ 40 nrad Hz , is dominated by intrinsic thermal noise. The
pendulum motion is also monitored by a commercial autocollimator, with ≈50 nrad resolution for both
twist and tilt modes, allowing calibration of the sensor by exciting large twist motion and purposefully
tilting the apparatus by a few µrad.
The facility is equipped with a home-made electrostatic actuation circuitry that, as in the scheme
proposed for LISA actuation, is integrated with the sensor bridge electronics to apply audio frequency
and DC voltages to the sensing electrodes. Audio voltages are used for PID control of the pendulum
torsional mode, while DC biases are applied for electrostatic characterization of the sensor electrodes
7
and to measure the test mass charge. It is worth noting that, as required for LISA, the electrostatic
actuation circuitry does not add excess noise to the sensor sensitivity.
The pressure is kept below 10-5 mbar by a vibrationally isolated turbo pump; use of an ion-pump has
been avoided to prevent electrical charging of the test mass due to electrons coming from the pump
itself. The measured net residual test mass charging rate, ≈ +1 e per second, was occasionally balanced
using electrons emitted by the hot cathode pressure gauge.
The entire experiment is enclosed in a thermally insulated room (Figure 5a), whose temperature is
controlled by a constant temperature water bath that stabilizes the air circulating inside a heat
exchanger. The torsion fiber tube is covered with an additional layer of thermal shielding, giving
higher temperature fluctuation suppression. Temperature fluctuations at the fiber tube are suppressed
by at least a factor 20 between 0.1 and a few mHz, and by ≈ 10 below 0.1 mHz. The sensor housing,
electronics box, vacuum vessel, fiber tube, thermal room, and lab temperature are continuously
monitored by Pt100 thermometers. The magnetic field is monitored by a three-axis 10 nT resolution
flux-gate magnetometer, placed in the neighborhood of the pendulum. The capacitive sensor itself
measures the platform tilt: a tilt of the apparatus along the axis η (θ) will cause a translation of the
pendulum relative to the sensor along x (y), determined by the ≈ 1 m fiber length, ∆x ≈ ∆ η × 1m. Most
measurements are automated by dedicated software, and all experimental data and possible
environmental noise sources are continuously recorded by a (in-house) data acquisition and control
system.
Figure 5: (a) the experimental apparatus inside the thermally isolated room.(b) The molybdenum-SHAPAL capacitive
sensor prototype designed in Trento and the titanium hollow TM used in the decribed testing.
Torque sensitivity limit and coupling to environmental noise sources
The torque sensitivity of a torsion pendulum is intrinsically limited by the mechanical thermal noise
with power spectrum S Nth (ω ) = 4k BT Γ [11][12] and by the additive readout noise Sφread (ω ) , which can
2.1.2
ωQ
8
be converted in an equivalent torque noise through the pendulum transfer function
2 2
⎤
⎡ ⎛
⎞
⎛
⎞
1
ω
2
F (ω ) = ⎢Γ ⎜1 − ⎜⎜ ⎟⎟ ⎟ + 2 ⎥
⎢ ⎜ ⎝ ω0 ⎠ ⎟ Q ⎥
⎠
⎦⎥
⎣⎢ ⎝
2
−1
to give an overall torque sensitivity:
S 1N/ 2 (ω ) = S Nth (ω ) +
Sφread (ω )
F (ω )
2
where ω0 is the resonance angular frequency, Q is the quality factor, I is the momentum of inertia and
Γ is the torsional spring constant.
This can be exceeded, however, by pendulum coupling to environmental noise. In Trento, an excess
noise with respect to the expected thermal noise floor was observed and consequently an activity
devoted to the characterization and suppression of the disturbances has been performed . The main
identified sources are floor tilt, magnetic noise and temperature fluctuations (the single test mass
configuration, with a compact highly symmetric design and hollow test mass, makes the gravity
gradient noise negligible). The power spectral densities of those parameters as measured in the test
facility in Trento are shown in Figure 6.
Figure 6 Spectral densities of important environmental disturbance sources. The upper panel shows the fluctuations of lab
foor tilt, as measured by the sensor; tilt noise levels are also representative of the orthogonal axis η. The middle panel
shows the fluctuations of one horizontal component of the magnetic field near the pendulum. The bottom panel shows the
temperature noise at different apparatus locations; the curve labelled “sensor" refers to a transducer inside the vacuum,
9
while the “fiber tube" thermometer is attached to the outside of the tube. The temperature data were sampled at 1/40 Hz,
whereas all other sensors where sampled at 10 Hz; the white level at 2.5 mK /sqrt(Hz) is the readout noise. The effects of
the stages of thermal insulation are clearly visible.
We characterize the coupling of these disturbances to the torsional mode of the pendulum with
experiments in which the external source was modulated at a high enough level to induce a well
resolved signal in the pendulum twist[13]. The coupling to each disturbance term was estimated by the
ratio of the induced torque to the magnitude of the input parameter. Monitoring the environmental
noise levels under normal operation conditions permits an estimation of each systematic effect
contribution to the overall torque noise as summarized in Figure 7.
Figure 7 Torque noise contributions from external environmental couplings, shown with the thermal noise, readout limit,
and raw torque data. The sharp peak at 0.5 mHz is an artifact of a test mass charge measurement performed during the 101
hour run, that is analyzed here by cutting it into three 120 ks windows. The Hanning windowing process leaves an artificial
peak near the 2 mHz pendulum resonance, due to the imperfect knowledge of the torsion pendulum transfer function,
affecting the tilt subtraction procedure.
In order to improve our estimate of the low frequency force noise data, we subtracted the effect of
coupling to floor tilt from the raw experimental data. The correction is performed by measuring the tilt
components and then converting them into a torque by means of the measured tilt-twist coupling
coefficients. The instantaneous coupling torque, Fourier transformed into the frequency domain, is
then converted into a twist angle and this is converted back into the time domain to be subtracted from
the raw angular time series. This procedure does not involve a subtraction of noise spectra, but only a
time series subtraction based on calculation of the instantaneous torque. The Fourier transform is used
only to convert the calculated torques into twist angles, accounting for the torsion pendulum transfer
function. The tilt measured with the sensor is compared with the output of the optical autocollimator,
in order to verify that the subtracted signal is a real apparatus tilt motion, rather than a “fake”
displacement signal coming from the sensor itself. The tilt correction procedure, whose results are
compared with the raw data in Figure 2 and Figure 3, leaves a torque noise which is only a factor 3 to
10
5 over the thermal noise in the mHz region, and in particular below 20 fN m
mHz, with a minimum of 4 fN m
2.1.3
Hz between 0.3 and 10
Hz at 3 mHz.
Recent upgrades
Some modifications of this torsion pendulum facility have been recently realized (Figure 8):
• A motorized rotation stage was introduced to support the capacitive sensor; by modulating the
sensor rotation angle with respect to the torsion fiber equilibrium position, we will characterize the
entire spring-like coupling between the test mass and the displacement sensor
• An apparatus consisting of an UV lamp and optical fibers, provided by Imperial College of London
(UK), has been installed on the sensor, to control the test mass electrostatic net charge q. In
combination with the technique to measure q, already successfully tested, this will constitute a
significant test of the Charge Management System, one of the key features of the LISA/LTP
gravitational sensors.
• A set of 4 heaters and thermometers was installed on the external walls of the gravitational sensor
chassis, with the aim of inducing and monitoring low frequency oscillating 4-fold symmetric
thermal gradients, searching for a coherent torque exerted on the suspended test mass, through
mechanisms like radiation pressure fluctuations, radiometric effect and temperature depending
outgassing rate.
• replacement of the gold coated W fiber with an uncoated fiber, with the aim of increasing the
torsion pendulum quality factor Q, reducing in this way the mechanical thermal noise.
• An electrostatic shield between the torsional member and some exposed dielectric surface of the
capacitive sensor was installed and has successfully removed, to a large degree, the trans-twist
coupling observed in the previous experimental run
A new testing campaign will start soon, and the expected outcome should include precious
experimental information regarding performance optimization techniques and design drivers for the
new facilities.
11
(c)
(b)
(a)
(d)
(e)
Figure 8 View of the current implementation of the inertial sensor integrated in the torsion pendulum facility, with the most
recent and significant upgrades. The key features are (a):heaters; (b): thermometers; (c):electrostatic shields; (d):UV fiber
for test mass charge control; (e): DC motor for remote control of the sensor rotational position, properly shielded to limit
the magnetic and electrostatic torque induced on the test mass.
2.2 Test facilities upgrades
As summarized in the previous section, torsion pendulums have already proven to be very useful in
measuring the stray forces and stiffnesses arising in LISA-like displacement sensors and
characterizing specific stray force sources.
However, progress with the current torsion pendulum is limited in several ways:
- the pendulum is sensitive to torques, rather than the translational forces most relevant to LISA
drag-free control
- the level of force noise that can be reached with the pendulum, limited by intrinsic thermal
noise in the torsion oscillator, is roughly two orders of magnitude above the target force noise
for LISA
- the torsion pendulum has only a single degree of freedom (DOF) that is sensitive to force
disturbances, while a LISA test-mass will be free-falling (or very weakly controlled) in all 6
DOF
We propose here a new generation of facilities that will address these limitations and thus improve
both the sensitivity of our ground based measurements and the degree to which they are representative
of flight experiments.
In the next section we will briefly describe the following upgrades:
¾ Torsion pendulum for testing directly the translational degree of freedom
12
¾ Higer sensitivity torsion pendulum
Facility with more than one soft force sensitive degree of freedom
2.2.1
Torsion pendulum for testing one translational degree of freedom
The current torsion pendulum facility is a single mass configuration that is sensitive to torques on the
test mass, not the translational forces most relevant to our specific application demanding near perfect
free fall along a single translational axis. Consequently it would not detect dangerous stray net forces
and the characterization of the random forces depends on the conversion through a model dependent
arm-length. As this distance varies depending on the type of disturbance under investigation and is
unknown for un-modeled stray forces, it is desirable to have a configuration sensitive directly to the
net forces along the translational axis relevant for LISA. We will use a 4-mass torsion pendulum
(shown schematically in Figure 9 left), where, by displacing the test masses from the torsion axis, the
pendulum is sensitive to net forces applied on a test mass.
In order to produce a torsion pendulum configuration sensitive to net forces, it is necessary to displace
the test mass from the torsion fiber axis. This can be done by adopting a simple Cavendish type
geometry. However, as the effective armlength of the displaced test mass increases, the pendulum
quickly becomes more susceptible to environmental gravity gradient noise[5]. It is thus necessary to
have a high degree of symmetry to suppress couplings to fluctuating low-order multipole. Figure 9
(right) shows a four mass design with quadrupole compensation and the predicted force sensitivity
[18].
Figure 9 Left: Conception of the four-fold symmetric torsion pendulum. The centers of the four hollow, cubic test masses
(edge length 4.6 cm) are displaced 11cm from the torsion fiber axis. The smaller cubes displaced along the y-axis (vertical)
serve to reduce the gravitational quadrapole moment and provide a reecting surface for an optical readout of the pendulum
twist. The positions of the displacement sensor electrodes (outlined) are shown surrounding the mass on the lower right.
Like the gravitational wave strain measurement, the pendulum twist is sensitive to forces along the x-axis. Right: Predicted
force sensitivity of the pendulum shown on the left (using a 53 µm diameter tungsten fiber of length 1 m and quality factor
Q = 4000) as compared to a single mass configuration. The four-mass design is directly sensitive to net forces, while the
single mass curve is interpreted from its torque sensitivity using an e_ective arm-length of 1 cm.
13
Figure 10 The vacuum enclosure designed for accomodating the 4-masses pendulum that will be operating in Trento in the
second half of this year.
The experimental apparatus (Figure 10), that will accommodate the first version of the 4-masses
torsion pendulum, is under construction and the first experimental run is expected in Trento the second
half of this year.
We note that an ideally 4-fold symmetric pendulum, with characteristic armlength R, would
have a response to gravity gradients, generated by a mass at distance r, that is proportional to
R4/r5. As such, increasing the armlength R to increase the pendulum force sensitivity is done at
the price of an increasing susceptibility to gravity gradient noise. Thus, for the 4-mass design
shown here, and for any other future multiple mass pendulum, an environment that is “quieter”
from the standpoint of mass motion can ultimately allow more sensitive measurements of the
intrinsic force disturbances relevant to LISA.
2.2.2
Higher sensitivity torsion pendulum
The pendulum thermal noise so far achieved leaves upper limits on stray force noise well above the
requested in-orbit performance levels. The thermal limit for force noise in a torsion pendulum
is proportional to (Φ/(C*B))1/2, where Φ is the loss angle and is equal to Q-1, B is the tensile breaking
strength (in Pascal) of the wire material and C is a safety factor (C<1) that expresses the percentage of
the breaking stress at which the wire is loaded.
The optimal material to realize the suspension should be a low dissipation material with a high and
reliable (C not too small) breaking strength.
Within the INFN LISA PF collaboration the group in Perugia has already performed a preliminary
study with the aim of singling out the fiber material. A study of the mechanical and dissipative
properties of short wire samples in three different materials has been carried out: Steel (C85),
Tungsten and fused silica. Fused silica has been identified as the most promising candidate.
14
The breaking strength of Fused Silica has been measured at: B(SiO2)=4.05±0.55 GPa. This value must
be compared with the breaking strength of C85 wires: B(C85)=2.90±0.02 Gpa and that of Tungsten:
B(W)=1.92 GPa. The Fused Silica breaking strength is the highest, but it shows also the largest
fluctuation. This is mainly because the Fused Silica breaking strength is dominated by surface defects.
For this reason the value of C*B has to be considered the same for the three materials.
The SiO2 loss angle is shown in figure 11 versus the frequency of the mode. It is clear that the Fused
Silica loss angle is more than two orders of magnitude lower than the C85 and W.
The different frequency behavior is due to the different thermoelastic contribution.
In figure 11, two different curves for Fused Silica fibers are shown. The curve with solid stars is
measured hanging the fiber with an intermediate mass to reduce the suspension recoil losses. Since a
small difference is measured between the SiO2 with and without the insulating mass, a low recoil loss
contribution is expected in the measurement apparatus.
From these studies the pendulum thermal noise of a suspension made of Fused Silica fibers is expected
to have a thermal noise more than one order of magnitude better as compared to C85 and Tungsten.
Figure 11 Measurements of the loss angle versus frequency. Squares: Tungsten wire. Plus: C85 wire. Circles: Synthetic
fused silica. Stars: Synthetic fused silica with isolation bob. The error bars are neglected because they are smaller than the
experimental point symbol size.
15
Improvement and automatization of the facility for the thin (from 20 to 100 micron) silica fiber
fabrication are in progress, and the efforts will be then focused to the identification of an electrical
conductive coating compatible with a high torsional quality factor.
The expected improvement in term of force sensitivity thermal limit with a Q of a million is shown in
the following figure, both for the 1 mass and a 4-masses configuration: a torsional Q of a million
would decrease the pendulum thermal force noise would decrease to order of 10 fN/root(Hz), a big
improvement that would leave us within an order of magnitude of the LISA goals.
Taking full advantage of the decreased level of thermal noise in a higher Q pendulum requires
correspondingly lower levels of all other environmental disturbances, including the temperature,
seismic activity, and gravity gradient effects already mentioned.
Figure 12: Force sensitivity thermal limit, both for 1-mass and 4-masses torsion pendulum configuration, with the tungsten
fiber currently used in Trento (Q of about 3000) and with a silica fiber with torsional quality factor of a million.
2.2.3
Facility with more than one “soft” force sensitive degree of freedom
The pendulum has a single “soft” or sensitive degree of freedom, whereas a free-falling inertial
references mass is free-falling (or controlled with a low frequency suspension) along all degrees of
freedom. To better represent the flight conditions in which the test mass is sensitive to forces along
different degrees of freedom, we are studying suspensions which have multiple degrees of freedom
with very low effective elastic constants and thus high sensitivity.
Specifically, a test facility with many soft force-sensitive degrees of freedom allows for:
• Measuring forces and stiffness simultaneously along different degrees of freedom
16
• Closing feedback loops on one DOF and measure the effects along another one
• Closing feedback loops simultaneously on more than one a degree of freedom
Advantages with respect to the single DOF test bench:
• More effective in identifying and debugging spurious effects and non -linearities
• Allows for testing of actuation cross talk with closed feedback loops: in particular, it allows to
measure the residual disturbance along the sensitive translational axis when we close the
control loop along the ϕ rotation (because is the control loop that will be used also in LISA)
• Allows for measuring the stiffness and cross-stiffness with closed feedback loops
• Verification of the dc stray voltage compensation tecnique simultaneously in different DOF
• Verification of the compatibility of the charge measurement by means of a dithering voltage
applied in terms of noise induced in x.
Disadvantages of the many DOF platforms, are the lower sensitivity, due to the high vertical
acceleration gravitational background, and the higher complexity.
At least two DOF should be translational degrees of freedom, and at least one of those should be very
weekly coupled to the rest of the system. The mechanical problem becomes to design suspensions with
very low stiffness and very low friction, with a resonant fundamental frequency in the range of few
mHz. Stability against temperature fluctuations and seismic noise are also a requirement of the final
design. Additionally, as with the four-mass pendulum discussed in section 1.1.1, the proposed multiple
DOF pendulums involve relatively large arms, and thus the design should maximize symmetry and
seek the environment with a minimum of nearby gravitational disturbances.
Several geometries are being investigated; in Figure 13 is shown the roto-traslational pendulum, that
has two soft degree of freedom, one rotational, around the vertical axis of the test mass, and the other
traslational, around the main vertical axis of the system. One of the most promising geometry is the
Roberts linkage, shown in Figure 14.
Figure 13 : Roto-traslational pendulum, 2 soft DOF
17
Figure 14: Roberts linkage [18]
The design of the Roberts linkage is such that in the ideal case (massless frame, and ideal constraints)
the suspension point P is only allowed to move in a plane; ideally, this device has a null resonant
frequency. More realistically, Roberts linkages have been already been designed and tested with
resonant frequencies in the range of the tens of mHz.
Figure 15: An implementation of a bidimensional Roberts linkage
Other schemes include a vertical soft degree of freedom obtained with leaf springs. In this case, the
leaf springs are pre-bent, in order to lower the vertical stiffness, still being able to hold the weight of
the test mass (Monolithic Geometric AntiSpring, MGAS)[20]. Systems like this have been tuned to
vertical frequencies of about 100 mHz.. This kind of vertical suspension is presently being used by
the TAMA interferometric antenna, and it has been chosen as the vertical suspension of the isolation
test tower being built in Firenze by our group.
18
Figure 16 Scheme of the MGAS leaf spring.
Soft horizontal, linear degrees of freedom can also be obtained with two orthogonal folded
pendula,(Watt’s linkage)[21][22]. A folded pendulum is made by assembling a positive pendulum and
an inverted pendulum: the restoring force is gravity, with a negligible contribution from springs at the
joints. These devices have been tuned down to about 10 mHz.
Figure 17: Scheme of Watt's linkage
Figure 18: (Right) Horizontal accelerometer based of the folded pendulum scheme; (Left) One of the elastic joints [22].
The final choice for the multi degree of freedom suspension will be taken on the basis of a careful
finite element analysis, which will also determine the internal modes of the system.
19
Finally, also electrostatic suspensions with more than one degree of freedom are being investigated:
they seem less promising, because of the relatively high weight of the test mass, and because of the
possible interactions with the capacitive sensors.
2.3
LNGS
As discussed above, environmental conditions appear to be the limiting factor of our current pendulum
performance and would represent even more serious problems for the next generation of higher
sensibility torsion pendulum and of multi-degree of freedom suspension systems.
A “quiet” site would significantly increase the probability to achieve the best performances and would
reduce the efforts needed to reduce the environmental disturbances effects in term of:
• shielding (temperature, magnetic field)
• active control (temperature, floor tilt)
• designing an apparatus with a high degree of immunity (Newtonian gravitational coupling).
These strategies look particularly challenging because of the very low frequency measurement band
and because they would increase the constraints onto the overall facility design.
The most relevant environmental noise sources are:
• Seismic noise: motion of laboratory can induce a stray torque on the pendulum through several
mechanisms [1].
Any tilt motion of laboratory floor can induce a torque on the pendulum for example through
any position dependent torque induced by the gravitational sensor itself, or the effect of linear
cross-coupling of suspension point tilt into pendulum twist. Linear micro seismic noise of the
suspension point of a torsion pendulum is a source of noise because of coupling of swinging
modes into the torsional mode. Also vertical microseismic noise can couple into pendulum
twist via nonlinearities in the fiber response to vertical spring-type modes of the pendulum[1].
In order to reduce these stray effects, damping of the simple pendulum mode and the vertical
spring-like mode should be provided together with a high degree of symmetry of the masses
suspended [4]. Both strategies add constraints to the pendulum design, and the high symmetry
requirement means usually to make the pendulum heavier at the expense of its sensitivity.
•
Gravity gradient noise
A source of torque noise for a torsional pendulum is the coupling of mass multipole moments
of the pendulum to gravity gradient fields. In order to reduce this effects the strategies is to
reduce the moments of the pendulum and to place the experiment where the gravitational fields
gradients fluctuations are small [3][5].
The compensation of the moment of the pendulum poses serious constrains on the overall
design and requires adding mass, at the expense of its sensitivity. Moreover, even if one
nominally cancels the low order moments “by design,” small imperfections will still create
coupling to gravity gradients fluctuations, thus necessitating more demanding machining and
assembly tolerances requirements.
As these effects are related to changing ambient mass distribution, one expects that lower
seismic noise, better environmental temperature stability and better isolation from weather
condition will produce a more benign “gravitational” ambient.
20
•
Temperature
On a large scale a variation in temperature can alter the geometry of the building, and itself
induce a motion of the ground and Newtonian noise.
Since the best approach to eliminating environmental disturbances in a measurement requires reducing
the effect at its source, the choice of the experimental site is important and must be made with
care.Low microseismic and newtonian noise, stability in temperature on time scales exceeding a day
qualify the site for the experiment, and the environmental stability of an underground site has been
already choosen as the most promising option with respect to these criteria for gravitational wave
detectors and torsional pendulum for experimental gravitational [16][23][24][1]. Laboratori Nazionali
del Gran Sasso are potentially very good candidate..
Measurements over many years show that, at low frequencies, tilt and microseismic noise is about 5
times lower than in Trento Laboratory.
x 10
-7
G ra n S a s s o M a y 2 0 0 0
4
A c celeration acc 2+acc1 [g]
3
2
1
0
-1
-2
-3
-4
0
50
100
150
200
250
T im e [ h ]
Figure 20: Tilt mesurements at LNGS; solid tides are visible
.
2
Figure 19 Tilt PSD at LNGS; the vertical scale can be read in rad / Hz
acceleration and inclination.
because of the coupling between horizontal
21
Freq
Hz
10-2
10-3
10-4
Trento
µrad/√Hz
0.04
0.3
3
Gran Sasso LNGS
µrad/√Hz
0.05
0.1
0.6
Figure 20 Comparison between tilt measurements in Trento and at LNGS
Figure 21 Tilt measurement in Trento (see also Figure 6).
More difficult is the assessment of Newtonian noise, and its measurement requires high sensitivity.
Models [16] suggest that a solid rock cave guarantees a gravity gradient noise level that is lower by
several order of magnitudes than that on the surface, and the equipment we propose to set up
at LNGS will be sensitive enough to verify the models.
Finally, the temperature is stable on a seasonal time scale.
Following these considerations, LNGS can be considered a very good site for the experiment.
22
2.3.1
Requests to the LNGS
After a visit to LNGS, we have identified a suitable area defined Nodo C, close to the interferometer
GIGS, as shown in the map.
Nodo C
Figure 22 Map of the LNGS; Nodo C is shown.
A number of upgradings of the Nodo C are anyway required:
•
•
•
an effective insulation of the area from the acoustic noise, due to human activity and
venting of the tunnels: two adjacent brick walls, separated by a layer of insulating
material; doors and windows must be designed accordingly.
decoupling of the lab floor from the floor of the halls: at present the floor is a continuous
concrete layer, that couples the entire area through the microseism due to human activity;
decoupling can be done by simply removing the floor in Nodo C.
remove the layer of pebbles under the floor and down to the solid rock; this is required in
order to grant a firm coupling to the rock, and to remove a relevant source of gravity
gradient noise.
In addition to all these specific requests, there is also the standard request of electric power (15 Kw
required), venting. A small crane will also be needed in the experimental area.
All the jobs, to be performed under the responsibility of the Laboratori, will be done in close
collaboration with our team.
In the area surrounding Nodo C a rather large flow of water is expected; since water flow in the
tunnels is stable on a seasonal time scale, it should limit the experiment. We are waiting for more
detailed information on the water flow, as soon as it will be available.
23
The experiment should last at least a few years, until 2006, with an highly advisable extension into the
following years (through 2008), aiming to address specific issue related to LISA (extension not yet
approved by INFN). Installation will last for a few months, while decommissioning will take a few
weeks; each run will last between one and two weeks, in presence of the experimenters; no emissions
are expected, apart from the exaust gases from rotary pumps.
The apparatuses will be constructed at the parent Laboratory and only transferred to LNGS upon their
completion; at that point we will request access to the mechanical and electronics workshops for minor
adjustments of the components and for regular maintenance. One or two rooms at the site in Assergi
are requested for the personnel present during the runs.
3
3.1
OPTICAL READOUT SYSTEM
Introduction
The usual solution for satellite drag free control, adopted as a reference solution also for
LISA, is the usage of capacitive sensors. In this case, the sensor is essentially a capacitor where one
of the plates is the surface of the test mass while the other one is connected to the spacecraft. In
particular, such a device has been developed, and successfully tested by the group of Trento
University and will be tested in flight in the technology demonstration mission LISA-Pathfinder. The
main disadvantage of capacitive sensors is the need for a very small free gap between the two plates
and then between test mass and spacecraft. The typical value for the gap is below 1 mm, while a gap
larger than a few mm is hardly compliant with the required sensitivity. The small free gap turns out in
strong sensitivity to net charge deposited (for example due to cosmic rays) on either the test mass or
the spacecraft. This puts severe limitations to the maximum acceptable rate of charge deposition and
imposes frequent discharge by UV flash lamps. Then it would be very helpful to substitute the
capacitive sensor with some alternative position sensor allowing increasing the gap up to several cm,
and relaxing consequently the specifications. The obvious alternative solution is some kind of optical
sensor [26,27,28]. It is worth noting that increasing the gap poses some technical problems because it
becomes difficult to use electrostatic actuation to control the position of the test mass and alternative
solutions, like radiation pressure actuation, should be investigated. In any case, an optical readout
system, as integration of the capacitive one, could be very useful also with a small gap. The main
point is the addition of some redundancy. Providing a back-up solution in case of malfunctioning of
the capacitive system after the launch, gives a considerable risk-reduction for the mission.
Furthermore the optical sensor is potentially more sensitive that the capacitive one, and less sensitive
to cross-couplings between (electrostatic) actuation and (capacitive) measurement.
Goal of this study is the direct application of an optical readout system to the LISA project
whose design is already quite advanced and plans the launch of a demonstration mission (LISApathfinder) in 2008 and the final one in 2012. Therefore, we can only take into account solutions that
are compatible with the actual design or require only very small modifications, while any substantial
modification would imply an unacceptable time dilation. In the following we will not examine
solution requiring increasing of the gap and radiation pressure actuation because not suitable for
24
LISA, although these subjects are relevant from the scientific point of view having in mind a possible
“follow on mission” [30].
We will only analyse in detail an optical readout system, based on the usage of optical levers
and position sensors. To evaluate the possibility to use such a sensor for LISA, we must verify two
main aspects:
1) The sensor must fulfil specifications on sensitivity and back action.
2) The sensor must fit in the present design of LISA.
Next section is devoted to the operation principle and to the estimation of the most relevant noise
sources limiting the sensitivity of the device. In next section, we report the experimental results
obtained so far in bench-top experiments at the Napoli section and describe next steps of
experimental activity. A section will be devoted to the feasibility of the integration of the optical
sensor in the present design of LISA; in particular, we will analyse and compare some alternative
solution pointing out advantages and potential problems.
In the end, we propose a possible planning for the development of the technology taking into account
the schedule of Pathfinder and LISA.
3.2
The optical readout system
Due to the extreme sensitivity required for GW detection, the specifications for the position sensor
are very stringent for both the displacement sensitivity and the back action, that is the unwanted force
that the sensor itself might apply on the test mass. According to the design of the antenna [26] the
required sensitivity along the most sensitive DOF (the interferometer optical axis) is 10-9 m/Hz1/2,
while the angular sensitivity is ~ 5·10-8 rad/Hz1/2. For the spurious forces, the upper limit is, 6·10-15
N/Hz1/2. These specifications, in principle not too severe for an optical sensor, become critical if we
take into account the low operation frequency (10-4 – 10-1 Hz) where thermal drifts of both mechanics
and electronics are dominant respect to sensor intrinsic noise.
Amongst the sensors proposed so far, some are based on interferometric readout; this solution,
which in principle can surely reach and overcome the necessary sensitivity has the disadvantage of
being, in general, more complex with respect to other kind of optical sensors. Furthermore, at the very
low frequency of interest, it is very likely that, rather than by the limiting noise of the detector (shot
noise for an optical sensor), measurement sensitivity will be limited by other sources of noise, like
thermal drifts in both mechanics and electronics. For that reason, we prefer, as reference solution, a
system based on optical levers and position sensors rather than interferometric sensors that are in
principle much more sensitive than required, but with the disadvantage of higher cost and complexity.
The principle scheme of an optical lever sensor is very simple. A laser beam, is sent trough a
single mode optical fibre, to the surface of the test mass. The reflected beam is detected by a quadrant
photodiode (or a position-sensing device (PSD)) that measures its transverse displacement.
The displacement (∆xs) of the spot on the sensor due to a longitudinal displacement (∆x) of the
test mass, depends on the incidence angle θ as:
25
∆x s =
∆x
sin(θ)
(1)
The displacement due to a rotation ∆θ around an axis orthogonal to the normal of the surface
depends on the optical lever arm h:
∆x s = 2 h ⋅ ∆θ
(2)
With a suitable combination of beams and sensors, all the six DOF of the test mass can be
detected. The optical fibre is necessary to reduce the beam jitter that would otherwise be dominating.
In order to evaluate the sensitivity, the first point to consider, it the noise that such a sensor can
reach in the frequency band of interest for LISA (from 100 µHz up to 100 mHz). As for any optical
sensor, the ultimate limit is shot noise that is expressed, for a position sensor, by [4]:
⎛ 633 nm ⎞
~
x sn ≈ 2 .8 ⋅ 10 −11 ⎜
⎟
⎝ λ ⎠
12
⎛ 1mW
⎜⎜
⎝ P0
⎞
⎟⎟
⎠
12
12
⎛ 0 .78 ⎞ ⎛ L ⎞
⎟⎟ ⎜
⎜⎜
⎟
⎝ η ⎠ ⎝ 1mm ⎠
⎤
⎡m
⎢⎣
Hz ⎥⎦
(3)
where λ and P0 are wavelength and power of the laser, η the quantum efficiency of the detector
and L the measurement range (that depends on the spot size if a quadrant photodiode is used and on
the detector size for a PSD). We can see that, with reasonable numbers for the relevant parameters, the
shot noise limit is well below the specifications.
26
LISA PF
Another important source of noise is the current noise In of the trans-impedance amplifier
used to read the photodiode current (assuming photoconductive operation). This imposes a limit
sensitivity given by:
12
~
L ⋅ In (f )
⎛ 0.43A / W ⎞⎛ 1mW ⎞⎛ 1mHz ⎞ ⎛ L ⎞
~
⎟
⎜
xI =
≈ 4 ⋅ 10 −10 ⎜⎜
⎟⎟⎜
⎟
⎟ ⎜
⎟⎜
α(λ ) ⋅ P0
⎝ α(λ ) ⎠⎝ P0 ⎠⎝ f ⎠ ⎝ 1mm ⎠
⎡m
⎤
⎢⎣
Hz ⎥⎦
(4)
~
Where α(λ) is the responsivity of the photodiode and we assume In (f ) = 10 −13 / f A/Hz1/2
that is the typical value for a widely used device (Analog Devices OP27EP). As we can see, also in
this case the sensitivity depends on laser power and measurement range (i.e. spot size). This source
of noise is likely the limiting factor for a sensor based on an optical lever. Other relevant noise
sources are mechanical vibrations and thermal drifts, which depend on the actual set-up and can be
reduced with a very rigid mounting and a thermal stabilization.
From the point of view of the back action, this is due essentially to the fluctuation of the
~
~
light power impinging on the test-mass ( F ≤ 2 ⋅ P / c ); assuming a conservative upper limit of
~
F ≤ 6 ⋅ 10 −16 N / Hz , we get for the relative power stability the limit:
~
⎛ 1mW ⎞
P
⎟⎟
≤ 10 −4 ⎜⎜
P0
⎝ P0 ⎠
⎡1
⎤
⎢⎣
Hz ⎥⎦
(5)
This limit is achievable by actively stabilizing the laser power.
3.3
Preliminary results
To verify the considerations of the previous section on the potential sensitivity of the optical sensor,
we performed some bench-top measurement at the Napoli section.
The set-up is shown in Figure 25. The source is a He-Ne laser coupled to a single-mode
optical fibre. The output power is about 340 µW. The laser beam is reflected by a piezoelectric
tilting mirror, split in two by a cube beam-splitter and detected by two quadrant photodiodes (QPD).
In this way, we can perform differential measurements of the displacement of the beam on the two
detectors and cancel the common effects due to thermal drifts and vibrations of the fibre and mirror
mounts. The only point where the symmetry is lost is the beam splitter itself, since any
displacement of this component will only affect the reflected beam leaving unaffected the
transmitted one. The entire set-up was closed in a box to reduce the effect of air movements and the
temperature variations. The spot size on the QPDs was about 0.5 mm.
LISA PF
Figure 23: Experimental set up. A fiber coupled He-Ne laser beam is reflected by a piezoelectric tilting mirror,
split in two by a cube beam-splitter and detected by two QPDs. The two sensors are mounted on micrometric
translators for calibration and fine positioning. The signal of QPD2 can be used in a closed loop configuration
for actively stabilizing the beam on position by acting on the PZT mirror.
3.3.1 Sensitivity measurements
We performed the measurement in two operation modes. In the first case, we just centered the two
QPDs (that are mounted on three DOF micrometric translators to adjust transverse positions and
optical path) on the beams. The difference of the signal of the two QPDs should provide the
incoherent sum of the noise of the two sensors and gives our limit sensitivity, while the two
independent measurements provides a measurement of the beam motion that is common for the two
sensors. A sinusoidal angular movement (at 50 mHz) is imposed to the PZT tilting mirror for
relative calibration of the signals of the two QPDs, while absolute calibration is obtained with the
micrometric translators. As usual, the position of a beam is obtained by taking the difference of the
signals of pairs of sectors of the QPD normalized to their sum in order to cancel the effect of
amplitude fluctuations.
In the second measurement mode, we used the signal of one detector (QPD2) to actively
stabilize the position of the beam by acting on the PZT mirror, and take the displacement as
measured with the other detector (QPD1). The servo-loop bandwidth is about 5 Hz.
The result of a typical differential measurement is shown in figure 26 ((a) spectrum and (b)
time domain). As we can see by looking at the calibration peak (50 mHz), in the difference signal
the common mode motion of the beam is cancelled by more that two orders of magnitude while the
wide band spectrum is reduced by a smaller amount. That means that we have reached a noise level
that can either be detector intrinsic noise or relative motion due to asymmetries in the set-up
(mainly BS) or uncorrelated drifts in the mounts of the detectors. In figure 26, the electronic noise
level, measures as the differential signal with the laser switched off, is also shown compared with
the expected value (solid line) computed according to formula (4).
The sensitivity is about 10-8 m/Hz1/2 at 1 mHz (one order of magnitude above the
specifications) and fall below 10-9 m/Hz1/2 above 10 mHz.
28
LISA PF
(a)
(b)
Figure 24 : (a) displacement noise as measured with the two QPD and from their difference compared with the
measured and expected electronic noise. The differential measurement is within the specification above 10 mHz.
(b) beam displacement measured with the two QPDs and their difference (time domain).
Figure 25 : : (a) displacement noise measured with the two QPDs when the beam is actively stabilized on QPD2. The
displacement measured with QPD1 is within the specification above 10 mHz. - (b) beam displacement measured with
the two QPDs when the beam is actively stabilized on QPD2. The residual displacement measured by QPD1 is less than
10 nm on a time of 6000 s.
The closed loop measurement is reported in figure 27 ((a) spectrum and (b) time domain). As
expected, the error signal (that is the one of the QPD used for the servo-loop) is reduced by a large
amount, while the other one (that is the actual sensor noise measurement) reaches a value similar to
the one obtained with the differential measurement. Also in this case, the sensitivity is close to the
one measured in open loop and reaches the specification above 10 mHz. Looking at the time
domain, we can see that the residual beam drift, as measured with QPD1, is less than 10 nm.
From the open loop measurement with the single QPDs (figure 26), and taking into account
a total optical path length of 235 mm, we can also put an upper limit to the angular jitter of the
beam at the output of the fiber to ~ 1.7·10-7 rad/Hz1/2 at 1 mHz and 1.7·10-8 at 10 mHz.
29
LISA PF
Figure 26: – Experimental set up for laser power control, the error signal is detected with QPD2 (taking the sum of the
four elements of the quadrant). The signal is filtered with an analogical integrator and used to drive the current of the
laser diode
3.3.2 Laser power stabilization (back action)
Laser power stabilization is another relevant point, studied experimentally to verify the
specifications posed in (5). In this case, the laser source is a laser diode coupled to a SM fiber
optics, emitting up to 10 mW at 635 nm. The laser power is regulated by adjusting the diode
current, while the laser head is temperature stabilized by a peltier cell to avoid thermal drifts, due
to current variations, that induce frequency shifts (not relevant in our case but important for
spectroscopic applications) and, most important, jumps of longitudinal laser modes that are a
source of amplitude noise. In our set-up (figure 28) we use as a sensor one of the QPDs. The error
signal is the sum of the current of the four elements of the detector. The power is stabilized with
respect to a voltage reference corresponding to a total power on the sensor of 1.25 mW. The
bandwidth of the integrative servo-loop is 5 Hz.
In figure 29, it is shown the spectrum of laser power fluctuations in open and closed loop,
compared with the specification defined in eqn. (5). As we can see, the open loop noise is above
specifications by more than two orders of magnitude, while with active stabilization there is a large
safety margin in the whole bandwidth.
30
LISA PF
Figure 27 : Spectrum of the laser power fluctuations in open loop (blue line) and closed loop (red line). With active
stabilization, the noise is well below the specification (horizontal line) at all frequencies.
3.3.3
Present status of experiments and future developments
Concerning the displacement sensitivity of the optical sensor, the measurements performed so far
are in agreement with specifications above 10 mHz. At frequencies between 1 and 10 mHz, the
noise is above the specifications by a factor < 10. In the present set-up, the sensitivity in not limited
by shot noise and amplifier current noise, but most likely by real relative displacements of the two
beams. Actually, the electronic noise is already within the specifications and it can be further
reduced either by increasing the power of the laser beam or reducing the spot size (see eqn. 4).
To improve the measurement with the aim of direct demonstration of the specifications, we
have prepared a new more rigid set-up that allows performing differential measurements with a
symmetric optical layout (within machining tolerances). The system is shown in figure 30. The
entire bench is obtained by a single stainless steel block where the interfaces for fiber couplers and
sensors are machined. At the center of the box, are placed mirrors that reflect the beams as the
surfaces of the test mass.
31
LISA PF
Mirrors
Optical
fiber
Position
sensor
3.3.3.1.1
Fiber
coupler
Errore.
Figure 28 : New rigid set-up for differential measurements.
Assuming an asymmetry of 2 mm, due to machining tolerances, and a thermal expansion
coefficient for steel of α ≈ 2 ⋅10 −5 we get, in presence of a temperature fluctuation δT , a
displacement noise δx T ≈ 4 ⋅ 10 −9 ⋅ δT . By assuming fluctuations amplitude of 10-2 K/Hz1/2 with a
white spectrum in the whole band, it should be possible to go down to the intrinsic noise of the
sensor.
It is worth of noting that the specification of 10-9 m/Hz1/2 refers to the most sensitive DOF
(that is the displacement along interferometer optical axis), while the ones for orthogonal axes are
relaxed by more than one order of magnitude and then are already fulfilled with the present
preliminary measurements. Furthermore, from (1) we can see that between test mass displacement
and spot displacement on the sensor there is a factor depending on the incidence angle (1/sin(θ))
that helps in further relaxing the specs. In the end, it looks quite likely that the proposed optical
sensor can reach the sensitivity required for LISA, even if a direct experimental demonstration is
necessary.
As already pointed out, it looks that the back action due to laser power fluctuation is not a
problem.
Once this set of preliminary measurements will be completed, it will be useful to proceed to
the development and test of a sensor operating on six DOF, to be integrated and tested in the new
torsion pendulum being completed in Trento or in the one that will be developed at Gran Sasso
National Laboratories. In this way, it will be possible to verify, in a configuration as close as
possible to the real one, the performance of the system in view of his integration in LISA.
3.3.4
Compatibility of the optical sensor with the LISA design
As already mentioned, once it is demonstrated that the OS can reach the required sensitivity,
still remains the problem of his integration in the actual design of LISA, that in already quite
32
LISA PF
advanced. Although the launch of the final mission in scheduled only for 2012, in 2007-2008 there
will be a technology demonstration mission (LISA-Pathfinder). The design of the payload of this
mission is already almost frozen and any further substantial change with respect to the technology
that will be tested on flight must surely be rejected. This holds, in particular, for the geometry of the
electrodes used for capacitive sensing and electrostatic actuation. Of course, this would be different
in case the present design shows serious drawbacks, but in this case, a more general reconsideration
of the whole mission would be necessary, with consequent unavoidable time delays.
Assuming, in the contrary, that the inertial sensor of LISA will be essentially very close to
the one to be tested on LISA-Pathfinder, there is the problem of studying the integration of an
optical readout system. In figure 31, it is shown a scheme of the IS (engineering model). As we can
see, the space available for the passage of light beams that should reach the test mass is extremely
small; this is the main problem to solve in order to find a reliable solution. A second relevant
problem is how to bring the optical beams in the vacuum chamber and let the reflected one to exit
(unless the detectors are placed inside the chamber itself).
Figure 29 : Scheme of the inertial sensor engineering model
3.3.5
Optical paths
For the optical paths, we studied several possible solutions. One possibility is to reach the
test mass using the space between the electrodes. In this case, the maximum aperture is 1.6 mm.
This imposes a maximum beam waist for the gaussian beams of about wo = 300 µm, (so that the
aperture is larger than 5 beam diameters). Assuming a total optical path inside the vacuum
chamber of p ~ 100 mm, we need an ungula accuracy in the positioning of the reflecting mirrors of
∆θ= wo/p = 3·10-3 rad ~ 0.5°. In figure 32, it is shown a possible solution of this type. As we can
see, we need a few reflections for both the input and the reflected beams (the last can be reduced or
33
LISA PF
eliminated by placing the sensor inside the vacuum chamber. The most delicate point is the correct
placement of the reflecting mirrors (not shown in the figure). Obviously, since there are many
reflections, the angular accuracy for each mirror should be defined taking into account is actual
position along the optical path. Furthermore, also the mounting looks quite difficult, even if
possible in principle, because the space for the mirrors is quite small (8 mm) for the vertical DOF
(z) due to the caging mechanism.
y, θx, θz
z, θx, θy
Figure 30 : optical paths with reflections on small mirrors and passage of the beams in the space between the
electrodes.
A second possibility is to pass, for the x and z faces of the test mass, through the frame
supporting the electrode housing, and using the electrodes themselves as mirrors to direct the
beams on the test mass. In figures 33 and 34, it is shown a possible solution based on this idea and
in which the beams enter the vacuum chamber through optical fibers.
34
LISA PF
(a)
(b)
Figure 31 : optical sensor with optical fibers and diodes directly mounted in the electrode housing. (a) front view, (b)
upper view.
The fiber output coupler, that is a pigtail with a green (graded index) lens, is directly
coupled on a suitable interface on the electrode housing. For the x and z faces, the output beam is
detected, after two reflections on the electrodes and one on the test mass, by a position sensor (a
Hamamatsu PSD with 4x4 mm2 sensitive area in the drawing) that is also placed on the electrode
housing. The incidence angle is 74° and 75° respectively. For the y surface, the fiber output beam
is directly impinging on the test mass thanks to a hole already present on the central electrode for
different reasons. In this case, the incidence angle is only 10°. With this configuration, the angular
35
LISA PF
tolerance can be considerably relaxed. The holes for the passage of the beams can be enlarged to φ
= 3 ÷ 4 mm. The limiting factor is now, for the beams reflected by the electrodes, the minimal
distance between the beam themselves and the edge of the test mass (dm ~ 2 mm). Still assuming
w0 = 300 µm, posing a lower limit for distance from the edge of 4·w0 = 1.2 mm, and considering an
optical path from the fiber output of p = 65 mm, we get ∆θ = (dm-4·w0)/p ≈ 1.2·10-2 rad ≈ 0.7°. For
the y beam the tolerance is still larger because the path is shorter (10 mm) and the limit is posed by
the diameter of the hole (∆θ ≤ (φ-4·w0)/p ≈ 0.18 rad ≈ 10°).
This second approach gives several advantages with respect to the other one, which uses
the space between the electrodes. A first point is the relaxed specification for the alignment of the
beams. A second aspect is that all the optical beams enter in the electrode housing on the same side
and, most important, that the upper and lower parts of the vacuum chamber, where the caging
mechanism is, are not used. As a last point, the beams reflected by the x and z faces of the test
mass, exit form the electrode housing converging toward an optical window, even if they don’t
center it completely, and this simplifies considerably the set-up if it turns out that the sensors must
be placed outside the vacuum chamber (this aspect will be analysed in a while).
36
LISA PF
Figure 32: Solid model of the sensor with optical fibers and photodiodes directly attached to the electrode housing.
37
LISA PF
3.3.6
Selection and positioning of the sensors
As previously indicated, we have two possible position sensors: quadrant photodiodes and
(QPD) or position sensing devices (PSD). The first is a photodiode with the sensitive area divided
in four elements, separated by a (non sensitive) gap ranging from 50 to 300 µm according to the
selected model, while the sensitive area ranges between few mm and few cm. The PSD is a
photodiode where there is a resistive layer between the photosensitive area and the electrodes; in
this way, the current in the single electrodes depends on the position of the incident beam across
the photodiode. A QPD is generally (with the same responsivity α(λ), power Po and beam
diameter wo), more sensitive than a PSD [29]. For a QPD the displacement sensitivity can be
approximated as dI/dx ≈ α(λ)·Po/wo while for the PSD we get dI/dx ≈ α(λ)·Po/L, where L is the
detector size. On the other side, when the spot size is comparable with the gap, with the QPD we
loose a considerable amount of light. Furthermore, the lost power, and then the total current of the
detector, depends on the beam position. This fact can become a source of displacement noise and
mainly it can introduce force noise (by radiation pressure) if we use the total power of the QPD for
stabilizing the laser current. In the end, the response of the sensors depends on shape and
symmetry of the incident beam. In the case of the PSD, although the intrinsic sensitivity is lower,
total current and sensibility are, in a first approximation, independent on shape and position of the
beam reducing the problem of back action at a negligible level. A residual dependence on shape
and position can arise due to disomogeneity of the sensitive area. This effect, which obviously is
reduced by increasing the beam size, must be taken into account and a direct measurement is
necessary before we choose the final detector, even if the previous arguments give an indication in
favour of the PSD.
Another relevant point is the positioning of the sensors. From the assembling point of view,
the most convenient solution is the placement of the sensors on the external part of the electrode
housing (as shown in figure 32 and 33). On the other side, we must consider that a photodiode, that
we assume in photoconductive operation with reverse bias of a few volts, dissipates a power that
can reach a few mW (depending on photo-current and reverse bias); this can give rise to thermal
deformation of the inertial sensor. A more conservative solution could be the placement of the
diodes on the internal wall of the vacuum chamber, and then a few centimeters away from the
electrode housing. In this end, if also in this case the dissipated power results too high, we can
place the detectors out of the vacuum chamber through an optical window with the help of a few
mirrors (as in figure 32).
38
LISA PF
Figure 33 : Typical reflectance curve for a gold coating (from Melles-Griot catalogue.
3.3.7
Laser Source
The most suitable source for the optical sensor is surely a laser diode (LD) coupled to a
single mode optical fiber (SMOF). This type of source gives a good efficiency (emitted
power/absorbed power) and the emitted power can be easily controlled by regulating the current
flowing in the LD. Concerning the wavelength, the choice is limited to the near infrared (780 ÷
830 nm) because at longer wavelength the responsivity of silicon detectors decreased while at
shorter wavelengths the reflectivity of test mass and gold-plated electrodes decreases (Figure 35).
In this frequency range, there are plenty of devices available on the market, with power up to
several tens on mW and is highly probable that space-qualified products are already available (to
be verified).
The operation power will depend on the trade-off of several requests, as explained in the
previous sections but is should not exceed ~ 1 mW for each beam incident on the test mass. With
this power, the maximum static force applied by radiation pressure is rather small (6·10-12 N) and
can be easily compensated elettrostatically.
3.3.8
Planning of the activities
As previously explained, for a correct planning of the activity we must consider the schedule
of the LISA project. The time limit for the completion of the activity is the Preliminary design
review of LISA (expected in 2008). At this time, there will be the definition of the overall design of
the mission, taking into account the results of the flight test of LISA-Pathfinder and the ground tests
performed in parallel. So all the technologies to be implemented should be already fully tested for
that time. Another important appointment is the Final Design Review of LISA-Pathfinder (2005).
This is the last time for defining possible modifications of the inertial sensor for the LTP mission.
This includes, for example, the holes for the passage of the optical beams and the interfaces for the
placement of fibers and sensors in the vacuum chamber.
39
LISA PF
Taking into account this time scale, we propose in the following a possible planning for the
period 2004-2007:
2004
• Completion of tests on optical sensor’s intrinsic sensitivity, with the rigid set-up and tests and
comparison of the different type of position detectors (QPDs, PSDs)
• Finalization of the overall optical layout (already quite advanced)
• Design and realization of a prototype 6 DOF optical sensor with an optical layout similar to
the one to be used for LISA.
2005
• Test on torsion pendulum of some important aspect of the optical sensors (sensitivity, back
action).
• Study of couplings among different DOFs.
• Finalization of the design of the mechanical interface in order to implement it, as far as it is
possible, in the LTP sensor.
• Definition of the characteristics of the final sensor (power, sensitivity etc.).
• Start design engineering model for LISA.
2006
• Tests on selected components (vibrations, shocks etc.)
• Search for space-qualified components for the optical sensor (laser, sensors, electronics, fibers,
vacuum feed-troughs etc.)
• Finalization of design and realization of engineering model (this part of activity must be
founded by ASI and performed in collaboration with specialized companies under our
supervision).
2007
• Test of the engineering model and finalization according the results.
This planning should allow having a design sufficiently mature for the 2008 review. The
most urgent points are related to sensors characterization and choice of laser source and its control,
because only when a configuration that allows to reach the required sensitivity we can go on with
the detailed design for the implementation in the LTP. This is why it is important to progress with
the tests as soon as possible (already in 2004) and not after 2005 with the torsion pendulum. In the
end, if both the bench-top tests and the ones on torsion pendulum give good results in a short time,
we could imagine to anticipate to 2005 the definition of the engineering model end his test to the
2006, and this could permit the integration (perhaps partial and for only few DOFs) of an optical
sensor in LISA-Pathfinder, if the planning of this mission is delayed for other reasons.
4
WORKPLAN/MILESTONES
Summarizing, the overall outline of the project consists of:
• Design and production of a 4-mass roto-translational pendulum (Figure 13). It will be an
upgrade of the 4-mass pendulum in the Lab in Trento; it will have fused silica fibers, and,
possibly, a tilt compensation device. It will be assembled and tested in Firenze, and then
moved to the LNGS, when the site will be ready; at the LNGS a full analysis will be
40
LISA PF
•
•
performed. The groups of Trento, Perugia, Firenze, and Roma will be responsible for this
item (par 2.2.1, 2.2.2, 2.3).
Design and production of a (lower sensitivity) many degrees of freedom system; it will also
be operated at LNGS. The groups of Trento, Firenze and Roma will be responsible for that
(par 2.2.3).
Production of the optical readout system; the responsibility for this item relies with the
group of Napoli (par 3).
In the second half of 2004, the groups of Firenze and Roma will take care of setting up the facility
at LNGS with the support of the LNGS itself, and will place order for the vacuum chamber that
will be large enough to contain (not simultaneously!) the multi-degree of freedom system and the 4mass rototranslational pendulum with the fused silica fiber. The vacuum chamber should be ready
by the end of the year. At the same time, they will design and make prototypes of the multi degree
of freedom system, and with the help and expertise of the Trento Group, will prepare the 4-mass
roto-translational pendulum, with the fused silica fiber that should be ready by beginning of 2005,
care of the group of Perugia, .
Finally, the group of Naples in 2004 will complete the tests on optical sensor’s intrinsic sensitivity,
with the rigid set-up and tests and comparison of the different type of position detectors (QPDs,
PSDs); will finalize the overall optical layout (already quite advanced), and will design and realize
a prototype 6 DOF optical sensor with an optical layout similar to the one to be used for LISA.
Here it follows the schedule of the works :
41
LISA PF
5
MEMBERS OF THE COLLABORATION AND FINANCIAL SUPPORT
At present, the scientists working on the INFN LISA- PF project and involved in the development
of test benches for testing drag-free performance in earth based laboratories are:
University of Trento and INFN, Trento:
Michele Armano;Daniele Bortoluzzi; Paolo Bosetti; Ludovico Carbone; Ilaria Cristofolini; Mauro
Da Lio; Rita Dolesi; Mauro Hueller; Stefano Vitale; William Weber;
University of Firenze, and INFN Firenze/Urbino:
Gianni Bagni; Giovanni Calamai; Katia Grimani; Ruggero Stanga;
University of Rome, INFN Rome, and CNR:
Valerio Iafolla; Sergio Nozzoli; Giuseppe Pucacco;
University of Naples, and INFN, Naples:
Fausto Acernese; Enrico Calloni; Rosario De Rosa; Luciano Di Fiore;
University of Perugia, and INFN Perugia:
Luca Gammaitoni; Michele Punturo; Helios Vocca.
According to an agreement with ASI, INFN will finance the project, within a total of 1400 kEuro,
for the years 2004-2006.
6
REFERENCES
[1]
LISA and its in flight test precursor SMART-2”, S.Vitale et al, Nuclear Physics B
110 (2002), 210 and references therin.
[2]
[3]
[4]
[5]
[6]
On determining G using a cryogenic torsion pendulum, R.D. Newman and M.K.
Bantal, Meas. Sci. Technol. 10 (1999), 445-453.
Prospects for terrestrial equivalence principle tests with a cryogenic torsion
pendulum, R.Newman, Class. And Quantum Grav. 18 (2001)2407-2415.
Why is G the least precisely known physical constant?, C.C.Speake and G.T. Gillies
Z. Naturf.A 42 (1987) 664
New tests of the universality free fall Y. Su et al., Phys. Rev. D50, 3614 (1994).
Gravitational sensor for LISA and its technology demonstration mission, R. Dolesi et
al, , Class. And Quant. Grav. 20 (2003), S99;
42
LISA PF
[7]
Testing LISA drag-free control with the LISA technology package flight experiment
D. Bortoluzzi et al, , Class. And Quant. Grav. 20 (2003), S89
[8]
The LISA Technology Package dynamics and control D. Bortoluzzi, M. Da Lio, R.
Dolesi, W. J. Weber and S. Vitale, , Class. And Quant. Grav. 20 (2003), S227
[9]
Achieving Geodetic Motion for LISA Test Masses: Ground Testing Results L.
Carbone, A. Cavalleri, R. Dolesi, C. D. Hoyle, M. Hueller, S. Vitale, and W. J.Weber, ,
Phyis. Rew. Lett. 91, 151101 (2003)
[10]
Possibilities for measurement and compensation of stray dc electric fields acting on
drag-free test masses W. J. Weber, L. Carbone, A. Cavalleri, R. Dolesi, C. D. Hoyle, M.
Hueller, and S. Vitale, , to be published in Advances in Space Research, COSPAR 2002
conference proceedings.
[11]
P R Saulson, Phys. Rev. D 42, 2437 (1990)
[12]
M Hueller et al , Class. Quantum Grav. 19, 1757 (2002)
[13]
“Upper limits on stray force noise for LISA” L.Carbone, A.Cavalleri, R.Dolesi, C D
Hoyle, M Hueller, S Vitale and W.J.Weber, Class. and Quantum Gravity (2004)
[14]
“Measurements of small forces in the physics of gravitation and geophysics” F.
Fuligni and V. Iafolla, 1997, Il Nuovo Cimento, vol 20C, 619.
[15]
“Experiemental gravitation and geophysics”F. Fuligni, V. Iafolla, V. Milyukov, S.
Nozzoli, 1997, Il Nuovo Cimento, vol 20C, 637.
[16]
“Tidal tilts observations in the Gran Sasso undeground laboratory” V. Iafolla, V.
Milyukov, S. Nozzoli, 2001, Il Nuovo Cimento, vol 24C, 263.
[17]
“Mining for gravitational waves”, R. DeSalvo, 2004 Aspen Winter Conference on
Gravitational Waves.
[18]
“4-Mass pendulum for ground testing of LISA displacement”, C.D. Hoyle et al (to
appear in the Proceedings of Marcel Grossman Meeting 2003)
[19]
“Passive vibration isolation using a Roberts linkage” F. Garoi, J. Winterflood, L. Ju,
J. Jacob, and D.G. Blair, 2003 Rew. Sci. Instr. Vol 74, 3487.
[20]
“Monolithic Geometric Anti-Spring Blades”, G. Cella, V. Sannibale, R. DeSalvo, S.
Màrka, A. Takamori, 2004, submitted to Class. Quantum Grav.
[21]
“Performance of an ultra low-frequency folded pendulum”, D. G. Blair, J. Liu, E. F.
Moghaddam, L. Ju, 1994, Phys. Lett A, vol 193, 219.
[22]
Thesis, A. Bertolini, 2001 Università di Pisa.
[23]
“Current Status of large scale cryogenic gravitational wave telescope”, K.Kuroda et
al, Class. Quantum Grav. 20 (2003) S871–S884
[24]
“Ultra-stable performance of an underground-based laser interferometer
observatory for gravitational waves” Shuichi Sato et al
[25]
“LISA sensitivity below 0.1 mHz”,P. L. Bender Classical and Quantum Gravity 20, S301
(2003).
LISA Laser Interferometer Space Antenna: a Cornerstone Mission for the
[26]
Observation of Gravitational waves, ESA-SCI(2000)11, July 2001.
[27]
M.P.Chiao, F. Dekens and A. Abramovici (2003), in Gravitational-Wave Detection
M.Cruise and P.Soulson Editors, Proceedings of SPIE Vol.4858 98.
[28]
F. Acernese, E. Calloni, R. De Rosa, L. Di Fiore, L.Garcia and L.Milano, Class.
Quantum Grav. 21 (2004)S261-S267.
[29]
E.Calloni, A.Brillet, C.N.Man, F.Barone, F.Fusco, R.De Rosa, L.DiFiore, A.Grado,
L.Milano and G.Russo Physics Letters A 193 (1994)15.
[30]
S.Phinney, “The Big Bang Observer: direct detection of gravitational waves from the
43
LISA PF
birth of the universe to the present”, proposal NASA VM02-0021-002
44
Codice
Esperimento
Gruppo
LISA−RD
2
NUCLEARE
PREVISIONE DI SPESA
In KEuro
ANNI
Missioni Missioni
2005
2006
TOTALI
Mod EC./EN. 6
45,0
45,0
62,0
60,0
90,0 122,0
Spese
Materiale
Affitti e
Trasporti e
di
di
facchinaggi
calcolo
consumo
55,0
142,0
240,0
55,0
102,0
128,0
110,0
0,0
0,0
0,0
244,0
368,0
TOTALE
Compet.
544,0
390,0
934,0
Struttura
TN
Codice
Esperimento
LISA−RD
Gruppo
2
N
1
2
3
4
5
6
7
8
RICERCATORE
Cognome e Nome
Qualifica
Dipendenti
Incarichi
Affer.
al
. gruppo
%
N
2
2
2
2
2
2
2
2
100
80
70
100
90
100
85
90
1
2
3
4
Ruolo Art. 23 RicercaAssoc
ARMANO Michele
BENEDETTI Matteo
BORTOLUZZI Daniele
CARBONE Ludovico
DOLESI Rita
HUELLER Mauro
VITALE Stefano
WEBER William
Dott.
AsRic
R.U.
Dott.
R.U.
AsRic
P.O.
AsRic
TECNOLOGI
Cognome e Nome
BOSETTI Paolo
CRISTOFOLINI Ilaria
DA LIO Mauro
MEZZENA Renato
Qualifica
Incarichi
Ass.
Ruolo Art. 23
Tecnol.
R.U.
R.U.
P.O.
Univ.
Dipendenti
N
TECNICI
Cognome e Nome
Qualifica
Incarichi
1 GENNARA Pierino
2 SALOMON Claudio
3 SILVESTRIN Gabriele
Denominazione
Univ.
Univ.
Univ.
Annotazioni:
mesi−uomo
Mod EC./EN. 7
%
Assoc.
tecnica
SERVIZI TECNICI
20
20
20
10
4
0.7
Dipendenti
Collab.
Ruolo Art. 15
tecnica
%
10
10
20
3
0.4
Codice
Esperimento
Gruppo
LISA−RD
2
NUCLEARE
Data
completamento
Descrizione
luglio 2005
Commissioning della facility per il multi DOF suspension system
marzo 2005
simulazione del processo di carica di famiglie dei raggi cosmici piu'
rare dell'1% in composizione.
ottobre 2005
realizzazione del multi DOF suspension system ()
dicembre 2005
preparazione sitoLNGS e inizio delle fase di testing della multi DOF facility ()
giugno 2005
simulazione dell'effetto di carica dovuto ai protoni solari energetici con l'uso della geometria reale di LISA
Giugno 2005
modellizzazione di emissioni di massa coronale deboli dal sole (al
minimo solare) in L1
Settembre 2005
modellizzazione di emissioni di masse coronali forti dal sole (al massimo
solare) sul piano dell'eclittica
novembre 2005
modellizzazione e simulazione di una eventuale misura di Space weather utilizzando i rivelatori di particelle sui tre
satelliti
giugno 2005
adattamento e produzione fibre per la facility del Gran Sasso
dicembre 2005
studio effetto coatings sulla dissipazione interna delle nuove fibre
dicembre 2005
test su pendolo di torsione del sensore ottico per caratterizzare sensibilità e back−action
giugno 2005
studio degli accoppiamenti tra i diversi DOF
giugno 2005
finalizzazione del desing dell’interfaccia meccanica per implementarla, fin dove possibile, in LISA−Pathfinder
dicembre 2005
definizione delle caratteristiche del sensore finale (potenza, sensibilità etc.)
luglio 2005
prima fase di test con pendolo a 4 masse
luglio 2004
prima fase di misure con facility per test mass release
luglio 2004
prima fase di misure con facility per test mass release
dicembre 2005
upgraded test facilities
prototipi sensore inerziale e procurement versione rappresentativa del flight model per LTP
Mod EC./EN. 8
Struttura
Gruppo
TN
2
Coordinatore: Daniele Gibin
COMPOSIZIONE DEI GRUPPI DI RICERCA: A) − RICERCATORI
Componenti del Gruppo e ricerche alle quali partecipano:
N.
Cognome e Nome
Dipendenti
Affer.
al
gruppo
Incarichi
Ruolo Art. 23 Ricerca Assoc.
I
1
ARMANO Michele
Dott.
2
2
BAGGIO Lucio
AsRic
2
3
BENEDETTI Matteo
AsRic
2
4
BONALDI Michele
Ric.
2
5
BORTOLUZZI Daniele
R.U.
2
70
6
CARBONE Ludovico
Dott.
2
100
7
DOLESI Rita
2
90
8
FALFERI Paolo
Ric.
2
9
HUELLER Mauro
AsRic
2
10
MION Alessandro
Dott.
2
100
11
POGGI Silvia
Dott.
2
100
12
2
70
30
13
VINANTE Andrea
80
20
14
VITALE Stefano
15
WEBER William
R.U.
P.A.
AsRic
P.O.
AsRic
100
100
80
40
50
50
30
100
2
2
85
2
90
Ricercatori
15
7.15 5.55 1.3
Note:
INSERIRE I NOMINATIVI IN ORDINE ALFABETICO
1) PER I DIPENDENTI
2) PER GLI INCARICHI DI RICERCA
3) PER GLI INCARICHI DI ASSOCIAZIONE
(N.B.NON VANNO INSERITI I LAUREANDI)
Mod G1
(N.B.NON VANNO INSERITI I LAUREANDI)
Indicare il profilo INFN
Indicare la Qualifica Universitaria (P.O. P.A. R.U.) o Ente di rappresentanza
Indicare la Qualifica Universitaria o Ente di appartenenza per Dipendenti altri Enti:
Bors.) Borsista; B−P−D) Post−Doc; B.Str.) Borsista straniero; Perf.) Perfezionando;
Dott.) Dottorando; AsRic) Assegno di ricerca; S.Str) Studioso straniero;
DIS) Docente Istituto Superiore
4) INDICARE IL GRUPPO DI AFFERENZA
Percentuale
impegno
in altri gruppi
Ricerche del gruppo in %
Qualifica
III
IV
V
Struttura
Gruppo
TN
2
COMPOSIZIONE DEI GRUPPI DI RICERCA: B) − TECNOLOGI
Qualifica
N.
Dipendenti
Incarichi
Ruolo Art. 23
Assoc.
Tecnologica
Cognome e Nome
I
1
BOSETTI Paolo
R.U.
20
2
CRISTOFOLINI Ilaria
R.U.
20
3
DA LIO Mauro
P.O.
20
4
MEZZENA Renato
Univ.
10 40 30
Note:
1) PER I DIPENDENTI
2) PER GLI INCARICHI DI ASSOCIAZIONE
Mod G2
Percentuale
impegno
in altri gruppi
Indicare Ente da cui dipendono, Bors. T.) Borsista Tecnologo
III
IV
V
Struttura
Gruppo
TN
2
COMPOSIZIONE DEI GRUPPI DI RICERCA: C) − TECNICI
Qualifica
N.
Dipendenti
Incarichi
Ruolo Art. 23
Collab. Assoc.
tecnica Tecnica
Percentuale impegno
in altri gruppi
Cognome e Nome
I
1
GENNARA Pierino
Univ.
10 10
2
GOTTARDI Fabrizio
Univ.
10
3
SALOMON Claudio
Univ.
10 10
4
SILVESTRIN Gabriele
Univ.
20
Servizi (mesi−uomo)
−− Vuoto −−
Note:
1) PER I DIPENDENTI
2) PER GLI INCARICHI DI COLLABORAZIONE TECNICA
3) PER GLI INCARICHI DI ASSOCIAZIONE TECNICA
Mod G3
Indicare Ente da cui dipendono
Indicare Ente da cui dipendono
III
IV
V
Struttura
TN
Gruppo
2
PREVISIONE DELLE SPESE DI DOTAZIONE E GENERALI DI GRUPPO
Dettaglio della previsione delle spese del Gruppo che non afferiscono
ai singoli esperimenti e per l'ampliamento della Dotazione di base del Gruppo
In KEuro
IMPORTI
VOCI
DI
SPESA
Parziali
Partecipazione a gruppi di lavoro e conferenze
Totale
Compet.
10,0
10,0
Partecipazione a conferenze e working groups
12,0
12,0
Materiali di consumo
10,0
Materiale
Consumo
Seminari
10,0
Seminari ospiti
2,0
2,0
3,0
3,0
Spese
trasporto
Pubblicazioni Pubblicazioni
Scientifiche
Consorzio
Spese
calcolo
Ore CPU
Spazio Disco
Cassette
Altro
Affitti e
manutenz.
apparecchiat.
Materiale inventariabile
20,0
Materiale
Inventariabile
20,0
Costruzione
Apparati
Totale
(1) Indicare tutte le macchine in manutenzione
Mod G4
57,0
Struttura
TN
Gruppo
2
PREVISIONE DELLE SPESE PER LE RICERCHE
RIEPILOGO DELLE SPESE PREVISTE PER LE RICERCHE DEL GRUPPO
In KEuro
SIGLA
ESPERIMENTO
AURIGA
DUAL−RD
LISA−RD
Totali A)
SPESA PROPOSTA
Miss.
interno
Affitti
Miss. Materiale di
Trasp.
Spese di
Mater.
Costr.
Seminari
Pubblicazioni
e Manut.
estero
cons.
e Facch.
calcolo
inventar. apparati
Appar.
TOT
Compet.
32,0
5,0
10,0
24,0
5,0
20,0
37,0
5,0
12,0
11,0
25,0
35,0
20,0
5,0
120,0
124,0
45,0
197,0
47,0
49,0
54,0
71,0
145,0
366,0
10,0
12,0
10,0
Totali B)
C) Dotazioni di
Gruppo
Totali (A+B+C)
Mod G5
57,0
61,0
64,0
2,0
2,0
3,0
3,0
20,0
91,0
57,0
145,0
423,0

ISTITUTO NAZIONALE DI FISICA NUCLEARE Preventivo per

Transcript

Documenti analoghi

Meeting with Magneti Marelli PowerTrain 16 Giugno 14.30, Aula di

Art of Noise - CieliParalleli

quantum 4 - goclever

Circolare n.395 Vicenza, 11aprile 2015 Agli

Documento3 (Page 1)

Villanova di San Daniele, 28 Marzo 2012

Spiagge Soul 2016: oggi Fabrizio Pocci, Earth

Untitled - Arredatutto.com

V`LUX Nekos