Suspension Upgrades for Enhanced Interferometers Giles Hammond Institute

Suspension Upgrades for Enhanced Interferometers Giles Hammond (Institute for Gravitational Research, SUPA University of Glasgow) 18 th - 22 nd May 2015 LIGO-G 1500598 GWADW Alaska 1

Overview • Motivation for warm and cold suspensions • Noise terms • Some possible topologies – Higher stress a. LIGO suspensions – Longer warm suspensions (40 kg -80 kg) – Cold upgrade scenarios – Hybrid suspensions • Summary 2

Motivation for Suspension Upgrades Warm upgrades (e. g. A+, ET-HF, Cosmic Explorer) • Fused silica • 80 kg - 160 kg test masses • 60 cm - 120 cm suspension lengths • 800 MPa – 1. 5 GPa stress in fibres Cold upgrades (e. g. LIGO Voyager, ET-LF) • Silicon at 20 K - 120 K • 143 kg/200 kg • test masses 60 cm/100 cm suspension lengths Warm upgrades with fused silica offer a well developed technology and improvements in strain sensitivity at low frequency of up to 3. • Cryogenic upgrades for LIGO Voyager offer potential improvements in strain sensitivity of 8. • A V Cumming et al 2014 Class. Quantum Grav. 31 025017 G D Hammond et al 2012 Class. Quantum Grav. 29 124009 3

Noise Terms • • Suspension thermal noise – uses a combination of analytical and FEA dilution – horizontal, vertical and violin thermal noise Coating thermal noise uses finite test mass correction (Somiya and Yamamoto, Phys. Rev. D 79, 102004, 2009 – optimised a. LIGO coatings (16 131 nm Ta 2 O 5 and 17 182 nm Si. O 2) – • Seismic Noise uses DCC note by Shapiro et al. (T 1300786) to estimate transmissibility (longitudinal and vertical) – uses BSC requirements for ISI seismic noise (actually slightly better at 10 Hz) – 4

1. Higher Stress (>O 1/A+) • LLO alog: posted 18: 59, Wednesday 29 April 2015 - last comment - 07: 14, Thursday 30 April 2015 (17946) • Keeping current a. LIGO test suspension geometry, and operating at 1. 5 GPa would push bounce/roll to 6 Hz/9 Hz, and violin modes to 680 Hz A Heptonstall et al 2014 Class. Quantum Grav. 31 105006 5

1. Higher Stress (>O 1/A+) Keep end section at 800 m, but thin middle section to 300 m (1. 5 GPa). Use 5 mm stock to improve dilution. • • • 1. 5 GPA is 3 safety factor 350 m suspension (1. 1 GPa) demonstrated at LHO during a. LIGO SUS weld training (April 2015) A Cumming et al 2009 Class. Quantum Grav. 26 215012 6

1. Higher Stress (>O 1/A+) a. LIGO suspension=5 10 -23/ Hz @10 Hz • Minimal upgrade, only change final stage fibre geometry • Ready by early 2016 with robustness testing (e. g. O 1 upgrade or A+) 7

2. Longer Suspensions (A+, Explorer) a. LIGO suspension=5 10 -23/ Hz @10 Hz Longer suspensions (final stage 1. 1 m) and higher stress offer further improvements for both 40 kg and 80 kg • 8

3. Cold Suspensions (ET-LF/Voyager) LIGO Voyager assumes radiative cooling ET-LF assumes conduction cooling ET-LF fibre: 1. 75 mm Voyager ribbon: 2. 5 mm 0. 5 mm Parameter Voyager ET-LF Geometry Ribbon Fibre Power (MW) 3 0. 018 L (m) 0. 6 1 Mass (kg) 143 200 Test Mass (K) 124 20 Pen. Mass (K) 77 4 thermal (mm) - 1 strength (mm) 2. 5 0. 5 1. 5 Stress (Gpa) 0. 3 9

3. Cold Suspensions (ET-LF/Voyager) • Just showing suspension thermal noise for LIGO Voyager and ET-LF • This is a best case as: – dilution will be lower for real ribbons/fibres ( 2 -3 when necks are included) – bond attachments and associated noise need to be included => community needs to work on robust modelling (more discussion in breakout session)

4. Hybrid Suspensions • Hybrid suspensions are those which utilise silica fibres but with a silicon test mass (e. g. G 1500312 -v 1) • Silica has broad dissipation peak at low temperature. But can you still benefit from lower temperature operation Smith et al. , G 1500312 • Travaso et al. , Materials Science and Engineering A 521– 522 (2009) 268– 271 11

4. Hybrid Suspensions • Similar to a. LIGO performance until 240 K, but then worse performance than a. LIGO f=10 Hz Hybrid a. LIGO • a. LIGO uses cancellation to requirement thermoelastic meet 10 Hz • For cold silica, need to increase fibre diameter to maintain cancellation=> dilution gets worse • T<240 K, thermoelastic dominates until much lower temperature=> this pushes up thermal noise • see talk by Marielle van Veggel (breakout session) • Challenges also with jointing materials with different CTE=>induced stress 12

Summary • There a variety of suspension topologies which improve thermal noise performance Parameter High stress (e. g. A+) Longer suspension (e. g. A+, Explorer) Cold suspension (e. g. Voyager) Hybrid suspension 10 Hz improv. 1. 25 2 -3 8 (Voyager), 60 (ET-LF) Not better than 1. 04 Hardware changes None modest-significant Bounce mode 6 Hz 5 Hz 21 Hz 6 Hz Violin mode 680 Hz 370 Hz 300 Hz 680 Hz Stress 1. 5 GPa 350 MPa(1) 1. 5 GPa Readiness <1 year 2 -3 year 3 -5 years • • (1): grown fibre tensile stresses will likely be higher To understand full benefit need to include quantum noise and Newtonian noise 13