Numerical Simulations of Stray Light in Virgo

In recent years, the interest in Gravitational Waves (GWs) detection has grown significantly, mostly thanks to the development of fine-tuned interferometers (namely the two LIGO and the Virgo experiments), capable of reaching very high sensitivities. Gravitational waves are ripples in the fabric of the spacetime that travel at the speed of light, and are produced when huge masses are accelerated or deformed. This happens in many astrophysical scenarios, including supernova explosions or gravitational interactions between black holes or neutron stars. They travel through spacetime, deforming it and making the bodies strained and distances stretched and squeezed alternatively. In spite of being produced by extremely violent events in the Universe, GWs are difficult to detect, because of both the considerable small strain signal which arrives on Earth from distant sources and thus resulting to be quite faint, and the fact that all measurements are affected by a variety of noises. A key towards an enhancement of the sensitivity of these detectors is to reduce the noises, and thus the detectors experienced an upgrade phase that led to the first detection of a gravitational wave signal on September 14th, 2015. Quantum noise stems from the quantum nature of light, and is driven by vacuum fluctuations of the optical field entering from the dark port of the interferometer. The effect of quantum noise is twofold: shot noise (which generates phase fluctuations of the optical field and is the main sensitivity limit at high frequencies), and radiation pressure noise (which generates position fluctuations of the suspended masses and contributes to the noise at low frequencies). It was first proposed by Caves that injected vacuum squeezed states can be used to reduce the measurement's shot noise, and thus to improve the sensitivity in the high-frequency band. This result was achieved with the German-British interferometer GEO600 at the AEI (Albert Einstein Institute) of Hanover and was replicated in 2013 with the LIGO interferometer at Livingston. Advanced Virgo joined the third observational run (O3) with an operating source of squeezed vacuum developed at the AEI: the injection of the squeezed vacuum field produced an improvement in the high-frequency sensitivity up to 3 dB (the horizon for a Binary-Neutron Star detection increased by 2-4 Mpc). On the other hand, quantum noise is expected to dominate the interferometer sensitivity over the whole frequency band, therefore a frequency-dependent squeezing is required in order to reduce radiation pressure noise as well; a way to achieve this is through a Filter Cavity: the current plan for Advanced LIGO, Advanced Virgo, and KAGRA is to construct an approximately 300 meters long Filter Cavity. Among the other approaches with respect to the Filter Cavity, the one involving the exploiting of a pair of squeezed EPR-entangled beams seems promising, as the frequency-dependent injection of squeezed vacuum is achievable without the need for an external cavity and instead it uses the interferometer itself as a Filter Cavity, thereby reaching an equivalent result with minimal additional optical components. Stray light injects additional noise into the interferometer as it can re-couple to the main beam when reflected by vibrating surfaces: this is visible as bumps and peaks on sensitivity at low frequencies. Moreover, it can also spoil the control signals. After the installations, a laboratory/commissioning approach (tapping and shaking tests) helps find mechanical resonances and identify the optical element(s) responsible for stray light noise, while numerical simulations with a raytracing software can provide estimates of the fraction of stray light that couples back into the interferometer, its main sources, and clues for mitigation strategies. Surface roughness of optics and dust contamination are two main sources of stray light in advanced GW detectors: stray light can not only contribute extra noise if it re-couples to the main beam when reflected by vibrating surfaces, but can also spoil the control signals of the interferometer. Given the extremely low roughness of the optics employed, dust contamination is critical as it can easily become the leading contributor to scattered light even in the clean environment of Virgo. Predicting dust contamination on the optics' surface based on environmental measurements with commercial particle counters is however very difficult and prone to big uncertainties. Instead, to monitor dust contamination in Advanced Virgo directly silicon wafers can be used as dust witness samples by placing them on the different optical benches and exposing them to the same local environment and activities as the optics under inquiry. The samples are then imaged with a custom photographic setup to thereby determine the amount of deposited dust as well as its distribution as function of particle diameter: these values serve as input value to create a scattering model for the raytracing software and hence to predict the amount of produced stray light.