Optical losses as a function of beam position on the mirrors in a 285-m suspended Fabry-Perot cavity

Reducing optical losses is crucial for reducing quantum noise in gravitational-wave detectors. Losses are the main source of degradation of the squeezed vacuum. Frequency dependent squeezing obtained via a filter cavity is currently used to reduce quantum noise in the whole detector bandwidth. Such filter cavities are required to have high finesse in order to produce the optimal squeezing angle rotation and the presence of losses is particularly detrimental for the squeezed beam, as it does multiple round trip within the cavity. Characterising such losses is crucial to assess the quantum noise reduction achievable. In this paper we present an in-situ measurement of the optical losses, done for different positions of the beam on the mirrors of the Virgo filter cavity. We implemented an automatic system to map the losses with respect to the beam position on the mirrors finding that optical losses depend clearly on the beam hitting position on input mirror, varying from 42 ppm to 87 ppm, while they are much more uniform when we scan the end mirror (53 ppm to 61 ppm). We repeated the measurements on several days, finding a statistical error smaller than 4 ppm. The lowest measured losses are not much different with respect to those estimated from individual mirror characterisation performed before the installation (30.3 - 39.3 ppm). This means that no major loss mechanism has been neglected in the estimation presented here. The larger discrepancy found for some beam positions is likely to be due to contamination. In addition to a thorough characterisation of the losses, the methodology described in this paper allowed to find an optimal cavity axis position for which the cavity round trip losses are among the lowest ever measured. This work can contribute to achieve the very challenging losses goals for the optical cavities of the future gravitational-wave detectors.