Virgo Logbook

One object that could possibly be guilty of the etalon-like modulation of the scattered light noise is the silicon beam dump on B1s. This device is clamped on the output-flange of the detection tower and absorbes around 2 Watt. We went to the detection lab, opened the acoustic box between the tower and the bench and took some pictures with the thermal camera, see fig 1. Some parts of the beam-dump are around 5 degrees hotter than the enivironment, especially the back-side and the screws that hold the copper support of the silicon wafers. Fig 2 shows a normal picture from the same angle.

To test the hypothesis, we detached the beam-dump and placed it for 10 minutes on the air-conditioning to let it cool down. After that, we re-installed it again, without unlocking the ITF. As shown in fig 3, this effectively reset the etalon-like modulation and around 20 fringes can be seen. It is thus very likely that the beam-dump is to blame. The exact mechanism of the etalon is not yet understood, but there might be some interference between 2 reflections inside the metal box or between one reflection in the beam-dump and the output window.

The optical path length change between 2 optical surfaces separated by 10 cm of aluminum is given by OPD = n_fringe * lambda / 2 = L * alpha * deltaT, where alpha is the thermal expansion coefficient of aluminum of 20e-6. Assuming a distance of 10 cm, a temperature increase of 2.6 degrees is required to explain the 10 fringes seen during a normal lock, which seems plausible. The effect of the dndT of air as given by Edlen's equation should be an order of magnitude lower.