Below is a summary of the scattered light investigations and mitigation actions performed in the detection tower on the SDB1 bench from Monday morning to Wednesday morning (Jul 9-11).

For the first time it was possible thanks to the preparatory work performed by the locking team last week (many thanks to Julia and Diego!) to perform this investigation while the ITF was locked (0.5 dark fringe) which allowed us to visualize the beams with a relevant mirror configuration (PR aligned) and with a lot of power on the main beam. Most of the scattered light inspections were performed with the bench in-air but suspended. During the installation of beam dumps we managed to block the bench in a position that allowed to have the beam still aligned on the optics, which helped in positioning correctly the beam dumps. Below is a list of problems that we observed and tried to address:

·         The most shiny source of scattered light on the bench was located on the large beam dump used to dump 90% of the B1s1 beam (OMC1 reflection). This was due to a damage on the Si plate of the beam dump (see Picture 1) which was presumably caused by flashes happening during ITF unlocks. In order to mitigate this problem, the Si plate has been flipped in its mount (the other side was clear from damage) and the optical path of B1s1 towards this beam dump was modified (see pictures 2 and 3): a diverging lens provided by the EGO optics group (many thanks to Eric, Gabriel and Andrea) was installed approximately at the former location of  the beam dump, and the beam dump was moved further way in such a way that the beam size reaching the Si plate is now 3 times larger (so approximately w=2.1 mm). Therefore the power density reaching this beam dump is now about 9 times lower. An additional glass beam dump was installed near the lens (picture 3) in order to dump the residual reflection from the Si beam dump.

·         Looking for the residual reflections from the B1p/B1s1/B1s2 minilink viewport we found that they are not at the expected location/height on the bench. This can only be explained by a wrong inclination of the viewport, which must be tilted mostly horizontally instead of being tilted vertically. Accordingly we installed two new glass beam dumps (see picture 4) to dump these reflections. The beam dumps originally foreseen to block these beams (which were supposed to be lower than the main beam – see picture 5) have been removed. There is one residual reflection which could not be dumped, as it reaches the top part of the mount of the Hartmann beam lens. A possible solution to be considered in the future would be to perform a black coating on this mount.

·         We realized that the beam reflected by the waveplate that we installed last February in front of the OMC was not dumped. A glass beam dump has been added to properly block this beam (se picture 6).

·         We found a beam hitting the east flange of the tower. Thanks to the help of Paolo who rotated the NI CP in TY, we could confirm that this beam is a reflection from the NI CP. The NI CP position was then adjusted in such a way that the beam is no longer sent to the tower wall, but instead is being dumped on the B1/B5 diaphragm.

·         Finally a ghost beam which was hitting one of the edge of the dichroic mirror mount was blocked by adding an extra glass beam dump (see picture 7). We do not know from where this beam is coming from, it is not moving when moving NI CP, WI CP nor SR in Ty.

Note that on the M2 parabolic mirror we visualized 3 spots: the main spot near the center of the mirror which should be the main beam hitting M2, another spot located at the bottom of the mirror, which we assumed to be the B5 beam, and another spot which we can see at the bottom left of the M2 mirror and which is in fact an image of the beam hitting the dichroic mirror.

In addition to addressing the issues mentioned above, we also installed a beam dump (see drawing 1) to block the beam which is supposed to be reflected by the B5 minilink viewport (but we could not see this beam), and we installed a beam dump inside the SQZ pipe (see drawing 2) to block the reflection from the SQZ mini-link viewport.

During the investigations of scattered light we broke one of the bench LVDT wire, which was repaired by Fabio (many thanks to him!).

After the installation of all these beam dumps, we removed some ballast mass attached below the bench. The total bench weight has increased by about 80 g.

The bench was then suspended with the help of Paolo and rebalanced.