The intervention in the SDB1 tower started on Monday morning (Jan 14) and the work was completed on Wednesday evening (Jan 16).

We started on Monday morning (Jan 14) by inspecting the bench for scattered light with an infrared viewer while the interferometer was locked at 0.5 dark fringe and with the bench still suspended. We could observe the CP reflections dumped on the B1/B5 diaphragm roughly at the same position as where they were in September when we adjusted the CP tilts (picture 1, taken through the IR viewer). As expected we observed a lot of scattered light on the dichroic mirror mount (picture 2, taken through the IR viewer) and on the M2 parabolic mirror mount (picture 3, taken through the IR viewer). Surprisingly we noticed that some light was going through the M2 parabolic mount (Pictures 4 & 5). This made us realize that the absorbing plate located at the rear side of the M2 mirror had not been mounted correctly during the replacement of the M2 mirror in November 2017. This is clearly visible on Picture 6, where we can see that the plate does not cover the whole rear side of the M2 mirror. Accordinly we remounted the plate correctly, which was sufficient to get rid of any light in tranmission of the mirror.

Once the scattered light inspection was completed, the bench was blocked in order to allow the forthcoming activities.

On Monday evening and Tuesday morning we installed a mirror polished steel diaphragm on the 3 inches dichroic mirror mount (see Picture 7). For this purpose the dichroic mirror mount had to be dismounted from the bench, after taking several references in order to ease the realignment of the mirror : 3 iris diaphragms were temporary placed on the bench along the itf beam path (1 iris before the dichroic mirror, 2 iris after) and a SLED was also installed on the bench with its beam pointing to the dichroic mirror and its reflection used as reference. The mounting of the diaphragm was then performed in the clean room. Unfortunately during this operation a screw broke and the broken part remained blocked in the 3 inch mount, preventing us from reusing this mount. Therefore the dichroic mirror was reinstalled on a spare 3 inch mount. This changed the position of the dichroic mirror but a correct alignment on SDB1 could be recovered thanks to the references previously taken. Picture 8 shows the dichroic mirror after installation of the diaphragm as seen through the IR viewer and with the interferometer locked at 0.5 dark fringe. An improvement can clearly be observed by comparing this picture with picture 2.

Tuesday afternoon was dedicated to the tuning of the birefringence of the OMC. A specific entry will be made on this topic. We also replaced the aluminium OMC covers with new covers coated with AnoblackEC (see pictures 9 & 10).

On Wednesday morning, we released the bench, rebalanced it and restore the controls. Overall we have reduced the bench weight by about 15 g.

Wednesday afternoon was dedicated to a final measurement of the OMC birefringence tuning before closing the tower (more details in the specific entry).

On Thursday afternoon, while the bench was already in vacuum (but the pumping not yet finished) we adjusted the balancing with the marionetta stepping motors in order to reduce the corrections in TX and TZ. MA TX was moved by -1000 and MA TZ by -1500 steps.

A few general remarks :

• We could work without nitrogen sent to the cryotrap viewport during the whole week. This is probably due to the lower humidity rate in the central building during winter time.
• We noticed that when the OMC slow shutter is in the closed position, the beam is quite low on the shutter mirror (although it is not clipped). The shutter could probably be moved a bit more.
• We noticed traces of what looks like contamination on the glass baffles installed on the west flange of the tower. Pressing a finger against one of this baffle removed the « dirtness » (pictures 11 & 12). We tried to clean the baffles with isopropanol and acetone but this did not work. It seems that a mechanical action is required to remove the « dirtness ». We checked a piece baffle stored in the clean room and noticed similar patterns on it. Antonino will analyze further this baffle.

For the birefringence tuning of the OMC, we moved the position of the SR lens towards west by 2.3 mm. This improves a bit the mode matching, although it is not sufficient to totally compensate the effect of the cryotrap windows. The SR lens will have to be put back at its nominal position (Sa_SR_F0_X = +100) when the viewports are reopened.