Part1 - OMC substitution

The damaged cavity (OMC #6) has been replaced by the spare (OMC #7). See entry 42687 for the birefringence tuning. Visual inspection shows an almost annular and white pattern on the surface of the burned OMC (see picture 1). We tried to clean this surface with first contact, but the pattern remains present after removal of the first contact. According to Eric Genin this pattern is different (at least in color) from what is usually observed on optics where dust has been burnt (dark traces). The damaged OMC has been shipped to LMA for further characterization.

The absorbing covers of both OMC have been removed and substituted by the uncoated aluminium covers that were in place before the intervention of July 2018. Visual inspection of the OMC1 absorbing cover (coated with Anoblack NI) did not reveal any obvious sign of damage on the coating. The OMC2 absorbing cover (coated with Anoblack EC) which is more promising in term of optical absorption is going to be characterized for outgassing by the VAC team.

A protection device has been installed on the bench to prevent the beam from reaching the OMC during lock acquisition. This device consists in a flip mirror, which is actuated by an Agilis motorized rotator. When the device is closed the mirror blocks the beam and reflects it towards the B1s1 large beam dump. When the device is open (Pic.2 and 3) the beam travels freely towards the OMC. Since the size of the beam reaching the mirror in closed position is larger than the B1s1 beam, we replaced the lens in front of the Faraday by a more diverging one (new focal length is -100 mm), which was provided by the EGO optics group. This allows to obtain at the surface of the beam dump a beam of 2 mm radius when the device is closed, and 2.6 mm radius when the device is open (B1s1 beam). We adjusted the position of the mirrors and of the lens in order to have the residual reflections of the beam dump also properly dumped. It has to be noted that the rotator has to be opened by 45° (or more) in order to guarantee that the reflection of the B1s1 beam on the lens is not clipped.

For the rotator we have cannibalized the waveplate rotator installed in Feb 2018 (entry 40626) in front of the OMC. The mirror installed on the flip device is a HR mirror coated for an AOI of 45° (S/N M101). The waveplate has been remounted on the former waveplate holder that was located at the output of the old Farayday Isolator until Feb 2018. This waveplate mount has now been installed in front of the OMC at the previous position of the rotator (Pic.4).

In order to provide the required in-vacuum pair of conductors for the driving of the rotator, we sacrified the cabling of the SDB1_M4 TY picomotor and used its twisted pair of conductors for the rotator. The EGO operation team prepared a dedicated in-air cable allowing us to connect the device to the commercial agilis driver located at the bottom of the SDB1 rack. The rotator could then be successfully actuated.

The modifications on SDB1 described above were performed with the bench blocked. After the tuning of the OMC birefringence was completed, the bench was rebalanced and suspended back. In total the bench weight has increased by about 10g.

Part 2- SDB1 Scattered light hunting

Scattered light investigations were performed with the ITF locked at 0.5 dark fringe without using the signals from SDB2. For these investigations the bench was suspended and locally controlled at a position allowing to have the bench coarsely aligned with the ITF beam. The following mitigation actions were performed :

• With the help of Valerio Boschi to act on the WI CP TY and TX positions, we could identify the residual reflections of the WI compensating plate on SDB1. These reflections are now properly dumped on the B1/B5 diaphragm, which was achieved by adjusting the TY setpoint of the WI F7 filter at -2650 urad, and the TX setpoint of WI CP at +305 urad. These setpoints must therefore be frozen at these values. It has to be underlined that we are now seeing 2 reflections from the WI as well as 2 reflections from the NI (after finely adjusting NI CP TY as well). All of them are hitting the B1/B5 diaphragm. They should correspond to the beams called B6, B9, B6p and B9p which were predicted in VIR-0267C-15.
• A piece of absorbing glass has been glued at the top of the rear side of the Hartmann lens mount (see Pic.5), in order to dump one of the residual reflection from the B1p/B1s1/B1s2 minilink viewport (more details about this viewport reflections in entry 42128).
• A ghost beam hitting the metallic ring surrounding the B1p/B1s1/B1s2 minilink viewport was found (see Pic.6 and Pic.7). This is a candidate for explaining how the 418 Hz resonance which was previously excited by tapping tests on the east flange of the tower (entry 41671) couples to the sensitivity curve (entry 42511). After checking the origin of this ghost beam on SDB1, it was identified to be the reflection of the main beam on the second AR surface of the SDB1_M3 mirror (which is the R=98.5% mirror used to get the pick-off of the beam before the OMC). If we consider the AR surface to be around 200 ppm, the beam represents about 3 ppm of the main beam, that is to say around 1 uW in dark fringe). A glass beam dump has been installed in front of the viewport in order to dump this spurious beam (see Pic.8).

We noticed the presence of an impact on the glass baffle installed near the west tower flange (see Pic.9, inside red circle). We could see that the residual transmission of the HR curved surface of OMC1 was hitting the glass baffle precisely on this impact, but the OMC was quite misaligned when this observation was made. It has to be noted that in steady state the power in the beam transmitted through the HR coated surface of the OMC should be only of the order a few microWatts which cannot damage the glass baffle. We suspect that this damage has been caused during powerful flashes, either by the beam transmitted through the HR surface or by a misaligned beam inside the OMC crossing the surface out of the HR coating. The scattered light induced on the baffle by this hole seems to be dumped by the meniscus lens diaphragm, therefore its impact on sensitivity should be very limited. We hope that the new strategy that we are putting in place to close the shutter earlier during unlocks will help preventing this type of issue in the future.

As a side note, Pic.9 also shows that a lot of dust particles are present at the edge of the aperture of the west flange baffle, which is certainly caused by the needed manipulations of the nitrogen setup and when we have to wipe the viewport full of condensed water.

Incidents to report :

• When we entered inside the DET tower on Tuesday morning (Sep 11), the air flux was not blowing inside the tower. We realized this issue only about 2 hours after the opening. The VAC team was notified and the problem was fixed. According to Marco Ciardelli the incident seems related to the DET tower valve which did not open following the standard procedure.
• During the scattered light investigation, there was an incident when we touched the bench by mistake while checking the positions of the CP reflexion on the B1/B5 diaphragm. It induced a rotation of the bench around the Y axis (TY rotation) that made the beam going onto the peek part that is used to maintain the small parabolic mirror M2 in its mount. The beam being very small here and with a power of about 5W, it burnt the superficial part of the peak (see pictures 10 and 11). After discussing with Antonio, we decided to leave it like this as it should not be critical for the vacuum and it would be a very risky and delicate operation to try to change it in situ or even just clean it with isopropanol.