Yano, N. Sato, Akutsu w/ help of Ushiba; following 20939.

Summary

Continued installing the invac high-power beam dump (Fig. 1); still on the way; confirmed the beam dump resides where the reflection beam from PRM could be caught within the PRM BF yaw actuation range. Note that the K400 flange in the +Y direction of the IFI chamber was sealed so you will no longer access IMMT2 or the other stuffs from there.

Final checks before sealing the flange

We slightly re-located the beam dump behind IMMT2; see JGW-T2112509-v1, where No. 6 is "the beam dump behind IMMT2" here, but the one in the drawing and the actual situation differs; see note below. Anyway, the current purpose of this beam dump is to catch two ghost beams from IMMT2; they are due to AR transmisson of the forward and backward beams at IMMT2. To prepare the backward beam, PRM was set to ALIGNED state. Fig. 2 shows two IR beams on the sensor card in front of the "beam dump behind IMMT2", which was not taken photo after it was re-located... sorry.

In the list of the modification in the IFI chamber, replacing the in-vac 1 inch mirror for retroreflection of the IMMT2 oplev to 2 inch one was included. But judging from the recent situation, I determined not to do this replacement. Thanks to set up several IR monitoring points (IMMT1 trans QPDs, PR2 trans QPDs, and maybe the one before IFI), we may not be worried so much about losing oplev beams like in O3GK.

Re-locating the high-power beam dump

The BF of the PRM suspension has limited yaw actuation range, so the largest angle that PRM can reflect the input beam back to this beam dump is limited. In the original design (see JGW-E2113385), the beam dump was expecting that the PRM would be able to reflect the beam about 10 cm yaw away from the main beam path around IMMT2. Considering that the distance from IMMT2 to PRM is about 5 m (see wiki), this is about 10cm/5m = 20 mrad separation. As this is light reflection matter, the mirror angle to achieve this separation should be around the half of this, so 10 mrad, but still large. On the other hand, note that this left mechanical edge of the beam dump is the closest entity to the main beam path at IMMT2, so I do not want to make it closer to the main beam path to reduce the required angle separation. After some discussion, we determined to slightly (~ 1cm) push the high power beam dump in the -Y direction, meaning the main beam path so that we can save a certain amount of PRM BF range (maybe Ushiba-kun would report later, hopefully). Fig. 3 was cut out from a movie taken today, so the forward beam from IMMT2 to PRM and the backward beam reflected from PRM are looking-like simultaneously seen maybe due to retarded response of the sensor card. Figs. 4 and 5 are also from the same movie but shaking the sensor card in front of the high-power beam dump. Thanks to the 1-cm shift of the beam dump, it seems the reflection beam could be mostly caught.

Sealing the flange

With a block chain, we hung the K400 (or phi-400) plate, and seal the flange opening of the chamber with a metal gascket. The claw clamps were tightened with a torque wrench; 20 Nm -> 40 Nm.

Then, we also set a VF150 flange made of copper (C1020; see JGW-E2113385). This is a heat path connecting in-vac and in-air environment. The VF150 flange was set along with a V175 viton o-ring (Figs. 6 and 7). Note that the screws for this flange were not yet tightened.

Checking the thermomteres

Connecting the in-air cables to the cryocon residing in the 2nd floor of the clean booth, we checked if the thermomters could be read properly with the cryocon, but failed. After some inspections, we found we did wrong work on this. Later we will revisit here to repair the things.

Tips for Guardian states

• Before starting these works, like yesterday, we zeroed the gain of IMC ASC and called IMC_LSC_LOCKED so that the IMC ASC loop won't go crazy every time we interupt the IR beam path. Also, IMMT2 suspension guardian was set to LOCK_ACQUISITION so that it could become free from oplev beam path cutting.
• Like yesterday, we worked along with monitoring PR2 trans QPDs (forward IR transmission beam) and IMMT2 oplev. Every time the displacement of the signal got larger, we stopped our work, sealed the area, called the IMMT2 suspension guardian to ALIGNED, and click zero history for the IMMT2 ADS both pitch and yaw, and then re-call LOCK_ACQUISITION to the IMMT2 guardian.

Notes

• Unexpected ghost beam found behind the IMMT2 trans beam dump. We looked for the birth point, and it seemed due to that the backward beam to POM1 in the MCF chamber was partly reflected at the window on the gate valve in-between the MCF and IFI chambers. It might be automatically disappear when the GV opens, so we do not care about this any more.
• In the report above, I referred to the beam dump behind IMMT2. In the document JGW-T2112509-v1, the corresponding beam dump No.6 only catches the forward beam passing through IMMT2, because we may like to extract the backward beam from IMMT2 to outside of the IFI chamber. But now, in a review, KAGRA determined not to extract this beam, so today we just set the beam dump to dump both the forward and backward beams.

Remaining works

• Design proper jigs for installing the thermometer for the IFI in-vac optical table.
• As described above, the VF150 copper flange (see JGW-E2113385) was not tightened. Note that this VF150 flange is now set with a V175 (maybe) viton o-ring. Hopefully this viton is a good high temperature tolerable one...
• Wrong pin assigns for the thermometer to cryocon.
• Peltier unit.
• Can the small rack on the 2nd floor can face in the different direction so that we can work more easily...