We suspected 1 thing essentially: the extra beam reflected by the input prism of the OMC could actually be clipped by the OMC box or Peltier structure (see optocad drawing, 1rst picture). A huge part of this scattered light could reach the aluminium structure supporting the suspended rectangular baffle in front of the bench (see mechanical drawing and 2nd picture), scattered again and recoupled in the ITF. There was indeed some YAG light reaching this aluminium structure visible on the images of one of the viewport cameras.
After some inspections, we believe that is not the path for the extra noise:
- the beam reflected by the prism is not clipped by any mechanics and is dumped by absorbing glass as on the optocad drawing. However, even if no beam are clipped, a lot of light (most probably scattered by the various optics on the beam path) is clearly clipped by the OMC box diaphragm and well visible on the edge of the Peltier support (see 3rd, 4rth and 5th pictures). This light is scattered in all directions.
- the YAG light (that was visible from outside the tower on the aluminium structure supporting the rectangular baffle) was not a "beam" reaching the structure but the image of the light scattered in all directions at the level of the OMC (visible only at a certain angle of view).
So, we looked for another possible culprit. We probably found it: the reflected beam of B1p on the AR coating of the ouput window is reaching the aluminium structure of the rectanguar baffles. Note that this baffle was slightly displaced during the last week intervention. The other reflected beams are dumped by the large baffles on the west wall of the tower.
We changed a bit the position of the 2 rectangular baffles (aluminium structures are on the west side of the baffle now; and B1, B1p and B1s AR window reflections are dumped by the second rectangular baffle): 6th picture. We recentered the circular and rectangular baffles respect to the beam.
Let's be patient until next week to see if that was the problem...
