Matt, Stefan

In an attempt to find our missing power, we decided to look for a change in the light scattered from the TMs.  The general idea was to use the camera images normalized to the input power, and see if the scattered light changed or increased faster than the arm power.  (A change in scattering could indicate heating of the mirror surface by a small absorber which leads to deformation and increased scatter T1000154).  This is similar to the OL test done earlier.

To investiage this and avoid camera saturation issues, I made a script to grab images of the TMs at a variety of exposure times (mexp.py).  This script takes 5 snapshots of each TM with exposure times from 30 to 10000 (micro seconds?).  I ran it during several power-up sequences to get images at 2W, 10W, 20W, 40W and 50W (all in /ligo/data/camera).

The accompanying matlab code loads these images and uses them to make a high dynamic range image (HDR) for each TM at each power level (see attached PNG for 40W images, and attached mat file for 2W and 40W data).

The overall conclusion is that the scattered light scales with the arm power.  That is, there is no evidence of increased scatter loss at higher power.

(About the HDR image: these use a log-color scale.  The ITMY scatter seems to be much larger than other optics.  ITMX has a strange crecent shaped collection of scatters.  The ETMs look like the usual stary night, though ETMY could use some focus adjustment.)