During the wind storms this week I collected data with a second STS at locations around the BRS and the GND STS that gets corrected for tilt, ranging from a huddle between the two STSs at 47cm apart to a maximum separation of 550 cm. For each location I found a period of the storm during which the average over 1664 seconds was close to 17 MPH. The figures show data only for these 17 MPH periods. I checked that in addition to averages within ½ MPH of 17, the extremes for each period were within a few MPH of each other.
Figure 1 shows the coherences between the two STSs (top plot) as well as the spectra when the movable one was closest or furthest from the wall (bottom plot). Orange, red, blue, black and thin black traces are for increasing separation. The coherence tends to drop with distance and drops as low as 0.75 at the two sites at 2.08 and 2.54 meters, blue and black. The spectra in the lower plot show that the tilt near the wall (dotted thick black) is more than twice the tilt only 2.54 m away (black). The thin black dotted line is for the furthest location from the wall, near the far pier of the BSC, and its tilt is about half of that at the BRS. The trend for greater tilts at sites closer to the wall is consistent. A better location for the BRS and the BRS-corrected STS might have been on the opposite side of the beam tube from the wall.
The top plot of figure 2 shows coherence between the BRS and each of the two seismometers. The best coherence that I got was for the red site, as close as possible and centered on the BRS. This red dotted coherence is better than the solid red coherence for the GND STS, taken at the same time. The coherence near the bases of the near and far piers was remarkably poor. This may explain why using the BRS to correct the chamber rather than the seismometer, such as BRS RX to ISI RX, would not be very successful.
The bottom plot of figure 2 shows the coherence between the Z-axes of the GND STS and the moveable STS. I include this because the effect of tilt on the Z axis is much less than on the horizontal axes so it tends to show the coherence lengths for the real displacements that we want to subtract using a seismometer. At 2.5 m from the GND seismometer, the coherence is down to 0.5 from about 0.3 to 1 Hz. This is also the case at the far blue piers on the opposite side at about 5 m away. It is unlikely that even the corrected GND STS helps much in this frequency band. Just as we are in the near field region for wind induced tilt, we are also in the near field region for wind induced vibration. The chamber and walls are only about 3 m from the wall that is bending according to the pressure distribution outside the building and the bending scale is meters.
For 1-5, ISI was set to off-line, HPI was nominal except that sensor correction was turned off. For 6, ISI was back on set for normal IFO operation
1) Huddled, igloos touching, 47 cm +Y of GND 5/6 3:00 to 5/7 00:00; 1664 seconds starting at 14:32:17, 6-26 MPH, ave 17 MPH
2) Huddled rotated by 10 degrees 5/7 04:00 to 5/7 16:00; 1664 seconds starting 5/7 at 15:02:17, 7-25 MPH, ave 17 MPH
3) Across BRS from GND and 20 cm -Y to center on Y of BRS, 133 cm distance, had to recenter so 5/7 19:00 - 5/8 23:50; 1664 seconds starting 5/8 20:56:17 4-32 MPH, ave 17 MPH OR 5/8 14:32:17 6-27 ave 17 MPH
4) Near -Y, +X leg of BSC10, +190 Y of GND, 208 total, 11:00 5/9 UTC to 15:00 5/9 UTC, if drift in Y-axis is OK, the start time could be extended back until 5/9 at 2:00 UTC; 1664 seconds starting 5/9 4:10:17, 4-37 MPH, ave 17 MPH
5) about 30 cm from edge of slab, 254 cm distance from GND, all in +X 16:00 5/9 UTC to 5/10 0:00:00; 1664 seconds starting 5/9 16:30:00, 5-30 MPH, ave 17 MPH
6) Near far (+Y, -X) BSC pier, +510 cm Y total 550 cm distance from GND. Start time 5/11 5:00; 1664 seconds 5/13 15:12:57, 7-31 MPH, ave 17 MPH
Robert & Hugh
