[Arnaud P, Stuart A] Following doors being re-hung on the HAM5 chamber earlier today (see LHO aLOG entry 12915), Phase 3a DTT TF measurements have been taken this afternoon for SR3 (HLTS) suspension as follows:- - SR3 M1-M1 undamped results (2014_07_22_1200_H1SUSSR3_M1_ALL_TFs.pdf) HAM5 ISI Status: ISI unlocked, no damping or isolation loops running. SR3 alignment: No offset was applied during this measurement. SR3 undamped measurements from above have been compared with previous measurements for H1 SR3 at Phase 3a of testing (allhltss_2014-07-22_Phase3a_H1SR3_UnDamp_ALL_ZOOMED_TFs.pdf). The plot key is as follows:- Blue Trace = Model (hltsopt_metal) Orange Trace = H1 SR3 M1 (2014−07−15_1089513028), (in-chamber, doors open) Black Trace = H1 SR3 M1 (2014−07−21_1090036316), (in-chamber, doors hung) Magenta Trace = H1 SR3 M1 (2014_07_22_1200), (in-chamber, doors hung) It can be seen that despite having gone in-chamber to improve the M3 alignment and attempt to mitigate any sources of rubbing (which we thought we had) the peculiar damping of the first vertical mode persists. Furthermore, comparing before (Black trace) and after (Magenta trace) there appears to be no improvement. Somewhat frustratingly, before closing up the chamber, higher resolution measurements of the vertical DOF TF had indicated that the potential rubbing issue had been alleviated (see SR3_V2V_Doors_Open_vs_Closed.png). We had considered the possibility of hanging the door may have changed the air flow due to purge air, so we carried out a measurement both with purge air on and off (SR3_V2V_PurgeAir_Off_vs_On.png), but again this failed to improve the Q of the first vertical peak. Finally, a measurement was made with damping loops on for comparison (see SR3_V2V_Damping_On_vs_Off.png), which seems stable. Also, last night Arnaud repeated the vertical TF while applying +/- biases to various degrees of freedom to try and check in which way SR3 could have been hung up, but to no avail. Attached below is a log of particle counts taken during the HAM5 work covering the duration 1035 to 1126 (local). All data, scripts and plots have been committed to the sus svn as of this entry.
Also, I attached a trend of the osem signals (calibrated in um) before/after opening/closing HAM5 chamber yesterday (when Matt went in again to back up EQ stops) to show that the top mass didn't move much respectively with the cage.
Furthermore, this V2V tf was retaken with successive +/-yaw and pitch offsets applied to the top mass. We didn't see any change (cf attached screenshot). Measurements of the lower stages will be running overnight on opsws1, which might give us some more information.
Just in case it was lost in the mire of my earlier alog....we did low resoution Tfs with the door off initially but these showed nothing conclusive if the problem had been fixed or not. This is why Stuart dcided to try higher resoluton TFS with the door off to see if we could see if the problem had beenfixed..and from these TFs we thought we had. They mystery is why when we put the door on has it changed again ?????..hmmm puzzler.
Check the trends on the middle mass and bottom mass as well. The problem may not be at the top mass. Looking at the vertical mode shapes from the HLTS model: Freq: 1.07 Hz 3.5 Hz 28.1 Hz Top: 0.3826 -0.9250 -0.0051 Mid: 0.6524 0.2644 0.7071 Bot: 0.6543 0.2729 -0.7071 The 3.5 Hz mode is nearly all top mass motion, and that looks fine in the measurements. On the other hand the 1.07 Hz mode is dominated by motion at the lower stages.