The welders completed the move of the HEPI Tubing and we restarted the pumps and topped up the fluid. We'll be in bleed mode for as long as allowable. To be clear, the Actuators at the Piers remain isolated from the circulating fluid and therefore are not operational. EricA/HughR
I've pushed an updated version of the h2iscey model, that includes some missing termination of unused channels. I have also rebuilt and installed this model, both without issue:
controls@h2build ~ 0$ cd /opt/rtcds/lho/h2/rtbuild-2.4.1 controls@h2build rtbuild-2.4.1 0$ make h2iscey command is make h2iscey ... Successfully compiled ********************* make succeeded log file is /opt/rtcds/lho/h2/data/buildlog/h2iscey_2012_06_23_15:06:52 controls@h2build rtbuild-2.4.1 0$ make install-h2iscey command is make install-h2iscey ... make succeeded log file is /opt/rtcds/lho/h2/data/buildlog/install-h2iscey_2012_19_23_15:19:33 controls@h2build rtbuild-2.4.1 0$
NOTE: I have not restarted the model. I'll leave it to someone on site to restart the model.
Tediously, moanfully and slowly we make progress.
ISC table was rebalanced.
There was an excessive roll in the upper mass, which was fixed by using the slide mass at the top.
Left and right BOSEMs were too low (light 100% cut off), so they were brought up.
Side BOSEM was off-centered and (barely) touching, restricting the upward motion of the upper mass, and it was centered.
Face BOSEMs were centered and adjusted too.
After doing these, the center of the telescope input aperture didn't move much if any, but the yaw angle of the thing somehow got better (why?). Yesterday the laser beam injected from the total station was not hitting the secondary, but now it's hitting. We're still off in yaw and need to use the pusher to rotate the cube.
At the request of Ken Mailand, we also spent some time making measurement for the position of the ISC table in relation to the ISI table so Ken and Calum can understand what could have gone wrong between the cookie cutter and the TMS telescope.
We started to use BOSEM readings. At first nothing worked as burtrestore was not properly done. I think there was one reboot about two weeks ago where people didn't care to burtrestore h2sustmsyepics, and after yesterday's reboot Jim burtrestored to the morning of Feb/21, but of course that didn't work.
Anyway, after Jim burtrestored to a known good time (Feb/03 morning), BOSEMs started making sense.
We cannot burtrestore from the EY CDS workstation. Why?
We still haven't connected QPDs.
Attached are plots of dust counts > .5 microns.
The FMY medm values had not recovered properly from some frontend reboots. The "h2susfmy_safe.snap" file was used in the '/opt/rtcds/lho/h2/cds_user_apps/release/sus/h2/burtfiles/fmy/' directory.
New .ini files for H2SUSITMY and H2SUSETMY (added channels), restarted dc0 at 15:31 PST Feb. 22, 2012
Hugh, Corey, Hugo
Some of the CPS-targets of HAM-ISI Unit #2 are scratched. Scratching patterns suggest that damages were caused by the insertion of a Teflon shim to check the sensor-gaps, which was performed prior to disassembly/reassembly. This test was waived for the two latest units tested at LHO and should be avoided.
Some targets have dents. The origin of these dents remains unknown.
Damaged targets will be sent back for re-machining.
None of this damage was seen during clean and bake. It's possible that the way we package these targets post-bake could be indited but I don't see how. Each target is cradled in contec wipes. Let us know if this appears to be a systemic issue.
I also think it is unlikely that the Clean and Bake process is responsible for the scratches but please keep an eye out as you unwrap target faces and let us know if the ConTec is not providing enough cushioning from the foil.
Freed up the TMS and balanced it.
It turned out that the height of the TMS was spot-on, but about 7mm off horizontally. What is the cookie cutter for, again?
Anyway, we pinned one corner of the cube, installed four temporary dog clamps with teflon tips, undid the real dogs until everything is loose, and used a pusher to tilt it.
After this adjustment, it was off by about 1mm in yaw, and still spot-on in pitch. We called it good.
Next, injecting the red visible laser beam from the total station, we can see that the angle is still far off in yaw (but pitch is OK).
We'll continue the angle adjustment today.
This is BTW a very slow, frustrating business.
We need to free the TMS suspension and uncover the ETM and ERM when initial aligners look at the alignment, then we need to cover the ETM and ERM and use EQ stops for the TMS, then squeeze ourselves into an extremely tight space between the stage zero and the cube, push the cube, and repeat.
We need a super strong helper or two whose shoulder is as narrow as 12 inches, chest as thin as 8 inches, 6 feet or taller, long but bendable arms. Anybody volunteering?
Attached are plots of dust counts > .5 microns. I have included a plot of H0:PEM-LVEA_DST15_MODE to show when the connections were lost due to this.
200W beam
Kyle, Gerardo Sensitive testing possible now that "gross" leak is fixed and pumping via turbo. YBM pressure ~8 x 10-7 torr. Turbo header valved-out, 100% of turbo throughput routed through leak tester -> Found a 1.1 x 10-7 torr*L/sec leak on untested 4.5" CF blank -> Torqued screws and now < 5 x 10-8 -> Found 1.7 x 10-7 torr*L/sec leak on coax feed-through appears to be left feed-through (of 2 on flange) and not CF joint-action? -> YBM viewports checked OK with indicated background 5 x 10-10 torr*L/sec at time of this testing. -> Replacement feed-through flange from Friday doesn't leak now but may in the future as I think the flange screws are bottomed out and not in tension. We may revisit this flange the next time the YBM is vented?
Both X1 test stand computers have been restarted in the staging building following a timing system interruption to move AC power for the timing master. All previously running models have been restarted.
Due to the need to move AC power from the old circuits to the new circuits in the MSR, the timing master was shut down this morning about 11:20 PST. The timing system interruption resulted in the need to restart all front-end computers. Work is complete, and all models have been restarted and burt restored to 11:00 PST on Feb. 21, 2012.
Daphen P. Doug D. Joe D. Thomas V. 8 assemblies for the custom viewports were put together and proof/leak tested referencing the documents listed below. A few minor adaptions needed to be made to the procedure for assembly, namely the torquing of the SHCS require two people, one must hold the assembly upright, while the other can torque the bolt down with 205 inch-lbs of torque (which is quite a bit, make sure you eat Wheaties beforehand). The knife edge on the bottom side of the flange must be protected from damage so we used dog clamps as spacers that prevent the knife edge from contacting the table top during the assembly phase. The optics were in general in good shape, except for a few particulate that were easily wiped off by drag wiping. The flange had a few burn marks from baking but Joe had said that's pretty common and alright. Documentation: As each viewport was assembled, we recorded the part numbers for the flange, optic, and clamp for each assembly. The total assembly part number is D1100999 and the S/N will be recorded in ICS as each assembly is put together successively (i.e. the first 8 assembled are D1100999 S/N 001-008). All of the assemblies were proof and leak tested by Doug, the inspection records will be attached to the assembly in ICS. I have many pictures of the event, but a few are attached.
TMS was badly crooked for some reason. Maybe it's different cable tension, maybe the fact that we changed the wire clamp hole to make it higher by 8mm.
After making various adjustment, TMS upper mass is now only pinned at one corner. We need to slide the adjusting mass a bit. Apart from that, no touching.
Before any adjustment, Doug measured the TMS height to be about 1.5cm higher than nominal. However, just by eyeballing, it looked OK. After our adjustment, TMS came down a bit (a few mm) but not 1.5cm. Thing is, it looks somewhat lower than ERM. not higher. This needs to be revisited before making fine adjustments to the balance.
As such, the first thing in the morning is to move the slide mass in the upper mass so nothing touches and measure the TMS height. If it is really 1cm or more high, we need to lift the tele and change the wire clamp position again.
Kyle, Gerardo, John Found gross leak on dual 3-pin actuator power feed-through on BSC8 (NW) -> Vented YBM, replaced leaker with one that was bench tested and found not to leak. As found, in-vacuum cables connectors were not screwed to feed-through connectors -> NOTE: flange-to-flange gap is uniform but conflat joint is not "metal-to-metal". Seems as if flange bolts are bottomed out in blind holes, tapped holes may be too shallow on this one flange(?) -> Pumping YBM overnight with purge/vent valve "cracked" open.
Results of inspecting the leaky feedthrough. It appears that the feedthrough was hit with enough force on the air side to break the weld on the vacuum side. Photos attached.
