There will be no LIGO Executive Committee scheduled Monday, November 6, 2006 due to the Staffing Committee meeting scheduled that morning.
A string of meetings has begun at the end of this week at MIT: a meeting between the Pulsar Group and some outside neutron star experts, along with face-to-face meetings of the various Analysis Groups. Then, Saturday and Sunday we hold the LSC Observational Results Meeting. On Monday and Tuesday, many LSC members will stay for a joint meeting with numerical relativists. Attending the f2fs and the LSC Meeting will be representatives from each of the Virgo analysis groups.
>From: Rod Luna <rluna@ligo.caltech.edu>
>From: Linda Turner - turner@ligo.caltech.edu>
>From: Cleveland Mak <mak_c@ligo.caltech.edu>
>From: "Funaro, Catherine" <Catherine.Funaro@caltech.edu>
>From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
>From: Florence Kaufman <fkaufman>
>From: Gina Salone <gsalone@ligo.caltech.edu>
>From: Ed Jasnow <jasnow@ligo.caltech.edu>
No report. We will be requesting input for the monthly report for the NSF next week.
>From: Cindy Akutagawa <cindy@ligo.caltech.edu>
>From: Bill Tyler tyler@ligo.caltech.edu
Nothing significant to report this week.
Those following our S5 progress will know that we have been plagued by
problems that began after an M4.5 earthquake under
The drastic measures included "burping" selected BSCs on both interferometers. The symptoms exhibited on H2 were that pendulum resonances of ETMy were anomalous and the biases needed to achieve the old TM alignment had changed drastically. Suspecting a charging issue as the origin, we decided to attempt to relax the charge separation by venting the ETMy chamber to a few Torr (the pressure range where the gas exhibits its maximum conductivity) and shaking the external supports for the stacks. We refer to this process as "burping". The first burping brought biases much closer to normal, but with some over correction and an increase in noise. So we ran a second burp procedure, which got the biases closer to the usual range and allowed us to get back near-7 Mpc inspiral range. We then declared victory (falsely, it turned out) and buoyed by our success we decided to burp H1 to see if we could recover inspiral range. The H1 burp and some loopology by Rana were both needed to recover old inspiral ranges, but we achieved 15 Mpc most all of last night.
Unfortunately, H2 has continued to have issues when it drops out of lock and the bad low-frequency noise returns. Dave Ottaway made a terrific call on SciMon duty and decided to try a different fix but reversing the sign on the OSEM damping servo causing the ETMy to swing with amplitudes of order 100 microns. This fixed the noise without venting to a few Torr, cheering many but also confusing electrostatics fans. Our experience now is that we have had roughly an incident per week with this problem occurring in H2, but we think we have a relatively quick fix as well. We have decided to stage for a full vent and entry of the ETMy to change to the new-style earthquake stops (which should mitigate charging problems from earthquakes), but to watch and hope that the rate of H2 problems remain tolerable without needing so disruptive and risky a vent.
For sure we will have a bad duty factor the week of Nov 6, because we will take down H1 on Monday morning to float the output table at the AS port. We expect an outage of order three days and then tweaking time to get back to normal. This change will improve the decoupling of coherent noise between H1 and H2 and should enable us to improve the WFS bandwidth. We expect this will enable better data quality during daytime running conditions.
- some VE gauges were slow to restart.
- cryopump autofill servos were confused for a while, probably due to overfill after the power came back.
- ETMY Bias module probably damaged due to line transients (replaced on 30 Oct)
- LDAS dataserver was damaged; repairs carried out over the weekend.
(Dwayne Giardina)
(
See Advanced LIGO
See Advanced LIGO
The problems we were having with arm-cavity visibility now appear to be solved. Akira identified them as a contamination issue, removed and drag-wiped the SAC mirrors, then reinstalled SAC and realigned. Visibility is much improved, and we expect to close the chamber and begin pumping out Thursday evening.
We have finished reinstalling all of the actuators, they now all have nearly correct gaps, so when we release the system this time it should be free.
Brian O'Reilly has been hare this week working with us. He arrived with two GS-13s in their pods. We have been fit checking them and will power them up and see how they perform next week. We hope to be releasing the system by this friday (11/3/06)
For the past two weeks, we have been testing new EarthQuake Stops for the LIGO Large optics designed by Doug Cook (LHO). These stops have glass tips, to alleviate the suspected charging of viton tipped earthquake stops presently installed in the interferometers. The new stops seem to fit well, and in repeated dynamic testing have not caused any damage to the suspended optic after impact of the optic on the glass tips. A prelimnary report and some movies of the testing can be found at:
http://emvogil-3.mit.edu/~sarin/EQStop/Telecon_Oct302006.pdf
QuickTime movies:
http://emvogil-3.mit.edu/~sarin/EQStop/WholeLottaShaking_1.mov
http://emvogil-3.mit.edu/~sarin/EQStop/WholeLottaShaking_2.mov
MPEG movies:
http://emvogil-3.mit.edu/~sarin/EQStop/WholeLottaShaking_1.mpg
http://emvogil-3.mit.edu/~sarin/EQStop/WholeLottaShaking_2.mpg
Stefan Ballmer
After addressing most of the review committees questions (more software injections, strong signal bias, insensitivity to cut) the stochastic Radiometer analysis got preliminary approval from the review committee.
Duncan Brown
Last week:
This week:
Kent Blackburn
TclGlobus 1.4.0 has been released to VDT team.
Still working on patching Globus Toolkit 4.0.3 source code with debugging statements to investigate LDAS managerAPI core dumps. I'm having a problem building the source code under Solaris platform. Still working on this issue.
Still fixing several issues when aborting partitioning workflow.
Current status of running HIPE with 24 partitions at 4 OSG production sites:
OSG_LIGO_PSU: partition 13
osg-gw-2_t2_ucsd_edu: partition 5
Purdue_ITaP: partition 8
UWMilwaukee: partition 12
It will take about another week to complete all of them. Auto-restarts and cleanup of intermediate GWFs seem working fine.
Working with Michael Samidi, tested the LIGO workflow planner with three and six full HIPE workflows. Three workflows appears stable on a dual processor running FC4. Six workflows exhaust the process ID space and Condor-G/DAGMan becomes unstable. More analysis is required to determine the trade-offs for supplying sufficient submit host resources for larger scale workflow submission on the OSG.
Applied patch from VDT team to fix updating grid-mapfile from a cron job in OSG 0.5.1
Working with Ewa Deelman at ISI, finalized plan for account/access support of ISI/Pegasus team for OSG extensions work at UCSD via the OSG VO.
Provided OSG-ITB with a detailed analysis of validation of OSG 0.5.1 sites with nanoHIPE. Analysis can be viewed at: http://osg.ivdgl.org/twiki/bin/view/Integration/SiteValidationTableITB051
Provided ISI/Pegasus team with a detailed analysis of nanoHIPE jobs.
Attended ITB, Pegasus and VDT telecoms.
Organized first OSG Finance Board Meeting to be held here at Caltech.
Contacted Jim Yeck regarding a position on the OSG Finance Board as the "liaison to the funding agencies". Jim has accepted.
Interviewed with Editor of the newly formed "International Science Grid This Week" newsletter. The first edition will have an article on LIGO and its use of Grid Computing Tools.
Continuing to work on OSG statements of work for the institutions receiving OSG funds
Lisa M. Goggin
Efforts to compile LDAS on an Opteron base system running Solaris 10 continue. Using SunStudio compiler suite, lib/general has been successfully compiled and all unit tests pass. This required modifications to configuration files and some source code. These changes have been tested to be backwards compatible with the GNU compiler suite.
Version 1.8.352 was used for system testing. When using tclglobus for the "All User" command, the final result does not get flushed from the tclglobus socket. Globus debugging was added and the log files are being analyzed. There were no other issues.
The file structure for doing system testing is being reworked in an effort to have all scripts and data needed for the system tests to be part of the LDAS installation process.
(Phil Ehrens)
(Erik Espinoza)
(Fred Donovan)
(Igor Yakushin)
(Dwayne Giardina)
(Fred)
(Dwayne)
Rejected: 21,786 (65 containing virus)
Accepted: 3,545
Total: 25,331
(
(Christine)
(Veronica)
(Week of Oct. 22)
(Bruce Sears)
-iLog Maintenance: (0.5 days)
General iLog maintenance (user adds, keyword adds, systems work, etc.)
(Mike)
(Christian)
(Larry)
From: Hiroaki Yamamoto hiro@ligo.caltech.edu
e2e weekly meeting
Sany and his company reported their study (see "Mechanical Simulation for advanced LIGO" section below).
Hiro tested the statespace module of the vertical mode of HAM-SAS to find unstable. Valerio had an idea what's wrong (reformulation to change force-input to position-input as an actuator, which was done to support e2e interface ), and he will circulate an update when he finish investigating the issue.
AdvLIGO LSC/ASC design using FP arm model with quad suspension (Osamu, Hiro)
We discussed how to reduce the pitch motion of the test mass on AdLIGO cavity. Rana suggested that local damping on upper middle mass and penultimate mass can be turned on during lock acquisition. We will try it soon.
Static IFO simulation (Hiro)
The requirements of surface quality with spatial frequency of < 5 1/cm were calculated. Using the mirror surface map of 4ITM061, the requirement comes out to be RMS < 0.7nm. An independent estimation using random noise with proper spectrum set the limit to be RMS < 0.7nm. They are consistent. The orange peel effect seems to be negligible when properly measured. The surface data measured at CIT after coating gives an estimation of the loss to be 2.4nm.
Fast simulation of Dual Recycled Michelson Cavity (Hiro, Keiko)
The approximate expressions of light fields for fast calculation of field evolution in a FP cavity were validated. Evolution of light fields in a FP cavity was calculated under the following conditions;
(1) One of the cavity mirrors has a tilt (~0.01 rad)
(2) One of the cavity mirrors is displaced from the resonant point (~1e-6m from the resonant point)
(3) One of the cavity mirrors moves at a speed (~1e-6 m/s).
The results passed self-consistent tests.
(1) Only one certain mode that is not affected by the tilt builds-up in a cavity.
(2) Appropriate amount of light power builds-up in a cavity. The power agrees with analytic predictions.
(3) Once the light power builds-up, then the power goes down because the mirror displacement increases as the time goes on.
The evolution when one of the mirrors is oscillating does not produce correct results and the code is being debugged.
Modeler - e2e simulation engine (Hiro, Bruce)
A new version of e2e code, 3.2.2 were released, which supports a new option for e2e spectrum analyzer to ease the study of unstable systems. Also setup were directories and binary codes for three sets of machines at CIT,
(1) 32 bit machines, halfdome, ren, etc
(2) 64 bit SUSE Linux, 2.3GHz : homam
(3) 64 bit Fedola Core 4, 2.6GHz : rigel1, nunki
To workaround the numerical library issue of new version of libm.so, an older version is placed in E2E_HOME/lib so that this library is used to build e2e codes.
Bruce is working on implementing a customizable primitive in modeler. Implementing a new self-contained data reference object in the modeler. These will be used to implement the new user defined primitives which can contain an arbitrary set of data I/O ports (in number and type.)
Mechanical Simulation for advanced LIGO (Sany and SLU team)
Now that our analysis shows that the effect of back reaction from the triple suspension to the HAM-SAS optical table is not substantial whether or not the suspension?s local damping is activated, we started to model the AdvLIGO Input Mode Cleaner. (At this point, we include the horizontal model of HAM-SAS only, hence the position and yaw degrees of freedom for the triple suspension. We can include the HAM- SAS vertical model when the corresponding state space model becomes available.) As the first step, we are currently optimizing the local damping control filter. The current version of our local damping servo senses the lowest mass's (optic) motion for position and yaw, and feeds back actuation forces to the respective degrees of freedom.
ALFI : e2e front end (Melody)
Continuing in the graphical implemention for the new User Defined Primitive.
Currently working on providing the default values for ports and parameter declarations.
From: Ken Mason <kmason@ligo.mit.edu>
Three bids were recieved for the design and fabrication of the HAM seismic isolation system. A bidder has been selected and approved. A contract will be sent within the next couple of days.
We have removed all 6 actuators and recentered them. The setup fixtures had to be shimmed to properly center the bobbins.
Brian O'Reilly is at LASTI test fitting the GS-13, L4-C and STS-2 pods to the BSC structure.
From: Riccardo DeSalvo desalvo@ligo.caltech.edu Subject: minutes HAM-SAS meeting 11/1
Date: Thu, 2 Nov 2006 07:06:32 +0100
Next week Virginio will come. Good time, because he can replace Riccardo while on committee duty at UniPi.
The oven has been finished, cleaned and delivered (photos uploaded) and is presently heating up empty and manually with some boiloff nitrogen (1 m3/hour) flow for cleaning up. The thermal time constant is about 6 hours. The peak temperature ~ 200 degrees, as desired, a little less than expected.
There are some excess losses from the bottom, not worth fixing (could apply an external blanket). As expected the oven is hot at the cap rim, where the stainless steel lining forms a heat bridge, and is just warm elsewhere. The pipe containing the cold leads of the resistors 1s unexpectedly hot (140 to 160 degrees) but well within its parameters (250 degrees for the TC leads, 400 for the heater leads). The graph shows 4 thermocouples, all in air, one 10 cm from the bottom of the oven, one half height, and one closer to the cap.
The central one is in the air gap containing the heating resistors below the floor. Only few degrees spread between the three relevant TCs.
The PID temperature controller is late, we will install it as soon as arriving or replace with a different model.
The local provider of GAS (Air Liquide) did not have a second dewar to replace the defective one that discharged the previous week end. We are presently running with a smaller dewar. We found a second provider which will provide a dewar thursday afternoon (optimistically) or Friday. They also have more so they can send in a second one when the first one gets low. And they cost 1/2 as much.
We also got Nitrogen instead of Argon for speed of delivery, this also is cheaper.
We are now ready to process large parts as soon as they get cleaned.
Now that the oven is working we cleared Soldi to deliver the large parts.
He is, and continues to be the slow link though.
We received the blade bending tools, general tools (wrenches, screwdrivers, stands, etc.) and some more filter parts, which were promptly baked and stored in the small oven, but we are still waiting for the silver plated screws and other cleaned screws as well as the filter bottom plates that were waiting for the large oven. Should get these Thursday. We are also waiting for the coils to come back from cleaning.
We expect to bake parts to mount the GAS filters over the week-end.
On Thursday we will also bring in for cleaning the spring box, base structure and IP parts.
The modifications of the elevator carts have started and are proceeding.
The belts have been replaced with dry chains, the gear with a non reversible one, which required larger diameter shafts. End stops, safety pins and masks have been added as well.
We had problems in understanding the boiloff of the "polluted" samples. We could not boil it off completely. It turned out that the flask remains full of vapours and when cooling it recondenses, we initially thought that we collected a LARGE amount of non evaporable junk, this also because every sample was yellowish.
It turned out that every bottle of isopropyl or ethyl alcohol, even analysis grade, has in the label something called APHA color 10 (see note below). We are told by the vendor that this is essentially uneliminable, no matter what grade of alcohol is provided.
The good thing is that the alcohol and its residue appears to fully evaporate and no other sign of the APHA color remains after a time. The two graphs in test_01 show a window painted with test isopropyl after external evaporation (which is fast) as inserted and a few hours later, when the isopropyl adsorbed on the surface (slowler to go away than the bulk evaporation) also evacuates carried by the N2 flow in the instrument.).
We also reconditioned the boiloff pot, eliminating the brine.
We are now ready for systematic tests, starting from a backlog of samples that we have in store.
The first one tested seems to show traces of polyethylene, which may come from the wipes.
More systematic results next week.
It will be tough to make quantitative assessment of pollutants, unless they are identified.
With identified pollutants we can make an amplitude calibration with a known amount, with non identified pollutants we can only guess.
We also found a baked bug when making samples for the FTIR, it is tough to make sure there are no insects.
If we get a steady flow of parts from Soldi, and find no problems with FTIR tests or further problems elsewhere we should be able to finish up by end of November.
The LVDT circuit to go into production tomorrow, waiting for cables.
Still to decide if to install the triple immediately or at a later stage.
From: Ben Abbott <abbott_b@ligo.caltech.edu>
From: Ken Mailand kmailand@ligo.caltech.edu
The CES shop has completing the last arm section, and is modifying the fixture rotator assembly X-axis plate to match the arm. The lift assembly frame is in the CES shop. The construction of the controller for the lift assembly linear drives is in progress. I'm working on the LASTI 'conveyor' structure for the sus installalion.
From: Chris Echols <cechols@ligo.caltech.edu>
1. Developing SolidWorks assembly D060502, Upper Mass Assembly, OMC. In progress.
2. Developing SolidWorks assembly D060306, OMC Suspension Assembly. In progress.
3. Designing layout for PSL beam steering into interferometer, in D060165, AdLIGO VE System Layout. In progress.
From: Bill Kells kells@ligo.caltech.edu
From: GariLynn Billingsley <Billingsley_G@ligo.caltech.edu>
I have sent out a review request for the RFP, SOW, drawing and specs for the Advanced LIGO polishing pathfinder. I hope to have comments next week and be able to release the documents before the 15th of November.
From: Helena Armandula ahelena@ligo.caltech.edu
Removed First Contact, after 1-1/2 months, from a coated surface previously characterized for optical absorption. Initial measurements indicate that the optical absorption has been reduced, test still in progress.
Working the details for the sputtering of the gold coating on LASTI's ESD (reaction mass).
From: Dave Reitze <reitze@phys.ufl.edu>
From: Michael Smith <smith@ligo.caltech.edu>
I completed a preliminary analysis of the scattered light noise in ADLIGO, using the transfer functions that Hiro created with Twiddle (T060073-00). The following noise sources, listed in their order of severity, considerably exceed the noise requirement of ADLIGO if mounted on the SEI platform:
AS GW signal focus lens, glint
OMC scatter
AS GW signal focus lens, scatter
OMMT-ISC lens, scatter
OMMT-ISC lens, glint
OMC transmitted signal lens, glint
AS GW photodiode, scatter
The good news is that these noise sources can be mitigated by the following:
1) suspend the OMMT-ISC, OMC, OMC trans, and OMC refl components with a single or double pendula SUS (this could be similar to the proposed AOS ETM telescope and TRANSMON table suspension)
2) mount all the AS GW signal detection components on an additional suspended optical table
With these mitigations, the SLC noise requirement of < 1/10 SRD displacement will be met.
From: Peter King <pking@ligo.caltech.edu>
The intensity noise of a Lightwave NPRO was measured with a suspended high power photodetector. The measured noise was somewhat higher - as I recall - than that measured when the photodetector was mounted to the optical table. Mechanical resonances of the crude gallows-type suspension showed up in the measured intensity noise but not in the measured dark noise. The suspended dark noise measurement is identical to that measured previously when the photodetector was mounted to the table.
The measured intensity noise suggests that the combination of beam pointing on the photodetector and the motion of the photodetector cross-couples over to intensity noise and that if the photodetector was to be suspended better control of its motion is required. At the moment the suspension wire is approximately 400 mm long.
Problems measuring the low-frequency dark noise of the Si-based photodetector last week, suspected to be due to popcorn noise, do not occur when the photodiode is replaced by an InGaAs one. No problems were encountered when measurements were performed on the power supply or the circuit minus the Si photodiode.
For additional information about this report, contact Albert Lazzarini or Phil Lindquist