LIGO Weekly Report

From: Rainer Weiss <weiss@ligo.mit.edu>
Subject: Notes from the LSC Executive committee meeting Jan 24,2003

to: LSC executive committee
from: R. Weiss January 27, 2003
concerning: Notes from the LSC Executive Committee meeting January 24, 2003

The meeting had a small attendance since many of the members were on travel
to the sites for the E9 run.

New authors for LIGO 1 papers
------------------------------
After I reported on my conversation with Jordan Camp at Goddard Space Flight
Center concerning the authorship of new members of the group on the major
LIGO 1 papers, the committee had an active discussion about this. I had told Jordan
that people now entering the Collaboration would not be authors on the general
LIGO 1 papers but clearly were to be authors on the specialized papers to which they
contribute. Gary suggested that a cutoff on authorship after a specific date was
one model for the Collaboration and that in other HEP collaborations there is a
trial period of, say. a year after which the new members could become full
authors of general collaboration papers. He offered to make a set of talking
points on this issue for the next LSC Executive meeting.

Action on shifts and other service functions
---------------------------------------------
Keith Riles and I sent a letter to all LSC PI's stating the policy on
assigning scientific monitoring shifts and other service functions. One of the
consequences of the letter was to determine the correct FTE count in each of the
groups. It seems that undergraduate students have been tallied in the count of
some of the university groups. The LIGO laboratory will make a uniform response
for all four of its sites.

LSC Support for the advanced LIGO proposal and submission strategy
------------------------------------------------------------------

A draft letter endorsing the advanced LIGO proposal was sent to all LSC PI's.
The letter is to be signed by all LSC institution Principal Investigators
and is addressed to Bev Berger and Tom Lucatorto. So far approximately 20 out
of 33 PI's have responded.

The proposal will (hopefully) be submitted to the NSF before Feb 1. The
strategy of whether it should be an MRE or Physics Division proposal will be
discussed with the NSF.

"Upper limit" group papers
--------------------------
Alan Wiseman and I will meet this week on the telephone with the 12 reviewers
assigned to work with the authors of the "upper limit" group papers. In a
meeting of the "upper limit" group chairs, held last week, the different
quality of the science results being reported by the groups was discussed.
As a means of dealing with the variance and also to give room for future papers,
Peter Saulson suggested that we consider writing a single paper composed of results
from all the "upper limit" groups.

Next LSC Executive Committee meeting
-------------------------------------
Friday February 28, 2003 at 11AM EST

LIGO I Construction/LIGO Laboratory Administration (Lindquist)

There was a site teleconferences held on Thursday, January 30, 2003. The following issues were among those discussed:

Civil Construction at Livingston--Interior walls are complete. Personnel are moving into the building. There has been somne activity regarding the traffic control gate. Recommendations are being sought.

Livingston Off-site Storage Building--The building is not very secure and is currently being used to store stuff that is no longer of any use. An exception is the vacuum equipment bake blankets, which are crated. Ed has been given authorization to surplus the beam tube bake power supplies. It was deemed more cost effective to procure new supplies than refurbish the old should the need arise.

Civil Construction at Hanford--NTD has been contacted for a drawing package for proposed modifications to the new Laboratory Building. No cast data is available yet..

Budget--Preliminary data for January indicates that we are running at approximately 91 percent of budget assuming $28 million funding. This does not include special items like corte optics, seismic remediation, etc.
No Meeting Next Week due to Aspen Conference, etc.

The list of current actions revised to reflect the status of open actions assigned through January 16, 2003 may be found at ACTION LIST.

Web pages for the DCC give simple how-to's for document numbering, easy access to the latest on-line documents, and search capabilities for the DCC database. Take a look. . .

WE 01/30/03	Packages	Faxes
In	24	38
Out	9	31

Advanced LIGO (Frey)

Weekly Advanced LIGO Project Controls meeting was not held this progress period. Next meeting is scheduled for Friday at 9:30am at ECR.

Project Web Site for posting schedule and progress related data continues to be updated with the latest and greatest.

Change Request CR-030002 to adjust the FY 2003 operating budgets to reflect actual staffing during the first quarter has been prepared. Additional information is required.

Quality/Safety (Tyler)

NOTE: The Hanford Safety Audit scheduled for January 30, 2003, was postponed. The Hanford Safety Audit has been rescheduled for the 24th of April, 2003.

LIGO Hanford Observatory (LHO) and Interferometer Operations (Raab)

Summary of Commissioning Activities at LIGO Hanford Observatory (compiled by M. Landry)
---------------------------------------------------------------------------------------
The E9 engineering run completed successfully on Monday 27th. This key dry run tested
S2-readiness for dmt monitors, astrophysical injection, and short-latency calibration.
Science-mode duty cycles for the Hanford ifo's were 54% (H1), and 58% (H2), although these
numbers are low due to calibration and astrophysical injection, and simple locked
duty-cycles were considerably higher. A comparison of E9 displacement noise in the
Hanford interferometers can be found at http://blue.ligo-wa.caltech.edu/ilog/pub/ilog.cgi?group=detector&date_to_view=01/25/2003&anchor_to_scroll_to=2003:01:25:06:46:56-landry_m
Early running was characterized by disturbing non-stationary noise, whereas the ifo's
closed the run on a twenty-hour quiet and coincident lock.

The AS path was redistributed to divert 15X more light to AS2 than to AS1. The
latter pd is now the acquisition photodiode, the former is the run-mode pd. The ASI
servo has been enabled with a 30Hz bandwidth, and we now land 60-80mW on AS2. The result
is improved noise above 200Hz, see
http://blue.ligo-wa.caltech.edu/ilog/pub/ilog.cgi?group=detector&date_to_view=01/30/2003&anchor_to_scroll_to=2003:01:30:18:43:07-daniel
Despite the better noise spectrum, we see a shelf of noise alternately appearing and
disappearing at 90-800Hz.

The 2k MC has had trouble relocking, breaking lock when the boost gain is touched. This
has been traced to an offset in the MCL path.

4k WFS work continues apace. Filters to compensate the optical levers have been loaded
onto the ETMs and RM, allowing an increase in bandwidth of the wfs loops. For details
please see: http://blue.ligo-wa.caltech.edu/ilog/pub/ilog.cgi?group=detector&date_to_view=01/29/2003&anchor_to_scroll_to=2003:01:30:03:03:39-luca

The 4k PSL FSS has been shown to be working at spec (flat within 2dB up to 100kHz, and no
more than 20 degrees of phase lag at 100kHz). An elog was posted:
http://blue.ligo-wa.caltech.edu/ilog/pub/ilog.cgi?group=detector&date_to_view=01/17/2003&anchor_to_scroll_to=2003:01:29:20:28:15-rick

LIGO Livingston Observatory (LLO) and Interferometer Operations (Coles)

Commissioning: Disassembled a pre-mode cleaner that was not properly aligned and reglued it. Continued to check the ISS servo with a the board that allows an EPICS control. The AC coupled photodetector oscillates at 500 kHz and the small millivolt offsets are amplififed by the large gain in the circuit to saturate the output. (Kovalik) A failed BS optical lever laser (old point source laser) and a failing ITMx optical lever laser (iFlex 1000) have been replaced with the last two remaining iFlex 1000 lasers that we have on hand. We replaced a failed ion pump on GV1. We have also updated Some of the ISCT table photographs. (Traylor, Hammond)

CDS: Worked with Alex to fix the DAQ flickering bit experienced during E9. Except for the above DAQ bit flickering all CDS systems worked through E9 without any breakdowns or problems. No reboots done during E9. Epics Log server is now running on LLO1 (CDS Server) to log errors and reboots of epics processors at LLO. This will help us to keep track of the epics system stability. Checked the Frame builders disk space to reconfigure it for S9 and the future. NOTE: Framebuilders LLOFb2 and LLOFb0 will be shutdown for disk reconfiguration work on next Monday and Tuesday - Feb 3 and 4, 2003. Moving directories from DTT and ops/controls home directory to a local disk on control4 to make space for S9. Reconfigured Big Brother and its web site for CDS Network to show the current CDS network status. Installed and configured a A1000 for disk to disk backup. Working on setting up scripts to do backups. (Chethan)

The following is my progress report,
1) Tidal servo operation is being automated. A script is being generated
that runs as a cron job that periodically creates current_tide.dat file for
tidal servo.
2) State Vector Diagnostics is being created to monitor the State Vectors.
3) Fixed the bugs in MC Autolock as per Rana's request.
Ash

LDAS admin:
1) replaced failed T3 disk;
2) generated 96 channel and AS_Q only RDS during E9;
3) had to troubleshoot LDAS when jobs got stuck in metadata API.
LDAS data analysis:
1) ran waveburst DSO during E9;
2) fixing bugs in waveburst DSO.
--
Igor Yakushin

GC: Wrote and installed a script over the weekend to monitor the cryo pump levels and several pressure gages. This script then emails Allen, myself, and the operators if there is a problem. It also emails Allen's cell phone. looking into more network problems with LSU. People have been having problems maintaining connections to various places. Also, some web pages are not coming up. Unfortunately, there does not seem to be a clear pattern to the problem. I am setting up a laptop for one of our visitors. I should have it ready by the end of the day today. (Shannon)

LDAS: Finished the first set of benchmarks on the evaluation units. I will run some more tests tomorrow. (Shannon)

Other: We had a nice celebration in Louisiana Wednesday evening in honor of Gerry Stapfer's retirement. Gerry's last day of work is Friday, Jan. 31. We wish him well in his retirement in Gold Beach, Oregon.

Detector/Technical Support (Coyne)

Seismic Upgrade Project

Hydraulic External Pre-Isolation (HEPI)

Dennis Coyne reporting

Electro-Magnetic External Pre-Isolator (MEPI)

Dennis Coyne reporting

CDS Software

Rolf Bork reporting

- Lori is working on a digital version of Input Optics WFS for 40m lab.

- I'm working on documenting all the code we have in place, starting with LSC.

- I finished the WatchDog Requirements document.

- Alex is trying to get a realtime version of Linux running for some of our

tests for future use. It looks like it may be working, but we need some more

testing. The problem we are trying to overcome is the ever increasing

requirements on our VME processors and the limited CPU speeds available on VME

processor boards. One possible solution is to have the VME CPU do only I/O,

passing off the data, via our 2Gbit/sec reflected memory network, to a higher

speed box for the actual digital signal processing. We could then go out and

buy a 2.8GHz CPU box, for example, to do this work. As newer, faster boxes come

along, we would not have to replace the VME unit.

CDS Hardware

Rich Abbott

No report

Jay Heefner reporting

LOS Coil DAQ: A single channel test board design has been completed and sent out for fab. It is due back on Monday.

Adv LIGO DAC Module: We have not received a revised delivery date for the prototype module. Jay has emailed them and called them several times, but they have yet to respond.

PSL

PeterKing

The PZT mirror of the remaining optically contacted pre-modecleaner was

removed from the body (which was dropped during the summer). The

pre-modecleaner will be rebuilt.

A composite amplifier was laid out and the board milled. Hopefully the

output noise of the amplifier will be satisfactory at 10 Hz for use in

other low-noise measurement tests.

Lee Cardenas

MOPA S/N 119 sent to be repaired on 8/16/02. Is back from
Lightwave Electronics. Laser is being installed inside the PSL enclosure and
waiting to be checked.
Several small tasks (new OTF cavity design, laser bases, new chamber etc...) are underway.

Optics Analysis

Erika D’Ambrosio

I made some more decomposition of the grids representing the beam resonating in

theFabry-Perot cavity with non-spherical mirrors, when there are distortions of

the surface and when the alignment of the mirrors is adjusted to compensate the

tilt included in the deformation maps, to double check all my results before the

final version of the paper on the flat topped beam is submitted. A short version

has started to be circulated by Kip to some members of the Ligo project who are

interested in our proposal.

Optical Contamination Cavities

Lee Cardenas, Liyuan Zhang

Preparing a test report for the results of measured absorption change and scattering loss change due to the introduction of hydraulic fluid into the chamber. The currently favored hydraulic fluid (Aquamil) has no observable change in absorption to a level of 0.2 +/- 0.2 ppm/year. The scattering loss

OTF Lab at Lauritsen ROOM 38
Cavity #3 : We have introduced a new hydraulic fluid test sample as an alternative to Aquamil. It is pumping as right now.

Cavity #2 Test cavity : The chamber is pumping with new cleaned mirrors ~ 70ppm each.
We are taking RGA measurements.

OTF Lab. (W. Bridge)
Contamination Cavity # 1
We have changed new cleaned 50 ppm mirrors for the 70 ppm.
We locked the cavity, the output power is higher than before but the noise level
has increased by ~20%. We are checking the electronics, cabling and so on as to find the cause.
So far was unsuccessful. Hopefully will solve this situation soon.

Absorption Test Measurement prototype is in progress.

>From: Bill Kells <kells@ligo.caltech.edu>
>
>Sick most of last week, but working on:
>
>1. Finished e2e study of the experiment that
>   A. Melissinos and B. Butler performed in Nov. at LHO.
>   The Parametric response of the locked ifo was
>   measured, with some unusual features (beyond what
>   is seen, and model corroborated for a single arm
>   cavity alone). An e2e model of this configuration
>   gives nothing unusual (jsut the single cavity response).
>   So now there is an interesting mystery to look into!
>
>
>2. Have been (with Erika D'a.) putting together a rough initial
> analysis of the thermal lensing data that Peter F. and I
> gathered two weeks ago at LHO 4k (eloggd). We will have a
> preliminary result this week.
>
>3. Took a hard look at the usefullness of single mode fibers
>   in GW interferometers. Big problems, which I could not
>   resolve to give any cogent talk in Aspen. So this has been
>   shelved and item 2. above substituted.

40 Meter Interferometer (Weinstein)

Thermal Noise Interferometer (Libbrecht)

2. identifying differences between NAC and SAC that would account for their (slightly) different noise curves, and

The first thing we looked for, as a possible dominant noise source, was acoustic coupling to the optics and laser outside the chamber. Gently touching any of these components produces a noticeable jump in the noise as displayed on a spectrum analyzer, but this test does not tell us if this noise source is dominant, only that it is present at some level. If acoustic coupling were dominant, we ought to be able to listen to the error signal, appropriately amplified and monitored with headphones, and hear ambient noises through the instrument. We set up a system with which we could do this, listening to NAC and SAC through separate stereo channels. We found that, while touching the table produces a prominent, loud signal in the headphones, ambient noises such as fans, voices, and work on the nearby MINOS project are not readily identifiable. SAC's error signal sounds like very stable static, but NAC's has intermittent chirps, the source of which we are currently trying to identify.

On comparing NAC and SAC, we have identified one major difference: The response of the RF photodiodes are quite different. We tuned them to be approximately the same several months ago, but now they appear to be off by quite a bit. SAC's photodiode is much more sensitive than NAC's. We are in the process of re-tuning NAC's photodiode to make it as sensitive as SAC's.

On increasing the power, we find that SAC's lock becomes unstable quickly. This is a servo-gain issue, not strictly a laser-power issue. Because of its large discriminant slope, due to both a very sensitive RF photodiode and good alignment, SAC runs out of gain margin quickly as the power is increased, and our electronic gain cannot be reduced far enough to compensate without increasing its noise.

NAC, however, has a much smaller discriminant, so we have a lot more room to turn up its power and compensate by turning the (higher) electronic gain down. We find, as we explore parameter space in this arm cavity, that we can increase the power without any problems as long as we keep the unity-gain frequency of the servo below about 500Hz. When the unity-gain frequency gets higher than that, we begin to excite a violin-mode resonance at 506.95Hz, and a high-Q mirror mode at 53.808kHz. We spent some time this week building a pair of notch filters to take the modes out, and we have not yet had the chance to try them in NAC's servo with high power.

LASTI (Zucker)

Data Analysis and Computing (Lazzarini)

Matt, Hiro and Biplab discussed about the comparison of WFS signals
from e2e with those described in 1998 paper by Fritschel et al. After
that Hiro described the work done by Bill Butler, Adrien and Fred Raab
on high frequency signals especially at the free spectral range frequency.
Bill Kells tried to reproduce their transfer functions using E2E's
modeler-freq.

(Biplab & Hiro) Differences between E2E and FFT results have been resolved.
They produce essentially same WFS signals but still there are some
differences with the '98 paper by Fritschel et al. Discussed these with
Daniel.

(Matt) Began testing of new E2E optics components (computation primitives
Vacuum and Mirror).

Alfi
----

(Bruce)
- Continued implementation of the saving of options, search
paths, and other information from session to session.
- Completed initial implementation of WebStart distribution/
execution system for Alfi.

(Melody)
- Started working on optimizing the modeler's FUNC_xxx modules.
Currently trying to understand the current implementation.

(Blackburn)
The performance of LDAS during the weekend long E9 Run was far less that
we had hoped for at this stage in LDAS code development. The LDAS system
at Hanford way about 12.5% of its jobs terminate in error conditions, while
at Livingston about 6.5% of the jobs terminated in error. This was compounded
by bottlenecks introduced by the failure modes which would delay subsequent
jobs resulting in nearly the same precentages of jobs at the two sites being
delayed access to the system. The primary failure modes were those reported
from previous weeks testing:
- 3600 second timeout for jobs in the frameAPI
- MPI_INIT errors when starting up parallel mpi jobs on Beowulf Cluster
- Deadlock condition between metaDataAPI and the lightWeightAPI
Of these, the deadlock was responsible for the majority of the problems. We
had hoped that this failure mode would not be exercised regularly during the
run. However, the Burst group introduced a new set of scripts which we didn't
have during our testing which regularly triggered this failure during the
first day. Saturday afternoon these scripts were modified by the Burst group
to not use the database and the problem was significantly reduced, though it
did continue to occur a time or two after the changes.
Compounding the problem was the fact that this deadlock required a human to
log into the LDAS system and kill the metaDataAPI process by hand. When the
network problems from the worm hit, we were unable to log on and kill the
deadlocked API, thus resulting in continued job failures until someone could
be contacted at the sight that was knowledgable on killing and restarting
processes in LDAS. NOTE: Normally the LDAS system would automatically kill
a crippled API and restart. Several changes to the OS and SSH were not being
properly mapped onto the new LDAS version causing these automated clean-ups
to fail and requiring human intervention. I (Kent) spent a considerable amount
of time "baby-sitting" the LDAS systems at LHO and LLO, staying up most of
Friday night battling the worm in an attempt to get onto Livingston and fix
the deadlock there. This deadlock occurred at Hanford also. Most instances
were caught and fixed in less than an hour. However, a few deadlocks lasted
several hours.

On the up side, we see jobs coming through from the upper limits groups right
from the start of the run. In fact most searches were tested and running at
about 11 AM on Friday. There were issues with the new scheme for using the
calibration information which caused most of the DSOs from LAL and LALwrapper
to not support on-line calibration. This was also compounded by the worm as
we were unable to "pull" source code and "push" new LAL/LALwrapper libraries
to the sites while the internet was out.

(Blackburn)
Since the outcome of the pour results for the E9 usage of LDAS, a very focused
protocol has been enforced for coding and testing in LDAS. As a result all of
the failure modes mentioned above now appear to be fixed. I am being
conservative and saying "appear" because I want to do more testing still. In
fact, we have been able to run the same jobs from the Burst Group which were
so effective at tickling these bugs at twice there peak rate on the LDAS-CIT
system with the new code base without seeing the problem over night.

There are a couple of new issues which we are now working on. The frameAPI
was seen to core dump twice in 28 hours of testing with the new code. Changes
have been identified in both the TCL and C++ layers of the frameAPI which will
be tested this evening.

We have now ran through all of our test suites twice since the end of E9 and
found only one un-expected failure in the frameAPI which has now been fixed
and confirmed to pass its test.

We also spent a significant effort developing new tests to further stress the
system and to more closely mimic the way the Upper Limits Groups will use LDAS
during the Science Run. We now have the driver scripts from several of the
Upper Limits Groups and have run then together with the createRDS commands on
a 3600 stretch of pre-E9 data without problems.

We are now back up to about 700 to 750 jobs per hour on the systems. This is
up from the 600 per hour seen prior to E9 but still down from the 900+ seen
with the 0.5.0 release of LDAS. The bulk of this slowdown is now believed to
be due to bugs in GCC 3.2.1 which prevent us from optimizing the code under
Solaris. A new minor relese of GCC (version 3.2.2) is expected out the day
before the science run. Not enough time to test and integrate it into LDAS,
but we are optimistic that it will solve our optimization issues.

(Philip Charlton)
Fixed complex() action in datacon so that
- meta-data and dynamic type is preserved.
- made second argument optional eg. complex(x) is the same as
complex(x,0).
- added documentation for same.

Added code and options for applying calibration to convert ADC counts to
strain (in addition to existing strain -> ADC counts code).
- (not in CVS yet)

Extended respfilt() function with extra optional parameter to indicate
"forward" or "backward" transfer function.
- (not in CVS yet)

(Dan Kozak)
* Installed Solaris 9 on ldas-archive.
* Worked with Sun/Qlogic to install 2Gb FC switch beta hardware.
I Am continuing to work with them to make this hardware work.
* Received confirmation that the 1600 9940 tapes ordered to archive
S2 data should be delivered to Caltech before the start of the run.

(Al Wilson)
* Minor tweaks with BigBrother at LHO and LLO (limits and site monitoring).
* Proving that a replacement drive (Maxtor 200GB) will work with the 3Ware
card in the Datacache systems. The results are the this drive will work.
Also after talking with the factory the Western Digital (all sizes) will
not work at this time.

(Stuart Anderson)
- Working with Sun to hopefully finalize the SAM-QFS license deal.
- Working with Dan Kozak to help evaluate new 2Gb/s Fibre Channel switch
and Host Bus Adapter equipment.
- Keeping the software synchronized on all of the LDAS systems as several
pre-releases of LDAS and LAL where tagged before, during, and after E9.

(Keith Bayer)
- Installed o/s on rambus unit.
- Completed burn test on both units.
- Benchmarked both ddr and rambus units.
- Troubleshot rambus hyper-threading on multiple o/s's.
- Purchased Intel D350EMV2 board as possible workaround to ht issue.

(Igor Yakushin)
- Replaced failed T3 disk;
- Generated 96 channel and AS_Q only RDS during E9;
- Had to troubleshoot LDAS when jobs got stuck in metadata API.

(Shannon Roddy)
- Finished the first set of benchmarks on the evaluation units.
I will run some more tests tomorrow.

(Greg Mendell)
- The tapecontrol script ran at LHO and LLO during E9, successfully
archiving the raw frame data during the run. The S2 configuration of
the script, which writes dual copies of the data to tape, ran without
problems. The tapes have been or will be sent to Caltech.
- Updated the rds loop script to run under monitoring by the Search
Summary Page. Ran the script during E9 to generate the level 1 RDS
frames using LDAS at LHO. The script was able to run in real time with
other jobs running on a fully loaded LDAS system. The E9 RDS frames are
visible to LDAS and available at LHO under /frame20/rds/E9.
- Replaced failed hard drives in the ldas gateway E450 and controlmon
computers.
- Currently working with Dave Barker to test a backup data archiving
solution for the mass storage room and to set up ldas VPN servers.

- This week I continued work on routines to convert data from
radio-pulsar catalogues into a form suitable for LIGO targeted pulsar
searches.

(Keith)
- Moved NIS+ server back to NW17 due to emergency power
- Disruption in NW22 last Saturday
- Ordered more IP addresses
- Ordered desktop for professor
- Patched several windows boxes
- Ordered UPS box for gc servers in NW17-011
- Ordered AP-2000 wireless hub to extend wireless to all area's of NW22

(Shannon)
- Wrote and installed a script over the weekend to monitor the cryo pump
levels and several pressure gages. This script then emails Allen,
myself, and the operators if there is a problem. It also emails Allen's
cell phone.
- Looking into more network problems with LSU. People have been having
problems maintaining connections to various places. Also, some web
pages are not coming up. Unfortunately, there does not seem to be a
clear pattern to the problem.
- Setting up a laptop for one of our visitors. I should have it
ready by the end of the day today.

(Christine)
- Repaired networking files and printers on two laptops.
- Created a new user account.
- Continued installing Solaris 9 on GC suns.
- Investigated more bandwidth problems on the WAN and some internal
problems with the ATM hub.

(Mike)
- Loaded two new laptops with GC/engineering software packages, which will
be used to swap out Helena Armandula & Rich Abbott.
- Worked on the servers getting an updated ghost image on all NTSRV's; I
still have one more server to update.
- When I ghosted Pictor I notice a lot of system errors that are being
logged. It seems that this server is running at an intermediate state. I
fixed what I could, but it looks like the daughter board is going out,
which means this server well have to be replaced. We have a replacement in
the process of being ordered.
- There is a problem with one of the third floor access points located in
West Bridge. I did some thorough troubleshooting on this access point, but
it is acting real flaky and might have to be replaced. This is the AP
located where the printers are, not the conference room AP, which is
running fine.
- The 5000n printer on the 3rd floor of West Bridge seems to have died I did
what I could to try and fix this but was unsuccessful. Larry also tried and
was unsuccessful, after completely taking the unit apart with the HPC rep.
the culprit paper jam was found. It was well hidden in the unit.
- Performed a lot of user support this week that included networking, viruses,
and hacking issues that I have successfully corrected.

(Veronica)
- Spent most of the time working on Advanced LIGO proposal.
- LIGO website: updated various portions (Undergraduate projects, Aspen,
document templates, conferences, media website). VHS tapes: ready to place
the order with the DMC as soon as I have payment details. Tracked down a
problem with a DCC document. Updated laboratory Chart.
- CaJAGWR website: taped a video of last seminar. Did the usual website
upkeep.

(Larry)
- Spent a great deal of time monitoring the systems during this last worldwide
worm attack. We were fortunate and did not have any major problems with the
equipment but access to the Caltech network through normal routes was
difficult for a good period of time. Still seeing some minor issues with
other locations.
- Worked on a number of procurements. Received a copy of the CISCO PO and will
be going over that with the Campus IT people to make sure we get everything
we need.
- Performed a new SUNpci install on a SUNblade box. The SUNpci1 card installed
OK but installing WIN2K as the OS takes a great deal of time and so far it
has some major issues when booting up. The unit also runs very slow. I will
be going back to windows 98 and it is recommended the the SUNpci II card be
used if WIN2K is to be installed.
- Working a number of network issues. Trying to track down a problem with one
of the edge switches. If the firmware upgrade doesn't resolve the issue we
will request a replacement for the switch.
- Worked with Mike on a couple of virus issues and PC reworks. Still adding
sites (4-6/day) to be blocked by the mail server.
- Started on the monthly backups. Lisa left everything in pretty good shape,
even made up the labels before leaving on vacation.
- Working on a IDEAS problem with Janeen. Also, checking on the licenses. I
have conflicting messages. One indicating our licenses should expire in the
next day or two.
- Trying to get the new matlab licensing sorted out. Once again our rep. has
changed so it is taking some time to get him up to speed.
- Regular user support and cleaning of accounts.

LIGO II/Advanced R&D (Sanders)

Received from MLD the annealed substrates coated with 30 layers of SiO2/Al2O3.
The parts were sent for "Q" measurements to Glasgow (P.Sneddon) and Syracuse (S.Penn). Also, four coated 1"dia. parts were sent to R.Route for absorption measurements.

Loading experiments (16 kg, different times, several days) on a silica/sapphire bond failed to show measurable creep at the 5x10^-13 m/sec level.
We'll be re-checking for creep after the same bond gets re-loaded while being heated to ~30.5 degrees C for 24 hrs.

The load on the SF4/silica bond has been increased to 8.5 kg (in sheer) without problems; the other bond continues under tension (20 kg). Next, the parts will be tested while the temperature is increased.

SolidWorks
Janeen and I are going to attend the Advanced course in April 2003, after Mike Lloyd's visit. (We also hope to attend the ANSYS course around the same time)

Assembly
Janeen and I have been looking at adding the specialised parts to the assembly and also creating drawings for these parts. For example rounded bolts, studding and alternates)

Mode Cleaner Suspension
I have been helping Mark and Janeen with the resonant testing of the Mode Cleaner structure, please see their respective weeklies.
With the help of Russell Jones in Glasgow Helena and I have been adding some upgrades to the coil assembly. These should be re-installed onto the suspensions later today.

Eddy Current Damper
Once the updated copper array has been machined Phil Willems and I will test the 2x2 array on a single pendulum in the Syncotron.

Recycling Mirror Suspension
Caroline, Norna and I held a conference call to discuss the parameters of this suspension and will hopefully have a finalised set by this time next week. Mike Plissi, Eoin and I also held a conference call to discuss Eoin's ANSYS model of a cantilever blade.
Mike Lloyd will visit Caltech from the U. of Glasgow from the 16th if February.

Two full size sapphire blanks have been delivered to Caltech from Crystal systems. Phil Willems will get them first for Q measurement.

OPTICS PARTS LIST
Received parts:
Re-worked half-wave plates were received from KLC, for the Faraday isolator.

Orders pending:
Input mode-matching telescope off-axis parabolic mirrors are expected end of Feb. from SORL; drive electronics to drive the x and y axes of the 2nd PZT vacuum IFO steering mirror are pending from Piezo-Jena. MMT telescope parts are being reworked. Alternative mirror holders for holding spherical mirrors are being fabricated. Some of the mechanical parts for magnified focus lens are still in process of being fabricated.

IFO MODE-MATCHING TELESCOPE
A correct drawing was received from Piezp-Jena for the hole pattern in the PZT mirror. Modifications to the MMT drawings were made and parts were sent out for rework.

FARADAY ISOLATOR
The Faraday rotator was received from EOT, and the actual part dimensions are different than the data sheet dimensions. Therefore, several fabricated parts of the Faraday isolator assembly will have to be reworked to accomodate the actual dimensions of the Faraday rotator.

AdLIGO Suspensions
Creating drawings for fabricated parts. We'll concentrate on updating the assembly in Solidworks soon.
Updated the osem cost spreadsheet with hybrid osem costs and sent it to Russell Jones in Glasgow for review.

Mark Barton brought a triple pendulum over to the machine shop and clamped it to the milling center (approx. weight: 55,000 lbs) to measure the structural resonances. The first mode was ~50 Hz. FEA modeling last June in Algor reported a first mode at our goal of ~150 Hz, for a simplified model. Last June, the first set of analyses came up with ~50 Hz, but it was said that Algor glitches when using mm, so cgs unit system was used (Solidworks model was converted to cm and imported into Algor.) The cgs model was the one with the 1st mode at ~150 Hz. I now feel all Algor results are suspect. It has been quite troublesome this past week when trying to remesh and reanalyze models.

Mark, Phil, Calum and I discussed the problem. Mark is working on a mathematical analysis to support the results. He's performing more experiments too. Calum designed some stiffeners, which when fabricated and installed, got the 1st mode up to about 72 Hz. In the meantime, Luke Williams confirmed the centerline to centerline placement of the MC1 and MC2, along with the 1ppm radius so that we can look at using 2" x 2" structural members (like the LOS) for the structures' legs, as opposed to the current 1.25" x 1.25" legs. The LIGO 1 LOS structures are very similar in footprint and have a measured 1st mode above 150 Hz. We will identify the mode cleaner parts that need rework/fabrication to fit into a new structure.

Larry Wallace made some changes to allow me access to IDEAS master series 9. I've imported a Solidworks model into IDEAS, via STEP, successfully. I'll run that analysis this afternoon.

Weekly Report for Week Ending January 30, 2003

Special Announcements:

LSC Issues (Weiss)