LIGO Weekly Report

There was a site teleconference held on Thursday, September 26, 2002. The following issues were among those discussed:

Tropical Storm Isodore--The storm was not bad in Livingston. The main impact was further east. The road did well, and power lines are intact. Livingston prepared a set of draft procedures for preparing the facility. Sections, especially concerning computing equipment, passwords, etc., drew lots of comment and good advice. It was proposed that we should formalize the procedure. It was also proposed, tongue deeply embedded in cheek, that we pass it along to Hanford to be used next time they have floods or hurricanes.

Budget--It was noted that new indirect rates have been issued for FY 2003 to cover increased medical benefits costs. I estimate that this will cost LIGO approximately $150K per year.

Equipment--We have distributed an email to LIGO.ALL describing Caltech Finance modifications to the ORACLE systems to enable them to capture equipment data directly from their files instead of using the E. Chargois database. The P-Card systems will be changed starting October 1 to guide people to the proper selection of procurement category. We will followup with selected individuals with the biggest impact.

Livingston Staging Building--A not-to-exceed purchase order is being prepared for the contractor, Excel, selected to do the finishing and furnishing of the staging building at Livingston. The building contractor has still not fixed the HVAC or the auditorium stage. Ed Jasnow will be traveling to Livingston in October to close out the building contract.

Hanford Laboratory Building--The contractor is working the punch list slowly, but promises action. Funiture is being moved into the building.

All Terrain Vehicle for Livingston--The first dealer selected attempted some slight-of-hand with the proposed price and options selected. A secondary dealer is being approached.

Use of Auditoriums by General Public--Ed Jasnow has prepared an email for NSF concerning the use of the auditorium by outside organizations and the notification of the DOE concerning the same. The action (#114) will remain open pending people having an opportunity to review it to assure that it is adequate.

Foreign Travel--Problems have been encountered by non-US citizens on foreign travel attempting to re-enter the country. Since the situation is so fluid and the jeopardy high, LIGO will not support travel outside of the country if the traveler does not have a green card (unrestricted access). When Irene gets back, she will be asked to go over all travel to collect list of currently planned foreign trips. If you are holding a foreign passport trips will be cancelled unless we know that we can get you back in.

LN₂ delivery--We discussed how to avoid empty LN₂ tanks at the sites. The problem in Livingston was blamed on faulty instrumentation. The instrumentation was replaced last week and calibrated yesterday. Interference with operations will continue to be minimized. Typical intervals between fills is three or four months.

The list of current actions revised to reflect the status of open actions assigned through September 26, 2002 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. . .

Advanced LIGO (Frey)

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

We held a LIGO Change Request Board (CCB) during the Executive Committee Meeting on Monday, September 23, 2002. The following Change Requests were approved:

CR-010012 Revision B	WBS 1.4.4.1	Closeout Construction Budgets for Initial Computer Equipment Complement at the Sites	P. Lindquist
CR-020012	OPs 1.6	Additional Funds to Complete 40-Meter Facility Upgrade	A. Weinstein
CR-020015	WBS 2.11, 3.11 OPS	Low Noise Dust Monitors (For Information Only)	D. Coyne

CR-010012 Rev B--Closeout Construction Budgets for Initial Computer Equipment Complement at the Sites.
CR-020012 Rev A--Additional Funds to Complete 40-Meter Facility Upgrade (Revision A divided the original request into immediate needs and longer term, more uncertain needs. The immediate needs were approved. The remainder will be placed in an FY 2002 Planning Package.

Quality/Safety (Tyler)

Continued working with Ruth Brambila to update as required, status of several JPL support Work Orders.

Worked with Ed Jasnow and Ruth, to establish an acceptable (to JPL), system for processing funding and JPL Requisitions for the Raytheon support CWO through IPAC. JPL had informed us that they were no longer able to provide this support. To replace the JPL support, LIGO needs to have a method for "local" direct access to the JPL financial system which IPAC has now agreed to provide. Distributed a memo naming the two IPAC people that will access the JPL system for LIGO.

Meanwhile working with Ruth to help resolve LIGO and JPL CWO funding discrepancies. This effort will require additional work by the IPAC folks to obtain prior years JPL billing history which is unavailable from the JPL folks. Ruth and others will then have to plow through the billing history data to discover where and why the funding discrepancies occurred and to figure out how and what is needed to reconcile the funds data.

Reviewed the status of LLO Safety Audit action items with Gerry Stapfer. A status report was distributed.

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

Summary of Commissioning Activities at LIGO Hanford Observatory (compiled by F. Raab)
-------------------------------------------------------------------------------------
(see ilog for details - if it's not in the ilog, it didn't happen...)

Daniel Sigg has improved DTT yet again with two new features. The first allows the export tool to export any trace displayed in the plot windows. This means that you can now export a calibrated spectrum! Previously, one could only export raw counts. We expect this will greatly ease the handling of reference curves, especially when the calibrations are changing over time. The other new feature, FOTON, provides a GUIfied tool for digital filter work.

2K IFO (H2)
-----------
Jay is up working with Richard and Josh to get the new digital controls working. It looks good for getting the input optics going, although Jay had some reservations about confusing signals from SM2 this morning. Once this is resolved it looks good to align and try locking the mode cleaner.

4K IFO (H1)
-----------
Some software issues still need resolution, but they do not prevent locking of the interferometer. We have had mixed results in transitioning to common mode and Dave and Rolf are contnuing to study timing issues that may be related. Bill Kells is visiting to study optical properties of the interferometers and we have a team working to nail down frequency noise issues before Francois Bondu's visit ends. We are giving these investigations first shot at H1. A one-arm frequency-noise measurement done this past weekend suggested that the frequency-noise improvement obtained in July by realigning the REFL ISCT path may have degraded significantly. We have also found that notches in the mode-cleaner servo are not what they used to be. These two issues are getting attention from the operators.

4K Frequency-noise Investigation & VCO Calibration
--------------------------------------------------
Francois Bondu, Luca Matone and Rick Savage

We continued this week with the frequency noise investigation. An old
simulink model has been modified to reproduce the total loop gain in the
frequency stablization servo. The model reproduces well the loop gain for
frequencies below 100kHz.

We also revisited the VCO calibration. First we calibrated the PZT
actuator of the PSL and found 7.17MHz/V. This was done with the PSL not
locked and ramping the PZT: the drive was compared to the separation
between the carrier and the sideband. We then locked the PSL and we drove
the VCO within the bandwidth of the FSS. The AOM would shift the laser
frequency, and the shift would be compensated by the FSS loop. For
frequencies below the PC/PZT crossover (40kHz), the PZT would correct for
the laser frequency shift. Using the PZT calibration, we inferred the VCO
calibration and we found a calibration factor for MC_F of 41Hz/cts at DC
with p @ 0.57Hz and z @ 40Hz. This calibration differs from the one
(30Hz/cts at DC ,p @ 1.6, z @ 40) previously used by a factor of 2x. We
measured the electronic gain used to drive the VCO and using the
calibration previously found, we built a simulink model that reproduces
well the different measurements.

Finally, we used the VCO calibration to estimate the frequency noise seen
by the MC -- the noise dips down to 1e-2 Hz/rHz at 1kHz. Details can be
found in the 02sep26 elog.

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

Tropical Storm Isidore: LLO was prepared for the approaching storm on Wednesday afternoon. Vacuum systems, computers, and electronics were put in a safe state and the area around the buildings was cleared of materials that could be blown around in the wind. We coordinated our activities with the Livingston Parish Office of Emergency Preparedness in both timing of the shutdown and in getting an all-clear to start back up. The storm was downgraded from a hurricane to a tropical storm, so the biggest threat to the area was that gusty winds and high rainfall could interrupt electrical service. The storm had been projected to pass right over Livingston Parish, but at the last hours before landfall veered eastward, so there was very little damage in the Livingston area. The site operation is being restored today (Thursday).

Interferometer: Installation and commissioning of the digital suspension systems for the large and small optics continued this week. The SOS check out was completed and locking of the mode cleaner began Wednesday. Check out of the LOS in the corner station got underway on Monday. Mike Zucker is here to assist with this. Andri Gretarsson continued studies of the wave front sensors, using an AM laser to check out the entire system by illuminating individual gains and comparing responses.

HEPI /LASTI
Assembly issues have arisen with the bypass of the welded actuator. Been
through several variations of a vacuum braze process which have all resulted
in thermal distortion of the diaphragm. We've redesigned components to use
a threaded, mechanical process to restrain the diaphragm. Machining today
and should have first article tomorrow. We need a satisfactory resolution
of this problem because it is preventing assembly of the remainder of the
actuator. Marcel is actively setting up the 3D TeamWorks site per Dennis'
instructions. He is also seeing to it that we file all HEPI drawings in the
DCC, at the present state of design. Marcel is also placing all design
documentation in a clear and understandable directory structure on a secure
LLO server (Basin?).
Hammond/Kern

ETF
Vendor did an abysmal job machining the Pods, as well as the restraining
mechanism for the geophones. We've also found that neither prototype nor
production Pods fit in the ETF. More time than we can spare right now will
be needed to get to the bottom of the interference so we're deferring this
until we complete the HEPI documentation.
Hammond/Kern

All of the ETF parts as well as the HEPI and MEPI systems have been loaded
into 3Dteamworks at the request of Dennis Coyne. (Our online collaboration
tool to be used by the different subsystem teams to interface the digital
models) The models of LIGO one from Hytec will soon go into 3Dteamworks,
but we needed more disk space from the vendor. Janine Romie and Callum
Torrie have offered a great amount of assistance with the upload of the
models as well as overall functionality. We have also talked about
configuration of the models and how to make the digital content available to
all interested parties while maintaining documentation control. After
conversation with Jonathan Kern, Janine Romie, and Deanna Chun, I have
started to develop a plan that will accommodate this intent. With the use of
PDMWorks (an online documentation control program that acts as a "vault" we
will keep the parts configured as "in-work", or released. The in-work parts
will reside in 3Dteamworks. The drawings in the vault will mirror the
drawings in the DCC with the appropriate designations and would follow the
same change cycle as the drawings in the DCC. I have developed parts
hierarchy for LIGO SEI that segregates the different components as well as
the different programs i.e. Advanced LIGO. External pre-isolation, etc...in
a clear and organized fashion. Jonathan has asked to put Readme text files
in each directory to serve as navigator. Much of the work is in the planning
stages and Jonathan has asked Shannon Roddy to Set up the PDMWorks part
server and I am working on making sure that the subsystems work as intended.
I will present the results of these efforts to Jonathan Kern by Friday, the
27th of September and then I will get the information to the usual
interested parties.
(Hammond)

Optics/Laser Safety
Assimilating all LIGO documentation that can be gathered, on which to base a
thoughtful analysis of the clipping/appodizing problems at the dark port.
Mike Smith has been very helpful in sharing his 40m experience with us. Next
step will be to perform a Gaussian wavefront analysis of the LLO 4k IO, and
probably design an improved ISC beam reducer telescope.
Kern

LDAS admin:
1) Applied for Cybernetics Annual Extended Maintenance for the tape
robot.
2) Studying our DB2 license agreement with IBM.
3) Had to shut down LDAS on Wednesday due to Isidore tropical storm.
Turned it back on today. Everything seems to be working OK (running
tests now) except for eth0 on the new RAID: it cannot see the outside
world. Troubleshooting it now.
4) Writing proposal for DB global use across LIGO lab sites.
Data analysis:
1) Writing waveDSO.
2) Learning about grid computing
(Igor)

On Wednesday we shut down all the computers and network facilities
at Livingston in preparation for Isidore. Today the storm has passed
and we are turning everything back on. Our servers are all currently
running, and we are in the process of turning individual workstations
back on. We have taken advantage of the server shutdown to make a low-
priority configuration change so that the last services provided by
our old Solaris 8 file server have been transfered to our new Solaris 9
server. (Evans, Roddy)

Detector/Technical Support (Coyne)

DETECTOR SCIENCE & ENGINEERING

For the two weeks ending 9/25

Seismic Upgrade Project

Hydraulic External Pre-Isolation (HEPI)

Actuator

Jonathan Kern, Marcel Hammond

Assembly issues have arisen with the bypass of the welded actuator.Been through several variations of a vacuum braze process which have all resulted in thermal distortion of the diaphragm. We've redesigned components to use a threaded, mechanical process to restrain the diaphragm.Machining today and should have first article tomorrow.We need a satisfactory resolution of this problem because it is preventing assembly of the remainder of the actuator.

SEI Solid Model Files

Marcel is actively setting up the 3D TeamWorks site.He is also seeing to it that we file all HEPI drawings in the DCC, at the present state of design. Marcel is also placing all design documentation in a clear and understandable directory structure on a secure LLO server (Basin).

Pump Station

Ken Mailand

System tests are finished, a motor control circuit is being fabricated and should be complete by week of 9-23.

The manifolds [5] with their mounting bases shipped to MIT, Drawings in dcc.

Shipment ot LASTI is likely to be delayed to make the prototype pump station leak free and ready for glycerin/water mixture; Final decision of the configuration/state of the pump station is expected 9/27. Final decision on the fluid choice is pending further investigation and may not be made until just before wetting the system at LASTI.

Electro-Magnetic External Pre-Isolator (MEPI)

See the LASTI report for details on the MEPI testing. Highlights:

·Prof. Samir Nayfeh and students for MIT/ME dept. are conducting a modal survey of the HAM support structure

·Rich Mittleman et. al. are debugging problems in the sensor interface board electronics; This has delayed system identification work

·Modeling on the dynamics and controls continues

CDS Software

Rolf, Alex and Lori have just returned from visits to both observatories.

The primary changes to the code after S1 were to make all the

digital filter systems as "generic" as possible, per request from Sigg,

Peter, etc. This involved:

-- Renaming most EPICS channels in the digital servo systems and

making the appropriate channels in the EPICS state code which

communicates into to/from these channels.

-- Matching up the front end code, such that channel differences

between various optics ended up looking like the same channel on every

EPICS screen.

-- "Master" EPICS screens, from which all optic screens could be

automatically generated for each IFO and each site.Besides the optic

screens, LSC and ASC screens are now also generated for all IFO in this

manner.

-- Code which is commonly used in all digital servo systems was moved

out to separate code objects, which are then compiled with each digital

system. Initially, various versions of these routines were embedded C

code in each front end.Examples of these objects are:

-- Code which performs the digital filtering.

-- Code which reads digital filter configuration files.

-- Code which knows how to handle/route GDS signals.

-- Pentek initialization software

-- Pentek timing card software.

As far as any "functional" changes, these were limited to:

-- Each filter can now have 10 second order sections (SOS). Previous

limit was 4 SOS.

-- Every filter bank in every digital servo system now has 1 GDS

excitation input and 3 GDS readout test points.To accomodate this in

the large optics controllers, software was added to the ASC frontend to

route GDS signals between the ISC control network and the DAQS network.

At present, the each optic is limited to one excitation signal and one

TP at a time (per original spec), but will be upgraded in the next few

weeks to allow for multiple TP up to six.

-- LSC and ASC output filter banks were removed.With the digital

suspension system, these outputs now feed a filter bank at the LSC/ASC

inputs of the optic control software.

-- Old CM+DET mode of LSC removed, as CARM error now sent to MC2.

-- Mode cleaner length servo signal is direct ADC input to MC2

controller.

-- Timing diagnostics were added to LSC frontend code to check timing

between LSC controller and all suspension controllers.

In general, a history of all CDS code modifications can be found in the

CDS branch of CVS, maintained by LDAS.The CVS tree can be accessed

from the ldas-sw.ligo.caltech.edu web page.Specific instructions were

entered in the LLO elog last night.There are also web page links from

the LLO and LHO CDS pages which show all code loaded on all CDS CPU,

along with the revision numbers.

In addition to this info, when large charges are made, such as this post

S1 change, I will email this group with a list of changes.As small

updates are made, they will have elog entries at the sites.

CDS Hardware

Jay Heefner reporting

·Low noise DAC: We have directed Freq Devices to proceed with the galvanic isolation option for the module. We have also asked them to use our two pin lemos as the clock input connector.

·FPGA Timing Module: Further development is awaiting the test board.

·LHO 2K DSC Upgrade: Have assisting in the testing of the 2K DSC. So far the LVEA system is 90% checked out and all small optics appear to damp. LOS testing and damping will continue tomorrow and Friday along with mid station testing.

Rich Abbott reporting

·Work progresses on the motor speed control servo for the seismic retrofit. All the readbacks are functioning and we are in the process of closing the digital feedback loop to the motor speed control unit.

·I am on travel at LLO and assisting with the trouble shooting of a power supply related multi-board failure, and in general, helping with the debug of the DSC.

Mohana Mageswaran

I am working on the changes and corrections of the existing Seismic Interface Box (sensor interface board for the seismic retrofit)

I am doing a schematic and a PCB layout of a voice coil driver test Board (adv. LIGO seismic system component).

Sander Liu

LEMO antialiasing filter chassis -shipped three new ones for LHO.

Micro-Seismic signal processor - In the process of building one new set for LHO.

In the process of testing a set of microseismic signal processor for LLO.

Obtained a quote to build longer data cables for the LEMO Antialiasing Filter chassis from SoCal Value Added for LHO.

DMT

John Zweizig

I continued to adapt the DMT for version 5 frames and test the

results. A new vers v2.3.0 has been tagged and seems to work on both raw

and process frame data in version 5 frames.

I have worked on looking through the data quality trends from

S1. I have attempted to refine the glitch frequency trends by using a

more intelligent pre-filtration which notches out the 60Hz and 180Hz lines

and uses a High-pass filter at 100Hz with a faster fall-off. This seems to

give larger rates as expected (the glitches aren't hidden under low

frequency noise), but the (L1) rates are still constant except where

non-science data sneaks in. The rds doesn't have the IFO state channel, so

this should be rerun from the raw data. I have also been working on

displaying the various IFO channel frequency band trends in a convenient

manner which will facilitate scanning for pathological periods.

PSL

PeterKing

After examining the remains of the optically contacted pre-modecleaner

at LLO - the one that was dropped in the summer - it appears that the

mirror attached to the PZT shifted alignment, whilst the optically

contacted mirrors and prism alignment was not changed.The PZT assembly

will have to be pulled off and re-aligned before being epoxied on again.

The curved mirror from the optically contacted pre-modecleaner #4 was

removed, re-aligned and re-attached because of a mistake I made during the

initial fabrication stage.The new alignment is much better.The output

from NPRO #259 was mode-matched into the pre-modecleaner and an initial

alignment was completed.The alignment will be worked on tomorrow.

The 80 MHz crystal oscillators from Wilmanco arrived.These will be used

as the reference for which the VCO phase noise will be measured against.

I have been looking at some Hamamatsu photodiodes for possible use in the

intensity servo.

Optics Analysis

Erika D’Ambrosio

I stopped working with William Kells and I made more simulations for the flat

topped beam proposed by the collaboration led by Prof. Thorne. I used the maps

Garilynn Billingsley gave us, to investigate how worse the contrast can be if

the mirrors' surface is not exactly the required one. Those imperfections also

bring a smaller reduction of the thermoelastic noise than in the ideal design.

Bill Kells

In transit to LHO, and just getting started with program to determine COC losses in H1/H2 ifos.

Optical Contamination Cavities

Lee Cardenas, Liyuan Zhang

OTF Lab. (Bridge)

Contamination Cavity # 1

Cavity still locked and continue taking ringdown measurements.

This cavity will be stopped as tomorrow due to a high absorption level (a few ppm). It is suspected that the mirrors aare contaminated by vacuum baking them in the chamber.

We'll start again this test(Mineral oil Vitrea 46).

Liyuan has accomplished working on the lab view program for the new OTF Lab.

Now we have a new lab view program running that we'll use it on the new set-up.

Liyuan is looking into the lab view program for the Scatterometer.

New OTF Lab at Lauritsen ROOM 38

Cavity #3 Reference Cavity still pumping down. Optical train ready.

Continue taking RGA measurements.It will be switched to the Ion pump.

Cavity #2 Test cavity

This cavity still pumping, optical train installation in progress.

Continue taking RGA measurements. I am still looking for a new roughing pump.

The old one needs new oil seal to be replaced.

I got the GPIB card already for the Lab View.It was given to me by Peter King.

This card has been tested and it works with our digital scope.

I am still looking for the lab view disc along with the license.

The NewOTF lab. is ready for operation.The Standard Operational procedure for both NPRO Lasers.

LIGO-100 mW and LIGO-700mW ND:YAG Lasers.was reviewed by Garilynn and it needs some

modifications to be completed.

Misc... tasks,ordering some new optics (mirrors, beam splitters etc.) for both OTF and PSL Lab.

40 Meter Interferometer (Weinstein)

Thermal Noise Interferometer (Libbrecht)

LASTI (Zucker)

Data Analysis and Computing (Lazzarini)

Simulation and Modeling (Bhawal)
E2E for LIGO I meeting
-------------------------
On Tuesday this meeting was convened by Dennis Coyne and David Shoemaker.
Matt presented the status of SimLIGO. There were discussions and feedback
on wave front sensors, phase map for mirrors, thermal lensing, intensity
and frequency noise, line noise and effects of the rotation of small
optics. One question was whether Hiro should implement the phase-map
calculation in current e2e environment or should wait until the introduction
of the new field model (being developed by Matt) which will have more useful
and realistic features (like mirror thickness, 2 surfaces etc). It was
agreed that it's better to implement it in the current environment for
starting studies of the problems at hand as soon as possible. There were
discussions on thermal lensing and whether there's need for repolishing
the recycling mirrors. The reliability of E2E's current approach to this
problem (that mimicks the effect by a variable refractive index) was also
discussed.

Thermal lensing
------------------
(Biplab) Using a simple matlab code, I generated plots for Mode-matching
radius of curvature (ROC) of the recycling mirror for given hot states of the IFO (determined by
recycling factor, bulk loss, surface loss etc). These preliminary
and simple calculations show that for the current parameters of 4k IFOs,
the ROCs of recycling mirrors are not so bad. Dennis, Garilyn, Matt and
I had a discussion on this. The predictive power of these simple
calculations needs to be tested by comparing with Bill Kells' old
analysis and some new Melody-studies made by Peter Fritschel.

Code development and maintenance
----------------------------------
(Biplab and Matt) Another bug in summation cavity was discovered
while studying common and differential mode of misalignments.
The bug was corrected and the changes are CVS committed.
Alfi
-----
(Bruce) Continued implementing cut and paste in Alfi5
(Melody)
- Continued working on the automated tester for alfi5.
- Worked on a new feature in alfi5 to export modeler files.

Software Systems (Blackburn)
Effort continues on the development of the new frameCPP I/O library. This
activity has fallen about a week behind. The complete functionality of the libarary
with regards to the specification is still not complete, primarily missing
compression support. However, integration into the frameAPI has already
begun. Testing performed late yesterday and last night suggests that the
frameAPI is pulling incorrect data out of the frame or that the data is
somehow mangled before it enters the DSO codes running on the Beowulf. An
unrelated bug in the frameAPI was also discovered this week which caused
data segments to be repeated in the output of the frameAPI and was detected
using filter algorithms in the dataConditionAPI.

A performance issue in the managerAPI was discovered this week which raised
the CPU load from the managerAPI to over 60% in a busy LDAS system. This was
quickly resolved with a slight tweaking of the code and the managerAPI is now
running at about 30% CPU load.

We are still experiencing orphan jobs, time outs and ocassional core files
in the new frameAPI code. These have been around since the performance of
the frameAPI was improved using multi-threaded calls. There has been some
improvements of these issues but they have not been eliminated. Investigations
continue.

A new version of LAL and LALwrapper were announced this week. They were
built and RSYNC'ed to all LIGO Laboratory LDAS systems.

Testing of LDAS under Redhat 7.3 is looking very promising at this time. There
were some minor issues to overcome and we still need to verify that the
results of dataPipelines are consistent with Redhat 7.1. However, it looks
like we will be able to support this newer version of that OS in the next
release of LDAS.

Greg Mendell of LHO gave the following (excerpted) report of his experiences
running LDAS this week:

The knownpulsardemod DSO ran (off and on) between Sept 20 2002 and Sept
24 2002, generating SFTs at LHO and LLO. Everything went very well,
with almost no errors except due to missing data. LDAS ran great!

There was one unusual error (that occurred twice), which may indicate
some problem in reading frames in a single job from QFS and NFS file
systems (with the latter exported from linux). This error should be
looked at more closely since it may cause jobs run by other DSOs to fail.

1. At LLO 680 jobs were run; 16 jobs failed. All the failure were due
to missing data on the disks. No failures were due to LDAS! (For
comparison, the failure rate within ldas for E7 SFT production was
1.9%.)

2. The SFT jobs input 2048 s of data. The mean time to run the jobs was
126 seconds, or 16.25x real time. (This is 16% improvement in
performance from E7.)

1. At LHO 1431 jobs were run; 21 jobs failed. All the failure were due
to missing data on the disks, except 2 that were due to the LDAS error
reported in Item 3 below. No failures were due to LDAS except possibly
these 2. (For comparison, the failure rate within ldas for E7 SFT
production was 1.8%.)

2. The SFT jobs input 2048 s of data. The mean time to run the jobs was
128 seconds, or 16x real time. (This is 15% improvement in performance
from E7.)

* Fixed a PCI conflict with GigE cards in the 5 new dual-Xeon LDAS servers.
* Continued archiving S1 to data to HPSS and SAM-QFS. The estimated completion
time is Friday morning (Sep 27).
* NFS mounted 16 1.1TB filesystems to the SAM-QFS server. The playback
of all S1 data to this disk storage will start on Sep 27 or 28.
* Finished upgrading the LDAS-TEST system to RedHat7.3 to evaluate this
is a suitable Operating System upgrade for all Lab LDAS systems coincident
with the next LDAS software release in a few weeks.

MIT
---
(Keith Bayer)
* Investigated bios versions for 3ware card in datawulf box.
* Received LDAS VPN boxes.

* Applied for Cybernetics Annual Extended Maintenance for the tape robot.
* Studying our DB2 license agreement with IBM.
* Had to shut down LDAS on Wednesday due to Isidore tropical storm.
Turned it back on today. Everything seems to be working OK (running
tests now) except for eth0 on the new RAID: it cannot see the outside
world. Troubleshooting it now.
* Writing proposal for DB global use across LIGO lab sites.

* Renewed Cybernetics' annual maintenance contract for the LDAS LHO
tape library/robot system.
* Node32 crashed over the weekend. These node have crashed in the past
and some have had motherboards, memory, and hard drives replaced.
Node32 had all of these replaced before its last crash. We temperatures
and humidity in the room have not gone outside of allowed tolerances,
they do fluctuate more at LHO in the staging building than we would
want. I restarted the burntest on nodes 25, 29, 32, and 48. If crashes
occur again, I may test these units outside of the staging building to
see if environment is a factor.
* Discussed with Otto the temperature and humidity situation in our
room, especially the need to perhaps add a humidifier to the room. We
have set the week of November 4 as the date to move the LDAS equipment
from the staging building to the new building.
* Generated SFTs on the S1 data for the periodic search and backed
these up onto tape (more about this is in a separate weekly report under
data analysis).

Data Analysis Activities (Lazzarini)
Shawhan:
* Fixed a bug in the LDAS job proxy server.
* Built new versions of GDS/DMT LIGOtools packages.
* Worked on design issues for a new trend-data viewing system.

Weinstein:
- Working on understanding the S1 calibration / response function
with Mike Landry (and with help from Philip Charlton).
Modulo the time dependence, I have it working and understood for burst
simulations.
- Developing more flexible burst simulation scripts for LDAS jobs,
in consultation with Peter Saulson.
- Testing new power DSO performance, in consultation with Erik and Patrick.
- Testing new slope DSO performance, in consultation with Ed.
- Working with Julien on whitening filters for S1.
- Working on understanding angular/polarization dependence
of IFO network (H1,H2,L1,G1...) coincidences.
- Gearing up for large-scale S1 burst simulation ldas jobs.

BAM (Bursts Analysis Meeting) #1 took place at MIT Sep 20-22.
Several aspects of the bursts analyses were addressed and streamlined.
All MIT Bursts group members (Stefan, Shourov, Laura, David and
Erik) contributed in almost 3 full days of discussion and hands
on work with Bursts DSO triggers and vetoes together with Finn, Ito,
Saulson and Sutton who flew at MIT. Members of the Bursts group,
including GEO colleagues, joined the activities via teleconferencing.
Specific tasks where the MIT group contributed included:
o importing GEO data to MIT, logistics and frame/calibration issues
addressing (Laura, Shourov)
o generation of vetoes and diagnostic plots (Stefan, Shourov)
o generation of DSO triggers and diagnostic plots (Laura, Erik)
BAM #2 is scheduled for Oct 4-6. Contact Saulson and Finn for details.

Mendell:
Generated the S1 LIGO 2048 second SFTs at the LIGO sites. SFT data for
H1 and H2 are at lho, and SFT data for L1 is at llo. The data is
available from LDAS and it has been backed up onto tape. Distributed
scripts, getsftdata.tclsh and ExtractILWDdata.tclsh, to the PULG group
to get SFT data (also in the /knownpulsardemod/scripts/ directory of the
cvs repository
:pserver:username@gravity.phys.uwm.edu:/usr/local/cvs/ldasmdc). The
scripts require an ldas password and the ligotools LDAS job package.

Below are GPS intervals for which S1 LIGO SFTs are available. The gaps
between intervals at LHO are due to reboots of the DAQ system and
represent missing raw data. Most of the gaps between intervals at LLO
occur for the same reason. However, at LLO some raw data from day 1 is
currently not on disk (though it is on tape) because of limited disk
space. This is mostly from times with Livingston was not locked, and so
will not significantly affect the periodic analysis, though eventually
we can get SFTs for this data too if necessary.

H P_LDAS_KPD_SFT_H1 714151644 714264284
H P_LDAS_KPD_SFT_H1 714266332 714333916
H P_LDAS_KPD_SFT_H1 714335964 714536668
H P_LDAS_KPD_SFT_H1 714538716 714860252
H P_LDAS_KPD_SFT_H1 714864348 715144924
H P_LDAS_KPD_SFT_H1 715146972 715189980
H P_LDAS_KPD_SFT_H1 715194076 715265756
H P_LDAS_KPD_SFT_H1 715267804 715400924
H P_LDAS_KPD_SFT_H1 715405020 715611868

H P_LDAS_KPD_SFT_H2 714150032 714264720
H P_LDAS_KPD_SFT_H2 714266768 714334352
H P_LDAS_KPD_SFT_H2 714336400 714537104
H P_LDAS_KPD_SFT_H2 714539152 714860688
H P_LDAS_KPD_SFT_H2 714864784 715143312
H P_LDAS_KPD_SFT_H2 715147408 715190416
H P_LDAS_KPD_SFT_H2 715192464 715264144
H P_LDAS_KPD_SFT_H2 715266192 715401360
H P_LDAS_KPD_SFT_H2 715403408 715620496

L P_LDAS_KPD_SFT_L1 714188064 714222880
L P_LDAS_KPD_SFT_L1 714243360 714433824
L P_LDAS_KPD_SFT_L1 714435872 714442016
L P_LDAS_KPD_SFT_L1 714444064 714695968
L P_LDAS_KPD_SFT_L1 714698016 714956064
L P_LDAS_KPD_SFT_L1 714958112 715171104
L P_LDAS_KPD_SFT_L1 715173152 715570464
L P_LDAS_KPD_SFT_L1 715572512 715580704

General Computing (Wallace)
MIT:
(Keith)
-Tested out VRVS audio w/ Larry. Things worked pretty good even though some
of the equipment was temporary for the test.
Hooked up VRVS to polycom phone system
(will test it on the VRVS Friday conference w/ Hanford)
-Ordered extra harddrives for ldas gc desktop machines
-Interfaced with PFSC regarding meeting space for NSF in Oct
-Received gigE card for lancelot (will install Friday?)
-Met with ESNET folks gathering some background networking info
-Talked with cogent - alternative network services are not available
at least in the near term.

Livingston:
(Tom)
-On Wednesday we shut down all the computers and network facilities
at Livingston in preparation for Isidore. Today the storm has passed
and we are turning everything back on. Our servers are all currently
running, and we are in the process of turning individual workstations
back on.
-We have taken advantage of the server shutdown to make a low-
priority configuration change so that the last services provided by
our old Solaris 8 file server have been transferred to our new Solaris 9
server.

(Shannon)
-Worked with Larry and finalized on the equipment to be installed for the GigE
network.
-Setup a couple of laptops and ran more network cables. Made a layout for
the new network.
-Ordered Solsoft software.
-Added more locations to be filtered on e-mail.
-Prepared for the storm.

Hanford:
(Christine)
- Set up the video conferencing computer in the new building auditorium.
- Purchased a new matlab toolbox from a third party vendor.
- Worked on some LHO GC Help web pages.
- Organized the OS upgrade for fortress, the LSC sandbox computer.
- Tracked down some orders that haven't arrived.
- The usual user support

CIT:
(Bruce Sears)
-General Ilog Maintenance:
       - Dealt with problems arising from the inadvertent mirroring
          of Ilog software and databases at Caltech.
        - Added users.
        - Troubleshot a problem with the LLO detector log.

(Mike)
-Reloaded Server with Win2K and some service packs, but have not had the time to
configure this as a webserver.
-Went over security logs and notice multiple attempts of unknown users
trying to hack into all NT servers. After speaking to Larry we decided to
tighten down on all servers with added security.
-Worked on servers trying to figure out what is going on with DCC batch
files that are not running as they are set up to do.
We are still looking into this matter.
-Upgraded Larry J.'s and Irena Petrac's, computers to 2000 Pro.
-Started loading a new High end engineering workstation.
-Worked on setting up a print server for Mondays Meeting.
-Larry and I ran two cable drops up to Gary Sanders office. The run was an
interesting one.
-Performed quite a bit of user support to attend to this week; that included
software/connectivity issues.

(Wendy)
-Continuing with the tape restorations.
-Worked on a couple of laptops.
-Assisted users.

(Lisa)
- Built a laptop for Matt Evans
- Ran some video patches on pollux
- Began looking into how to do certificate based access on the mail server
- Worked with Larry on resolving a DCC database problem.
- Still working on getting phone lines on the modem pool moved to digital lines.

(Veronica)
- LIGO website: finished Press and Media website, with a few
changes/additions.
Posted updates to Internal Bulletin Board, meetings, fellowships pages.
- LSC website: posting the late-arrival transparencies from the ast LSC
meeting as they keep coming in.
Updated MOUs, talks webpages.
-Looked into a problem with the LSC progress reports database. There were
reports that the progress report online submit form "messes up" the report
data. Need to look at the example to better understand what's going wrong;
waiting for a sample report from Irena. In the meantime, combing through
the database scripts to see what could have caused the problem.
-Working with Caltech Digital Media Center on processing a videotape of the
public lecture from the August LSC meeting.
-LDAS website: made a new webpage with the updated ilog links. Modified
the index.cgi script to reflect the changed pointer to ilog links.
-Project Science website: posting updates as they arrive. Working on
redesigning the layout of the original website to better incorporate its
growth.
-CaJAGWR website: checking if the site content is indexed in the new
search engine that Caltech is launching on 30 Sept.

(Larry)
-Worked with Donna and Mike on getting things ready for different presentations
and a upcoming conference. Along those same lines working out some of the
logistics with Keith for the NSF review.
-Worked on a number of documents for Albert.
-Assisted a number of users with various program setups and diagnostics of their
applications.
-Debugged a few e-mail issues and worked on a couple of DCC issues.
-Assisted a couple of other groups on Campus with their Sun orders.
-Worked/working with the electrical shop in getting more power into the computer
room. Hopefully, that will finish up soon.
-Worked a couple of network issues. Updating the campus DNS tables for
Livingston for some reason took longer than normal. Also, checking in to some
performance issues on the connection to the CIT wan. We had a few machines
getting very poor performance while others did not. Switching the port
connection to the Caltech backbone appears to have fixed the problem but we are
still checking things out to find out why this problem just appeared.
-Reviewed upgrades with Mike. He is really getting those done in a efficient
manner.

LIGO II/Advanced R&D (Sanders)

This week I hacked around with my modularized Matlab suspension model,
corrected a number of bugs and created a few useful derivative models,
including one with non-diagonalizing (pure SISO) active damping and one
with eddy current damping. I also dug out my old notes on Mal Gray's
interferometric sensor and started refamiliarizing myself with it. Phil
Willems and I plan to get it working again and see if it can't be made to
be less sensitive to tilt of the test object by incorporation of a corner
cube target.

I have been in contact with Kodak about ion beam etching for the purpose of creating Mexican Hat shaped surfaces.
Kodak does have the ability to etch very large substrates, greater than 2 meters. The standard approach is a raster scan with a ~1cm ion beam, using an input map along with known removal rates.
To date, Kodak has little or no experience in etching or polishing sapphire.
Kodak has an extensive large-aperture metrology capability, the current best measurement is .01 waves rms at 633 nm or 6.3 nm rms deviation from the ideal surface. Compare with 0.8 nm requirement for LIGO1
Kodak prefers to do full processing on optics rather than just ion beam etch and so would suggest that they bring the part to flat before ion etch. Polish on fused silica, for example, currently yields better than 5 angstroms microroughness.
Kodak is willing to do some investigative work, under contract, to determine the feasibility of providing optics with a flat top or Mexican hat-shaped surface.

I still have not tracked down the reason why the MV162 input/output
controller times out when downloading the vxWorks kernel. I hope to sort
this out when Jay gets back in town. Otherwise all other cards check out
fine.

A set of laser evaluation criteria was written by Ivo Zawischa. The
criteria will aid the decision process for the high power laser downselect.
The selection rules are not finalized yet.

Suspension
With the help of Janeen, Betsy and Helena we put the triple pendulum back together on Friday. It is currently suspended at ~0.3mm from the desired position. Betsy set up an optical lever arm for us.
Work is underway to compare the heights of the various stages to numbers from the SolidWorks and an AutoCAD assembly.
A third assembly will be carried out early next week.

Parts
Parts for a basic catcher and additional parts, some recommended by our various visitors, are being made at both campus machine shops and one local machine shop.

SolidWorks
I have been working on updating the mode cleaner drawing tree and completing, adding nuts, blots and comments, some of the assemblies associated with the mode cleaner.

3D TeamWorks
Marcel Hammond has now set up the SEI 3D TeamWorks site. We both help collaborative sessions on Monday and Tuesday lunchtime in order to demo and practise with the software.
Russell Jones and I are using the SUS 3D TeamWorks as the home location of files associated with prototype eddy current dampers for use on the Glasgow triple and the Caltech triple.

AdLIGO Suspensions
Business as usual.
Working on the viewgraphs for the technical review on October 8th.

Akiteru, Virginio (Hongo)
Vacuum problems and one contact problem in one tower, probably on filter 1, breaking vacuum to correct problems.

Alessandro
Completed setup of chemical lab, purchased chemicals plus nickel target for sputtering (with Brett’s equipment).
Prepared 12 electrodeless nickel Phosphorus plated samples at different deposition periods for process calibration, tomorrow will measure thicknesses in DSEM and will check composition with EDS.
After will proceed to deposition of pure sputtered Nickel and possibly evaporation of the braze layer.
Built pulsed Power Supply duplicate for electro-polishing of the MoRuB flex joint profile.

Stoyan
Analysis of the stress strain data increasing stress cycles, evidence of slipping on every cycle.
Ready to test new splat pistons when splatter free.

Maddalena, Barbara, Stoyan, …..
Splatting. Repaired splatter again, splatting away, made 14 and 15, back at 17 to generate more samples replacing the ones used in testing.

Barbara
DTA making more tests, buut always troublesome.
Looking at earlier data, identified two well repetitive peaks, probably due to BCC and Boride crystal precipitation out of the alloy. Baking samples at corresponding temperature to check on X-ray scattering for appearance of crystallite.

Xavier
Studying stress concentration in MoRuB flex joint for variable curvature radia raccording between the 50 and the 10 micron thicknesses.
Analysing heat conductivity data

Weekly Report for Week Ending September 26, 2002

Special Announcements:

LSC Issues (Weiss)

LIGO I Construction/LIGO Laboratory Administration (Lindquist)