|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday March 18, 2002 will be:
(Meeting time: 10:30 am Pacific Time)
Open meeting 10:30 - 11:30
Special Items:
Transition to Scientific Running (LIGO M020136-A-M)
Our prior memorandum (LIGO M010216-A-M) of August 2001, appended below, laid out the strategy for transition from commissioning to scientific operations, and interleaving these operational modes. In this memo, we define the progression of early scientific running in additional detail and create a basis for planning the important data analysis and interferometer commissioning and development work between the scientific running periods.
We have just completed a very successful E7 engineering run, as called for in the attached memo. Analysis of the data from this run and further interferometer commissioning and modification, at least partially based upon the E7 experience, is underway. We have carried out a set of "lessons learned" discussions in the Laboratory Executive Committee meetings, and a discussion with the LSC Executive Committee. The LSC has also held an important teleconference to expose the status of the data analysis effort with the E7 data. The E7 run also involved coincidence running with the GEO 600 interferometer and the Allegro bar detector. We have consulted with our partners in these efforts, as well, regarding accomplishments and the perspective for the future. These have all been quite constructive and the scientific, technical and operational issues have been examined and priorities that guide the next steps have been discussed.
Our overarching mission is to accomplish the scientific reach planned for the LIGO interferometer system and to exploit the system, with the LSC, to accomplish the science. Given this priority, we structure the next steps in the program around the progression of scientific running periods and the scientific reach for each of these periods. The development of the data analysis and detector systems, and the conduct of the scientific data analysis will then be guided by these science runs.
Our plan for the next steps includes:
Following the E7 Engineering Run -
1. The March LSC meeting should be the point at which we reach consensus on the sensitivity and noise curves from E7, as called for in the memo below. This consensus should help us benchmark E7 and the set the mark for the S1 run as we laid out in the earlier memo.
2. By the end of May, the analysis working groups should be able to complete much of the targeted data analysis tasks, as we called for in the memo below.
The Upper Limits Runs S1 and S2 -
3. Given the lessons learned from E7, the remaining high priority interferometer installation and rework tasks, and the goal to initiate science running described below, we will schedule the S1 run to begin at 8:00 am Pacific time on Saturday, June 29 and to be completed at 8:00 Pacific on Monday, July 15. This run will be the joint responsibility of the Laboratory and the LSC. The sensitivity goal is a two site coincidence with 3 interferometers running and the achieved scientific reach (volume searched x observation time in coincidence) should be an order of magnitude better than achieved in the E7 run. At least one interferometer at each site should be operated in the full recycled configuration.
4. The S2 run will have a goal of at least an order of magnitude improvement in scientific reach (volume searched x observation time in coincidence) beyond S1 and should follow successful completion of analysis of the S1 data. Considering the estimated time to carry out the data analysis of the S1 data, and the interferometer commissioning and development tasks, we will schedule the next science run to begin at 8:00 am Pacific Friday November 22, 2002 with completion at 8:00 on Monday, January 6, 2003.
These two runs will complete the upper limit running and the orientation for the LIGO running experience. We believe that this should lead to a broad set of new publishable limits, well beyond what has been previously published.
Extended LIGO Search Runs
5. Installation of the additional seismic filtering at LLO will be scheduled to follow the completion of the S2 run. A plan for the major additional interferometer commissioning, modifications and rework is under development with the goal of supporting the next major science run after S2 in the middle period of 2003. This significant shutdown period will be intended to achieve readiness for an extended science run with a significant step in sensitivity.
6. The S3 run will mark the beginning of true search running, representing a step beyond setting upper limits on selected gravitational wave searches. S3 will be intended to accomplish a real search for gravitational waves with significant astrophysical significance. We expect to schedule S3 to commence about June 27, 2003 and this run will be planned for several months duration.
7. During 2003 and 2004, we will plan to run in this search mode for at least 50% of the calendar time, followed by the planned one year integrated LIGO science run at design sensitivity. This science run will be completed prior to proposed major interferometer replacements.
With this framework, we will now develop a more detailed set of merit figures for setting goals for these runs. These planned runs will now be used to structure the detailed LIGO program plan for LDAS development, detector modeling and diagnosis, interferometer commissioning, modification and revision. This program planning will provide a clear structure for the work involved in interleaving interferometer development and improvement with progressively more ambitious science running.
_________________________________________________________________________________________________________________________________
Prior Memo
LIGO Commissioning and Transition to Scientific Running (LIGO M010216-A-M)
During the past 18 months, LIGO installation has been essentially completed and commissioning of subsystems and complete interferometers has been initiated. We have achieved first lock in the Hanford 2k interferometer, operated the Livingston 4K in recombined mode, carried out several successful engineering runs, encountered and overcome several technical issues and at least one instance of force majeure. It was very encouraging to observe the reestablishment of lock in the fully recycled mode of the 2k following the extensive rework necessitated by the earthquake and the planned replacement of the suspension optical sensors. Early running of the Hanford 4k mode cleaner with the new sensors, digital controls and higher power is another very positive indication that our commissioning is progressing. Nevertheless, we have a great deal of work ahead in completing the commissioning of the three interferometers and initiating scientific observation.
The plan we established last year called for a coincidence run this year as a significant milestone, for the use of engineering run data to exercise the complete analysis pipeline from acquired data to a calculated upper limit, and for initiating committed scientific running in 2002 interleaved with progressive noise and availability studies. We defined this strategy recognizing that once the interferometers operated in stable configurations for significant blocks of time, scientific data collection and analysis and the detector characterization that is part of this process would be of critical importance to bringing the interferometers to full scientific utility. The strategy requires that interferometer commissioning should evolve from fully focused on technical issues to balanced data and detector studies as part of the smooth progression to science.
As part of this strategy, the Laboratory and the LSC have organized the data analysis working groups and the upper limits groups with the goal of initiating the scientific analysis team efforts. Proposed physics goals have served as foci for the activities of these groups. The groups have already carried out successful Mock Data Challenge activities. Our plan has been to initiate the prototype data analysis with the upper limits groups using engineering run two-site coincidence data this year.
This memo provides additional definition to this commissioning and early science program for the next year. We list below several goals and the major planning milestones around which we should structure the details of the program.
The major goals of this program are:
1. Bring the LIGO interferometers, and our data acquisition and analysis capability rapidly to the ability to collect and analyze coincidence data to address the scientific goals of LIGO. Advance both the interferometer commissioning and the commissioning of our analysis capability in a mutually balanced manner.
2. Emphasize the goal of initiating the scientific running in 2002, with the subsequent scientific running interleaved with instrumental sensitivity studies until the instrumental limits are achieved.
We define Engineering runs (E) as data taking runs that are primarily motivated by the Laboratory detector improvement program and not by scientific goals leading to publications. Science runs (S) are intended to produce a data product for the LSC with scientific goals and resulting in publications.
With this background, we define the following major program milestones:
1. The LIGO Laboratory will carry out the E7 run before the end of the year. We anticipate that the run will take place during December and will be scheduled for two full weeks. The run is an engineering run and will be the responsibility of the LIGO Laboratory. The precise schedule for the run will be defined by the commissioning directors and the observatory heads, Stan, Rai, Dennis, Fred and Mark. The Laboratory will operate at least one interferometer at each observatory with the earnest goal to operate all 3 instruments. At least one operating interferometer should be in full recycled configuration so that the interferometer data is representative of a complete configuration. The goal is to achieve significant individual periods of locked coincidence data and a total locked duration that is significant. Sensitivity of the interferometers will not be guaranteed except that it will be the result of our best effort. While the E7 run is the responsibility of the Laboratory, participation in all of the run activities is open to the LSC members in the LIGO I working groups. The data collected will be available to all of the upper limits groups. The analysis goal is to produce sensitivity curves and information on noise by the end of February and full execution of the analysis to demonstrate our ability to do an end-to-end analysis by the end of April. If the E7 data run is not successful in producing data capable of accomplishing the run period goals, the Laboratory will repeat the E7 run in January 2002.
2. The S1 run will be held in May 2002. The prime purpose for this run is to carry out the first scientific searches. This run will be the joint responsibility of the Laboratory and the LSC. The sensitivity goal is a two site coincidence with 3 interferometers running and the achieved scientific reach (volume searched x observation time in coincidence) should be an order of magnitude better than achieved in the E7 run. At least one interferometer at each site should be operated in the full recycled configuration.
3. Between the E7 run and the S1 run, the commissioning teams are to have priority with the interferometers. Any data provided will be collected for instrument characterization purposes.
4. In order to accomplish the goals of the E7 and S1 runs, decisions will be required on scheduling implementation of digital controls for suspensions, coil driver augmentation in LLO, tidal servos, common mode servos, etc. The details of of these implementations will be defined by Stan, Dennis, Rai, Fred, and Mark representing the commissioning and observatory management, consistent with the goals and milestones above.
5. The S2 run will be scheduled with at least an order of magnitude improvement in scientific reach (volume searched x observation time in coincidence) beyond S1 and following successful completion of analysis of the S1 data. We anticipate future interleaving of E and S runs with significantly improved scientific reach will continue until design sensitivity and reliability are obtained and an extended science run is motivated.
Barry Barish and Gary Sanders
no report
LIGO Operations--Administration
The list of current actions revised to reflect
the status of open actions assigned through March 7, 2001 and notes for
that meeting may be found at ACTION
LIST.
From: Ed Chargois <chargois_e@ligo.caltech.edu>
>From: Linda Turner - turner@ligo.caltech.edu>
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. . .
ACCOMPLISHMENTS
ACTIVITY
| 03/14/02 | Packages | Faxes |
| In | 51 | 40 |
| Out | 12 | 33 |
From: Esther Cunningham <esther@ligo.caltech.edu>
Press here for ACCOUNTS
PAYABLE HISTORY DATA .
From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
From: irena@ligo.caltech.edu (Irena Petrac)
CONSTRUCTION:
SUPPORT (Baldon, Torres, Lloyd, Tischler)
Progress Period from 03.08 to 03.14
Accomplishments:
The Annual Report for the Construction Project as of the end of November has been delivered to the NSF.
A quarterly report for the Construction effort is now due. I have started to assemble material for it. The goal is to prepare a short report comparable to the old Monthly Reports without molesting the task managers to heavily.
The following change requests have been submitted:
| CR-010012
Revision B |
WBS 1.4.4.1 | Closeout Construction Budgets for Initial Computer Equipment Complement at the Sites | P. Lindquist |
| CR-020002 | WBS 1.3 (OPS) | CDS Remote Control Room | D. Coyne |
Press for the latest Contingency Needs Projection.
From: Cindy Akutagawa <cindy@ligo.caltech.edu>
The first phase of the fire suppression system installation in room
215 Synchrotron was completed. Also included in this first phase
work was extension of the raised floor and addition of an equipment ramp
and safety rail. Along with the HVAC ducting modification, a problem
with the thermostat control system was also corrected (this was apparently
one of the reasons for the "rapid" heat up rate observed by LDAS folks).
General Items:
--------------
(F. Raab)
The walls are going up on the OSB East Building. Ground shaking from construction has decreased so that it is no longer preventing us from operating the 2K interferometer during the day. We did lose two days of operation due to high winds. We have rotated our operator coverage to 7am-2am to increase overlap with the scientists working on commissioning. We had attempted a 24-hr commissioing schedule, but we found that we had spread science resources too thin to be effective.
Detector: The mode cleaner in loop frequency noise was measured and compared to the in-loop frequency noise of the frequency stabilisation servo. The two agreed above about 10 kHz. We now have a calibration for the in-loop frequency noise of the mode cleaner which needs to be checked against the frequency noise seen by an interferometer arm.
Some work on ISCT4 to reduce to the beam size going through the electro-optic shutter required some realignment to be done on the table. (Joe Kovalik)
Re-zeroed all optical levers on large optics and re-calibrated dark levels on the QPD's for each OPLEV. Received a tutorial on ion pumps and other vacuum situations for operator 'need to know' situations. (G.Traylor, H.Overmier)
Optics and Installation: Had a training session with the installer of the Laser Safety Interlock system and am learning to administer the software. We're using the access system and no serious faults have been uncovered, although it appears I need to lengthen the access interval at some of thetables because the access times out before users open the table doors (unless they're quick). Completed the first of the core optic scatterometers and installed it at ETM-x. We're using a temporary channel for data collection. I'm working on three more for the cavity optics. (Jonathan Kern)
CDS: Epics Security databases and Security configuration files created.
Checking to see if we can incorporate it without booting the processors.
If not we will have to do a "boot-a-thon" sometime during Rolf's visit
next week. Upgraded the OS and setup an Ultra 10 for the gateway for Epics
security. The gateway is not still connected to the outside world. Working
with Dave on the VPN's (Virtual Private Network) for Virtual Control Rooms
to be setup at Caltech and MIT. Ordering a Sun Blade 100 for setting up
the test VPN network. Working on the Pentek test code . Found problems
with booting of the new pentium. Experiencing problems with GPS Clock
Driver for the pentek testing when booted
with a baja.(Chethan)
LDAS: Configured several more experimental replication scenarios. Finished
running burntest on beowulf nodes: no errors were found. Applied patches
to all LDAS Solaris machines. Although it looks like
some problems with SunFire still are not resolved. Discussing with
LDAS group how to set up a central metaserver. Learning LAL. (Igor)
General Computing: In addition to the preparations for the conference
here next week, we are in the process of upgrading our Solaris user file
server to a Blade 1000. We are taking advantage of this upgrade to
also upgrade many of our freeware software packages. The new server will
be faster than theold server, and will also have more disk space for the
user files. (Tom Evans)
|
Commissioning:
|
|
·FM
dewhitening filters successfully engaged after bug fixes to the LSC front
end code. No effect on the ifo noise level, though (as expected).
·Main
modulation frequency changed to match mode cleaner resonance. Over time
the MC length had drifted so that the modulation freq was about 700 Hz
off of the MC resonance. The modulation freq was changed by this amount,
and all demodulation phases were shifted by 11 deg to account for the change.
This eliminated some excess RF AM noise seen earlier on the beam reflected
from the recycling mirror (ifo misaligned).
·A
second LSC photodetector was added to the AS port, to be used initially
as a monitor. To free up a dedmodulation board to use with this detector,
the demod board that has been used to generate the 2*omega signal (SPOB)
is being replaced with a new '2*omega' demod board built for this purpose.
This changeover is almost complete.
·A measurement was made of the transfer function of frequency fluctuations of the input light to the anti-symmetric port. The behavior is not understood at all frequencies (see elog for details), but it is consistent with the earlier conclusion that we are not limited by frequency noise (at least below a couple of kHz).
Based on this
summer's LSC characterization effort, we modeled the 2km IFO during E7
with the goal to reproduce the AS_Q/ETMX Transfer Function. The result
of the model was then compared to the measurements. We obtained very good
agreement between the two without adding any artificial parameters,
just our knowledge of the system (an overall gain adjust was necessary).
The phase measurement has shown much more delay than expected (4-5 samples,
compared to the expected 1-2 samples).
The agreement
gave us confidence in the model, which is relatively simple, runs under
Matlab and assumes no cross-talk between the CARM, DARM, MICH and PRC paths
(e2e will have to deal with this issue). We are now able to restrict the
parameterizations for the 2k E7 fits: the initial values given to Minuit
(minimization function software for chi2 running under root) should be
set to
-
zeros at 100mHz, 100mHz
-
poles at 10Hz, 10Hz, 20Hz, 100Hz and 180Hz
Below 100Hz, the transfer function behaves according to -1/E, where E is the electronic TF (mainly some of the digital filters enabled in the LSC screen), whereas above 100Hz, the product -P*C dominates, where P is the pendulum TF (resonance at 0.74Hz, Q=10) and C is the cavity TF (p=180Hz).
·Locking.
After the successes reported last week, more difficulty has been experienced
in getting the instrument locked. There have since been some changes to
photodetector gain/attenuations, as both saturations and very low signal
levels were found to exist. Parameter tuning is still in progress.
·Noise studies. Continuing characterization of the electronics noise in the LSC & DSC systems, concentrating on the LVEA electronics. Some findings in the elog, but the full set of measurements, when finished, will be collected into a separate report.
The output matrix filters were tested when the 4k Xarm was locked. Stan lowered the MC crossover so the beam bouncing on ETMX was visible by the camera. We recorded the PSD pitch/yaw signals from the quadrant photodiode placed on the transmitted beam with/without the filters enabled. The enabling of the filters had a dramatic effect: the bouncing was not visible by the camera anymore. The pitch RMS (integrated down to 2Hz, as measured by the quadrant photodiode QPDX) decreased by ~10x.
·Noise
budget: It is now firmly established that the 1/f^3 noise seen in the differential
strain spectrum between approximately 60 to 300 Hz, when the common mode
servo is engaged, is not due to the electronics in the LSC chain. This
includes the photodetector preamps, the RF demodulators, the pre A/D conditioning
filters, A/D//D/A process, the post D/A filters, LOS controllers and the
coil drivers. We are encountering a new noise source. The noise experienced
previously due to ground loops between the de-whitening filters and the
LSC input has been eliminated by well balanced buffers and it has become
possible to use two stages of de-whitening and gain the reduction in noise.
The in-situ noise of the LOS controllers has been measured at the 150 to
200 nv/sqrt(Hz) level. This is currently not the limit to performance but
needs to be fixed otherwise the interferometer will not improve below 1000
times the scientific requirement at 100 Hz. The "dark" noise has also been
measured and is about a factor of 30 below the current performance. Measurements
of the noise in the common mode servo electronics are in progress.
·Wavefront
sensors: The beam size has been reduced at the electroptic shutter Pockels
cell which has resulted in reduced wavefront distortion. The antisymmetric
wavefront sensor (WFS1) is producing usable signals.
·Mode
cleaner and PSL: The careful measurements of the servo transferfunction
have been completed and are being incorporated in the model for the common
mode servo. The search for the source of the glitches in the frequency
noise of the laser, observed during E7, continues.
·Mirror scattering: The first of four telescopes to long term monitor the scattering by the test masses was installed. The data will be referenced to the transmitted cavity power and archived.
·A design review date has been set: Friday 4/12 from 9:00 am PT to ~11:30 am PT. Review documentation should be available by 3/26 (some sooner). The design requirements document may be released this week. A review committee has been selected.
·The schedule is being reviewed and revised; Will be released by 3/26.
·The spare fine actuator components (one chamber’s worth, with the exception of drive electronics for the PZT actuators) have been shipped from Hanford to Livingston. This is in anticipation of potentially installing them onto one ITM chamber to measure the effect of reduction in the stack resonances on an arm cavity (at 1.2 and 2.1 Hz with resonant gain stages). These spare mechanical components are from the prototype and have differences from the production system.
·Hytec has been requested to provide a quote to provide additional fine actuator systems for two chambers. Quote is expected next week.
·The
counter-wound coil springs are still acceptable for the HAM chambers. Stiffness
and strength testing on these prototypes will begin this week.
·The
bypass network has been received and will be integrated into the actuator
for testing.
·Testing
of new accumulator continues.
·SEI
Hydraulic Pump Station progress is good. Expect to complete initial assembly
in a couple of weeks. All parts have been received.
·wired
up LVDTs and Actuators on test HAM table:
5LVDT OK, one to be fixed tomorrow morning (one broken wire), all
actuators seem good.
·made
small tennis ball test and found about 3 Hz vertical main resonance and
2.2 horizontal (very very coarse).Tomorrow
will start frequency scans from actuator to LVDT in different combinations,
to see if we can identify better the modes.
·We
have 6 actuator drivers but we are short of LVDT drivers, hope that the
CDS shop can deliver soon.
·Virginio and Luca working on the D-space controller, if we get the electronics, and with luck we may have the full system cabled to the D-space before LSC meeting.
The kapton “paint”, which was to be used on the voice coil actuators being assembled in Italy was allowed to get too warm in transit unfortunately. As a consequence more kapton “paint” has been ordered.
The mechanical parts have been detailed in drawings and some are already into manufacture in Italy.
·Continued
work on new ASC code; last part to complete is addition of microseismic
feedforward. Hope to finish tomorrow for installation at LLO next week.
·Requests
are in for LSC code changes for addition of second AS photodiode; hope
to get to that on Friday.
·Lori
is working with Mohana on testing the new Pentek clock modules.
·New
EDCU is ready for test; Lori and Hongyu will work this next week.
·Alex is working on Framebuilder mods to read DMT trends; Frame5 code is complete, pending testing by LDAS.
·Dave Barker is continuing the EPICS security upgrade; should be completed in a few weeks. Plan is to upgrade LLO to the level at LHO next week.
·Alex has developed a method to tag CDS object code and filter parameter files so that one can more readily identify the machine configuration. This will be tested by Rolf soon.
Worked on analysis and writeup of a draft RFI/EMI mitigation plan, which is being reviewed by the team.
·EO
Shutter Controller (Sander Liu): Fabrication on three sets of circuit boards
is now complete. Some chassis parts were found to contain mistakes and
were returned to the manufacturer for correction. We are still waiting
for the high voltage power supplies ordered from UltraVolt to arrive.
·LEMO Antialiasing Filter Chassis (Sander Liu): Phoenix is putting together four circuit boards. They should be ready early next week.
·Common Mode Servo: Richard M. has gathered documentation on the required board changes (as implemented on the 2k interferometer) and the interconnection/cabling diagram for the 2k interferometer. Rich A. will implement the board changes for the LHO 4k and make appropriate record documentation changes.
·GPS Clock Driver Board: 4 of the new boards will be shipped to LLO on Monday to support installation of the combined WFS and Optical Lever system. An additional 18 boards are ready for test and 22 are in the build process.
·Fine
actuator control summing module: Plan to install at LLO next week.
Lightwave
have apparently decided not to repair NPRO #393 to the 1-W level and only
re-aligned the laser to deliver 500 mW.I
had hoped to find the contract specifications on this particular laser
in order to resolve this issue.
After trying the slow loop PID state notation code on another MIPS processor, the code ran fine for two days before I decided to pull the plug on it. Checking the files loaded by the PSL processor indicated an inconsistency with iocCore, one of the Sup files, seq, vxWorks and vxWorks.sym. Whilst the file sizes were the same, the checksum was different. The corresponding files were copied from LHO and loaded onto the PSL IOC. So far the slow loop code has been running for over 4 hours without any hangups, whereas previously it died after an hour.
I am continuing
to work on sideband imbalance, introducing the arms and I have been running
several different cases, changing one thing at a time with a view to identify
what is critical and what is not. For example for a symmetric interferometer
there is no sideband imbalance even when the arms are a little off from
the resonant condition for the carrier.
Now I am analyzing data for a one-dimensional case that is no higher order modes are involved. I am studying the impact of little adjustments in the length of the long arm cavities for a "not exactly anti-resonant condition" for the sidebands. The interferometer is a symmetric one since the offsets are the same in the two arms, However there is a significant sidebands imbalance.
Contamination
Cavity # 2.
Still locked
with 120 milliwatts of power into the cavity and 85milliwatts of transmitted
power.
several ring
down measurements performed as to be ~ 37us.each
time and the beat frequency ~ 74.4 MHZ each time at TEMoo.
We have also
taken beat frequency measurement at TEM01.Today
is our first official day of measurement of ring down and beat frequency
after we took care other factors that made the cavity more stable( we heat
the cavity a few Centigrades above ambient temp.) We will continue taking
measurements every day from now on for about two to three weeks.
Contamination
Cavity # 1
This cavity
has been upgraded and It was locked before but now it is in standby.We
will continue our efforts in this cavity as soon as we free up some equipment
needed. (filters and photodiode detectors has been ordered andwaiting
on)
GariLynn Billingsley
MCCM2K03,
MCCM4K04 and MCCM4K05 were received from REO and are being inspected.
SPETM01 has been measured after coating and shows streaks similar to those we have seen before on a few of the Core optics. The optic still meets the specification for rms roughness over the central 80 mm. Helena believes these may be removable and will attempt to clean the streaks.
On the analysis front, I used a recently upgraded version of the PSLmon glitch finder to generate trends of the glitch frequencies in the AS_Q channel as well as L1:PSL-ISS_ACTM_F. Since the glitches seen by Peter Saulson on this channel come in trains, I made two trend, one limiting glitches to 2 ms and the other limiting the glitches to 100ms. In effect, the first should show the individual glitches and the second should count the trains. In fact the results weren't as expected, and I haven't interpreted them yet. As Peter observed, the glitch rate varied greatly over the playground data sample. However, I did not see the significant correlation between the PSL glitches and the AS_Q glitches that Peter Observed. This may have to do with the definition of PSL glitches which may include many glitches of other types than those Peter was looking at.
no report
> LASTI (MacInnis, Mason, Mittleman, Ottaway, Rollins, Shoemaker, Zucker)
> =======================================================================
>
> VACUUM ENVELOPE:
We have vented for installation of accelerometers and geophones to
characterize the stack
and support structure. These items were prepared as Class B hardware
to
prevent
contamination of the vacuum envelope.
Wiring is foil-wrapped and passed through the
open 60" access door to preamp electronics in an outside rack.
> EXTERNAL PRE-ISOLATORS:
>
> In situ BSC stack characterization (Rich and Myron):
All equipment arrived from LLO and LHO. Some last-minute fit problems
with
the
shakers required
machining adapter plates and raising the softwall cleanroom by about
a foot
to get
clearance. These delayed installation a few days such that we
may miss our
"window"
to get the testing done and the BSC buttoned up before people drift
off to
LSC. As a
contingency we've resolved to disconnect our cabling and reinstall
the BSC
door temporarily
to minimize potential for atmospheric contamination (we are particularly
concerned as we lack
a positive-pressure clean air flow system).
> Pre-isolator mechanical design/fab (Ken):
A design of the pre-isolator with the electomagnetic actuator has been
created and presented at last week's SEI meeting.
The main Housing assembly drawings and many of the machined drawings
have been taken to three fabrication shops for quoting. Quotes
have
been promised by this friday.
An assembly drawing of the hydraulic HAM pre-isolator has been
completed.
> PSL (Ottaway and Rollins)
The bandwidth of the FSS board has been found to be limited to around
200
kHz, for oscillation free operation. This is a little strange as two
independent measurements of the open loop gain suggest that a bandwidth
of
a least 750 kHz should be possible. These measurements were made in
two
ways, the first involved closing the FSS loop with a unity gain frequency
of 150 kHz, then measuring the closed loop response from 50 kHz to
10 MHz.
The open loop response is derived from this. The next method involved
using the PMC cavity as an analyser to measure the frequency response
from
a signal injected into the test 1 input of the FSS board. This should
have
included all of the relevant frequency response of all the components
in
the loop except the Ref Cavity Pole and the linewidth of the PD, which
were later accounted for. It does not appear as if any of the electrical
components are saturating.
In addition to this it was noticed that the phase and amplitude response
of the fast correcting pockel cell were changing with time, this was
causing stability problems with the cross over between the PZT path
and the EOM path. This was significantly improved by realigning the
pockel
cell in the laser path to minimize the RFAM generated at the modulation
frequency (35.5 MHz) used to lock the PMC. However, it has been noticed
that residual instability still remains as the FSS Phase Correcting
path
periodically becomes dominated by a 10kHz signal that slowly comes
and
goes (timescale of order minutes).
Simulation and Modeling (Bhawal)
We are organising e2e school at LLO conference room on 18th March,
Monday
from 9am to 6pm. All are welcome.
Details are available in e2e homepage http://www.ligo.caltech.edu/~e2e
LIGO I simulation system
--------------------------
(Matt) System testing continues, along with infrastructure (modeler,
alfi) debugging.
I have also spent time writing my presentation for the upcoming E2E
school.
LIGO: cold to hot
-------------------
(Biplab) found a bug in calculation that includes phase-offsets in
recycling cavity.
Working on that. Continuing other studies related to transition from
cold to hot state.
Code development and maintenance
---------------------------------
(Hiro) Working with Matt to make modeler compatible with the new LIGO
simulation package.
Alfi
----
(Bruce)
- Fixed cyclic include bug.
- Implemented Save Copy As.
- Fixed misc PRs.
(Ed Maros)
Worked on enhancements to build procedure to support JAlfi
(Melody)
- Finished the node settings window for alfi5.
Also worked on fixing multiple selection and menu problems.
LIGO Data Analysis System
Software Systems (Blackburn)
We were not in a position to have the code freeze this Wednesday due
to a dozen open
issues in code development that are critical to this release. The code
freeze has been
rescheduled for the Monday after the LSC meeting.
Began announcing to the LIGO/LSC LDAS users community the discontinuation
of the
shared user accounts within LDAS. This will start with the removal
of the ldas_mdc user
this week. The LSUG group was also informed of the intend to re-issue
ALL LDAS user
account for the Science 1 run.
Research and prototyping associated with the implementation of a "federated"
database
for all LDAS systems continues. This effort is being headed up at LLO
with guidance
from CIT.
The GCC 2.95.3 standard library string class has been modified to make
it thread safe
after it was identified as the primary case of the dataConditionAPI
failure. No test results
are available at this time detailing any improvements from this change.
cntlmonAPI and utilities: fixed db2utils and logscan email that were
broken by changes
in gpstime; fixed deletion of core files to also delete .file files
metadataAPI: verify core dump is caused by metadataAPI not catching
a database open
error exception when the database server is down when requesting uniqueIds;
added code
to do a reconnect if the database server is down and then restarted;
fixed metadata pipeline
macros to call reattach directly to prevent loss of return value in
macro; verified changes
worked by pipeline testing.
cmonClient updates: changed insertion page to allow choice of insertion,
queries or both;
clean up graph data handling; patched emu-graph to be closer to x-axis
values (gpstime) to
reduce excessive looping when figuring out x-axis tick marks; minor
enhancements to
tapecontrol per Greg's request; minor fixes to various pages to keep
up with ldas changes.
Wrote a sample client and server for Ed and Isaac to explain how the
Tcl channel version
of the data socket is supposed to work from Tcl which has now been
implemented in the
generic LDAS library.
Development continues in the dataConditionAPI on the new shift/interpolation algorithm.
Discovered that dataConditionAPI ingestion routines for objects originating
from the
metaDataAPI were no longer returning the names of ingested objects
(PR #1367).
Frame API was packing detector structures in outgoing ilwd objects in
individual containers. I
changed this so they are packed into a single container as required
by datacond et al for
treatment as a linked list on ingestion.
Added new state variable ::${jobid}::usrcmdinput and function mgr::getCommandInputText
for
use by the controlMonitorAPI. This procedure returns the user command
name and arguments
on a per-job basis.
Phil Ehrens, Philip Charlton and Isaac Salzman met to discuss the syntactical
details and
interoperability of frameAPI resampling and the selection of data subsets
by time offset and
duration.
Implemented simultaneous concatenation and resampling of data in the
frameAPI. Formerly concatenated
objects which were time-sliced from multi-second frames were packaged
as multiple objects, they are
now properly concatenated into single logical objects when contiguous
segments of data exist. Time
metadata are correctly updated to reflect the data in the output.
Implemented correct time slicing of ADC data for the getFrameElements
user command. Formerly the
time metadata added to the ADC on conversion to ilwd was taken directly
from the parent frame header.
Values are now updated to reflect the actual data array in the ilwd.
Worked to make certain that all frameAPI data sent to the dataConditionAPI
were packed into containers
at the proper nesting level and with the appropriate metadata for correct
ingestion.
Wrote generic API function for use by the controlMonitiorAPI which allows
discrimination between user
keyed GPS and UTC time values.
The eventmonAPI now has support for the added detector geometry structure.
It has passed unit tests.
The support for Frdetector geometry in the dataConditionAPI is still
under development.
frameAPI/framecpp: Fixed problem report #1383 concerning corrupted data
when doing ilwd->frame->ilwd
conversion of FrProc complex data. Removed exception specifications
from frameCPP. Removing exception
specifications from frameAPI.
A 'watch jobs' utility has been added to 'guild' to display a list of
jobs currently running on an LDAS
system. At present, it only works with the ldas-dev system; it will
work with the other LDAS systems
after the next release of LDAS. Also, 'guild' has been modified to
allow users to limit queries to a
particular "user tag", which is a short string that users can specify
in association with one or more
LDAS jobs to facilitate retrieving the results.
Investigated the behavior of LDAS when reading frame files produced
on little-endian machines. Both
frame i/o libraries (FrameLib and frameCPP) have been found to have
internal inconsistencies which
are being addressed.
Hardware Systems (Anderson)
Caltech
-------
(Dan Kozak)
* Continued HPSS migration from Redwoods to 9940s.
* Installed OpenSSH-3.1p1 on ACSLS (the Ultra 10 in Booth). This
took
more time than one would expect since the machine has no compiler
and
the method of just copying the binaries from lugh which worked
for the
old ssh didn't. Finally ended up installing pkg-get (from
bolthole.com) and using it to get the binaries in pkg format
from
sunfreeware.com.
* Dealt with another crash of the Ultra 10 managing the STK silo. Haven't
heard back from Sun. Talked to STK field service (Gary
Vogt) and told
him to try to get in touch with them and/or have them call me.
Told him
that the next failure will cause us to swap the machine out
for one of
our own.
* LIGO SAN testbed:
* Installed SAM-QFS. Working on configuration, specifically
getting
SAM-QFS to reliable talk to ACSLS (the silo software).
(Al Wilson)
* Cleaned out 215 Synchrotron.
* Brought the ldas-test system back online.
* Installing Big brother on ldas-test.
* Tested the backup of the new systems.
* Testing the temperature script on dataserver@LHO.
* Added new version of openssh to linux boxes at Caltech
also upgraded glibc to 2.2.4-19.
* Installed motif and nedit rpm's on linux server boxes.
(Stuart Anderson)
* Brought the initial ldas-cit system on-line. It currently consists
of 3 Sun servers with no Beowulf cluster yet.
* Upgraded the ldas-test and ldas-cit Solaris servers to the latest
Solaris 8 Maintenance Updated (sol8mu7).
MIT
---
(Keith Bayer)
* Loaded a specific set of playground data on metaserver (~80GB).
Livingston
----------
(Igor Yakushin)
* Configured several more experimental replication scenarios.
* Finished running burntest on beowulf nodes: no errors were found.
* Applied patches to all LDAS Solaris machines. Although it looks like
some problems with SunFire still are not resolved.
* Discussing with LDAS group how to set up a central metaserver.
* Learning LAL.
(Shannon Roddy)
* Ordered and received the 12-pair 50um Fiber to run to the new building
for LDAS.
Hanford
-------
(Greg Mendell)
* Richard McCarthy has obtained three quotes for the purchase of 12
pair
of armored 50 micron multimode fiber to go between the current
office
building and the new building. Once the distance through the
conduit
between the buildings has been accurately measured (which should
happen
within the next few weeks) he will purchase the fiber.
Data Analysis Activities (Lazzarini)
Mendell:
1) David Chin reported that corrected clean locks were posted at
http://tenaya.physics.lsa.umich.edu/~dwchin/LIGO/E7/LockSegments/.
Thus, loop scripts that use the knownpulsardemod DSO were restarted
Wed
Mar 13 to generate SFTs at LHO and LLO. The scripts use the clean locks
to generate quality channel information. They will run for the next
7 to
8 days.
2) Resolved issues with AEI group as to how to center frequency bins
on
exact targets with known pulsar frequencies.
3) Test input of SFTs using the frameAPI. This works, but more testing
is needed to get SFT data into the dataconAPI and into the
knownpulsardemod DSO.
Charlton:
This week I spent some time (~ 1 day) prototyping a Matlab script for
looking at components of the 2-phase fn FCT lying in the tau0-tau1
plane.
The implementation is not efficient because of the difficulty of working
with 3D matrices in Matlab, so it is best suited to looking at
small "postage stamps" of the FCT.
The function has the form:
H = fct2d(h, M, L, k)
where
h is the input data
M is the number of bins in the tau0 direction
L is the number of bins in the tau1 direction
k is a particular time index
that is, the function returns the plane
H_{k k0 k1} where k is fixed, 0 <= k0 < M, 0 <= k1 < L
A simple extension would be to allow subranges of indices in M and L
to be
specified.
Shawhan:
Worked on improving the procedure for calculating the response function
of an interferometer from the raw calibration information. In
particular, figured out how to fit simultaneously to amplitude and
phase data using MINUIT in ROOT.
Verified that the E7 RDS (reduced data set) files being produced by
Isabel Leonor are constructed correctly, complete, and can be read
by LDAS.
Overhauled the LIGOtools "Frequently Asked Questions" web page (at
http://www.ldas-sw.ligo.caltech.edu/ligotools/faq), adding several
new write-ups.
Helped a number of LSC members solve problems retrieving LIGO data with
the getFrames utilty.
General Computing (Wallace)
MIT:
(Keith)
-Installed a new mail-server.
Livingston:
(Tom)
-In addition to the preparations for the conference here
next week, we are in the process of upgrading our Solaris
user file server to a Blade 1000.
-We are taking advantage of this upgrade to also upgrade many of our
freeware
software packages. The new server will be faster than the
old server, and will also have more disk space for the
user files.
(Shannon)
-Spent this week preparing for the LSC. Working on fiber connections
between the two buildings for the network. Set up three wireless
access
points, 24 port 10/100 switch, projectors, laptops, etc for the
conference. Working on finding a hard to find fiber patch cord
to hook
up the network.
Hanford:
(Christine)
- Continued setting up new Win2k PCs and transferring user files from
old
PCs to new ones. I have completed 5 out of 9.
- Helped users learn how to use Win2k OS.
- Helped CDS with purchasing equipment and supplies.
- Various other support to users and assistance to CDS.
CIT:
(Veronica)
- LIGO website: posted various updates on Conferences/Talks, Elba GWADW,
Internal Bulletin Board, LSC, MOUs. Tracked down and corrected a misplaced
entry in the LSC talks database.
- Attended a teleconference with Gary and Sydney on the Elba GWADW
preparation.
- CaJAGWR: Videotaped the talk by Braginsky and Khalili. Compressed
and
posted video and viewgraphs of Braginsky's/Khalili's talk.
- Reading through CVS and GNATS manuals.
- Working on the webpages for the Advanced Suspension group.
(Mike)
-Loaded all GC software and configured security plus added critical
updates
and service pack 2 for two laptops, these were 2000 professional loads.
-Loaded a webserver on top of NT4.0 but this box has problems loading
active
server pages, which looks like a hardware problem.
-Having a big problem with a NT4.0 server that is running IIS 4.0 it
keeps shutting its services off. I have installed multiple hotfixes
and
plus installed the IIS lock down tool but now this server seems to
be
having some hardware issues. We have a replacement server ordered and
will
be transferring this website over to the new server.
-Working with Larry testing the new edge switches WBridge, 40Meter,
Wilson
House, to get ready for Friday nights switch over.
-Started working security issues on NTSRV's that have been suggested
by (ITS) this is real time consuming, but hope to have this finished
up by
next week.
(Lisa)
- Janeen's external hard disk lost its disk descriptors. Larry
and I tried to
re-label it but to no avail. We have a good backup of Janeen's
work, so this is
inconvenient, not tragic.
- Installed orinoco access points to cover the 2d floor of bridge and
the other
half of the third floor. Reprogrammed all of the access points
with a new mac
address list. This list has been sent out to sysadmin so they
have a copy at
the sites.
- Working with Bryan Johnson of Audits on our security audit.
- Built a new Blade 100 for Gary.
- Will be working on the luna swap out for Wilson House the new unit
has
arrived.
(Larry)
-The testing for the new GigE backbone has gone well. The board in
slot 2 of the
core switch was bad. Replacing it also allowed us to tighten down the
fans and
other items that loosened up during the burn-in of the unit. It now
appears that
everything is a go for the swap out this Friday night.
-Worked on a couple of the servers. Mainly debugging and upgrading
some of the
patches. Still have a couple of major rebuilds to perform on the new
servers.
-Ordered server replacements for a number of the PC web servers.
-Working on a number of orders with SUN for the matching grant program.
-Assisting Mike and Veronica on various web issues. Mike will be handling
more
of the hardware and OS issues concerning the PC Web servers and Veronica
will
continue with the WEB pages and associated applications. There are
plans for
cross training but time limitations may keep that to a minimum. Also,
spent time
working with Mike on getting the CIT Oracle install for the PC's to
work
correctly.
-Went over some of the security audit issues. So far nothing major.
Most of the
procedure items are covered in the GC computer usage policy document
that is now
going through the system.
-Worked on some of the logistical issues for the LSC.
-Sent a copy of the JPL drawing formats to Janeen.
-Worked a number of hardware issues with Mike and Lisa. Most were PC
related.
From: Mark Barton <mbarton@ligo.caltech.edu>
This week I did a lot of work with my toy model to
check that the
calculation of thermal noise. Most of the thermal noise in the pendulum
mode is expected to come from the elasticity and associated dissipation
due to bending of the wires near the flexure points. The rather complex
code for the wire bending was indeed predicting the right amount of
noise
relative to analytic expressions, which is pleasing given that it hadn't
received any validation in the comparisons to the Matlab model. However
the older code for the plain longitudinal stretch of the wires was
producing spurious damping, and it turns out that "dissipation dilution"
hadn't been correctly allowed for. The reason it was going wrong was
rather interesting. First it turns out that when you analyze a pendulum
system in a normal mode approximation, the physical reality that the
mass
actually moves in a curved path and has gravitational restoring force
is
implicitly replaced by a fiction that the mass travels in a purely
horizontal path and has elastic restoring force. The mode shapes and
frequencies come out the same because there's a numerical coincidence
between the restoring forces in the two scenarios. However this makes
it
look as if energy is being cycled into and out of the lossy elastic
potential when in fact it's not. Second, the loss that a spring with
damping factor phi contributes depends on how it's deployed. If you
put
it under a large static tension and then use the sideways restoring
force
(as in a violin mode), the effective phi is amplitude dependent and
goes
to zero in the limit of small amplitudes. In fact it's this tension
effect, not the losslessness of gravity that causes the low-loss nature
of a pendulum. Gravity merely acts as a tensioning agent, and could
be
replaced by a lossy spring attached between the mass and some point
a
distance below without spoiling the Q significantly. The fix was to
do
the first stage of the calculation a second time with the tension in
the
wires relaxed, and to use as a base for the damping calculation only
the
components of elasticity that persist with no tension (e.g., the vertical
elasticity of the bounce mode).
From Benno Willke:
LSC LASERS WORKING GROUP
Minutes of telecon Mar 14 th 2002:
participants: morning call: Peter Veitch, Jesper Munch, Maik Frede,
Martina Brendel, Ralf
Wilhelm, Rene, Ivo Zawischa, Michèle Heurs, Benno Willke
evening call: Rick Savage, Peter King, UF, David Shoemaker, Benno Willke
1) updates
Adelaide:
- laser head is together, problems with mounting the waveguide that
allows spreading out
the pump distribution in the horizontal direction while keeping it
confined in the vertical
- expect laser experiment to start next week
LZH
- modelling continued, surprising result that thermal aberations are
similar for transversal
and longitudinal pumped rods
- laser diodes for 1200W of pump power delivered, 4 from 30 diodes
had to be returned
due to different problems (short between heat sink and TEC, Thermistor
failure)
- started experiments with longitudinal pumped water cooled rod (300W),
due to so far
unclear reasons the undoped end cap of a rod cracked at a position
where modelling
shows only a small temperature rise and no big temp. gradient
Stanford
- Lightwave amplifier was delivered
- due to missing cabling no experiments started so far, cabling is
now in hand
- plan to set up chain LIGO-laser(8W) – Lightwave-amplifier (20W) –
Stanford Amplifier1
(35W) – Stanford AmplifierII (100W) within the next month
2) schedule
- Benno announced that the high-power-stage decision milestone was
delayed until end of
July, Adelaide and Stanford said that this timescale is reasonable
to get a 100W level laser
up and running and characterize it
From: ctorrie <ctorrie@ligo.caltech.edu>
Suspensions
Janeen and I held a tele-con with various people from Caltech, Stanford
and Glasgow to discuss the requirements for the design of the OSEM for
LASTI.
We held a follow up tele-con with Russell and Alastair in Glasgow re.
update on the new hybrid coil former/head for LASTi. (Blade wire clamps
and assemblies were also discussed
I have been working on 2 sets of assembly drawings in SOLIDWORKS for the parts that went to the Physics machine shop and have been continuing with my weekly machine shop class.
I have been discussing with Norna plans for both my visit to GEO in May and her visit to Caltech in April.
Janeen and I have sent drawings for the new coil former/head and surface mounts for the LED and PD's to Glasgow. These are to be put with drawings of the coil clamp etc.. and a quote will be obtained for the new prototype this week.
Janeen and I have been discussing in more detail the planned experiments
and there breakdown for planned 3 SURF students.
From: Helena Armandula <ahelena@ligo.caltech.edu>
Advanced LIGO SUS
Silicate bonding
Had prints made and obtained quotes to manufacture parts to be silicate
bonded and be used to investigate bond creep.
Heavy glass
The polisher manufacturing samples made from VIOX # 24935 is having
a hard time with the substrates' surface staining. They will be delivering
the polished substrates next week.
From: Janeen Hazel Romie <romie_j@ligo.caltech.edu>
AdLIGO Suspensions
Still waiting on two vendors' quotes for maraging steel for suspension
prototypes. Two have already come in.
Working with GEO people on controls prototype sensor/actuators along
with a possible laser diode approach to noise prototype sensors. The controls
prototypes shall have sensor/actuators that use the magnets and coils of
the GEO design and the LED and photodiode boards of the LIGO I osems.
Working with Calum on integrating the MC design components.
AdLIGO PSL
Peter King
The amplifier, refrigerated bath and laser power supply arrived at
Stanford. After a few questions about how to connect the various
cables and the optical setup, the amplifier is in the process of being
aligned with the output of the 10-W laser.
For additional information about this report, contact sanders@ligo.caltech.edu