|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday May 21, 2001 will be:
(Meeting time: 10:30 am Pacific Time)
Open meeting 10:30 - 11:30
Special Items:
The new LIGO Newsletter can be found
at:
http://www.ligo.caltech.edu/LIGO_web/0104news/0104main.html
Report on the Engineering Run E4
no report
WBS 1.2 LIGO Operations--Administration (two
weeks)
There was no site teleconference held on May 17, 2001. A site teleconference was held on Thursday, May 10, 2001. Discussion included:
From: Ed Chargois <chargois_e@ligo.caltech.edu>
Ed is currently in Hanford conducting a property audit. This status report is from last week.
>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
| Contractual/Procurement |
26
|
| Drawings |
150
|
| Engineering Documentation |
15
|
| Graphics/Presentations |
26
|
| General Correspondence |
5
|
| Management/Policy |
12
|
| Publications/Formal Notes |
3
|
| Technical Notes, Informal |
9
|
|
|
161
|
| Packages | Faxes | |
| In | 34 | 29 |
| Out | 13 | 38 |
Press here to access the DOCUMENT CONTROL CENTER WEB PAGE.
From: Esther Cunningham <esther@ligo.caltech.edu>
Press here for ACCOUNTS PAYABLE HISTORY DATA.
From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
Last Week
This Week
Last Week
This Week
From: irena@ligo.caltech.edu (Irena Petrac)
Last Week
Support (Wood)
Progress Period from 05.04 to 05.10
Accomplishments:
Accomplishments:
WBS 1.4.1.2 Project Controls (LIGO Construction)
We have established a partial schedule for the review of MRE cost estimates for selected systems over the next two months. Reviewed Core Optics Components Monday, May 7, 2001. Reviewed PSL Thursday, May 17, 2001.
Still working with David Shoemaker to set up assignments for Advanced R&D Annual Report. We should be submitting this report (necessary to get FY 2001 Advanced R&D funding) over the next month or so.
A LIGO Change Control Board was
held on Tuesday, May 8, 2001. The following Change Requests were
approved:
| CR-000018 | WBS 1.1.4 | Curbing for Service Roads at Livingston | G. Stapfer |
| CR-000019 | WBS 1.2 | Additional Lab Equipment | D. Coyne |
| CR-000020 | WBS 1.1.4 | Staging Building and Renovations to Existing Building--Livigston | F. Asiri |
| CR-010001 | WBS 1.1.4 | Return of Unused Construction Budget to Contingency for Civil Construction | F. Asiri |
| CR-010002 | WBS 1.2.1.1.6-7 | Return of Unused Construction Budget to Contingency for ISC Design | S. Whitcomb |
| CR-010003 | WBS 1.2.1.1.9 | Return of Unused Construction Budget to Contingency for Detector Systems Engineering | S. Whitcomb |
| CR-010004 | WBS 1.2.3.1 | Return of Unused Construction Budget to Contingency for Physics Environment Monitoring | S. Whitcpomb |
I have distributed minutes (LIGO-M010087-00-P).
Note also that we have scheduled a Technical Review Board (TRB) on May 18th at 10:00am PDT to discuss design and placement aspects of the OSB II building at Hanford. Otto is distributing materials and will contact key people to ascertain where materials should be sent and telephone connections established.
Press for the latest Contingency Needs Projection.
From: Kris Duncan <kris@ligo.caltech.edu>
From: Ed Jasnow <jasnow@ligo.caltech.edu>
General Items:
--------------
(F. Raab)
Installation/rework is still ongoing (see below). Some frequency transients
have been observed on the 4K PSL and work is underway to cure the problem.
The DAQ software has been upgraded.
On 5/21 I will get a briefing on the proposal for the Maiden Wind Farm
atop
Rattlesnake Mountain. On May 12 (Saturday), LIGO headlined the Astronomy
Open
House at the University of Washington. (What a difference a decade
makes!) On
5/22 LIGO will sponsor "A Night of Science" at Gladstone High School,
where
students will give presentations to the community on their LIGO-related
research
projects.
Installation & Rework:
----------------------
(D. Cook)
The Y-arm spool has been
reinstalled. We realigned the ITMy-2k yesterday
morning as it didn't recovery after having the controller disconnected
and
reconnected. After re-aligning the ITM the OSEM 50% open light voltage
settings came back very close to the original alignment values(~2mV)
which
verified the previous alignment. I followed up by switching out the
total
station for the COS laser AC and projected the reticle pattern
through the
to the Y-arm PO view port to verify the alignment. We closed out HAM9(RM),
BSC7(ITMx,FMx), BSC4(BS) and BSC8(ITMY&FMy) today. All beam dumps
are in
place and exit checks completed. The doors were placed back on HAMs7,8,9
and BSC 8.
The IO alignment and mode
cleaner alignment needs further work so HAM7&8
doors were hung temporarily for the power outage on Friday and will
be
reopened on Monday to continue the alignment. This could
pose a week
delay in pumping the corner station and delay some of the needed equipment
at LLO.
We will stage for the ETMy-4k
optic and ETM telescope installation on
Monday in hopes to complete the 4k installation. Optical lever installation
is on going for both IFOs.
2k-Input Output Optics work
-----------------------------------
(Corey Gray, Betsy Weaver)
This week's goal was to close off the IOO portion of re-work/earthquake
repair on the 2k before "lunchtime" on 5/17 so that everything could
be
sealed up for the 5/18 power outage.
First off, on the 2k PSL table, we installed our "checked-out"
Brewster-angle Polarizer. The periscope PZT Mirrors were also
installed
(many thanks to Nergis!). This was done w/o altering the "pointing"
of our
input beam into the vacuum system.
Work was then focused on "close-out" work for the IOO. We clamped
all of
the HAM7 optics and performed tasks such as CO2-cleaning the optics,
NO2-deionizing of the optics, removal of targets, installation of a
baffle,
and setting of the safety stops. Unfortunately, when all the
optics were
free-swinging and the alignment was checked--it was seen to be noticeably
off. So much so, the pitch/yaw bias values were several factors
different
on some optics! This will be revisited next week when the system
is opened
up again.
E4 operation at LLO was the focus of activity during the early part
of the week. We locked the interferometer in a recombined, but not recycled
state and attempted to run in this configuration from Friday at Noon until
Monday at noon. It has typical experience here that we have not been able
to lock the interferometer during the day because of greater seismic activity
and so have done much of our commissioning activity at night. The E4 run
provided the opportunity to assess whether construction activity on site
was the predominant source of this seismic noise since there was no construction
activity on site during the weekend. We found that we were unable to lock
during the 8-5 pm portion of the day on Saturday either, but on Sunday
could lock the entire day. The ability to lock seems to correlate very
well with fluctuations in the seismic background in the 1-3 Hz region.
We know that there are some problems with the small optics suspensions
which enhance the coupling of these optics to ground motion. For example
MC2 has very low Q values (<1K in one case) for some of its modes, and
the ring down is amplitude dependent, so this cause of this behavior needs
to diagnosed and mitigated as part of the vent we are planning next week.
Prior to venting, we will attempt to tilt the HAM4 optical table using
elements of the course actuator system to align the output Faraday isolator
to eliminate the obvious vignetting.
On Thursday, we held a meeting with Barry Barish, representatives of
the NSF, and various representatives of La state education, such as some
local teachers and school board superintendents, the Vice Chancellors of
LSU and Southern Unversity, and the Vice President of LSU. Also present
were the State EPSCoR officer, a member of the board of regents, and program
officers from the NSF-funded state systemic initiative and Mississippi
Delta rural systemic initiative. There seems to be broad based interest
in establishing an educational outreach center in association with LLO
and collaboratively utilizing it in partnership with local educators.
| Installation&
Commissioning:
Livingston |
Other Science/EngineeringActivities:
Issues/Concerns |
See also the Installation web page
Alignment of the IO optics to the interferometer is underway, with the goal of completing it before the power shutdown Friday.
Chamber exit checks have been completed for the 2 km BSCs and HAM9. The Manifold spools have been replaced and doors are being reinstalled.
We had more than 16 spirited visitor on site during the run, most of them giving shifts. Interested scientist were (nearly all the time) present in the control room besides the shift takers, pursuing their projects, detector characterization investigations and/or making sure that the detector operates in its best state. Some of them prepared preliminary reports by the end of the run. Most of these documents are accessible through the E4 web page ( http://blue.ligo-wa.caltech.edu/engrun/E4/ ), while the others will become accessible soon (http://blue.ligo-wa.caltech.edu/engrun/E4/Results ). There were numerous DMT monitors running and producing valuable information during the full time of the run, mostly connected to detector diagnostic studies and investigations. Their result will be posted at the E4 page as soon as they become available.
We recorded ~400Gb of Reduced Data Set frames with only a couple of lost frames from the Friday evening locks. LDAS did not report any data loss. The correlation group created a smaller reduced dataset, which includes the most interesting environmental channels from both observatories.
I would like to thank everybody participating in the run for their hard and good work which made this effort successful and worthwhile!
Footnote: We observed significant changes (http://blue.ligo-wa.caltech.edu/engrun/E4/Results/SeismicActivity/ ) in the LLO seismic background in the region around 1Hz, consistently starting around 6AM and lasting until about 5-6PM. This pattern is true for all weekdays and somewhat surprisingly for Saturdays, while Sunday is usually fairly quiet. We had significant problems obtaining lock during these noisy periods. The origin of this noise is not clear yet, but wide range of possible sources can be blamed and need to be investigated (ranging from on site construction to highway noise). It is very likely that the planned improvements for the upcoming vent and more experience with the interferometer will greatly improve this problem. However, until we are absolutely sure that the improvement will be adequate to achieve the desired 95%+ duty cycle, it might be advantageous to track down whether we deal with local noise and/or outside noise. It is very likely that we can address the noise source if it is under our control, but it is unlikely that we can deal quickly and easily with outside sources.
The losses in the arm cavities were estimated.
The apparent losses measured at the antisymmetric port are dominated by
the beam being cutoff by the Faraday isolator. The losses measured at the
symmetric port, where the beam is smaller, give upper limits of the loss
per mirror of 70ppm.
The arm lengths of the cavities were measured in terms of the RF modulation wavelength. The measurements give equal arm lengths with an uncertainty of 30 cm. There is a diagreement of 2 meters between the absolute lengths measured and those tabulated from mensuration during installation. We are investigating this discrepancy.
GariLynn Billingsley
Four LIGO 1 End Test Masses shipped to REO for coating. These
had been well characterized in their uncoated state, we hope to gain insight
into effects of the coating process when we measure them after they are
returned from REO.
Though the Gaussian beam for driving the interferometer is chosen to have the best overlap with the "reshaped field" that should be excited, the gains are not the expected ones. More specifically the coupling between the recycling cavity and the arm cavity is right, but I am looking for the cause of relevant additional losses which reduce the stored power, in comparison of what should be using spherical mirrors.
A major cause was identified and this gave a performance of 75%, while I was stuck at values around 30% of the nominal gains.
I will do more investigation to understand what choice in the design is needed to get things better.
Still haven't been told about NPRO #259 nor 10-W laser #110.
I troubleshooted a few problems previously reported with the 80 MHz VCO boards. The major problem that I saw was that some polarized de-coupling capacitors were incorrectly soldered in. I made the required alterations and the boards seemed to perform as expected.
The super-polished substrates ordered from Mindrum Precision have slipped delivery, apparently because of more urgent jobs from within the project. They should be shipped out by the end of this month.
Lee Cardenas
Rick Karwoski
Intensity stabilization: This week was spent integrating temporary
circuitry into the servo electronics to simulate the mode cleaner.
The system is currently setup at Lauritsen undergoing some initial testing.
The next step will be to close the loop including the MC.
Ben Abbott
The intensity servo DC photodiode board is currently complete.
It has been fully tested, and retrofitted with potentiometers that enable
my signal to have 0 DC offset going into Rick's box.
Coming to the end of the testing on the 4k LSC boards. We have still to test one photo-diode that was not functioning properly, two whitening filter boards that are currently under test. We plan on shipping everything except the photo-diodes to LHO on or before the 18th of May (Friday) such that it is there for Rich's visit to LHO next week.
I first made a quick test for interference between processes on the Linux platform. To do this I ran 11 monitors (all the same) simultaneously on the Linux node. This used about 80% of both CPUs. There was no obvious interference between the processes in that the CPU usage scaled approximately linearly with the number of processes.
I then made a very preliminary benchmark of the relative processing power of delaronde (Sun E-450, 4x400MHz) and alvar (Linux 2x1000MHz) by running two monitors (BitTest and PSLmon) with the same configuration files simultaneously on the two platforms. The two monitors were chosen because they make different demands on the CPU. BitTest loops over many (274) channels and performs simple processing (calculate the average and rms, look for stuck bits) on the data. PSLmon, on the other hand , performs an FFT on fewer (29) channels and calculates the power in several frequency bands of each channel. The raw results are as follows:
BitTest PSLmon
delaronde16.3% 19.8%
alvar 14.0% 16.1%
-------------------------------------
ratio 1.16 1.23
The numbers here are taken from the "top" utility and scaled as
necessary to give the fraction of a single CPU used by the monitor process.
>From this you can see that the speed of one alvar CPU is about 1.2 times
the speed of one delaronde CPU. This is considerably worse than what one
would expect naively from the ratio of clock frequencies (1000/400 = 2.5).
This may be due, in part, to the fact that the Linux system has to swap
many bytes when reading the input frames or to a significant difference
in the way the top utilities assign system time. A large systematic difference
is suggested by measurements of the data pusher process which takes much
more time on the Linux system (7.1%) than on the Solaris system (0.3%)!
Even if the uncertainties in the comparison (which tend to make the Linux system seem slower) are ignored, the Linux system would still win out in terms of $/MFLOP.
One should remember that this is a very rough measurement of the relative
performance of the two machines. The fact that the Linux system must do
a great amount of extra work to reorder data bytes obscures its performance
for very complicated monitors. It would be very interesting to find out
where the cycles are being used by the Linux machine. It is likely
that a little optimization of the byte swapping algorithm and a better
understanding of the timing calculations will lead to a measurement that
is closer to the ratio of clock frequencies.
This week we continued working on lock acquisition in our arm cavities.
After trying various servo configurations without success, we decided
that
a more detailed understanding of the physics of the system might be
in
order. To that end we have been both modeling the system mathematically
and measuring some of the more prominent mechanical resonances in a
test
setup outside the vacuum chamber. The modeling, basic though
it is, and
not yet complete, has yielded some good insights into the mechanics
of
control theory. Many thanks to Rick
Karwoski for helpful discussions on this! For measuring mechanical
resonances, we have started by suspending a dummy test mass (of the
same
dimensions, mass, and moments of inertia as our test masses) with the
same
kind of wire used for our test mass suspensions. With this setup
we
were able to easily measure the fundamental violin resonances at ~500
Hz.
We also got some good advice from Stan Whitcomb, who helped us eliminate
one source of noise on the test cavity error signals: a 2.42MHz signal
that
comes from beating between our mode cleaner sidebands and those used
for
our test cavities. Many thanks to Stan
for this and lots of other useful
suggestions!
We are still trying to acquire lock in a test cavity using a variety
of
strategies, in addition to our modeling and resonance measurement work.
One observation we have made several times is that, late at night when
ground motion is low, the system will occasionally pass very slowly
through
resonance with the servo loop closed. However, the system does
not even
briefly acquire lock, even when our calculations indicate that it should.
We do not yet know why this is the case, and we are actively investigating
this unexpected behavior.
On other fronts, we fixed a crossover resonance in the mode cleaner
servo,
caused by the accidental introduction of a pole with an SR 560. We
also
modified the mode cleaner servo to use separate amplifiers for the
mirror
and PZT path, so that the crossover is adjustable.
We re-checked the laser power and found that, with the laser turned
up to
550 mW, about 350 mW is making it into the vacuum chamber. Some
gets
diverted to the PSL, but most of the missing light gets eaten by the
Faraday isolators.
no report
Simulation and Modeling (Bhawal)
Thermal Lensing
---------------
Last week, Hiro attended the Advanced LIGO optics modeling meeting
at
MIT. e2e does not have thermal effect yet, but based on the
discussion with Ray, it turned out (1) modal model with mode up to
4
is enough to simulation semi quantitative results and (2) most of the
Melody code can be used, possibly together with some run results
(like temperature at fully loaded time).
Generalised LIGO model
----------------------
(Hiro) keeps working to complete the generalized version of Han2k to
be
easily customizable for other IFOs. This version of the package will
need a newer version of modeler (better macro handling) and alfi
(better multi line code handling) to make it easy to customize.
Mode-matching
-------------
(Biplab) Studying mode-matching problem when input beam is not perfectly
matching with cavity eigenstates. Started with FP cavity. When mirrors
face each other, primitives and summation cavities produced same results.
When the end mirror faces outside, there is a small difference in results.
Also studying this issue in recycling summation cavity.
Had discussion with Erika on FFT runs. Started playing with FFT code.
Finally it would be used to compare with E2E's results and to make
quantitative judgements on questions like how many modes would be sufficient
for accurate simulation of the degenerate recycling cavity of LIGO.
Pre Stabilised Laser
--------------------
(Biplab) Studied with Rick frequency sensor noise from transverse motion
of RFPD in PSL.
Code development
----------------
(Andrea) I started to implement in the E2E the capability of producing
frames. I am also continuing to collaborate with VIRGO, supporting
the
tuning of the mechanical simulation in response to new measurements
of
the control transfer functions.
Alfi
----
(Ed Maros) Worked on block quoting that does not require escape characters.
(Bruce) Continuing work on implementation of connection junctions.
LIGO Data Analysis System
Software Systems (Blackburn)
The Inspiral MDC began this week. There is a very strong representation
from the LSC this go round including UWM, PSU, LHO, and CIT. Overall
the
Mock Data Challenge is going quite well and the new LDAS dataPipeline
is
truly being stress tested. Lots of little problems have been identified,
but nothing that is a show stopper. Being tested on the LSC side are
the
new release of the LAL library and the inspiral and fct search codes.
The
biggest problems have been with file I/O from the dataConditionAPI
which
is not thread safe and is rooted deep in the gcc distribution, along
with
an NSF delay on the beowulf nodes which causes jobs to start up to
slowly
for the LDAS codes monitoring the wrapperAPI. The testing is about
half
way finished with another day of testing tomorrow. It looks like we
will
be able to finish on time with only minor issues with documentation.
The
most challenging tests will be performed tomorrow when blind searches
for
inspiral waveforms will be carried out.
Andrea Vicere had two meetings with Sanjeev Dhurandhar, who is visiting
the LIGO laboratory. They are working on how to adapt his proposal
for
coalescing binaries coherent network analysis to the limitations of
the
inter-observatory bandwidth: the purpose is to understand how much
sensitivity is lost by splitting the analysis in two steps, one local
to
the observatories, the other combining only those results which
are above threshold in the local analysis. The next step is to
understand what are the practical consequences on the DA pipeline.
Work is continuing this week on a new framebuilder at the sites. Also
in the background of this MDC progress is being made on implementing
a
threaded version of the XML library in the lightWeightAPI.
Peter Shawhan attended the annual conference of the International
DB2 Users Group, and learned a lot more about database design,
administration, and performance tuning.
Hardware Systems (Anderson)
no report received
Data Analysis Activities
Vicere'
* Data analysis: I worked to support the MDC for the production of
data
in frame format.
Some changes to the library were requested in order to
increase the
flexibility of inputs and outputs.
* Network analysis: I had two meetings with Sanjeev Dhurandhar, who
is visiting
the LIGO laboratory.
We are working on how to adapt his proposal for
coalescing binaries coherent
network analysis to the limitations of the inter-observatory
bandwidth: the
purpose is to understand how much sensitivity is lost
by splitting the analysis in two
steps, one local to the observatories, the other combining
only those results which
are above threshold in the local analysis. The next step
is to understand what
are the practical consequences on the DA pipeline.
Lazzarini
* Sanjeev Dhurandar is visiting for a period of 6 weeks. He is collaboratiing
with several people in the area of improved hierarchical search algorithms
that take advantage of a decimation-in-time step in order to reduce the
NlogN costs of the first step FFTs. The approach he is exploring is founded
in several discussions he had with people in Millikan last August when
he visited briefly. He has been discussing this work with Benoit Mours,
who has been developing a different variation on the same theme. Sanjeev
is also working with a student on implementation of a network based analysis
for multiple interferometers. He has been working with Andrea Vicere' on
this during his visit.
* This week I met with Massimo Tinto from JPL. He and Tom Prince have
discussed an exchange of effort during the next period of time while Tom
serves as JPL Chief Scientist. Massimo would spend time at LIGO on data
analysis activities to and to help supervise the student/postdoctoral scholar
group we share with Tom. Massimo is interested in understanding how the
systems we have been developing may be relevant to LISA for data analysis.
General Computing (Wallace)
MIT:
(Keith)
-Installed and configured first cluster machine
-Installed and ran memtest86
-Installed redhat7.0
-Installed redhat7.1
-Replacing failed HDD on print server / code
repository
-Moved out machines from computer lab for one
time floor waxing
Livingston:
-Worked the E4 run.
-Ordered a number of new computers to handle different services.
Hanford:
(Christine)
- Investigated and requested quotes for new tape robot, disk raid
system, Sun Solaris compatible CDRW and compute farm. Decided
to buy a
Exabyte 215M tape robot, Electronix RaidWeb Arena II disk raid system
8X75 Gb drives, Yamaha CD-RW drive, and Sun Blade 1000. Started
procurement process for each of these.
- Prepared a FY2001 expenditure report.
- Prepared for power outage to site on Friday 5/18/01.
CIT:
(Barbara)
- Created and installed a labor resource summary report for the Costbook.
Created and installed a new version of the detail report that pages
through
the activity sheets for a WBS or system. Imported the PSL activity
sheets
and line items.
- Finished a first draft of the LDAS hardware block diagrams.
Next will
set them up for review.
- Installed the April newsletter and made other web site changes to
Talks,
MOUs, ...
- Found a potential site map creator for the LDAS site.
(Lisa)
- Finishing up install of forte6 compilers on sirius
- configured a time server on rana to support the VME crates.
- finished up with the CDS ultra 60 for mit.
- made a pile of dns changes
- locked account jugas and deleted alias ugas_j.
- finished getting an OS and drivers installed on a tecra 8000 laptop.
- got the daily and weekly backups for ldas back on track.
- usual round of user support.
(Sam)
-Setup for senior Ditch-Day
(Larry)
-Worked on some financial issues with Albert and company.
-Fixed a number of PC issues. Most were setups that needed to be repaired.
Others were just loose cables.
-Worked on WAN documentation.
-Restored a number of directories for different people.
-Regular fixes and moving of equipment.
From Bill Kells:
Was At Mit most of last week for Adv. LIGO modeling.
So far this week have been finishing up on
repercusions from considerations there.
From Janeen Romie:
New osems for 40m, LASTI, PNI etc.: Pal Pilot has delivered the damaged
pigtail connectors here today. We will cancel the contract with them.
I
will fix the connectors and give the new ones along with the rest of
the
components to Southern California Braiding (our second quote) for
fabrication.
Advanced LIGO: Holding another telecon tomorrow morning with Norna
Robertson and Phil Willems
about advanced suspensions documentation. Helena and I are working
on
the schedule in preparation for the review June 18th.
From Rich Abbott:
Obtained more information on the capacitance position detectors. I have
narrowed down the choice of vendors to one, and have received a detailed
report on a solution that is advertised to fill our needs. Pricing and
physical dimensions are now in house. I spent some time on the phone with
them talking about implementation details. I am now starting to work on
the voice coil actuator and the ADC requirements for reading the output
of the capacitance position sensors.
From Peter King:
2.2 40m Lab & LASTI
Lee has installed the reference
cavity in the 40m Lab. It should
be connected to the roughing pump some time tomorrow. Other components
should be installed by the end of the month. Or at least, that
is the
intent.
Apparently the reference cavity vacuum chamber shipped for the LASTI
PSL
had a little misadventure. Lee has held back shipping the other
vacuum
parts until we hear whether or not the chamber is okay.
Had a quick meeting with Alan about who was providing what on the table.
2.3 Advanced LIGO PSL
I found out that the person
that I have been dealing with for getting
a good price for the pump laser diodes from Coherent, opted for a career
change.
I have not had a chance to contact the new rep. about our requirements.
From Mike Smith:
40 METER
Optomechanical layout:
Video cameras for viewing the faces of SRM, PRM, and BS mirrors were
relocated to the top of the input chamber with appropriate steering mirrors
in the BS chamber. Ion pumps and associated plumbing were added. BS PO
ISC table was moved closer to the beam tube. Main laser power attenuator
was added to the PSL table. Cable tray and rack locations were moved to
reflect the as-built configuration.
ADVANCED LIGO
Cost Data for the Stray Light Control and PO Mirror Assy and Telescope
sections of the AOS Task was input into the LIGO cost book.
From: Riccardo DeSalvo <desalvo@ligo.caltech.edu>
Florian arrived in Pasadena, he is a NC
machine industrial control engineer student. Virginio is training
Florian on our DSP control system using the old IP tower and
implementing damping. They call things with different names!!
Weekly 15 May 2001
Florian arrived in Pasadena, he is a NC machine industrial control
engineer student.
Virginio is training Florian on our DSP control system using the old
IP
tower and implementing damping. They call things with different
names!!
Florian is expected to move to Japan in 2 weeks to boost Akiteru’s
operation.
Virginio
Working on the old tower control system. Modified LVDT assembly
to
allow better nulling of the rest signal.
Need to change tuning platform for accelerometers.
Alessandro
Tests on Ultra Sound cutter.
Riccardo
Still no bid from Sonic mill for sapphire machining.
Rewired large part of the creep facility data acquisition system to
eliminate ground loops and disconnect with 10K all ADC inputs. The
noise
effects are reduced but the ADC still see the switching of the small
fan
in all ADC channels as a level. Problems in the DAQ card?
Gone to Japan.
Akiteru, Kenji and 2 undergrads.
Magnet coil actuators in the IP are more than 10 times too strong.
Smaller magnets will reduce by a factor of 5, another factor of 2-4
will
come from wiring less spires on the coil and the rest by reducing the
gain of the driver.
Accelerometers in new R/O configuration are much easier to operate
and
robust, performance with present sensor is 5 10-5m/R(Hz) at 1 Hz.
May need to implement the more sensitive capacitors.
The IP is too good, the attenuation could be sensor limited above 1
Hz
Tuned completely one tower, all sensors and actuators characterised.
Working on the separate Suspension characterisation. Measuring
the
reaction of the actuatio
Problem with soldering Tungsten wire in the miniGAS springs,
the solder
does not wet the tungsten, and slips from the soldering cup.
Problem
temporarily solved with a kink before soldering, but will need to coat
the wire. This caused 2 weeks delay.
From: David Shoemaker <dhs@ligo.mit.edu>
Subject: SWG: agenda and updates, telecon Friday 10 May 8.30 pacific
SWG telecon: Agenda, May 11, 8.30 am US pacific time
Schedule and Lab information
Potential collaboration with Boris Lunin
Coating losses characterization program
Discussion of updates below
=============================
Updates:
SAS monthly report May 2001 - Riccardo DeSalvo
Following the mechanical installation of the SAS towers in Japan the
sensors and actuators have been installed, cabled and calibrated in
one tower (including the double pendulum and its voice coils).
Ready
for the second. The suspensions actuation is being characterised.
The program (tight) is to have a working interferometer by end of
June.
We are slowly getting some test sapphire flex joint cut by ultra sound
machining. If immediately successful, during the summer we will
get
some diving board like (flex joint with a small mass) to be tested
for
Q factor and strength and thermal conductivity. The ultrasound
machining technique presently under test is the master scheme, in
which a tool shaped like the negative of the part to be obtained
slowly eat into the sapphire. The tool also slowly wears down
and the
cut will not be perfectly straight. We also expect problems when
the
tool exits on the other side that may break the joint. It is
not yet
the technique of ultrasound machining demonstrated by ONERA in which
the tool is always floating over the part being machined (therefore
applying zero DC effort) and the tool wear is continuously monitored
and corrected for. This will require the purchase of a dedicated
machine. The NC ultrasound mill, operated with gasoline based
(as
opposed to water based) slurry, as those used to polish hygroscopic
crystals, may be able to cut quartz flex joints without introducing
water stress in the surface.
After the very successful tests of the horizontal accelerometers in
Japan, Alessandro is starting the development of vertical ones to
eventually subtract gravity gradient noise from advanced suspensions.
===================
Stanford SWG Telecon Monthly Report 3rd May 01 S. Rowan, R. Route,
M. Fejer and colleagues
A) Coating loss studies Measurements have been
made of the Q factors of modes of two LIGO I size test masses. The
first mass was a polished but uncoated spare end test mass, the second
was a coated folding mirror. Three modes of the uncoated mass
were
found to have Q's of approx 10^7. When the corresponding modes
of the
coated mass were measured one was found to be at a level of
approx. 10^7 whilst the other two modes had slightly lower Q's.
To
interpret these results requires that the identification of the shapes
of these modes - David C. at Glasgow is running his FE model to do
this. In addition it should be noted that the masses were polished
by
two different vendors and thus may have surface layers with different
properties. An ear was silicate bonded at Stanford to the coated
fused silica sample being studied at Syracuse, to allow this sample
to
be re-suspended via a silica suspension so that further measurements
of the Q of the sample can be made.
(B) Silicate bonding JH and HA
visited Stanford to participate in the silicate bonding of attachments
to 3 of the final GEO optics (two far mirrors and a spare) and their
corresponding intermediate masses. The optics were polished by General
Optics and coated by REO. Several items of interest - GEO have
adopted a new design of ear - so that one, larger area ear, rather
than two smaller ears, is bonded to each side of a mass. This reduces
the number of bonds to be made and includes a larger safety factor
in
terms of the bond strength. Sodium silicate solution was used for the
bonding rather than potassium hydroxide solution. Tests at Glasgow
suggest a level of bond strength close to a factor of 2 stronger in
shear/peeling than that found using KOH solution. The ears were
bonded to flats polished on the sides of the optics. The flatness of
the areas to be bonded was checked by looking at the interference
fringes formed between a reference flat and the bonding surface. On
two of the coated masses distortions of the fringes could clearly be
seen, mostly occurring towards the edges of the flats closest to the
HR coated face of the optic. It seems likely that this was due to
overspray of the coating from the front (HR) faces of the sample -
the
fixture used to protect the bonding surfaces may have had small gaps
allowing small amounts of coating material to leak through. Even
with
these distortions the ears appeared to bond successfully to the masses
- however small imperfections are visible in these bond. 3 pairs
of
sapphire samples were bonded together using sodium silicate solution
as part of some work with the TAMA project to investigate the use of
this jointing technique in constructing all crystalline test mass
suspensions
============================
TNI -- Ken Libbrecht
At the TNI we now have the vacuum chamber closed and pumped, the mode
cleaner locked, and the test cavities aligned. We can produce error
signals in both test cavities, and we are currently working on locking
schemes. (We had previously locked a test cavity after bypassing the
mode cleaner, which allowed us to use the fast PZT input on the
laser. We are trying other schemes to produce a sufficient servo
bandpass for the lock to catch.) Crystal systems dealt us an
additional 4-month delay in producing our sapphire test masses, and
they now say they'll ship in September. If all goes well this
shouldn't impact our schedule too badly.
===================================
Phil Willems
Fiber research
--------------
Sadly, I must retract my announcement at
the LSC meeting of having observed nonlinear thermoelasticity. There
was a silly math error in the theoretical prediction. The Q values
and
(dE/dT)/E values are still good. Work continues here at Glasgow on
measuring the Q of the vertical bounce mode of fused silica to check
for stress dependence of the loss. We expect to install suspensions
next week.
Silicate bonding
----------------
Our tests of sapphire/sapphire and sapphire/silica bonds have been
compromised by poor surface flatness at the edges of our 1/2" sapphire
substrates. They are being reworked for uniform flatness.
============================
Report from Glasgow for Suspensions Working Group Telecon on Friday
11th May 2001.
1) Bonding Issues in GEO. (see also Sheila Rowan's report from
Stanford).
We have adopted sodium silicate solution for bonding the ears (prisms)
to GEO mirrors. The bonds, on measurement, are close to a factor of
2
stronger in shear/peeling than with potassium hydroxide solution.
We have increased the bond area from 1 cm^2 on each side of the 6 kg
mass to 3 cm^2 to increase the safety margin to a factor greater than
10. This has been done because of the increased stress on a bond if
a
fibre breaks and the mass 'bounces'. Measurement of the mechanical
loss on samples bonded with silicate suggests that there should be
no
thermal noise penalty even with the increased area.
During April at Stanford Sheila Rowan, Helena Armandula and Jim Hough
bonded ears (prisms) on to three end mirrors and three intermediate
masses. Some problem was experienced with two of the end mirrors due
to overspray of coating on to the flats; however the silicate appeared
to react with the overspray and allow bonding.
The mirrors and intermediate masses are being sent to Hannover, and
the intention is that Jim and Geppo and Phil will go to Hannover in
the first week in June to hang two end mirrors in GEO on silica
suspensions. Other mirrors ( near mirrors, power recycling mirror)
are
already on wires and beamsplitter is expected to be hung on wires this
week. First operation of the Michelson will be tried once the end
mirrors are in place. The expectation is that the near mirrors and
the
final good beamsplitter will be bonded and hung on fused silica fibres
later in the year.
2) Prototype Quadruple Suspension We have now taken delivery at
Glasgow of the majority of parts for the all metal quad. pendulum and
reaction pendulum. The support frame and the blades are still to come
(expected this week and beginning June respectively). Everything will
be sent out to MIT in June. Calum and Norna have fixed up travel
plans- arriving MIT late June staying for ~ 1 month.
3) Design issues A meeting was held last week w/Phil, Ken and Calum
to
talk through design issues for the recycling mirror suspensions. These
may be tightest to fit into HAM tanks due to large size (26.5 cm
diameter). Phil is running through Geppo's thermal model to get
initial parameters, and then we will address the mechanical model.
=====================
Helena Armandula
At Stanford, participated on the silicate bonding of prisms to GEO's
intermediate masses and mirrors. The purpose of the visit was
to be
able to learn the process and transfer the technology to Caltech.
Expect to return on July 23rd. to finish the attachment of prisms to
the rest of the GEO optics.
=====================
SWG monthly progress report for the Syracuse group submitted 9 May
2001 by Peter Saulson
Steve Penn has confirmed that flame-polishing a
superpolished surface can dramatically reduce the surface friction.
A
thin disk that had Q's in the range of 4 to 6 million showed Q's up
to
26 million after an aggressive flame polish. These numbers have been
confirmed after resuspending the sample. Unfortunately, the surface
now has terrible optical quality. We (with help from Dave Reitze's
group and the MIT group) are looking into the possibility that a
localized repair process, such as by scanning a high-power CO2 laser
beam over the surface, might give the high Q without damaging the
optical quality. Josh Smith has extended his measurements of
losses
due to silicate bonds, on samples supplied by Glasgow. On a second
bonded sample, one of the modes (flexing in the plane of the bond)
shows a Q of 110,000. This compares to 200,000 or a bit more for the
first sample he tried. A control sample (of the same glass polished
and shaped in the same way but not bonded) had a Q of over 9
million. Josh will try to get the orthogonal mode of the second sample
measured this week, before he leaves for a summer working with
GEO. What he doesn't get done will be picked up by a summer
student. (There is also a new sample bonded with a sodium silicate
solution, handed to us by Jim Hough at the LSC meeting.) Andri
Gretarsson has been improving his violin mode monitoring setup. He
has
written new data acquisition code, and has also built a servo to track
the pendulum mode of the fiber. Now he can take full advantage of the
high sensitivity of his sensor without the dynamic range limitations
that had plagued him in the past. Scott Kittelberger has built
and
tested a proper anti-aliasing filter, and completed a new round of
systematic tests. He is about to start comparing anelastic signals
from a variety of fused silica samples. Sasha Ageev is digging
into
Algor, to try to set up a model for the anelastic effect in sapphire
that will properly account for thermoelasticity.
===========================
LASTI -- Mike Zucker
Configuration: Ken Mason is working on the optical layout for the
recently proposed dual-cavity configuration. This plan will
permit tests to more closely approach the advanced LIGO displacement
noise target by increasing the feasible cavity beam size.
The sticky part is the need to get four beams (two
'arm-cavity' beams and the two beams of a 15-meter triangular
mode cleaner) to fit through our 30" beam tube. So far it looks
feasible in AutoCad.
Cleanrooms: The custom BSC cleanroom/work platform and "cartridge
installation" craneable cleanroom arrived from the vendor and were
installed by our rigging subcontractor. Fit is excellent and
the dome
flange-level work platform feels solid. Curtains have been installed
and the blowers and lights are being wired up for testing. We are
are modifying the safety rails slightly to facilitate
temporary opening of gates for passage of crane loads.
Electronics: We received more cross-connect components, power supplies
and two more powered Eurocard crates from Caltech.
PSL: Dave Ottaway joined us this week and will begin organizing the
assembly and test of the PSL. The reference cavity vacuum chamber sent
from Caltech arrived seriously abused by UPS; we are evaluating if
they did any irreversable damage .
Lab space/SEI externals: the "Stiff" LIGO II active seismic
isolator prototype was crated and will be loaded on a truck bound for
Stanford today. The cleanroom for HAM19 is complete and operational.
The
remaining HAM19 seismic supports were staged for installation and a
parts inventory completed. BSC fixtures were cleaned and staged for
upcoming support installation. We are overdue for fine alignment of
the BSC piers, plan to take this on next week after the cleanroom is
in.
For additional information about this report, contact sanders@ligo.caltech.edu