Press here for ACCOUNTS PAYABLE HISTORY DATA.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday July 10, 2000 will be:
CANCELLED DUE TO GWIC TRAVEL
No report this week.
WBS 1.2 LIGO Operations--Administration
From: Ed Chargois <chargois_e@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. . .
From: Cleveland Mak <mak_c@ligo.caltech.edu>
From: Esther Cunningham <esther@ligo.caltech.edu>
Press here for ACCOUNTS
PAYABLE HISTORY DATA.
From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
No report this week (vacation).
Rita Torres
Graph provided
by Cindy Akutagawa
Please note that this report covers only three days due to the four day Holiday Weekend.
Progress Period from from 6.30 to 7.6
Accomplishments:
The following change requests have
been submitted:
| CR-990028 | WBS 1.1.3 | Beam Tube Enclosure Closeout | F. Asiri |
| CR-000005 | WBS 1.2.1 | Upgrade Pre-stabilized Laser | S. Whitcomb |
| CR-000006 | WBS 1.2.1 | Re-polish Core Optics Components | S. Whitcomb |
| CR-000007 | WBS 1.2.2 | Replacement of Optical Lever Lasers | S. Whitcomb |
| CR-000008 | WBS 1.1.4 | Cameras and Projection System at LIGO Livingston Observatory | F. Asiri |
| CR-000009 | WBS 1.1.4 | Cameras and Projection System at LIGO Hanford Observatory | F Asiri |
| CR-000010 | WBS 1.2.2 | Redesign Suspension Controllers | S. Whitcomb |
| CR-000011 | WBS 1.2.2 | VME Development System and Spares | S. Whitcomb |
| CR-000012 | WBS 1.2.2 | ASC/LSC Rework | S. Whitcomb |
Copies of these change requests have been distributed to memebers of the LIGO Change Control Board (See LIGO-M000176-00-P).
Press for the latest Contingency Needs Projection.
From: Kris Duncan <kris@ligo.caltech.edu>
No report this week.
OPTICS/COC INSTALLATION: Beamsplitter and ITM-y have been installed and aligned since last Thursday. ITM-x installation is in progress. (Jonathan Kern)
PSL: We are replacing all the New Focus mounts with Optosigma ultrastable mounts to reduce acoustic sensitivity on the PSL table. We are currently realigning the reference cavity. (Joe Kovalik)
Administrative: Review of site expense of past year (7/99 - 6/99) indicate
transactions ~ $50K in total have been mis-allocated to Site Support (5L503).
Request put in to Caltech Financial Account to do a cost transfer from
LIGO.00005-3.8. (Coles, Wascom)
All items covered elsewhere.
All items covered elsewhere.
| Installation
& Commissioning:
Livingston |
Other Science/Engineering
Activities:
Issues/Concerns |
See also the Installation web page
We are currently looking
at two different designs for the AC OSEMS. The first is the suppressed
carrier proposed and tested by Rai Weiss. The other is an amplitude stabilzed
oscillator design. We are still trying to have a few in place by late summer.
No report this week.
Not much to report this week due to the Independence Day holiday.
Having
gotten the test cavity in our photothermal noise experiment to resonate
we
have been looking for noise sources. We found some noise in the
mirror
clamps in the test cavity and are in the process of replacing them
with a
new design. We also ordered some new mounts for the input optics
on that
experiment, again to reduce mechanical noise and make alignment easier.
No report this week.
Biplab worked with Brad Zamft (Rick's summer stundent) working on
the
shotnoise calculation of the reference cavity. The e2e generated half
the
amount of shotnoise compared to the value calculated by E. Black, but
further discussion among Biplab, Hiro and Eric resolved the discrepancy
and the e2e prediction turned out to be correct. Now, the complication
caused by the higher order mode is being investigated.
Luca studied the effect of the misalignment on the locked Michelson
system.
He is going to work with Matt to study the real impact of the effect
on
the
stability of the lock and is going to estimate the accuracy of the
prediction.
Hiro kept working on the parser/runtime-compiler. Now local functions
can
be included in the formula.
G.Cella sent us a data file of the mirror motion of the reference
cavity.
This is the convolution of the measured motion of the PSL table top
and
the MSE simulation of the reference cavity. With the up coming e2e
module to generate mirror motion based on the data file, the effect
of
the table motion on the optical signal will be simulated to study the
PSL noise.
Ed completed the make environment of MSE.
Bruce kept working on cleaning up the alfi code.
Test results for ingestion of data using this version of LDAS are
consistent with older tests.
Test results for translation of frames into ilwd objects are also
consistent with older tests.
Test of the transmission of ilwd objects through sockets are not
yet complete. However, preliminary results are showing slightly
higher results. These results are slightly higher than is to be
expected from 100bT ethernet and testers are reviewing code to
see if a rate computational error can account for the roughly 15%
higher than expected results.
A new socket communication timeout problem showed up on the Hanford
system between the controlmon(Intel) and dataserver(Sun). It is not
clear yet if this is isolated to these two boxes or exists between
any pair of Intel and Sun. Test are under way to determine if it is
software or hardware related.
The Guild user interface has new TCL/TK functionality to support a
request for a list of available channels in frames.
The framebuilder is being upgraded to write out version 4 frames. This
will be very important for testing LDAS's version 4 frame functionality.
The dataConditionAPI group meet and discussed readiness for the Mock
Data Challenge at the end of the month. The largest remaining obsticles
are the completion of the command chain interpreter in the C++ layer
and
the development of the TCL layer for the API. The TCL layer will require
the most work to have ready for the MDC.
The new release of LDAS is being delayed until more information about
the
ilwd socket transmission tests are available. The results of these
tests
should be available by the end of this week.
Hardware:
>From Greg,
1) The shelving for the beowulf cluster at LHO has been sorted
out.
The necessary pieces to build a least 3 shelving units (4 shelves per
unit) are here, or are being sent here from Livingston.
2) The Sun D1000 288 GB disk unit was incorrectly installed on
the ldas-wa
dataserver on June 30 and will be re-installed shortly.
3) Upgrades of the OS to RedHat Linux 6.2 on the ldas-wa linux boxes
has
begun.
4) The moving of the cds racks across the mass storage room to make
way
for the beowulf shelves has begun today (7/6/00).
>From Omar,
1) Upgrading beowulf server and nodes to redhat 6.2
2) Moving system critical software to local disks from nfs to improve
overall
stability.
3) Continuing work on ups shutdown.
MIT:
Working on getting the new computer room arranged to handle future
installs.
Hanford:
Installed the latest version of gcc compilers on the license server.
Worked on several procurement issues for new software and hardware.
Provided Sys. Admin. support for the many summer students and visitors.
Livingston:
Working on getting the FORE equipment s/w upgraded.
Finally, received some quotes from Bellsouth concerning upgrading the
network
connection to a higher data rate. We've asked them to also give us
a cost
estimate for OC3 connection speeds.
CIT:
(Barbara)
- Added features to the CostBook web forms for Capital Projects.
Added
Location field to Equipment forms. Made a number of changes related
to
hourly rates.
- Created/installed web page describing Hanford fire from Fred's e-mail
and
photos.
- Made minor updates to LDAS web site.
- Made a temporary fix for a problem on the NT DCC server. The s/w
will hang
once in awhile, presently looking for a patch to resolve the issue.
(Suresh)
- Installed some application software (Exceed 6.2 and Microsoft Office
97) on
a new PC to be used by Rick Karwoski. Working on to install ITS site
licensed
sofwares (Matlab, Mathematica etc) over the network.
- Fixed system and printing problem in Sun system tiger.
- Worked on user account and aliases related problem.
(Samantha)
Hobbled around until discovered the joys of rolling chairs, almost
as good
as rollarblades, although there is not quite the fine control.
- Completely rebuilt the computer the operates the DVD and the
Scanner. The operating system was complete messed up, thus we started
at
the beginning with an FDISK and moved forward from there. Loaded
the
operating system again, started loading the software, and then ta da,
got
to start over again.
- Building a test machine to try out a new way of backing up NT
servers. Currently in the process of making sure everything works with
the current operating system before it gets annihilated in favor of
NT
SERVER. Also making sure box is stable to use for future software
testing
- Burned some CDs that have been needing to be burned, configured the
software
that was going on the CD.
- Randomly decided to visit Eudora web site, discover that Eudora is
now
up to 4.3, as soon as the computer has decided that it has been formatted
enough times and that it really would like to have an operating system
installed on it that will refrain from crashing every ten minutes will
test out the new Eudora 4.3 and see what new and cool features that
it
has. Hopefully it will make tunneling easier.
- Started to write up some documentation. Discovered that through
my
Caltech career I have lost any ability that I might once have had to
write
proper English. I no longer recall how to structure sentences
and worse
yet I no longer recall how to fully spell out words. The spelling
deficit
could be overcome if I reverted to the tried and true way of abbreviating
everything, but concluded that only I would know what the abbreviations
stood for, thus this would render the documentation useless.
(Lisa)
Worked a number of issues at Wilson House. Working on getting more
information
for the modem pool replacement. Putting in more documentation for existing
processes performed by the Sys. Admin. group. Working on resolving
some CADENCE
licensing issues.
(Larry)
Working on getting a few of the items on the large SUN order expedited.
Yes, it
has gone through but most items are not scheduled to ship until August.
Working on some of the network issues for upgrading the network at
Hanford,
Livingston and CIT.
Wrapping up items for the 01 General Computing budget plan.
Ordering a few more items to get the Wilson House network connection
moved over
to single mode fiber all the way from end to end. Presently, there
is a
combination of multi-mode and single mode.
So far, the testing of the GNATS system for GC problem reporting has
hit a few
bumps but we will continue with it for a few more weeks before deciding
if it is
a useful tool or not, for our purposes.
David Reitze wrote:
LIGO II:
- Worked on LIGO2 concept drawing for Tanner
- Worked on a length sensing and control scheme for LIGO II and
conducted
measurements regarding thermal lenses in TGG crystals.
- LIGO Input Optics Reference Design Study document is online
at
~tfrey/index.html under L2 Proposal
Data link.
From: Sam Richman <srichman@ligo.mit.edu>
MIT two-stage isolation protoype (Sam Richman, Jamie Rollins, Jon How)
We now believe that the substantial tilt coupling seen while driving
with horizontal forcers at low frequencies is caused by the anisotropic
nature of the blade springs. If one pulls horizontally on the bottom
of the flexure that connects the tip of the blade spring to the suspended
stage, the spring displaces vertically by different amounts depending
on which horizontal direction one pulls. This seems to explain the
large tilt effect, and why it is the same size with all three horizontal
forcers.
Jon How (MIT aero/astro prof.) has used Jamie's TF measurements to construct
state-space models of the system, and we will now try to develop controllers
based on these models.
_______________________
SWG telecon and Monthly Progress summaries, Thursday, July 6, 8.30
pacific, one hour
Agenda:
1) Lab/Organizational issues
2) Suspensions
a) suspensions per se
b) fibers fabrication
c) fibers noise performance
3) Test mass materials and integration
a) coating losses
b) bulk losses: sapphire, fused silica
c) bonding, welding
4) Thermal noise measurement
a) TNI near-term
b) plans long-term
5) Isolation systems
a) stiff system
b) soft system
6) LASTI; coordination of efforts in a general sense
=============================================================
Some notions of significant events in SEI-SUS over the coming years to discuss. These are my interpretations of dates and places; please chime in with changes or misconceptions in our discussion or beforehand.
- a round of SEI (isolation) prototyping on the HAM design is in planning, to be performed at Stanford; assembly will start in January '01 with testing starting in February '01
- a vacuum-ready SEI HAM system design will start halfway through the
process above, which will arrive at LASTI in Oct
'01
- a vacuum-ready SEI BSC (test mass isolation system) will follow, date TBD, ideally ~6 months later
- the SEIsmic isolation systems will be first tested stand-alone, to determine if the controls are ok and to measure isolation transfer functions and residual optics platform noise (using seismometers)
- suspension design work to be reviewed in 4q00, with GEO leading the effort
- prototyping of suspensions is on-going at GEO, with significant interchange LIGO-GEO as LIGO people come up to speed on the design and develop fabrication techniques
- by 4Q01, we will come to close on many detailed aspects of the suspension design: test mass materials properties, fiber/ribbon choice and fabrication technique, bonding/welding development, characterization of the violin string excitation. The Preliminary Design Review will take place, and the LIGO Lab will start to lead the suspension final design and fabrication work.
- a few (2 or 3) 'controls' SUS MC (mode-cleaner suspension, steel wires, dummy masses but with a mirror attached) can get to LASTI in October '01, and a 'controls' SUS TM (test mass) quadruple shortly thereafter.
- then a suspension will be mounted on the SEI system to ensure mechanical fit, and controls compatibility
- a cavity will be build up of (maybe) one TM suspension on a BSC SEI, and two MC suspensions on a HAM SEI, to test overall controls.
- sapphire and fused silica are being pursued, with fused silica either a fallback or 'step one' of a staged approach. A choice will be made in early '02, with trial and special-purpose test mass optics fabrication started
- the TNI will have made tests of sapphire substrates by early '03 to check on the thermoelastic noise model, might be moving on to different spot sizes
- suspensions with the final test-mass material come to LASTI for characterization and noise testing in 3q03
- Lastly, there will be 'first article tests' of the first units running
off the production lines; this is just months before installation at the
observatory starts, and mostly checks installation procedures and fit-checks
of the production isolation and suspension units.
=============================================================
GEO/Norna Robertson:
Glasgow GEO suspensions group - SWG report, July 1 2000
1) Visit from Andri Gretarsson for a week to discuss losses in fused
silica, sensors for violin mode damping and experiments to determine
coating loss. Together we bonded together 3 halved silica rods for
bond
loss measurements to be carried out in Syracuse later in the summer.
2) During the course of a series of bonding experiments/evaluations
we
discovered that superpolished ears from GO were not bonding to
superpolished flats from GO, after standard cleaning procedures. This
situation was rectified by pre-etching the surfaces either with
concentrated detergent of saturated Ammonium difluoride solution for
half
an hour after which bonding could be carried out as normal. (Geppo
Cagnoli
and Jim Hough) This has been discussed with Phil Willems and Helen
Armandula at Caltech.
3) While considering the LIGO II suspension design for the requirements
document and the conceptual design document we realised that for a
fixed
sapphire mass there is an optimum mirror diameter to minimise the effect
of
thermoelastic damping. For a mass of 30kg Geppo Cagnoli has calculated
this
to be approx. 28.5 cm. If we use 32 cm diameter we must increase the
thickness and the mass for optimum solution. Thus consideration must
be
given urgently to what mass we should design for LIGO II.
4) Algor is being used to calculate mode shapes to allow better
interpretation of the effect of dielectric coatings on the losses at
the
resonant modes of a silica mass (David Crooks and Peter Sneddon).
5) Algor is also being used for the preliminary modelling of twisted
silica
ribbons (David Crooks)
6) Calculations on the thermoelastic effect arising from change of
modulus
with temperature are close to a conclusion (Geppo Cagnoli and Phil
Willems). This has implications for pendulum thermal noise in LIGO
II design.
7) Preliminary quadruple design for a 32 cm sapphire mass has been
carried
out. (Calum Torrie et al).
Next month
1) Gregg Harry visits mid-July for two days for further discussions
on
silica losses etc.
2) Sheila Rowan is bringing back sapphire mass with bonded silica cone
for
loss measurements
3) Sheila Rowan is also bringing back silicon masses for loss testing.
4) Work will progress on construction of new fibre and ribbon pulling
machines.
5) Further work on overall suspension design, incorporating considerations
resulting from choice of mass, choice of lower frequency cut-off, and
effect of new thermoelastic damping calculations.
=========================================================
Stanford/Sheila Rowan:
Interim report on LIGO II related work.
(1) In a parallel set of experiments with the Glasgow group a set of polished fused silica rods has been silicate bonded for use in a collaborative experiment with Syracuse on measuring the loss associated with silicate bonding
(2) A fused silica ear has been bonded to a thin fused silica plate in order to make a low mechanical loss suspension for use in the coating experiment at Syracuse
(3) Discussions with Phil W. and colleagues in Caltech to assess plans
for further bonding studies for both silica/silica and silica/sapphire
bonds.
Some preliminary silicate bonds between one inch diameter silica/sapphire
substrates have been made. These will be strength tested at Caltech using
a modified version of the strength tester there.
(4) Work has continued with Kenji Numata of the TAMA project on understanding the finite element modeling of an-isotropic materials with a visit by him to Stanford on 22nd June.
(5). A fused silica attachment has been bonded to the sapphire mass
described in previous reports, for mechanical loss studies. This is now
setting and Q measurements will start in the next few weeks.
Plans for the coming month:
(1) Attendance at the MG9 meeting in Rome followed by a lab. visit to
the
VIRGO group at the University of Perugia
(2) Initial Q measurements of the silicate bonded sapphire sample
(6) Continue with the Q measurements of single crystal silicon samples.
===========================================
Caltech/Phil Willems:
Silicate bonding:
-----------------
We have tried to understand the structures seen in silicate bonds and
what
causes them. The easily visible crystals appear to be KOH precipitating
in gaps
within the bond. Fainter, fernlike structures that I call 'rivers'
appear to be
channels by which water escapes from the interior of the bond to the
edges.
Sometimes a diffuse etching is also visible. Sheila Rowan has found
that these
features are largely absent if the substrates are extensively rinsed
in DI water
before bonding. We are also investigating this. We have two of Sheila's
most
recent bonds on loan which we are studying optically to try and measure
the
thickness and refractive index of the bond layer.
The bonds made at Caltech to compare strength with the bonds made at
Stanford
will be tested today. We are working the kinks out of our fixture to
test how
bonds cure under LIGO II loads. We are also modifying our strength
tester to
break 1" silica/sapphire bonds that Sheila Rowan has provided.
We have found that silicate bonds that are broken cleanly while still
new and
then cleaned and rebonded are substantially stronger the second time.
This may
be due to the cleaning or due to a preliminary etch of the surfaces
by the first
bond. This seems to corroborate the finding in Glasgow that etching
off the
superpolish can make the difference between bonding and not bonding.
Fused silica fibers/ribbons:
----------------------------
The calculations of the nonlinear thermoelastic damping are essentially
done and
the paper is being written up. Published values for materials parameters
of
fused silica indicate that the optimum thermal noise is achieved at
250 MPa
static stress, not including surface effects.
The strength of our fused silica ribbons is not satisfactory for LIGO
II. The
strength values average around 400 MPa in direct tension. Bending tests
show
strength values around 3 GPa, so the glass itself seems to be good-
the problem
could be isolated flaws or details of the tension test or something
else. We
have tried chemical etching and flame polishing without much effect.
I am still investigating CO2 laser welding as a technology for LIGO
II.
============================================
Syracuse University/Peter Saulson:
Syracuse group's monthly progress report to SWG
submitted by Peter Saulson
4 July 2000
1. Q of fused silica rods
We are far along in the drafting stage of a paper on the observation
of Q of up to 57 million in a 3 mm rod of Suprasil 2. It will be shared
with SWG shortly, and submitted to LSC review.
A big unknown is the relative importance of the remaining surface
losses. Astonishingly, the Q for this sample falls almost on the line
of Q vs. diameter for the smaller samples measured by Gretarsson and
Harry 1999. A naive attempt to subtract out surface losses by taking
seriously the old surface loss model would imply bulk loss of Suprasil
2
well below 1e-8. We don't really believe this -- an equally likely
possibility is that the surface losses in this sample were made very
small by the flame polishing. The Glasgow group has seen substantially
lower surface losses than we had previously, adding to the plausibility
of this alternative interpretation.
We are getting ready for measurements of Q vs. temperature. This
is interesting because Boris Lunin (Moscow State Chemistry Dept.) has
seen a Q peak around 105 degrees C. His samples showed Q = 30 million
at room temperature, 50 million at the peak. His measurements were
made
at 8.4 kHz, in the 1st longitudinal mode of bars. He and we have
corresponded about the likelihood of our seeing higher Q at 100 C vs
the
alternative hypothesis that our Q peak is really at room temperature,
due
to some structural difference between his glass samples (4 different
kinds) and ours. (Our measurements are also at substantially lower
frequencies, below 1 kHz up to a few kHz.)
2. Anelastic effect in fused silica
We can now reliably subtract out the spurious signal at the rigid
body mode frequencies, the main new systematic effect coming from
suspending our sample.
Steve Penn is now checking his fitting routine by using it to analyze
fake data that he has generated, containing not only signal and Gaussian
noise but all of the awkward stuff that we also see. So far his code
is passing almost all tests (it gives the right value for phi in
every case, so far), but there is still a bit of uncertainty about
whether
errors are being estimated correctly in all cases. This is important
because our biggest present worry about the experiment is scatter between
measurements that don't appear to come from our Gaussian noise.
We are about to try a data run in vacuum, to see if it makes any
difference. It would be surprising if it changed phi, but might make
a difference in thermal drift of fixed birefringence.
3. Mechanical dissipation in coatings
Gregg Harry shipped our large silica sample to REO for coating.
N.B.: REO has recently been very accomodating.
Thinner samples are being prepared for measurements. Sheila Rowan has
bonded an ear on one sample. (Thanks!) A second sample will have a
post
welded on it by Syracuse glass blower John Chabot, just in case it
will
work OK.
4. Violin mode monitor
Andri Gretarsson has just returned from a pleasant and very useful
trip
to Glasgow. He discussed this experiment there, and received a special
expression of interest (and offer of help) from Geppo Cagnoli. Andri
learned of a clever Glasgow idea for using a rotatable glass block
to
allow the laser beam to track the motions of the fiber while staying
fixed on the output photodiode. We will have some telephone discussions
soon between Syracuse and Glasgow, to work out the details of cooperation.
Andri is now testing a new quieter laser, to reduce excess laser
noise. The biggest technological challenge on the horizon will be
adding vibration isolation to his setup, so that violin modes are
excited by thermal noise only.
5. Miscellaneous work
Scott Kittelberger is building up his expertise with Algor. His
current test problem is to model the distribution of strain in the
cylinder-on-cylinder geometry we use now for our anelastic aftereffectmeasurements.
=====================================
Penn State/Gabriela Gonzalez:
=================================================
TNI/Eric Black
Having successfully locked our mode cleaner, we are now preparing our
test
cavities for installation. This month we have been designing
lock-acquisition servos and assembling hardware, as well as looking
into a
second set of mirrors to allow us to study noise in cavities with different
finesses.
We have also been considering the feasibility of smaller-scale experiments
(smaller than the TNI, that is) to measure excess noise in seismic
isolation systems and some of the new noise sources Prof. Braginsky
has
predicted.
==========================================
Caltech/Riccardo Desalvo:
==================================================================
JILA/Tuck Stebbins:
May Progress Report: During the past month, we have worked on constructing
a
test rig to evaluate cross-coupling in the electromagnetic forcers
used in
the two stage active isolation system. In the process of fabricating
the
test rig, we developed and tested a new design for a non-contacting
imaging
position sensor which would meet the design requirements of an active
LIGO
II isolation system. The test rig is nearly complete. We have also
supported the control system work at MIT on the two stage active prototype
that we delivered in March.
June Plan: In the next month, we anticipate finishing the fabrication
of the
forcer test rig and the evaluation of the forcers. We will continue
to
support control system development on the prototype at MIT. We expect
to
begin project planning and early design work on the next generation
of
prototypes.
============================================================
Stanford/Brian Lantz:
==============================================================
LSU/Joe Giaime:
LSU:
Excess noise test facility: Data acquisition computer and various other
instrumentation received and being initially assembled. Quiet granite table
under test. (M. Kingham, Dan Busby)
microseism feed-forward:doing calculations on sensor noise and setting
up to test Guralp on temperature-instrumented block (Y Yamada, summer surf
student)
seismometer test report generated. (W. Johnson): http://sam.phys.lsu.edu/johnson/SeiFeb2000/LivingstonLSCfinal.pdf
LIGO-II SEI team work plan and budgeting (J. Giaime)
====================================================================
MIT/Sam Richman:
(Sam Richman for Jamie Rollins, Shourov Chatterji, Jon How, Joe Giaime)
Shourov has finished constructing 12 channels of forcer current drivers,
and has verified their basic operation. The power transistors were getting
too hot at high inputs, so he is making a better cooling scheme and will
operate at a lower supply voltage. Jamie is setting up the two Siglab units
on loan from the West coast; these will be used for measuring multiple
TFs simultaneously.
Jamie completed a set of open-loop transfer function measurements from
forcers to various sensors on the upper stage, and has Jon used these measurements
to construct state-space models of the system. We now have a better handle
on one significant effect: when driving a horizontal forcer at very low
frequencies, the stage tilts by an amount only 16 dB less than if it were
driven by a *vertical* forcer. We now have a good working hypothesis to
explain this effect. The anistropic behavior of the blade springs, coupled
with the fact that our flexures are mounted vertically offset from the
blade tips by a significant amount, causes a coupling to tilt when the
stage moves horizontally. Applying a 15% correction with the nearby vertical
forcer reduces this tilt by more than an order of magnitude at 0.1 Hz.
This has allowed us for the first time to close a horizontal loop using
a broadband seismometer. The seismometer signal is blended with that from
the corresponding position sensor at about 70 mHz.
Jamie has completed a rig for testing blade springs used in the two-stage
prototype. We have found it difficult to get repeatable (better than 7-8%)
measurements of the spring rate, possibly due to imperfect clamping. The
frequencies and Qs of the parasitic modes are being measured. For the smaller
blade, the lowest-lying parasitic mode is at about 200 Hz.
In the next month, we will largely focus on developing and testing
controllers using the models Jon has constructed.
============================================
MIT/Mike Zucker:
Vacuum/infrastructure: Continued monitoring of vacuum performance.
Ordered special high-flow hydrocarbon absorber, desiccant and HEPA
filter modules for air backfill. Ordered process control relay card
for Granville gauge controller to implement additional layer of
failsafe on turbo gate valve.
Planning:
1) Worked with the suspension and seismic isolation groups to find
the best way to stage the testing at LASTI. Looks like a HAM seismic isolation
test, followed by Mode-cleaner suspension testing, is the way to start.
BSC seismic isolation follows, with a test of a Test-Mass suspension. The
simplest controls and moderate-sensitivity performance test would be performed
in a triangular cavity, with one test-mass suspension (on a BSC isolation
system) and two mode-cleaner suspensions (on a HAM isolation system).
2) Looked at common ground with the Caltech 40m interferometer improvements to reduce the proliferation of designs. We hope to have nearly identical pre-stabilized 10-watt LIGO I-style lasers, suspended steering mirrors, data acquisition, diagnostics, and controls systems.
3) Continued to weigh advantages and disadvantages of further testing
at LASTI of a LIGO II laser and/or Mode Cleaner. No conclusions.
Mark Beilby/Penn State:
SWG Progress Report, 7/3/2000
Tuning the excitation stage, Aran Glancy (REU student), Mark Beilby
(Post
doc), Jerod Caligiuri (undergraduate student):
Our excitation stage (for exciting the top mount of our pendulum in
a
controlled manner) is now fully operable in all 3-axes. This stage
was
initially unable to move to our specs of 10 microns up to 100 Hz in
the 2
horizontal axes (vertically this spec was met). It was discovered that
a
spacer (inserted by the manufacturer and covered with black shrink
wrap)
between the driving PZT and a ball bearing, which makes contact to
the
surface to be moved, was causing the ball bearing not to touch the
end of
the PZT. After this spacer was removed on all PZTs that had it (the
vertical PZTs had no spacer and ball bearing), the stage now moves
to spec
in the horizontal directions. The stage currently has resonances, which
start at about 15 Hz. Stiffening the components and reducing the mass
of
the stage are now being investigated, so as to push the lowest resonances
above about 100 Hz, which is the approximate frequency up to which
we
would like to make measurements on our pendulum.
Tuning the electronic mirror controllers, Amber Stuver (graduate student),
Gabriela Gonzalez, Glancy:
Tuning of the OEMS controllers, which we have from the MIT PNI, is
now
completed. Damping of the pendulum in all six degrees of freedom has
been
obtained. Preliminary measurements of the pendulum transfer functions
have now begun, while improvements are also being made to the excitation
stage as mentioned above.
Modeling the suspension system, Sean McReynolds (graduate student),
Gonzalez, Beilby:
A Maple model has been completed which gives the transfer functions
in all
six degrees of freedom for a single pendulum hanging by 2 wires from
cantilever blades and perfectly aligned. The mode frequencies obtained
match a previous Matlab model by Gonzalez, along with the measured
mode
frequencies of our pendulum. Work has begun to include misalignments
in
the model, so a prediction can be made for cross-couplings.
Data Acquisition System, Caligiuri, Gonzalez:
Work is continuing on writing software for an improved data acquisition
system. In particular, the DAQ is being setup, so that the computer
can
be used as spectrum analyzer in a sine swept mode for up to eight
channels.
Research in better position sensors: Matthew Knee (REU student), Beilby:
Tests are being conducted using many different types of LEDs in a position
sensor similar to that was used in the MIT PNI, a fin with a slit between
an LED and a split-photodiode. Initial tests are studying the sensitivity
of the sensor to motions perpendicular to the most sensitive direction
as
a function of the beam pattern of the LED. Initial results are
encouraging; our best results show the calibration of the sensor vs.
distance perpendicular the sensitive direction does not change by more
than 2% per mm. These results have been obtained for unlensed or nearly
omni-directional LEDs. The change in calibration vs. distance
perpendicular the sensitive direction is much worse for lensed or more
beamed LEDs, and the curves of position vs. photodiode difference current
are not even linear in some cases. These results improve over a past
preliminary test on one OESM from the MIT PNI whose calibration of
the
sensor vs. distance perpendicular the sensitive direction changes by
about
10% per mm. Next, work will begin on comparing this split-photodiode
sensor to a single photodiode shadow sensor.
Characterizing the seismic environment: Eric Rossi (REU undergraduate),
Gonzalez:
Work is continuing on characterizing the seismic environment and obtaining
a good baseline for the seismic noise. In addition, the seismic noise
is
being studied as a function of time of day and weekday vs. weekend.
Comparing different controller topologies:
The electronic shop in the Physics Dept. at Penn State is nearly done
with
the electronics to control the pendulum via point-to-point.
===================================================================
Brian Lantz/Stanford:
We have been working through the design, cost, and schedule of putting
the
HAM reference design into the ETF.
We have adding the tilt-horizontal coupling of the horizontal inertial
sensors to the model, and are now using that to try and diagnose some
interesting effects of the double active stage at MIT. That model
functionality will then be used optimize the design of the ETF prototype
(with direct carry-over to the reference design).
The work on the hydraulics continues. We have been measuring various
parameters of the actuator, such as the hydraulic resistances of the
various valves and flow restrictions, and calibration of the holding
force
of the actuator. We are in the process of adding both passive filtering
and
active pressure control to the pump.
=============================
Virginio Sannibale: Caltech/SAS
July 5th ,2000
Caltech SAS Seismic Group Short Report.
1- New MGASF: Tuned the 1.5 mm MGASF filter at 220 mHz. Measured the
transfer
function with 1.5mm thick blades. Measurement done between filter body
and
the payload gives 53dB of attenuation a@ 10Hz internal modes frequencies
of
blades are lower than the 2mm blades (49 Hz, 206 Hz, 440 Hz) as expected.
Thermal drift of the working point has been measured more accurately
using
just the day night temperature excursion. With a filter tuned at 230mHz
we
got 1mm/K. Increasing the vertical mode frequency the dz/dT decrease
roughly linearly. A frequency tuning not around the minimum of the
frequency versus working point graph, can significantly reduce the
dependency on the temperature.
Filter body internal modes frequency has been characterized also. The
spectral mode content starts at about 220Hz.
2- Creep Blade Facility almost at the last stage of the installation.
Temperature stability over one day test completed, with reasonable
result (fluctuation of about +-100mK). DAQ systems under test, Position
sensor with ~10nm/sqrt(Hz) sensitivity below 1 Hz under calibration.
Improved the driver electronic noise by a factor 2 and improving the
position sensor geometry, an overall factor 10 could be achieved.
3- IP Threshold feedback and Fine Alignment Actuators (stepping motors):
2channel driver
prototype successfully tested. Communication between the DSP via VME
dedicated
board to the driver will be start soon. Some solution to decrease
the stepping motor noise are under evaluation.
4- Simulation : Program to determine MGASF blade shape analytically
to make the
stress uniform, under test. MGASF simulation to predict the thermal
behavior
under first test also.
5- Accelerometer : second prototype of the folded pendulum accelerometer
is
under manufacturing First prototype of the PCB of accelerometer driver
is
already available.
Vertical accelerometer design based on the MGASF is under evaluation.
6- Control: optimization of the diagonalization procedure for the position
sensor of the IP. Study of the effect of the thermal drift in the
diagonalizaltion procedure. Found problems on measuring the transfer
function with IP tuned below 30mHz probably due to the noise source
generator.
-7 Small IP prototype: started the characterization of the small inverted
pendulum prototype internal mode stability working point etc. Oil
bearing IP shaker, under test.
For additional information about this report, contact sanders@ligo.caltech.edu