e2e LIGO-1 modeling meeting on
June 26, 2001
Contents
High
Priority issues
Summary
of achievements since the last meeting on May 22
Progress
of detector experts support
COS: Mike Smith
SEI/SUS: Mark Barton
COC: Garilynn Billingsley
PSL: Peter King
IOO: Dave Reitze
ISC/IOO: Nergis
PEM: Robert Schofield
Hiro’s priority : Things to be done asap with the assistance of the hardware experts (1+2+3 will enable e2e to generate today's noise curve)
1.
LSC model : Luca is working on it. Luca needs data and various information, and
people's assistance are very much appreciated when Luca needs help.
2.
Mechanical model : Data of the seismic motions at LHO and LLO.
Hiro asked: How can we test the suspended mirror motion, or how well do we know the mirror motion ? Ans: Use LIGO
3.
Laser noise : amplitude and frequency noise of the laser going into the mode
cleaner and same noise at the output of mode cleaner. Beam jitter to the PR
mirror.
- Use MC_F for the output of the PSL (input to the MC); use an Arm
Cavity to best measure what is coming out
- Frequency noise down to 15 Hz is believable; below that, cannot
distinguish from mirror motion. However, those signals still interesting
4.
validation of the modal model : Biplab is working on it. Same as 1. above.
The
following is the Hiro’s summary of achievements since the last meeting on
May 22nd. The updated action item document is available http://www.ligo.caltech.edu/~e2e/actions/e2eActionItems.pdf.
1.
There was a meeting to discuss the direction of the LSC/ASC modeling. The
minute is http://www.ligo.caltech.edu/~e2e/actions/LSC_ASC_05-30-01.html.
Luca
has started working on the LSC modeling, which will include major electronic
noises, digitization, whitening and de-whitening. Andrea is going to work with
Luca to implement the ASC model which will use simple 3D mirrors (based on
Malik's calculation) and WFS.
The mode cleaner will be simulated by a triangular mirror system to simulate the amplitude noise induced by the mode cleaner, but the alignment degree of freedom may not need to be simulated.
2.
There were a few meetings to understand the stack transfer functions, with Mark
Barton and Dennis. Ed Daw has a webpage related to this http://lsuligo.phys.lsu.edu/edaw/homepage/homepage.html. It seems the transfer
function delivered by HYTEC matches reasonably well with the data.
3.
Garilynn will provide reflection and transmission maps by the end of August
with proper orientation (upward direction pointing to z direction). The map
will be used to calculate the mode decomposition matrix. We need to choose a
proper extrapolation of the measured data to the entire mirror.
4.
Based on the Bill Kells study of the thermal effect, the naive idea using an
effective refractive index will be a good starting point, but we need to
understand which quantity can be simulated quantitatively and which only
qualitatively. E.g., the sensitivity could be calculated with good accuracy but
the SB power in Michelson could be off by several 10%.
5.
Andrea has written a program to convert the e2e output to a file with Frame
format. The library is flexible and can be used various other purposes. http://www.ligo.caltech.edu/~avicere/Fbs/doc/html/index.html
6.
Biplab is working to validate the modal model implementation using various
methods. This work will use the real data and FFT simulation.
7.
Dave Reitze has provided us IOO cavity parameters.
8.
Hiro is working on the new release of Han2k package. That will use the latest C
code, the same one used at Hanford for the lock acquisition, and will have
several changes to make it easier to use and maintain. The release schedule has
slipped to the middle of July.
9.
Virginio has a 3d mirror model based on MSE, which gives good results matching
with data. This has not been integrated in the e2e framework yet, but it is very
close. It may be better to adopt it soon in stead of using ad-hoc tricks, like
the bounce mode simulation.
Progress of detector experts support:
COS:
Mike Smith
Scattering:
is there a way for e2e to include scattering from baffles? A detailed model of
the baffles is not useful now, but having a way to put in a net scatter and
motion from an object on 1) SEI table 2) Vac tube wall 3) in-air photodiode
would be useful. P5
Mode
distortion: a model for the way the alignment signals are compromized by
imperfect telescopes and vignetting. P10
UPDATE:
Scattering:
I
have analytical expressions for calculating the light power that is scattered
and/or reflected back into the interferometer mode from each of the principal
external optical paths--baffles, beam dumps photodetectors, glint surfaces,
etc. I also have models for calculating the expected phase noise spectrum and
displacement noise spectrum caused by each external optical path.
Mode
distortion:
I
have an analytical expression which predicts the cross-coupling between the
displacement and angular components of the wavefront sensor signals, as a
function of the amount of astigmatism in the PO optical train. It should be
possible to add the effects of vignetting to this model.
The
principal scattered light noise source is the antisymmetric port photodetector
surface that is directly connected to the ground motion. The scattered light
noise model could be easily validated by vibrating the photodetector with a
known amplitude spectrum and measuring the IFO displacement noise spectrum. A
similar experiment was performed at TAMA to determine the scattered light noise
contributed by the beam tube surface.
I
propose that Seiji and I perform the photodetector experiment at LHO later in
the summer when he is here.
***
Try shaking the periscope as a way to modulate the optical path for all
components on the ISC table
---------------------------------------------------------------------------------------------------------
SEI/SUS:
Mark Barton
a set
of measured transfer functions (ground motion to optic motion) would be the
best point of departure for the combined SEI-SUS system, and calculated
transfer functions for SUS. Once this is implemented, we will be able to judge
if we need more subtlety; if so, it will probably be in SUS and can be built in
later on. With the suspension, either a detailed MSE or a cruder mass and
string model would probably give us a good quality model, and I feel more that
we want to have a model confront the experiment at the level of the suspension
than in the stack -- we are more likely to find a subtle problem in the noise
performance in the suspension than in the stack, and would get help from the
detailed model comparison.
SEI-SUS
measured transfer functions along optical axis: P1
SEI-SUS
measured transfer functions along other axes: P6
SUS
modeling more subtle than 1/f^2: P4
UPDATE:
I've
packaged up my Mathematica model of the suspension and given it to Hiro. It's
not suitable for direct incorporation, but it does calculate many parameters
that may be useful in a more computationally efficient model.
I've
also packaged up the Hytec SEI model and given it to Virginio. I think the
combination of these two models is likely to be adequate for the purposes of
e2e. It's already been determined that the measured stack transfer function for
both HAM and BSC stacks is extremely close to that predicted from the model.
One could get a further slight improvement with the full model that Hytec used
internally, which incorporates frequency dependent damping for the coil
springs. However the full model depends on software which Hytec claims as
proprietary, and high-level negaotiations would be required to get it released.
HIRO:
Hytec XF -- fairly accurate for
stack
---------------------------------------------------------------------------------------------
COC:
Garilynn Billingsley
Lumped
parameters: transmissions, best estimates of losses, absorption, for each
optic. P2
FFT:
optical model of complete interferometer using measured phase maps. P6
UPDATE:
By
the end of August I will provide for each optic of each interferometer
phase
map (with tilt removed)
transmission
loss
estimate
absorption
(for RM, BS and ITM only)
*** Extrapolation
to the edge of the mirror is a challenge
---------------------------------------------------------------------------------------------------
PSL:
Peter King
Control
systems: diagram and parameters P1
Lumped
optical parameters: intensities at several points, mode matching: P2
Characterization:
intensity noise and frequency noise spectra (MC or better yet km arm
measurement) P3
UPDATE:
>>>Control
systems: diagram and parameters P1
This
is available on Rich Abbott's home page. Whether or not it's in a form suitable
for the e2e folks I'll have to find out.
>>>Lumped
optical parameters: intensities at several points, mode
>>>matching:
P2
For
LHO I will ask Rick Savage for these. For LLO I will measure them as we re-lay
the table down here. I hope to have these numbers by the end of June.
>>>Characterization:
intensity noise and frequency noise spectra (MC or
>>>better
yet km arm measurement) P3
Ditto.
The MC or arm spectra for the intensity noise and frequency noise will most
likely be done around mid-to-end of July, depending on when things get up and
going at LLO. For LHO I will have to ask Rick again.
-----------------------------------------------------------------------------------------------------------
IOO:
Dave Reitze
Lumped
parameters: intensities, mode matching, length, finesse, etc. P1
UPDATE:
Here
is a 1st crack at input to the IO e2e modules (see attached 'e2e.pdf'). I've
color-coded in red the parameters that I think are important. As you'll see
from the attached file, I am still gathering some information from people. Also
and more importantly, the rebuilds of the LHO 2k and LLO 4k mode cleaner will
result in new measurements of the MC parameters, but you can take these as a
starting point. Still missing is detailed information about jitter, separation
distances and some MMT mirror data, as well as other things I haven't thought
of.) When you get to a point where you need more info, let me know and we'll
iterate...
and
more recently:
-
provided Hiro with cavity parameters for the IO mode cleaners. Still mining
some data on mirror reflectivities.
-
clarified some data with Hiro concerning jitter suppression.
***
Jitter not easy to measure; telescope suspect; some recent measurements in
ilog.
------------------------------------------------------
ISC/IOO:
Nergis
IO:
length and alignment control systems diagram and parameters P2
LSC:
control systems diagram and parameters (operational and locking) P4
ASC:
control systems diagram and parameters P6
UPDATE:
I
have been working with Luca at the observatory to make sure he has access to
the most current transfer functions for the LSC loops. He has also begun
measuring electronics noise levels in the (analog) LSC modules on the LHO 2k.
We held a meeting 05/22/01 to lay out a plan. There we decided that the LSC was
the right place to start, the ASC would follow. There was a discussion about
A2D and D2A model but we can proceed with a measured level for now since we are
not quite at the theoretical limit for our ADCs.
-------------------------------------------------------------
PEM:
Robert Schofield
First
check with Matt Evans and other e2e to see what has already been done, formats,
etc. in integrating the seismic environment into the locking modeling
Seismic
noise ('normal' conditions) from each seismometer in all DOF P1
Seismic
noise at several representative accelerometers on each test mass support beam
P3
Statistics
of seismic noise P5
Acoustic
spectra next to PSL and relationship to PSL frequency P6
UPDATE:
Integration
of environmental effects into the e2e model.
At
David's suggestion, I have contacted Matt Evans and am learning what has been
done to include seismic noise in the locking modeling.
In
conjuction with another project, I am currently putting together seismic and
other environmental noise spectra that should be useful for the e2e project.
David
S. and I have discussed first steps in compiling and measuring environment to
interferometer transfer functions.
I
hope to make some progress in this effort during the next month.