Detector Watch 5 ( Revived! )

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,578
Credit: 306,914,202
RAC: 178,279
Topic 192563

Well I've finally found some time to attempt a revival of this thread(s). :-)

It is a continuation of this one.

Looking at the Livingston logs there is an occasional mention of 'Kleine Welle'. This means 'wavelets' in English, I think first mentioned as such in the setting of optics theory by either/or Young/Huygens - when referring to re-radiation of light waves by atoms, the summation of which giving rise to a new wavefront.

My reading seems to indicate that for LIGO the idea is to examine the data from the detectors with respect to a model that has 18 ( yes, eighteen ) parameters! If an excess of energy arrives within a given time period in a particular frequency band then this may be characterised. This is an attempt at modelling the 'burst' sources of gravity waves. This is incorporated into the Data Monitor Tool ( DMT ) aspect of operations which deals with online/realtime data ( ~ < 10 seconds old ). Presumably then this comment from Monday 02/04/07:

Quote:
'The KleineWelle glitch monitor was lagging behind by 4-5 hours in updating data.'

indicates some laggard behaviour...

The Solomon Islands' earthquake ( with sadly a tsunami ) had knocked the IFO's out for a number of hours ( + aftershocks ), it was an 8.0 on the Richter.

Mr Dan Hoak is always good for a quote:

Quote:
....He was using the Hilbert-Huang transform to pick out noise in different frequency bands. Here, I'm using the Hoak-Hilbert-Huang method...

:-)

This refers to an application of a method, distinct from the ( Fast ) Fourier and Kleine Welle stuff, to examine the data for 'non-linear, transient signals'

There was also a Hindu-Kush earthquake ( 6.2 ) on 02/04 also:

which looks to be a USGS plot, the contours showing propagation times of the P-waves ( pressure ) and the 'antipodeal shadow' which is characteristic of earthquakes.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,578
Credit: 306,914,202
RAC: 178,279

Detector Watch 5 ( Revived! )

Livingston:

Now here's such a nice image I've put it on my desktop. :-)

which Dan Hoak at Livingston produced by taking a line by line scan using the photodiode reading at one of the beam examination/takeoff points ( AS1 in this case ) using a mirror there that can be rotated in pitch and yaw axes. It ought to be a circular image but is oblong as probably voltage/angle bias is not the same ratio for each axis.

Earthquakes, deliveries, mowers, trains, 1st graders and their buses contributed to lock loss. Network outage from a backhoe blade.

For those that can make the sense of it, there are entries relating to Perl scripts and Matlab too..

I think I'll look up those 'violin modes' that are so frequently mentioned and get back to you. Alternatively, Chipper might .... :-)

Hanford:

Now here's some more nice images, showing pleasing trends at H1:

Note H1 has touched 16 Mpc on occasions!

and H2:

The Thermal Compensation System for H1 - which helps the end test mass mirrors maintain optimal optical shape by selective warming of mirror areas using a separately aimed CO2 laser - is playing up. Thought to be either a hardware ( 'frozen' motor controller ) or a software ( Bad Burt Restore? ) problem that was resolved by a reboot or two. Gee, my home computer works just like that too ... :-)

Earthquakes of various sites and sizes impacted on lock.

Some other computer issues:

Quote:
Since h2susepics is recently failing in a new and wonderful way, with the process not gracefully terminating, but lurking in the background. In this case, if a new process is started, then after the burt restore the medm screens start flashing their LEDs. The new wiki entry explains how to check the old process is no longer running, and to kill it if it is still there

This is reminiscent of Windoze too ... :-)

Cheers, Mike.

( 09/04/07 - edited for typos )

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Chipper Q
Chipper Q
Joined: 20 Feb 05
Posts: 1,540
Credit: 708,571
RAC: 0

RE: I think I'll look up

Quote:
I think I'll look up those 'violin modes' that are so frequently mentioned and get back to you. Alternatively, Chipper might .... :-)


Off the top of my head, the violin modes are related to the suspension wires that are used to hold the optics in position. Just like the strings on a violin, there will be sets of frequencies that are resonant with a specific wire (including a fundamental frequency, and then also all the harmonics of that fundamental), depending mostly on the length of the wire (string length), but also on the materials that the wire is made of (which define the variables for string stiffness and string mass), and the manner in which it's affixed (defining the string tension).

So the suspension wires in the LIGOs, stretched tightly from positioning the optics (including the Beam Splitter and End Test Masses [and others]), will have unique resonant frequencies in the same way that violins (or any stringed instruments) have. But what does this mean for the scientists? Well, the suspension wires serve two very important purposes: they not only help to isolate the optics from external vibrations, they are also required to align the laser beam. So knowing the violin mode of a specific wire is necessary not only for determining the proper control signals to use to align the beam or dampen vibration, but special attention is required to see that additional noise isn't being introduced (coupled?) via those unique resonant frequencies from any other sources in the environment; and considering all the harmonics, that's a lot of bases to cover.

In addition to the violin modes, there are also pendulum modes to contend with, since the optics are suspended...

(Thanks for the nudge also, Mike. I hope the above is right, and in all events it's good way for me to be sure I'm learning what I'm checking; most grateful for the opportunity!)

Chipper Q
Chipper Q
Joined: 20 Feb 05
Posts: 1,540
Credit: 708,571
RAC: 0

If you do a Google search

If you do a Google search using "ligo violin mode", you can see right away that there's been a great deal of work regarding the testing of different materials needed to meet the requirements of supporting the mirrors (also referred to as test masses, or optics) in the interferometer of a gravitational wave observatory.

For Initial LIGO, materials like tungsten and niobium were tested to see which material made the best wire. For Advanced LIGO, the testing included different shapes for the wire, like using a bar shape, and a ribbon shape. The material for the ribbon is fused silica (which turned out to work better than everything else that was evaluated, if I'm not mistaken).

Now, these violin modes, these natural resonances (like a string tuned to play a specific note) that arise from the type of material, the unique dimensions of it, and from the tension placed on it, result in an unwanted musical instrument of sorts, but with a profound difference - if you pluck a guitar string, it will keep on vibrating, and you can continue to hear the sound for some period of time after the pluck, and it takes a while for the sound to fade (a lot longer if it's an electric guitar!).

So if the suspension of the Advanced LIGO should get accidentally plucked, which could happen if the mirror bumps against a stop during an earthquake, or (as has been most often the case with GEO 600) it can happen if the alignment control is driven at frequencies around the violin modes, for how long will the plucked "note" sustain?

Well, the plucking is technically referred to as excitation, and the sustain of the note is known as the decay time: the note could last as long as 3 months (!), quite different from any guitar :)

Quote:
The natural decay time (constant) of the violin modes could be as high as ~3 x 10^9 cycles, 8 x 10^6 s or 3 months (although it will probably be much shorter due to inevitable coupling with lossy parts of the suspension). ...
Assuming that excitation of the ribbons will be a reasonable rare event (daily or less frequent), the tolerable decay time is as much as a few hundred seconds ... This fits well with experience on GEO 600 where ~ 600 Hz violin modes ... affect detector performance for less than one hour after considerable excitation.

(quoted from here, pp. 2-3)

I recalled seeing a picture of the End Test Mass for the y-arm of H2 at Hanford in the elogs (see Corey Gray's entry on 12/07/06) where one of the suspension wires is visible; the whole picture is over 3 MB (too large for ImageShack) so I copied & pasted from it to make these (I added the arrow pointing at the wire in the 2nd picture; click the thumbnails to see the full size pictures) -

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,578
Credit: 306,914,202
RAC: 178,279

Thank you very much Chipper!

Thank you very much Chipper! :-)

Yep, I went and looked up a violin web site or two. Alot of interesting physics and maths from such a simple thing as a string tied at it's ends. It's important for LIGO because:

- you're gonna get them, there are strings with tied ends ( or equivalents ) aplenty.

- as noted they can store/carry energy long into the future. For instance, since the mirror suspensions are enclosed in vacuum then it won't be friction with air that settles them down. It'll be vibration into the mounts plus electro-magnetic effects.

- 'coupling' is the term that indicates energy flow to/from the wires. The last time I studied this in detail ( 20+ yr ago ), the maths was hard for even simple cases....

I've been some what flumoxed trying to follow some recent attempts by a team at Livingstone as regards what appears to be a new technique for bringing the IFO into lock. From earlier today :

Quote:
We have locked the IFO on DC light using the pOMC. Best ever DC-locked IFO inspiral range: 250 kPC. Next stop, Andromeda.


DC means 'direct current', or a static test if you like. It is not a science mode scenario, the description of efforts is entertaining ( well it sounds like fun to me ), and here's a nice pic ( from the Anti-Symmetric port Photo-Diode ):

a pretty flower arrangement ( six fold? ) produced by resonance between a given frequency and multiples. Who says you can't get art from science! :-)

Hanford, both H1 and H2, has had some particularly awesome ( 90% + ) duty cycles, and an unusual cause of lock failing:

Quote:

The HAMMER facility called at about 7:05 am, to inform us that they were going to explode a bomb before 7:30 am, did not ask how big the bomb was.

The effect was an AS5 trigger event on H1, ending segment H1-3252, H2 was able to ride through the noise generated by the bomb.

'Hammer' ( Hazardous Materials Management and Emergency Response Training & Education Center ) is nearby. Also nearby is Energy Northwest ( a nuclear power station ) which has large cooling towers with fans:

Speaking of energy, it's an oft mentioned phrase 'noise budget' - this means to what mode/area/source do we attribute energy within the arms? Most of it is 'noise': meaning not the signal of interest, but the spacetime deflections are within the hub-hub ( it's E@H's job to tease it out ):

Here you can see a legend on the upper right showing who's who. Like any accounting you don't want to miss/lose anything nor double-count either. :-)

Here's some representative Figure's Of Merit showing good range ( ~15Mpc for H1, ~7Mpc for H2 ), excellent triple co-incidence, and beautiful sensitivity:

They are really reeling in huge slabs of high quality data that we at E@H will be working on for years! :-)

Now here's a better version of the beam:

- it's these photons there that are telling us about the bending of spacetime, LIGO converts them to an electrical current ( photodiode ), thence to numbers ( analog to digital conversion ) and into the raw data files.

Cheers, Mike.

(edit) Replaced the flower with a better one. Also the Energy Northwest plot shows a peak at ~ 2.3 Hz - probably the 'thrum' of fans.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Dan G.
Dan G.
Joined: 10 Feb 07
Posts: 39
Credit: 97,382
RAC: 0

Hi, it seems the link you

Hi, it seems the link you provided:

http://www.ligo.org/pdf_public/camp02.pdf

is password protected. Can you tell me where I can download this document? Thank you!

Cheers,

_dan

Chipper Q
Chipper Q
Joined: 20 Feb 05
Posts: 1,540
Credit: 708,571
RAC: 0

RE: Hi, it seems the link

Message 61790 in response to message 61789

Quote:

Hi, it seems the link you provided:

http://www.ligo.org/pdf_public/camp02.pdf

is password protected. Can you tell me where I can download this document? Thank you!

Cheers,

_dan


It looks like all the papers posted for comment at the LSC have been password protected. I was able to find the one you're asking about here...

Dan G.
Dan G.
Joined: 10 Feb 07
Posts: 39
Credit: 97,382
RAC: 0

Hey, thanks!! I needed that

Message 61791 in response to message 61790

Hey, thanks!! I needed that paper.

Cheers,

_dan

Quote:
Quote:

Hi, it seems the link you provided:

http://www.ligo.org/pdf_public/camp02.pdf

is password protected. Can you tell me where I can download this document? Thank you!

Cheers,

_dan


It looks like all the papers posted for comment at the LSC have been password protected. I was able to find the one you're asking about here...


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.