Detector Watch

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

RE: Reality check

Quote:
Reality check please..... :-)


I think your first analysis was correct. The log entry you quoted had a UTC-stamp of 20/10 07:09:54, and it was in the 'LHO Eve Shift Summary'. So we need to see the State Vector plot with T0 >= 20/10 07:09:54 UTC to get the data that includes that Evening Shift, and the most recent one (when I checked) was T0 = 20/10 00:49:20, which includes only about the first couple of hours of the Evening Shift at Hanford. I hope I figured that right; bit of a lag between the shift summaries, and production of the Figures of Merit?

Interesting to note that there was also a GRB mentioned in both sets of logs. The Swift satellite detected it at 19/10 04:19:06 UTC, but I haven't seen a burst description posted (at the GRB Real-time Sky Map) yet.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,051
Credit: 109,287,102
RAC: 52,158

RE: I think your first

Message 47947 in response to message 47946

Quote:

I think your first analysis was correct. The log entry you quoted had a UTC-stamp of 20/10 07:09:54, and it was in the 'LHO Eve Shift Summary'. So we need to see the State Vector plot with T0 >= 20/10 07:09:54 UTC to get the data that includes that Evening Shift, and the most recent one (when I checked) was T0 = 20/10 00:49:20, which includes only about the first couple of hours of the Evening Shift at Hanford. I hope I figured that right; bit of a lag between the shift summaries, and production of the Figures of Merit?

Interesting to note that there was also a GRB mentioned in both sets of logs. The Swift satellite detected it at 19/10 04:19:06 UTC, but I haven't seen a burst description posted (at the GRB Real-time Sky Map) yet.

Hmmmm... sigh.... now I've just been reading this, & I found this:

Quote:

Interpreting the BLRMS Figure of Merit at LHO

Fatter lines are lower frequencies, frequencies are grouped by thickness and color

Red: 0.1 to 0.3 - microseismic band, also shows earthquakes a little
Black: 1 to 3 Hz - distant anthropogenic
dashed: EX should pick up vitrification plant work best
solid: EY should pick up city best, also, trucks bleed through from higher band
Blue: 3 to 10 Hz - trucks, explosions at YFC, close antropogenic
dashed: LVEA
solid: Y-end - almost all spikes are trucks on 240
Magenta: 10 - 30 - motors, local events - normally you cant see these, but if something bad happens they stick up

NOTE: the non-physical event (seismic activity can spike up, but not down) at -6.7 is a DMT glitch
NOTE: T0 is actually T at -12

( My bold emphasis ..... )

Cheers, Mike.

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

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

I recall reading that list at

I recall reading that list at least 3 different times, and somehow never made the connection in the note you emphasized. So are the additions I made to this State Vector plot correct? Reality check check please! :)
[img]http://groups.msn.com/_Secure/0RQAAAAkV*uUzp2CptTBO0yi6DYjZCb1XaY*0yESzpHcWiy3BE*zNCV8eChYFI*dvkVgaPSPuuUcFLKJikkhJ0H1WStn7SO9vZVnwdUlP*CQ/StateVec.gif?dc=4675594411539957356[/img]
That's assuming the Evening Shift is from 4pm to midnight - I think the shifts are longer than 8 hours, and overlap, but hopefully that's the 6-hour block in question.

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

Another GRB mentioned in the

Another GRB mentioned in the logs on the 21st. Been reading some of the links Mike posted about them. They are detected at a rate of not quite 1 per day, but there are estimates that for every GRB that's detected, there are 450-500 that occur without being detected. Since gravitational waves are thought to be associated with both types of GRB (short or long duration), the ratio of 450:1 increases the odds for the number of potential signals that could happen within the range of the LIGOs.

The search for GW signals in the LIGO data can be concentrated near the times when these GRB triggers occur. The benefit of this type of search ('triggered') is that not all of the LIGOs need to be in Science Mode. But for 'untriggered' searches it's important for more than one of them to be in SM. They had a good Owl Shift on the 21st at Hanford (log entry from 21/10, just after 8am local):

Quote:
H1 was in SM the entire shift (yay!).
. . .
H2 almost made it, but fell out of lock in the last 20 minutes.
Returning to SM was done on the second attempt.

@Mike: Looking forward to your next 'closer look' in the detector watch, and eager to assist with any data-mining or reverse-engineering! :)

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,051
Credit: 109,287,102
RAC: 52,158

RE: Another GRB mentioned

Message 47950 in response to message 47949

Quote:

Another GRB mentioned in the logs on the 21st. Been reading some of the links Mike posted about them. They are detected at a rate of not quite 1 per day, but there are estimates that for every GRB that's detected, there are 450-500 that occur without being detected. Since gravitational waves are thought to be associated with both types of GRB (short or long duration), the ratio of 450:1 increases the odds for the number of potential signals that could happen within the range of the LIGOs.

The search for GW signals in the LIGO data can be concentrated near the times when these GRB triggers occur. The benefit of this type of search ('triggered') is that not all of the LIGOs need to be in Science Mode. But for 'untriggered' searches it's important for more than one of them to be in SM. They had a good Owl Shift on the 21st at Hanford (log entry from 21/10, just after 8am local):

Quote:
H1 was in SM the entire shift (yay!).
. . .
H2 almost made it, but fell out of lock in the last 20 minutes.
Returning to SM was done on the second attempt.

@Mike: Looking forward to your next 'closer look' in the detector watch, and eager to assist with any data-mining or reverse-engineering! :)

Well ( so none of us would die wondering! :-) ) I contacted Bruce Allen - who very kindly confirmed for me that the T0 does in fact refer to the LEFT-hand side of the horizontal/time axis where it is marked '-12'. So this analysis is indeed the correct one. Hence the '0' mark on the RIGHT-hand side of the horizontal/time axis is in fact at T0 + 12 hours.
Phew!
So now we can all correlate the data confidently in the time domain! ( I'd been having trouble with that, which led to my doubts you see ... )

They are indeed having a much better run at all the LIGO's, higher duty cycles ( % of time in science mode ) with good and long triple co-incidence periods ( when all three are simultaneously locked ).

The log is revealing about the GRB's ( Livingstone 21/10/2006 ):

Quote:

1539 We have had a GRB Event Notification, and will disallow injections
for an hour.

1541 Another GRB Event Notification.

1558 Another GRB Event Notification, but this is apparently a duplication
of the previous notification.

1607 The AS_TRIGGER tripped and took the IFO out of Science Mode.
There was no obvious seismic cause.

1608 The trigger has been reset and the IFO is back in Science Mode.
This is segment S5-03451.

1640 The GRB Stand-Down period has ended, and we are allowing
injections again.

This would strongly suggest a main purpose of the GRB alarm line is to veto injections. So what's that about then? .... well, lets have a closer look :-) ( you twisted my arm Chipper! ):

The interferometers are very complicated gadgets, with the aim of examining the heavens, not the Earth. Alot of the time they do seem to function as ( fairly expensive ) seismographs, train detectors, anthropic monitors and weather stations! In other words noise - generally defined as the wiggles on the output of the detectors which are not of interest to astronomy.

Let's get down to photon level and take a trip through a LIGO.

- we are born in the NDYag laser which powers the whole interferometer, light wise. This is a solid state device which, when suitably encouraged, will allow emission of monochromatic photons - of a single frequency/wavelength/energy. It is sort of like graduating from a school where everybody is taught the same, and leave with the same qualificiations ( or are kicked out if not ). There is in fact some spread in these characteristics, as the 'purity' is limited by many factors, but ultimately by the Uncertainty Principle. The wavelength is 1.062 um ( ie. about a millionth of a metre ) by the way - in the infrared range.

- we progress to the 'mode cleaner', an arrangement of mirrors that trims the pack of photons a bit ( I've yet to delve deeply here ), and yields a more exactly defined crew emerging. A 'polishing' process.

- now to the beam splitter which, as the name suggests, flicks some photons up one arm and the rest up the other.

- let us say I go up the Y-arm, eventually to hit the mirror at the End Test Mass ( ETMY ) some 4km away and then return.

- upon return ( to the corner station area or LVEA ) I will meet some of my 'siblings in arms' ( chuckle/nudge ) and compare notes about our various journeys.

**** Each of us photons are in fact particles - we are whole lumps ( never fractions ) and if travelling together arrive in some integral number of lumps. But we each have a unique 'internal' timer which oscillates. Visualise it like a watch with a single hand that goes around and around endlessly, and the number of times per second it completes a circuit is the photon's frequency. This is called the phase of the photon. It is extremely important that all us photons start our journey synchronised in phase ( just like in the war movies when they synchronise watches before the mission ), and have the same intrinsic 'rate' at which our watches tick. The wavelength is in fact the distance they will travel during a single full circuit of the sweeping hand of their watches. If you multiply the frequency by the wavelength you get the speed of light! ****

- now some of the photons are going to travel in such a way as to not accumulate the same change in phase as others! Some will have spun their wrist-watch-hands around by a lesser or a greater amount - maybe differing by a whole number of circuits, fractions of a circuit, or whatever.

- when they hit the detectors ( eventually producing an electrical signal which we keep as data for analysis ) then the response of the detector depends on the combination of the phases from many photons. If two photons return to LVEA with synchronous phase they add to the response, if they return to LVEA with anti-synchronous phase ( watch hands diametrically opposite ) then they subtract from the response. Hence the effect of any one photon is amalgamated with that of others - this is called interference and devices which exhibit this are called interferometers.

**** Strictly speaking interference would occur even if each photon travelled entirely alone, but we'll skip the spookier quantum stuff today..... :-) ****

So, you may ask, what affects the change/progression of the phase of a photon as it traverses the LIGO's? Now for some relativity ( gasp! ).....

- basically there's a thing called the metric. This is an all purpose ruler for surveying the spacetime neighbourhood, recording events etc..... ( it is actually a set of numbers combined in a special way ).

- light has the special property that when we use this ( Minkowski ) metric, a quantity called the 'proper time' is always zero. Light is said to follow the 'null' geodesic/path. This is an accomodation of the comments that 'light travels in straight lines' or 'light takes the path of least time' that you may have heard.

- if a pure Minkowski/flat metric only applied to the LIGO's we wouldn't have much to talk about ... interesting lights, pretty machines, shiny floors etc... :-)

- however, distant and unbelievably violent events ( fortunately for us not too close ) produce an alteration of spacetime which propagates to us ( a 'gravity wave' ), and alters the metric hereabouts. But even if the metric alters, light still goes via the null geodesic/path in spacetime, and that accomodates that metric change. For many different waveforms, directions, polarities etc of the gravity waves there will be a detectable difference in the accumulated phase change between the X-arm and the Y-arm of an interferometer due to the fact that the metric alteration is ( slightly ) different between the arms.

Asleep yet?

Finally returning to the injections. These are deliberate, but well honed, disturbances to the interferometer initiated by the LIGO researchers so that it's response to said nudges can be studied in fine detail. This enables proper characterization of the non-gravity wave factors that cause the photons to be measured as having a phase differential. Hence this injection program/process helps eliminate false alarms ( ie. we thought we found a gravity wave but there wasn't actually one ) and raises the odds of producing solid detections ( by suggesting design or implementation changes to minimise noise ).

So by veto-ing/disallowing any of these tests/injections during the period of time we are hoping to detect the gravity ripples from a Gamma Ray Burst ( GRB ) source, we don't 'pollute' the chunk of that part of the data set. Otherwise it would be a bit like recording over a favourite TV show.

Keep in mind that many, if not most/all, of us have likely crunched work units containing injections. Think of it as an ongoing quality control regime which has the aim of ensuring confidence in the whole LIGO program.

Whoa there Mike! That'll do ya...

( See what happens on my Sundays... thoughts turn to E@H )

Cheers, Mike.

( edit ) If you see any acronym containing 'DARM' or similiar it means 'differential arm' - ultimately it would refer to this phase comparison as above. Also 'burp' or 'burping' refers to opening up an interferometer segment for inspection/servicing, where the vacuum is interrupted. Sort of 'passing wind' perhaps ...... could think of worse monikers .... :-)

( edit ) There's a bounty of other outcomes for our photon friends - absorbtion for instance - that I've omitted.

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

AnRM
AnRM
Joined: 9 Feb 05
Posts: 213
Credit: 4,346,941
RAC: 0

Great work guys....much

Great work guys....much appreciated....Cheers, Rog.

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

RE: So now we can all

Quote:
So now we can all correlate the data confidently in the time domain!


I actually missed an hour of one of the World Cup games (error somewhere in the time-zone differences between East/West coasts, London, and Germany), so my confidence level basically looks like the LIGO microseismic plots - a little shaky! :)

It occurred to me that the issue of timing has a several good possibilities for future "Mike's Closer Look" topics. It's fairly straightforward calculating offsets like differences in geographic time-zones. But consider how much harder is the scientist's work:
-Assuring the accuracy of timing between the LIGOs in order to achieve 'triple-coincidence',
-Achieving long intervals of time in 'triple-coincidence' (ideally a whole year, for S5),
-Accounting for an offset in time between the 24-hour UTC clock, and motion of the celestial sphere (just like the E@H screensaver) due to things like the Earth's rotation, tilted axis, and elliptical orbit) whose value changes each day (well, constantly). This means that objects of interest to LIGO scientists rise and set in the sky (on the celestial sphere) a few minutes earlier (or later) each day, depending on which day of the year it is. An approximation of the offset, over the course of a year (length of S5), looks like this:
[img]http://groups.msn.com/_Secure/0SwDNAnQXPJOpByaGcnzLw7HU0xc2GTBI0ssePbCLwSg0ExnUM!Ffj**w9BDPSbJdcmUyl5I!uHNnQpyKCpQwKg7hCqzgD9EPKjhQih*wihcus6A4PVuZeQ/EquationOfTime.gif?dc=4675594713455122469[/img]
(Plot mined from Wiki's Equation of time) :)
- - -
They had to replace an optical lever in Hanford's H2 Y-arm (FMY). Not sure yet if 'oplevs' are for just steering the beam, or more primarily to isolate (or dampen) the optical elements (several in each arm) from seismic. But I discovered how to calibrate them, after a little diggin :)

Quote:
So by veto-ing/disallowing any of these tests/injections during the period of time we are hoping to detect the gravity ripples from a Gamma Ray Burst ( GRB ) source, we don't 'pollute' the chunk of that part of the data set. Otherwise it would be a bit like recording over a favourite TV show.


You've a keen eye for the detail, Mike, and the injections with their possible veto makes sense.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,051
Credit: 109,287,102
RAC: 52,158

Good stuff, Chipper. Yes I am

Good stuff, Chipper. Yes I am preparing a time discussion....

H2 is on the fiddle again, ETMY ( end test mass Y arm ) suspect - this time without any evident seismic jolt. This is a good example to discuss:

Horizontal axis is the frequency in logarithmic scale - meaning equal distances along the scale refer to equal multiples - going from 10 Hz to 10,000 Hz. This way of viewing helps keep the data visible on the page without having a really wide monitor, or scrolling. :-)

The left side vertical axis is our IFO signal, but indicating how much movement there is for each frequency value. This scale is also logarithmic, but note those negative powers of ten ( which are the reciprocals/inverse of positive powers of ten ). Ten to the power of twenty is a really, really big number and thus ten to the power of minus twenty is a really, really small one. This is the level of sensitivity required for detecting gravity waves!

Think of it like the tuner on a radio, so for instance those black spikes coming up off the black fuzzy curve would be like radio stations. Imagine turning the dial and suddenly hearing an increase in the sound from the speaker at those peaks.

Now there are lots of contributors here. The upper right legend give the acronyms for various sources. Each contribution is coloured differently. A few examples:

DARM - differential arm ( the science data )
Seismic - ground movements
SusTherm - thermal vibrations from the gear suspending the mirrors
Shot - photons arriving in bunches rather than perfectly regularly
Goal - the original design specification ( called SRD elsewhere )

Basically the bad boy is the black fuzzy line ( starting up top under the 'z' of Hz ) sliding down towards the lower right, then turning upwards and more gradually rising to the middle of the right side. The range, at 1.2 Mpc, is poor - 6 to 7 is the expected/good value.

The ETM's are well studied, in that by design and observation their response to disturbance is known. One method of teasing out the source of problems is to examine how they swing freely - that is without any influence by the usual damping and other machinery. They have a pendulum action with roughly one Hertz as the natural frequency they want to oscillate at, in various modes. By comparing current swinging rhythms with reference measurements ( presumably made when things were going well ) then some deduction as to the cause of the problem may be made. If mosquitos could survive in a vacuum then it could be as little as one of them landing in the wrong spot to cause a problem!

Cheers, Mike.

( edit ) Also the fundamental limit on measurement, Heisenberg's Uncertainty Principle, as it applies to this device isn't 'far' away here - about 10 to the power of minus 23 I think.

( edit ) Sorry, I should have emphasised that it is the part of the black curve to the left of it's turning/lowest point which is the bad bit. It should flop down against/near the Goal curve - which it does match fairly well to the right of the low point.

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

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

RE: H2 is on the fiddle

Quote:
H2 is on the fiddle again, ETMY ( end test mass Y arm ) suspect - this time without any evident seismic jolt...
...The ETM's are well studied, in that by design and observation their response to disturbance is known. One method of teasing out the source of problems is to examine how they swing freely - that is without any influence by the usual damping and other machinery. They have a pendulum action with roughly one Hertz as the natural frequency they want to oscillate at, in various modes. By comparing current swinging rhythms with reference measurements ( presumably made when things were going well ) then some deduction as to the cause of the problem may be made. If mosquitos could survive in a vacuum then it could be as little as one of them landing in the wrong spot to cause a problem!


After not finding any obvious reasons for the lock loss and subsequent drop in range to 1.2 Mpc, that's exactly the kind of teasing they tried (during the 'late aft/early owl shift', on ETMy), along with some nudging, referred to in the log as 'Exciting the Optic', and it did the trick! Back up to ~6.5 Mpc in a matter of minutes! Less than a mosquito, perhaps a microscopic dust particle wedged in the works?

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,051
Credit: 109,287,102
RAC: 52,158

RE: Back up to ~6.5 Mpc in

Message 47955 in response to message 47954

Quote:
Back up to ~6.5 Mpc in a matter of minutes! Less than a mosquito, perhaps a microscopic dust particle wedged in the works?


Yeah.... I was thinking of an Intrepid Mozzie Duo - Neil Strongarm & Buzz Dieldrin - finally landing on the Big Shiny Lump, only to be bumped off before they could say 'the Ornithopter has landed'.... :-)

If you compare this the blue trace with yesterdays black one, then you can see how DARM has fallen back towards the design curve ( SRD ).

You may be wondering why they don't just go in and casually manhandle the mirror etc. It takes ages to realign and re-settle the whole array - so that may be a backward step, not to mention the risk of introducing some new error. However, see as follows:

Quote:
Whatever it is, again the only two possibilities are the driver electronics chain or the ETMY/cage itself.


Quote:
The driver electronics chain failure is hard to imagine (but it's possible to think of some conspiracy).


Quote:
will be a vent of the 4k vertex section up to ~5 Torr in an attempt to discharge any charge on the 4k ITMs


Quote:
As for the 2k ETMy 2Mpc-6Mpc bi-polar issue, there are plans to go in and look at ETMy in a few weeks while the 4k is down for ISCT4 table floating work. The idea is that we will find something wrong with the suspension which will be an easy fix (loose stop tip, etc.). We will probably take this opportunity to replace the stops on this suspension with the Enhanced LIGO version of stops. TBD.


Quote:
We're still not committed to full venting MY


Now here's an interesting interaction between H1 and H2, due to the fact they share the same tunnel:

Quote:
2k IR is bad because the 4k suspensions are swinging around, flashing 2k light back into the 4k.


IR is infrared, as the lasers are in that range ( 1.064 um ) so this is not visible.

and a helpful warning too:

Quote:
TC Olympic Railway called this morning to report that there would be a few hours of train traffic around Energy NW at 10am. They report 75 cars and a few 3 locomotives.


which I guess we'll see on the seismics.

Cheers, Mike.

( edit ) My apologies for not mentioning Livingstone who, despite suffering from trains, construction and earthquakes, are back at 7Watt laser power for the full 13 or so Mpc. Wasn't there a movie called Planes, Trains & Automobiles? Steve Martin & John Candy?

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

Comment viewing options

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