Slashdot Mirror
Gravitational Wave Detection Imminent?
Seumas Hyslop writes "The UK Telegraph is reporting that we may finally have equipment that are sensitive enough to measure gravitational waves, which are incredibly small and have evaded detection despite the theories that they are present as a way of explaining gravitational effects. Basically, a laser beam is split into two branches that are sent down two identical 2000 feet long tubes and back again via mirrors. Assuming the two arms remain exactly the same distance, they will cancel each other out. But the scientists think that the beams will interfere with each other owing to the effect of gravity, meaning the length of the branches is altered and a gravitational wave has been detected."
Bring out your gravity surfboard and roll on!
bash$
Maybe the gravitational waves changed the i into an a.
i won't even get into it.
anyways, the purpose of the interferometer is to measure the differential gravitational strain between two remote masses. as a gravity wave passes (supposedly), two masses will be driven to oscillate in quadrature with one another. that means that relative to some fixed point, one mass will be drawn closer, and at a right angle another mass will be pushed further away. IIRC.
luckily a michelson interferometer is a great way to detect these small changes, where the remote masses are mirrors. the extremely long beam paths increase the sensitity of the device. and two remote locations are needed for local error cancellation. if you have three locations (there is a LIGO opening in louisiana soon. uh, maybe) then you can actually do gravitiational wave astronomy.
probably some LIGO person will write a better explanation, but it's late.
m
Gravitational Radiation - the cosmological reference, not the meteorology ones.
Some other gravitational wave detection projects
Some anomalies in gravity theory
and, of course, Einstein@Home
Mongrel News all the news that fits and froths
Both sites are asking for public help processing the data, via a special screensaver called Einstein@Home.
--Greg
The interferometric GW detection systems have been under development for quite a while. These include the LIGO project in the US, the GEO in the UK/Germany, and Australia and I believe Japan and Italy have their own versions. LIGO started collecting data a couple of years ago. So now the guys in the UK turned on their instrument.
So what's the big deal?.. Well, there isn't one. Today's instruments are pretty damn bad. I don't remember the numbers, but you'd have to run them for quite a few decades in a row for a good chance to observe one event (it would have to be something big falling into a black hole somewhere relatively close to us, or a major supernova, or something equally rare.) Essentially, you are trying to measure a ludicriously small displacement (10^-16 cm) of a macroscopic object.
The good thing is, technology is continuing to improve, increasing the sensitivity. Furthermore, there's hope (subject to funding) of creating a space-based version of the experiment by bouncing laser beams between three satellites millions of kilometers apart. So is the GW detection imminent?.. Considering the scale and cost of the projects, it better be, but I (being a scientist and all) prefer to steer clear of that word. So provided the funding doesn't get cut, we'll very likely detect gravitational waves in a few years. But be prepared to wait.
For more deets, check out www.ligo.caltech.edu
i will try again here. so one case of the "gravitational wave" theory is that when two black holes spiral around one another (or any two large masses), they will emit energy in the form of gravitational waves, like two boats circling in a lake. physicists would like to detect this energy.
let me digress for a second to radio. normal EM radiation is in the dipole form. which means the radiation makes charges (electrons in an antenna) oscillate up and down. gravitational waves (i think) hit us in the higher order quadrupole mode, which instead of "up and down" is more like "in and out". or taking a circle and squishing it along one axis, and then the other.
so lets say you are standing on a field. then you have two stones hanging on strings, one 100m north, and the other 100m east. when a gravitational wave passes, if you were God, you would be able to notice that the north stone was pushed closer while the east one was pushed away, then the east one was pulled toward you while the north one was pushed away.
to detect this *infinitesimally* small force, you replace the rocks with mirrors. and put the mirrors in vacuum to prevent them being jittered by air molecules and strange index of refraction effects with the air. then put the mirrors really far apart to increase the relative sensitivity to the same strain.
now take a laser beam, split it where you stand and send half the beam to each mirror. the beam then returns to you, you recombine it at the same beamsplitter, and the photons in the laser beam will interfere. whether this interference is *bright* or *dark* depends on the relative path length difference of the two arms.
you can detect changes on the order of 1/100 wavelength (actually, much less, but that's more complicated) which is about 1e-8 meters. since the interferometer is 2e3 meters long, that means you can detect a fractional change of about 1 part in 1e11. but it's actually crazy better than that due to many smart inventions the LIGO people created about locking optical cavities. you get the idea.
so then you watch your interference as a function of time, then go to your astronomy books to see what events should create gravitational waves at the frequency you have observed them.
in a nutshell.
m
ps. analogy: a radio telescope uses electronic amplifiers to measure the induced motion of electrons from EM waves : a GW telescope uses a high finesse optical cavity to measure the induced motion of masses from gravitational waves
I'm still awake and really should be sleeping, but instead I'll simplify even further, for the first year physics guy. Great description, by the way... I didn't know gravitational waves were supposed to be quadrupole.
Any accelerating charge (an electron for instance) will create an electromagnetic wave. A radio transmitter basically causes electrons in its antenna to oscillate at a particular frequency, and this produces radio waves at that frequency. Theoretically the same thing should hold for mass and gravity. If you cause a mass to accelerate (like the charge) then it should produce gravity waves (like the radio waves). Because gravity is so extraordinarily weaker than electromagnetism, the waves are correspondingly smaller, so very difficult to detect. Einstein says gravity causes space-time to curve, so passing gravity waves should stretch and squish space-time a little bit as they pass. Unfortunately you need to be able to measure distances really precisely.
An interferometer is how you do it. You send out two in phase light beams, bounce them off a mirror, then recombine them. If they travelled exactly the same distance then they should still be in phase (peaks and troughs line up) so they'll reinforce each other. If they travel slightly different distances then they won't be quite in phase anymore and the intensity of the recombined beam will be a bit less than it was originally.
So now you send the beams off at ninety degrees to each other and see if the ratio of the distances they travel changes. It will of course, due to all kinds of things, but maybe one of those things is passing gravity waves. So you have detectors on different continents and correlate their measurements. Local things (tiny earthquakes, people walking around above the detector, somebody turning on their washing machine down the street) will not be recorded by both detectors. Things like gravity waves will.
One more interesting thing you can do -- if you have more than two detectors, by watching when the waves are recorded by each detector you can measure the speed of the wave... the speed of gravity, and you can tell what direction the wave came from.
Simplified lots, and I should be sleeping, so that was probably full of errors and you should pay attention to the parent instead, but that's probably as simplified as it can get.
METERS
try reading http://en.wikipedia.org/wiki/LIGO
maybe the european one is 2000 feet, but not the two in th US. actually, the full length of each arm is 4000 m. i've been to the facility, touched a beamtube, drove to the end. meters.
m
Do you seriously think they might have forgotten about callibration? Do you think whoever is in charge of this thing is that dumb? By all means, if you do, pick up a telephone, call them and shout "Remember to do some form of callibration!!!". Be sure to be very emphatic. Science will thank you.
It's not an error, don't you know that in US english on the Internet any vowel can replace any other ?
The few readers who will actually know what "immanent" means will also know that it was actually supposed to be "imminent", so no harm done. The rest will just see it spelled as usual.
Live with the times !
May contain traces of nut.
Made from the freshest electrons.
in order for this to work you need both 2000ft arms to be the same EXACT length.
:-) and you'd take a long term average of the results to find a distance that has the highest peaked output and call that the centre baseline.
I presume that they have some way of adusting one path - you simply adjust it to peak brightness / least inteference. Then when something happens, it'll be a different distance either way and you'll see a null, or at least a drop.
If you can't get a peak because the damn thing is jiggling all over the place, then it's working
You are in a twisty maze of processor lines, all alike.
There is a lot of hype here.
LISA satellites need to be stable to within 1 nm per root Hz of bandwidth. (It's been a while since I worked on it, so someone else is welcome to explain what exactly this means.) Suffice it to say that this is a tractable problem, and I would argue no more difficult than the Advanced LIGO designs currently being implemented. And you get more bang for the buck in sensitivity.
Please show me a good reference for LIGO expected detection rates. This is taken from a popular book, but the numbers agree with what I remember hearing from those working on LIGO.
Supernova (within our galaxy)
1 to 3 per century
Black Hole/Black Hole Merger (300 million light-years)
1 per 1,000 years to 1 per year
Neutron Star/Neutron Star Merger (60 million light-years)
1 per 10,000 years to 10 per century
Neutron Star/Black Hole Merger (130 million light-years)
1 per 10,000 years to 10 per century
Source: Einstein's Unfinished Symphony: Listening to the Sounds of Space-Time by Marcia Bartusiak
Here's the deal with local sources: their masses are tiny compared to astronomical sources!
But here's a more local source that we have detected: the moon. The moon causes tidal deformations in the Earth's crust, which LIGO (disclosure: I am involved with the LIGO project) and the other large scale interferometers (GEO, VIRGO, TAMA) have to subtract out in order to see anything besides the moon.
Essentially, to make gravitational waves large, the conditions which need to be satisfied are 1) large amounts of matter 2) moving quickly. Things which satisfy this are: supernovae core collapses which are sufficiently non-axisymmetric, compact (eg. black hole or neutron star) binary system decay, and maybe some events we don't yet know of.
-Leo
This is in reply to this post and a number of others on the same topic.
... so they can correlate with those sources.
:)
Major sources of noise: seismic, acoustic, photon shot noise, thermal noise.
1) Acoustic noise: the entire beam tube system is in vacuum, so the only mechanical vibrations can be coupled in through the mirror supports, which are suspended on thin wire. The pendulum created by the hanging mirror essentially creates a mechanical low-pass filter which reduces the effects of noise above about 10 Hz. The gravitational wave projects (on Earth, not talking about LISA here) are mostly interested in frequencies around a few hundred to a few thousand Hz.
2) Seismic: this can cause pretty large displacement. Each of the mirrors (on its' hanging suspension) is sitting on a system of masses and springs (three levels) which creates a third order lowpass filter which further reduces noise.
3) Photon shot noise: this rises with frequency; essentially, photons are uncorrelated random events which create a Poisson noise distribution. In a Poisson distribution, the standard deviation of count rate is equal to the square root of the count rate, so the variance is decreased by decreasing count rate at the detector. This is why the interferometric detectors operate "in null," meaning they keep the mirrors at a differential path length which is equal plus or minus integer multiples of wavelengths. This way, the output at the point where they interfere is kept dark. The idea is that it's easier to detect a difference between 0 and 1 than between 100 and 101. (There is a ton of feedback to keep the whole system in null. Read up on Pound-Drever locking to understand it.)
4) Thermal noise: the surface of the mirror is made of atoms which jiggle in random Brownian motion. This is unavoidable unless the mirror is cooled sufficiently, which is difficult to do because of how well isolate the mirrors are. However, the Brownian motion can be averaged out over a large area by making the laser's spot size large.
So they've thought about it a little bit. And they are also measuring other non-detector channels like seismic activity and acoustics near the detector and wind speed and
The NSF doesn't go around giving millions to any old project
-Leo
Not sure about GEO600, but the LIGO interferometer uses a simple solution: build two observatories on opposite sides of the country, and if only one detects a signal, it's almost certainly spurious. I'm guessing that since TFA says GEO600 will come online at the same time, it'll just be treated as another part of the same array for those purposes.
This is sqrt(not) a sig.
This is a reply to this post and some of its' ancestors.
Gravitational waves are predicted to weakly interact with everything which is matter-energy. For that matter, gravity interacts with itself (which is why GR predicts black holes and other such singularities). However, in the weak-field regime (that is, space-time is flat except for a deviation which is orders of magnitude less, meaning we can take the leading term in the expansion, so the theory is linear), gravitational waves just pass through everything. Since they pass through things, their energy falls off like the square of distance from the source. In the strong-field interacting picture, they certainly should exhibit non-linear exotic behaviour, but those are precisely the parts of GR we are trying to probe with LIGO.
The exchange between matter-energy and curvature (gravitational waves) that you are thinking about is from the latter to the former, but just think about the former being turned into the latter - that is the prediction of the source of gravitational waves. However, it works both ways.
On the levels at which LIGO hopes to detect gravitational waves, we will see about 10^53 gravitons. I am quoting this figure without understanding where it comes from, since we certainly don't have a quantum theory of gravity. But gravity is predicted to be quantum in nature as well, but we won't see the quantization from where we stand.
One of the ancestors addressed the issue of measuring while your meter stick is being squashed and expanded, and another about the local speed of light. These issues are related. One of the postulates (argue argue whatever) of GR is that the speed of light is constant in every frame, and it has the same constant value compared between all frames. Light is the perfect meter stick (or clock) for making measurements with.
I had the same thought about measuring the arm lengths as you did for a while until I started taking GR. Here's how the thought goes: "If space is being stretched, and a meter stick is sitting in front of my face, I will always see the meter stick as being one meter long." Here's what GR predicts: the proper distance between free test masses sitting in space as a gravitational wave passes by will exhibit the increase in the X, decrease in the Y and vice versa oscillation pattern. To measure this, you need to use something free, like the mirrors at the ends of the beam tubes (they are really only free in one dimension). To measure distance, you can't use a meter stick, because it is not an ideal measuring device which you need to measure space with in GR. The ideal device is light. To think about it without resorting to a meter stick increasing and decreasing, think about the light travel time. Since light has a constant speed in all frames, if the proper distance is what is really increasing (disregard what happens to the meter stick, since it is made up of fallible matter and might stretch along with space, but light won't), then it will take longer to go down one arm and shorter down the other. Therefore one arm will add phase relative to the other, which will no longer perfectly interfere at the end.
-Leo
That's the most rediculous thing I've ever heard!
Bogtha Bogtha Bogtha