- #1
Enthalpy
- 667
- 4
Hello everybody!
Extreme interferometers like Ligo, Virgo, Leo600, Tama300 try to detect gravitational waves
http://en.wikipedia.org/wiki/Gravitational-wave_detector
and ground movements are one difficulty for them. The mirrors are suspended in several stages to insulate them, sometimes actively.
In addition to suspension, I suggest to measure the ground's movements by other means, and identify by how much these movements transmit into the measure at all frequencies, then subtract this contribution as estimated from the measured ground movements and the transfer function. This so-called adaptive filter is commonly used in acoustics, one method being the Kálmán filter
http://en.wikipedia.org/wiki/Kalman_filter
which routinely attenuates noise sources by 40dB.
In a first method, triaxial accelerometers can measure the ground movements. At least three pieces on a triangle plus one at depth (earthquakes are deep) would pick the distant noise sources' direction, which influences the effect on the interferometer. A too wide basis might be less good against near sources - perhaps. Dedicated sets of sensors can pick noise made by known sources like a machine, maybe a road.
Accelerometers are straightforward hence may be used already; I didn't see them mentioned after short reading. The following one is not shown on the interferometers drawings I saw and could be new. My sketch omits the second arm, the interference components, and all refinements.
I propose to add auxiliary beams between mirrors hold at an earlier stage of the suspension. These would pick ground movements almost as the main beams do, easing the cancellation, but upstream the mechanical filter hence more strongly. The auxiliary beams pick gravitational waves as well, but this contribution is strongly attenuated by the transfer function that mimics the mechanical filter.
Being more shaken, the auxiliary beam is built less sensitive than the main one, by using fewer bounces, a longer wavelength... Different wavelengths would help sort out both beams; consider my "evanescent wave optical filter" for strong stopband attenuation. A new interferometer design like Tama300 has it easier.
This looks useful. Have you seen it used at a gravitational wave detector?
Marc Schaefer, aka Enthalpy
Extreme interferometers like Ligo, Virgo, Leo600, Tama300 try to detect gravitational waves
http://en.wikipedia.org/wiki/Gravitational-wave_detector
and ground movements are one difficulty for them. The mirrors are suspended in several stages to insulate them, sometimes actively.
In addition to suspension, I suggest to measure the ground's movements by other means, and identify by how much these movements transmit into the measure at all frequencies, then subtract this contribution as estimated from the measured ground movements and the transfer function. This so-called adaptive filter is commonly used in acoustics, one method being the Kálmán filter
http://en.wikipedia.org/wiki/Kalman_filter
which routinely attenuates noise sources by 40dB.
In a first method, triaxial accelerometers can measure the ground movements. At least three pieces on a triangle plus one at depth (earthquakes are deep) would pick the distant noise sources' direction, which influences the effect on the interferometer. A too wide basis might be less good against near sources - perhaps. Dedicated sets of sensors can pick noise made by known sources like a machine, maybe a road.
Accelerometers are straightforward hence may be used already; I didn't see them mentioned after short reading. The following one is not shown on the interferometers drawings I saw and could be new. My sketch omits the second arm, the interference components, and all refinements.
I propose to add auxiliary beams between mirrors hold at an earlier stage of the suspension. These would pick ground movements almost as the main beams do, easing the cancellation, but upstream the mechanical filter hence more strongly. The auxiliary beams pick gravitational waves as well, but this contribution is strongly attenuated by the transfer function that mimics the mechanical filter.
Being more shaken, the auxiliary beam is built less sensitive than the main one, by using fewer bounces, a longer wavelength... Different wavelengths would help sort out both beams; consider my "evanescent wave optical filter" for strong stopband attenuation. A new interferometer design like Tama300 has it easier.
This looks useful. Have you seen it used at a gravitational wave detector?
Marc Schaefer, aka Enthalpy