summaryrefslogtreecommitdiffstats
path: root/physics/LIGO
diff options
context:
space:
mode:
Diffstat (limited to 'physics/LIGO')
-rw-r--r--physics/LIGO48
1 files changed, 48 insertions, 0 deletions
diff --git a/physics/LIGO b/physics/LIGO
new file mode 100644
index 0000000..23fba64
--- /dev/null
+++ b/physics/LIGO
@@ -0,0 +1,48 @@
+=======================================================================
+LIGO: Laser Interferometer Gravitational Observatory
+=======================================================================
+
+.. warning:: This is a rough work in progress!! Likely to be factual errors, poor grammer, etc.
+
+.. note:: Most of this content is based on a 2002 Caltech course taught by
+ Kip Thorn [PH237]_
+
+Noise Sources
+~~~~~~~~~~~~~~~~~~~~~~
+
+For initial LIGO, seismic noise dominates below about 60Hz, suspension thermal
+noise between 60 and 180Hz, and radiation pressure shot noise above 180Hz.
+
+Sensitivity
+~~~~~~~~~~~~~~~~~~~~~
+
+Advanced LIGO will use 40kg sapphire test masses with sensitivity of about 10e-19 meters: 1/10000 of an atomic nucleus, 10e-13 of a wavelength, and half of the entire mirror's wave function.
+
+LISA
+~~~~~~~~~~~~~
+
+5e6 km separations between three spacecraft, 1 (astronomical unit, ~1.5e8 from the sun. 1 watt lasers.
+The heterodyne detection is of the beat frequencies at each spacecraft of the
+two incoming beams. Doppler shifts of spacecraft must be taken into account,
+due not only to sun radiation pressure etc, but varying gravitational fields
+from planetary orbits.
+
+The test masses inside LISA should be free falling and have relative
+separations stable to 10e-9 cm (10e-5 wavelength of light).
+
+LISA's sensitivity is in the milihertz regime.
+
+.. note: (insert LISA noise curve?)
+
+Data Analysis
+~~~~~~~~~~~~~~~~~~~~
+
+Using matched filtering (eg, take cross correlation between two waveforms,
+integrating their product), frequency sensitivity will be around the inverse
+of the number of cycles of waveform (for LIGO, around 20,000 cycles, for LISA
+around 200,000 cycles).
+
+This technique requires known, theoretically derived waveforms (within
+phase/amplitude). There are other methods when we don't have good guesses
+about the waveform we are looking for...
+