![Modeling of optical scattering using the Bidirectional
Reflectance Distribution Functions of Advanced LIGO test masses
Hunter Rew1
and Joseph Betzwieser2
1) College of William and Mary 2) California Institute of Technology
LIGO-G1400923
Introduction
test mass
test mass
test mass
test mass
light storage arm
photodetector
laser
beam
splitter
light storage arm
Figure 1: Diagram of a laser interferometer.
The Laser Interferometer Gravitational-Wave Ob-
servatory (LIGO) contains two Fabry-Perot cavities
as seen in Figure 1. As light reflects between two test
masses, a small percentage is scattered in an unin-
tended direction due to surface imperfections. This
scattering results in power loss and noise
in the interferometer (IFO) which must be ei-
ther accounted for or corrected[1]. Measuring opti-
cal scattering requires a power measurement of the
light scattered to a known area, distance, and angle.
This can be accomplished directly either through
the use of a camera with a known conversion from
power incident on the lens to intensity of light in
the image or with a photodiode (PD). The scat-
ter is measured by the Bidirectional Re-
flectance Distribution Function (BRDF) and
is given by[2]:
BRDF =
Ps
Ω × Pi × cos(θs)
(1)
It is a measure of the ratio of power which is scat-
tered (Ps) per solid angle (Ω) to power incident (Pi)
on the reflecting surface.
Figure 2: How light scatters from an imperfect surface[3].
LIGO Test Masses
Figure 3: Photograph of installed ITM.
Distance from optic center (m)
Distancefromopticcenter(m)
Power Distribution on ITMY (clean)
−0.5 −0.4 −0.3 −0.2 −0.1 0 0.1 0.2 0.3 0.4 0.5
−0.5
−0.4
−0.3
−0.2
−0.1
0
0.1
0.2
0.3
0.4
0.5
Figure 4: Model of power scattered to ITMY.
Figure 3 Shows the input test mass (ITM) that the laser passes
through to enter a Fabry-Perot cavity. Once the laser enters a
cavity it reflects, and scatters, back and forth between the two
test masses (TMs) until light exiting the cavity is equal to the
entering light[1]. To prevent scattered light from interfering with
the main beam, structures, called baffles, are placed between the
TMs. These baffles each contain 4 PDs. The final power circulat-
ing within the cavity is given by Equation 2[1].
Pcavity =
2F
π
× PITM (2)
Here F is the cavity finesse (416 for Advanced LIGO) and PITM is
the power entering the cavity through the ITM. Once the TMs are
aligned and the cavity power has stabilized, the IFO is said to be
in lock. The path and power distribution (W × m−2
) of the light
can be modeled using the Static Interferometer Simulation (SIS)
package, which uses phase maps of the TMs taken before instal-
lation. Figure 4 shows a head on view of this power distribution
in log scale at the input test mass of the y-arm (ITMY).
Comparing Data with the Model
Data was taken for each set of PDs surrounding each TM to
find the power scattered to their solid angles. Power incident
on the TMs was calculated with Equation 2. This was done for
every lock period of Advanced LIGO. With this data, BRDF
values for each individual lock were calculated using Equation
1 and averaged to produce the statistics in Table 1. As can be
seen from the large standard deviations, the BRDF fluc-
tuates significantly from lock to lock. The features
of alternating peaks and troughs in Figure 4 encompass the
region where the PDs are located. It is easy to imagine that,
with a slight change in alignment, the light seen by a PD could
change significantly. The figure to the right shows how the
measurements compare to the SIS modeled values for scatter
towards ITMY. There are two important points to note from
this plot:
• The measurements are comparable with the model
• Scatter from the installed TMs is larger than the model for
all angles
The latter indicates that the fine structure of the TMs was al-
tered during installation, likely from accumulation of particles.
PD BRDF (Ω−1
) SD (Ω−1
) SEM (Ω−1
)
1 1160±40 680.57 74.26
2 8.0±0.4 3.33 0.37
3 14.5±0.6 10.65 1.19
4 170±20 75.11 8.56
Table 1: Mean of measured BRDFs from ETMY with
standard deviation and standard error of the mean.
4 5 6 7 8 9 10 11 12
x 10
−5
0
0.2
0.4
0.6
0.8
1
1.2
x 10
−8
Scatter angle (radians)
Fractionoftotalpowerscatteredtoangle
Comparison of Measured and Modeled Scatter by Angle (ITMY)
Measured scatter
Modeled scatter
Figure 5: Measured and modeled fraction of power
scattered to PDs.
Conclusion
The measured BRDF of each test mass varies sig-
nificantly from lock to lock. From analyzing mod-
els produced by SIS, this variation is likely the re-
sult of alignment changes. We found that the
fine structure of the test masses was al-
tered during the process of installation,
resulting in larger power losses at all mea-
sured angles compared to the model. More
work should be done comparing measurements to
the model in order to develop new phase maps which
reflect the conditions of the installed test masses.
Additional Information
Read my paper Contact me
References
[1] Peter R. Saulson.
Fundamentals of Interferometric Gravitational Wave
Detectors.
World Scientific Publishing, 1994.
[2] John C. Stover.
Optical Scattering: Measurement and Analysis.
The International Society for Optical Engineering, second
edition, 1995.
[3] Bsdf.
http://en.wikipedia.org/wiki/File:
BSDF05_800.png, 2006.
File: BSDF05 800.png, GNU Free Documentation
License.
Acknowledgements
This research was conducted under Caltech’s Summer Under-
graduate Research Fellowship program at the Ligo Livingston
Observatory and was funded by the National Science Founda-
tion. Special thanks to Hiro Yamamoto of Caltech for his work
on SIS.](https://image.slidesharecdn.com/156cf473-a4d6-4005-8779-35a7f600edd5-150908083410-lva1-app6891/85/Poster_Modeling_of_Optical_Scattering_in_Advanced_LIGO-1-320.jpg)
Poster_Modeling_of_Optical_Scattering_in_Advanced_LIGO
- 1. Modeling of optical scattering using the Bidirectional Reflectance Distribution Functions of Advanced LIGO test masses Hunter Rew1 and Joseph Betzwieser2 1) College of William and Mary 2) California Institute of Technology LIGO-G1400923 Introduction test mass test mass test mass test mass light storage arm photodetector laser beam splitter light storage arm Figure 1: Diagram of a laser interferometer. The Laser Interferometer Gravitational-Wave Ob- servatory (LIGO) contains two Fabry-Perot cavities as seen in Figure 1. As light reflects between two test masses, a small percentage is scattered in an unin- tended direction due to surface imperfections. This scattering results in power loss and noise in the interferometer (IFO) which must be ei- ther accounted for or corrected[1]. Measuring opti- cal scattering requires a power measurement of the light scattered to a known area, distance, and angle. This can be accomplished directly either through the use of a camera with a known conversion from power incident on the lens to intensity of light in the image or with a photodiode (PD). The scat- ter is measured by the Bidirectional Re- flectance Distribution Function (BRDF) and is given by[2]: BRDF = Ps Ω × Pi × cos(θs) (1) It is a measure of the ratio of power which is scat- tered (Ps) per solid angle (Ω) to power incident (Pi) on the reflecting surface. Figure 2: How light scatters from an imperfect surface[3]. LIGO Test Masses Figure 3: Photograph of installed ITM. Distance from optic center (m) Distancefromopticcenter(m) Power Distribution on ITMY (clean) −0.5 −0.4 −0.3 −0.2 −0.1 0 0.1 0.2 0.3 0.4 0.5 −0.5 −0.4 −0.3 −0.2 −0.1 0 0.1 0.2 0.3 0.4 0.5 Figure 4: Model of power scattered to ITMY. Figure 3 Shows the input test mass (ITM) that the laser passes through to enter a Fabry-Perot cavity. Once the laser enters a cavity it reflects, and scatters, back and forth between the two test masses (TMs) until light exiting the cavity is equal to the entering light[1]. To prevent scattered light from interfering with the main beam, structures, called baffles, are placed between the TMs. These baffles each contain 4 PDs. The final power circulat- ing within the cavity is given by Equation 2[1]. Pcavity = 2F π × PITM (2) Here F is the cavity finesse (416 for Advanced LIGO) and PITM is the power entering the cavity through the ITM. Once the TMs are aligned and the cavity power has stabilized, the IFO is said to be in lock. The path and power distribution (W × m−2 ) of the light can be modeled using the Static Interferometer Simulation (SIS) package, which uses phase maps of the TMs taken before instal- lation. Figure 4 shows a head on view of this power distribution in log scale at the input test mass of the y-arm (ITMY). Comparing Data with the Model Data was taken for each set of PDs surrounding each TM to find the power scattered to their solid angles. Power incident on the TMs was calculated with Equation 2. This was done for every lock period of Advanced LIGO. With this data, BRDF values for each individual lock were calculated using Equation 1 and averaged to produce the statistics in Table 1. As can be seen from the large standard deviations, the BRDF fluc- tuates significantly from lock to lock. The features of alternating peaks and troughs in Figure 4 encompass the region where the PDs are located. It is easy to imagine that, with a slight change in alignment, the light seen by a PD could change significantly. The figure to the right shows how the measurements compare to the SIS modeled values for scatter towards ITMY. There are two important points to note from this plot: • The measurements are comparable with the model • Scatter from the installed TMs is larger than the model for all angles The latter indicates that the fine structure of the TMs was al- tered during installation, likely from accumulation of particles. PD BRDF (Ω−1 ) SD (Ω−1 ) SEM (Ω−1 ) 1 1160±40 680.57 74.26 2 8.0±0.4 3.33 0.37 3 14.5±0.6 10.65 1.19 4 170±20 75.11 8.56 Table 1: Mean of measured BRDFs from ETMY with standard deviation and standard error of the mean. 4 5 6 7 8 9 10 11 12 x 10 −5 0 0.2 0.4 0.6 0.8 1 1.2 x 10 −8 Scatter angle (radians) Fractionoftotalpowerscatteredtoangle Comparison of Measured and Modeled Scatter by Angle (ITMY) Measured scatter Modeled scatter Figure 5: Measured and modeled fraction of power scattered to PDs. Conclusion The measured BRDF of each test mass varies sig- nificantly from lock to lock. From analyzing mod- els produced by SIS, this variation is likely the re- sult of alignment changes. We found that the fine structure of the test masses was al- tered during the process of installation, resulting in larger power losses at all mea- sured angles compared to the model. More work should be done comparing measurements to the model in order to develop new phase maps which reflect the conditions of the installed test masses. Additional Information Read my paper Contact me References [1] Peter R. Saulson. Fundamentals of Interferometric Gravitational Wave Detectors. World Scientific Publishing, 1994. [2] John C. Stover. Optical Scattering: Measurement and Analysis. The International Society for Optical Engineering, second edition, 1995. [3] Bsdf. http://en.wikipedia.org/wiki/File: BSDF05_800.png, 2006. File: BSDF05 800.png, GNU Free Documentation License. Acknowledgements This research was conducted under Caltech’s Summer Under- graduate Research Fellowship program at the Ligo Livingston Observatory and was funded by the National Science Founda- tion. Special thanks to Hiro Yamamoto of Caltech for his work on SIS.