Bespoke GW Models

Finesse can be extended with additional packages to model particular gravitational wave detectors. Two such packages are finesse-ligo and finesse-virgo. Additionally, packages exist to combine Finite Element Analysis (FEA) of the test masses, with optical simulations of the interferometer. One example is the test-mass-thermal-state package.

This example uses finesse-ligo to obtain the transfer function from Meters of test mass motion into Watts of optical power for the LIGO Livingston Observatory.


This example is based on the LSC example in the finesse-ligo documentation. However, the code here is not automatically tested, as this would require making finesse-ligo a finesse dependency.

Please see the finesse-ligo documentation for up-to-date information.

Importing finesse-ligo

If it is installed, finesse-ligo can either be imported with finesse.ligo, or it can be imported directly with import finesse_ligo. Both cases should import the same code:

import finesse
import finesse_ligo
import finesse.analysis.actions as fac

from finesse_ligo.factory import aligo
from finesse.plotting import bode
from finesse_ligo.suspension import QUADSuspension

finesse-virgo works similarly. We then proceed to build the Finesse model. A parameter file is distributed with finesse-ligo that contains the best estimate of all parameters in the interferometers. In this case, we will load the parameter file for the LIGO Livingston Observatory:

factory = aligo.ALIGOFactory("llo_O4.yaml")

The ALIGOFactory function can accept either the name of a parameter file, or, a path to a parameter file, stored somewhere on your system.

Choosing what to model

It is computationally expensive to model the entire interferometer. Usually, one wants to model some subset of the physics and some subset of the ports. We now choose what to model using the factory function:

# Make a HOM model
factory.options.QUAD_suspension_model = finesse_ligo.suspension.QUADSuspension
llo3 = factory.make()
llo3.modes("even", maxtem=2)

In this case, we have chosen to model an interferometer, with suspended optics and the HG00, HG20 and HG02 optical modes.

Initialising the locks

A real interferometer has a real operating point enforced by feedback control. It is critical to find this operating point using simulated locking actions. In this case, the locks and gains are known and are distributed with finesse-ligo:, aligo.DARM_RF_to_DC()))

Modelling the Interferometer

We are now in a position to do some modelling. Such as plotting the the transfer function:

analysis = fac.FrequencyResponse(
    np.geomspace(0.1, 10e3, 300), "DARM.AC.i", "AS.DC", open_loop=True

For further information, please see the documentation for the extension packages, or see the :doc:`../getting_started/getting_help`_ page.