Forgot your username or password?
Document Actions
Alignment Simulation Software
Discussion on which software we need to simulate alignmnent
2009/03/26
participants:
EVO: A. Freise, F. Terenzi, M. Pichot, F. Bondu
EGO: M. Mantovani, A. Vicere, G. Losurdo, M. Punturo
* Several questions in A. Vicere's talk:
- Model for multi payload suspension?
M. Mantovani: yes, we need to understand how to control these suspensions
- Multi-mirror cavities: effective model (total length) or really several mirrors?
M. Mantovani: real poly mirror suspension: better, it allows to understand mode displacement
- Infinite light speed in recycling cavities: is it OK?F. Bondu: the linewidth of the recycling cavities is much larger than the detection bandwidth and than UGF's
=> OK for SBs, TBC for carrier
* Highlights by A. Vicere:
-siesta can output mechanical ABCD matrices
-Some workforce may be available when VSR2 starts.
-Siesta probably not appropriate tool for exploratory studies (several NDRC configurations)
-Implement new payload: 3-4 months if full time
-Implement Signal recycling cavity: less clear
Discussion
----------
1. Siesta or Finesse?
- - - - - - - - - - - -
A. Freise: E2E can save the state space, and has parallelization
A. Vicere: Saving state space available, may be not tested. Parallelization would require a computing scientist. A gain of 10 might be possible.
A. Freise: We need both time domain and frequency domain simulations. Pool of real users is small.
F. Bondu: Time domain simulation: accurate mechanics and control; non linear effects
Frequency domain: all linear questions
A. Vicere: Cross checks possible with two models
A. Freise: Time domain software useful for students
Conclusion: We need to develop and use both.
2. Simulation for alignment
- - - - - - - - - - - - - - -
M. Mantovani: need simple model to start with, in the static situation
A. Freise: I thought about radiation pressure in Finesse, in the framework of T. Corbitt, calculating both sidebands simultaneously.
Problem is to code optics in C. Would take 2 months for a trained engineer, 6 months including training and debugging.
I wrote a proposal; if funded, it would start autum 2010. I propose help if Virgo provides a person to spend several months (full time or integrated time) on this in the next 12 months.
A. Vicere: for siesta, the path to implement mechanics is clear. For optics, may take 1 year?
Conclusion: F. Bondu identifies a 6 month task in WBS, task attribution TBD by VSC.
Some work may start during VSR2.