|
Size: 2779
Comment:
|
← Revision 31 as of 2015-03-16 18:41:46 ⇥
Size: 2162
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 5: | Line 5: |
| ||Name||Platform||Domain||Capabilities||Author||Web|| ||Optickle||Matlab||Frequency Domain||Radiation Pressure, 01 mode||Matt Evans||--([[http://ilog.ligo-wa.caltech.edu:7285/advligo/ISC_Modeling_Software|Password protected]])--|| ||Finesse||C?||Frequency Domain||Higher order modes||Andreas Freise||[[http://www.rzg.mpg.de/~adf/]]|| ||e2e||C++||Time Domain|| ||Hiro Yamamoto||[[http://www.ligo.caltech.edu/~e2e/]]|| |
||Name||Platform||Domain||Capabilities||Author||Web||Model|| ||Optickle||Matlab||Frequency Domain||Radiation Pressure, 01 mode||Matt Evans||--([[http://ilog.ligo-wa.caltech.edu:7285/advligo/ISC_Modeling_Software|Password protected]])--||on Hongo svn|| ||Finesse||C?||Frequency Domain||Higher order modes||Andreas Freise||[[http://www.rzg.mpg.de/~adf/]]||?|| ||e2e||C++, JAVA||Time Domain|| ||Hiro Yamamoto||[[http://www.ligo.caltech.edu/~e2e/]]||?|| == Mode match tool == LIGO's favorite is Nic Smith's "A la mode": http://www.ligo.mit.edu/~nsmith/alm/alm.html |
| Line 13: | Line 18: |
| I am comparing several simulation tools with theoretical line using LCGT 2007 parameters. I used here FINESSE, Thomas, Oprickle. Except for FINESSE, all results matched pretty well within 0.1%-1% accuracy. FINESSE seems to have a bug of a factor by sqrt(2), but it can be compensated later if you know this information. | Several simulation tools with theoretical line using LCGT 2007 parameters are compared here, FINESSE, Thomas, Oprickle. Except for FINESSE. |
| Line 16: | Line 21: |
| *FINESSE: Developped by Andreas Freise. C based. 'ps2S' and 'scale meter' commands are used for the unit in m/rHz and 'fig' command are used to produce excitations in frequency domain. Compensation factor *sqrt(2) was needed to match other results. This could be a bug. *Thomas: Developped by Thomas Cobit. Visual C++ besed, running only Windows machine. This is the first simulation tool which can calculate quantum effect using two-photon mode. *Optickle: Developped by Matt Evans. Matlab based, very flexible simulation tool. You can use a full set of Matlab functions. This tool also can calculate full quantum effect. |
*FINESSE: 'ps2S' and 'scale meter' commands are used for the unit in m/rHz and 'fig' command are used to produce excitations in frequency domain. FINESSE seems to have a bug of a factor by sqrt(2), but it can be compensated later if you know this information to match other results. *Thomas: Developped by Thomas Corbit. Visual C++ besed, running only Windows machine. This is the first simulation tool which can calculate quantum effect using two-photon modes. *Optickle: Very flexible simulation tool. You can use a full set of Matlab functions. This tool also can calculate full quantum effect. Optickel RF calculation seems to have a bug of a factor by sqrt(2) for the radiatoin pressure noise, but DC readout calculation is OK. This bug can be compensated to put 2*sqrt(2) instead of 2 on the setMechTF function but radiation pressure noise for DC readout will be larger by a factor of sqrt(2) in this case. |
| Line 20: | Line 25: |
=== FPMI, no loss case for LCGT parameters === {{attachment:FPMI_noloss.png}} [[attachment:20090709_simall_FPMI_noloss.zip]] For shotnoise calculation here, just very hevy masses are used in Thomas tool and Optickle. === PRFPMI, no loss case for LCGT parameters === {{attachment:PRFPMI_noloss.png}} [[attachment:20090709_simall_PRFPMI_noloss.zip]] === BRSE, no loss case for LCGT parameters === {{attachment:BRSE_noloss.png}} [[attachment:20090709_simall_BRSE_noloss.zip]] === BRSE, no loss case with radiation pressure noise for LCGT parameters === {{attachment:BRSE_noloss_rad.png}} [[attachment:20090709_simall_BRSE_noloss_rad.zip]] Optickle uses vacuum injections from dark port and other lossy ports on this model due to introduce quantum noise, so Optickle result with RF detection is worse than other calculation like Thomas's DC readout. Calculation above this model assumed ideal flat shot noise, so all the results were matched well. |
|
| Line 53: | Line 30: |
| [[attachment:20090709_simall_BRSE_loss_rad.zip]] | [[attachment:20090930_simall_BRSE_loss_rad.zip]] Optickle RF result is lower by a factor of sqrt(2) than other results due to the bug explained above. |
Simulation Tools
Name |
Platform |
Domain |
Capabilities |
Author |
Web |
Model |
Optickle |
Matlab |
Frequency Domain |
Radiation Pressure, 01 mode |
Matt Evans |
on Hongo svn |
|
Finesse |
C? |
Frequency Domain |
Higher order modes |
Andreas Freise |
? |
|
e2e |
C++, JAVA |
Time Domain |
|
Hiro Yamamoto |
? |
Mode match tool
LIGO's favorite is Nic Smith's "A la mode": http://www.ligo.mit.edu/~nsmith/alm/alm.html
Comparison of simulation tools for LCGT
Several simulation tools with theoretical line using LCGT 2007 parameters are compared here, FINESSE, Thomas, Oprickle. Except for FINESSE.
Theoretical: just analytical way shown in LCGT design document on page 14.
- FINESSE: 'ps2S' and 'scale meter' commands are used for the unit in m/rHz and 'fig' command are used to produce excitations in frequency domain. FINESSE seems to have a bug of a factor by sqrt(2), but it can be compensated later if you know this information to match other results.
- Thomas: Developped by Thomas Corbit. Visual C++ besed, running only Windows machine. This is the first simulation tool which can calculate quantum effect using two-photon modes.
- Optickle: Very flexible simulation tool. You can use a full set of Matlab functions. This tool also can calculate full quantum effect. Optickel RF calculation seems to have a bug of a factor by sqrt(2) for the radiatoin pressure noise, but DC readout calculation is OK. This bug can be compensated to put 2*sqrt(2) instead of 2 on the setMechTF function but radiation pressure noise for DC readout will be larger by a factor of sqrt(2) in this case.
BRSE, with loss, radiation pressure noise for LCGT parameters
20090930_simall_BRSE_loss_rad.zip
Optickle RF result is lower by a factor of sqrt(2) than other results due to the bug explained above.