We were adding a new filter module for annular CO2 heating to the TCS simulation model. Both filter modules are added to the substrate diopter calculation, but their inputs are switched using the CO2 mask. For now the same filter as for central heating has been loaded. This needs to be updated.
The calibration factors are -11uD/W and -13uD/W for ITMX and ITMY annular heating, respectively, see alog 66617.
We also noticed that the filter module name for H1:TCS-SIM_ITMX_SUB_DEFOCUS_FULL_SINGLE_PASS was too long and shortened it to H1:TCS-SIM_ITMX_SUB_DEFOCUS_FULL_SGL_PASS. However, we added the following EPICS channels H1:TCS-SIM_ITMX_SUB_DEFOCUS_FULL_SINGLE_PASS_OUTPUT and H1:TCS-SIM_ITMX_SUB_DEFOCUS_FULL_SINGLE_OUT16 back in so medm screens and trend will stay unchanged.
The following 2 lines have been added to the CDS_CA_COPY guardian list:
- H1:TCS-SIM_ITMX_SUB_CO2_MASK_SEL H1:TCS-ITMX_CO2_FLIP1MON
- H1:TCS-SIM_ITMY_SUB_CO2_MASK_SEL H1:TCS-ITMY_CO2_FLIP1MON
I used the CO2 profile measurement in aLOG 65382 to estimate the average radial CO2 intensity profile on the CP. The spatial scale on the CP was determined by using the 16 axial nodes in the HWS measurement in aLOG 65436 as fiducials to estimate the diameter on the CP.
From this, I ended up with an average radial profile for intensity for the annular CO2 laser - which, when normalized to deliver 1W, looks like the following.
I put this into a 2D axially symmetric model of the coupled ITM+CP system in COMSOL and worked out the transient response for the defocus. The response looks like the following:
From this response, we estimate the steady-state defocus and can fit a weighted sum of exponential decays to the normalized response. The fitted time constants, tau, correspond to the poles in the transient response function, where the poles are 1/(tau*2*pi).
The COMSOL model predicts peak after about 2-3 hours and a slow decrease over 24 hours to settle around -9uD/W.
Steady state defocus = -8.91uD/W
amplitude = [-1.571, -0.846, 1.417]
tau = [798.6s, 7756.6s, 26491.9s]
poles = [199.30uHz, 20.52uHz, 6.01uHz]
Normalized response = 1 + sum[ amp(ii)*exp(-t/tau(ii) ) ]
The transient output of the annulus can be estimate by a weighted sum three single-pole LPF, where the weighting for each pole is just the "a" values given in the previous aLOG.
These are single-pass values.
Attaching the analysis used.
Note on the absolute power level for this simulation: it's a little unclear how to normalize the amount of power in the CO2 intensity profile. There is the power on the thermopile to consider, the aperture of the CP to consider (some power is lost as it falls off the edge of the CP) and the missing bits of the intensity on the paper on the camera. This will result in a calibration error in the magnitude in the simulation - the time constants should be relatively unaffected though.
Translated this into the following filter function: zpk([2.837e-05;2.691e-06],[6.01e-06;0.0001993;2.052e-05],1,"n")
Attached is the filter step response as given by foton in blue (red is the previsouly loaded filter.
New filters loaded into H1:TCS-SIM_ITMX_SUB_DEFOCUS_CO2_AN and H1:TCS-SIM_ITMY_SUB_DEFOCUS_CO2_AN at FM6.