Is there a way to evaluate an optical equation with PSD?
The method proposed by one of your coworkers appears to be a good one: Evaluate the LOS equation with post26 calculations and then do an RPSD of the value. This preserves the signs of the individual terms. The 1 sigma value can be obtained as the square root of the integral of the RPSD. Here are some of the verifications I made: 1. The POST26 1 sigma of the difference between the corners matches the difference of the POST1 1 sigma values. Mode 1 dominates this quantity and the corner displacements are both positive, so subtracting 1 sigma values in POST1 is OK in this case. 2. The individual POST26 variables stored with NSOL contain the modal value for the degree of freedom being stored. 3. When these individual variables are combined the new variable has the same structure as the original variables and correct sign. 4. Preservation of the phase relationship between modes is observed in the dip in the RPSD at MODE 4 (572 Hz) in the attached Powerpoint. The attached listing shows that the combined result for mode 4 has the opposite sign of mode 1, and therefore subtracts from the combined RPSD. |
||
|