An error function minimization approach for the inverse problem of adaptive mirrors tuning
2014
Adaptive x-ray optics are more and more used in synchrotron beamlines, and it is probable that they will be considered
for the future high-power free-electron laser sources, as the European XFEL now under construction in Hamburg, or
similar projects now in discussion. These facilities will deliver a high power x-ray beam, with an expected high heat load
delivered on the optics. For this reason, bendable mirrors are required to actively compensate the resulting wavefront
distortion. On top of that, the mirror could have also intrinsic surface defects, as polishing errors or mounting stresses. In
order to be able to correct the mirror surface with a high precision to maintain its challenging requirements, the mirror
surface is usually characterized with a high accuracy metrology to calculate the actuators pulse functions and to assess its
initial shape. After that, singular value decomposition (SVD) is used to find the signals to be applied into the actuators,
to reach the desired surface deformation or correction. But in some cases this approach could be not robust enough for
the needed performance. We present here a comparison between the classical SVD method and an error function
minimization based on root-mean-square calculation. Some examples are provided, using a simulation of the European
XFEL mirrors design as a case of study, and performances of the algorithms are evaluated in order to reach the ultimate
quality in different scenarios. The approach could be easily generalized to other situations as well.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
11
References
4
Citations
NaN
KQI