One-turn matrix in the presence of the RF field?

[gpark]

I have to replace the actual lattice by a one-turn matrix to deal with multiple turns (more than a thousand). The lattice has a RF cavity. In this case, when I read off a one-turn matrix from .mpr file, use EMATRIX to define a new lattice, and benchmark beamdynamics against the actual lattice, they were different.
Is this because the one-turn matrix is not a correct replacement of the actual lattice when there is RF field? (The longitudinal sector of the one-turn matrix looked a little funny), or there must my mistakes in dealing with the script?

Thanks,

[michael_borland]

The recommended way to set up the single-turn matrix is using the ILMATRIX element, since it allows including details (such as chromaticity) in a fashion that is appropriate to long-term tracking. You can find a guide to doing this in the examples files, under ILMatrixFromTracking.

Once you've done that, you should be able to perform tracking and verify the tunes and chromaticities, among other properties, compared to element-by-element tracking.

Some other examples show use of the ILMATRIX method for collective effects studies. See, for example, storageRingRfNoise, APSRing/ionEffects1, or multibunchCollectiveEffects/APS-24Bunch-CBI.

--Michael