Touschek lifetime and +/- momentum aperture

[NicolaCarmignani]

At ESRF we are doing some lattice studies for our current and upgraded storage rings, using Elegant to compute the momentum aperture and resulting Touschek lifetime.
Elegant computes the positive and negative momentum aperture to be used in the lifetime calculation. The corresponding scattering rates should then be added to get the total scattering rate (1/tau = 1/tau_p + 1/tau_n).
However, looking at the touschekLifetime C code, it seems that the minimum of the momentum aperture is used.
We have cases with big enough asymmetry between delta_p and delta_n that this difference in evaluation of lifetime has significant impact.
Did we read the source code correctly?
Are we right in thinking the lifetime should be computed including the positive and negative terms separately?

Nicola Carmignani
Boaz Nash

[xiaoam]

Nicola,

You are right. We did assume symmetry (using the minimum value) momentum aperture in our code for reducing calculation work. Thanks for point out the need to treat negative and positive part separately. The code has been updated and should be realeased next time. Anybody need the code before the new realease date can send me an email.

Thank you very much,

aimin

[hywel]

That's really helpful. We're about to do some +/- Touschek calculations with asymmetric momentum aperture too, so this is a handy update when it comes.

[michael_borland]

Since it seems this is of some interest and urgency, we'll try to push out the update in the next week or so.

--Michael

[xiaoam]

A quick solution to this problem is as following:
Assume the input file twiss: a.twi, aperture: a.mmap
sddsprocess a.mmap pos.mmap -redef=col,deltaNegative,"deltaPositive -1 *"
sddsprocess a.mmap neg.mmap -redef=col,deltaPositive,"deltaNegative -1 *"
touschekLifetime neg.out -twiss=a.twi -aperture=neg.mmap ...
touschekLifetime pos.out -twiss=a.twi -aperture=pos.mmap ...
1/lifetime = 1/(2*lifetime_neg) + 1/(2*lifetime_pos)

This only works for command line calculation. For Monte Carlo simulation inside elegant, you have to wait for the new release.

Thanks,

aimin

[hywel]

That's a great fix, thanks!