The OPAL Discussion Forum

[<suzie.sheehy AT stfc.ac.uk>]

Hi all,

Could someone please explain how the 'correlation' parameter in the 
distribution function works?

This hasn't been an issue for  me until now as I've only used distributions 
where the twiss parameter alpha=0, so no correlation was necessary. However 
now I'd like to input a beam with alpha non-zero if possible. Ideally I'd 
like to just input the required emittance and twiss parameters, but this 
doesn't seem to be possible (yet?).

I am trying to use a binomial distribution, m=0.5, 1.0 or 2.0. The manual 
doesn't make it clear whether this is even possible, but the 
'distribution.cpp' source code around line 2069 – seems to be implying that 
this can be used for a binomial distribution.

Below is my distribution input, based on the fact that in the source code:
corr_i = -alpha_i/sqrt(beta_i*abs(gamma_i))

Dist1:DISTRIBUTION, DISTRIBUTION=BINOMIAL,
sigmax= 5.1063219705e-03, sigmapx=1334.48363408, corrx=-0.554335172211,mx=1.0,
sigmay=  25.0e-03, sigmapy=28.0, corry=0.0,my=1.0,
sigmat=   7.0724423802e-03, sigmapt=695.650263656, 
corrt=-0.733389020876,mt=1.0;

However this doesn't seem to work in the x (horizontal) plane. Regardless of 
what I put for 'corrx', my distribution is not tilted in the x/xp plane.
It seems to tilt OK in the z (vertical) plane though.

Suzie
-- 
Scanned by iCritical.