Asymmetric linear optics with quadrupole fringe fields

[wei0852]

Dear all,

I'm trying to calculate linear optics with quadrupole fringe field integrals.
I noticed the linear optics comes out asymmetric, which is weird because the lattice elements are still symmetric.
I also checked the result of elegant, which gave more reasonable, symmetric linear optics.

Here are the MATLAB scripts to reproduce the result:

test_fint_asymmetry.txt

sevenBA.txt

ati = atindex(THERING1);
for j=1:length(ati.Q1)
    THERING1{ati.Q1(j)}.fringeIntM0 = [-0.005772312523804592 4.41250468033291e-05 -4.787169625888183e-07 6.715551627497556e-09 8.917345548597336e-08];
    THERING1{ati.Q1(j)}.fringeIntP0 = [0.004789430951781449 0.0001128040953067462 5.806584758045332e-06 5.691680359138711e-07 2.630893399461595e-07];
    THERING1{ati.Q1(j)}.FringeQuadEntrance = 2;
    THERING1{ati.Q1(j)}.FringeQuadExit = 2;
end

Did I mess up something in my settings?

ELEGANT:

[lfarv]

@wei0852 : thanks for documenting the problem. #1081 is an attempt to fix it. It's waiting for @carmignani's approval but you may test it in the meantime.