Foxification of e-lenses (final) #1056

amereghe · 2020-02-24T15:14:47Z

This PR aims at finalising the foxification of e-lenses, as discussed with @rdemaria recently.
For Gaussian and uniform electron beam distribution, I have added the explicit expression for R1=0, both in the FOX and in the tracking part.
In addition, for the Gaussian distribution, I have added a linearisation of the kick of the full lens when r<1E-7 mm.

I have added few tests, each checking that the de-tuning for a single full e-lens with a specific profile and kick is sound.
The check is done reducing all the beam distributions to a constant one, and comparing to the tune-shift of a quad with the same normalised kick.

At the same time, I allow the use of the elens in the very degenerate case of the wire.
I have added a test for that - checking only kick during tracking and in FOX, but no specific test on tune-shift.

At the same time:

more human readable output at various levels;
removed bug on Gaussian and radial profiles, where the normalisation geometrical factor was wrongly taken into account twice.

… distribution read from file AT the same time: * added units to debug echo of input parameters; * more readable code for calculating theta_r2;

* handling of empty lines in files of radial profile; * change of printout in echo of radial profile + warning about point at R=0.0; * added debug lines in fox calculation with radial profile;

…sian profiles

…g of geometrical normalisation factor

These follow the addition of a warning line when echoeing the profile in STDOUT

…enses, Rref must be introduced as reference radius at which the theta_ref is computed, instead of R2 and theta_r2

amereghe · 2020-02-26T12:07:11Z

Hi @rdemaria ,
I think this is the final version.
I haven't updated the original elens tests - I will wait for your green light before doing that, such that I update the tests only once.
That's why circleci fails.
I will do the same with the documentation.

Please find here the updated .pdf with all the math - now it should be correct.
I think that the most interesting slides for you are 2 and 6.
Moreover, you can find on slide 8 the tune shifts given by the different types of e-lenses tested as calculated analytically and as computed by FOX.
Are they acceptable?

…n repo space occupancy

amereghe · 2020-02-27T15:27:12Z

ok - I have updated the slides, showing also that (last slide):

theta_prime(r) at R1 for the uniform distribution is different if the lens is hollow or full;
theta_prime(r) at R2- for the Gaussian distribution tends to the limit at R2+ if R2 tends to infinity.

I have also compiled a more detailed cross-check of the tune shifts computed by FOX for the uniform, Gaussian and radial distributions, evaluating what happens also when the kick is lowered by a factor 10 - please find the table here. The tune shift predicted by SixTrack agrees much better to that predicted analytically, showing that the implementation is sound and the tests use quite strong kicks. I will re-set the tests to use the weaker kicks, then.

@rdemaria , if this is all fine for you, I will then update the documentation and the elens tests that were already existing, and then merge :)

…ear tune-shifts with equivalent quadrupolar gradient

…red by a factor 10

… with radial profile from file nearby r=0

amereghe · 2020-03-02T14:02:25Z

ok, @rdemaria , I am done here.
You can accept the PR.
I will draft all the math in a specific report, and add in the documentation a reference :)

tpersson

It looks good to me, but I haven't checked in details. The things I checked all looked good!

amereghe · 2020-03-05T11:36:45Z

Hi @tpersson ,
Thanks for the feedback. Shall I wait for your detailed review or shall I merge the PR?
In addition, FYI, following the suggestion of @rdemaria , I am de-facto drafting a report on the physics behind the implementation, but the repo will come much later than this PR...
Thanks!

amereghe · 2020-03-09T13:25:25Z

ok to merge, as agreed with @tpersson

amereghe added 8 commits February 20, 2020 13:05

better handling of elens kicks during tracking

7ad11ea

propagating to the FOX part changes from the previous commit

0996938

accepting and using degenerate e-lenses to a wire

c956c1b

removing wrong calculation of theta_r2 for Gaussian lens and for beam…

a2f5797

… distribution read from file AT the same time: * added units to debug echo of input parameters; * more readable code for calculating theta_r2;

Minor changes:

7c55384

* handling of empty lines in files of radial profile; * change of printout in echo of radial profile + warning about point at R=0.0; * added debug lines in fox calculation with radial profile;

further debug messages in elens_kick_fox

251cb48

adding dedicated tests for fox calculation in case of radial and Gaus…

5652ba0

…sian profiles

same fort.2 in single FOX tests

7f31482

amereghe requested review from rdemaria and tpersson February 24, 2020 15:15

amereghe added 5 commits February 25, 2020 16:47

fixing expected kick in elenses for test ramp_DYNK, following cleanin…

890e0a8

…g of geometrical normalisation factor

fixing plotting script parsing radial profile as echoed by 6T

7737445

These follow the addition of a warning line when echoeing the profile in STDOUT

allowing for a WIRE type of elens

3f853b1

In order to have the WIRE as a fully consistent degenerate case of el…

5b25327

…enses, Rref must be introduced as reference radius at which the theta_ref is computed, instead of R2 and theta_r2

adding test of degenerate elens to a WIRE

15483d7

changing type for a couple of fort.2 files in elens tests, to cut dow…

338640f

…n repo space occupancy

amereghe added 8 commits February 27, 2020 16:38

updating elens FOX tests of tune-shift to smaller kicks, and more lin…

79663e2

…ear tune-shifts with equivalent quadrupolar gradient

updating README file of elens tests with the kicks for FOX tests lowe…

6ad9b95

…red by a factor 10

forcing 3-points interpolation in case of tracking/FOXING a full lens…

f197424

… with radial profile from file nearby r=0

updating elens tests

d61de1a

Merge branch 'master' into eLenses_step3

e7492ae

updating CHANGELOG.md

3e87058

updating user manual

3f4098c

updating physics manual

14da71c

rdemaria approved these changes Mar 4, 2020

View reviewed changes

tpersson approved these changes Mar 5, 2020

View reviewed changes

amereghe merged commit 4d51731 into SixTrack:master Mar 9, 2020

amereghe mentioned this pull request Mar 12, 2020

Foxifying chebyshev polynomials maps #849

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Foxification of e-lenses (final) #1056

Foxification of e-lenses (final) #1056

Uh oh!

amereghe commented Feb 24, 2020 •

edited

Loading

Uh oh!

amereghe commented Feb 26, 2020 •

edited

Loading

Uh oh!

amereghe commented Feb 27, 2020 •

edited

Loading

Uh oh!

amereghe commented Mar 2, 2020

Uh oh!

tpersson left a comment

Uh oh!

amereghe commented Mar 5, 2020

Uh oh!

amereghe commented Mar 9, 2020

Uh oh!

Uh oh!

Foxification of e-lenses (final) #1056

Foxification of e-lenses (final) #1056

Uh oh!

Conversation

amereghe commented Feb 24, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

amereghe commented Feb 26, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

amereghe commented Feb 27, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

amereghe commented Mar 2, 2020

Uh oh!

tpersson left a comment

Choose a reason for hiding this comment

Uh oh!

amereghe commented Mar 5, 2020

Uh oh!

amereghe commented Mar 9, 2020

Uh oh!

Uh oh!

amereghe commented Feb 24, 2020 •

edited

Loading

amereghe commented Feb 26, 2020 •

edited

Loading

amereghe commented Feb 27, 2020 •

edited

Loading