search

MethodicalAcceleratorDesign / MAD-X

MAD-X repository
Other
49 stars 39 forks source link

makethin issue when using TWISS module #1130

Closed gusimon closed 2 years ago

gusimon commented 2 years ago

Hello everyobdy,

While using a script to obtain a matched sliced ttbar lattice given by Léon Van Riesen-Haupt I came across an issue. I tried to use makethin in order to slice my rfcavities, dipoles, quadrupoles and sextupoles (the aim here is to slice my multipoles before launching TWISS and then to analyse the tune of the beam). The lines I'm talking about are between lines 85 and 101. And when running madx, just after the beginning of the makethin sequence and before the TWISS sequence. The program crashes and gives me the negative drift error shown in the picture attached.

Léon wasn't sure about the issue so he advised me to address the issue here. Sorry if It's not relevant and thank you in advance.

rematch_makethin.madx.txt makethin_error

rdemaria commented 2 years ago

It could be a bug, that might be solved by adding seqedit, sequence=fcc_p_ring;flatten;endedit; just before or just after makethin.

HelmutCERN commented 2 years ago

I had a quick look at this and noticed inproper used of MAKETHIN in several places: 1) select, flag=makethin, class=rfcavity, slice = 1, thick = true ; Thick cavity slicing is not implemented. The FCC-ee cavites are actually thin (have no length). If they would be long you could turn of slicing by selecting 0 slices. 2) SELECT, FLAG=makethin, SLICE=3; No element is selected here. What do intend to do ? 3) MAKETHIN, SEQUENCE=FCCEE_P_RING, STYLE=TEAPOT, MAKEDIPEDGE=false; here you slice the already sliced sequence. This should really not be done.

As a test I removed 1 - 3 and ran without error.

gusimon commented 2 years ago

I did the same and it worked fine, thank you very much !