Progress with MAD-NG and highlights from IPAC23
LNO section meeting on Progress with MAD-NG and highlights from IPAC23
News:
The LHC is running well and the ATLAS/CMS luminosities appear now to agree within 1-2%. See also yesterday's LBOC.
MAD-NG progress, Laurent slides
Laurent gives a detailed presentation on BENDS in MAD-NG and compares with PTC (exact = true, time=true, icase = 56). He comments on different methods used for Bend maps, Drift-Kick-Drift (DKD), Thick-Kick-Thick (TKT) and Kick-Map-Kick (KMK). For PTC SBEND and (true) RBEND he uses the “Old” thick maps based on E. Forest, rather than the “New” method based on D.Sagan that suffers from major numerical instabilities for small bending angles. He first concentrates on the SBEND (curved sector bend) and observes generally excellent agreement between PTC and MAD-NG to the 1e-13 level and good convergence of thin versus thick maps models.
He then illustrates and discusses the validation including misalignment of the different types (curved, true, parallel) of RBENDs and length parameters defined in MAD-NG, and reminds that MAD-X translates RBEND into an SBEND with adjusted pole face angles (equal to half of be bending angle). Many improvements were also done on the MAD-X PTC level to enable comparison with MAD-NG that will be made available as pull-request to MAD-X github.
Rogelio asks about the status and schedule for the next MAD-NG release. Laurent says the next release implementing the features discussed today should be ready in about a month. He also plans to make a presentation on the quadrupole and the solenoid in MAD-NG soon.
Solenoid definition, Helmut, slides
Triggered by MADX issue#1147 opened by Riccardo, Helmut prepared a pull request to warn and if possible remove inconsistencies in solenoids specified on the MAD-X input level. So far, MAD-X uses a single element type SOLENOID that is handled in the TWISS module in separate routines as thin solenoid if the length parameter L=0 and as thick solenoid otherwise.
Riccardo proposed to look as alternative to dedicated SOLENOID attribute checking/fixing into adding a new THINSOLENOID element defined with KSI and LRAD. Helmut said he is interested to look into the implementation of this alternative proposal as a modified or new pull request.
Rogelio stresses the importance of documenting well the solution that will be chosen. Laurent doubts the need for a new element and says MAD-NG just checks/warns for zero length and non-zero strengths in all elements and gives the exemple of the quadrupole that checks for nonzero k1,k1s,k0,k0s when l is zero.
Riccardo will also ask the FCC community for feedback.
Highlights from IPAC23, LNO participants, slides
The LNO participants of IPAC23 all agreed that they found this years IPAC particularly interesting, stimulating and well attended (1600 participants). Rogelio mentions excellent CERN presentations (LIU - Mallika, HL-LHC, Oliver) and the progress on high field magnets recognized with a prize for Xingchen Xu.
As illustrated in his slides, Felix got the impression that the accelerator field is really booming, in many large and in particular also in many smaller very interesting projects. A guide on the many new projects can be found on the web. The planned EIC collider appears to be very interesting and challenging, and there is also a new linear collider concept, the Cool Copper Collider or short C3. There appears to be much interest and progress in accelerator driven systems (ADS), and we can expect to see in future a multitude of very innovative compact accelerator systems.
Riccardo made pictures of some interesting poster contributions and mentions “fringe field effects in solenoid rings”, “generalized gradients in BMAD”, and a library for magnetic field processing of interest for construction of maps with multipoles.
Rogelio comments that Christoph Heßler / GSI, previously CERN/AWAKE showed a nice graphical interface to display, change and rematch lattices.