# 2018-05-25

From Administrative

**Director - M. Spata**

*No report*. Acting Accelerator Department Head

**Accelerator Physics & Acting CASA Department Head - T. Satogata**

*No report*.

**Accelerator R & D - Y. Zhang**

Fanglei Lin

- Continued exploration of JLEIC collider rings' geometry and presented in the JLEIC R&D meeting.
- Implementing vertical doglegs in the electron collider ring.
- Working on the pCDR documentation.
- Miscellaneous: participated HE-JLEIC discussion, attended CASA group meeting, attended the JLEIC R&D meetings, attended JLEIC detector meeting.

Vasiliy Morozov

- May 12 - May 25
- JLEIC IR design: editing of Guohui's IR design paper completed, first draft of the IR section for the pCDR completed, development of the detector background section outline
- JLEIC polarization: teleconference with Anatoliy's group, work on the polarization section of the pCDR, report on the effect of the potential energy increase on the ion polarization scheme
- Preparing and giving talks on the JLEIC design and R&D at the JLEIC advisory committee meeting and INFN delegation meeting
- Collider rings' design: implementation of the compact solenoid decoupling scheme developed by Anatoliy's group
- Geometry tagging LDRD: working on getting the Sartre event generator working on JLab's farm

- May 26 - June 8
- Writing of pCDR sections
- Collider ring design consistent with the available footprint

Amy Sy

- Working on the closeout report for skew parametric-resonance ionization cooling (Skew PIC). Summarizing the work that was done and results obtained.
- Working on the JLEIC ion sources chapter for both the pre-CDR and the ICFA beam dynamics newsletter contribution.

Guohui Wei

- On Leave

Yuhong Zhang

- Explored RHIC magnets for the JLEIC ion collider ring
- Organized study on high-energy JLEIC, worked on several issues (ion injector, baseline, luminosity estimate)
- Worked with Fanglei on figure-8 footprint
- Attended several meetings to discuss preCDR
- Organized writing a contribution to ICFA Beam Dynamics newsletter
- Continued to write preCDR

**Computational Physics - Y. Roblin**

Kirsten Deitrick

- Last two weeks:
- Magnetized beam (LERF PoP):

- -re-phasing for FEL boosters after correcting buncher setting
- -discovering the correct off-crest phase for booster cavities
- -zeroing after FEL booster
- -discontinuing work on UITF; since it's being installed in CEBAF, it can't be used in the LERF PoP

- Straight Merger at CBETA:

- -beginning data analysis, getting GPT decks set up

- CSR:

- -continue working on KEK CSR tech note (minimal)

- eCloud:

- -pull ion beam sizes/parameters from presentations; feed into PyECLOUD
- -email CERN questions related to PyECLOUD results
- -email people (J.L. Vay, M. Furman, Amundson, Crittenden) about codes (WARP+POSINST, POSINST, different POSINST, ECLOUD)

- Impedance:

- -general meeting
- -smaller meeting specific to pre-CDR documentation
- -schedule next impedance meeting (June 25th)

- Meetings:

- -Impedance (general)
- -Impedance (pre-CDR)
- -CASA
- -Cooler x2 (verbal preliminary on Straight Merger at CBeta results, -)
- -CBETA weekly meeting x2
- -JLEIC x2 (verbal preliminary on Straight Merger at CBeta results, -)

- Misc:

- -Open House training
- -paper maintenance (finding new, organizing all)
- -helping Sajini with automation and space charge settings for ASTRA (Magnetized GTS)

- Next two weeks:
- Magnetized beam (LERF PoP):

- -evaluate skew quad position/strength on correcting booster asymmetries
- -start scan of charge/length/size

- Straight Merger at CBETA:

- -continue data analysis
- -begin rough simulations (actual simulations, if I can get the field map from George Biallas)

- CSR:

- -continue working on KEK CSR tech note (minimal)

- eCloud:

- -check with Vasiliy about other beam sizes / magnet strengths (outside of arc cell)
- -run cases with variations
- -start looking at instability thresholds?

- Meetings:

- -CASA
- -Cooler
- -JLEIC (maybe)

Yves Roblin:

- CEBAF:
- -chicane remodel
- -BTEAM
- -autosteer specs
- -spin calculations for fall2019, writing document for new spin tool

hc: 10573.4 MeV/c hb : 10575.8 MeV/c ha: 2218.39 MeV/c

- -note on ced2elegant floor calculation
- -estimates /simulations for 2R coupling
- -optics cleanup, CED bug squishing
- -preparation for half energy (SL off ) run for Hall A experiment , ARC1/ARC2/1C/3C strings cross-calibrations
- -ELEGANT update on ops to version 34.0 (Anthony did the update, I did the checking)

- JLEIC:
- -starting pulling together stuff for the pCDR beam-beam

- OTHER:
- -comp. phys. journal review, Nim A review.

Dennis Turner

- CEBAF:
- -CED2ELEGANT. Fixed a bug where termination points in the Hall D proper would lead to incorrect lattices. Tracked down to a bug in the CED API.
- -CED2ELEGANT. Fixed a bug where the extraction magnets MYA's would appear in recirculating lattices. Tracked down to a bug in the ZONE definitions in the CED.
- -AutoSteer. Started working on AutoSteer from the prototypes and notes From Y. Roblin
- -CED2ELEGANT. Adding support for generating the starting floor coordinates and angles in the generated command files

He Zhang

- Fast multipole method. The paper on Traceless Cartesian Tensor FMM for Coulomb Kernel was finally accepted. Now the paper is in pressing. An online version is available here:

https://www.sciencedirect.com/science/article/pii/S0021999118303280?via%3Dihub A parser was added to the kernel-independent FMM code. Now the code can read a kernel from a text file and perform the calculation without recompiling. Computation time is slightly increased but acceptable.

- I started to write the paper on the kernel-independent FMM based on the Cartesian tensor and Differential Algebra. Some data was generated. However, I noticed that the efficiency is not as good as expected, although DA package has been revised for better performance. After profiling the program again, it is identified the bottle neck is the Multipole-to-Particle (M2P) operator and the Particle-to-Local (P2L) operator for ill-separated boxes. Usually the dominating operator is Multiple-to-Local (P2L). But in this case, the M2P and P2L operators includes many DA calculation for high order derivatives of the kernel function and the DA calculation is time consuming, these two operators cost much more time than the M2P operator. Possible solutions are (1) use single level FMM, which does not need M2P and P2L operators, instead of multi-level FMM; (2) use some DA approximation around the box center for the high order gradient calculation; and (3) further improve DA efficiency. To test solution (1), I revised the FMM framework so that a full tree of boxes for a given level can be generated. For a uniform distribution, a single level FMM is about 20 times faster than a multi-level FMM. This confirms that the computation bottle neck is due to the M2P and P2L operator. But the efficiency of a single-level FMM reduces very fast when the particle distribution deviates from the uniform distribution. Solution (2) will bring new errors. It will need testing to see how much the error will be increased and how much the efficiency can be improved. As to solution (3), the DA calculation efficiency can be improved by faster indexing of the DA element. But I think what we can gain is limited.
- To demonstrate the kernel-independent property, we should present the calculation results for various kernels. For 2D solutions of the modified Laplace equation, the modified Stokes equation, and the modified Navier equation, one needs to calculate the first and the second kind modified Bessel functions K0 and K1. The current DA package cannot do this calculation. After some investigation, I found K0 and K1 can be calculated by the rational approximation method (Pavel Holoborodko), which is implemented in the boost library. The method should be adaptive to DA calculations.

pCDR writing

- A draft including the following topics has been finished: Multi-Stage Cooling Scheme, Simulation Program (JSEPC) Development, DC Cooling in the Collider, Bunched Beam Cooling in the Collider Ring. The following two sections will be added soon: DC Cooling in the Booster Ring (Needs some simulations for this topic.), Optimization of the electron cooling efficiency.
- Others
- -Reviewed two papers
- -Work plan in the following two weeks
- -Continue writing and polishing the pCDR report
- -Some coding for the kernel-independent FMM
- -Dust particle transport
- -Understand the dispersive cooling theory

**Diagnostic Development - K. Jordan**

*No report*.

**LERF - S. Benson**

*No report*.