DESY FEL Seminar Progress in Optics Setup / Pt.1

Transcription

1 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 1 DESY FEL Seminar Progress in Optics Setup / Pt.1 Mathias Vogt (MFL) & Johann Zemella (MPY1) Reminder: Motivation Reminder: Matching/Re Matching : Methods Reminder: Matching/Re Matching : FLASH Sections Side track: The Asymmetric Super Gaussian Typical Injector Matches 2018 First Injector Matches 2019

2 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 2 Reminder: Motivation BACCA cavity (3 GHz/warm) RF stations D chicane 3.9GHz SCRF 1.3GHz SCRF RF gun (1.3GHz/warm) injector laser (1&2&3) 5.6 MeV 146 MeV RF station 1.3GHz SCRF FLASH INJECTOR S chicane 450 MeV RF stations separation/collimation beam dump LOLA sflash (seeding) 1.3GHz SCRF FLASH MAIN LINAC CRISP (FL2) beam dump SASE undulators THz undulator Albert Einstein 350 MeV to 1250 MeV FLASH2 FEL user halls Kai Siegbahn pre ionization laser (25 TW) FLASHForward Exp. FLASH3 PETRA III RF GUN: space charge dominated regime (lowest E) Design Optics : starts exit of solenoid 1GUN zero current limit (no SC) valid only un compressed DBC2: match beam from RF GUN into optics in linac Dominant sources of optics perturbation (un compressed): ACC2 Badlands till FL2020+ energy profile (!!!) : k 1 := ( B y / x)/(p e) Necessity of re match constraints (Finally: correct for SC in compressed beam!)

3 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 3 Matching / Re matching of Beamlines: Basic Concepts (1) incoming beam reference point for initial cond. s_0 matching section perturbed reference point for measurement s_1 measurement section perturbed s_2,...,s_n outgoing beam = = modified perturbed linear optics! beam ellipse: z T B(s) 1 z = ɛ β(s) α(s) B(s) = α(s) γ(s) map M b a (k i,..., k j ) from s a to s b through quads Q i to Q j w/ strengths k i to k j. z(s b ) = M b a (k i,..., k j ) z(s a ) B(s b ) = M b a B(s a ) M T b a. linear in B, quadratic in M β b α b = T b a β a α a γ b γ a. T b a = T b a (k i,..., k j )

4 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 4 Matching / Re matching of Beamlines: Basic Concepts (2) incoming beam reference point for initial cond. s_0 matching section perturbed reference point for measurement s_1 measurement section perturbed s_2,...,s_n outgoing beam = = modified perturbed for each transv. plane (X & Y) : 1: measure beamsizes σ i σ(s i ) = β i ɛ, i = 2,..., n 4 Twiss parameters at reference point s 1 as least square solution of: σ 2 2 σ 2 n = ( T 2 1 ) 1, ( T n 1 ) 1, β 1 ɛ α 1 ɛ γ 1 ɛ 2: transport β 1, α 1, γ 1 backwards β 0 α 0 γ 0 = T 0 1 β 1 α 1 γ 1 gives measured initial cond. β 0, α 0, γ 0 Remark: T i 1 (i = 2,..., n), and T 0 1 are assumed close to! 3: match measured β 0, α 0, γ 0 into β 1, α 1, γ 1 β 1 α 1 γ 1 = T 1 0 (k i,..., k j ) using quads Q i to Q j β 0 α 0 γ 0 Remark: βγ + α 2 1 needs at least 4 quads for both planes (X & Y)

5 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 5 The (n 3) Screen/Wire Method incoming beam reference point for initial cond. s_0 reference point for measurement s_1 matching section > 4 quads (better 5 or more) measurement section outgoing beam s_2,...,s_n ; n >= 4 typically s 1 s 2 best: n 4 ;(n = 3 : no errorbars) screen: measure X & Y simultaneously optimum phase advance per screen/wire : 45 (0, 45, 90, 135 w/ 4 stat.) - fit (1:) might yield negative β 1 ɛ What can go wrong? : (apart from broken equipment... ) - large E spread & spurious dispersion - beam distorted / broken into beamlets how do σ s compare for totally different beam shapes? - screen : coherent OTR go to minimum E spread (m.o.l. on crest) & beautify beam (steering,... )! - small charge small signal

6 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 6 The Single Quad On Single Screen/Wire Method incoming beam reference point for initial cond. s_0 reference point for measurement matching section s_1 > 4 quads (better 5 or more) s_1_max measurement section entrance plane of scanned quad s_2 outgoing beam max. possible downstream s 1 is entrance of scanned quad concept: changed focusing shifts beam waist through s 2 (simplified evaluation for 1-quad-on-1-screen e.g. Minty & Zimmermann) for thin-lens quad: σ 2 2 (k 1 ) should be a parabola ( ) screen: measure X & Y simultaneously iff optics suited for (halfway) symmetric scan: (k 1 0 X & Y beam waists symmetric around s 2 scan in cycle direction! What can go wrong? : (apart from broken equipment... ) - Same as w/ (n 3) screen/wire method - parabola ( ) not sufficient can still yield imaginary β 1 hard to judge quality of scan until finally evaluated - FLASH: watch injector cycle vs. linac cycle

7 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 7 Modified Optics Upstream Screens During Measurement incoming beam reference point for initial cond. s_0 matching section perturbed reference point for measurement s_1 modified optics for measurement perturbed s_1_old measurement section s_2,...,s_n outgoing beam modified perturbed - while reducing number of quads for actual matching, modify optics between s 1 and s 2 + to optimize measurements (from ) i.e. for symmetric beam waist for single quad scan. can be applied to all before mentioned techniques since optics between s 1 and s (old) 1 modified from matching section prolonged further upstream See Johann s Pt.2!

8 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 8 Matching/Re Matching in FLASH/1/2 (1) modified QScan Q3DBC3 >OTR5DBC3 modified QScan Q8TCOL >OTR9TCOL ACC1 UBC2 DBC2 ACC2/3 UBC3 DBC3 ACC4/5/6/7 TCOL injector match / 4 Screen DBC2????? 4 Screen SFUND SFELC ORS SFUND SMATCH 5 6Wires UNDU UNDU MultiQuadSingleScreen Q1/2/3DBC3 > OTR5DBC3 FL2EXTR FL2SEED 4 Screen FL2SEED Injector Match: 4 OTRs in DBC2 5 quads upstream BC2 (try preserving waist in last dipole) all of them can be used after recent magnet upgrade (TQAS) in triplet! & 5 independently powered quads DBC2 : no longer needed! nice beam required or no convergence used to converge typically after 3-4 iterations recently (January 2019) lousy convergence, asymmetric reconstructed initial conditions, asymmetric and too large emittances??? but SASE works fine...

9 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 9 Matching/Re Matching in FLASH/1/2 (2) modified QScan Q3DBC3 >OTR5DBC3 modified QScan Q8TCOL >OTR9TCOL ACC1 UBC2 DBC2 ACC2/3 UBC3 DBC3 ACC4/5/6/7 TCOL injector match / 4 Screen DBC2????? 4 Screen SFUND SFELC ORS SFUND SMATCH 5 6Wires UNDU UNDU MultiQuadSingleScreen Q1/2/3DBC3 > OTR5DBC3 FL2EXTR FL2SEED 4 Screen FL2SEED Screen at SFUND: too small beam images at high E Symm. Single Quad Scan Pt.2 (JZ) 6 matching quads in ORS 5 6 wires in UNDU: servers being ported to (XFEL) µtca standard 6 matching quads in SFELC/SMATCH undulator match might violate LOLA constraints 4 screens FL2SEED: 6 matching quads in FL2EXTR and FL2SEED suffers extremely from spurious dispersion & from too small beam images at high E work in progress Symmetric Single Quad Scan

10 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 10 Sidetrack: Asymmetric Super Gaussian FEL beams: transv. spatial densities often non-gaussian Extension: asymmetric super Gaussian see how well it fits! Asymmetric Super Gaussian: [x/y] ρasg (q) := 1 N (s,,r) exp 12 q q0 s (1± ) ± := q q0? + : (in c notation :-) q := x, y qo R, s > 0, < 1, r>1 N (s,, r) so R [x/y] ρasg (q) dq 1 that σq2 = R [x/y] (q q0 )2 ρasg (q) dq, etc. r

11 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 11 Typical 2018 Injector Match (after 4 iterations) : 0.4nC/1.2mm, 308.5A, 52MeV/m, min-e-spread in BC2 4DBC2 6DBC2 8DBC2 10DBC2 β T = 2.46 m α T = ε T = 0.48 um β M = m α M = ε M = um β T = 2.66 m α T = 1.31 ε T = 0.52 um β M = m α M = ε M = um m P = 1.05 λ P = 1.36 m P = 1.01 λ P = (αt x + βt x )/ (εt βt ) (αt y + βt y )/ (εt βt ) x/ (ε T β T ) y/ (ε T β T )

12 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 12 Typical 2018 Injector Match (after 4 iterations) FLASH1 β (m) Unix version 8.51/17 02/11/ β x β y s (m) δ E/ p 0c = 0. Table name = TWISS

13 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 13 Typical 2018 Injector Match (after 6 iterations) 4DBC2 6DBC2 8DBC2 10DBC2 β T = 2.46 m α T = ε T = 0.49 um β M = m α M = ε M = um β T = 2.66 m α T = 1.31 ε T = 0.51 um β M = m α M = ε M = um m P = 1.00 λ P = 1.03 m P = 1.00 λ P = (αt x + βt x )/ (εt βt ) (αt y + βt y )/ (εt βt ) x/ (ε T β T ) y/ (ε T β T )

14 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 14 Typical 2018 Injector Match (after 6 iterations) FLASH1 β (m) Unix version 8.51/17 02/11/ β x β y s (m) δ E/ p 0c = 0. Table name = TWISS

15 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 15 same conditions as Remeasured night, (file restored) 4DBC2 6DBC2 8DBC2 10DBC2 β T = 2.46 m α T = ε T = 0.48 um β M = m α M = ε M = um β T = 2.66 m α T = 1.31 ε T = 0.53 um β M = m α M = ε M = um m P = 1.01 λ P = 1.13 m P = 1.01 λ P = (αt x + βt x )/ (εt βt ) (αt y + βt y )/ (εt βt ) x/ (ε T β T ) y/ (ε T β T )

16 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt One Single Rematch FLASH1 β (m) Unix version 8.51/17 09/11/ β x β y s (m) δ E/ p 0c = 0. Table name = TWISS

17 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 17 Remeasured after One Single Rematch 4DBC2 6DBC2 8DBC2 10DBC2 β T = 2.46 m α T = ε T = 0.48 um β M = m α M = ε M = um β T = 2.66 m α T = 1.31 ε T = 0.52 um β M = m α M = ε M = um m P = 1.01 λ P = 1.16 m P = 1.00 λ P = (αt x + βt x )/ (εt βt ) (αt y + βt y )/ (εt βt ) x/ (ε T β T ) y/ (ε T β T )

18 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 18 First 2019 Injector Match Campaign (Jan 2019) I sol /A GUN-GRAD.SP init MiMaAmpl init emitt final MiMaAmpl final emitt m/mv x/y x/y /µm x/y x/y /µm / / / / / / / / / / / / /1.1* 0.9/ / / / /0.91 *: had to stop before convergence was reached

19 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 19 L2, 0.4 mm, 52 MV/m, A, orbit optimized; n plane emit (um) MiMaAmp comment 0 x: start 0 y: start 1 x: converging 1 y: converging 2 x: converging 2 y: converging 3 x: worse! 3 y: slightly worse 3a x: consistent to 3 3a y: consist. 3 3b x: consist. 3/3a 3b y: consistt. 3/3a 2a x: back to 2 / cons. 2 2a y: back to 2 / cons. 2 3c x: unfortunately cons.. 3/3a/3b 3c y: cons.. 3/3a/3b x: even better 5 y: even better 6 x: slightly worse 6 y: slightly worse 7 x: same 7 y: xtrmly good 8 x: BEST STOP 8 y: GOOD

20 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 20 First 2019 Injector Match Campaign (Jan 2019) L2, BSA: 1.3mm, gun: 52 MV/m, 0.8deg, I sol =310.3A final emit x/y /µm: 0.51/0.86 ; final MiMaAmpl x/y: 1.47/1.10 L2, BSA: 1.2mm, gun: 52 MV/m, 0.8deg, I sol =310.3A final emit x/y /µm: 0.57/0.87 final MiMaAmpl x/y:1.56/1.02 L2, BSA: 1.1mm, gun: 52 MV/m, 0.8deg, I sol =310.3A final emit x/y /µm: 0.61/0.88 final MiMaAmpl x/y: 1.33/1.04 L2, BSA: 1.0mm, gun: 52 MV/m, 0.8deg, I sol =310.3A final emit x/y /µm: 0.63/0.83 final MiMaAmpl x/y: 1.39/1.11 Luckily : SASE is was very nice in 2019!!! work in progress!

21 DESY FEL Sem / M.Vogt (MFL) J.Zemella (MPY1) / Optics Setup Pt.1 21 Than you for listening to part one! Next: part two!!!