M2 Tuning procedure for COMPASS

  Lau Gatignon / SL-EA


This document is in preparation

THE MUON BEAM

After having verified all the items on the M2 checklist, the tuning is done in three successive stages :

1) horizontal and vertical dipole steering,
2) horizontal and vertical quadrupole focussing,
3) halo optimization.

The procedure is the same for the different optical modes.

For beam tuning it is conventient to look at a picture of the optics. Such a picture can be copied from the Web (with FTP). The optics drawing (Postscript files) comes in two parts:

  • Hadron section from T6 up to the hadron absorber,
  • Muon section from the hadron absorber to the experiment.
  • 1. Dipole steering

    Dipole steering is done in two steps :

    i) with a "monochromatic pencil beam":

    - close the pion momentum slit COLL's 1 and 3 to ± 2 mm , giving a momentum spread Dp/p = 5 mm/% );
    - close the muon momentum slit SCR 4 and 5 to ± 20 mm , giving a momentum spread Dp/p = 12 mm/%);
    - adjust beam flux with COLL 2 (4) and SCR 3 to obtain beam profiles in the MWPC's with sufficient statistics.

    ii) with the full momentum acceptance to correct for second order centroid shifts :

    - COLL's 1 to 4 and all SCR's set to nominal values.

    The steering procedure described in points (4.1.1) to (4.1.5) below is followed through for the two steps in turn.

    1.1. Definition of central momenta

    A setting of the muon beam is characterized by two central momenta p(p) and p(m). The pion central momentum p(p) is defined in the horizontal plane of the hadron section as the momentum of a pencil ray from the target which is on axis in collimators 1 and 3 for nominal setting of BEND 1. The muon central momentum p(m) is defined by the momentum station magnet BEND 6 (vertical plane) as the momentum of a pencil ray which is on axis at SRC 4 or HOD 1 and at HOD 4.

    1.2. Steering into and through the decay FODO channel

    Horizontal plane : Steer with TRIM 1 to centre the beam in MWPC 1 and with BEND 3 to centre the beam at MWPC's 3, 5 and 7.

    Vertical plane : Centre the beam in MWPC 2 with BEND 2 and use TRIM 2 to centre at MWPC's 4, 6 and 8.

    In case of difficulties to centre the beam at all four positions simultaneously, optimize at the entrance and exit of the FODO channel (MWPC's 1, 2 and 7,8). Do not forget to move the chambers out of the beam after this part of the tuning procedure as to reduce radiation damage to a minimum!

    1.3. Steering onto the hadron absorber

    The Beryllium absorber has a diameter of 55 mm, not much larger than the muon beam, so that careful steering is important. The vertical wire chamber MWPC 10, just downstream of BEND 4, is from now on centered onto the nominal beam axis.

    Horizontal plane: steer with TRIM 3 to centre beam at MWPC9,
    Vertical plane : steer with TRIM 4 to centre onto MWPC 10.

    Go back to 'normal' muon beam operation, i.e. switch on BEND's 4 and 5.

    1.4. Steering through the Beam Momentum Station (BMS)

    One pair of steering dipoles are used in each plane and the two sets of wire chambers MWPC 11 to 14 are available to control the steering. The phase advance between the spectrometer magnet BEND 6 and the two horizontal (TRIM 5, TRIM 6) and two vertical (BEND 4/5, TRIM 7) steering elements allow for orthogonal steering in both planes . The two pairs of wire chambers are symmetrically planed 24.5 m upstream and downstream from the mid-point of the spectrometer magnet and allow for direct control of the position and angle of the beam in BEND 6 (or the position in the four hodoscope planes).

    Vertical steering : Control the beam angle at BEND 6 with BEND 4 and/or 5 by adjusting the peaks in the beam profiles in MWPC 12 and 14 to the same value. Control the beam position at BEND 6 by centering the two distributions using TRIM 7.

    Horizontal steering: Use wire chambers MWPC 11 and 13, control angle with TRIM 5 and the position with TRIM 6.

                  	Position at Bend 6   	Angle at BEND 6

    Horizontal TRIM 6 TRIM 5 Vertical TRIM 7 BEND 4/5

    1.5. Steering to the experiment

    The horizontal steering to the experiment is done with BEND-7 and BEND-9. Center first the beam on MWPC-17 with BEND-7, sensitivity 1.6 mm/mrad, i.e. 36 Amps/mm at 100 GeV/c. If necessary the beam position at MWPC-17 can be adjusted with TRIM-6. Then center on the NA58 target using an experimental scaler and BEND-9. Sensitivity 26 mm/mrad, i.e. 1.8 Amps/mm at 100 GeV/c. The horizontal centering of the beam at the NA58 target is done with BEND-8, sensitivity 6.8 mm/mrad or 8.5 Amps/mm.

    In case the target dipole is switched on, its field must be compensated with a linear combination of BEND-7 and BEND-9. The sensitivities are as follows:

    Steering element Effect at center BEND-9 Effect at NA58 target
    BEND-7 7.3 mm/mrad 43.413 mm/mrad
    BEND-9 - 25.92 mm/mrad

    Therefore the linear combination (in BL) BEND7 - 1.67 BEND-9 changes the angle at the NA58 target without affecting the position. A 1 mrad kick at BEND-7 leads to a 2.75-1.67*1.867 = -0.377 mrad change at the NA58 target.
    In terms of currents, the following linear combinations are relevant:

    B7 - 1.374 B9 : 0 mm/mrad 280 A/mrad at NA58
    B7 - 1.208 B9 :0 mrad/mrad20.8 A/mm at NA58
    B7 alone :
    2.5 A/mm at NA58


    2. Quadrupole tuning (high-energy mode)

    2.1. Tune acceptance stage and match to FODO decay channel

    a) Tune vertical focus at BEND 2

    - Close momentum slit COLL 1(3) to ± 2 mm (Dp/p = 5 mm/%);
    - Measure vertical beam profile by scanning with COLL 2 and 4, count on TRIG 1,
    - Displace focus downstream and remeasure profile: Reduce Q3 and Q6 currents (in absolute value) by the fractions indicated below to move the vertical focus downstream by 1 meter:

    			Q3	by 	2.20 permille
    			Q6	by	4.53 permille 
    

    - Choose quadrupole current giving smallest spot.

    b) Check horizontal dispersion compensation at BEND 3

    - Close COLL 2 and 4 to ± 10 mm;
    - Open COLL 1 to ± 50 mm;
    - Scan 5 mm slit of COLL 3 at several points over the full aperture of BEND 2 (± 50 mm) and observe horizontal beam profile in MWPC l;
    - If the profile centre in MWPC 1 changes significantly for large positive and negative Dp/p, correct dispersion compensation with QUAD 8 (to increase the horizontal dispersion downstream of BEND 3 by 1 mm/% an increase of the absolute current of Q8 by 5.4% is required). The dispersion is correctly compensated if the beam at MWPC 1 moves by 10% of the displacement at the collimators.

    2.2. Tune phase ellipses at entrance of decay FODO

    To minimize beating along the FODO channel the horizontal and vertical phase ellipses (x, x', y, y') should be upright at the start of the FODO decay channel. This can be controlled in two steps.

    a) Maximize beam size at FODO entrance

    Measure horizontal and vertical beam size at MWPC 1 and MWPC 2, respectively, and increase to fill available aperture.

    sx = ± 60 mm (aperture of QUAD 11)
    sy = ± 28 mm (aperture of BEND 3)

    For an orthogonal movement and in order to keep the dispersion correction unchanged, QUAD 7 to QUAD 11 in the horizontal plane and QUAD 8 to QUAD 11 in the vertical plane must be used together. The appropriate coefficients are listed below (in permille per mm increase):

    	Quad	Horizontal	Vertical	
    	Q7	  -51	 	   -	
    	Q8	 14.6	 	  8.4		
    	Q9	 -3.5		  3.1	
    	Q10	  2.2		 -4.6
    	Q11	  1.1		 -1.1	
    

    b) Minimize beam angular spread at FODO entrance

    As MWPC 3/4 is placed at 90o phase advance downstream of MWPC 1/2, the beam size in the former is a measure of the beam angular spread at the position of the latter.

    Minimize horizontal and vertical beam size at MWPC 3 and MWPC 4, respectively, in order to minimize beam angular spreads at the FODO entrance (or to rotate phase space ellipse to an upright position).

    This (as under point (a)) is done with a appropriate combination of current changes in QUADs 7 to 11. Check that beam size in MWPC's 1 and 2 do not change. In particular watch beam intensity, as in particular vertical losses in this region of the beam line can lead to HALO in EHN 2, which cannot be influenced later on. The coefficients (again in permille per mm increase in MWPC 3+4):

    	Quad	Horizontal    Vertical	
    	Q7	  -44	         64
    	Q8	   12	          3
    	Q9	   17	         47	
    	Q10	  -59	       -156	
    	Q11        54	        -42	
    

    An iteration between points (a) and (b) may be necessary to find best conditions.

    2.3. Focus onto the absorber

    The quadrupoles QUAD 19 to 21 must be tuned to focus the muon momentum p(m) onto the absorber. Note that MWPC 9+10 are aligned on the undeflected pion trajectory (i.e. as if B4 were switched off).

    - Switch off BEND 4 + 5
    - Tune the horizontal focus with QUAD 20 and 21 to minimise spot size in MWPC 9 by reducing the fields in Q20 and Q21 (in absolute value) by the following fractions

    		Q20	by	1.79 percent
    		Q21 	by 	3.67 percent
    
    - Tune the vertical focus by minimising the spotsize in MWPC 10, reducing the fields (in absolute values) by fractions:
    		Q20	by	2.11 percent
    		Q21	by	19.3 percent
    
    - Switch on Bends 4 and 5 again.

    As the flux of pions is about 20 times larger than the muon flux (upstream of the absorber), it may be useful to tune the focussing with pions. In that case the tuning of the focus should be preceded by the following steps, strictly in the order described:

    - Scale currents of Q17 to Q21 up to p(p)
    - Close SCRAPERs 2 and 4 to ± 20 mm
    - Close COLL 2 and 4 to reduce the flux to acceptable rates (± 5 mm each, to start with, increase flux if necessary to obtain reasonable MWPC profiles).
    - Move out all absorber modules (1 to 9) in TUNE/SPECIAL/ABS (this may require intervention of a absorber hardware specialist).

    After tuning of Q20-Q21, go back to normal muon beam operation and scale Q17-Q21 to muon beam energies.

    2.4. Tune beam waists at the momentum station

    a) Check vertical dispersion compensation.

    A crude check of the dispersion compensation in the vertical plane at BEND 4 can be made by observing the centre of the vertical beam profile in MWPC 14 for different openings of the muon momentum slit SCR 4. Vary SCR 4 from ± 20 mm to ± 100 mm (R36 = 13 mm/%). If the dispersion is found to be only partly compensated, correct with QUAD 24 and QUAD 28 ( Q28 ~ Q24 x 33.69/24.0).

    b) Tune verticale focus at BEND 6.

    Use QUAD 29 and 30 to displace focus at BEND 6. Observe vertical beam profiles in MWPC 12 and 14. Minimize beam spot in both chambers (or use hodoscope phase plot to upright horizontal phase-space ellipse at centre of BEND 6). Reduce

    			Q29	by	3.88 percent
    			Q30	by	45.0 percent
    
    (in absolute value) to move the focus 1 meter downstream.

    c) Tune horizontal focus at BEND 6.

    Like point (b), but use MWPC's 11 and 13. The focus is moved downstream by 1 m by reducing Q29 and Q30 (in absolute values) by the following fractions:

    			Q29	by	3.80 percent
    			Q30	by	22.6 percent
    

    2.5. Tune to focus beam at polarised target and polarimeter.

    Q33 - Q34 may be tuned empirically by minimising the near beam halo with respect to beam intensity. In fact the passage of the beam through the small aperture quadrupoles Q35 and Q36 is rather critical in terms of halo. The effect on the focus is rather small, to the extent that Q33 and Q34 are much weaker than Q35 and Q36.

    If the beam at the target is too large, a variation of Q36 alone may lead to a better compromise between spot size and divergence.

    3. Halo optimization

    3.1. General remarks

    The natural muon HALO in the M2 beam reaching the experimental hall and integrated over a typical area of 8 x 5 m around the beam (corresponding to the large veto wall that existed in experiment NA 2) is about 75 % of the beam intensity. Of these muons about 13 % originate from along the decay FODO, about 80 % from the absorber region, further 3 % from the second FODO and the last 3 % of HALO muons leave the beam aperture in and downstream of the momentum station.

    To make the beam a useful tool for counter experiments this HALO must be reduced to several percent (typically < 15 %) of the beam intensity. This is done with specially designed magnetic collimators (scrapers). These are 5 m long magnetised iron blocks with a toroidal field of ~1.5 T and a remotely adjustable field free gap for the beam . They are supposed to scrape off the muons surrounding the "good" beam and deflect them away-from the beam line. Seven of these scrapers are installed along the beam line, two with horizontal jaws , five with vertical jaws. Six scrapers (2H + 4V) are to be found around the absorber, the source of the majority of HALO muons, one downstream of the momentum station. Their action is enhanced by two sets of Magnetised Iron Blocks (MIBs) following the scrapers.

    The stray-field inside the aperture of the magnetic scrapers has an appreciable quadrupole component, which can perturb the muon beam. For the nominal setting of 100 A (15 KG), the perturbations are small and can, for momenta > 100 GeV/c, be neglected.

    3.2. Magnetic scraper control

    Each magnetic scraper has five variables to control :

    - one current (nominal setting 100 A),
    - 4 motors, 2 per jaw i.e. the jaws can be adjusted in position and angle. The range for the jaw openings is from ± 25 to ± 100 mm.

    Important : The jaws can only be moved with the magnetic field in the scrapers switched off.

    3.3. HALO tuning

    As the HALO is a very complicated phenomenon depending on a large number of interlinked parameters, it is rather difficult to predict a detailed tuning procedure. However, some possible steps of a tuning procedure are listed below:

    In the "non-scrape-position" the jaws of the first 5 (eventually 7) scrapers are set at such a position, that they just do not scrape into the muon beam but intercept as much as possible the HALO surrounding the beam. A possible procedure to establish these conditions is as follows:

    a) Open all scrapers to the starting-position given in the table below:

    	Scraper 1	± 35 mm
    	Scraper 2	± 40 mm
    	Scraper 3	± 60 mm
    	Scraper 4	± 80 mm
    	Scraper 5	± 45 mm
    	Scraper 6	± 55 mm
    	Scraper 7	± 70 mm
    
    b) Ajust one scraper after the other in the order along the beam.

    c) Move upstream end of first jaw (MOTOR) towards centre and observe beam intensity on the NA58 monitors. Move until beam flux reduces, then 2 mm back. (If you want to be more careful plot intensity against position and move 2 mm back from the point where flux starts decreasing). Do the same with downstream end of jaw. This will establish the angle of that jaw. Do the same for the second jaw.

    d) Now leave the angle fixed and move one jaw after the other (POSITION) to find positions where they just do not reduce beam intensity.

    e) Observe effect on the HALO on downstream HALO counters and experiments veto counters.

    After the complete procedure the HALO at the entrance of EHN 2 according to HALO simulation calculations should be down to ~ 10 % of beam flux if integrated over the total area of the veto counters. The distribution of origins for these HALO muons along the beam line has now drastically changed compared to the distribution for the natural HALO (with no scrapers at all).

    As expected, mainly the HALO muons originating from the absorber region have been reduced, whereas HALO muons originating from downstream the absorber region are, at the position of the scrapers, still inside the beam phase space and have therefore not been attacked.

    The last two scrapers (SCR 6,7) have not been included so far as their good (or bad?) effect on the HALO depends on the experiment in question. Their main effect is to spoil-out the HALO very near to the beam. However, their bending power is not strong enough to bend these HALOs completely out of the sensitive area of the experiments.

    The family of HALO muons originating from upstream and along the decay FODO channel and still arriving at the experimental hall within 8 x 5 m2 around the target has now increased to about 30 % of the total HALO. Most of these muons reach the hall in a direct way above the beam line and are therefore out of reach for the HALO scrapers. The only way of attacking these muons is to reduce the number of losses along the FODO channel by careful steering and by cutting down the beam dimensions to the useful acceptance as early as possible:

    a) Close down the acceptance collimators COLL 1 to 4 and observe beam flux and HALO at experiment, optimize against both.

    b) Now redo the steering into and through the decay FODO channel. Observe beam flux and HALO.

    c) Try to optimize the focii at the FODO entrance against beam flux and HALO.

    To further attack the HALO comming from the absorber region and from the momentum station it will be necessary to further close the scraper jaws and therefore at the same time reducing the muon beam flux. The answer to the question, which scraper is the most efficient in optimizing the quotient of HALO reduction over beam flux reduction will critically depend on the particular HALO problem under study and must be left to these studies. However, some general remarks apply:

    a) The further upstream the scraper used the more magnetized iron is seen by the intercepted HALO muons and consequently the larger is their total deflection angle.

    b) As muons, which are still inside the beam aperture at the scraper positions, but leave the aperture further downstream to become HALO, must populate the outer parts of the 6-dimensional phase space, it is most favorable to hunt for them at positions where the beam is large, e.g. large dispersion. This condition is best fullfilled by SCR's 3 and 4, and HALO calculations have indeed shown these two scrapers to be somewhat more efficient than others.

    4. Quick Monitoring of the beam

    A short procedure exists for quick monitoring of the performance of the beam.


    THE HADRON BEAM


    This section is in preparation.

    The hadron optics is a more styraightforward optics, as shown in the optics diagram on the Web.

    1. Preparations

    To avoid excessive rates in the experimental hall, which may cause radiation alarms and, even worse, damage detectors in the experiment, it is important to
    1. Select a 40 mm target head in T6,
    2. Close COLLs 1 to 4 to ±1 mm.
    Only then put the absorbers to OUT position.

    The intensity at the experiment should never exceed 108 particles per burst.
    In 1999 we plan not to exceed 107 particles per burst in hadron mode.

    Open all scrapers to ± 90 mm.
    Open COLL-5 to ±90 mm;
    Move MWPC 1 to 4 into the beam (e.g. by typing FOR I=1,4; SET TXWCA(I)=1).

    2. Definition of central momenta.

    The fron end serves to define the beam flux and to define the central momentum of the beam. The momentum defining bend is BEND-1, which should thus always left at nominal values. This is important also because BEND-1 and QUADS 1 to 6 are in common with P61, which runs sometimes in parallel with the M2 beam.
    The vertical beam position at the collimators is an image of the beam at T6 and cannot be changed. It may therefore be necessary to tune the centre of the gaps of COLLs 2 and 4. COLLs 1 and 3 should remain centred around their nominal position, thus allowing a correct definition of the central momentum.

    3. Steering into and through the first FODO channel.

    Horizontal plane: Center the beam with TRIM-1 onto MWPC-1. Then center with BEND-3 onto MWPC's 3, 5 and 7. If necessary, iterate TRIM-1 current to compromise between good steering on MWPC1 and MWPC5+7.

    Vertical plane: Center the beam in MWPC2 with BEND-2. Use TRIM-2 to cxenter in MWPC 4, 6 and 8.

    4. Steering into and through the second FODO channel.

    Horizontal plane: Center onto MWPC-9 with TRIM-3.
    Set TRIM-6 to 0 Amps. Then center with TRIM-5 onto MWPC-11. Finally center the beam in MWPC-13 with TRIM-6.
    If necessary iterate between TRIMs 5 and 6 to get good steering both in MWPCs 11, 15 and 17.

    Vertical plane: Center onto MWPC-10 with TRIM-4.
    Close COLL-5 = ±5 mm and scan BEND-5 through COLL-5 onto TRIG-1. Re-open COLL-5 to get a reasonable flux.
    Then center the beam onto MWPC 18 with TRIM-7. Check the profile on MWPC-12 to 16. If necessary, iterate with BEND-5 and TRIM-7 to get a reasonable compromise for the steering at COLL-5 and the four chambers in question.
    Try to keep BEND-6 at nominal current as the experiment uses its field (monitored with a NMR probe) as reference for its momentum measurement.

    5. Steering to the experiment.

    Horizontal plane: Center the beam onto the counters of the experiment (EXPT-1), using BEND 7.
    If necessary, the angle of incidence at the experiment can be adjusted with a linear combination of BEND-7 and BEND-9. The exact steering coefficients depend on the final magnification at the NA58 target, which still remains to be defined.

    Vertical plane: Center the beam at the experiment with BEND-8, counting on EXPT-1.

    6. Focussing at the experiment.

    Before focussing, the beam should be operated at nominal intensity. Q35 controls mainly the horizontal focussing, whereas Q33 and Q36 affect mainly the vertical plane.



    Last updated : 16 July 1999 by Lau Gatignon