View source for G2p other optics plan

==Normal HRS optics + Target Field optics==

I think there is an important factor to find out: How large is the effective
magnetic field region. If it is small, or quantitatively

typical length * bending angle (rad) * 4% (acceptance) < 1mm,     (1)

then the optics can be decoupled to two parts with small mixing:
I  normal HRS optics from an effective target
II bending in the target region



I will continue assuming Eq (1) is valid, then here is my PRELIMINARY
thinking on optics runs.

1. Calibrate part I, HRS optics w/ septum at 1GeV momentum. 
Start with an initial optics database like PREX or APEX for replay. 
Target field is turned off here. Single C foil target is used
       
 1.1 Sieve OUT, Raster OFF then ON (as large as possible), take data
 at quasi-elastic kinematics. This step is for vertex calibration of full
 acceptance, which is going to be used for clean up sieve pattern with vertex cut.
       
 1.2 Sieve IN, set beam at elastic kinematics, perform delta scan, dp
 = +4% -> -4%, 2% per step. 1.1 & 1.2 calibrate dp, th & ph for full acceptance.
       
 1.3 Large Raster ON. repeat 1.2, dp = 0%. This is to calibrate the
 raster correction. If the correction have strong correlation with dp
 (traditionally not), then dp scan will also be needed.

2. Calibrate part II
       
 2.1 Scan target field strength 0T -> 5T. Take elastic data, dp = 0%
 with Sieve IN, raster off. The goal is to get a correction function on dp,
 th & ph, which is a function of TargetField/HRS_p0 to leading order. You may
 need a target field simulation to help you get the function.
 Note: correlation between optics observables introduced by the target field
 is small for your case of small scattering angle.
       
 2.2 You may need to scan dp and/or raster for step 2.1, depending on
 how strong the correction function is correlated to dp/beam position,
 relative to your precision. A calculation/simulation before the experiment
 will tell.

3. Repeat calibration at 2GeV. The calibration will be easier since you
 already have a good optics database to start with. You can reduce the number
 of data point. Then build the optics as a function of (HRS_p0,
 TargetField/HRS_p0)

4. repeat for the 12 degree case.

Above might not be a full list. Someone will need to look carefully into
this business.

The above calibration is mainly relative to HRS central. It will be
important to get the absolute horizontal angle which is related to Q^2.
Multiple-target-mass method might not be enough since the momentum transfer
is small (need to calculate). 2GeV beam with proton and Ta target might be
just enough to give a 1% precision for 5 degree scattering angle. I GUESS
good survey on Sieve and beam position may give a precision of 1mrad.

Hope this note helps. I am also glad to answer questions if needed.


Cheers,

Jin