[d2n-analysis-talk] Bend-down momentum calculation

[Diana Parno]

Hello all,

At today's meeting, there were some questions about the bend-down momentum calculation in BigBite. In particular, Matt had noticed a few-MeV discrepancy in the location of the elastics peak (is that right, Matt?) for bend-down electrons in the positive-polarity configuration.

First, for anyone following along at home, the optics code used in our replays may be found in

/work/halla/e06014/disk1/diana/replay_64/20110507/replay_64/Round2Build/d2n_analysis/d2n/bigbitelib/THaOpticsE06010.cxx 

(The whole set of code used in the farm replays is in /work/halla/e06014/disk1/diana/replay_64/20110507/replay_64/Round2Build/d2n_analysis/ .)

You can look at line 2243 for the beginning of the proper bend-down momentum calculation, which takes as input the vertex calculation and the projection of the trajectory onto a lot of different planes (the bend_y locations). If you can follow the code up to that point, then you have superpowers. Note that the momentum variable is named "radius".

The first two lines constitute the first-order momentum calculation, which was done by Xin from (as I recall) positive-polarity data taken at 45 degrees. I did not redo this the way that I did for the bend-up optics.

What I *did* do was strip out all the "higher-order" corrections that had given us so much trouble for the bend-up optics. I also redid the low-momentum correction to remove the discontinuity in the momentum distribution, in parallel to the bend-up optics. (Basically, I took Xin's correction and gave it a new functional form in a transitional region, to smooth out the discontinuity.) I also removed the sketchy correction for energy loss. Line 2271 uses the radius variable to assign the three components of the reconstructed momentum; no changes after that make a difference.

You can check out this 2010 progress report for the discussion and justification of these changes: https://hallaweb.jlab.org/wiki/images/3/37/Parno2010_11_11.pdf

If we want to check the bend-down optics, we can retrace Xin's steps and look at our one-pass positive polarity data. Runs 1270-1273 were taken at one pass, with an H2 target, with the BigBite polarity positive and in coincidence with the LHRS. The sieve was apparently in. A quick check shouldn't take very long. If we discover a problem, then of course the timescale is more open-ended.

I missed the discussion of this issue during the break, so I'm not sure how serious it is (or may be). What shall we do next?

Best regards from sunny Seattle (really!),
Diana