hi everyone,
sorry if some of you are receiveing this for a second time. there have
been some problems with mail on athena, so i thougt i'd send it out
again to make sure everyone saw it, and to get it archived on the
blasttalk list.
thanks,
jason
-------- Original Message --------
Subject: proton asymmetry! (fwd)
Date: Wed, 20 Aug 2003 06:55:46 -0400 (EDT)
From: Haiyan Gao <gao@tunl.duke.edu>
To: Charles Seely <seely@mit.edu>
Jason,
Did you receive this message I sent out last night? I did not
receive it myself. If not, can you send this message out for me please?
Thanks,
Haiyan
---------- Forwarded message ----------
Date: Tue, 19 Aug 2003 22:00:42 -0400 (EDT)
From: Haiyan Gao
To: Zilu Zhou
Cc: Jason Seely , Michael Kohl ,
Douglas Hasell , Ricardo Alarcon ,
Antje Bruell ,
Baris Tonguc ,
Bill Bertozzi , Ed Booth ,
Tancredi Botto , Jo van den Brand ,
Henk Jan Bulten , John Calarco ,
Max Chtangeev , Benjamin Clasie ,
Chris Crawford , Adam DeGrush ,
Karen Dow , Dipangkar Dutta ,
Evgeni Tsentalovich , Manouch Farkhondeh ,
Octavian F Filoti ,
Bill Franklin , Eugene Geis ,
Shalev Gilad ,
Willy Haeberli ,
Bill Hersman , Maurik Holtrop ,
Ernie Ihloff ,
Peter Karpius , James Kelsey ,
Hauke Kolster , Aaron Maschinot ,
June Matthews ,
Kevin McIlhany ,
Nikolas Meitanis , Richard Milner ,
Jack Rapaport , Bob Redwine ,
Adrian Sindile ,
Simon Sirca ,
Tim Smith , Stan Sobczynski ,
Tong-Uk Lee , Chris Tschalar ,
Bill Turchinetz ,
Tom Wise , Yuan Xiao ,
Wang Xu , zhangchi ,
Vitaly Ziskin , Townsend Zwart ,
Akihisa Shinozaki
Subject: proton asymmetry!
Hello Zilu and others,
Yes, indeed a sign error was found in the interference
(GeGm) term in calculating the ep elastic asymmetry. I made this sign
error
in the proposal and everybody has been using the asymmetry expression
given in the proposal in doing their ``independent'' asymmetry
calculations. I am fully responsible for this.
The good news is that the data now look more promising. Please see the
attached asymmetry plot from Jason.
The data are
for runs from this past saturday up through 4am this morning. these
runs were taken with the strong holding field (250 Gauss).
The following cuts were made:
coplanarity cut: phi_l, phi_r within 10 degrees of each other
z_vertex: z_l, z_r within 10 cm of each other, both z_l and z_r between
-15cm - 15 cm.
missing momentum < 0.15
As you can see, the right detector results look promising. There are
questions concerning the left detector results. From this plot, it is
clear as Zilu pointed out that the asymmetry is bigger for target
polarization along the electron beam direction. Although the
asymmetry difference
between the zero degree and the 45 degree is not that dramatic, it may
still be a good idea to change the target polarization direction to zero
degree in order to have a statistically conclusive determination of the
target polarization in a short time. Note the overall dilution factor
Jason used for the beam and the target polarization was 0.14.
Concerning the target spin orientation for the proton f.f. ratio
experiment, we are re-evaluating the optimized spin angle and
hope to be able to report in a day or so. Thanks very much for your
patience. Zilu, thanks again.
Best regards, Haiyan
On Mon, 18 Aug 2003, Zilu Zhou wrote:
>
> Hi Jason and all,
>
> as i promised, i would like to write a long email here.
> for some of you who do not deal with polarizations and
> electron scattering, you could ignore this email entirely.
>
> first, i noticed from recent emails from aron, vitaly and
> others, that it was believed the asymmetries would flip signs
> between left and right. this is the same problem as in jason
> and adrian's codes. namely, there is a common bug in codes
> in calculating the asymmetries as function of spin-angle.
> i also believe, this caused some miss-understanding in the
> actual target polarization in comparison with "theory".
>
> before i go further, i would like to point out the bug first:
>
> asym ~ -[ term1 sine(theta_s) + term2 cosine(theta_s) sine (phi_s)]
>
> term1 ~ vt'* fm*fm
> term2 ~ vtl'* fm *fc
>
> common mistakes: a) missing the "-" sign in fm (see Donnelly
> and Raskin page 273)
> b) missing the "-" sign in vtl'(see their page 256)
> from the Q^2 negative values...
> if one misses one of the two "-" signs, it will cause cancellations
> in the theta_s 0 to 90 degree quadrant, and enhancements in other
> quadrants...
>
> i do not know which one you are missing, but most probably, b).
>
> now, many of you would get surprised: how come the asymmetries
> now do not flip signs from left to right, or from A_perpendicular
> to A_parallel? but my question to you, who told you so
> in the beginning? my evidence to enhance what i am saying:
>
> 1, this bug does not cause problems in extracting the proton
> f.f. as one only deals with A_para vs A_perp normally
> and consistently with "wrong" codes, in a pure kinematics.
> it only causes problems when one does large acceptance measurements
> as you have with blast, when spin alignments are not
> exactly parallel and/or perpendicular.
>
> 2, over years, e-p, e-d and e-he3 elastics asymmetries have
> become solid 3-D models deeply buried in the back of my mind.
> i know what is what and what the distributions are.
> betting against me when you do not have much experience
> is your bad choice. so, i was surprised to get challenged
> last friday.
>
> 3, if you do not believe me, here are other evidences:
> a, e-p elastic data from our PRL 89 012001(2002).
> here, A_para flips sign due to theta_s = 213 deg (anti-parallel)
> it is big, since it is almost parallel to the beam-pipe, not
> exactly A_parallel.
> b, as i mentioned, tones of data from hall-B with NH3
> target with polarization along the beeam pipe (antiparallel)
> if you could not find plots (as the elastic channel
> was used as their polarization monitor and normalizer),
> you could try Alex Skablin's thesis. Alex was Bertozzi's
> student did his thesis on the NH3 run.
> i will come back to explain why the elastic has flipped
> asymmetry vs the delta peak later. but remember the anti-parallel
> configuration, which defines the "+" definition.
>
> c, data or calculations from d(e,e'p)n at pm=0, which
> is equivalent to e-p elastic.
> Aedv(90,0) data from nikhef is in PRL 88, 102301 (2002).
> Aedv(90,0) and Aedv(0,0) calculations from Hartmuth are
> in the original blast proposal (see 91 version). the 96
> version does not have Aedv(0,0) as Ricardo dropped it out
> for some reasons.
> here, both Aedv(0,0) and Aedv(90,0) are negative at theta_pq_cm=0.
>
> d, calculations on he3(e,e'p)d from Laget at pm=0 which
> is also equivalent to e-p elastic with about -33% polarization.
> Ax' and Az' are both positive. some of these can be found
> in blast proposal (early version) or the first workshop,
> or Bill Hersman's hall-A proposal, or NIKHEF H.R. Poolman's thesis.
> but Richard knows well enough without plots.
> Ax' = A_perp and Az'= A_parallel. they are both positive at pm=0
> because, in the ground state, the proton is about -33% polarized
> AS SEEN FROM THE pd TWO_BODY BREAK_UP. (watch my capital
> letters for enhancement.) if the proton is 100% postively
> polarized, the asymmetries should be both negative.
>
> so, if you do not believe me, you should be satified by now
> with data, and citations to Donnelly, Arenhovel, and Laget.
> therefore, you should by now realize that it is in the first
> and 3rd quadrants the 2 terms in the asymmetry are enhancing
> each other. and both A_perp and A_para are negative in nature
> for the e-p elastic. and the beam-pipe direction gives you
> roughly the negative maximum.
>
> e, i have not seen the latest blast results. but if i take
> adrian's plots a few days ago, i would say both asymmetries
> were about -0.05 in average. A_left and A_right should be
> about -0.27 and -0.35 in theory for a pure 40deg scattering.
> then, if you believe you have a 65% beam polarization,
> and 45% atomic fraction, you would conclude that atoms are
> about 70+% polarized. (you have to do monte carlo to correct
> the angle acceptance dilutions, as you know, to get a true answer.)
> this is not terribly bad to compare with a good abs with
> a good MFT and second sextupole which gives about 88% maximum
> for the pure atoms.
>
> f, i would like to come back to explain why in a) and b) the
> elastic asymmetries are big positive vs negative in the delta.
> remember, it is anti-parallel to the beam pipe, so, the asymmetries
> are positive not negative for the elastic. as you all know,
> proton is spin 1/2 and delta is spin 3/2. the both reactions
> (e-p elastic and p-->delta excitation) are spin-transfer.
> in the former, it is spin flip, while in the latter it is
> spin excitation from 1/2 to 3/2. so, in the former, the proton
> "-" spin state has enhanced cross-section, while in the latter,
> it is the "+" spin state has enhanced polarized photon absorption
> cross-section. i hope this explains it in a hand-wave way.
>
> in anyway, the asymmetry should be like the following (with phi_s =0
> and a 40 deg scattering):
> corresponding to your asymmetries
> when target polarization is pointing to 45 left
> theta_s Asym
> 0 -0.23
> 10 -0.35 corresponding A_right
> 55 -0.45 corresponding to beam pipe settings
> 90 -0.33
> 100 -0.28 corresp your A_left
> 123 -0.14 corresp to nikhef A_perp
> 213 +0.38 corresp to nikhef A_para
>
> the above list is rough. i do not recall which beam energies
> were used, could be a mix of 720, 850 or 880. i wrote down
> on a scratch paper yesterday evening. and today, i could not
> logon to my MIT computer anymore and i could not double-check them.
> but they are about right. and if Jason corrects the bug, Jason
> can produce nice plots with everything.
>
> somehow, i do not see latest results on the total e-p data.
> could someone send me one copy to verify whether it is about
> -0.05 for both sectors similar to Adrian's earlier plots? thanks.
>
> kind regards. and i hope this email helps.
> Zilu
>
-- -------------------------------------------------------------------- jason seely 26.650.b massachusetts institute of technology 77 massachusetts avenue cambridge, ma 02139-4307email: seely@mit.edu phone: 617.253.4772/6734 html: web.mit.edu/seely/www --------------------------------------------------------------------
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST