Zilu,
Thanks for the prompt comprehensive response. I assume the appropriate
people will address these issues. One basic issue is the luminosity we
are taking data at. There seems to be a factor of about 4
difference between what Genya reports and what Chi reports. What do
the other channels (e,e'p) and (e,e'n) tell us? Vitaliy and Aron
when you compare yields to Monte-Carlo what luminosity do you deduce?
Richard
On Fri, 23 Jul 2004, Zilu Zhou wrote:
> Richard,
>
> i understand your concern. here i repeat and summarize all the
> possibilities we discussed
> in past a few emails:
>
> 1, i doubt flux of deuterium atoms is currently 3X10^16. or at least, i
> should say
> i doubt the average flux of deuterium atoms in the 180kC data was 3X10^16.
> Genya means 2 substates or 3 substates? if he means 2 substates for
> deuterium
> with 3X10^16, then everybody believes the abs hydrogen performance is
> currently
> at 6.4X10^16? please show me a number from intensity measurement.
>
> 2, the 180kC data gave average Pzz 38 %. (this might be wrong, see below #9.)
>
> 3, i suggest Chi should have done the projections, in stead of using 180kC
> averages,
> using current the best numbers. i am not sure which way he did.
>
> 4, missing acceptance in phi-angles? missing e-d events due to strong
> cuts and detection inefficiencies. (we knew many lost events at lowQ
> side due to
> normal bending and energy losses. and we knew at highQ side, Tofs
> would not work
> well for event selections.)
>
> 5, running with 4 states flipping, running more with Pzz=-2, instead Pzz=+1
> which
> is equivalent to "empty targets".
>
> 6, running with 12 mm diameter cell. and please check 3X10^16 atoms/s.
> running with
> 45deg spin angle instead of 32. (the t20 quality will be low once you
> committed
> running it parasitic to Gen. but i understand the compromise.)
> at 32 deg the d-function co-efficiencies in front of t20 are much
> smaller than 1:
> Q2(GeV/c) R20_para R21_para R20_perp R21_perp
> 0.146305 0.360 -1.207 -0.408 0.587
> 0.205382 0.414 -1.190 -0.431 0.512
> 0.280800 0.498 -1.149 -0.463 0.375
> 0.375944 0.593 -1.080 -0.490 0.197
> 0.518211 0.696 -0.974 -0.500 0.021
> 0.681249 0.861 -0.684 -0.443 -0.272
> R20_para should be close to 1, while R20_perp should be close to 0.5.
> R20_para should be higher at 45 deg. running with R20 low is equivalent
> to low Pzz (at this case, about half for most of Q^2 points for parallel
> kinematics.)
>
> 7, i suggest somebody else besides myself and Simon to re-run e-d monte
> carlo to get
> numbers for BLAST... Massi also did it for Ricardo's 97 TDR.
> but i am not sure what you want to get into. blast design geometry was
> known,
> e-d A(Q^2) was known, 1X10^32 and 70% were just numbers, and e-d
> elastic kinematics
> is co-linear so that there is no phase-space complication. (in
> retrospect, i would say,
> in terms of counting rates, the difference between existing 180kC data
> and existing
> A(Q^2) published data is only the blast performance and analysis, or
> bugs in error
> computations.)
>
> 8, i am not sure i want to drag Vitality's report into this discussion
> unless we know exact
> neutron numbers and unless we talk about the same numbers. i think
> citing d(e,e'p)
> comparison which would be more closer to e-d.
>
> 9, possibilities in existing error computations. i have not found time to
> cross-check Chi's
> computations. somebody else in blast must do this.
> Chi, please give me a call. in your email to me yesterday, why you
> divide A_exp by Pzz
> (which you meant 3X 0.37.6 = 1.13?) again to get A_th? since you get
> Pzz from A_th,
> there is a double count.
>
> 10, also i did not include systematics in the proposal. but now here, one
> double counts
> lowQ bin statistics error to get error in PZZ into error equations.
> i am not sure where Chi did it in this way for the 180kC data and
> mislead projections.
>
> yes, if i believe the above, and if i glue the above together, then i would
> not doubt an order of
> magnitude in fom deficit. i remember Chi said the counting rates was a
> factor of 7.7 away.
>
> btw, what happened to the recoil detectors for lowQ (at least the nikhef
> ones) and what happened
> to running with reversal bending scenario?
>
> i do not know what else to say.
>
> Zilu
>
> ps: my new home number in Norwood 781-762-5572. or my cell 203-733-0912.
> Chi, please give me a call sometime in the weekend on asymmetry
> error propagation.
>
>
>
>
> At 09:13 AM 7/23/2004 -0400, Richard Milner wrote:
> >Zilu,
> >I continue to lack an understanding of why the reported
> >uncertainties on preliminary data and projected uncertainties for 300-600
> >kC for the T20 measurement are so much larger than the proposal. We are
> >talking about at least an order of magnitude in figure of merit. I have
> >last night read your proposal again. The stated assumptions are
> >luminosity = 1 X 10^32 atoms/cm^2
> >pzz=0.7
> >
> >This morning I have gone over the best understanding of the ABS operating
> >parameters with Genya. Since the last shutdown Genya reports
> >
> >flux of deuterium atoms = 3 X 10^16 atoms/sec
> >target thickness = 8 X 10^13 atoms/cm^2
> >luminosity at 100 mA = 5 X 10^31 cm^-2 s^-1
> >pzz =0.6 to 0.7
> >
> >If the calculations of the proposal uncertainties is correct, then we
> >should be projecting uncertainties much closer to those in the proposal.
> >The GEn analysis supports the target parameters reported by Genya as
> >Vitaliy is reporting uncertainties consistent with the proposal.
> >
> >It is essential that we understand this. Genya is looking into the
> >calculation of the uncertainty.
> >
> >Richard
> >
> >
> >
> >On Wed, 21 Jul 2004, Zilu Zhou wrote:
> >
> > > not only LADs bars but also double Ohio-walls for Gen?
> > > this would make sense together with e-d elastic in term of
> > > luminosity & polarization numbers.
> > >
> >
> >--
>
--
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:31 EST