Primary Collimator Live Page
GEM rotator movie #1
GEM rotator movie #2
Tony's Status Report
Klaus's Status Report
Yongguang's Status Report
Qweak Simulation Forum
Keep Out Zones a la Allena
Previous Reports
Virginia Tech Update for Tracking Group,
Goal of this discussion: To present the arguments to adopt a (nearly) final location and shape of the cutout for the
primary collimator. Adopting a particular scheme will then allow each group to check the implications for their detector
and then we can iterate to the final design.
We are basically proposing that we adopt the downstream location. The pros and cons of this location are discussed
near the end of this text.
First we review how we got to this particular recommendation.
1. For reference, the scheme we are proposing is shown below (Figure 1).
The minitorus shown is somewhat larger than previously. Its coils have a diameter of 87.5 cm (rather than the original 52.5 cm)
BUT we have moved them downstream to accomodate this size increase. The upstream end of the minitorus is 33 cm further
downstream than in the previously shown design. This leaves more space to find a suitable location for the GEMs. (In fact,
in the scheme shown the GEMs could be located at the same place where they were located in the Boston collimator scenario).
2. Determining the "limits" of theta (upper and lower) and phi:
a) Accomodation of the main torus support structure: We have assumed that the main torus and its support structure is fixed
and cannot be modified. Both the upstream and downstream collimator choices were cut out to allow 2 cm of clearance all
around the scattered electron beam profile as it traverses the main torus (See those plots). Unless other constraints
come up, we believe that this will set the limits on lower theta and phi for our experiment.
b) The choice of upper theta boils down to what can be accomodated on a reasonable width (< 18 cm) quartz bar at the focal
plane. Figure 2 shows how the error on Qweak changes as the acceptance is decreased by increasing lower theta or decreasing
upper theta (for a handlebar detector).
|
|
Figure 2a - The error on Qweak as a function of delta theta for increasing lower theta. |
Figure 2b - The error on Qweak as a function of delta theta for decreasing upper theta. |
Clearly, a lot of the upper theta stuff is not even on the detector, so we chose cutoffs of 4 degrees for the downstream version
and 1.5 degrees for the upstream version (using Figure 2b). This can be seen clearly in the following series of snapshots of
different theta ranges at a given z location for the downstream collimator (See those plots).
So this basically sets the limit on upper theta for the experiment.
3. With the above choices, the rates, error on Qweak, etc. are summarized in rows 14 and 15 of the table on the primary collimator
"Live" page. The error on Qweak for a upstream collimator is always higher than for a comparable downstream collimator; it is well
understood why this is the case.
4. The extreme and normal rays for the two cases are (See Figure 3):
| |
|
Figure 3 - Definition of extreme and normal angles. The normal minimum and maximum
|
5. Region 3 considerations: The beam profile at region 3 looks similar for both downstream and upstream choices (See the plots).
From this, region 3 folks can see what size the active area of their detectors would need to be to accomodate this. Also, we considered
what the implications would be for a rectangular rather than handlebar detector in row 16 of the table on the primary collimator
"Live" page.
6. Region 2 considerations: With the redesigned mini-torus the downstream option looks viable. The Moller rates at the Region 2
chambers are now 105 kHz/nA (compared to to 3063 kHz/nA with minitorus off) for the downstream collimator and 124 kHz/nA
(compared to 3410 kHz/nA with minitorus off) for the upstream collimator. So with the downstream collimator we could comfortably
do the calibration runs with beam currents in the 5 - 7 nA range. The new minitorus design bends the e-p elastics by about 0.6 degrees.
(Note: We can run with minitorus on or off. The minitorus off numbers for the downstream collimator are shown in row 17 of the table).
The choice of whether to run with minitorus on or off for the production run is still to be made, but I don't think it has to affect this
decision. Also, the scattered electron profile fits through the mintorus (See the picture).
7. Region 1 considerations: The new minitorus location allows the GEMs to be located at the original "Boston collimator" location
(downstream arrow in Figure 4) if that is desired.
The profile for the downstream collimator at that location is shown below (See Figure 5). A 16 x 16 cm2 GEM should fit.
Figure 5 - The ep peak profile at z = -543.17 cm, or the old location of the GEMs.
The profiles at the more upstream z location that had been advocated recently (upstream arrow in Figure 4) are shown below for the
downstream and upstream versions of the collimator (See Figure 6a,b). It is unclear that the rates per unit area that the GEM
would experience at this location are tolerable.
|
|
Figure 6a - Accepted ep peak at z = -584 cm for downstream collimator. |
Figure 6b - Accepted ep peak at z = -584 cm for upstream collimator. |
8. Summary of the tradeoffs between upstream versus downstream.
a) For collimators that accept the same "extreme" rays, the downstream collimator always gives a smaller error on Qweak.
(favors downstream)
b) Mechanical tolerances are ~ a factor of two looser for the downstream relative to the upstream collimator (favors
downstream)
c) The acceptance over the length of the target is flatter for the downstream versus the upstream collimator (favors
downstream)
d) A downstream collimator would be visible to survey, and it would not get so activated that no one could ever survey
it again. Also, one could imagine making two downstream collimators (with different acceptances) and changing them out
relatively easily. (favors downstream)
e) The tolerances for helicity-correlated beam motion and size changes are much looser for the downstream collimator
(Jim's Report) (favors downstream)
f) Having the mini-torus upstream of the primary collimator has been pointed out as a concern. If we choose to run with
mini-torus on during production running, one could worry about fluctuations in mini-torus current causing fluctuations in
the position of the scattered beam spot at the entrance to the primary collimator. This concern has been addressed by
Dave Mack, and it appears not to be a problem. (wash - doesn't favor upstream or downstream)
g) For the upstream collimator, we could potentially have a 10 x 10 cm2 GEM work if it could tolerate the rates at the
upstream z location. This would require an actual shielding and beampipe scenario to be worked out to show that this is
possible.
9. Based on the above, we would recommend adopting the downstream collimator.
10. What remains to be done:
If we adopt the downstream solution, then Virginia Tech first needs to do some minor tweaking on the shape of the
cutout and the thickness of the collimator (since we have decided to adopt the lead-brass alloy). We need a few
days to do that, and then we will release the beam "profiles" at the various locations to people to finalize the
design of their components. At that point, each group needs to look at their detector in this scheme, and see how
the tradeoffs will work out. We may need to iterate back and forth a little bit. The ultimate final goal is to
get all of our pieces into the JLAB 3D model to do a final check for interferences.
Some considerations in the final iterations:
a) Shape of focal plane detector: (handlebar, v-shape, rectangle): I suspect this will partly boil down to a choice
of what inelastic fraction we are willing to accept. As Peter Bosted has pointed out, perhaps 0.02% is too conservative.
If we back off on that, then maybe we can have a simple rectangular bar. (Also, Dave Mack has requested that we study
the effect of changing the target-QTOR distance to see if it changes over which theta region the phi "de-focussing"
occurs).
b) Area near target: VT group will provide beam profiles to Greg Smith and the JLAB designer so they can design a
beamline, target chamber flange and window, and first "cleanup" collimator. They will determine if our current low
theta limit is actually realistic.
c) Region 1: We think there is enough room now that the GEMs can find a suitable location from the point of view of
rate in their chamber. If the GEMs can successfully work at the "original Boston collimator" location, then we may
be able to move the mini-torus a little further upstream to give more "head-room" to the Region 2 chamber operation.
Tony requests that the magnetic field at the location of the GEMs is less than 50 Gauss.
d) Region 2: We have a good idea of the rates in our chamber due to Mollers. But we also need to check that the rates
due to showering in the primary collimator (just upstream of our chambers) are acceptable. This should be easily
controllable by making the primary collimator thick enough, but we need to do the study to determine how thick it needs
to be.
e) Region 3 chambers: Region 3 needs to look at how big this implies that their detectors will need to be and indicate
whether that is economically feasible. (But I actually don't think this profile is any bigger than we have been
discussing since the time of the proposal).
Figure 7 - The accepted ep peak at the GEM is shown in red, superimposed on the "real" ep peak at that location (black).
One aspect of the downstream collimator that we realized after this meeting was that for the GEMs and the minitorus the
actual dose the electronics and/or support structures will see is larger than it would be for the upstream primary
collimator. This is because the cleanup collimator upstream will have a larger acceptance so that it doesn't cut into
the acceptance for the experiment (See Figures 1 and 4). If the upstream collimator were the primary collimator,
rather than a cleanup collimator, the acceptance would be the acceptance of the experiment.
Figure 8 - The accepted ep peak at the minitorus is shown in red, superimposed on the "real" ep peak at that location (black).
The cleanup collimator can be made to have a slightly smaller acceptance; it hasn't been fully optimized yet.
Figure 9 - Upstream of QTOR with dimensions.
