I am using Yongguang's executable, bash files, and geometry files.
Also, the formula Yongguang used to calculate uncertainties, which I used to find the uncertainties reported in this status report, underestimates the uncertainty. After we are able to reproduce the results, we will look into this further.
"Case 1" is reported to be the standard VT collimator from line 20, with the addition of a 50 cm thick lead shield wall with a window having 2 cm clearance. On his computer, I found a geometry file called "qweakfinal.euclid_vt." This geometry file had a shield wall with 2 cm clearance + 4 cm tungsten, so I tried this and also changing it to 2 cm clearance as he proposed it was in his report.
Table 1 shows the rates for
Table 1: Comparison of "Case 1" results.
| Elastic Photons | Elastic Photons from coll. region | Inelastic Photons | Inelastic Electrons | Moller Photons | Moller Electrons | |
| A | 0.21% | 0.038% | 0.014% | 0.12% | 0.44% | 0.183% |
| B | 0.2091% +/- 0.0052% | 0.0383% +/- 0.0018% | 0.0138% +/- 0.0003% | 0.1194% +/- 0.0022% | 0.4366% +/- 0.0216% | 0.1831% +/- 0.0259% |
| C | 0.2017% +/- 0.0051% | 0.0378% +/- 0.0020% | 0.0188% +/- 0.0004% | 0.0738% +/- 0.0018% | 0.5320% +/- 0.0254% | 0.2432% +/- 0.0322% |
Important Point: Yongguang claims the window with 2 cm clearance yields the rates reported. However, I found that this was not the case. The rates reported for "case 1" in his 12-30-05 status report come from a window design that has 2 cm clearance and 4 cm tungsten lining the edge of the window.
"Case 2" is reported to be different from case 1 by:
and should have a shield wall window with 2 cm clearance. Yongguang does not provide or reference this geometry file. However, in his report Yongguang links the geometry file for case 3 (see below). The only difference between case 2 and case 3 is the lintel. Ideally, we should be able to get a geometry file for case 2 by commenting out the lintel in case 3.
- moving the primary collimator 25 cm upstream
- changing the primary collimator to 5.25 cm thick tungsten
- adding 11.2 cm thick non-defining lead shielding behind the primary collimator
- reducing theta_max to 10.25
The case 3 geometry file does not work, so hence case 2 does not work yet. Once we get case 3 to work, case 2 should follow easily. To see a table comparing the results that Yongguang reports as well as two versions I tried click here.
"Case 3" is reported to be different from case 2 by the addition of a 5.25 cm thick x 16 cm high x 56 cm wide tungsten lintel positioned at Z = -80 cm.
As mentioned before, this geometry file is in his report. However, this geometry file does not produce the rates reported. I tried two configurations
Table 2: Comparison of "Case 3" results.
| Elastic Photons | Elastic Photons from coll. region | Inelastic Photons | Inelastic Electrons | Moller Photons | Moller Electrons | Yongguang | 0.16% +/- 0.0045% | 0.004% | 0.013% +/- 0.00032% | 0.041% +/- 0.0014% | 0.11% +/- 0.01% | 0.125% +/- 0.036% |
| Katherine A | 0.1709% +/- 0.0050% | 0.0018% +/- 0.0003% | 0.0175% +/- 0.0004% | 0.0689% +/- 0.0018% | 0.1117% +/- 0.0118% | 0.1184% +/- 0.0265% | Katherine B | 0.1705% +/- 0.0052% | 0.0023% +/- 0.0004% | 0.0139% +/- 0.0004% | 0.1171% +/- 0.0023% | - | - |
Yongguang created variations of "case 3" with different, realistic shield wall configurations (some amount of lead + concrete). He started with case 3 (primary collimator moved 25 cm upstream, lintel present) and tried three configurations:
These cases are from the 2-6-06 status report. Everything is the same as case 3 except the shield wall. Cases 4 a-c have
where the thickness of the tungsten shielding varies. Case 4 d has the same setup except the window has 4 cm + 2 inch clearance. The thickness of the tungsten "tooth" for each case is:
He has the geometry for case 4d on his report and these rates are reproducible. From looking at his rates we conclude that a) and d) are essentially the same and better than b), c), and case 3.
Conclusions