Present: Chris Allgower, Andy Bacher, Scott George, Paul Pancella, Tom Rinckel, Mark Pickar, Adam Smith, Ed Stephenson
The meeting came to order at 3:08 pm
QUAD #1 COOKED
Tom Rinckel reported that due to a failure of safety interlocks, quad #1 on our channel received 200 Amps of current without cooling water flowing, for an undetermined period of probably over an hour. This occurred a few days ago during the current Tagger run on the Cooler. A computer glitch unintentionally set the DAC of a multi-channel power supply which is shared between our magnet and devices currently in use. It is not clear why lack of water flow or (eventually) high temperature indicators failed to cut off the current. This was formerly the K300 quad.
Result was one very hot quad. Stagnant water in the coils apparently boiled and blew off the polyflo supply lines. The painted surface of the coils is discolored, and there may be cracks in the epoxy potting. The nearby dE1 thin scintillator approached its melting point, and has taken a non-planar shape. The extent of damage to the quad is unknown at this time, since the Cooler is still running.
After the tagger run, it will be necessary to replace the water fittings, put some power in the quad and test for shorts in the coils. There will probably be an opportunity before then to remove dE1 and determine if the light guides are okay. Chris will take care of that.
OTHER HARDWARE
Due to the busy shop and vacation schedules, major welding on the target box will not begin for about three weeks. Jack Doskow has taken over the task of designing a new exit snout from the 6 degree magnet chamber. Chris will also work with him on the geometry of the cosmic ray paddles so the lead glass stand designs can be finalized soon.
We should try to identify some manpower for the tedious job of applying the magnetic shielding to the phototubes of our lead glass counters.
Tom has drafted his install list for removing the tagger systems and completing the full installation of CE78. He will go through one more iteration before distributing it to the group for comments. It looks doable.
SIMULATIONS, ANALYSIS
It has been determined that the foils used in the wire chambers are twice as thick as previously assumed. Earlier calculations of energy loss also fudged the gas mixture, so that actual eloss is expected to be greater than predicted earlier. This agrees with Pancella's analysis of the test run data, at least qualitatively. Despite this correction, GEANT gives no indication that the downstream scintilla- tors need to be made any larger. Pickar has started to set up his GEANT calculation for the 3He case, in order to compare directly with data.
Adam reports that he tried several strategies to improve the TOF resolution in the test run data, with no positive results beyond the correction for vertical position in dE1 previously seen by Pancella. He will document the specific attempts for the group.
Scott George informed us that the large centroid drifts previously reported in the TURTLE model of the channel have been traced to an error in the inputs. This error was caused by believing (as advertised) that TURTLE and TRANSPORT use exactly the same inputs to define a beam line. In fact, there is at least one subtle difference which comes into play when a bending magnet follows some energy loss. TRANSPORT and TURTLE now agree in terms of the centroids of track distributions, and the reason for the observed lowering of the cone at the end of the channel is still unknown.
Ed tried to refine our time resolution goal by calculating the expected time spread with a Vavilov distribution. A monoenergetic 3He beam undergoing energy loss develops an asymmetric distribution of TOF due to this straggling effect. Just by taking the fwhm of this distribution and dividing by 2.345, he expects an equivalent sigma of 0.4 ns for this spread, much larger than the 0.1 ns we were hoping for. This intrinsic width is thus a major component in the observed sigma of 0.81 ns in the test data, and we probably shouldn't hope to reduce this very much. We have yet to determine how this will affect the resolving power of the alpha experiment.
The question of whether to segment dE1 is still pending. An REU project some years past took hard data on the rate characteristics of our PM tubes, but that data did not make the final report. Ed asked if anyone has the logbook from that investigation (by Damon Spade?).
Pancella presented results of his revised analysis of test run data. The two major changes are a slight revision of the beam energy, due to the effect of the measured target position on the circumference calculation, and Chris Allgower's determination of the scale factor for the timing histograms.
A complete reanalysis of the cone sizes as a function of Cooler RF results in a closed orbit circumference of 86.786 meters. Pancella claims this is the final answer (for May, 2001) and that the uncertainty is on the order of +/- 3 mm. Using this value and the orbit frequency (accurate to 10 Hz) results in beam energies that are accurate at the 20 keV level. Our six measurements between 199 and 200.5 MeV agree at this level, with no convincing indications of residual systematic effects.
With confidence in the beam energy, and our very clean 3He data, simple kinematics allows us to determine how much energy was lost by the 3He before beginning the flight path used for the TOF measurement. This energy loss is rather large for 3He, and a significant correction to the missing mass calculation.
Paul then showed the algorithm he currently uses for this calculation along with all assumptions and results at one beam energy. A measured value of 21.2 +/- 0.2 channels/ns is now used to scale all time spectra, and is no longer a variable parameter. The amount of energy loss is now the most important adjusted parameter. It was correctly pointed out that the energy loss determined by Pancella's method really includes some of the loss at the exit of the quad pipe in some funny average, since there is considerable flight path between there and the dE2 scintillator. Thus it is no surprise that he concludes the 3He are still lower energy than Pickar now predicts.
The momentum distributions of the 3He and the unseen pion are very asymmetric (favoring faster pions in the lab). Pickar says that this is in fact the way the cross section goes, favoring pions forward in the center of mass. The measured missing mass comes in a very narrow peak centered on the expected value. It is asymmetric and not well fit by a Gaussian, with a fwhm of about 300 keV/c2. It cuts off sharply on the heavy side, with no events above 135.9, but tails off on the low side down past 128 MeV/c2. (This for run 49, 200 MeV beam and 1.325 degree cone.)
IMMEDIATE FUTURE
Ed would like to ascertain who is going to do what over the next couple of months, as some collaborators leave and others arrive. A new post-doc, John Olmsted, will be joining the group in mid- September. Ideally, we should have a means to transfer all knowledge from Pickar, Pancella, and Smith to Olmsted, despite the negative overlap time at IUCF. Chris and Ed are probably the logical conduits. Chris will try to take over the GEANT calculations, which Pickar believes will be in usable form before he leaves.
Those who are returning to academic positions are requested to transmit their schedules over the next year to Andy for planning purposes.
The meeting adjourned at 4:40 pm.