Present: Chris Allgower, Andy Bacher, Chris Lavelle, John Olmsted, Paul Pancella, Mark Pickar, Tom Rinckel, Adam Smith, and Ed Stephenson
SUMMARY OF THE LAST RUN
Abstract:
Following a second luminosity calibration, production data for d+d -> alpha+pi0 was taken at the higher energy of 231.8 MeV for more than 9 days. Several dozen alpha+pi0 candidate events were recorded. This part of the run was ended by a short in the winding of the septum magnet. We switched to the Cooler A-region and polarized deuteron beam for a measurement of the cross section and analyzing powers in d+d elastic scattering. In addition to checks of the polarization using an H2 target and HD normalizations for the cross section, we obtained 4 days of d+d data with vector and tensor polarized deuterons.
This run concluded operations with the IUCF Cooler.
Detailed summary:
In the hope that we would be able to see a change in the alpha + 2*gamma "background" underneath the pi0 peak, we elected before the run to raise the beam energy to 231.8 MeV, corresponding to a cone angle of 1.75 degrees. We expected that the rate of pi0 events would increase, despite the problem that part of the cone would be cut away on the side toward the exiting deuteron beam.
The "sweet spot" in the Cooler setup at the top of the ramp was not as easy to find because the maximum lifetime did not, at least initially, correspond with the position of the gas jet. We chose to run at the jet location. We again saw the presence of a large deuteron background if the beam was steered to the right side of the jet, and we were able to operate about 3 mm from this edge. We scanned gas jet pushing pressure and confirmed that the most efficient operation was obtained with a lifetime at the flattop of about 100 s. Most of the running was with a larger beam current of 2 mA or more at the start of the cycle.
After checking with hydrogen gas to make sure that the luminosity systems were again functional, we repeated the calibration with HD molecular gas. A subsequent analysis showed that the results were consistent with the previous calibration in June at the level of 3%. The only problem was a shift in the time spectra caused by a faulty connection on an ECL cable. The calibration was run with a lower gas jet target density so that HD events could be taken with no prescale. At the end of the second calibration, half of the HD gas remained. Following the calibration, we found it necessary to warm the gas jet nozzle to 200 K in order to clear it of frozen material.
We switched to production running late on July 11.
It was noted that during the flattop, the comparative rates for the two luminosity systems and channel trigger events varied with respect to each other. This suggests that there may be some spreading of the effective target region during the time that the target is on.
Running went smoothly until rising outside temperatures began to cause problems for Cooler operations. The first incident was on July 17 when the building air conditioning was off for a time and temperatures inside the Cooler enclosure rose above 98 F. This caused a loss in gain for the delta-E1(b) PMT and the failure of one of the PCOS-III latch and delay modules. We decided to leave the PMT gain at the new value.
The second series of incidents began on July 20 when the water cooling towers outside the building were no longer able to keep the chilled water system at its setpoint of 80 F. As the operating temperature increased to 87 F, the first casualty was the compressor for one of the T-region cryopumps. For a while, we were able to get the temperature of its compressed helium heat exchanger down enough so that intermittant running was possible. Finally we decided to disable the temperature sensor and run with it hot. As this situation developed on July 21, we also experienced fluctuating timing for the delta-E1(a) PMT system, and the delta-E2(b) gain suddenly doubled. The delta-E1(a) readout was moved to another discriminator channel, and the delta-E1(b) gain was changed realizing that this would cause a break in the time offsets for that channel.
Running got underway again late on July 21, but ended abruptly at 2:11 on July 22 when the septum magnet had a catastrophic short in the lower half of its windings. This short did not appear to be a consequence of the other issues associated with the hot weather. Like a short that appeared during the original mapping of the magnet, this melted the copper tubing to the point that a serious water leak appeared. A repair was not possible, so this ended the production running for d+d -> alpha+pi0. We think that the production running generated a set of data with statistics comparable to the June run. The initial plan had been to switch to the d+d elastic scatteing measurement three days later.
Replay of the data during the run showed that, with the timing parameters used in June, the width of the pi0 peak was about 4 MeV FWHM (compared to 0.7 MeV before). Some increase was expected because of growing effects from the dispersion of the 6-degree magnet. The cause continues to be investigated. A replay using the deuteron group as a time marker shows that there are slow drifts in the time-of-flight as well as discontinuities associated with the changes in PMT gain noted above. So a run-by-run compensation may be required.
With the demise of the septum magnet, we chose to move the startup of the elastic scattering measurement forward by three days. The A-region silicon detectors were installed, and the polarized source was put back into operation. The beam energy was not changed. Tuning of the beam resulted in excellent background conditions in the A-region with little effort.
The first day was devoted to testing the two triggers: (1) two forward prongs in the PINTEX detectors to observe mainly d+p scattering in the angular region where the analyzing powers are large, and (2) one prong in the forward detectors in coincidence with one in the silicon detector array that surrounds the tube that confines the target gas. In both cases, the two prongs were required to be on opposite sides of the beam. The software was assembled from units developed by the PINTEX group for prior experiments. Online, gains were adjusted so that both d+p and d+d events could be recorded. Particle identification loci were clear in all detectors.
Initial measurements with the BLPL polarimeter showed that the polarization was satisfactory except for the weak field transition. Adjustments to the RF power level brought it back into reasonable operation. The first tests with an H2 target showed that the polarization survived injection and ramping in the Cooler. At first this appeared to contradict vertical tune measurements that showed we were crossing an intrinsic depolarizing resonance, but those tune measurements later appeared to be at best ambiguous.
The large-area PINTEX detectors were originally designed for the lower circulating currents characteristic of cyclotron injection. In this application, event rates were large and limited finally by the ability of the online software to process events coming over the network from the VME crate. If this limit was exceeded, the front end would go dead until processing caught up. We operated just under this limit for the remainder of this run. This needed careful management of the beam current and target density, as well as a watchful eye to make sure that other uses of the Zeus computer (including data compression and backup) remained at a sufficiently low priority.
We operated with one unpolarized and four polarized states (vector +, vector -, tensor +, and tensor -). Once per day, we took 4-6 hours of data with a hydrogen target to calibrate the polarization of the deuteron beam using measurements from RIKEN. Some reduction was noted online because background is not removed from the polarization calculations. Otherwise we ran with D2 gas. The polarization cuts also yielded something like the same asymmetries with a deuterium target because they registered quasi-elastic scattering.
At the beginning and the end of the four days of d+d production running, we used HD gas and an unpolarized beam to make the cross calibration of d+d against d+p. There appear to be several regions in which these two processes are distinguishable. In addition, an alpha-particle radioactive source was used to generate reference spectra for the silicon detectors.
By gating on deuterons in both the forward and silicon detector arrays (and including coplanarity cuts), we were able to define online an angular distribution that should resemble that for d+d elastic scattering. The data appears good between laboratory angles of 12 and 40 degrees, outside of which there are slow cutoffs in the acceptance of the forward detectors from all parts of the gas target tube. The cross section appears to have a "single scattering" distribution out to about 15 degrees (lab), after which there is a steady fall with a much smaller slope. We hope that binning of the data in 0.6 to 2.0 degree sections will lead to errors in the analyzing powers that are less than 0.01 at forward angles and 0.03 near 90 degrees (c.m.).
The run was brought to a close on Monday morning, July 29. The last Cooler beam circulated at 9:22 a.m. A measurement of the beam polarization at the output of the linac continued until 13:37. These measurements were the last for IUCF Cooler operations.
GROUP MEETING NOTES
Chris Allgower will be in charge of dismantling the electronics and gathering all equipment that needs to be returned to its rightful owners. This includes electronics from Fermilab and Brookhaven, as well as the Pb-glass and PCOS (from the tagger) from Hal Spinka at Argonne. Chris will contact Hal for details.
Final measurements of the locations of the luminosity scintillators is getting underway. Ed has unwrapped all scintillators, and measured the 44-degree pair against the frame of the target box. The main uncertainty here is the solid angle of each double scintillator. Work on the forward scintillators will start when the shielding wall at the northeast corner of the Cooler is unstacked.
An analysis subgroup has been formed, consisting of Andy Bacher, John Olmsted, Mark Pickar, and Ed Stephenson. [Adam Smith will join while he remains in town.] The purpose will be to define and carry out the steps in the analysis of the d+d -> alpha+pi0 data. Some work may be carried on by Adam and Paul during the fall.
It was decided that no "preliminary" cross section values will be quoted until the analysis process is complete.
We discussed what publications might come from this. These include:
The meeting was adjourned in time for our luncheon end-of-run celebration.