During the last several days a decrease in the number of "hits" in the beam hole blocks of the LGD was observed. Radiation damage is the obvious suspect. During a downtime period yesterday, I shutdown the LGD and removed a tube/base unit from a cell bordering the beam hole. The glass was observed to be noticably discolored, confirming the radiation damage hypothesis. Quantitative conclusions cannot be reached based on visual inspection.
I've since studied the effected blocks with an offline analysis program. Two postscript files are linked below, the first page in each shows the distribution of energy deposited in a block near the beam hole. This distribution is fitted to the sum of two exponentials of the form A*exp(x*k). The coefficient -1/k sets the horizontal scale of the distribution and is taken to be a measure of the amount of light deposited (on the average) on the photocathode. The two component nature of the distribution is apparent from the figure. Both exponentials were used to characterize the light collection with consistent results.
The 2nd through 9th pages of the files below show the dependence of -1/k as a functon of run number and integrated beam hours for each block on the edge of the beam hole. Examination of these figures leads me to the following conclusions:
1) Radiation damage is a cumulative effect. There seems to be no good evidence for a sudden loss of gain. A small, local trend near run 7800 (beam hour 75) might be significant but it also seems that damage continues after that point. (For those of you on site, this contradicts my statements in the logbook, numerical studies convinced me I was mistaken)
2) Radiation damage has resulted in a loss of gain of at most a factor of two. Comparison of the maximum and minimum for the plots will convince you.
3) Increasing the voltage on the tubes near the beam hole seems to have recovered the lost gain. The last point on each plot (deduced from run 8171) shows this to be the case. The voltage on all the tubes considered here was increased by about 100 volts immediately prior to run 8134 (about beam hour 300)
4) This radiation damage, while worrisome, does not constitute an emergency at this time. I urge you to study these results and reach your own conclusions. I shall continue to monitor this effect in the future.