At Maryland, I am mainly doing the GEANT simulations to study the
missing mass resolution (due to target related stuff such as the windows and
He3 gas) and the distortion of large missing mass spectrum due to collimators
punch-through. In this report I will mainly concentrate on the collimators
punch-through.
(1) An 1cm thick active NE102 collimator was placed in front of each of the 
    tungsten collimator. The energy loss in the active collimator is more than
    2 MeV (for both electrons and protons). Cuts on active collimator are
    very effective in removing the punch-through events. The main question,
    of course, is whether the active collimators can survive the high rates
    enviornment. Bogdan does not believe they will. He suggests to use PbWO4, 
    for example. The reason is that dE/dx for protons would be larger and one 
    can set a higher electronic threshold to suppress the low energy electrons.
    If we were to construct the active collimators with NE102, Bogdan said
    he would not be interested in. I will also do the simulation with PbWO4 just
    to compare.
(2) IN VCS experiment, Luminita used the target trace back to eliminate the
    punch-through events. She first computed the distance (in the reaction
    plane) from the intersection between both arms to the beam line, and then
    computed the difference between this quantity and the "true/known"
    interaction point using information from raster. I have done something
    similar to what she did and found this method to be effective also in
    removing most (but not all) of the punch-throught events. In discussing
    with Arun last week while I was at JLab for the users meeting, he suggested
    that if one is to trace back only to where the collimators are (instead of
    all the way back to the target), one should be able to remove the
    punch-through events "completely". I verified this idea with John LeRose
    and he agreed. I will implement this idea in my GEANT simulation. I also
    suggested to Luminita to try this idea with her data from VCS. I got the
    feeling that she was not very enthusiastic. However, she promised to make
    the most useful data set available to us and we can try to replay then
    ourselves. 

	I will show the simulation results during the July 21 meeting.
						George