Email from Roger:

Date: Thu, 30 Sep 2004 14:39:45 -0400
From: Roger D. Carlini 
To: Dave Mack , Mark Pitt ,
     Jim Birchall , Juliette Mammei ,
     Klaus Grimm , Neven Simicevic ,
     Greg Smith , Allena Opper ,
     Tony Forest , Norman Morgan ,
     Mike Finn , David Armstrong ,
     Yongguang Liang , Shelley Page ,
     Richard Jones 
Subject: Re: Qweak working group - today's meeting

Hi Folks:

I looked at Juliette"s Sculptured versions of Neven's downstream collimator.

Is the internal Bremsstrahlung on or off?

I have attached a cartoon of various "sculptured collimator" concepts. It looks to me from Juliette's rate versus theta 
that a simple "stepped" collimator may be too crude. The theta versus rate curve looks like it may have too 
drastic/sharp a fall off. I thought the basic idea was to do something like what I sketched as "corner clipping".  Also, 
the last figure in the pdf file of shapes is what I suspect might work the best - which expands the phi coverage at low 
theta and restricts it at large theta - in a smooth manner.

Just a thought, the reason that any shapped collimator can only do so much towards cleaning up the in-elastics is 
obviously that it is before the magnet. If we were to put an additional shapped cleanup collimator just infront of each 
bar this should be very effective as long at it were thick/dense enough to stop 1 Gev e- and not shower along the edges.

Finally, Paul confirms with TOSCA that the large divergence is phi really there.  Also, the addition of even significant 
negative curvature on the upstream edges of the magnet (or a powerful opposite polarity toroid) would only have about a 
10% effect on the image width at the bar. So no free lunch! I just wanted to double check these conclusions already reached.


Best Regards

Roger


Roger's attachment:


Collimators.pdf