[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CDR chapter 4



Jim, see my comments below.

Jim Stewart wrote:
> Hello Richard
>
> I am thinking a bit about what to say about the electron beam.
> I want to say something about how the beam spot size effects GlueX.
>
> To this effect I spent a few minutes thinking about diamond.
>
> You are proposing to use 20um thick diamonds. The reason is that, the 
> thinner
> the diamond is the better, since the multiple scattering is less. We 
> will never get thin
> enough that the coherent effect in the crystal is reduced. Secondary 
> effects are heating
> and radiation damage. The cooling was calculated for 15 um and 3uA and 
> was not
> found to be a problem. Radiation damage is the big open question.
>
> The radiation length for crystal diamond is X_0 = 12.13 cm. (PDG number)
> A 20 um diamond is then 20 um / 12.13 cm/X_0 = 1.65\times10^{-4} 
> radiation
> lengths thick.
>
> The nominal beam current is 3 uA or 3\times 10^{-6} C/s. Which 
> corresponds to
> 2\times 10^{13} electrons/second.
>
> In the last version of the report it is stated that the best estimate 
> for the allowable
> charge passing through the sample is 0.25 C/mm^2. The beam spot size 
> with Yves'
> new tune is 0.82 by 0.36mm. This gives a total area of 
> A=pi*a*b=pi*.41*.18
> A=0.23mm^2.
>
> The time we can run with one spot location depending on spot size is:
> T_damage(spot_size)=0.25(C/mm^2)/3*10^-6(C/s) =8.3*10^4 seconds/mm^2
>                                = 1 day/mm^2
> This means with a spot size of 1/4 mm^2 we can run 6 hours. Have I 
> made a mistake?
Actually it is closer to 5 hours with this reduced beam spot size.
> To run 14 days with a single beam position we need a 14mm^2 spot. This 
> would
> correspond to a circle of radius 2mm. That is roughly the size of one 
> diamond.
> A diamond should last longer than 2 weeks shouldn't it?
No, that is the correct order of magnitude.  Keep in mind that these 
estimates are very crude, none of them actually published as far as I 
know.  The one that I use was quoted in SLAC talks and proposals I have 
seen, referenced to a SLAC internal report Schwiters et.al., 1970 that I 
have never seen, probably in some archive somewhere at Stanford.  The 
same report [reportedly] claims that damaged diamonds were reclaimed 
using annealing in an oven and recycled 2-3 times before the accumulated 
damage was too bad to anneal out any more.

Our plan needs to include radiation damage assessment.  I think we can 
do that as a part of our two-year run plan at low intensity.  At startup 
we will run a factor 1/10 in intensity, which means our diamonds will 
last 10 times as long.  If we have 2-3 good diamonds mounted in the 
goniometer during year 1 then we can do spot moves after varying amounts 
of integrated dose, and combine running with radiation damage 
assessment.  At the end of the first year we would then have enough 
samples to map out the radiation damage history for those diamonds, and 
refine our plan for diamond inventory management.  Ten diamonds for a 
year of running is not an unrealistic estimate based on what we know, 
but as I said, that is only reliable within a factor 2-5.
> Have I made a silly mistake?
No, in my presentation to the 2006 tagger review, I said 3000 hours per 
crystal at 10^7 tags/s, 300 hours at 10^8 tags/s.  300 hours is 12 days, 
so I think we are in agreement.

-Richard J.