Optimization Procedure

Some text describing our optimization procedure.

Here is the summary of our approach. Please let us know if you think we are missing something important.

We assume that the main torus and all of its support structure is fixed. We have chosen two z locations to focus on for the primary acceptance defining collimator- the "upstream" location near where the original "Boston" collimator was and the "downstream" location near where the collimator was place for the jeopardy proposal. In both cases the collimator is initially "sculpted" to produce a beam profile that just makes it through all the main tourus support structure with 2 cm to spare around all edges. This will tell us what is the maximum amount of scattered particles we can get through the current main torus. Then we start to add in the other constraints:

* lower theta probably in reality will be set by the minimum tolerable beampipe plus shielding size and not the main torus support * upper theta will probably in reality be set by the constraint of a 2 meter quartz bar and reasonable focussing for elastic/inelastic separation * etc....

The goal is that when we are done, we will be able to say exactly what is the limiting factor for each of our dimensions (upper theta, lower theta, limits in phi, etc.). So we'll have a good basis for understanding why we ended up with the collimator we did.

We will also provide comparisons between various scenarios, so that we can make a quantititive judgement on the impact of various equipment choices and evaluate the tradeoffs. As an example of that, I mention the following:

* Mini-torus: It could be (we don't know for sure yet) that the mini-torus coils or its support structure may be the limiting thing on some aspects of the acceptance. If that were the case, we would try to generate the numbers for the case where the mini-torus was NOT the limiting thing (by thinner coils for example). In such a case, the Moller suppression would not be as good, but we would know quantitatively what the increase in FOM is and we would have the information available we need to evaluate tradeoffs.

* GEMS: It could be (we don't know for sure yet) that whatever is determined to be the optimum collimator (given all other constraints) will yield a beam profile that is larger than a 10 x 10 GEM. If that is the case then will will also compute the figure of merit for a collimator that would fit on a 10 x 10 GEM, so we have a basis of comparison.



Detailed plan:

Plan for finishing the final primary collimator design (first generated on Tuesday April 12; modifications added on Tuesday April 26)

1. Put in the critical elements of the main torus support structure. This is primarily the coil holders at the upstream end of the main torus. (We also found it was important to include some of the pieces of the downstream support structure).

2. Choose two z locations to focus on as the final primary collimator z locations. The criteria for determining these z locations will be:

a) For the "downstream" location: There should be enough room for:

* 5 cm of clearance between the most upstream part of the main torus support structure and the downstream side of the last ("cleanup") collimator

* room for the region 2 chambers to fit between the primary defining collimator and the cleanup collimator

* we will assume each of the collimators is 15 cm thick; That 15 cm corresponds to: 15 cm = 1 inch Al + collimator material + 1 inch Al so it is assumed that the actual collimating material is about 12.5 cm thick. Note that 20 radiation lengths of some typical collimating materials are: lead: 11.2 cm commercially available copper-lead alloy: 21.6 cm densimet: 8 cm (Just lead would be okay for the two "cleanup" collimators, but the latter two precisely machineable materials would be needed for the precision "primary" collimator)

b) For the "upstream" ("Boston") location. The exact location of this would be defined by leaving enough room for

* the GEMS upstream (?) of this collimator

* the mini-torus

3. For each z-location, we will pick a mini-torus location and determine the maximum size that the mini-torus coils and supports can be. This will hopefully be done in such a way that it is the coil holders at the upstream end of the main torus that ultimately limit the acceptance.

4. Once the above are set, we will determine the maximum possible acceptance for each z-location (with mini-torus on and off). By maximum possible acceptance, we specifically mean the following:

* lower theta will be the theta that just clears the the coil holder edge at the input to the main torus (or it may also be set by the minimum size beampipe we can accept)

* the limits in phi will be determined by the coil holder pieces at the upstream end

* the upper limit in theta will be more difficult to set. Presumably when theta gets big enough, the focus will become too poor or the inelastic separation will be too bad (or it may be determined by the cross bar at the downstream end of the main torus support).

5. After the "maximum possible" acceptance in step 4 is defined we will back off to "maximum reasonable" acceptances defined as follows:

a) Turn the hydrogen and aluminum in the target to vacuum (but leave multiple scattering and bremsstrahlung turned on so the collimator will work). This will hopefully allow us to define a relatively clean scattered electron profile at the input to the mini-torus.

b) Once step a) is completed, we will structure the collimator (both "Boston" and "downstream") such that the scattered electron profile has 2 cm of clearance all around at the main torus and at the mini-torus. This will define the maximum acceptable collimator sizes for the "trapezoidal" shaped collimator holes.

c) At each z location, we will also investigate the largest "square" collimator we can make subject to the above clearance criteria. By this we mean a collimator that has a larger phi range at smaller theta.

6. For each case ("Boston" and "downstream"), we will also determine the maximum collimator size that will allow the entire scattered electron profile to be accepted by the GEMs. For the lower theta here, we will ask Tony what is the smallest theta (for his z location) that he is comfortable with.

7. For all of the cases above, we will make tables of rates, FOM, etc. for the three most popular detector shapes (rectangle, v-shaped, and handlebar).

8. We will try to summarize the pros and cons of each location.

9. After a couple favored options are determined we will put in all the other necessary secondary collimators to make sure things haven't changed too much.

10. The "2 cm from the edge of support structures" is a fairly arbitrary criteria. At some point we will try to determine the actual radiation dose on the nearest pieces of support structure from the simulation.