DL.AC.UK>
To: <FFAG-E-MODEL
LISTSERV.CCLRC.AC.UK>
Subject: Second iteration of engineering design
Date: 2005 August 03; Wednesday 15:59
Hi all,
Based on the latest information we have, our engineers have produced a
second iteration of a possible engineering layout for the FFAG electron
model.
The drawings are based on:
- 15.9m circmference
- 42 cells
- FD structure a la PAC 2005 paper
- 1.3 GHz ELBE-type cavities
I enclose PDFs of the latest engineering layouts. Broadly speaking,
there are two possible schemes, for which we can summarise the benefits
and disadvantages as:
Scheme 1: Circular cross-section pipe. This has the advantage of perhaps
being a bit cheaper, and maybe keeps the horizontal aperture in the F
element small. However, it means that the vertical aperture in the D
elements is the same as the horizontal. I have heard that this is a big
problem in terms of supplying the required 0.2 T poletip field. Note
that in order to keep the fabrication sensible each vessel (2 cells
minus one cavity) is made up of three straight pieces welded together.
Depending on the bend in the Fs this will slightly increase the required
apertures. Also note that there are tapers to reduce the aperture to
that of the cavities, one of which is asymmetric (a cone cut not
perpendicular to its axis). There are therefore no steps at the flange
between vessel and cavity.
Scheme 2: Rectangular box-section fabrication. The philosophy is to
allow the D vertical aperture to be reduced to a smaller value whilst
managing the complexity of the required transitions and keeping the
fabrication sensible. Again, the vessel is made of several straight
pieces, which again means the actual horizontal aperture is slightly
larger than that required in a couple of places (see dotted lines on the
layout). However, the fabrication requires that the vessel in the F
elements has a square rather than circular cross-section. If the F is a
conventional-looking quadrupole then this will increase its required
inscribed radius. To assist in magnet design consdirations, illustrative
cross-sections for the F and D areas have been shown on the layout
drawing (erl-180-10147).
One issue in Scheme 2 is that of the tapers. These can either be
constructed from flat plates, which gives a rectangular aperture
cross-section at the end flanges, or they can be spark-eroded to provide
a transition from a rectangular to circular cross-section (possibly a
bit more expensive). The choice I guess depends on how important it is
to avoid steps at the flanges into and out of the cavities - these steps
would be a few mm in height against 25.4mm bore for the cavity. I would
appreciate any input people have on this.
At the moment, I would say that Scheme 2 is the preferred option.
Also enclosed are concepts for the girder arrangements and a view of the
complete accelerator.
Other questions that we have at the moment:
1. What is the present design of D dipole that we should be working to?
Our engineers are starting to look at the 200mrad injection/extraction
lines. One issue is clearance past the dipoles. A revised layout will be
circulated soon which also includes rotating the cavities so that the
tuner is at 45 degrees down, to help the beam clear the input couplers.
Does anyone see any problem with this as a philosophy?
2. For the F element, what should we be looking at? Quad, dipole etc.? A
best guess as to the appearance of this element would be great. Does the
increased aperture keep this element feasible? (it seems so)
3. From the engineering layouts, where do people think the BPMs should
go? Note that there is only 1 bellows per 2 cells (or around 70cm) -
still 21 bellows which is plenty! This means that for position stability
the BPMS are best next to the cavities, but what about RF noise?
4. What should the OTRs look like? I imagine a transverse sliding frame
to image the beam at different distances. The opening angle of the OTR
radiation lobes is severe at 10-20 MeV (I've not seen that mentioned
elsewhere yet), which means that the collecting optics have to be placed
close to the OTR screen to get a sensible S/N. Who is looking at this?
Also, how many OTRs do we need? i.e. can we just replace some of the
pumping crosses with OTRs?
Best,
Hywel
-------------------------------------------
Hywel Owen h.owen
dl.ac.uk
Accelerator Science and Technology Centre
CCLRC Daresbury Laboratory
United Kingdom
Tel/Fax: +44 1925 603797/603192
-------------------------------------------
"Errors using inadequate data are much less than those using no data at
all."
-- Charles Babbage
![[pdf]](/uknf/hidden/file_pdf.png)
erl-180-10142 Scheme 1 Circular Cell.pdf