Overview of the meetings of the CLIC parameter working group
First meeting, 12 January 2001
Jean-Pierre Delahaye presented some
slides
to question the centre-of-mass energy, the luminosity, the gradient and the
frequency of the main CLIC parameter list.
After a lively discussion no conclusion was drawn sofar. A
clarification of the requirements from the physics side is necessary.
Second meeting, 2 February 2001
Conclusions on frequency and centre-of-mass energy
Regarding the last meeting and the meeting on the future of CERN,
the following conclusions were drawn (see Daniels
slide). The working group recommends that we
concentrate our work on a centre-of-mass energy of 3TeV in order to meet the
needs of the experiments. The acceleration frequency should remain 30GHz.
Possible determination of the gradient
The final centre-of-mass energy will be a function of the gradient, fill
factor, site length, length of the beam delivery system and the overheads
for single-bunch beam-loading compensation, BNS damping and energy feedback.
Most parameters seem relatively certain,
see here. One is thus able to determine the
necessary acceleration gradient from these parameters.
Length of the beam delivery system
Frank presented two possible options for the final focus system, one is
3.3km long the other one about 1km. In the first one, one could hopefully
include the collimation of transverse tails. In the latter one, one might be
able to implement part of the transverse collimation. Also
energy collimation is necessary. Using an NLC lattice as a basis Thys has
found that the system needs to have a length of 4km. One might hope to
implement the second half of this system into the long final focus system.
Therefore two main options for the beam delivery region exist.
With the long final focus (3.3km) one has 2km of energy collimation,
assuming that the second stage of the energy collimation and the transverse
collimation can both be implemented in the final focus system. The total length
per side is a minimum of 5.3km.
With the short final focus system (1km) one needs 4km of
energy collimation. While it may be possible to shorten this a little,
transverse collimation is needed.
In both cases additional space for the matching is not yet included.
Franks summary gives a slightly different
results partly because he now includes the diagnostic section.
The working group therefore suggests (including Frank) that the target value
for the total length of the beam delivery system (of both sides) should be
10km.
Possible implication for the gradient
To achieve the required centre-of-mass energy for a total length of 40km,
the gradient would thus have to be in excess of
137MV/m.
Determination of parameter sets for lower energies
Two main different approaches
to determine the parameters at lower energies
exist. The working group suggests to use the one which concentrates on a
centre-of-mass energy of 3TeV and derives the other parameters from there.
Call for collection of critical parameters
Daniel asked the members of the working group to help to identify the
critical parameters for different parts of the machine.
A non-complete list for the main beam is given
here.
Third meeting, 2 March 2001
Limitations of the Site Length
Michel Poehler gave an overview of the possible limitations of the site length.
If one wants to be close to the CERN site, the length can be up to 40km without
too much of a problem. At each end an extension of another 5km might be
possible. But this might lead to severe technical problems and might increase
the cost significantly. A site length of more than 50km seems excluded.
It was therefore agreed to keep a site length of 40km. If one wants to go
beyond a centre-of-mass energy of 3TeV, one might consider some extension.
Limitiations of the Gradient
Walter presented our knowledge of the gradient limitations. The pulsed heating
by the RF will induce stress in the structures. The acceptable limit of this
stress seems to be given by the yield strength. Depending on the treatment of
the copper the yield strength varies strongly. For cold worked copper and for
some copper alloys 150MV/m seems well possible. The discussion of the RF
breakdown and the damage gave the same result, 150MV/m is not excluded.
Conclusions Drawn
A consensus was reached to propose to stay with 150MV/m and a site length of
40km. It would be very important to test experimentally whether the damaged
structures are damaged during processing or continously. Due to limited
resources it is likely that this tests can be done only when CTF3 and/or the
gyroklystron is running.
Fourth meeting, 16 March 2001
Status of the Damping Ring
John presented the current status of the damping ring. By increasing the
beam energy in the damping ring and with the help of wigglers it was possible
to achieve the required parameters.
The vertical emittance can be 3nm which is the lower value of the range
specified in the parameter table of the CLIC-report. The horizontal emittance
can be as low as 200nm which is a factor 2 better than required.
Not all instabilities have been investigated so-far, but the intra-beam
scattering has improved drastically so that it will not prevent us from
achieving the target parameters. More studies
have to be done in order to ensure that they can also be achieved with
imperfections.
Sixth meeting, 16 March 2001
Drive Beam Parameters
Roberto presented the status of the drive beam parameters (his transparencies
are here).
We have a fairly good idea how to derive the parameters for the drive beam
from those for the main beam. The total length of the accelerator determines
the total length of the drive beam pulse at production. The pulse length
necessary in the decelerator is determined by the fill time of the main linac
structure and the main linac pulse length. This in turn defines with the
selected multiplication factor the number of drive beams. The main linac
accelerating structures also determine the necessary RF power which in turn
defines the energy per drive beam. The choice of the drive beam
intensity is taken to get the stablest deceleration. Finally a cross check
needs to be perfomred to test limits from coherent synchrotron radiation,
normal synchrotron radiation, combiner ring impedance requirements, drive
beam accelerator stability and so forth.
The optimisation of the beam intensity has been done using an ad-hoc scaling
law based on thre existing structures ate 15mm, 20mm and 40mm aperture.
It has been proven useful but should be tested in a more systematic way.
Based on this scaling it was demonstrated that the highest intensity is the
most stable condition. The limitiation on the intensity arises from the
single bunch charge limit in the drive beam injector. It was concluded that
it would be extremely useful to do this investigation using systematic
calculations of the decelerating structures.
The four and six waveguide structures have a potential problem with dipole
modes at frequencies significantly below the fundamental. Also the resonant
damping seems to lead to problems, the two programs which calculate the
structures do not agree. The situation needs some clarification.
Other structures are under investigation, for example one with an axial
symmetry.
In the discussion it was also very clear that we have difficulties in changing
the parameters as long as we do not know which horizontal emittance might
be achievable in a damping ring.
It was concluded that, it might be a good idea to go through the
exercise of assuming a new accelerating structure with smaller a/lambda.
This would allow to understand the implications in some more quantitative
way.
Seventh meeting, 8 June 2001
Rediscussion of Parameters at Different Energies
Gilbert presented some transperencies to rediscuss the choice of parameters at
different energies. The NLC study claims that at Ecm=3TeV a total length of
the two beam delivery systems can be limited to 5km. Following the method to
determine the gradient established in the second meeting, this would allow
to reach E_cm=3TeV in the available site length with a gradient of only
120MV/m, see slide from second meeting.
Gilbert also suggested to reduce the bunch distance from 20 to 16 buckets,
the feasibility of this had been demonstrated by Daniel in the RF power source
meeting. Gilbert also suggest either to keep the charge per bunch the same
which allows to reduce the pulse length by 20%. The drive beam sector length
would be reduced from 625m to 500m and the energy gain per sector would
be about 45GeV. The other solution may be to keep the pulse length constant
but to reduce the charge per bunch by 20%. The energy gain per drive beam
sector would then be about 56GeV.
Gilbert also presented two possible low energy options. As presented in the
fifth meeting (minutes confidential sofar), one needs only one drive-beam
accelerator, which feeds both linacs. In one scheme the gradient in the
main linac structures is reduced by a factor sqrt(2) by reducing the
drive beam current by the same factor. This allows to reach an initial drive
beam neergy that is larger by a factor sqrt(2) the the nominal value. As
a result, the drive beam can be used for two nominal drive beam sector lengths.
The natural energy step in this case is 100GeV per side. One could split the
one drive beam acclerator into two at half the energy each. This would
lead to an energy step size of 50GeV per side. The same could be achieved using
the double pulse technique described in the fifth meeting, which would save
one of the two drive beam accelerators.
By reducing the bunch spacing, keeping the bunch charge, decreasing the
gradient to 120MV/m and using double pulses one could obtain steps of 45GeV per
side.
In the discussion it was emphasised, that a Z0 factory can also be easily
built if the natural energy step is more than 45GeV per side. It remains to be
investigated if the bunch charge can be kept constant at the lower gradient.
Ninth meeting, 24 August 2001
Gradient Discussion
Frank discussed the necessary length of the beam delivery systems. Currently
a solution exists with 12.4km length for both sides. If one reduces leaves one
of the two energy collimation stages out this could be reduced to about 7.6km.
Using a non-linear system there is hope to be able to reduce the system length
to 4-6km. It would then be possible to use a gradient of only 120MV/m and still
to reach a centre-of-mass energy of 3TeV.
In the further discussion it was questioned whether the best approach would be
to adopt 120MV/m in the design and change all parameters accordingly or
to stick with the old parameters. In the latter case more resources could be
put into the improvement of the design and to optimize basic choices,
as for example the accelerating frequency. The general agreement was that if
one can define a good study of these choices which can give results by the end
of next year it would be reasonable to concentrate on this.