Beam distribution on the collimators
for various accident cases

In this page, the particle distributions on the LHC collimators can be found for various accident scenario cases at injection (450 GeV) and at top (7 TeV). These distributions should be used as input for the FLUKA simulations of energy deposition in the collimator jaws of IR7.

For each considered cases, the matched beam distributions are given at the impact locations on the collimator. "Matched" means that the bunch particles have to correct transvrerse phase-space distribution (x, x', y, y') expected at the LHC. This is suitable for the FLUKA simulations that contains all the magnetic elements of IR7. The optics parameters are given at the beginning of the collimators and hence are suitable for the FLUKA tracking inside the collimator and in the downstream beam line.

Different failure scenarios are considered for injection and collision cases, as presented below. It is noted that the cases to be considered are several. As a guideline for the first simulations to perform, a priority list is provided in order to select the scenarios which are (expected) to be the most interesting.

Additional beam distributions at other locations then the ones considered below can be obtained upon request to Stefano Redaelli.

(1) Injection energy (450 GeV)

The worst considered scenario is the case of a full injection batch impacting on a primary or a secondary collimator. Since the injection form the SPS is done in the vertical plane, injection errors are likely going to generated large vertical oscillations. However, the cases of horizontal and skew errors will also be considered.
The damage level of the proton beam depends on the beam energy and size. In order to consider the worst case, the collimators where the beam size is smallest shall be taken into account. In the following graph, the transverse beam area is shown for all the collimators of IR7. The phase-I collimators are given in red.

Transverse beam size in mm2 as calculated for each collimator at IR7.
The optical beam size is considered (the contribution from dispersion is neglected.

Horizontal and vertical beam distributions with impact parameters of 1 sigma and 5 sigmas are considered for various collimators. For a typical beam size at 450 GeV of ~1mm, this corresponds to impact parameters of approximately 1 mm and 5 mm. The nominal gap settings of the collimators are assumed: 6 sigmas for the TCP's and 7 sigmas for the TCS's. The particle distribution are given at the beginning of each collimator.
The case of skew collimators is obtained similarly to the other cases but requires off-setting the bunch distributions in x and y simultaneously. This is done by following the same gap definitions used in SixTrack.

Parameters of beam distributions

(2) Collision energy (7 TeV)

The considered scenario is a wrong firing of the horizontal extraction kicker with nominal 7 TeV beam stored in the LHC. In this case, the various bunches of the beam experience different kicks during the kicker field ramp and are spread downstream of the kicker with different amplitudes. The normalized horizontal position of the bunch centroid versus time at a pi/2 phase advance downstream of the kicker is shown in the following graph.

Normalized horizontal bunch position as a function of time at pi/2 downstream of
the kicker in case of wrong firing at top energy. The nominal bunch spacing
of 25 ns is considered. The re-triggering time is 700 ns. Data by R. Assmann.

The kicker mis-firing is modelled by using the measured rising time and by assuming a re-triggering time of 700 ns (see Chapter 18 of the LHC design report). The extreme cases of first and last kicker module are simulated and the worst case is considered. The bunch amplitudes versus time are calculated at the worst location downstream of the kicker, i.e. at a pi/2 phase advance downstream of the kicker. Since various operation consitions must be considered, it is assumed that any collimator may be at a pi/2 phase advance from the kicker when the mis-firing occurs.

Correspondingly to the previous graph, the expected particle distribution versus x coordinate is given in the following plot. The distribution of hitting particles depend on the local optics parameters. As an example, the case of the first primary collimator (TCP.C6L7.B1) is considered. Here, it assumed that this collimator is at a pi/2 phase advance downstream of the kicker.

Transverse beam distribution at the first horizontal primary collimator (TCP.C6L7.B1) due to a kicker mis-firing at 7 TeV. It is assumed that collimator is at a phase of pi/2 downstream of the kicker. In the upper horizontal axis, the beam amplitudes in unit sigma are given.

In order to calculate the number of particles that can hit the collimator jaw, it is assumed that particles with amplitude above 10 sigmas are efficiently absorbed by the dedicated protection devices (TCDQ) whereas the particles with amplitude below 10 sigmas circulates in the ring for one full turn before they are extracted at the next passage trhough the kicker. It is assumed that the collimators may be set with a depth of 5 sigmas (the nominal settings of primary and secondary collimators are 6 and 7 sigma but some operational margin is assumed). Therefore, particles with amplitudes between 5 and 10 sigma might hit the collimator.

Zoom of the particle distribution of the previous graph for beam amplitudes below 55 sigmas. The particles that may hit the collimator are the ones with amplitude between 5 and 10 sigmas.

Since various operation consitions must be considered, it is assumed that any collimator may be at a pi/2 phase advance from the kicker. Therefore, any collimator can be hit by mis-kicker particles of amplitudes between 5 and 10 sigmas. The matched particle distributions should be generated at each collimator locations. In practise, only the case of horizontal collimators seems relevant. The skew collimators can also be hit by particle and, depending on their location, the corresponding effect could be relevant. This should be investigated.

It is noted that, since the collimators can be located either at at pi/2 or at 3/2pi phase advance from the kicker, both collimator jaws can be hit by mis-kicked bunches. One case or the other can be obtained by reflecting the particle distribution which are given below.

In the following input files, bunches with amplitudes between 1.5 sigmas and 15 sigmas have only been considered because bunches outside this range do not significantly contribute to the particles that might hit the collimator jaws. For each bunch, a Gaussian distribution is assumed.

Parameters of beam distributions

  S. Redaelli, Thursday, 05. April 2012 11:07 +0200