Pbar yield optimisation based on codes written by:

S. Van der Meer;
C. Hojvat & A. van Ginnekan;
S. Hancock & T. Sherwood;
N. Mokov (an early version of the MARS code);
N. Walker.

CERN PS production beam: 26 GeV, 1 to 1.5 x 10¹³ ppp,  5 bunches in 500 ns.  Repetition period 2.4 s.

Pre-focusing by quads - beam spot > 2 mm diameter.

Pre-focusing by lithium lens (20 mm dia, ~650 Tm^-1) much better, beam spot = 2 mm diameter (99%).

Target: Cu, Pb or W in graphite, air cooled, with magnetic horn (7ms risetime)  -  Cu retained for operation.

Target damage - fracture of Cu rod after several days of operation, but pieces contained by graphite surround with ~ 30% loss of yield that was then stable during months of operation.

Yield improvement programme started in 1980 for pbar and pions (because of interest in using long straight-section of AA downstream of the target area as a neutrino source). The latter was soon abandoned because of fears that it might interfere with the pbar collider work. Fermilab Li lens tested - 20 mm dia, 500 kA peak, pulse rise time 160 ms , transformer ratio 24:1, gradient at 60% peak (best linearity) 600 Tm^-1

REXCO used to study shock-waves in solid targets - predictions looked rather wild probably due to use outside valid parameter space. At least one prediction: that the Cu target material would suffer from reduced local density was invalidated by measurement of pion yield.

Li lens collector needed shorter, closer target. Chose iridium or rhenium and ran into containment problems Two steel containers suffered fatigue failure at downstream exit window. Problem solved by using Ti alloy container with improved cooling, (lens parameters: 20 mm dia, gradient 700 Tm^-1).

Enhanced yield could be obtained using pulsed-current target. Increased target length, i.e. lower Z material for improved yield since depth-of-focus problem is eliminated, (target parameters: 3 mm dia, 150 kA, pulse rise time 10 ms, gradient 1.3x10⁴ Tm-1)

Rigid Cu-Be rod - failed after ~6 h

Ag rod contained in high-pressure vessel - melted at downstream end

Fe rod contained in improved high-pressure vessel - survived test run.

Liquid metal targets were looked to as a solution to the radiation and shock-wave damage effects, but containment of the liquid metal (medium to high Z) appeared to be too difficult within the scope of our programme. This was particularly so for pulsed-current targets where metal electrodes were needed. A mercury jet target was designed and a laboratory prototype operated well. It was abandoned because of toxicity and also because it could not be adapted to pulsed-current mode (gliding discharge: all the current flows along a surface layer (Jaeger et. Al, Graz University).

Optimisation of yield from solid targets was studied by various people using several codes. This led to the 36 mm Li lens pulsed at 1.3 MA (from Novosibirsk) - gives better depth of focus for the same gradient as the 20 mm lens. Predictions validated, but lens and transformer damaged during test run, (lens parameters: 36 mm dia, 1.3 MA peak, rise time ~1 ms, gradient at 75% peak 600 Tm^-1)

Plasma lens tested in pbar target area - performance as good as 20 mm Li lens but maintenence needs greater. Theoretical improvement over Li lens due to reduced pbar absorption was not achieved (due to lack of time?).

Final solution: Ir target in water cooled Ti alloy container with 20 mm Li lenses upstream and downstream (the former was never installed due to downsizing of the pbar programme). Addition of the third type of pulsed current target (see above) would have given a factor of two yield enhancement.

Beam transport: immediately after collector lens - fully rad-hard quads and dipoles, then magnets with radiation resistance > 10⁹ Rad. Losses of antiprotons at large betatron amplitudes (within the first few ms) was extensively studied and some improvements resulted, but overall yield into the machine acceptance was never better than ~30% of the anticipated full-aperture yield based on extrapolation of the yield measured at 20 % of the machine acceptance (using collimators). Best yield into ACOL was: 8x10^-6 pbar/p. The measured pbar emittances at 3.5 GeV were ~ 1.5x10^-4 m. Missing factor never better than 3 to 4.