7.1 Van de Graaff accelerator operations and development

D.T. Corcoran, G.C. Harper, C.E. Linder, A.W. Myers, T.D. Van Wechel and W.G. Weitkamp

This year we found a leak in an old, five-minute epoxied half-inch NPT joint in a stripper box flange. This joint did not noticeably leak below 140 psig tank gas pressure, but above 140 psig tank gas pressure the HE ion gauge reading increased linearly with tank gas pressure up into the 10^-6 Torr range. When the NPT hole was filled and welded closed, the beam tube vacuum stayed well down in the 10^-7 Torr range to 225 psig tank gas pressure.

We developed unstable operation at 8.5 MV during an experiment this year. Local, small-scale sparking seemed to occur around the LE chain pellets as they left the terminal for a distance of 5 or 10 feet (as seen from a top port). When the LE chain charge was increased this sparking increased in brightness and frequency, and tank sparks were triggered. Upon opening we found that the upper LE pick-off pulley¹ had thrown its rubber tire, and the aluminum rim was being rubbed by the pellets. There were tiny metal chips dusted over the terminal, beam tube, and column. Replacement and cleanup allowed normal, high-voltage terminal operation.

During another experiment, a 1.5 Hz ripple with amplitude 1 to 4 µamp was observed on the high energy chain current meter. Visual inspection of the chain showed a side-to-side harmonic oscillation coincident with this ripple. The low energy chain seemed to have a smaller oscillation at twice the frequency. The frequency of the HE chain was strongly affected by added manual downward pressure on the counterweight. The carriage seemed to be sticking, so that the counterweight could not move it freely, and the chain had some link-joints that moved stiffly. LPS2 was applied to all bearings and links of both chains. After this treatment there was absolutely no observable ripple on either LE or HE chain current meters at 10 kV charge (in air).

Tank sparks have repeatedly taken out the pelletron charging resistors, high voltage charging supplies, and RG-8/U cables at the low energy end, with the negative supply being the most frequent victim. Spark bar gaps were measured to be within tolerance. High-power, high-surge, ceramic composition resistors were installed in place of the wire wound series charging resistors provided with the pelletron. These have withstood repeated tank sparks without being destroyed. Additional transient suppression has been added to the charging supplies at both ends of the machine. This is in the form of series impedances and shunt clamps attached to wires from the high voltage sections of the supplies into the electronics.

Current sensors for both high and low energy columns were connected to the ADC card of the terminal computer as a diagnostic. This extra column current measurement helps us to see whether the beam is hitting one of the beam tubes.

The stripper canal actuator was driven past its limit due to a faulty diode in the motor circuit. This destroyed the special, high-pressure bellows seal. The high cost of an equivalent bellows necessitates re-engineering of the motion feedthrough mechanism. The canal has been left out, as no current experiments require gas stripping.

The Leybold TMP-450 turbo-molecular pump at the low energy end, which has failed twice in the past, failed for a third time this year. We installed a greased bearing TMP-360 on this line. This allowed us to eliminate a 90° elbow, giving us about the same pumping speed at the beam tube.

Detailed, illustrated procedures were written this year for several of the gas handling operations and vacuum system operations involving the tandem. These have been included in the general operations manual for ion sources and beam transport and also posted at the appropriate systems.

We are experimenting with the use of corona points made from Thoriated Tungsten TIG welding electrode material ground with a toolpost grinder to a long taper. As there has been no observable degradation in sharpness since their installation ultimate point life cannot be accurately estimated, but it appears to be very long.

The terminal ³He ion source was installed, used, and removed, and the Van de Graaff was restored to tandem operation successfully three times during the last year.

During the year from March 1, 1995 to February 29, 1996 the tandem pellet chains operated 3162 hours. Additional statistics of accelerator operations are given in Table 7.1-1.