A prototype C(6+) injector using a laser ion source has been developed for a compact synchrotron dedicated to carbon ion radiotherapy. The injector consists of a laser ion source and a 4-vane radio-frequency quadrupole (RFQ) linac. Ion beams are extracted from plasma and directly injected into the RFQ. A solenoid guides the low-energy beams into the RFQ. The RFQ is designed to accelerate high-intensity pulsed beams. A structure of monolithic vanes and cavities is adopted to reduce its power consumption. In beam acceleration tests, a solenoidal magnetic field set between the laser ion source and the RFQ helped increase both the peak currents before and after the RFQ by a factor of 4.

A simplified UV-preionized long-pulse TEA CO2-laser amplifier is proposed, which is powered by a common Marx generator for both the preionization arc discharges and the main laser discharge. The measurements of small-signal gain reveal that there exists an optimum capacitance for the capacitors which are provided in the preionization circuit for the control of preionization level. The physical interpretation for this is also given. A small-signal gain higher than 3%/cm and an extraction energy of nearly 14 J/l (at the input energy level of nearly 1 J) have been achieved. It has also been demonstrated that the amplifier can amplify a long laser pulse of the order of microseconds without serious distortion of pulse waveform.

Experimental studies were performed on the interactions of a CO2 laser beam of long pulse duration (∼50 ns FWHM) and peak power up to 4×1012 W/cm2 with plasma blobs produced by this laser from a massive planar target of aluminum or polyethylene. Special attention was paid to the laser reflectivity (R) due to the stimulated Brillouin backscattering (SBS), in particular to the correlation between the SBS reflectivity and the e-folding scale-length for spatial plasma density variation (L n ). It is shown that (1) the experimental L n -values are consistent with the theoretical ones predicted from an ellipsoidally-expanding self-regulating plasma model, and (2) taking those predicted L n -values into account, the experimental R-values agree well with those estimated from an existing simulation model, which is restricted to a relatively low percentage (15–20%) even for longer pulse duration.

No abstract

No abstract