Most laser wakefield accelerator (LWFA) experiments to date have operated in the self-modulated (SM) regime and have been self-guided. A channel-guided LWFA operating in the standard or resonant regime is expected to offer the possibility of high electron energy gain and high accelerating gradients without the instabilities and poor electron beam quality associated with the SM regime. Plasma channels such as those produced by a capillary discharge have demonstrated guiding of intense laser pulses over distances of several centimeters. Optimizing the performance in a resonant LWFA constrains the on-axis plasma density in the channel to a relatively narrow range. A scaling model is presented that quantifies resonant LFWA performance in terms of the maximum accelerating gradient, dephasing length, and dephasing-limited energy gain. These performance quantities are expressed in terms of laser and channel experimental parameters, clearly illustrating some of the tradeoffs in the choice of parameters. The predicted energy gain in this model is generally lower than that indicated by simpler scaling models. Simulations agree well with the scaling model in both low and high plasma density regimes. Simulations of a channel-guided, self-modulated LWFA are also presented. Compared with the resonant LWFA regime, the requirements on laser and channel parameters in the SM regime are easier to achieve, and a channel-guided SM-LWFA is likely to be less unstable than a self-guided SM-LWFA.

• Received 4 October 1999

DOI:https://doi.org/10.1103/PhysRevE.63.036502