Abstract
Theory and experiments show that vacuum fluctuations in quantum gravity can be abnormally strong, also at the micrometer or nanometer scale, for the following reasons: (1) the Einstein action is not positive-definite; (2) it is the only possible effective gravitational action; (3) quantum mechanics, in the form of the Feynman path integral, must apply to it, because any natural process is the result of all its possible quantum amplitudes; (4) due to (1), there are important non-classical virtual gravitational field configurations which can agree on a common phase. These field configurations can only interact directly with coherent matter, but can emit virtual gravitons which are absorbed by ordinary matter. All this makes possible, in principle, a vacuum thruster much more efficient than those based on the electromagnetic Casimir effect. We give an estimate of its efficiency based on the mentioned microscopic processes and on some parameters observed in experiments on anomalous forces with superconductors. With the observed energy efficiency of the order of 10-4 and an electric power of 10 W per kilogram of mass to propel, we find that a velocity of 0.1c can be reached in ca. 10 years. Possible improvements and practical limits are outlined. We discuss how the concept of ideal vacuum should be modified in order to allow a consistent description of these phenomena.