Context . Full sky coverage adaptive optics (AO) on extremely large telescopes requires the adoption of several laser guide stars as references. With such large apertures, the apparent elongation of the beacons is absolutely significant. With a few exceptions, wavefront sensors (WFSs) designed for natural guide stars can be adapted and used in suboptimal mode in this context. Aims . We analyse and describe the geometrical properties of a class of WFSs that are specifically designed to deal with laser guide stars propagated from a location in the immediate vicinity of the telescope aperture. Methods . We describe, in three dimensions, the loci where the light of the laser guide stars would focus in the focal volume located behind the focal plane where astronomical objects are reimaged. We also describe the properties of several types of optomechanical devices that act as perturbers for this new class of pupil plane sensors, through refraction and reflections. We refer to these as ingot WFSs. Results . We provide the recipes both for the most reasonably complex version of these WFSs, with six pupils and, for the simplest one, only three pupils. Both of them are referred to on the basis of the European Extremely Large Telescope (ELT) case. We outlined elements that are meant to give a qualitative idea of how the sensitivity of this new class of sensors compares to conventional ones. Conclusions . We present a new class of WFSs, based on an extension to the case of elongated sources at a finite distance of the pyramid WFS. We point out which advantages of the pyramid can be retained and how it may be adopted to optimize the sensing procedure.

Ingot-like class of wavefront sensors for laser guide stars