When a two-body system is bound by a zero-range interaction, the corresponding three-body system -- considered in a non-relativistic framework -- collapses, that is its binding energy is unbounded from below. In a paper by J.V. Lindesay and H.P. Noyes it was shown that the relativistic effects result in an effective repulsion in such a way that three-body binding energy remains also finite, thus preventing the three-body system from collapse. Later, this property was confirmed in other works based on different versions of relativistic approaches. However, the three-body system exists only for a limited range of two-body binding energy values. For stronger two-body interaction, the relativistic three-body system still collapses. A similar phenomenon was found in a two-body systems themselves: a two-fermion system with one-boson exchange interaction in a state with zero angular momentum J=0 exists if the coupling constant does not exceed some critical value but it also collapses for larger coupling constant. For a J=1 state, it collapses for any coupling constant value. These properties are called "critical stability". This contribution aims to be a brief review of this field pioneered by H.P. Noyes.
The results of accurate variational calculations for the ground-state binding energies of several
N-boson systems in two and three dimensions are reported; the systems specifically considered are small (
N=2,3,4,and5) clusters of
4He atoms interacting through pairwise Lennard-Jones potentials. Evidence is found strongly suggesting each set of systems have a common functional relationship linking the binding energies and the De Boer quantum constant near threshold for self-binding. These calculations also confirm the existence of a unique threshold value of the interaction strength for all
N-boson systems in
2D. In addition, the linear relation between the binding energies of helium clusters present in
3D holds ture in
2D as well.
We estimate the ratio
R=g3/g2 of the critical coupling constants
g2 and
g3 that are required to achieve binding of two or three bosons, respectively, with a short-range interaction, and examine how this ratio depends on the shape of the potential. Simple monotonous potentials give $R\ensuremath{\simeq}0.8.$ A wide repulsive core pushes this ratio close to
R=1. On the other hand, for an attractive well protected by an external repulsive barrier, the ratio approaches the rigorous lower bound
R=2/3. We also present results for
N=4 bosons, sketch the extension to
N>4, and discuss various consequences.