A limit of nonplanar 5-body central configurations is nonplanar (Q6193976)

The authors, here, by the term ``limit'' mean the limit of a converging sequence in the space $(\mathbb{R}^p)^n\times \mathbb{R}^n$, where $p$ is the dimension of the position space, $(q_1,q_2,\dots,q_n)\in (\mathbb{R}^p)^n$ together with positive masses $(m_1, m_2,\dots, m_n)\in \mathbb{R}^n$. The dimension $d$ of a configuration is the dimension of the space generated by the vectors $(q_2-q_1, q_3-q_1,\dots,q_n-q_1)$ thus $d\leq p$ and $d\leq n-1$. Each position tends to a limiting position end each mass tends to a limiting mass, which is assumed to be positive. The corresponding result is true for $n\leq 4$, unknown for $6\leq n\leq 945$, false for $n=946$. If $n=5$ there are two independent vertical eigenvectors with nonzero eigenvalues, since the maximal number of such independent eigenvectors is $n-3$: ``A 5-body central configuration which is the limit of a sequence of 5-body central configurations of dimension $d$ is of dimension $d$.'' The authors, in this article made use the Moeckel [\textit{R. Moeckel}, Math. Z. 205, No. 4, 499--517 (1990; Zbl 0684.70005)], and Moeckel and Simo [\textit{R. Moeckel} and \textit{C. Simó}, SIAM J. Math. Anal. 26, No. 4, 978--998 (1995; Zbl 0830.34032)], who proved that, while continuously changing the masses, a 946-body planar central configuration bifurcates into a spatial central configuration. The authors show that this kind of bifurcation does not occur with 5-bodies. Question 17, also, in the list [\textit{A. Albouy} et al., Celest. Mech. Dyn. Astr. 113, 369--375 (2012)], is thus answered negatively. They first recall the construction of the Winter-Couley matrix and of the shifted Winter-Couley matrix, and they will show in a future work that a nontrivial degeneracy of the latter has surprising consequences if $d=n-3$. (Data sharing was not applicable to this article as no datasets were generated or analyzed during the current study).