Affine structures of decomposable solvable groups (Q2178530)

The notion of \textit{braces} was introduced by the author in [J. Algebra 307, 153--170 (2007; Zbl 1115.16022)] as a tool for solving the set-theoretic Yang-Baxter equation. In [Ann. Comb. 23, 391--416 (2019; Zbl 1458.20025)], he defined the notion of \textit{affine structures} and showed that a brace with adjoint group \(G\) is equivalent to an affine structure on \(G\). It was also shown that if \(G=AB\) for subgroups \(A,B\) with \(A\cap B=1\) (so \(G\) may be viewed as a bi-crossed product \(A \bowtie B\)), then affine structures on \(A\) and \(B\) satisfying some compatible conditions may be amalgamated to an affine structure on \(G\). All of these are reviewed in Sections 1 and 2 of the current paper. Suppose that we are in the above setting and \(A\bowtie B\) is endowed with an affine structure amalgamated from those of \(A\) and \(B\). In Theorem 1, it is proven that if \(\delta : B\times B\longrightarrow A\) satisfies suitable conditions, then one may modify this affine structure on \(A\bowtie B\) to obtain another affine structure. Let \(A\bowtie_\delta B\) denote \(A\bowtie B\) but endowed with this new affine structure, which the author calls the \textit{shifted} bi-crossed product of \(A\) and \(B\). In the language of braces, \(A\bowtie_\delta B\) is obtained from \(A\bowtie B\) by modifying the addition using \(\delta\) while leaving the circle operation unchanged. When \(\delta = 0\) is the trivial map, one gets back the ordinary affine structure on \(A\bowtie B\). When \(B\) acts trivially on \(A\), one recovers the \textit{asymmetric} product introduced in [\textit{F. Catino} et al., J. Algebra 455, 164--182 (2016; Zbl 1348.20002)]. When \(A\) acts trivially on \(B\), one recovers the \textit{lower shifted} semidirect product introduced in [Ann. Comb. 23, 391--416 (2019; Zbl 1458.20025)]. Thus, this construction via \(\delta\) extends and unifies these other product constructions that were previously known. In Theorem 2, all bi-crossed products \(A\bowtie B\) arising from \textit{bicyclic} braces \(A,B\) (i.e. both the additive and adjoint groups of \(A,B\) are cyclic) are classified.