Abstract
The stress relaxation of concentrated polymer solutions after double‐step strain is analyzed by the model of Doi and Edwards (DE model). The theory explains the experiments of Osaki et al., who showed that the Bernstein–Kearsley–Zapas (BKZ) constitutive equation works reasonably well when the successive deformations are applied in the same direction, but that it fails when the direction of the second deformation is reversed. This experimental result has been considered contrary to the DE model, since the previous theory of Doi and Edwards provides a BKZ‐type constitutive equation. In the present paper the discrepancy is reconciled by showing that (i) the source of the discrepancy is a mathematical approximation used in the previous DE theory, and that (ii) if this approximation is removed, the agreement between the theory and experiment becomes fairly good.

This publication has 17 references indexed in Scilit: