Slow transients and metastability in wormlike micelle rheology

The steady-state nonlinear rheology of wormlike micellar systems is thought to be subject to shear banding (the underlying shear stress vs. strain rate curve sigma((gamma))over dot is nonmonotonic). Shear banding may result in a plateau (sigma((gamma))over dot = sigma(p)) in the measured flow curve (at controlled mean strain rate (gamma)over dot). We present new rheological data for aqueous CPyCl/NaSal (100 mM/60 mM). Steady-state flow curves published previously for this system (Rehage H. and Hoffmann H., Mol. Phys. 74 (1991) 933) have since been interpreted as shear banded flow with ''top-jumping'', in which the steady-state shear rate (gamma)over dot(1) in the low shear band is the largest possible ((gamma))over dot(1) = (gamma)over dot(1)(max), sigma(p), = sigma(max)). That would rule out the existence of a metastable branch with a stress larger than a,. We show that such a branch does, however, exist (for temperatures in the range 20 - 25 degrees C). Similar results are found for a 100 mM/75 mM system. The time scale for relaxation of a metastable state onto true steady state flow, tau(ss), is far longer than the Maxwell time of the fluid; this is consistent with shear banding. We observe tau(ss) similar to ((gamma)over dot - (gamma)(c))(-p) in the metastable regime ((gamma)over dot greater than or equal to (gamma)(1)), with p an exponent that depends on composition and temperature. The ''critical'' shear rate (gamma)(c) is in some cases less than (gamma)over dot(1) so that no actual divergence of tau(ss) occurs. In at least one case, though, there is evidence for a physical divergence ((gamma)over dot(c) > (gamma)over dot(1)) accompanied by a small window of shear rates, (gamma)over dot(1) less than or equal to (gamma)over dot less than or equal to (gamma)over dot(c), for which tau(ss) is effectively infinite. In some respects the observed behaviour resembles that reported previously (Berret J.-F., Roux D.C. and Forte G.; J. Phys. II France 4 (1994) 1261) for equimolar CPyCl/NaSal in 0.5 M NaCL. Those results were interpreted in terms of nucleation and growth of a shear-induced nematic phase. However the same explanation is unlikely for the low weight fractions (phi less than or equal to 5%) used in our study.