In contrast to recent publications suggesting that particle cluster fo
rmation alone can play an important role in the shear thickening flow
behavior of concentrated colloidal suspensions, we believe that there
is little if any substantive evidence to prove it. To support this vie
w, we use data from various studies, including data from studies concl
uding that layered flow is not involved. One reason for the confusion
seems to center around the inability of various light-scattering and n
eutron-scattering techniques to show particle layering before shear th
ickening when the layers are not well defined. In this regard, one sho
uld understand that layered flow san occur without rigorous ordering o
f particles within the layers, and as the flowing suspension approache
s the point of instability, the hydrodynamic forces driving for the in
stability will jostle the particles within the layers sufficiently to
make it even harder to see the layering and any ordering, if it exists
, within the layers. Having these views, we argue that the process des
cribed by Hoffman (1972, 1974) for shear thickening is still applicabl
e with refinements. The major refinement is the idea that, after the h
ydrodynamic forces cause the instability which breaks up the layered f
low, particle jamming probably involves cluster formation both with an
d without particle contact. Particle roughness and angularity will fac
ilitate the contact. Finally, we agree with various authors who argue
that the best chance of finding shear thickening in concentrated collo
idal suspensions without layer formation lies in Brownian hard-sphere
suspensions, but the evidence given for it so far is not definitive. (
C) 1998 The Society of Rheology.