Several methods for measuring the compressive strength of strong particulate gels are available, including the centrifuge method, whereby the strength as a function of volume‐fraction is obtained parametrically from the dependence of equilibrium sediment height upon acceleration. The analysis used conventionally due to Buscall & White (1987) ignores the possibility that the particulate network might adhere to the walls of the centrifuge tube, even though many types of cohesive particulate gel can be expected to. The neglect of adhesion is justifiable when the ratio of the shear to compressive strength is small, which it can be for many systems away from the gel‐point, but never very near it. The errors arising from neglect of adhesion are investigated theoretically and quantified by synthesising equilibrium sediment height versus acceleration data for various degrees of adhesion and then analysing them in the conventional manner. Approximate correction factors suggested by dimensionless analysis are then tested. The errors introduced by certain other approximations made routinely in order to render the data‐inversion practicable are analysed too. For example, it shown that the error introduced by treating the acceleration vector as approximately one‐dimensional is minuscule for typical centrifuge dimensions, whereas making this assumption renders the data inversion tractable. © 2016 American Institute of Chemical Engineers AIChE J, 63: 1520–1528, 2017
Read full abstract