The state parameter (y) measures the distance from the current state to the critical state line (CSL) in the
compression plane. The existence of a correlation between both the peak angle of shearing resistance (�#
$ )
and peak dilatancy and y is central to many constitutive models used to predict granular soil behaviour. These
correlations do not explicitly consider particle crushing. Crushing induced evolution of the particle size
distribution influences the CSL position and recent research supports used of a critical state plane (CSP) to
account for changes in grading. This contribution evaluates the whether the CSP can be used to calculate y
and thus enable prediction of the peak angle of �#
$ and peak dilatancy where crushing takes place. The data
considered were generated from a validated DEM model of Fontainebleau sand that considers particle
crushing. It is shown that where y is calculated by considering the CSL of the original uncrushed material there
can be in a significant error in predicting the material response. Where the CSP is used there is a significant
improvement in our ability to predict behaviour whether the CSP is accurately determined using a large
number of tests or approximated using crushing yield envelopes. It is shown that the state parameter
calculated using the previously available definition can give a false sense of security when assessing
liquefaction potential of potentially crushable soils. The contribution also highlights the stress-path
dependency of the relationship between �#
$ and y whichever approach is used to determine y