Potassium
dihydrogen phosphate (KDP) is an important nonlinear material due
to its excellent physical and optical properties. However, it is also
a difficult-to-machine material due to its complex anisotropic microstructure.
To better understand the deformation mechanisms under external stresses,
this paper aims to carry out systematic nanoindentation simulations
using molecular dynamics (MD). To facilitate the structural characterization
of KDP, a machine learning-based method was developed. The results
showed that the subsurface damage is obviously anisotropic. On the
(001) surface, both tetragonal and monoclinic phases appear simultaneously
and part of the monoclinic phase transfers to the tetragonal phase.
The generated phases close to the surface undergo amorphization and
are squeezed out to form pile-ups. On the (100) surface, however,
an orthorhombic phase emerges directly from the original structure
rather than transforming through the monoclinic phase. No amorphization
happens and no pile-ups appear in this case. The first “pop-in”
in the load–displacement curve of nanoindentation signified
the emergence of phase transformation under the combined hydrostatic
and shear stresses. After unloading, the recovery of the deformed
KDP is also anisotropic. The maximum recovery takes place when the
indentation is on the (100) surface.