The effects of non-hydrostatic compression and applied electric field on the electromechanical behavior of poled PZT 95/5-2Nb ceramic during the F{sub R1} {yields} A{sub 0} polymorphic phase transformation

We conducted hydrostatic and constant-stress-difference (CSD) experiments at room temperature on two different sintered batches of poled, niobium-doped lead-zirconate-titanate ceramic (PZT 95/5-2Nb). The objective of this test plan was to quantify the effects of nonhydrostatic stress on the electromechanical behavior of the ceramic during the ferroelectric, rhombohedral {yields} antiferroelectric, orthorhombic (FE {yields} AFE) phase transformation. We also performed a series of hydrostatic and triaxial compression experiments in which a 1000 V potential was applied to poled specimens to evaluate any effect of a sustained bias on the transformation. As we predicted from earlier tests on unpoled PZT 95/5-2Nb, increasing the stress difference up to 200 MPa (corresponding to a maximum resolved shear stress of 100 MPa) decreases the mean stress and confining pressure at which the transformation occurs by 25--33%, for both biased and unbiased conditions. This same stress difference also retards the rate of transformation at constant pressurization rate, resulting in reductions of up to an order of magnitude in the rate of charge release and peak voltage attained in our tests. This shear stress-voltage effect offers a plausible, though qualitative explanation for certain systematic failures that have occurred in neutron generator power supplies when seemingly minor design changes have been made. Transformation strains in poled ceramic are anisotropic (differing by up to 33%) in hydrostatic compression, and even more anisotropic under non-hydrostatic stress states. Application of a 1000 V bias appears to slightly increase (by {le}2%) the transformation pressure for poled ceramic, but evidence for this conclusion is weak