Engineers have established a streamlined displacement self-sensing framework for piezoelectric ceramic stack actuators that accurately reconstructs dynamic mechanical motion directly from electrical measurements without external hardware sensors. By developing extended piezoelectric constitutive equations, the research team decoupled electromechanical responses into distinct dielectric polarization components and linear piezoelectric coupling terms, overcoming the heavy computational burdens that historically plagued charge-based tracking. The new analytical architecture eliminates secondary instruments by extracting free-end output displacement directly from driving voltage and total accumulated charge flowing through the multilayer ceramic structure. Under rigorous experimental evaluations spanning driving potentials up to seventy volts and excitation frequencies between one and fifteen hertz, the resulting five-parameter formulation achieved an average relative estimation error below five point seven percent.
Despite these outstanding electromechanical traits, the ferroelectric ceramic materials integrated within these actuators display severe intrinsic nonlinearities that produce a pronounced hysteretic relationship between applied excitation voltage and resulting physical stroke. Self-sensing methodology resolves this spatial conflict by treating the actuator simultaneously as an active force generator and an intrinsic position sensor, decoding mechanical displacement directly from accessible electrical terminal variables.
Early implementations of piezoelectric self-sensing fundamentally relied on classical linear constitutive equations, which assumed that the total electrical charge within the actuator consisted solely of an applied free charge alongside a strain-proportional polarization charge. Conventional schemes employed analog bridge circuits to separate these charge components, yet their measurement precision deteriorated rapidly because the equivalent capacitance of ferroelectric ceramics fluctuates substantially under changing preloads and dynamic deformations. Furthermore, because classical constitutive laws neglect polarization hysteresis and nonlinear electroelastic coupling relative to varying electric fields, subsequent researchers incorporated complex phenomenological operators, such as the Prandtl-Ishlinskii or Bouc-Wen formulations, into electromechanical models. Other investigators explored impedance analysis or high-order differential equations with nonlinear capacitance approximations, but these alternative techniques suffered from excessive instrumentation expense, noise sensitivity, and limited physical interpretability.
Under this derivation, the material is initially subjected to a unipolar electric field under mechanically unconstrained conditions, separating total electric displacement into a pure dielectric polarization term and a piezoelectric-coupling displacement term. Because non-180-degree ferroelectric domain switching completely dominates under unipolar excitation, the piezoelectric-coupling electric displacement remains strictly synchronous and proportional to piezo-induced strain, even when the underlying crystalline conversion displays pronounced hysteresis. In the second analytical step, mechanical stress is applied under an invariant electric field, dividing the elastic response into an uncoupled elastic strain and an electromechanical coupling strain that mirrors the piezo-induced electric displacement. By mathematically isolating purely dielectric permittivity and purely elastic compliance from energetic coupling processes, the extended equations demonstrate that core electromechanical conversion maintains an intrinsically proportional relationship throughout dynamic non-steady states.
Translating these microstructural constitutive equations into a workable engineering model required adapting internal tensor variables to macroscopic observable quantities for a multilayer stack actuator possessing one clamped end and one mechanically free boundary. The researchers related internal electric displacement, strain, and electric field strength directly to total measurable charge, free-end axial displacement, and applied excitation voltage. In dynamic operation with time-varying excitation potentials, inertial effects generate internal mechanical stresses that oscillate synchronously with driving voltage, while the free end experiences zero boundary stress. To capture dynamic energy dissipation accompanying rapid polarization switching, the researchers integrated the Debye relaxation model, which mathematically unifies linear rate-dependent losses, including dielectric dissipation, mechanical damping, and piezoelectric conversion loss, into an equivalent electrical transfer function. Furthermore, to eliminate an unmeasurable time derivative of dielectric polarization charge from the governing equations, an analytical convolution integral based on unit ramp voltage excitation was introduced, yielding an analogous, closed-form time-domain expression for dielectric charge accumulation.
By integrating the Debye-derived dielectric charge equation with a constant term accounting for rate-independent energy losses such as electrical leakage currents, the authors synthesized a compact, unified displacement self-sensing equation. This final formulation establishes a direct algebraic relationship where dynamic physical displacement depends exclusively on instantaneous total charge, applied terminal voltage, an exponential voltage-dependent saturation term, and an offset constant.
The experimental validation platform constructed to evaluate the proposed self-sensing formulation utilized an industrial packaged piezoelectric ceramic stack actuator measuring five by five by ten millimeters in outer dimensions. Manufactured with one hundred individual ceramic layers, the test actuator possessed a nominal static capacitance of zero point seven microfarads and an operational maximum displacement of ten micrometers at a rated driving potential of one hundred and twenty volts. The actuator was driven through a commercial high-voltage power amplifier possessing an electrical bandwidth of one kilohertz, while baseline true displacement was captured via an external noncontact capacitive sensor featuring a linear measurement span of two hundred micrometers and a five-kilohertz bandwidth. Driving waveform synthesis and synchronous three-channel data acquisition were coordinated through a National Instruments data acquisition board operating sixteen-bit analog-to-digital and digital-to-analog converters with sampling frequencies up to two hundred and fifty kilosamples per second.
The team implemented an upgraded Sawyer-Tower capacitive divider configuration, placing a thirty-three-microfarad reference capacitor in series with the actuator ground path. Furthermore, an integrated overvoltage suppression sub-circuit built upon a three-terminal adjustable regulator was wired between the primary drive loop and amplification electronics, shielding sensitive measurement hardware against severe voltage spikes caused by potential ceramic breakdown or dielectric aging.
To extract the five governing model parameters reliably without encountering numerical singularities, the authors designed a multitone excitation signal synthesized within a computerized control environment. The multi-frequency driving potential combined three distinct sinusoidal voltage harmonics at one hertz, ten hertz, and fifteen hertz, with respective peak-to-peak amplitudes of fourteen point four volts, nineteen point two volts, and thirty-eight point four volts, while ensuring composite voltage never exceeded seventy volts. This composite excitation thoroughly stimulated both quasi-static hysteresis loops and frequency-dependent damping dynamics across five hundred synchronized temporal data samples. The unknown parameters were identified using a nonlinear least-squares algorithm governed by a trust-region reflective optimization scheme, minimizing the sum of squared differences between measured capacitive sensor positions and model-predicted displacements. Correlation curves verified that while voltage-displacement and voltage-charge plots displayed pronounced, matching hysteretic loops, the correlation between total accumulated charge and displacement remained exceptionally linear, validating the fundamental premise of the extended constitutive model.
Experimental validation under increasing sinusoidal excitation profiles confirmed the exceptional tracking precision of the self-sensing model across its entire operational envelope. Across validation trials conducted at one hertz, five hertz, and fifteen hertz with peak-to-peak voltages progressively increasing in ten-volt increments up to seventy volts, root-mean-square displacement errors remained consistently below zero point one micrometers, recording values of zero point zero eight four micrometers, zero point zero nine eight micrometers, and zero point zero nine one micrometers. To evaluate accuracy without distortion from near-zero ill-conditioning, the team reported threshold average relative errors excluding displacements under five percent of peak-to-peak stroke, measuring mean relative errors of four point three percent at one hertz, five point six percent at five hertz, and five point two percent at fifteen hertz. Practical implementation remains bounded by quasi-static charge drift below one hertz and structural resonances above fifteen hertz, establishing clear paths for future research into refined energy-loss modeling and data-driven parameter optimization.
Subject of Research: Displacement self-sensing in piezoelectric stack actuators using extended constitutive equations
Article Title: Displacement self-sensing of piezoelectric stack actuators based on extended piezoelectric equations
Article References: Zhang, D., Chen, Y., Zhu, H., & Zheng, H. (2026). Displacement self-sensing of piezoelectric stack actuators based on extended piezoelectric equations. Discover Sensors, 2(1), Article 27. https://doi.org/10.1007/s44397-026-00059-9
Image Credits: AI Generated
DOI: 10.1007/s44397-026-00059-9
Keywords: piezoelectric stack actuators, displacement self-sensing, constitutive equations, electromechanical coupling, hysteresis compensation, charge measurement, Debye relaxation model, Sawyer-Tower circuit, capacitive displacement sensor, parameter identification, nonlinear least squares, trust region reflective algorithm, threshold average relative error, vibration suppression, nanoscale positioning, dielectric polarization, smart materials, micro-nano manipulation
News Source: Denise Maddox. (November 13, 2026). Extended Piezoelectric Model Achieves High Precision Displacement Tracking Without External Sensors. Scienmag.



