Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2611.03485

Identifying the Shear Piezoelectric Coefficient using Physics-Informed Neural Networks under Exact Boundary and Initial Constraints

Suhas Suresh Bharadwaj
View original source ↗Read source document ↗Export metadata ↓

Abstract

The shear piezoelectric coefficient $e_{15}$ governs the shear-mode electromechanical coupling in piezoelectric micromachined ultrasonic transducers (PMUTs) and surface acoustic wave (SAW) devices, but measuring it typically requires destructive resonance testing on dedicated specimens. On simulated data, $e_{15}$ can instead be recovered non-destructively from sparse, noisy subsurface wavefield observations using a constrained physics-informed neural network (PINN). By building exact spatial distance functions and temporal multipliers directly into the network output, the clamped, grounded boundaries and resting initial state hold automatically, leaving the loss with only the interior equations of coupled electro-elastodynamics and the data misfit. Non-dimensional scaling brings all governing coefficients within one decade of O(1), and the setup removes the boundary and initial-condition penalty weights that soft-constraint methods must tune. In a six-way comparison (plane-strain PZT-5H, manufactured wavefield, 30 dB noise), a soft-penalty baseline with adaptive weights converges to an erroneous $e_{15}$ in four of the five runs (51.7% mean error), and fixed penalties reach an error of 0.85% only after a second-order optimization stage. The constrained formulations reach mean $e_{15}$ errors between 0.15% and 0.51% with held-out displacement errors below 1%. Within these variants, the spatial multiplier gives rapid first-order convergence, and the quadratic temporal multiplier lowers the early-time electrostatic residual by a factor of 6.8 versus a linear one. On finite-element data along one subsurface line, the full formulation recovers $e_{15}$ with 1.19% mean error at 30 dB and keeps it between 0.65% and 2.71% across a 40 to 0 dB noise sweep. Finally, training on displacements alone still recovers $e_{15}$ with mean errors of 0.45% on finite-element and 1.64% on manufactured data.

Explore related subjects

physics.comp-phcond-mat.mtrl-scics.LGphysics.app-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Suhas Suresh Bharadwaj. 2026-11-02. Identifying the Shear Piezoelectric Coefficient using Physics-Informed Neural Networks under Exact Boundary and Initial Constraints. https://arxiv.org/abs/2611.03485

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Advanced computational methods for simulating the classical hydrogen atom in stochastic electrodynamics

Stochastic electrodynamics (SED) is a theory of nature seeking to derive key aspects of quantum mechanics and quantum electrodynamics from a purely classical physics perspective. While SED has been successful in many areas, it has had mixed success when addressing the classical hydrogen atom bathed in classical electromagnetic zero-point radiation. Because analytical approaches are largely intractable, advanced numerical simulation methods are required. To address this, we develop a pseudo-wave formulation that integrates the solid angle in k-space utilizing symmetry, resulting in a significantly reduced memory footprint and lower computational overhead for generating the zero-point radiation spectrum. This optimization is paired with a logarithmic frequency grid, a sliding window tracking the instantaneous Keplerian frequency, and a Tukey smoothing function to eliminate discrete force discontinuities. Integrated via an adaptive RK4 algorithm with step-doubling across highly dense grids of up to 100-million bins, these techniques demonstrate the asymptotic convergence of ensemble stability. We show that increasing grid resolution in the high-frequency domain systematically mitigates artificial electron decay, yielding a simulated probability density function that closely aligns with the quantum mechanical ground state P(r). Crucially, the simulations exhibit stable, low-eccentricity orbits without any observed atomic ionization events.

physics.comp-ph↗
arXiv2026

Quantum simulation of dissipative acoustic and elastic waves by Schrödingerisation: absorbing boundaries and explicit polynomial-size circuits

Explicit quantum circuits for the acoustic and elastic wave equations, obtained by Hamiltonian simulation of their first-order forms, conserve energy and therefore describe closed domains: an outgoing wave reflects off the edge of the computational region or, with periodic boundaries, wraps around and re-enters from the opposite side. We open the domain with the simplest physical absorber, a graded sponge layer in which every wave field is damped at a rate that grows towards the outer edge, and implement the resulting dissipative dynamics on a quantum simulator by Schrödingerisation, which turns a non-unitary linear evolution into a family of unitary ones by means of a single auxiliary variable. We derive the whole chain from the partial differential equation to the circuit, in one, two and three dimensions. An independent finite-difference solver confirms that damping all wave fields absorbs far better than the common velocity-only damping mask. For the sponge the quantum analysis is sharp: recovery is exact on the whole non- negative half of the auxiliary axis at every time; that axis can be truncated periodically at a length set by the largest damping rate and the simulated time; and the success probability is a fixed constant, about 0.4, times the fraction of the energy not yet absorbed, with only one or two auxiliary qubits measured. On the circuit side, the damping step, which generic synthesis compiles into a number of controlled-NOT gates exponential in the number of qubits, factorises exactly into polynomially many multi-controlled phase gates. Gate-level simulations on up to 25 qubits agree with the classical solutions of the damped wave equations.

physics.comp-ph↗
arXiv2026

Maximal correlation and optimal transformations of probability distributions in tensor-train format: a variational algorithm

The correlation between two subsystems $A$ and $B$ of a many-body system is usually measured through a pair of observables chosen in advance. The Hirschfeld-Gebelein-Rényi maximal correlation removes the choice: it is the largest correlation coefficient between any function of the configuration of $A$ and any function of the configuration of $B$, and the two functions that reach it, the optimal transformations, are the observables that carry the correlation. Both follow from a singular value decomposition (SVD) of the joint distribution normalized by its marginals, a matrix whose dimension is the number of configurations. We show that this SVD can be computed variationally for a classical distribution of discrete variables, such as the Boltzmann distribution of a spin model, given in the tensor-train (matrix-product) format, without ever forming the matrix. The optimal transformations are sought as tensor trains; each step updates a single core in closed form, and the inverse of a small matrix contracted from the distribution and the remaining cores performs the division by the marginals core by core. Every iterate is a feasible pair, so the objective is a lower bound on the maximal correlation that increases monotonically, and the cost grows linearly with the number of variables. The algorithm reproduces the exact SVD on small systems, recovers the known nonlinear optimal transformations of synthetic distributions of up to ten angular variables ($10^{15}$ configurations), and, for the bipartite antiferromagnetic model with $2^{32}$ configurations, returns an optimal transformation that is a function of the group magnetization, linear above the transition and sign-like below it: the staggered order parameter is read off from the distribution instead of being postulated.

physics.comp-ph↗