GARRF Foundation Logo

Gopalkrishna Advanced Rural Research Foundation India

Dr. APJ Abdul Kalam Research Zero Funding Initiative | Virtual AISIA Lab

Our First AI PhD Student Agent for EMI Technique

AGENT "ARJUN"

Dedicated to 3D Annamdas and Soh Formulation (Nanyang Technological University - NTU Singapore) | Universal Continuum Mechanics for 1D, 2D & 3D Solids

SHM Simulation Parameters

PZT-5H Coupled (3D Tensor)
PZT Sensor Location (x / L) 0.40 L
Damage Defect Location (x / L) 0.55 L
Left End (0.05L) Midspan (0.50L) Right End (0.95L)
Damage Severity Progression 50% Moderate
0% (Baseline Healthy) 50% 100% (Critical)
AI PhD Tutor: ARJUN ONLINE

Need guidance on the 3D analytical impedance tensor, boundary reductions, or why 1D models fail at ultrasonic frequencies? Ask ARJUN directly.

RMSD Metric 14.82 % Root Mean Square Deviation
MAPD Metric 9.31 % Mean Absolute % Deviation
Peak Δf Shift -3.40 kHz Stiffness Softening Shift
Correlation (CC)
0.912 Moderate
Waveform Cross-Correlation

3D Waterfall Admittance Signature (Conductance G vs Frequency vs Damage)

→ X-Axis: Frequency Sweep (20 kHz to 180 kHz)
↑ Y-Axis: Conductance G in milli-Siemens (mS)
↙ Z-Axis: Progressive Damage Steps (Baseline → 100%)
Max G
Min G
Structural Physical Twin Schematic Interactive Defect Halo
Golden PZT Patch Damage Notch Acoustic Wavefront
Conductance Spectrum G(ω) [mS] Baseline vs Damaged
— Baseline (Healthy) — Current Degraded State

ARJUN — AI PhD Student Agent

Specialized in 3D EMI & SHM

Dedicated to 3D Annamdas & Soh Formulation (NTU Singapore)

Try asking:

Namaste! I am ARJUN, AI PhD Student Agent for EMI

I have synthesized the complete doctoral research and mathematical formulations of Dr. Annamdas Venu Gopal Madhav and Prof. Soh Chee Kiong from NTU Singapore.

Whether you are an undergraduate, doctoral scholar, or practicing structural integrity engineer, ask me anything regarding 3D electromechanical admittance Y(ω), dynamic mechanical impedance tensors, PZT-structure interaction, boundary traction reductions, or damage metrics (RMSD, MAPD, CC).

3D Electromechanical Impedance (EMI) Formulation Suite

Authored & Formulated by Dr. Annamdas Venu Gopal Madhav | Virtual AISIA Lab

Physical Breakdown: The True Advantages of Dr. Annamdas Venu's 3D Formulation

Classical 1D models (Liang, Sun, Rogers) and early 2D models assume low-frequency global bending or 1D axial displacement. However, high-frequency structural health monitoring operates under completely different continuum wave regimes.

1. High-Frequency Ultrasonic Breakdown

At ultrasonic interrogation frequencies (30 kHz to 400 kHz), the acoustic wavelength collapses into millimeters:

Wavelength λ = Wave Speed / Frequency ≈ 5000 m/s / 100 kHz = 50 mm (5 cm)

Because the wavelength is comparable to or smaller than the thickness and width of the beam, plane-section assumptions fail. High-frequency Rayleigh surface waves, thickness-shear modes, and Lamb waves are excited, which 1D and 2D models cannot predict.

2. Elimination of Artificial 1D Constraints

A real square PZT patch expands along both planar directions through piezoelectric strain coefficients d31 and d32.

  • 1D Models: Artificially force d32 = 0, causing 15% to 30% resonance frequency errors.
  • 3D Formulation: Fully couples planar strains (S11, S22), shear strain (S12), and Poisson's ratio lateral breathing (νp, νs).

3. Internal & Depth-Specific Defect Sensing

1D and 2D models compress geometric thickness into neutral axes. Consequently, they cannot distinguish between:

  • A surface scratch vs a deep internal void.
  • Top flange corrosion vs bottom rebar debonding in concrete.

The 3D impedance tensor captures stress distributions across the z-axis, allowing detection of sub-surface delamination and through-thickness cracks.

4. Adhesive Bonding Layer Mechanics

Early models assumed "perfect bonding" with zero adhesive thickness. Under high-frequency cyclic strain, severe normal peel stresses (σzz) and shear-lag develop at the transducer edges.

Dr. Annamdas Venu's 3D formulation models the 3D elastodynamic boundary layer of the epoxy adhesive, preventing sensor debonding misdiagnoses.

Comprehensive Comparison Matrix: 1D vs 2D vs Annamdas & Soh 3D Model

Feature / Capability Classical 1D Beam Models 2D Plate Models Annamdas & Soh 3D Master Model
Cross-Axis Coupling (d31 vs d32) Ignored (Forced d32 = 0) Partially coupled Fully coupled with Poisson's ratio
Thickness & Shear Modes Completely missed Missed (Plane stress assumption) Accurately captured
High-Frequency Validity (>100 kHz) Breaks down (Severe error) Limited frequency range High accuracy across broad sweeps
Internal / Sub-Surface Sensing No (Averaged over cross-section) No (Mid-plane only) Yes (Depth-sensitive 3D tensor)
Adhesive Edge Peel Stress (σzz) Not modeled Limited approximation Fully modeled in elastodynamics
Universality (1D, 2D, 3D Solids) Only for thin beams Only for thin plates Universal continuum mechanics model
Virtual AISIA Lab | Gopalkrishna Advanced Rural Research Foundation, India