Couplant-induced stress evaluation errors in ultrasonic measurement: multivariable quantification and differential wedge mitigation

Longitudinal critically refracted (LCR) waves can offer efficient near-surface stress assessment, but suffering from couplant-induced inaccuracies, which undermine its reliability. This study quantitatively evaluates the effects of couplant type, dosage, and pressure on time-of-flight (TOF) stability, and the stress error propagation through synchronised TOF-digital image correlation (DIC) measurements. It shows that only 20 µm reduction in honey couplant thickness under a 2–3 kg preload shifts the TOF by 7.8 ns, which illustrates an equivalent 170 MPa error in stress estimation. Highviscosity couplants display nonlinear TOF-pressure hysteresis, while under cyclic loading oil and gel couplants producing TOF variability, fundamentally undermining reproducibility. To mitigate artefacts, a dual-receiver differential wedge is developed, suppressing common-mode TOF shifts by 90% for viscous couplants under dynamic loading. However, errors persist due to irreducible dose-pressure asymmetries, constraining practical accuracy to 50 MPa. Consequently, the strict couplant state invariance (±0.05 ml dosage; ±0.2 kg pressure) between calibration and measurement can hardly achievable. Our findings establish operational protocols for industrial applications tolerating ± 50 MPa uncertainty while advocating couplant-free technologies for precision-critical scenarios.