Answers are grounded in this paper and cite their sources.
High-throughput roll-to-roll (R2R) manufacturing of proton exchange membranes (PEMs) for fuel cells and electrolyzers demand rapid, non-destructive inline quality control of membrane thickness and proton conductance uniformity. Traditional contact-based techniques are slow, destructive, and constrained to offline testing; the contactless high-frequency eddy-current testing (ECT) offers a compelling alternative. This work outlines a continuous R2R monitoring architecture in which the measured sheet conductance is compared with its environmentally predicted value. This work demonstrates the metrological feasibility of deploying non-contact ECT sensor arrays for continuous web quality screening in industrial PEM manufacturing. This was accomplished by leveraging a laboratory-scale metrological proof-of-concept for inline R2R ECT integration. We repurpose the inductive front-end of a 13.56 MHz radio frequency (RF) bridge to characterize Nafion™ membranes under uncontrolled laboratory drift, implementing an ordinary-least-squares (OLS) calibration model that compensates for environmental baseline shifts. Within the calibrated humidity range, the thickness of membranes on held-out days was predicted with an RMSE of 0.48 mil, whereas in an independent experiment at a humidity outside that range the error rose to 1.49 mil, showing that measurements must be made under controlled humidity. Furthermore, we establish an inline proton conductance calibration framework. By computing theoretical baseline sheet conductance ( G s , i = σ ( RH i , T i ) · x i ) using the Springer et al. constitutive σ Springer ( RH i , T i ( K ) ) model and regressing the measured voltage response of the PEMs against G s , i , we extract an origin-constrained, Springer-referenced coil responsivity of K coil = 2158.2 ± 46.4 V · S − 1 for unstacked solid membranes. Interstitial air gaps in stacked configurations attenuate electromagnetic coupling, causing a 27 % responsivity drop.
No takes yet. Share an insight, caveat, or question.
Odusanya et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: