Abstract
Computational modeling and simulation are commonly used during the development of cardiovascular implants to predict peak strains and strain amplitudes and to estimate the associated durability and fatigue life of these devices. However, simulation validation has historically relied on comparison with surrogate quantities like force and displacement due to barriers to direct strain measurement-most notably, the small spatial scale of these devices. We demonstrate the use of microscale two-dimensional digital image correlation (2D-DIC) to directly characterize full-field surface strains on a nitinol medical device coupon under emulated physiological and hyperphysiological loading. Experiments are performed using a digital optical microscope and a custom, temperature-controlled load frame. Following applicable recommendations from the International DIC Society, hardware and environmental heating studies, noise floor analyses, and in- and out-of-plane rigid body translation studies are first performed to characterize the microscale DIC setup. Uniaxial tension experiments are also performed using a polymeric test specimen to characterize the strain accuracy of the approach up to nominal stains of 5%. Sub-millimeter fields of view and sub-micron displacement accuracies (9nm mean error) are achieved, and systematic (mean) and random (standard deviation) errors in strain are each estimated to be approximately 1,000μϵ. The system is then demonstrated by acquiring measurements at the root of a 300μm-wide nitinol medical device strut undergoing fixed-free cantilever bending motion. Lüders-like transformation bands are observed originating from the tensile side of the strut that spread toward the neutral axis at an angle of approximately 55°. Despite the inherent limitations of optical microscopy and 2D-DIC, simple and relatively economical setups like that demonstrated herein could provide a practical and accessible solution for characterizing cardiovascular implant micromechanics, validating computational model strain predictions, and guiding the development of next-generation material models for simulating superelastic nitinol.