In vibration-based high cycle fatigue testing, a base-excited plate is driven at a high frequency resonant mode until failure. In one vibration-based method involving a cantilevered square plate, a mode often referred to as the “two-stripe” mode is sometimes used because it exists at high frequencies and produces large uniaxial bending stresses along the free edge that are suitable for fatigue testing. The purpose of this work is to precisely investigate how the dimensions of a more generally rectangular plate influence performance when driven at the two-stripe mode. Included are the results of many thousands of modal analysis simulations. From these simulations, general trends with respect to resonant frequencies, frequency isolation, and stress fields in the plate are examined. Results of select geometries were then experimentally validated using a 1000 lb shaker. It is generally shown that, compared to square plates, rectangular plates with 1.37 length-to-width ratio exhibit more favorable stress distributions and frequency isolation. Recommendations are also given for how to quickly select preferable plate dimensions when planning a test based around the operating frequencies of the test setup.

A multiple-insert carrier plate assembly has been developed to increase the throughput of vibration-based fatigue testing. Typically, in vibration-based fatigue testing, a cantilevered rectangular plate is excited in a high-frequency resonant mode until failure. In this manner, an S-N curve for bending fatigue can be constructed. The concept behind vibration-based fatigue testing has been improved by employing a U-shaped carrier plate and fastening multiple smaller specimens into the channel of the carrier plate, such that fatigue tests can be conducted in tandem. However, because of the presence of a variable stress field over the surface of the assembly, the strain in the specimens is unequal. To account for this unequal strain distribution, a framework is provided for estimating how damage accumulates within each specimen. Using this framework, the efficiency gained by adopting a multi-insert approach is evaluated. By optimizing the strategy with which specimens are replaced as they fail, a throughput 5.49 times greater than sequential testing is shown to be feasible.

No abstract

Digital Image Correlation (DIC) is a non-contacting, camera-based technique that calculates full-field displacements and strains by comparing digital images taken before and after an object is deformed. During a vibration-based fatigue test, DIC has an advantage over strain gages in that it is non-contacting and does not accumulate damage during the test. In this work, DIC was implemented to build strain-velocity calibration curves as an alternative to strain gages. First, a curve fit was applied to DIC displacements and strains along the free edge of the plate using an approximate solution for the mode shape of a cantilevered plate. In total, the curve fits were applied to three sets of DIC data: (i) the raw strains calculated with DIC; (ii) the in-plane U-displacements from which the raw DIC strains were computed; and (iii) the out-of-plane W-displacements observed in the direction of motion. Second, classical plate theory was used to calculate strains by taking derivatives of each of the applied curve fits. Third, the peak strains from each curve fit were used to build the strain-velocity calibration curves. Further, a Monte Carlo Method uncertainty analysis was performed to estimate the uncertainty of the curve fitted DIC and strain gage measurements. Of the three curve-fits, the DIC strains derived from the out-of-plane displacements provided the most precise measurements relative to a strain gage at all excitation levels used to build the calibration curves.

No abstract