• Johansen, Nicolai Frost-Jensen

Abstract

In this thesis, two issues relating to conventional rain erosion testing are addressed: Firstly, erosion performance of a coating system has traditionally been expressed in hours to failure. Secondly, in an R&amp;D A/S whirling arm rain erosion tester (RET) droplet impacts are distributed randomly; thus, information from the individual impact is lost. <br/><br/>The first problem with the time-based performance metric is that these results are often not reproducible on other testing set-ups. The second problem is that repeated droplet impacts at a single point with speeds over 100m/s are almost impossible and impractical for erosion testing. This type of high strain rate impact fatigue cannot be performed on traditional cyclic fatigue testers. <br/><br/>To solve the first problem of making RET data reproducible standards and best practices such as the ASTM G73-10 and DNVGL-RP-0171 exist. These are used to evaluate results from the new R&amp;D A/S style RET, however, neither provide a full framework for rationalising erosion performance.<br/>The second problem of isolating the effect of a single impact has previously been attempted by water jet based testers. These are capable of impacting a single point with a jet or mist of water, however, all of these testers are still only approximations of actual droplets.<br/>In this thesis, the first problem of making the RET data more reproducible, was solved by combining methods from both standards with a statistical approach to the data analysis. To solve the second problem of repeated impacts, the water droplet impact was substituted by an impacting polymer ball, fired by the newly developed Single Point Impact Fatigue Tester (SPIFT). By firing polymer balls at the target coating a high strain rate impact can be generated. In traditional cyclic<br/>fatigue this would result in unnatural heating of the sample, which is avoided in the SPIFT by a combination of impact rate control and forced air cooling. <br/><br/>By using both the new SPIFT and the RET, three different coating systems were tested at different droplet sizes and impact speeds. Afterwards, all samples were evaluated using the new methods to evaluate the data and construct SN Curves. The SPIFT was used to establish a link between internal heat generation, resulting from impact, and the dynamical mechanical properties of the material.<br/>The new evaluation methods allowed comparison of results from all the different tests and testers. We expect this method to significantly increase the usability of RET data while making the data more suitable for predicting actual erosion lifetime. The SPIFT provides a new tool for evaluating fatigue performance of new coating systems, as well as providing more insight into the nature of impact fatigue. Comparing RET and SPIFT, a tentative severity factor of F₀ = 0.25 between<br/>RET and SPIFT was found. It was found that impact heating could be correlated to the material property tan at high frequencies using time-temperature superposition DMA data.

Topics

• impedance spectroscopy
• polymer
• fatigue
• ultraviolet photoelectron spectroscopy