they have the ability for coordination motion along the chain backbone14. Under conditions where the test rate is fast relative to this relaxation time the molecules do not have time to displace during loading. Under conditions where the test rate is very slow the material can relax to the loading resulting in sigmoidal curves. The relaxation modulus is variable and depends on both the strain rate and temperature, which in this case is 10°C. In terms of rain erosion behaviour this demonstrates the importance of the timescale of the stress applied as it results significantly differing responses. This also raises the question if the coatings themselves undergo any stress relaxation over their lifetimes due to residual stress from cure or flexure of the blades. Following Zeltmann's method the predictions of elastic modulus are evaluated as the secant modulus at 2.5% strain from the stress-stain values generated from the relaxation function using: [3] where s, e' and t represent stress, strain rate and a time variable used for integration, respectively. Using this procedure, the elastic modulus at any strain rate can be calculated. The relaxation modulus was finally used to yield a linear relationship predicting the actual Young's modulus of the PU LEP over a range of strain rates (Figure 7). Owing to the manner in which the data are calculated, the relationship breaks down at the extremes as there may be other mechanisms that occur outside of our testing ranges and so are not entirely reliable. Young's moduli obtained using conventional quasi-static test methods for selected commercial PUs and it is clear that the Young's modulus varies significantly from quasi-static test conditions (0.278 GPa at 1 s-1) to more representative high strain rate impacts (106 GPa at 108 s-1). These differences could have implications on the lifetime prediction of coating systems such as the calculation of the water hammer pressure in the Springer damage model15. Increases in modulus result in increased water hammer pressures that may exceed the yield strength of the material. Alternatively this may cause mismatches between layers that alter the ratio of wave reflected and transmitted thought the multilayer system. However, it must be noted that this transformation assumes a linear material and relies upon the master curve data which can also w w w. s a m p e . o r g May June 2021 Journal.indd 23 Figure 7. Predicted Young's modulus of the PU LEP as a function of strain rate. be inaccurate due to the assumption of TTS such as missing certain transition outside the temperature range tested. This work requires validation using high strain rate methods such as Split-Hopkinson pressure bar which can operate at strain rates of up to 105 s-1. Advanced Composite Training for Engineers Basic to advanced design & analysis for structures & repairs Go to www.abaris.com for a full list of courses! * Composite Training * Onsite Training * Consultation We can come to you! +1.775.827.6568 admin@abaris.com Advancing Composite Technology Since 1983 M AY/J U N E 2 0 2 1 | SAMPE JOURNAL | 23 4/8/2021 12:32:28 PMhttp://www.abaris.com https://www.sampe.org/