Abstract
The discrepancy between lab accelerated life tests (LALTs) and field exposures of organically coated aerospace alloys subjected to corrosion is a well-known problem. For example, some Mg-based chromate-free primers for AA2024-T351 perform well in field corrosion testing but perform poorly in LALTs such as the ASTM B-117. Conversely, some primer coatings on metals have been found to perform well from a corrosion standpoint in LALTs but poorly in the field. Currently, it is not well understood whether various differences in environmental severity factors (ESFs), such as chloride and UV, cause such discrepancies. A lack of understanding of how ESFs affect the coating/substrate system is a consequence of previous studies’ reliance on low fidelity interrogation methods, such as visual inspection and coating gloss measurements, which do not adequately capture the entire range of interactions between ESFs, the coating and the substrate. Subsequent LALTs are developed by trial and error. This research seeks to take the first steps to understand the influence of selected specific ESFs on metal corrosion and polymer degradation in scribe creep.
To begin to understand the influence of environmental severity factors on underpaint corrosion, we compare the similarities and differences in corrosion and scribe creep results from standard LALTs, field sites and lab full immersion tests (FIT). Ultra-high molecular weight epoxy resin (Poly(Bisphenol A-co-epichlorohydrin) glycidyl end-capped (C18H22O3)n•C22H26O4; CAS No. 25036-25-3,trade name Eponol) coated AISI 1018 steel samples (UNS# G10180; 0.15% C, 0.7% Mn, Fe; wt. %), with controlled scribes to expose the bare metal, were used in all tasks. In this initial investigation, comparisons between standard LALTs, FITs and field exposures were made using a suite of high-level surveillance methods: electrochemical impedance spectroscopy (EIS), Fourier Transform Infrared Spectroscopy (FTIR), X-Ray diffraction (XRD), and Raman spectroscopy. These methods maximize the ability to detect corrosive degradation both to the bare substrate at the scribe and to the substrate under the coating, as well as improve detection of degradation of the coating itself. By elucidating corrosive changes at a high level that could be missed with low fidelity surveillance techniques, these methods also improve our ability to make comparisons between LALT, FIT and field tests, and demonstrated a methodological improvement upon previous studies. Tests were conducted according to the LALT standards and were interrogated at predetermined time points (i.e., 0, 1, 3, 5, 10 and 15 days) during a total exposure time of 15 days. Briefly, results from this work demonstrate that there is a positive correlation between mass loss on bare 1018 steel samples and scribe creep length on coated steel for both lab- and field- exposed samples. Additionally, there is an inverse correlation between scribe creep length and the low frequency electrochemical impedance of the coating for coated steel near the scribe for lab and field samples. Concerning field vs. LALTs, the results demonstrate that LALT’s without UV radiation do not correlate as well with field exposure in terms of polymer coating degradation. This points to the need for UV radiation to damage the polymer coating. Comparison of LALT results to full immersion tests (FITs) helped to clarify the mechanism of scribe creep in these lab and field exposures.
LALT, FIT and field comparisons demonstrate that while the effects of ESFs on the corrosion of bare steel are straightforward, their effects on the scribe creep of coated steel are more complex. Therefore, systematic variation of ESFs in non-standard LALTs and statistical analysis of the resulting data were conducted to better understand the effects of ESFs on scribe creep of ultra-high molecular weight epoxy resin coated AISI 1018 steel. The role of ESFs in scribe line corrosion and corrosion of “intact” organic coatings was studied using 3D optical microscopy, EIS, XRD and Raman microscopy while parallel studies at the University of Southern Mississippi examined coating degradation with FTIR. Samples were exposed to various levels of ESFs in a set non-standard LALTs, in a fractional factorial design. The effect of UV light, temperature, relative humidity, chloride, and wet/dry cycling on underpaint corrosion and degradation of electrical properties of organic coatings were investigated using the high-level surveillance methods noted previously. Differences in corrosion morphology, corrosion products formed, and rates of scribe creep were examined as a function of ESFs with the objective to produce a parametric model that could relate scribe creep length to ESFs. It was found that temperature and cycling had the largest effect on scribe creep. Temperature and cycling in combination interacted to produce an effect on scribe creep that was greater than the additive effects of each.
Finally, the data from the non-standard LALTs was used to develop an empirical model of scribe creep of Eponol coated 1018 steel. This is the first empirical model to relate scribe creep to ESFs for an organically coated steel system known to the author, similar to those constructed in the past for bare steel. The model assesses the relative strength of individual and combined ESFs on scribe creep of coated steel. The model found that cycling had the biggest effect on scribe creep. The model also indicated the possible interaction between temperature and cycling. Field behavior was subsequently predicted using the model. By highlighting which ESFs most greatly impact scribe creep, the model provides guidelines for creating future LALTs that better mimic important aspects of the field environment. The model also serves as a template for creating scribe creep models for other commonly used coating schemes. Finally, the model highlights ESF interactions that should be further investigated in future research.
Overall, this body of work adds to the understanding of the effects of ESFs on scribe creep and the relationship between anodic wedging and cathodic delamination on scribe creep. This work also demonstrates the applicability of fractional factorial design for quickly and efficiently investigating multiple ESFs in relatively few experiments. Insights gained can be used in the construction of future LALTs.
