Abstract… · Web viewKeywords: Coatings, Hydrolysis, Kelvin Probe, Corrosion Monitoring, Non-Destructive Evaluation, Polymer Degradation, Polyurethane Abstract Degradation of coatings

NON-DESTRUCTIVE EVALUATION OF URETHANE-EPOXY COATING SYSTEMS USING A PORTABLE SCANNING KELVIN PROBE

Douglas C. Hansen, University of Dayton Research Institute

John Griffiths, Bio-Logic Science Instruments Ltd

Robert Roberts, Bio-Logic USA

Maria Inman, Faraday Technology Inc.

Keywords: Coatings, Hydrolysis, Kelvin Probe, Corrosion Monitoring, Non-Destructive Evaluation, Polymer Degra-dation, Polyurethane

ABSTRACT

Degradation of coatings on aircraft, while not visible on the surface, can be the result of changes in physical and chemical composition, leading to premature coating failure. Standard lifetime predictions fail to estimate the func-tional lifetime of urethane coatings on aircraft due to the wide variety and combinations of environmental factors that trigger degradation. To better understand the role that environmental conditions play in coating performance, a non-destructive technique is necessary to quantitatively assess the degree of degradation/reversion. The scan-ning Kelvin probe has demonstrated sensitivity to changes in alkyl chains and terminal groups of polymers, deter -mined the interfacial diffusion of water through epoxy coatings, and detected corrosion under coatings. These ca-pabilities make the scanning Kelvin probe a compelling approach to detecting and characterizing the degradation behavior of multi-layer urethane and epoxy coating systems as well as corrosion occurring at the coating-metal in-terface. A portable SKP system is being developed for detecting and identifying the degradation of urethane-epoxy coating systems, which could be used in the field for rapid coating evaluation to enable implementation of condition-based maintenance. Early targeting of areas of an aircraft that require coating refurbishment will de-crease the need for complete regular overhauls of painted structures.

INTRODUCTION

High performance aircraft are subjected to harsh environmental factors during operation including impacts with water and other particles, as well as high temperature and humidity.1 Rain erosion coatings, made from polyurethane, are used to protect the metal surfaces and other underlying coatings from these damaging ele-ments by absorbing the force of impacts and providing a barrier against corrosive substances. Over time these coatings break down and may fail in flight by delamination, leading to unplanned downtime and expensive refin -ishing procedures.1 Coatings that are exposed to high temperature and humidity are particularly susceptible to degradation,1 leading to the assumption that the cause is hydrolysis. These types of decomposition lead to a breakdown of the polymer network structure and reduce the strength of the coating, increasing the likelihood of failure.1

1

Paper No. 2019-0000

2019 Department of De-fense – Allied Nations

Technical Corrosion Con-

Performance degradation of urethane and epoxy coatings on high performance aircraft due to chemical reversion, while not visible on the surface, can be the result of changes in physical and chemical composition, thereby lead-ing to premature coating failure. Standard lifetime predictions fail to estimate the functional lifetime of urethane coatings on operational aircraft due to the wide variety of environmental factors and combinations of factors that trigger degradation at different rates. To better understand the role that environmental conditions play in the over -all lifetime performance of these coatings, a non-destructive technique is necessary to quantitatively assess the degree of degradation/reversion that can occur. Currently no non-destructive method exists for accurately detect -ing when a particular aircraft coating will fail prior to degradation being visible. Therefore, there is a need for a reli-able non-destructive method to detect the changes in the surface coating that indicate when it is close to failure or has begun to fail.

The Scanning Kelvin Probe (SKP) has long been known to have the ability to detect the change in Volta potential of metal substrates underneath polymer coatings.2,3,4 More recently the Kelvin probe has been used to measure changes in the surface potential portion of the work function measurement. Hansen et al. 5 demonstrated the abil-ity to detect the changes in length and conformation of DNA molecules using an SKP system, I.R Peterson 6 de-tected work function differences on gold plated slides covered with a monolayer of phospholipid, Bastide et al. 7 observed changes in the work function of photoactive organic layers on semiconductor surfaces upon exposure to light, and Grundmeier et al.8 demonstrated that the SKP could detect the diffusion of water and ions through adhe-sive coatings and how they interact with the polymer metal interface. All of these show the sensitivity of the SKP to changes in surface material and indicate that the SKP should be able to detect chemical changes in a de -graded polyurethane coating as it undergoes hydrolysis. The scanning Kelvin probe (SKP) has demonstrated it -self to be sensitive enough to detect changes in the alkyl chain lengths of polymers and their terminal groups as well as determine the interfacial diffusion of water through epoxy coatings as a function of the chemical structure and functional groups present within the coating. It has also demonstrated its capability to detect corrosion under coatings at the coating-metal interface. These capabilities make the use of the scanning Kelvin probe technique a compelling approach to detecting and characterizing the degradation behavior of multi-layer urethane and epoxy coating systems as well as any corrosion occurring at the coating-metal substrate interface.

The Scanning Kelvin Probe

The scanning Kelvin Probe (SKP) is a non-destructive inspection technique that has been used to investigate the electrochemical properties of surfaces, coatings and polymers over the last 20 years.2,9 The SKP is a non-contact, non-destructive scanning probe instrument that measures the work function difference between conducting, coated, or semi-conducting materials and a vibrating capacitance probe. Through an applied swept backing volt -age, the work function difference is measured between the scanning probe reference tip and sample surface. The work function can be directly correlated to the condition of the surface being investigated.

A unique aspect of the SKP is its ability to make measurements in a humid or gaseous environment. This feature lends itself to a fieldable application, unlike more traditional electrochemical measurements, which require direct contact with a surface in an aqueous environment.

The SKP technique has been used to measure the difference in the work function of polymers adsorbed to solid substrates by using a backing potential to null the current between the sample substrate and a vibrating probe that is electrically connected to the substrate and positioned close to the sample surface. This difference in work function between the non-contact scanning Kelvin probe and the underlying substrate is defined as the contact potential difference (CPD).10 The utility of the SKP technique in the laboratory to characterize changes in

2



molecular conformation in adsorbed organic layers at the air/solution interface and on solid substrates has been demonstrated11,12 and changes in the alkyl chain lengths of polymers and their terminal groups have been determined using the SKP technique.13 Figure 1 shows a SKP instrument and close up of the Kelvin probe above a sample surface.

The principle of operation of the Kelvin Probe is illustrated in Figure 2 (left – right). Three energy level diagrams depict two metals having a constant separation and work functions and Fermi levels of (1, E1) and (2, E2), re-spectively. The first diagram in Figure 2 (left) shows the metals with no electrical contact and differing Fermi lev-els. When electrical contact is made (center diagram) the flow of charge allows the Fermi levels to equalize and gives rise to a surface charge. The potential difference, VC, is related to the difference in work function:

Figure 1Overall view of the scanning Kelvin probe system, and a closer view of the probe above a sample.

-eVC + 1 - 2 1

where e is the electron charge. The inclusion of an external backing potential, VB (right diagram) allows for a nulling of the surface charge at a unique point where VB = - VC. This point represents the work function difference between the two materials. When capacitance is a minimum, the sample potential, V B is varied automatically to ensure that the null point is always maintained and thus the potential at which nulling occurs equates to the work function difference between the probe and sample. This approach to measuring work function has been used for quite some time to measure the Volta potential difference or contact potential difference (CPD) between a probe tip and a surface of interest. The Volta potential difference is defined by IUPAC as the electrical potential differ-ence between one point in vacuum close to the surface of a metal and another point in vacuum close to the sur-face of another metal where both metals are uncharged when brought into electrical contact.14 The contact poten-tial difference is analogous to the Volta potential, defined as being the potential difference between a conductive scanning tip and a conductive sample in air.9 The ability of the SKP technique to resolve potential differences be-tween a conductive scanning probe tip and a conductive sample in air at ambient conditions and temperature is defined by the relationship:

φ = ψ + χ 2

where the inner Galvani potential (φ) is defined to be equal to the sum of the Volta potential (ψ) and the surface potential (χ). In this relationship, the Volta potential is defined as the work required to bring a charge from infinity to the near surface without being influenced by surface dipoles that are responsible for the surface potential on the sample or probe tip. Further examination of this relationship can be expressed as:

Φ = μe + eχ 3

where the work function (Φ) is the sum of the inner chemical potential of an electron (μ e) and the energy required

3



to move an electron across the interface of the metal (charge, e, times the surface potential, χ). Therefore, when considering two separate areas on a sample surface, Equations 2 and 3 can be combined:

-Φ1 + εχ1 = -Φ2 + εχ2 or ρΦ1 - Φ2 = ε(χ1 – χ2) = ΔΦ 4

showing that variations in the measured CPD represent the spatial differences in the work function of the material as a result of surface potential changes. It is clear then that Volta potential differences are the result of galvanic activity involving an electrolyte, whereas contact potential differences arise from differences in work function be-tween the probe and the sample substrate. The CPD of self-assembled monolayers (SAMs) of alkanethiols on gold has been shown to correlate with the difference in the dipole-charge distribution between the SAMs and the surface potential of the gold.13 Indeed, Hansen15 demonstrated that the SKP technique is capable of discriminating between layers of a nitrocellulose paint based upon changing dielectric strength. Since the dielectric strength of a polyurethane or epoxy coating is directly related to the amount of water content contained within the coating, and one of the possible depolymerization (or reversion) mechanisms involves hydrolysis, it should be possible by the SKP to detect, localize and quantify the amount of degradation occurring. The dielectric constant of a polymer is linearly related to the molecular polarizability and density of the dielectric material.16 In addition, the ability of the SKP to detect the corrosion/degradation of metal substrates and/or corrosion product underneath polymer coat-ings is well known,2,17 and therefore this technique is ideal for the corrosion detection of fillers and pigments within the coating systems.

Figure 2

Schematic illustrating the principle of operation of the scanning Kelvin probe system

EXPERIMENTAL PROCEDURE

Panel Coating Systems and Exposure Procedures

Early stage laboratory work demonstrated the ability of the bench-scale SKP system to differentiate between a polyurethane rain erosion coating and an epoxy primer coating system, and to differentiate between a degraded and non-degraded polyurethane coating, epoxy primer and polyurethane/epoxy primer stack up coating system.

4



This demonstration provided the basis for the development of the alpha-scale portable SKP system. This preliminary work was done in the laboratory using a bench-scale SKP system (Figure 1) on aluminum alloy panels (3″ x 6″) of 2024-T3, cleaned with Prekote® and coated with:

MIL-PRF-23377 Epoxy Primer- Deft 02-Y-40 0.6-0.9 mil MIL-PRF-85285 Rain Erosion Coating – Caapcoat B-274 11-13 mil

Separate sample substrates with only primer or rain erosion coating over primer were constructed (Figure 3, Left image), as well as a complete coating system stack up (Figure 3, Right image). The complete coating system stack up samples were sheared in half after the coatings had fully cured to allow for the determination of the CPD of the primer, CPD of the topcoat, and CPD of the primer-topcoat system stack up, respectively, before and after exposure. Shearing the coated sample in half provided 2 identical samples for each exposure test condition and control for subsequent analysis by SKP. During the SKP analysis, the separate halves were electrically connected with conductive tape.

Figure 3Photographs of Al-2024 panels coated with MIL-PRF-23377 Epoxy Primer and MIL-PRF-85285 Rain Erosion Coating.

The coated panels were exposed to elevated temperature (121°C) and high humidity (100% RH) conditions in an autoclave to induce degradation of the polyurethane - epoxy coating material. The coated panels were exposed for 4, 8, and 12-hour time intervals and the amount of degradation was determined by SKP analysis as compared to an identical non-exposed control sample.

RESULTS

The SKP technique was used to qualitatively and quantitatively determine the integrity of the coating and primer within a coating system stack up. CPD measurements were made using a scanning Kelvin probe modular component of the M470 scanning probe system (Bio-Logic USA, Knoxville, TN) following methods described previously.17,18 The M470 SKP system utilizes a vibrating, shielded 500 micron diameter Pt probe tip at a working distance of 80 microns from the sample surface to indirectly measure the contact potential difference (CPD) between conducting materials and a sample probe by use of a swept backing potential. In addition, this system has the capability to measure the surface topography of a sample, utilizing an optical surface profilometer with a 650 nm laser using laser triangulation.

5

Sample with Primer-Rain Erosion Stack-up before sectioning for exposure

Rain Erosion Coating

Primer



A detailed measurement of the sample surface geometry including surface roughness statistics is generated. The surface geometry profile is then used to program the surface scan of the sample using the scanning Kelvin probe, allowing the Kelvin probe to remain at a constant height above the sample during the entire scan. This technique eliminates any contribution that changes in the probe height will have on the CPD values measured during the SKP scan. A total of three Kelvin probe area scans were done on each zone of each coated 3″ x 6″ test panel. Optimization of the measurement parameters such as probe height, probe vibration amplitude and electrometer gain setting were determined for the SKP system using a non-exposed coated panel as a baseline control. Subsequent measurements were done on coated 3″ x 6″ panels that have been exposed to a high temperature (121oC) and high humidity (100% RH) environment. For each panel/exposure, a total of 9 separate measurements were done for statistical analysis of CPD values for each sample and between each sample measured.

Figure 4 shows an example of the results obtained, for a non-exposed panel. The data shows a clear difference in CPD between the primer, and the rain erosion coating. A quantitative difference of 0.207 eV was measured between the two polymer coatings.

Figure 4Qualitative and quantitative differences in CPD between a primer and a rain erosion coating (actual sample shown in

Figure 3 left). The panel was not exposed prior to the measurement.

Comparison of SKP Characterization of Exposed and Non-exposed Polyurethane Coating System Panels

4 Hour ExposureSample 398-A1B-001A/001B was used for the first exposure test. The two sides 001A and 001B were initially connected but were cut down the middle making two separate samples (these will be referred to as 001A and 001B). Sample 001A was exposed to 121°C and 100% RH for four hours. The pair was then reconnected with conductive tape and scanned in the SKP. Figure 5 shows preliminary results that demonstrate qualitative and quantitative differentiation between non-exposed and exposed polyurethane/epoxy primer stackups. There was a difference in CPD of 0.239 eV between the non-exposed (001B) and exposed (001A) coating. Furthermore, a dif -ference in CPD of ~ 0.150 eV was observed between the exposed coating that had adhesive residue from the tape that covered the coating during exposure and the non-taped surface. This observation may lead to applica-tions for determining not only the degradation of a coated surface, but also identification of any surface contami -nation.

6



Figure 5Preliminary scan: Qualitative and quantitative differences in CPD for a primer/rain erosion coating stack-up (actual

sample shown in Figure 3 right) that is pristine or has been exposed for 4 hours at 121oC and 100% RH.

In the preliminary scan, the occurrence of tape residue was not anticipated, so the scan did not cover much area of the degraded coating. Therefore, additional scans were run over a much broader area of the exposed coating stack-up, in the area of the white box in Figure 6.

Figure 6Scan area of 35 mm x 30 mm (white box, step size of 250 m) on a primer/rain erosion coating stack-up: 001B and

001A. The area to the right of the center line was exposed for 4 hours at 121oC and 100% RH (001A).

Figure 7 shows the qualitative (top) and quantitative (bottom) differences in CPD for a primer/rain erosion coating stack-up (actual sample scan area outlined in white box in Figure 6) that is pristine, has tape residue, or has been

7



exposed for 4 hours at 121oC and 100% RH. Similar to the data in Figure 5, there is a ~0.2 V change in the ex-posed surface (note: at this point we are not controlling the humidity of the air in contact with the sample).

Figure 8 shows optical micrographs of representative samples of the pristine and 4 hr exposed surfaces. Note, Figure 8 (top) was taken from the taped area but it was visually identical to the pristine surface. Figure 8 (bottom) is of the coating after 4 hours of exposure at 121oC and 100% RH; the circles are likely water marks from conden-sation in the autoclave. The darker spots in the exposed coating area (Figure 8 (bottom)) may be indicative of the beginning of coating reversion, but more data is needed. Metallic specks in the coating have also been observed to cause a drop in work function.

Figure 7Qualitative and quantitative differences in CPD for a primer/rain erosion coating stack-up (Figure 6) that is pristine,

has tape residue, or has been exposed for 4 hours at 121oC and 100% RH.

8



Figure 8Optical micrographs of selected areas of the coated panel shown in Figure 6. Top micrograph is representative of the tape residue area, and the appearance of the pristine surface is very similar. The bottom micrograph shows the coat-ing stack-up surface that has been exposed for 4 hours at 121oC and 100% RH. The red cross hairs in each image in-

dicate the location of the Raman IR measurements. The white scale bar is equivalent to 5 microns.

Figure 9 focuses on the change in CPD for the primer/rain erosion coating stack-up (Sample in Figure 6) including a potential defect on the right side of the image. At this potential defect, a large spike in work function was ob-served. It should be noted that this defect might have been present before the exposure.

9



Figure 9Line scan across the pristine surface, tape residue, and exposed surface, including a possible defect in the exposed

coating surface (far right). The line scan shows a sharp increase in the work function at the defect.

The preliminary data collection and analysis described above was done upon a representative line scan between the exposed and non-exposed samples. Subsequent analysis of the data was done on the delta of the average work function values within a scan area between the non-exposed and exposed 4 hour samples, as this provided a more accurate analysis. Mid-sample area scans are taken of all samples to avoid possible edge effects from the masking tape used to protect the cut edges during exposure.

8 Hour ExposureThe sectioning of the sample 398-A1B-002A and B was performed in the same manner as the 4 hour exposure sample, 001A and B. This time 002A was left unexposed and 002B was the exposed side. In Figure 10, scuff marks with significant dirt inclusions from the shearing process used to section the sample can be seen especially on the unexposed sides most notably on the side along the cut as well as in the middle. This contamination shows up in the SKP scan as two different signals with an overall increase in work function on the lower half of the scan especially on the unexposed side, as well as some lower work function points which are caused by metallic flakes in the coating. The contamination was strongly attached to the surface and could not be removed by brushing or air blasting. However, this did not prevent the scanning of the sample with the Kelvin probe and resulting in mean-ingful data. In the Kelvin scan (Figure 11) there was an observed increase in work function of 0.074 eV from 002A (average of -0.911 eV) to 002B (average of -0.837 eV).

10



Figure 10Sample (top) before and (bottom) after 8 hour exposure.

Comparison of 4 and 8 Hour ExposuresThe first goal was to determine if the change in work function between the two exposed samples 001A (4 hour ex -posure) and 002B (8 hour exposure) was more than just an artifact of the probe height or scanning at different times. Several 45 mm by 12 mm Kelvin area scans were performed with these two samples together (Figure 12), with 001A on the left and 002B on the right. The average work function for the 4 hour exposure was -0.634 eV (Figure 12a) and with the tape line (Figure 12b) -0.507 eV. The 8 hour exposure side had a tape work function of -0.611 eV (Figure 12c) and the exposed surface was -0.732 eV (Figure 12d). This shows that there is a 0.1 eV dif-ference between the two exposures. Interestingly this holds true for their tape lines as well.

11



Figure 11Kelvin scan of 8-hour exposure.

12



Figure 12Kelvin scan of degraded sides 001B (left) and 002A (right).

The two unexposed sides were tested to determine if they exhibited similar values. The non-exposed side of the 4-hour exposure sample, 001B, had an average work function of -0.969 eV and the non-exposed side of the 8-hour exposure sample, 002A, had an average work function of -0.939 eV. It is worth noting here that the work function of 001B decreased by ~0.120 eV from its initial scan a month before, whereas 002A changed by only 0.028 eV. Primer Comparisons for 4 and 8-Hour ExposuresTo further understand what was causing the shift in work function, identical aluminum panels with only primer ap-plied were exposed for either 4 or 8 hours at 121 C and 100% RH. This investigation showed that the primer lay -ers were not applied evenly. Looking at the unexposed primer sides (Figure 13, top) the primer applied on the left side is noticeably thinner with areas of visibly exposed substrate than the right sample. The Kelvin area scan ( Fig-ure 13, bottom) shows that the left panel exhibits much more variation than the right. The average work function of the left side is -0.987 eV (no red spot), -0.962 eV (with red spot) and the right side average is -1.01 eV.

Figure 13Unexposed primer image (top) and Kelvin scan (bottom).

The 4 hour primer exposure (Figure 14) had a change in work function of 0.123 eV with an average work function for the unexposed surface (Figure 14, top left) of -1.06 eV and 0.937 eV for the exposed surface (Figure 14, top

right).

13



Figure 14Kelvin scan of an unexposed (left) and 4-hour exposed (right) primer.

The 8 hour exposure (Figure 15) had a more complicated pattern of degradation with at least two distinct regions on the exposed side (left). The average work function of the blue section was -0.985 eV and the green/red section was higher at -0.808 eV. Both of the sections combined had an average work function of -0.897 eV. The unex-posed section (right) had an average work function of -1.06 eV. From the two average surfaces, there is a differ-ence of 0.163 eV in work function. However, it could be as small as 0.075 eV or as large as 0.252 eV depending on which section is considered.

14



Figure 15Kelvin scan of an 8 hour exposed (left) and unexposed (right) primer.

Curing Samples

When the unexposed sides of the 4 and 8 hour samples were scanned together it was noticed that the work func -tion of the 4 hour sample, 001B, changed significantly more from its original scan than the unexposed 8 hour sam-ple, 002A. A possible explanation of this was that the 4 hour sample had not fully cured at the time the first Kelvin area scan was made, which was taken only 2 weeks after the coating was applied. The 8 hour sample was not exposed and tested until over a month and a half later. If this is true then there could be many different effects go-ing on; the REC could undergo a different reversion (or hydrolysis) reaction or the same at an accelerated pace. It could also use some of the energy from the autoclave to cure instead of degrading, there could be more water ab-sorption because not all the bonds are completely formed and this would make the difference in work function measured between the two unexposed 4 hour sample sections 001A and 001B inaccurate.

An additional test being run is examining any effect(s) that curing of the polyurethane coating has on the work function and Raman values of the sample surfaces. To test this, newly coated samples were scanned with the SKP and Raman spectra were taken every 2 weeks following the application of the coating to see if there are any observable changes over time. The first set of data has been taken for both Raman and SKP with the average work function being -1.14 V. This is significantly lower than the current unexposed samples (~ -0.9 V). The current hypothesis is that the work function will increase over time and reach the value of the samples that were coated 5 months ago.

15



12 Hour Exposure

Sample 004B (Figure 16b,c) was exposed under the same conditions as the previous samples (121°C and 100% RH) for 12 hours. The surface potentials (Figure 16, bottom) for the different areas were: a) = -0.649 V, b) = -0.406 V, c) = -0.568 V, and d) = -0.501 V. The 12 hour exposure results show that there is a relatively small change of 0.081 V between the unexposed surface (Figure 16a) and the exposed surface (Figure 16b,c). This is comparable with the 8 hour exposure and much lower than the 4 hour exposure (Figure 17); the error bars repre-sent the standard error of the mean, and therefore the average value for the 4 hour sample is statistically different from the 8 and 12 hour samples, whereas there is no statistical difference between the 8 and the 12 hour sam-ples.

Figure 16Sample 004A/B image (top) and Kelvin scan (bottom); a) unexposed b) tape line c) 12 hr exposed and d) crater.

16



Figure 17Work function difference between exposed and unexposed samples. The error bars represent the standard error of

the mean value (SEM).

The 12 hour unexposed side (Figure 16a) has a higher work function (more positive values) than the 4 and 8 hour exposed samples (Figure 18), and this may have been caused by contamination (Figure 19b) on the edge of sam-ple 004A (Figure 16a). The higher work function likely caused the smaller observed difference in work function in the 12 hour sample. When a larger scan (Figure 20) was performed on sample 004A, the contaminated area in the previous scan (Figure 20b/Figure 19b, same area as presented in Figure 16a) exhibited a much higher work function than the sample surface located farther away from the cut edge (i.e. more towards the middle) as shown in Figure 20a.

Figure 18Work function measurements (surface average) of exposed and unexposed samples

17



Figure 19Sample 004A: (a) Clean (middle of the unexposed sample) and (b) Contaminated (edge of the unexposed sample -

Figure 16a).

Figure 20: Sample 004A: (a) middle of sample and (b) edge of sample in previous scan (as shown in Figure 16a).

The 12-hour exposure also showed much more significant cratering than the 4 and 8-hour samples, with some craters nearly reaching the primer layer. When looking at the exposed surface (Figure 16c) in comparison to the craters (Figure 16d) with the SKP, there is a difference of 0.111 V between the two (or 0.148 V difference from the unexposed side (Figure 16a)).

When considering the Raman analysis, the observed decrease in the amplitude of the peaks at 1600, 1440, and 1350 cm-1 represent changes in the urethane stretching (C=O), methylene asymmetric bend (R-CH2-R), and ester symmetric stretching [(C=O)-O-C], vibration bands, respectively (Figure 21).19,20,21,22 A decrease in the amplitude of these peaks between the 4 and the 8 and 12 hour exposures suggests that hydrolysis of the ester groups, rather than chemical reversion of the polymer coating, is occurring with increasing exposure to the elevated temperature and relative humidity conditions.23 The observed trend of decreasing amplitude of these peaks for the 4, 8 and 12 hour samples appears to correspond with the decreasing trend of the difference in work function (Figures 17 and 18). While these observed differences are most pronounced between the 4 and 12 hour exposures, it should be noted that no craters were observed on the 4 hour exposure samples. It is possible that the craters observed on the 8 and 12 hour samples are indicative of significant chemical and physical changes to the polymer coating, which are reflected in the differences observed for both the work function and Raman analyses. Therefore the re -sults presented suggest that the SKP may be capable of detecting hydrolysis in the polyurethane coating.

18

b)a)



Figure 21Difference in Raman spectra between the crater bottom and the area adjacent to the crater on the 12-hour sample as

well as the 4 and 8 hour exposure samples.

Portable Scanning Kelvin Probe - Alpha-Scale PrototypeThe alpha-scale prototype of the portable SKP was designed to enable characterization of coatings in the field. Considerations included probe tip design, method of maintaining the required standoff from the surface of the structure being assessed, instrumentation needs, and method of portability, as well as operating parameters such as the speed of the measurement, maximum area that could be assessed, resolution, and optimum gap. The por-table SKP was designed to:

Scan the probe across the surface of the sample in a raster form whilst synchronously measuring the work function, such that an image of the work function difference with respect to the sample surface could be ren-dered on the display of a computer.

Allow the user to adjust the parameters of the work function measurement and the scanning profile to control the process.

Be field deployable, meaning that it should be suitable for use on an external structure in dry open air condi-tions. The portable SKP could be taken to a surface of interest and utilized at its location.

The orientation of the portable SKP with respect to the surface of interest could be operable in any plane rela-tive to the ground, i.e. the instrument can scan when mounted horizontally, vertically, inverted or any other such orientation.

The portable SKP will be “placed” on the surface of interest, for the purposes of measurement. The surface may not be flat, it may include a radius of curvature. The portable SKP will measure the work

function difference irrespective of the curvature within defined limits. The alpha-scale portable SKP will not include battery power, but shall be powered from a conventional mains

supply. This In principle an external battery pack could be used with the system if a mains supply was un-available.

The scanning of the surface should be as rapid and efficient as practical to minimize the measurement dura-tion.

Figure 22 shows the design of the alpha-scale portable field-deployable alpha-scale SKP system. The probe can identify degraded areas in horizontal and vertical axes of a coating system stack over an area of greater than 200 x 200 mm.

19



Figure 22CAD design of the portable SKP scanning head

There are four 4-inch rubber cups on the corners to hold the scanning head to the device under test. The rod sticking out from each side controls the suction pump: press 3 or 4 times and it adheres to the surface. Each rub-ber cup is specified to hold 18 kg (40 lbs.). The load capability of each suction cup on the surface of interest is:

Z axis to failure 194 lbs. X and y axis 130 lbs.

The frame is built from aluminum extrusion so that it’s easy to replace and adjust as needed to vary the shape of the scanning head. Ball screws have been used to allow more rapid scanning – 5 mm per revolution. Cable chains enable scanning over a longer distance with collapsible cable trays. At the top right of Figure 1 is the um-bilical cable that will go to the control unit.

When designing the system, the engineers at Bio-Logic kept as much as possible on the scanning head to mini -mize cabling to the control unit. The green control boards are included in the scanning head and can be readily

20



changed in case of breakage or revisions. The box to the left of the green boards is a USB hub. The front-end electronics will be part of the probe and the cables will terminate at the electrometer end so that it can easily be replaced. There is a cover, that can’t be seen in the drawing. Figure 23 shows the completed scanning unit, and computer and electronics contained within a rack mount.

Figure 23Photographs of the alpha-scale portable scanning Kelvin probe and control system.

A Beta-scale prototype of the system is under consideration, based on feedback from the initial experiments with this system. This second, and future, versions may address some of the following questions:

1. Addition of height tracking / OSP to the system2. Improved PID control 3. Improved suction cups and orientation (gimballed, height adjustable) or manipulative robotic arm (instead

of current x-y rail scanning system)4. Addition of a stop on the vertical position, so that it doesn’t pull the scanning head down too close to the

test surface

Portable Scanning Kelvin Probe– Alpha-Scale Preliminary TrialsThe system shown in Figure 23 was delivered and installed at the University of Dayton Research Institute (UDRI), and a successful demonstration of the technology was held at the UDRI facility for personnel from Wright-Patter -son AFB. The portable SKP system was demonstrated in both horizontal and vertical orientations. In the horizon-tal configuration, i.e., sitting on top of the laboratory bench, the system was first tested on a calibration panel

21



shown in Figure 24, left. Figure 24, right clearly shows that the bare steel rectangle was identified using the porta-ble SKP system. Similar results were obtained using the bench-scale M470 SKP system.

Figure 24Left) Galvanized steel panel with a small section of the zinc coating removed, exposing the steel substrate (black

square) and a yellow box showing the scanned area, and Right) a Kelvin scan of the calibration panel area with the red box identifying the bare steel.

For further demonstration of the capabilities of the portable SKP system, an SCCS (Figure 25) was scanned with the SKP system in the horizontal orientation, to demonstrate that the portable SKP could perform rapid measure -ments that could elucidate features on the part surface and under the coating. As shown in Figure 25, there are rivets along the center line of the part and along the center seam, and there are holes in the panel at either end of the SCCS. The Kelvin scan in Figure 26 clearly identifies the rivets beneath the coating, the center seam, and the holes at either end. Future work will use the SKP system to measure the CPD of an SCCS that has been exposed to harsh conditions in an accelerated test.

22



Figure 25Close-up of the center portion of the SCCS, showing the rivet pattern beneath the coating system, and holes at each

end of the middle section, and inset) a photograph of the entire SCCS.

23

Note: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coatingNote: some of the rivets beneath the coating



Figure 26Kelvin scan of the center portion of the SCCS, identifying the rivets beneath the coating, the holes at either end of the

SCCS and the center seam.

For testing in the vertical orientation, a stainless steel panel was fixed to the side of the laboratory bench at UDRI and the portable SKP was mounted using the suction cups in the vertical direction (Figure 27). Parallel strips of copper tape and silver paint were placed on the surface of the stainless steel panel (Figure 28, left). Figure 28, right shows the Kelvin scan of the area of the panel with the copper tape and silver paint, and the difference in work function between the substrate and each strip can been seen.

24

Center seam

End hole

Rivets



Figure 27Left) Stainless steel panel fixed to the lab bench at UDRI, and Right) the portable SKP system mounted with suction

cups to the stainless steel panel in a vertical orientation.

Figure 28Left) Strips of copper tape and silver paint on the stainless steel panel (4 to 5 cm in length), and Right) Kelvin scan of

that area of the stainless steel panel clearly identifying the copper tape and silver paint.

The measurements done in the UDRI laboratory using the alpha-scale portable SKP system are very preliminary, but show the utility of the measurement technique and the ability to scan in unconventional orientations. The scans presented above were measured in 10 minutes, showing that rapid measurements could be made of the

25



surface condition. Future instrumental development and careful selection of the probe tip size may decrease these scanning times further.

Finally, the portable SKP system underwent preliminary testing Hangar 4 Restoration Complex of the U.S. Air Force Museum at Wright-Patterson AFB (Dayton, OH). The portable SKP system was tested on two aircraft on site, the F-15 “STREAK EAGLE” and the MC-12. Both of these aircraft provide an opportunity to demonstrate the ability of the prototype to make NDE and NDI measurements on “real world” USAF aircraft surfaces for opera-tional and field testing. Figure 29 shows some of the photographs from the test day. The results of these mea-surements will be shown in future publications.

Figure 29Left) the alpha-scale portable SKP secured to the side of the F-15 “STREAK EAGLE. Middle) the alpha-scale portable SKP system on the top wing surface of the F-15 “STREAK EAGLE, and Right) the alpha-scale portable SKP system

secured to the side of the MC-12 aircraft.

CONCLUSIONS

The objective of this technology development program is to develop and commercialize a rugged, portable scan-ning Kelvin probe system for quantitatively assessing the degree of degradation of coating stackups on aircraft surfaces. To date, laboratory testing has proven the concept of using a scanning Kelvin probe to monitor degrada-tion in a coating, with the results indicating that the degradation of the polyurethane coating is due to hydrolysis in-stead of chemical reversion. An alpha-scale prototype of the portable SKP system has been designed and built, and testing in the laboratory and in the field in both horizontal and vertical orientations.

ACKNOWLEDGEMENTS

Funding for this work is gratefully acknowledged from Air Force SBIR Phase I Contract Number FA8650-16-M-5065 and Air Force SBIR Phase II Contract Number FA8650-17-C-5074, under the management of Dr. Nicholas Wilson (AFRL/RXSS, Wright Patterson AFB). The technical contributions of David Borth and Dr. Ron Zeszut are greatly appreciated. We would also like to thank the Restoration Division of the National Museum of the U.S. Air Force (NMUSAF/MUR) and museum volunteers for their time and support of the effort.

26



REFERENCES

27



1. Tiong U.H., and G. Clark, “Fatigue: The structural environment as a factor affecting coating failure in aircraft joints” Procedia Engineering, 2 (2010) p. 1393-1401.2. Grundmeier, G., K.M. Juttner, M. Stratmann, “Novel Electrochemical Measurement Techniques in Corro-sion,” Vol. I & II, Materials Science and Technology: A Comprehensive Treatment: Corrosion and Environmental Degradation, eds P. Haasen, E.J. Kramer, R.W. Cahn. Weinheim: Wiley-VCH, (2000).3. Grundmeier, G., W. Schmidt, M. Stratmann, “Corrosion protection by organic coatings: electrochemical mechanism and novel methods of investigation,” Electrochimica Acta, 4:5 (2000) p. 2515-2533.4. Hansen, D.C., H.S. Isaacs, G. Adzic, J. Gitto III and F.J. Martin, “Scanning Kelvin probe measurements for the detection of corrosion processes beneath applied paint coatings on aluminum alloy and steel substrates,” Polymer Preprints, 45:2 (2004) p. 203-204.5. Hansen, D.C., K.M. Hansen, T.L. Ferrell, and T. Thundat, “Scanning Kelvin Probe Characterization of Hemocyte Deposited Ceramic Films on Metallic Surfaces,” Langmuir 19 (2003) p. 7514-7520.6. Peterson, I.R., “Kelvin Probe Liquid-Surface Potential Sensor,” Rev. Sci. Instr. 70:8 (1999) p. 3418 - 3424.7. Bastide, S. D. Gal, D. Cahen, L. Kronik, “Surface Photovoltage Measurements in Liquids,” Rev. Sci. Instr. 70, (1999) p. 4032-4036.8. Grundmeier, G., M. Stratmann, and K. Wapner, “In situ infrared spectroscopic and scanning Kelvin probe measurements of water and ion transport at polymer/metal interfaces,” Electrochimica Acta 51, (2006) p. 3303-3315.9. Hughes, T.H. and A.E. Muster. (2006) J. Electrochem. Society 153:11, B474.10. Janata, J. and M. Josowicz (1997) Analytical Chemistry 69:9, 293A.11. Peterson, I.R. (1999) Rev. Sci. Instr. 70, 3418.12. Bastide, S., D. Gal, D. Cahen, L. Kronik. (1999) Rev. Sci. Instr. 70, 4032.13. Lu, E. Delamarche, L. Eng, R. Bennewitz, E. Meyer and H.-J. Guntherodt. Langmuir 15 (1999): 8184-8188.14. McNaught, A.D.J. and A. Wilkinson, ed. IUPAC Compendium of Chemical Terminology. 2nd. Malden, MA: Blackwell Science, Inc., 1997.15. Hansen, D.C. 2005 Tri-Service Corrosion Conference. Orlando, FL.: NACE International, 2005.16. Hansen, D.C., K.M. Hansen, T.L. Ferrell and T. Thundat. Langmuir 19 (2003): 7514-7520.17. Grundmeier, G., W. Schmidt and M. Stratmann. Electrochimica Acta 45 (2000): 2515-2533.18. Wapner, K., M. Stratmann and G. Grundmeier (2006) Electrochimica Acta 51, 3303.19. P. Vandenabeele, Practical Raman Spectroscopy: An Introduction (Chichester, United Kingdom: Wiley, 2013).20. G. Socrates, Infrared and Raman Characteristic Group Frequencies: Tables and Charts (Chichester, United Kingdom: Wiley, 2001).21. K. Bruckmoser, K. Resch, Macromol. Symp. 339, 1 (2014): p. 70-83.22. P. Larkin, IR and Raman Spectroscopy: Principles and Spectral Interpretation (Oxford, UK: Elsevier, 2011).23. Borth, D.J., E.B. Iezzi, D.S. Dudis and D.C. Hansen. Corrosion 75 (2019): 457- 464.

Documents

Abstract… · Web viewKeywords: Coatings, Hydrolysis, Kelvin Probe, Corrosion Monitoring, Non-Destructive Evaluation, Polymer Degradation, Polyurethane Abstract Degradation of coatings