You are on page 1of 8

ME TECHNICAL PAPER w x

From Materials Evaluation, Vol. 70, No. 6, pp: 757764. Copyright 2012 The American Society for Nondestructive Testing, Inc.

Fluorescent Nondestructive Testing Sensitivity Improvement through Choice of Safety Glasses


by Richard D. Lopez*

ABSTRACT

A quantitative evaluation of the relative effect of various safety glasses on inspection sensitivity was performed. This work characterized lens transmission spectra, determined penetrant and magnetic particle test media fluorescent emission spectra, measured ultraviolet A radiation source emission spectra, and determined the effect of lens choice on the fluorescent luminance of typical test media. Experimental data showed that not all lenses are equal, and one amber lens allowed a greater amount of fluoresced light from indications through to the technicians eyes, compared to its clear counterpart. KEYWORDS: fluorescent, penetrant, magnetic particle, safety glasses, sensitivity, clear lenses, amber lenses.

Introduction Fluorescent nondestructive testing (NDT) test media, which include liquid penetrants and magnetic particles, absorb energy from an excitation radiation source (exciter) and emit fluoresced visible light. Specimen geometry or surface finish can be such that excitation radiation is reflected into the technicians eyes. When strong blue light or ultraviolet A (UVA) reaches the eye, test sensitivity suffers, and the technician is exposed to hazardous wavelengths (Clarke, 1954). Protective eyewear generally has clear lenses, although amber lenses also have their place in NDT. Clear lenses block only ultraviolet (UV) radiation, and allow all visible light to pass. Amber lenses block UV radiation and some visible light, transmitting only blue-green visible light and longer wavelengths. Amber safety lenses may be found in industrial fluorescent NDT, but their use has not been widely adopted. Selecting the optimal safety glass lens will maximize inspection sensitivity because indication luminance and probability of detection are directly proportional. Experiments were performed to determine which safety lens was best when utilizing fluorescent test media and common UV excitation sources. This paper represents a portion of the work reported in the authors masters thesis, which focused on the effect of excitation source and filtering safety lens on fluorescent NDT (Lopez, 2010). Health and Safety UV radiation is generally defined as electromagnetic radiation with a wavelength between 100 and 400 nm; photons in this energy range can harm biological cells. The small subdivision of UV radiation utilized in fluorescent penetrant testing and fluorescent magnetic particle testing (MT), known as UVA, was long considered completely safe for extended exposure to the technician. UVA radiation has a wavelength between 315 and 400 nm. Recent research on UVA radiation exposure suggests that a measureable health risk is present, although this risk varies greatly with photon wavelength (Diffey, 2002). Possible deleterious effects of excessive UVA exposure are cataracts, retinal burns and skin cancer (Reed et al., 2009). These photobiological effects have been studied and defined by health and safety organizations. One domestic authority, the American
JUNE 2012 MATERIALS EVALUATION

* John Deere Moline Technology Innovation Center, 1915 Scholl Rd., 283 ASC-II, Ames, Iowa 50011; (309) 749-9337.

757

ME TECHNICAL PAPER w x safety glasses

Conference of Governmental Industrial Hygienists (ACGIH), maintains and periodically updates guideline values and indices for exposure to hazardous chemicals and physical agents (ACGIH, 2004). One of ACGIHs guidelines offers recommendations on safe exposure to UV radiation, visible light and infrared radiation. In NDT, there are two main concerns when working with non-ionizing radiation: retinal photochemical injury due to blue light exposure, and extended UV exposure of the skin and eyes. Blue light hazard pertains to wavelengths between 400 and 500 nm, and this action spectrum peaks between 435 and 440 nm (Algvere et al., 2006; Zuclich et al., 2005). The ACGIH UV hazard action spectrum (S) is defined between 180 and 400 nm, and peaks at 270 nm (ACGIH, 2004; Schmidt, 1975). Radiometers and luxmeters used in NDT employ integral sensors, which are assemblies consisting of a photodiode, aperture, filter set and light diffusing cover (Xu and Huang, 2000). Specially filtered integral radiometer sensors that replicate blue light hazard and ACGIH UV hazard action spectra are both commercially available.
Proposed Benefit Amber lenses have been promoted for fluorescent NDT usage for decades (Betz, 1969; Hagemaier and Bowles, 1979; Holden, 1983; Lopez et al., 2006; Ness and Moss, 1996). These lenses are mandatory for blue light exciters; however, their use with UVA sources has not been widely adopted. The

purpose of this paper is to show that amber lenses improve fluorescent indication contrast when using UVA exciters. Figure 1 illustrates the change in visual appearance of indications on a penetrant testing and monitoring (TAM) panel irradiated with a typical micro-power xenon light (MPXL) exciter. In this case, a UVA-blocking photographic filter was replaced by a #2E amber filter. Background brightness decreased while indication brightness remained qualitatively unchanged with the amber filter. The signal-to-noise ratio improvement was solely because of the amber (blueblocking) filters ability to absorb the reflected excitation energy before it reached the camera. While indication luminance appeared to be unaffected in this photographic experiment, a thorough and quantitative analysis was required to determine if this was indeed the case.
Experimental Overview Experiments were conducted to evaluate the effectiveness of amber safety lenses in improving the sensitivity of fluorescent NDT. This work established the wavelengths emitted by eight UVA exciters, determined the visible light wavelengths emitted by fluorescent test media, estimated the transmission characteristics of four common safety glasses, and measured actual fluorescent luminance of typical test media. Luminance measurements were made on bulk test media samples with, and without, the filtering effect of safety glasses.

(a)

(b) Figure 1. Five fluorescent penetrant testing indications on a testing and monitoring panel irradiated by a micro-power xenon light source and photographed while changing only the filter: (a) a clear ultraviolet A (UVA) filter was replaced by a (b) UVA and blue light blocking #2E amber photographic filter.

758

MATERIALS EVALUATION JUNE 2012

Ultraviolet A Exciters Excitation sources utilized in this study included a filtered medium-pressure mercury vapor lamp, a filtered short-arc mercury vapor lamp, two filtered MPXL sources, an integrally-filtered (phosphor-coated) low-pressure mercury vapor fluorescent tube-based exciter, and a selection of UV light-emitting diode (UV-LED) sources. This overall group of exciters represented the variety of exciters used with, and around, fluorescent NDT activities. A UV-visible (UV-Vis) spectrometer is a device that characterizes the emission of a UV or visible light electromagnetic radiation source in counts per integration time, relative irradiance or absolute irradiance versus wavelength. The price of such spectrometers varies widely, and the low-cost option for obtaining spectral data is a polychromator spectrometer, which contains a linear charge-coupled device (CCD) array and stationary grating. These polychromator spectrometers cost a fraction of the price of quality double monochromator spectrometers. High-end devices, however, offer optimal signal-to-noise levels, dynamic range and data resolution. To capture emission spectra from each exciter, a CCD array polychromator spectrometer was coupled to a 230 m core extreme solarization-resistant optical fiber. This spectrometer was configured with a 10 m slit and a 600 mm1 grating groove density. Such a configuration allowed for rapid collection of emission data over a range of 250 to 800 nm, with a maximum sensitivity at 400 nm and approximately 1 nm resolution. As expected, the spectra emitted by the UVA sources varied dramatically. Emission spectra (Figure 2) were gathered in relative irradiance mode, which canceled out much of the inherent electronic noise and system error associated with the spectrometer components and optical fiber. Table 1 summarizes the peak or mathematical central emission wavelength, as well as each exciters spectral width in terms of a full-width at half-maximum (FWHM) value. FWHM is simply the spectral emission width, in nanometers, at half the height of its peak. The historical standard medium-pressure mercury vapor exciter was centered at 365 nm with the narrowest FWHM of 2.6 nm, while the fluorescent integral-filtered tube-based exciter was centered at 351 nm with the broadest FWHM of 37.5 nm. The MPXL exciter, which has proven to be quite successful in NDT applications, was shown to have a broad emission spectrum mathematically centered at 372 nm. While the MPXL source emitted some violet visible light, it also emitted longer wavelength red visible light. This red light emission was facilitated by the transmission characteristics of Woods glass UVA filters, which allow radiation with wavelengths between 700 and 840 nm to pass. Exciter irradiance and illuminance were measured with high-quality integral sensors. Because none of the exciters emission spectra matched the spectrum that was used to calibrate the high-quality integral sensors, spectral mismatch

Figure 2. Normalized relative irradiance from common ultraviolet A exciters, as captured with zero boxcar smoothing and 25-scan averaging using a polychromator spectrometer.
TABLE 1

Central or peak emission wavelength, and full-width at halfmaximum (FWHM) for each exciter Exciter
Mercury vapor Micro-power xenon light Integrally-filtered tube-based Ultraviolet light-emitting diode 1 Ultraviolet light-emitting diode 2 Ultraviolet light-emitting diode 3 Ultraviolet light-emitting diode 4 Mercury short arc

Wavelength (nm)
365 372 351 367 368 371 366 366

FWHM (nm)
2.6 36.3 37.5 7.4 11.0 14.0 8.4 9 .0

correction factors were calculated for each exciter/sensor combination so that absolute measurements could be obtained (Envall et al., 2006; Field and Emery, 1993; GuggHelminger et al., 2004; Larason and Cromer, 2001; Reed et al., 2009; Xu and Huang, 2000). Spectral mismatch measurement error is unavoidable without correction factors. Absolute irradiance was an important measurement, because with it one could ensure that test media samples were equally irradiated by the variety of exciter types. Illuminance varied between exciter types, but no measurement exceeded the 21.5 lux limit called for in typical industrial standards.
Test Media Fluorophores, which are fluorescent dyes or pigments, absorb specific wavelengths of excitation energy and quickly release that energy in the form of longer wavelength visible light. Modern fluorescent test media are generally composed of at least two fluorophores: an optical brightener that absorbs UVA energy and emits blue light (with a peak near 425 nm), and another fluorophore that absorbs blue light to emit the
JUNE 2012 MATERIALS EVALUATION

759

ME TECHNICAL PAPER w x safety glasses

characteristic yellow-green color common to most fluorescent test media (Graham, 1967). This teamwork approach to fluorescence is known as the cascade effect. A spectrofluorometer is a device capable of determining fluorescent excitation and fluorescent emission spectra of test media. A fluorescent emission scan irradiates the sample with a single wavelength of light, while scanning across a specific wavelength range for the emission of longer wavelength fluoresced light. This test irradiated samples with 365 nm energy, while measuring the intensity of light emitted with a wavelength between 400 and 700 nm. Fluorescent emission spectra for 190 fluorescent penetrant and magnetic particle samples were obtained experimentally with a UV-Vis spectrofluorometer. These samples represented 41 different products, and included up to 13 different production years of some individual products. Samples were held within disposable poly(methyl methacrylate) cuvettes to avoid cross-contamination. Fluorescent emission results for 32 varieties of fluorescent penetrant testing test media showed that penetrants generally acted similarly. With the exception of Mfr D-Pen 1, penetrant test media were shown to have peak fluorescent emission wavelengths between 499 and 518 nm (Figure 3). Three representative penetrants (Mfr D-Pen 7, Mfr A-Pen 2 and Mfr B-Pen 2) were chosen for further work based on their fluorescent emission characteristics. This small subset of products all exhibited relatively small optical brightener emission peaks near 425 nm. Mfr B-Pen 2 exhibited peak emission at 499 nm; Mfr A-Pen 2 peaked at 509 nm; and Mfr D-Pen 7 peaked at 514 nm. Considerable variance in the relative intensity of an emission peak near 425 nm can be seen in Figure 3. This variance was analyzed, and an inverse relationship between the 425-nm optical brighteners fluorescent emission peak height and the sensitivity level of the penetrant was noted. Liquid penetrant testing media are classified by sensitivity,

Normalized optical brightener amplitude

0.8 0.6 0.4 0.2 0 0.5

Penetrant sensitivity level


Figure 4. Normalized height of the 425 nm optical brightener fluorophore peak noted in spectrofluorometry data versus penetrant sensitivity level. R2 = 0.7222.

Figure 3. Average normalized fluorescent emission spectra for batches of 32 different penetrant test media, where multiple batches of a given product are represented by a single data set.

which varies from the lowest level, level one-half, to the highest sensitivity, level 4, penetrants (SAE, 2006). A primary difference between penetrant sensitivity levels is the concentration of the more expensive cascading fluorophore (Graham, 1967). Test medium Mfr D-Pen 1, which was the lone level onehalf sensitivity water-washable penetrant included in this study, had the only fluorescent emission spectrum in which its 425 nm optical brightener peak was higher than its 508 nm cascading fluorophore peak. This suggested that when cascading fluorophore concentration was low, peak height of the optical brightener emission increased. Conversely, when concentration was high, essentially all of the optical brighteners 425 nm fluorescent emission was absorbed by the cascading fluorophore and converted into 510 nm light. In Figure 4, a plot of penetrant sensitivity level versus 425 nm optical brightener peak height showed that the relative height of the optical brightener peak scaled inversely with cascading dye content. The low sensitivity of this test medium likely equated to a reduced concentration of the expensive cascading dye, which then allowed the low-cost optical brightener to dominate the emission spectrum. A trend in the location of the peak emission wavelength was also noted when peak fluorescent emission wavelength was analyzed versus penetrant type (Figure 5). Post-emulsifiable (also known as Method B or D) penetrants generally had shorter wavelength emission peaks compared to the common and the biodegradable water-washable (also known as Method A) penetrants. This difference was deemed statistically significant through the use of a Students t-test analysis. Convention holds that the same yellow-green fluorophore is common across all penetrant brands, so it is feasible that chemistry variations between post-emulsifiable and waterwashable penetrant types were the cause of this fluorescent emission peak shift. Variation in fluorescent emission peak

760

MATERIALS EVALUATION JUNE 2012

Emission wavelength peak (nm)

520

515

510

505

through 2007 batches of Mfr D-Mag 1 was positioned between 519 and 522 nm. A small emission peak blue shifting (a shift to shorter wavelengths) was noted in the 2008 batch of Mfr D-Mag 1 with respect to other batches of the same product, and this batch peaked at 516 nm. The average of Mfr D-Mag 2 emission spectra peaked at 521 nm, while the average for Mfr B-Mag 1 was at 522 nm. Samples from Mfr B-Mag 1, Mfr D-Mag 1 and Mfr D-Mag 2 were selected for further experimental work based on their emission spectra.
Post-emulsifiable Water-washable Biodegradable water-washable

500

Type
Figure 5. One-way statistical analysis of the peak fluorescent emission wavelength versus penetrant type, showing that post-emulsifiable penetrants generally had a shorter-wavelength emission peak than water-washable and biodegradeable waterwashable penetrants.

versus manufacturer and sensitivity level were not statistically significant. Peak emission wavelength is important when considering the effect of filtering safety glasses because a portion of the light emitted by an indication may be removed from view. The luminance of shorter-wavelength fluorescent emission spectra would be decreased to a greater extent than longer-wavelength spectra. A filters cut-on wavelength is defined in this paper as the shortest wavelength at which transmittance exceeds 50%. Fluorescent emission spectra from 21 samples from five varieties of MT media exhibited little variance between products, and the general shape of emission peaks for MT media was identical with the exception of Mfr B-Mag 1 (Figure 6). Peak emission for the 1996

Figure 6. Average normalized fluorescent emission spectra for batches of magnetic particle test media, where multiple batches are represented by a single data set.

Safety Glasses Fluorescent NDT is performed with the aid of safety glasses to protect the operator from adverse health effects and to increase an indications contrast ratio (indication luminance versus background luminance). While photochromic (auto darkening) and neutral density glasses are poor choices, longpass-filtering glasses, which may be clear or amber in color, are preferred. Amber glasses block UV radiation and begin to transmit light at a longer wavelength than clear glasses. The point at which the amber lens begins to transmit, as well as the maximum transmittance, varies by manufacturer. Many amber safety glasses available for NDT applications today are designed for use with blue light exciters, and their filtering cut-on wavelengths are longer so that the blue excitation energy is effectively blocked. Safety lens transmission spectra affect background luminance, as well as the luminance of fluorescing test media during an inspection. Full transmission spectrum characterization of a curved safety lens would require the use of a spectrometer coupled to an integrating sphere. An integrating sphere would capture all transmitted light regardless of the refracted angle caused by the curved safety lens. Such a system was not available for this experimental work, so approximate transmission spectra were captured using equipment on hand. An unfiltered, 35 W MPXL lamp was utilized as a transmission source, as this lamp offered strong irradiance between 260 and 800 nm. The polychromator spectrometer described above was again utilized, this time in transmission mode. The spectrometer was set to a 3 ms integration time, 150 scan averaging, and 1 nm boxcar averaging. The radiation source and optical fiber were rigidly held during experimentation to avoid positioning error. Calibration of the spectrometer in transmission mode set transmittance to 100% for all wavelengths. When a filter was introduced into the light path, its optical properties altered the spectral intensity of light reaching the detector. Characteristic transmission spectra for each filter could then be observed based on how they absorbed radiation from the source. While spectra gathered in this manner were noisier than those that would be gathered using the preferred equipment, they compared favorably with information available from the lens manufacturers.
JUNE 2012 MATERIALS EVALUATION

761

ME TECHNICAL PAPER w x safety glasses

Figure 7. Combined spectrometer and spectrofluorometer data showing why a contrast ratio improvement is possible with the optimum amber safety lens: clear safety glasses transmit a greater portion of excitation energy, and some amber lenses transmit less of the visible light emitted by fluorescent indications.

Figure 8. Transmission spectrum of amber lens 1 safety glasses overlaid by the fluorescent emission spectrum of six representative fluorescent test media.

With approximate safety lens transmission curves, one may consider the effect of glasses on the luminance of typical fluorescent test media. The relative positions of emission spectra from NDT exciters, transmission spectra of typical safety lenses, and fluorescent emission spectra of test media all interact to produce a given contrast ratio (Figure 7). A clear lens absorbs UVA (315400 nm), but allows some excitation energy to reach the technicians eye. Amber lenses all absorb UVA and many blue light hazard wavelengths (400500 nm), but some amber lenses absorb a greater amount of visible light emitted by a yellow-green fluorescent indication. A well-selected pair of amber safety glasses, such as amber lens 1 evaluated in this study, would block all reflected excitation radiation while transmitting the majority of the fluoresced light emitted by an indication. A summary of the filtering safety glasses evaluated in this study is presented in Table 2.

Amber lens 1 qualitatively appeared to have the highest maximum transmittance, while amber lens 2 appeared to have the lowest maximum transmittance and the longest cut-on wavelength. As previously shown, the peak fluorescent emission wavelength varies between test media options (Figures 3 and 6), and this variation would be expected to affect the luminance of indications depending on the chosen safety lens. As an example, Figure 8 overlays the fluorescent emission spectra of six test media samples atop the amber lens 1 transmission spectrum. Visually, one would expect that the cut-on wavelength for amber lens 1 would have a negligible impact on indication luminance, and that the slight increase in maximum transmittance could offer an advantage over the clear lens. Amber lenses 2 and 3 (Figure 7) would be expected to result in lower indication luminance values for a given set of conditions.
Luminance Measurements Experimental work was performed to measure the filtered and unfiltered luminance of six varieties of test media when each was exposed to irradiation from UVA exciters. Bulk samples of test media were used rather than crack indications. This evaluation technique was chosen for two reasons: the luminance of actual penetrant indications decreases quite rapidly with time when exposed to irradiance levels of 50 W/m2 or greater; and the curved surface of safety glasses refracts light, which greatly decreases measurement repeatability on small crack indications when manually holding the filtering lenses within the optical path (Lopez, 2006; Lopez et al., 2006). Luminance was measured with a fully automatic filter photometer, which employed a 0.5 aperture, automatic neutral density front filter and photopic rear filter set. Test media, as described above, were chosen based on their fluorescent emission spectra. The general setup for these

TABLE 2

Safety lens color, shortest transmission wavelength and approximate cut-on wavelength where the 50% transmittance threshold was reached Safety glasses
Clear lens Amber lens 1 Amber lens 2 Amber lens 3

Transmission begin (nm)


390 409 488 469

Cut-on wavelength (nm)


401 469 513 501

762

MATERIALS EVALUATION JUNE 2012

experiments is presented in Figure 9. Exciters were positioned at varied heights above a black felt-covered stand. The removable felt-covered stand held test media samples at exactly the same height as the integral sensors, so that the distance between the exciter and sensor/sample remained constant. Standoff distance of the exciter controlled irradiance, and black felt on the removable stand minimized background luminance. The photometer was positioned and rigidly held with the aid of a geared tripod head. Figure 9 also offers two views of test media samples through the photometer. Photobleaching of fluorescent penetrants was a potential source of error in this work, and efforts were made to minimize its effects. Photobleaching is a time-dependent decrease in fluorescent luminance, which can be minimized by reducing irradiance, or by reducing the interaction between an excited fluorophore and air. Past work showed that holding a thin air-free layer of penetrant between two UV grade fused silica windows offered such protection for irradiance levels at least as high as 200 W/m2 (Lopez, 2006; Lopez et al., 2006). UV grade fused silica is transparent to UVA and visible light, and thus ideal for this work. A sandwich-like configuration was utilized for all liquid penetrant samples (Figure 9c). Magnetic particle samples, which presented less photobleaching concern due to their

(b)

(a)

(c)

Figure 9. General setup for luminance experiments showing: (a) the relative position of photometer, sample and exciter; (b) close-ups of a typical magnetic particle test medium; (c) and a typical penetrant test medium sample when viewed under ultraviolet A excitation.

resin-encapsulated nature, were simply placed upon one fused silica window in an open configuration (Figure 9b). Four exciters were selected for this work: a mediumpressure mercury vapor, an MPXL source, an integrallyfiltered tube-based exciter, and a UV-LED source. These exciters could be positioned at various heights to expose test media samples to an absolute irradiance of 10 to 90 W/m2. Luminance of the black felt increased with irradiance, and it is known that fluorescent NDT relies on the contrast ratio between an indication and its surroundings (Schmidt and Robinson, 1984). Luminance of the black felt was therefore subtracted, through the use of equations obtained from bestfit trend lines, from test media luminance measurements specific to the test conditions. This subtraction technique facilitated a fair analysis of the observed contrast for all exciters and test media and provided a measure of signal strength versus background noise. The fluorescent emission spectrum of a test medium is independent of excitation spectrum; therefore, experimental results with various exciters could be compiled to determine an average response for each filtering lens (Lakowicz, 2006). Photometric luminance measurements were made with, and without, safety lenses positioned in the optical path of the photometer. The amount of variance for a given filter, when all exciters were considered, was minimal (3 to 6% difference), and this variance was likely within the bounds of experimental error. Experimental results shown in Table 3 established the relative effect of various lens when measuring the luminance of each test media sample. While correlation was not perfect, the general trend was that luminance from test media with shorter-wavelength emission peaks (Mfr B-Pen 2 and Mfr A-Pen 2, for example) were more attenuated by amber lens 2 and amber lens 3 than longer-wavelength peak emission test media (Mfr B-Mag 1 and Mfr D-Mag 2, for example). Data show that amber lens 1 always allowed a greater or equal amount of fluoresced light emitted by a test medium to reach the detector, as compared to its clear counterpart. In industrial applications this would result in more light from an indication reaching the technicians eyes and a higher expected probability of detection. Amber lenses designed for use with blue light exciters (amber lenses 2 and 3), which were shown to have a longer cut-on wavelength in Table 2, allowed less fluoresced light to reach

TABLE 3

Percent of unfiltered luminance of each fluorescing test medium when viewed through four different filtering lens options
Filter type
No filter Clear lens Amber lens 1 Amber lens 2 Amber lens 3
* Actual results

Mfr B-Pen 2
100% 91% 92% 70% 74%

Mfr A-Pen 2
100% 89% 92% 75% 78%

Mfr D-Pen 7
100% 87% 89% 74% 77%

Mfr B-Mag 1
100% 88% 88% 80% 82%

Mfr D-Mag 2
100% 92% 92% 80% 83%

Mfr D-Mag 1
100% 84% 87% 74% 78%

JUNE 2012 MATERIALS EVALUATION

763

ME TECHNICAL PAPER w x safety glasses

the detector. In general terms, it could be assumed that indications viewed through amber lens 2 and amber lens 3 would be 6 to 22% dimmer compared to indications viewed through amber lens 1.
Conclusion Safety glasses are required for fluorescent NDT to protect the technician from harmful UV radiation, but that is not the sole reason behind their importance. Inspection sensitivity can be maximized with an amber lens by preventing UVA, as well as a portion of the visible light spectrum, from reaching the technicians eyes. One must choose an amber lens carefully because the long-pass filtering, cut-on wavelength varies with model. A longer wavelength cut-on will result in less of the visible light emitted by yellow-green fluorescent indications reaching the eye. Cut-on wavelength is key when the goal is a signal-to-noise ratio improvement. Background luminance (noise) reduction is based on the relative positions of the safety lens transmission spectrum, and the test mediums fluorescent emission spectrum. Qualitative spectrometry results showed that the cut-on wavelength of a clear safety lens was 401 nm, while three common amber lenses varied between 469 and 513 nm. Experimentation with 190 different fluorescent penetrant and magnetic particle media samples revealed that penetrant peak emission wavelength ranged from 499 to 518 nm, and magnetic particle peak emission wavelength ranged from 516 to 522 nm. Improper amber lenses degrade inspection sensitivity by blocking light from the indication, and some were shown to be 6 to 22% worse than the top performer. The optimal safety lens choice (amber lens 1) allowed significantly more fluoresced light emitted by the test media to reach the detector, as compared to other amber-colored lenses evaluated. Amber lens 1 also outperformed its clear counterpart in most cases. This increase in light throughput was due to its cut-on wavelength and high transmittance. When more light from an indication reaches the technicians eyes, especially when a higher signal-to-noise ratio is attained, higher probability of detection is expected. When utilizing typical fluorescent magnetic particles and fluorescent penetrants of level 1 sensitivity and higher, use of a safety lens with a transmission spectrum like amber lens 1 is highly recommended to maximize inspection sensitivity.
REFERENCES

ACGIH, TLVs and BEIs Based on the Documentation of the Threshold Limit Values for Chemical Substances and Physical Agents & Biological Exposure Indices, American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio, 2004. Algvere, P.V., J. Marshall and S. Seregard, Age-Related Maculopathy and the Impact of Blue Light Hazard, Acta Opthalomologica Scandinavica, Vol. 84, No. 1, 2006, pp. 415.

Betz, C.E., Principles of Penetrants, 2nd ed., Magnaflux Corporation, Chicago, Illinois, 1969. Clarke, J.E., Factors in the Use of Black Lights in Fluorescent Inspection, Nondestructive Testing, Vol. 12, No. 4, 1954, pp. 2125. Diffey, B.L., Sources and Measurement of Ultraviolet Radiation, Methods, Vol. 28, No. 1, 2002, pp. 413. Envall, J., L. Ylianttila, H. Moseley, A. Coleman, M. Durak, P. Krh and E. Ikonen, Investigation of Comparison Methods for UVA Irradiance Responsivity Calibration Facilities, Metrologia, Vol. 43, No. 2, 2006, pp. S27S30. Field, H. and K. Emery, An Uncertainty Analysis of the Spectral Correction Factor, Proceedings of the 23rd IEEE Photovoltaic Specialists Conference, Louisville, Kentucky, 1993, pp. 11801187. Graham, B.C., Mechanisms Contributing to Fluorescence and Visibility of Penetrants, Proceedings of the Fifth International Conference on Nondestructive Testing, Montreal, Canada, May 1967, pp. 225233. Gugg-Helminger, A., W. Dhn, S. Fenk and B. Angelo, Spectral Matching of Actinic Functions, Spectral Mismatch Correction Factors and Calibration of UV Detector Calibration, Proceedings of the 196th PhysikalischTechnische Bundesanstalt Seminar Workshop jointly with the DAfP Traceability in UV Dosimetry Applications and Requirements, Karlsruhe, Germany, 22 September 2004. Hagemaier, D. J. and D. Bowles, Evaluation of Magnetic Particle Inspection Oxides, Materials Evaluation, Vol. 37, No. 11, 1979, pp. 4753. Holden, W.O., UV/Black Light Measurement for NDT, Materials Evaluation, Vol. 41, No. 3, 1983, pp. 244, 246, 248249. Lakowicz, J.R., Principles of Fluorescence Spectroscopy, 3rd ed., Springer, New York, New York, 2006. Larason, T.C. and C.L. Cromer, Sources of Error in UV Radiation Measurements, Journal of Research of the National Institute of Standards and Technology, Vol. 106, No. 7/8, 2001, pp. 649656. Lopez, R.D., L. Brasche and D. Eisenmann, UV-A Induced Fade in Fluorescent Penetrant Tests, The NDT Technician, Vol. 5, No. 3, 2006, pp. 14. Lopez, R.D. UVA-Induced Fade of Penetrant and FPI Indications, Review of Progress in Quantitative Nondestructive Evaluation, Vol. 25A, D.O. Thompson and D.E. Chimenti (eds.), AIP Conference Proceedings, Melville, New York, 2006, pp. 590597. Lopez, R.D., Comparison of Radiation Sources and Filtering Safety Glasses for Fluorescent Nondestructive Evaluation, ProQuest Dissertations and Theses, Accession Order No. AAT 1487993, 2010. Ness, S., and C.E. Moss, Current Concerns about Optical Radiation Safety in Fluorescent Magnetic Particle and Penetrant Methods, Materials Evaluation, Vol. 54, No. 3, 1996, pp. 364367. Reed, N.G., S. Wengraitis and D.H. Sliney, Intercomparison of Instruments used for Safety and Performance Measurements of Ultraviolet Germicidal Irradiation Lamps, Journal of Occupational and Environmental Hygiene, Vol. 6, No. 5, 2009, pp. 289297. SAE, SAE AMS 2644E, Inspection Material, Penetrant, SAE International, Warrendale, Pennsylvania, 2006. Schmidt, J.T., Black Light for Inspection Use, Materials Evaluation, Vol. 33, No. 11, 1975, pp. 21A23A, 26A, 35A38A. Schmidt, J.T. and S.J. Robinson, Penetrant Fluorescence Measurement, 1982 Model, Materials Evaluation, Vol. 42, No. 3, 1984, pp. 325332. Xu, G. and X. Huang, Characterization and Calibration of Broadband Ultraviolet Radiometers, Metrologia, Vol. 37, No. 3, 2000, pp. 235242. Zuclich, J.A., F.H. Previc, B.J. Novar and P.R. Edsall, Near-UV/Blue Light-induced Fluorescence in the Human Lens: Potential Interference with Visual Function, Journal of Biomedical Optics, Vol. 10, No. 4, 2005, 044021.

764

MATERIALS EVALUATION JUNE 2012

You might also like