Optical Radiation Transmission Levels Through Transparent Welding Curtains

Tests with blue, green, gray and yellow welding curtains from commercial sources indicate yellow curtains transmit a larger component of hazardous wavelengths than the others, although the magnitude of optical energy transmitted by yellow curtains is small

BY C. EUGENE MOSS

ABSTRACT. Results are described for communication among workers and poraled, and Wilson Sales Company. tests conducted on 25 transparent with management. Industrial use of As far as it can be determined these welding curtain samples of four colors the transparent curtains, however, has were the only known manufacturers of from five manufacturers to document resulted in questions concerning the transparent welding curtains at the spectral transmission values of curtains protection they afford workers and time of testing. used in welding environments. The observers outside the immediate Some samples had been previously conclusion based on the results of welding area. However, the published acquired which enabled transmission these tests suggests that most transpar­ literature, excluding manufacturers' tests to be conducted on aged ent curtains will afford adequate specifications, for such curtains is curtains. In addition, a used bicolored protection from optical radiation in limited." For this reason the National welding curtain was obtained from an the ultraviolet region, but that all of Institute for Occupational Safety and industrial welding facility. Samples them will transmit high percentages of Health undertook tests to determine from this particular curtain permitted infrared radiation. optical radiation protection provided spectral transmission comparisons to by transparent welding curtains: this be made on a before and after clean­ Introduction paper describes the results of the tests ing condition. performed. Table I categorizes the samples Numerous skin and eye injuries have tested according to manufacturer, col­ occurred in the welding environment Apparatus and Methods or, thickness, and age. from exposure to ultraviolet, visible, and infrared radiation (hereinafter de­ Acquisition of Welding Curtain Samples Transmission Measurement System noted optical radiation).17 A control measure that has been used to reduce Samples of transparent welding cur­ All spectral transmission measure­ the incidence of injuries to bystanders tains were requested from all those ments were performed with a Perkin- consists of using welding curtains manufacturers who were identified Elmer Model 323 recording spectro- made of metal or flame-resistant fabric through a search of the welding trade photometer. This system has a low that are generally opaque to most opti­ literature. Twenty-one curtain samples sensitivity level which makes it a valu­ cal radiations. of various colors and sizes were able tool to document low transmis­ received. A relatively recent innovation in sion values. The spectrophotometer is Manufacturers that provided sam­ curtain material has been the use of a double-beam single-pass recording ples were Davlynne Incorporated, FIRL transparent colored curtains made monochromator that uses deuterium Industries Incorporated, Frommelt In­ from polyvinyl chloride plastic film and tungsten lamps as the optical radi­ dustries, Singer Safety Products Incor- and at least three other components. ation sources. The detector used in the These components are a color dye, system consists of a photomultiplier tube for the ultraviolet and visible flame retardant chemical, and an Paper to be presented at a symposium regions, and a lead-sulfide cell for the ultraviolet absorber. The intensity and sponsored by the AWS Safety and Health spectral distribution of transmitted Committee during the AWS 60th Annual infrared region. The spectral range of optical radiation through a given Meeting in Detroit, Michigan, April 2-6, the spectrophotometer was 210 to curtain depends, in some manner, on 1979. 2600 nanometers (nm), i.e., 8.3 x 10" such factors as optical density, thick­ to 102.4 x 10 " in.* C EUGENE MOSS is a Health Physicist, U.S. ness, and color. Public Health Service, National Institute for One of the main advantages of using Occupational Safety and Health, Cincin­ *7 nanometer (nm) = I Xo 70 " m = 39.4 X such curtains is the increased visual nati, Ohio. 10 " in. = 70 angstroms (A).

WELDING RESEARCH SUPPLEMENT I 69-s does not vary too much, since the Table 1-Data on Tested Curtain Samples intensity of transmitted radiation is

1 dependent on thickness. To check Thickness""" how well the manufacturer controlled Sample Color cm ,ge in years the thickness of the curtains, transmis­ A Yellow 0.0351 > 1 sion measurements were compared B Green 0.0356 > 1 among samples cut from separate C Gray 0.0328 > 1 sheets of the same curtain type. As part D Yellow 0.0361 > 1 of this test a measurement was made E Green 0.0328 > 1 to obtain the thickness of each sample F Yellow 0.0335 > 1 with a micrometer. The average thick­ G Yellow 0.0368 < 1 H Green 0.0312 < 1 ness is shown in Table 1. Gray 0.0297 < 1 Blue 0.0312 <1 Transformation Technique K Green 0.0305 < 1 L Yellow 0.0307 <1 A program was developed for a M Gray 0.0328 <1 Hewlett-Packard Model 9830A com­ N Green 0.0307 < 1 puter to transform transmission values O Yellow 0.0358 <1 from the three separate optical radia­ P Green 0.0300 < 1 tion regions into one continuous curve Q Yellow 0.0353 <1 on semi-logarithmic graph paper. This R Yellow 0.0310 > 1 format permitted better comparisons S Blue 0.0297 > 1 among the various samples. T Yellow 0.0328 < 1 W Gray 0.0371 > 1 X Green 0.0351 <1 Results Yellow 0.0310 >1 z A typical sample spectral transmis­ '•'Thickness given is average of three measurements sion plot as obtained from the spectro­ '"2.54 cm = 1 in. photometer is shown in Fig. 1. Note After the recommended warm-up letter, placed into an envelope until that the three optical regions were period, the spectrophotometer was the order of measurement was deter­ divided as follows: ultraviolet 210-360 calibrated at the lowest wavelength mined by a random number process, nm, visible 340-700 nm, and infrared value in each of three optical radiation and measured. Each of the samples 600-2600 nm. regions at both the zero and 100% was positioned at the entrance slit of The original data, while informative, transmission level. The instrument was the detector chamber. Care was taken have several disadvantages in their set for 100% transmission and exam­ during the tests to position all samples existing format. First, curves shown in ined for departure from 100% across over the cell window in the exact same this manner are difficult to read with the entire spectral range of all regions. manner by the use of adhesive tape samples that have unusual transmis­ A zero-check was made at the end of strips. Cloves and tongs were used to sion properties. Second, comparisons each region to determine the magni­ minimize contamination. A plot of with samples having different spectral tude and direction of any drift correc­ percent transmission vs. wavelength transmission values are laborious. Fi­ tions. was made on every sample. nally, the curves do not follow the A rectangular-shaped segment, ap­ standard division of optical radiation proximately 25 X 50 mm (1 X 2 in.) regions which are ultraviolet, less than Reproducibility was cut from the center portion of 400 nm, visible 400-760 nm, and every sample, assigned an arbitrary It is important that curtain thickness infrared, greater than 760 nm. These

z o CO s z <

z Transmission Range Optical Region LU 0-7% Ultraviolet (210-360 nm) U 0-700% Visible (340-700 nm) 0-700% Infrared (600-2600 nm)

NANOMETERS Fig. 1—Typical spectral transmission curve of welding curtain obtained for ultraviolet radiation is 1% and for visible and infrared radiation is 100%

70-s|MARCH 1979 IOOO 1300 IOOO 1300 1600 1900 2200 NANOMETERS NANOMETERS Fig. 2—Typical modified spectral transmission plot Fig. 3—Composite of green curtains (samples B, E, H, K, N, P and X)

difficulties can be averted by using the previous mentioned transformation sion tendency in this area while the during the time of study. Therefore, format technique. When this format is yellow and blue samples demonstrate one should not assume only aging used, a composite curve such as in Fig. a very sudden and distinct transmis­ caused the effect observed. Note that 2 results. sion change around 500 and 800 nm, with the same curtain, sample trans­ There is a small error in the compos­ respectively. mission values appear to decrease as ite curves due to the difference in 4. All curtain samples transmit in­ the contaminant level increases—Fig. sensitivity of the two sequentially used frared radiation. For example, the blue 10. spectrophotometer detectors. The er­ and yellow samples will transmit A theoretical relationship exists be­ ror associated with this factor is about about 90% in the infrared (IR) region tween spectral transmission levels and 5% and occurs at the junction of the while the gray samples transmit the thickness of the curtains provided that visible-infrared region. least at 20%. the curtain tint and its concentration In order to determine the degree of Samples of the same color and are the same. This relationship is spectral transmission uniformity in the approximately equivalent thickness, shown as equation (1): manufacturing process, a comparison but differing in age, generally show I (\) = \e(\)e-*«>* (1) was made among all curtain samples variation of transmission levels over of the same color. Figures 3-6 show the measured optical radiation region In this equation, which is often sample data according to color. Sev­ of less than 5% (Figs. 6, 7, and 8). referred to as Lambert's Law of eral characteristics are apparent from However, Fig. 9 suggests that aging Absorption, l„ is the optical radiation these figures: factors may be quite important for incident upon the welding curtain, I is 1. Except for the yellow curtain yellow curtains in the ultraviolet (UV) the optical radiation transmitted samples there appears to be relatively region. Figure 9 also shows the older through the curtain, x is the curtain's good consistency in transmission val­ samples permit more transmission thickness, and k is the absorption coef­ ues, i.e., less than 10% variation over near 400 nm. In fact, four of the five ficient. the measured optical region. The new yellow samples, from different Equation (1) can be rewritten as: yellow samples appear to give consis­ manufacturers, have a high absorption tent values, except in the UV region. level of around 400-500 nm, whereas log, -kx (2) 2. All samples experience a drop in none of the older samples show this (T) transmission around 1700 nm. This characteristic. particular decrease in transmission is Since none of these curtain samples A comparison between two curtains probably due to curtain composition. are older than one year, this particular (A and B), identical except for thick­ 3. The shape of the transmission apparent aging effect could be of ness, can be made by taking the ratio R curve undergoes drastic change be­ major importance. It must be noted, of equation (2) for the two curtains: tween the visible and IR regions. For however, that there is no assurance example, the green and gray samples that sample composition or manufac­ R = show a gradual transmission progres­ turing methods had not changed (,)^-(T)

WELDING RESEARCH SUPPLEMENT I 71-s io V

-

z o / /

- fjt z LU u \

lllllll 100 400 700 1000 1300 1600 1900 NANOMETERS Fig. 4—Composite of yellow curtains (samples A, D, F, G, L, O, Q, T Fig. 5—Composite of gray curtains (samples C, L, M and W) and Z)

7 IOO 400 00 IOOO 1300 1600 1900 2200 700 IOOO 1300 I60C 1900 2200 NANOMETERS NANOMETERS Fig. 6—Composite of blue curtains, j is the old curtain and S is the Fig. 7—Comparison of old and new gray curtains (samples C new curtain. and M)

72-s | MARCH 1979 old

io IOOO 1300 00 400 700 IOOO 1300 1600 1900 2200

NANOMETERS NANOMETERS Fig. 8—Comparison of old and new green curtains (samples B Fig. 9—Comparison of old and new yellow curtains (samples S and X) and Q)

(3) Figure 11 suggests the critical wave­ plates, it is informative to note the lengths that cause ocular and skin knXB results (see Table 2). For example, the biological effects, but does not give comparison indicates that certain Under the assumption that k = k a b information on the energy necessary transparent curtains exceed the ab­ equation (3) becomes: to produce these effects. Not shown is sorption properties required of a shade R =^± (4) the American Conference of Govern­ 2 lens. It should be noted that all mental Industrial Hygienists recom­ curtain transmission values in the IR 2 This means that transmittance is a mended exposure level of 1 mW/cm region exceed the values given in ANSI function of thickness only. If, howev­ for the wavelength region between Z87.1. Also, note that only the green er, equation (4) is not valid, then the 320 and 400 nm. and yellow curtain transmittances ex­ difference in transmittance has to be From a protection viewpoint, the ceed the UV value listed for shade 2. In explained by more than consideration best curtain would minimize or elimi­ particular, 5 of 11 yellow samples and 3 of thickness. To confirm this observa­ nate hazardous wavelengths (i.e., of 7 green samples do not meet shade tion, transmission values from curtains shaded area). In almost every curtain 2 criteria. Of interest here is that 7 of having the smallest and largest thick­ tested, wavelengths less than about the 8 that failed were new samples. In nesses of each color were compared. 340 nm were greatly suppressed. This no way should this comparison The results indicate that within experi­ fact is very important, since it tends to remotely suggest that it is appropriate mental errors differences in transmit­ suggest that most transparent curtains for the transparent welding curtains to tance could be explained by thick­ give adequate protection in the UV be used as a replacement for welding ness. region. However, most curtains also safety lens. transmit at wavelengths beyond 400 A summary spectral transmission nm which is in the shaded area. There­ Discussion plot depicting various curtain samples fore, as far as overall protection from is shown in Fig. 12. Since most samples In determining the degree of protec­ ocular and skin hazards is concerned tested demonstrated little transmis­ tion offered by transparent curtains, it (i.e., minimizing the shaded area), the sion difference between old and new is necessary to define the spectral grey curtain would appear the best curtains, then Fig. 12 was developed zones that can produce potentially choice. However, if visibility of the using results from only new samples. hazardous optical radiation biological workplace is paramount, then one Figure 13 shows results provided by effects. These zones are shown as the might consider the yellow curtain. one of the manufacturers. Note that shaded area in Fig. 11. Figure 11 was The transmittance results can be Figs. 12 and 13 generally agree with developed by combining data from compared with the ANSI Z87.1 stan­ each other except for the blue curtain 9 10 recent publications. - The shaded dard for welding filter plates." Al­ transmission values. In this case the area indicates the relative spectral though this comparison is not neces­ blue curtain values are smaller in the hazards from broadband non-coher­ sarily valid since transparent plastic 500-to-800 nm region than those ent sources, such as a welding process. curtains can not be considered filter reported by the manufacturer.

WELDING RESEARCH SUPPLEMENT I 73-s . after ' before

aOO 500 400 500 600 TOO NANOMETERS Fig. 11-Envelope (shaded area) of potential hazardous broadband non-coherent optical radiation

The results shown in Fig. 12 could be Fig. 10—Comparison of yellow curtain of vital safety importance in selecting a (sample R) before and after cleaning transparent curtain for use in other than welding applications (i e , shield­ ing a UV curing system) However, since these curtains are intended for use in the welding environment, the spectral distribution of optical radia­ tion from various welding processes transmitted through different curtains is of occupational concern. Using current published welding data'2 and the spectral transmission values in this paper, it is possible to estimate the approximate distribution of optical radiation transmitted through trans­ parent curtains. Figure 14 shows the actual and approximate optical radiation spectral IOO 400 700 IOOO 1300 1600 1900 2200 irradiance values from a typical weld­ NANOMETERS ing process. Figure 15 shows the product of the approximate irradiance and transmission values; it was devel­ Table 2 —Transmittance Test Results of Weld ng Curtains Compared with Values Given in ANSI Z87.1 oped using the same curtain samples discussed in Fig. 12, with the exception of the yellow curtain. Since this paper Spectral transmittance Infrared finds that older yellow curtains trans­ at selected wavelength, % transmitt ance mit more than new curtains, only older Sample color 313 nm 334 nm 365 nm 405 nm > 760 n n, % yellow curtain transmission values are Shade 1.5 lens 0.2 0.8 25. 65. 25 shown in Fig. 15. Shade 2.0 lens 0.2 0.5 14. 35. 15 A Curtain yellow 0.03 0.04 10. 23. 86.5" B Curtain green 0.01 0.02 O.I 2. 43." Conclusion C Curtain gray 0 0 0 3. 28." With this particular welding process, 7. 78.'" D Curtain yellow 0.03 0.05 5. both old and new yellow curtains E Curtain green 0 0 0.1 1.6 43.'" transmit more optical radiation energy F Curtain yellow 0.35 0.7'"' 16.'" 12. 79." G Curtain yellow 0 0 10. 22. 94." than other curtains. H Curtain green 0.04 0.06 0.5 2.0 48." While the magnitude of optical I Curtain gray 0 0 0 2.0 23." energy transmitted by a yellow curtain I Curtain blue 0 0 0 0.1 80.'" is small, the spectral distribution K Curtain green 0.27'" 0.42 1.0 3. 51." contains a larger component of haz­ L Curtain yellow 0.25"" 6.'"' 3. 0 89." ardous wavelengths than other cur­ M Curtain gray 0 0 0 1. 20." tains. Obviously, additional factors N Curtain green 0.34"" 0.54"" 0.9 2.1 41."' such as distance from welding arcs O Curtain yellow 0.23"" 2.0 0 81." 5 ,., shielding mediums, welding pro­ P Curtain green 0.33"" 0.55"" 1.0 2.3 42." cesses, etc., have to be considered Q Curtain yellow 0.22"" 6.0'" 2.5 0 82." before one could recommend a partic­ Rr Curtain yellow 0 0 0 0 78." R„ Curtain yellow 0 0 0 0 89.'" ular curtain for a certain welding S,, Curtain blue 0 0 0 0 68.'" event. However, evaluating welding

S;, Curtain blue 0 0 0 0 82." curtains according to welding pro­ T Curtain yellow 0.17 3.6"" 2.0 0 86."' cesses seems more prudent and proper W Curtain gray 0.01 0.01 0.04 3.0 51." rather than allowing workers/manage­ X Curtain green 0 0 0 2.0 20." ment to select one merely on the basis Z Curtain yellow 0.02 0.03 7 14. 86."' of eye appeal. In the near future a more comprehensive report will be "Indicate values greater than those for Shade 2.

74-s I MARCH 1979 ACTUAL "DATA

400 480 560 NANOMETERS Fig. 14—Absolute spectral irradiance at 1 m for GTAW welding process, on steel, 275 A, 'Ar, In. (7.6 mm) arc length, helium shielding gas

1000 1300 1600 1900 NANOMETERS Fig. 12—Spectral transmission measurements found in this paper

400 500 WAVELENGH, Fig. 15-Approximate spectral distribution ot optical radiation (shaded areas) transmitted through various transparent welding curtains from GTA welding process

600 800 IOOO I I400 I600 WAVELENGTH(NANOMETERS) 9. American Conference of Governmen­ Fig. 13—Spectral transmission measurements reported by manufac­ tal Industrial Hygienists, 1977, Threshold turer (modified) Limit Values for Chemical Substances and Physical Agents in the Workroom Environ­ ment, P.O. Box 1937, Cincinnati, Ohio 45201. published on this type of evaluation. It 3. El Gammal, M. Y., Soliman, A. M. and 10. Pitts, D. O, Cullen, A. P., and Hacker, must be kept in mind that other Mostafa, M. S., "Actinic Conjunctivitis—The P. D., "Ocular Ultraviolet Effects from 295 nm to 400 nm in the Rabbit Eye," DHEW curtain selection criteria, such as being Effect of UV and IR Radiation on the Eyes of Welders and Glassblowers," Bull Ophth. (NIOSH) Publ. No. 77-175, 1977. National able to see through the curtain, may Soc. Egypt, 66/70:41, 1973. Institute for Occupational Safety and have to be considered. 4. Russ, D. S., "The Short-Term Effects on Health, Cincinnati, Ohio 45226. While this paper has dealt with Health of Manual Arc Welding," I. Soc. 11. American National Standards Insti­ transparent curtains, it should be Occup. Med., 23/3:92, 1973. tute (ANSI) 1968, American National Stan­ mentioned that much more work 5. Hanene, E., and Gutschmidt, E., dard Practice for Occupational and Educa­ needs to be done with reflective char­ "Squamous Cell Carcinoma in an Electric tional Eye and Face Protection, ANSI Z87.1, acteristics of opaque curtains. Welder," Berufsdermatosen, 24/5:119, 1430 Broadway, New York, New York 1976. 10018. 6. Filipiakow, Z., "The Influence of 12. Marshall, W. |., et al., Nonionizing Reterences Welding Arc Radiation on the Picture of the Radiation Protection Special Study No. 42- 1. Ruprecht, K. W., "Foveo-Maculopathy Eye Fundus," Klin. Oczna, 40/4:529, 1970. 0312-77, Evaluation of the Potential Retinal Resulting from Arc Welding," Zentralblat 7. Goldman, H., "Genesis of Heat Cata­ Hazards from Optical Radiation Generated Arbeitsmid, 26:200, 1976. ract," Arch. Ophth., 9:314, 1933. by Electric Welding and Cutting Arcs, 1977, 2. Clark, B. A. )., "Welding Filters and 8. "What's Happening with Welding USA Environmental Hygiene Agency, Aber­ Thermal Damage to the Retina," Australian Curtains," Welding Design and Fabrication, deen Proving Ground, MD (NTIS No. ADA j. Optom., 51:91, 1968. |une:98, 1978. 043023).

WELDING RESEARCH SUPPLEMENT I 75-s