Aesthetic laser marking assessment

Aesthetic laser marking assessment

Optics & Laser Technology 32 (2000) 187±191 Aesthetic laser marking assessment T.W. Ng a,*, S.C. Yeo b a Faculty ...

409KB Sizes 0 Downloads 4 Views

Optics & Laser Technology 32 (2000) 187±191

Aesthetic laser marking assessment T.W. Ng a,*, S.C. Yeo b a

Faculty of Engineering, Engineering Block EA-07-32, National University of Singapore, 9 Engineering Drive 1, Singapore 117576 b Philips Singapore Pte. Ltd., Engineering Department, 259 Jalan Ahmad Ibrahim, Singapore 629148 Received 15 November 1999; accepted 28 March 2000

Abstract In this paper, the CIE color di€erence formula was applied to evaluate four types of material surfaces; anodized aluminium, stainless steel, poly-butylene tetra-phthalate (PBT), and phenol formaldehyde, marked using a Nd:YAG laser, and viewed under three common modes of illumination; tungsten, ¯uorescent and daylight. The color di€erence values were based on the spectral re¯ectance readings obtained from a spectrophotometer. Each material exhibited di€erent color di€erence trends in relation to marking speed for the di€erent modes of illumination. Nevertheless, general comparisons could be made in terms of operational marking speeds and the maximal color di€erence values for each material. 7 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction Laser marking provides a unique combination of speed, permanence and versatility. It can generate considerable savings in reduced manufacturing and tooling costs, elimination of secondary processes and consumable material, and reduced inventory and maintenance downtime. The technique has been used extensively in the production of indelible and legible alpha-numeric characters and logos for product identi®cation and traceability in the semiconductor industry [1±3]. Recently, there has been interest in using laser marking to improve the aesthetic appearance of marked images, thereby increasing a product's perceived value [4,5]. Laser marking is essentially a thermal process that employs a high intensity beam of focused laser light to create a contrasting mark on the material surface. Beam-steered marking employs mirrors mounted on high-speed, computer-controlled galvanometers to direct the laser beam across the target surface (Fig. 1). Each galvanometer provides one axis of beam motion in the marking ®eld. A multi-element ¯at-®eld lens assembly subsequently focuses the laser light to achieve high power density on the work surface while main* Corresponding author. Fax: +65-777-3525.

taining the spot travel on a surface. As the target material absorbs the laser light, the surface temperature increases to induce a color change in the material. Currently, laser marking systems in industry employ either carbon dioxide (CO2) or neodymium-doped, yittrium-aluminium-garnet (Nd:YAG) lasers. Previous work have indicated that Nd:YAG lasers produce superior mark legibility on surfaces such as ceramics and plastics [5]. In semiconductor marking, quality is assessed by comparing legibility characteristics such as contrast, width, depth and micro-cracks. These characteristics are usually obtained using techniques such as optical microscopy, ultrasonic microscopy, electron microscopy, surface roughness measurements and contrast evaluation devices. In aesthetic marking, however, the colorimetric dissimilarity between the marked and unmarked surface is the crucial factor that determines acceptance or rejection. In this work, we examine the level of laser marking using the CIE color di€erence formula. This formula is based on spectrophometric measurements derived from marked and unmarked surfaces. The contribution of di€erent illumination conditions was also included in the analysis.

0030-3992/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 3 0 - 3 9 9 2 ( 0 0 ) 0 0 0 4 0 - 2


T.W. Ng, S.C. Yeo / Optics & Laser Technology 32 (2000) 187±191

photopic vision. There are three types of cones in the human retina that are sensitive at three di€erent wavelength bands. Although the rods provide a spectrally di€erent receptor than the cones, there is overwhelming evidence [6] that when levels of illumination are high enough, color vision is basically a function of the cones alone. The Commission Internationale de l'Eclairage (CIE) [7] has devised a standard of tristimulus values X, Y and Z to describe color matching. These values can be determined using the color matching functions, x-, y- and z-, which are variables dependent on the wavelength as shown in Fig. 3, via the following relations: X ˆ K…R1 S1 x 1 ‡ R2 S2 x 2 ‡ R3 S3 x 3 ‡ . . .† Y ˆ K…R1 S1 y 1 ‡ R2 S2 y 2 ‡ R3 S3 y 3 ‡ . . .† Fig. 1. Optical beam steered laser marking system using computercontrolled galvanometers.

2. Color vision basics A diagrammatic representation of a human-eye cross section is shown in Fig. 2. Most of the optical power is provided by the curved surface of the cornea. The cornea and lens acting together form a small inverted image of the outside world on the retina. The two types of light receptors present in the retina are the rods and cones. The function of the rods in the retina is to give monochromatic vision under relatively low levels of illumination. Viewing under such conditions is generally known as scotopic vision. Scotopic vision occurs via the absorption of light in a photosensitive pigment called rhodopsin. The function of the cones in the retina, alternatively, is to give color vision at normal levels of illumination. This form of viewing is called

Fig. 2. Cross-sectional diagram of the human eye.

Z ˆ K…R1 S1 z1 ‡ R2 S2 z2 ‡ R3 S3 z3 ‡ . . .†


where K is a constant, R the spectral re¯ectance factors, and S the relative spectral distribution. The value of S is dependent on the light source used. In the adaptation process of the visual system, colored objects undergo appreciable changes in color appearance as the light source is changed. To provide some form of uniformity, the CIE has introduced standardization to distinguish between illuminants [8]. Three of the common light sources are tungsten ®lament lamps, ¯uorescent lamps and daylight. The spectral power distributions for them, as de®ned by in the CIE standard, are shown in Fig. 4. 3. CIE color di€erence formula Colors di€er in both chromaticity and luminance. Hence, some method of combining these variables is required. One of the popular color spaces is the CIELUV color space [9], which de®nes

Fig. 3. The CIE color matching functions for a standard colorimetric observer.

T.W. Ng, S.C. Yeo / Optics & Laser Technology 32 (2000) 187±191


Fig. 4. The representative relative spectral power distributions of tungsten, daylight and ¯uorescent lamps.

L ˆ 116…Y=Yn †1=3 ÿ 16 u ˆ 13L …u 0 ÿ un0 † v ˆ 13L …v 0 ÿ vn0 †


where un0 , Yn and vn0 are the values of u ', Y and v' for the appropriate chosen reference. The values for u ' and v ' can be found from the tristimulus values using

Nd:YAG laser. A standard line matrix was marked on the plates for measurement. The lines were marked at di€erent speeds to produce a variation in marking quality. The laser marking speed has a direct relation to the laser marking time for a standard power setting. The spectrophotometer used for color measurement was a Perkin Elmer Lamba 10 model. The range of measurement of the spectrophotometer was from 400± 700 nm.

u 0 ˆ 4X=…X ‡ 15Y ‡ 3Z † 5. Results and discussion 0

v ˆ 9Y=…X ‡ 15Y ‡ 3Z †


If the di€erences of color on two surfaces are desired, the color di€erence is found using DE uv ˆ ‰…DL † 2 ‡ Du † 2 ‡ …Dv † 2 Š1=2


From the known spectrophometric signature, Eqs. (2)± (4) then allow for a quantitative value to be obtained to compare the colorimetric visibility of any lasermarked portion of surface with its adjacent unmarked section.

The CIE color di€erence values were calculated and plotted at various marking speeds for the di€erent materials used as well as under the three illumination modes mentioned. Fig. 5 describes the case for anodized aluminium. The curves for the di€erent illuminants were roughly of the same magnitude but tended to exhibit a wavy trend. Hence, it was dicult to ascertain the presence of particular optimal marking

4. Experimental Four sample plates of di€erent material were used in the investigation of this work: (a) anodized aluminium; (b) stainless steel; (c) poly-butylene tetra-phthalate (PBT); and (d) phenol formaldehyde. For (a), the aluminium substrate was immersed in an electrolyte consisting of a mixture of weak acids. This resulted in the formation of an electrically insulating ®lm of aluminium oxide. For (b), the stainless steel sample used was SS304 with low carbon content. Samples (c) and (d) were both plastics with PBT being a thermoplastic and phenol formaldehyde being a thermoset. All sample plates were marked using a 1.5 W output

Fig. 5. Color di€erence values obtained at di€erent marking speeds and types of illumination for anodized aluminium.


T.W. Ng, S.C. Yeo / Optics & Laser Technology 32 (2000) 187±191

speeds wherein the color di€erences were maximal. In the case of daylight and tungsten illumination, there appeared to be minor reducing color di€erence trends in relation to marking speed. With ¯uorescent illumination, this trend did not appear to exist. For anodized aluminium, marking speeds around 30 m/s appeared to be normal. Fig. 6 shows the results obtained for stainless steel. In the case of daylight and ¯uorescent illumination, an increasing color di€erence with marking speed trend was uncovered. While this might appear to be favorable in a sense that faster marking throughputs could be implemented, there was a possibility that marking speeds beyond 10 m/s might produce uneven lines. Hence, the results should not be extrapolated beyond the range described. In the case of tungsten illumination, there appeared to be a minor dip in color di€erence at marking speeds around 5 m/s. Regardless of the type of illumination, marking at speeds of around 5 m/s appeared to be conventional for stainless steel. The case for PBT is given in Fig. 7. The color di€erence values appeared to increase with marking speed until a plateau was attained at a marking speed of around 200 m/s under tungsten and ¯uorescent illumination. The curve for daylight seemed to follow the same trend; except that a higher marking speed (around 320 m/s) was needed to reach a similar plateau value. Hence, with the exception of marking speeds exceeding 300 m/s, color di€erence was more distinct when viewed under the arti®cial illuminants (tungsten and ¯uorescent). The other noteworthy observation was that marking speeds of 200 m/s and above should be used to obtain optimal visibility markings under tungsten and ¯uorescent illumination. In the case of daylight illumination, marking speeds above 300 m/s was ideal. Fig. 8 shows the results for phenol formaldehyde.

Fig. 6. Color di€erence values obtained at di€erent marking speeds and types of illumination for stainless steel.

Fig. 7. Color di€erence values obtained at di€erent marking speeds and types of illumination for PBT.

The color di€erence values increased with marking speeds and reached a plateau at a marking speed of 200 m/s under daylight illumination. With tungsten illumination, a peak di€erence value approximately equal to the case of daylight was achieved at a marking speed of around 100 m/s. A similar trend was obtained for the case of ¯uorescent illumination. Nevertheless, the color di€erence values were markedly lower. This interesting result indicated that laser markings at any speed were more distinct when viewed under tungsten as compared to ¯uorescent illumination. Another interesting outcome was that marking at speeds below 200 m/s tended to produce relatively low visibility markings when viewed under daylight illumination. In a manufacturing environment, marking speed is an important issue as higher marking speeds generally mean higher productive output. Based on the results

Fig. 8. Color di€erence values obtained at di€erent marking speeds and types of illumination for phenol formaldehyde.

T.W. Ng, S.C. Yeo / Optics & Laser Technology 32 (2000) 187±191


Fig. 9. Comparative maximal color di€erence values for the di€erent types of marked surfaces and illumination modes.

found here, marking speeds could be generally classi®ed as low for stainless steel (5 m/s), medium for anodized aluminium (30 m/s), and high for PBT and phenol formaldehyde (300 m/s). A graphical comparison of the maximal color di€erence values computed for the various materials and illumination modes is depicted in Fig. 9. It could be seen that the variation in the color di€erence values was mainly minimal when di€erent illumination modes were used. In general, these values could be classi®ed as low for anodized aluminium (2±3 units), medium for PBT and stainless steel (25±60 units), and high for phenol formaldehyde (85±190 units). High color di€erence values typically indicate higher visibility markings.

6. Conclusions In this work, the CIE color di€erence formula was used to assess laser markings. Four types of materials marked at di€erent speeds and observed under three di€erent illumination modes were analyzed. The plots of color di€erence value against laser marking speed revealed di€erent trends for the materials and illumina-

tion modes applied. In terms of operational laser marking speeds, stainless steel could be classi®ed as low, anodized aluminium as medium, and PBT and phenol formaldehyde as high. In terms of maximal color di€erence values, anodized aluminium could be classi®ed as low, PBT and stainless steel as medium, and phenol formaldehyde as high. References [1] Holizinger G, Kosanke K, Menz M. Printing of part numbers using a high powered laser beam. Optics and Laser Technology 1973;5(6):256±65. [2] Healy Jr LH. Laser marking in the electronics industry. In: Proceedings of the IEPS, 1984. p. 392±8. [3] Pecht M, editor. Handbook of electronic packaging design. New York: Marcel Dekker, 1991. [4] Alexander DR. Laser marking using organo-metallic ®lms. Optics and Lasers in Engineering 1996;25:55±70. [5] Stevenson RL. Beam steered laser marking of plastics. Plastics Engineering 1997;February:23±5. [6] Hunt RWG. Measuring colour. New York: Ellis Horwood, 1991. [7] CIE Standard on Colorimetric Observers, CIE S002, 1986. [8] CIE Standard on Colorimetric Illuminants, CIE S001, 1986. [9] CIE Supplement No. 2 to Publication No. 15, Recommendations on uniform colour spaces, colour di€erence equations, and psychometric terms, 1978.