Three-dimensional non-destructive optical evaluation of laser-processing performance using optical coherence tomography

Abstract

We demonstrate the use of optical coherence tomography (OCT) as a non-destructive diagnostic tool for evaluating laser-processing performance by imaging the features of a pit and a rim. A pit formed on a material at different laser-processing conditions is imaged using both a conventional scanning electron microscope (SEM) and OCT. Then using corresponding images, the geometrical characteristics of the pit are analyzed and compared. From the results, we could verify the feasibility and the potential of the application of OCT to the monitoring of the laser-processing performance.

Introduction

The development of ultra-short pulse lasers has increased the fundamental understanding of interaction behaviors between a laser beam and a material, and the advance of a high-power laser technology has facilitated technical developments in laser-based material processing. With the development of high-powered and short-pulsed lasers, conventional material processing typically done using mechanical tools has instead been performed by commercial-grade lasers such as CO₂ and Nd:YAG lasers, providing advantages of cost-effectiveness, high precision and non-contact processing [1], [2].

Though the laser-based material processing has the potential to present a high degree of freedom in the control of fabrication processes, unwanted sample deformations commonly occur during various laser processing such as engraving, marking, cutting, bending, and welding due to the improper usage of laser irradiation conditions. To prevent quality degradation of the laser processing, many researchers have focused on the development of diagnostic techniques by employing optical, acoustic, thermal, and computational methods. For instance, an acoustic wave may be used as a means of detecting the condition of a weld in a workpiece [3]; the acoustic wave originates from the shockwave created when a plume is rapidly expanded during the welding process. A pyrometer, an infrared radiation sensor, is used to monitor surface temperature variation in laser brazing [4]. In addition, laser-induced ultrasonic waves show great potential for use in the examination of welding conditions [5]; the application of a voltage or a current Hall effect transducer is also available for monitoring welding conditions [6]. Other, more commonly exploited monitoring methods include optical and visible approaches.

Optical inspection methods for laser processing have generally utilized a photodiode or photodiode array with light sources ranging from UV to IR [7]. In many cases, the plasma caused by a highly focused laser beam is measured using optical inspection methods, and the information of sputter and bead shape could be obtained from this measurement [8]. The spectral response to a different colored light source, which is dependent on the welding condition, has been measured and analyzed to determine the correlation between a processing mechanism and an optical signal variation. In addition, spectrally analyzed information obtained using laser spectroscopy [9], [10] such as laser-induced breakdown spectroscopy and laser-induced fluorescence spectroscopy has been used to monitor the quality of a laser-based cleaning processing.

Compared with the above approaches, visual methods utilizing a CMOS or CCD camera can acquire more comprehensive information regarding surface deformations [11]. Although direct observations using an optical microscope or other visual methods can be used to view a three-dimensional profile of a laser-processing region, it is difficult to analyze the two-dimensional cross-sectional or three-dimensional structural information of deeply engraved samples because of limitations in observable depth. Though the prior methods could be applied to investigate laser processing, a more precise evaluation of laser-based processing performance is required before a cross-sectional view of physically bisected workpiece can be observed; to this end an optical microscope or a scanning electron microscope (SEM) is suggested [12]. However, a physically bisected workpiece would probably reveal distorted results based on the fact that it is likely that the workpiece would be damaged or spoiled during the cutting process. Another disadvantage is that this method could not be applied to the real-time monitoring of the laser-based processing and could require a time-consuming preparation process.

As a promising optical imaging modality, optical coherence tomography (OCT) has the potential to provide a high-resolution non-destructive cross-sectional image by using a white-light interferometry scheme [13], [14], [15]. The use of a white-light or broadband source provides a high axial spatial resolution of up to submicrons, achieved by coherence gating. The exploitation of interference effects makes it possible have a measurement range of up to several millimeters along the depth of a sample. Using an optical scanner over the sample, OCT could reconstruct the cross-sectional image of the sample without any incision. When the detailed tomography of a sample is obtained, OCT does not require any physical contact or destruction of the sample. Until now, most interest in OCT research has been focused on the in vivo imaging of biomedical samples of retinas [16], skin [17], and blood vessels [18]. In spite of fewer applications in material characterization, as compared with OCT use in biomedical fields, the application of OCT for the non-destructive imaging and examination of non-biomedical samples has continuously increased due to inherent advantages of OCT; the ability to view the resolvable anatomy of biological tissues at a micron-scale.

In many cases, the primary task of an OCT application is the imaging of structural information. For instance, results pertaining to the detection or examination of the subsurface cracks in the material of ceramics have been reported [19]. The imaging of the internal microstructures of polymer matrix [20], injection-molded plastic parts [21], and paper [22] has also been demonstrated. Besides conventional cross-sectional imaging, the functional imaging of a strain-mapping required for the stress analysis has also been presented using polarization-sensitive OCT [23].

In this paper, we present our results related to the use of OCT to monitor the performance of the laser processing of an acrylonitrile butadiene styrene (ABS) plastic. ABS plastic is currently widely used in numerous applications, but due to its high thermal sensitivity, the performance of laser processing on an ABS plastic sample can be significantly degraded when improper irradiation conditions are applied. For these reasons, the monitoring of the performance of laser processing is required as part of the investigation into how to optimize processing conditions.

In the investigation of performance variation in a sample, an ABS plastic having high sensitivity to laser operational conditions is selected as the most appropriate material. Initially, the ABS plastic sample used in the experiment has a plane surface with high reflectivity. Though various plastics have been imaged using an OCT system, it is rare to find OCT images of laser-processed plastics that were used for an evaluation of laser-processing performance. Previously, the cross-sectional images of the surface deformation formed on an glass and ABS plastic samples by laser processing were reported [24], [25]. Also, microscopic images of pits and rims on an ABS plastic created by a Q-switched Nd:YAG laser were compared with the corresponding OCT images, which were found to be well matched [25].

However, more detailed analysis of the laser processing on materials such as a change in geometrical shape, an increase in depth, and rim formation of a pit due to laser-processing conditions have not yet been performed. To appreciate the difference in the laser-processing performance for different processing conditions, we performed OCT imaging on a number of samples while changing operational parameters such as pulse energy and the repetition rate of a Q-switched Nd:YAG pulse laser; this laser is commonly used for laser processing in industrial and clinical fields [26]. By reconstructing a series of two-dimensional tomographies, we generated three-dimensional images of deformed samples on ABS samples deformed during laser processing. Through the presentation of these three-dimensional images, we could investigate the dependence of the laser processing on operational parameters such as single pulse energy, repetition rates, and number of pulses, factors that were used to explain why the laser-processing performance of identical materials may vary. Then, we compared the OCT images to SEM images.

Although in previous studies, two-dimensional images obtained with a time-domain OCT system are presented, in this paper we present three-dimensional image to enhance the availability of information integrated from the two-dimensional OCT images. And although a quantitative analysis is not dealt with in the paper, more plentiful information for evaluating laser-processing performance can be analyzed. Finally, contrary to other papers [22], [25], this paper employs laser processing using a pulse laser, which displays a significantly smaller pit and rim size.