The use of areal surface topography analysis for the inspection of micro-fabricated thin foil laser targets for ion acceleration

Abstract

This paper proposes a novel approach to the characterization of critical dimensions and geometric form error at micro and sub-micrometric scales, suitable for application to micro-fabricated parts and devices. Thin foil laser targets for ion acceleration experiments are selected as the test subject in this instance. The approach is based on acquiring areal maps with a high-precision 3D optical interferometric profilometer and on processing the surface topography data with novel techniques obtained by merging knowledge and algorithms from surface metrology, dimensional metrology and computer vision/image processing for quality inspection. This new approach allows quantitative measurement results for the thin foil laser target to be obtained, and shows promise for being applied to a wide array of similar problems involving quality inspection of micro-parts and devices, and to structured surfaces in general.

Feature-based characterisation of signature topography in laser powder bed fusion of metals

“…Data pre-processing consists of levelling, removal of non-measured points (voids) and removal of spike-like measurement artefacts. Levelling is implemented via leastsquares mean plane subtraction; voids are removed by replacement with weighted interpolation of valid neighbours [12]; and spike-like measurement artefacts are identified as local outliers and removed by interpolation of neighbours [25]. In Figure 5, an extracted region taken from the first test dataset is shown, highlighting the identification of three spatter formations.…”

Review of feature boundary identification techniques for the characterization of tessellated surfaces

“…Notable examples include: linear interpolation [? ], median interpolation [11,12], splines [??] and kriging [13].…”

“…The typical solution to this problem is to remove form and long wavelength components from the topography, which can be done by subtracting a smoothed version of the original surface from the original surface itself, the residual containing only the higher spatial frequencies. The smoothed topography can be obtained by polynomial fitting on the original, although this technique does not work well in presence of steps and other sharp discontinuities, or by using a moving median filter [11,12]. Another recent approach by Le Goic et al is to use discrete modal decomposition [15].…”

“…They are most effective when used to detect sharp edges, such as steps, and have been used extensively in the field of computer vision [38,39]. Senin, Blunt, and Tolley [12] have recently devised a technique for the identification of thin foil laser targets for ion beam acceleration experiments, which at one step uses a height thresholding operation to discriminate between the target and background surfaces, as shown in figure 4. For tessellated surfaces, Kong et al [34] have used a height threshold to identify the lenses and separate them from the background in a micro-lens array.…”

“…Due to texture parameters being calculated on each pixel using information from a given amount of neighbouring pixels, a sharp transition in the topography (e.g. a step) may produce various results in terms of the transition of the local texture parameter, since at each position the Figure 4: Height thresholds can be very effective when the surface contains step-like features [12]. Figure 5: When the surface contains shallow features with not sharp edges height thresholding often produces many false features [9].…”