ISO-Standardized Filtering for DektakXT Stylus Profilers

When various parameters of a surface - such as roughness, step heights, or depths - are measured by any metrology method, only a representation of the surface characteristics is obtained. Nevertheless, the potential of proper filtering for data analysis, carried out in accordance with established ISO standard techniques, can never be underestimated while aspiring to provide the most reproducible and highly accurate results for a measuring system.

Bruker has provided ISO compatibility of the two-dimensional (2D) profile ISO 4287 and 4288 standards to their versatile Vision64^® software, which is used to power the DektakXT^® StylusProfiler. This article discusses the setup and application of standardized filtering methods, as well as the implementation within the Vision64 software for a surface roughness application example on a silicon wafer.

Stylus Measurement as a Representation of Surface Profile

The image and 3D illustration of a machined metal surface depicted in Figure 1 represent the data collection process through stylus trace on a true profile, by carrying out filtering to produce different profiles of interest.

Figure 1. Pictorial representation of the effects of filters to raw (total) surface profile.

The first profile generated is known as “total profile” and is the representation of the surface. The “primary profile” is then produced by applying the short cutoff filter with a cutoff spatial frequency of λ_s to the total profile. The short cutoff rejects specific spatial frequencies. These frequencies are considered to be the stylus deformation and noise. It is to be noted that for this to happen, the radius of the stylus tip is anticipated to be smaller than the short cutoff frequency.

When the primary profile is produced, the roughness long cutoff filter with a cutoff spatial wavelength of λ_c is applied. Once the roughness cutoff is applied, the resulting rejected spatial information - the spatial frequency content with a wavelength less than the roughness cutoff - represents the “roughness profile.” The information that passes without being rejected represents the “waviness profile.” Filtration of the waviness profile can be performed one more time, at a longer cutoff wavelength to isolate the waviness as well as the form of sample being tested.

Here, it should be noted that the stylus trace analysis follows standard rules regarding the traversing and evaluation length, with definitions in relation to the cutoff spatial scales of interest. Figure 2 illustrates a general trace detailing the sampling and evaluation lengths.

Figure 2. Total profile with divisions into sampling, evaluation, and traversing lengths.

The “evaluation length” is defined as the length along the stylus scan direction utilized to assess a specific portion of the “traversing length.” The “sampling length” is either less than or equal to the evaluation length. In the case of the total and primary profiles, the sampling length is equal to the evaluation length. In the case of the roughness and waviness profiles, the sampling length is equal to the filter cutoff wavelength that is used to isolate the roughness from waviness (λ_c).

Surface Texture Parameters – Filter per ISO 4287, 4288 and ASME B46.1

For characterizing the surface profile, the surface texture parameters are calculated using the preferred filters. Hundreds of parameters have been defined for industrial use, of which many also appear in the standards. The surface texture parameters of the surface profiles can be divided into three categories:

• Height parameters that characterize the average, valley, and peak of the ordinates
• Spacing parameters that determine the spacing of the valleys and peaks of the surface profile
• Hybrid parameters that combine the spacing and height data obtained from the profile.

These parameters are differentiated by adding prefix letters, where P stands for the total or primary profile, R stands for the roughness profile, and W stands for the waviness profile. The calculation method earmarked by ISO 4287 for specific parameters is the computation of values over the entire evaluation length. However, for other parameters, the method is the computation of the values within a sampling length. The ISO 4288 standard and the ASME B46.1 standard modify this method of computation to include estimates of parameters that are calculated over a single sampling length versus the average value of parameters that are calculated over all available sampling lengths within the evaluation length. The way in which these filters and computations are applied depends on an easy method provided by the Vision64 software on the DektakXT platform.

Figure 3 shows a representation of the separation of a profile into different spacing segments, based on upper and lower limits.

Figure 3. Segments of a profile showing distance between peaks and adjacent valleys, as well as identification of additional peaks based on ISO 4288 standards.

Vision64 software (Figure 4) ensures that the default settings fully conform to the ISO recommended standards set for the measurement analysis. Figure 5 shows an example of 2D profiles scanning a silicon wafer for surface roughness.

Figure 4. Filtered roughness dialog settings in Vision64 software for DektakXT.

Figure 5. Example 2D profiles on a silicon wafer showing the four profiles (total, primary, waviness, and roughness) according to prescribed filter settings.

The specified cutoffs λ_c and λ_s are used to apply the digital filters, where λ_c and λ_s respectively denote the long and short wave cutoffs. The following two equations detail the mathematics behind modifying a particular signal component by using this filtering (Gaussian-filtered roughness and waviness). At λ = λ_c, λ_s, the amplitude of a specific component of the signal is attenuated by 50%.

Conclusion

The application of filters using Vision64 software is a straightforward task to measure various test articles on the DektakXT platform. The cutoff wavelengths defined by ISO or ASME can be implemented in a simple manner by checking a box. In addition, the software allows a user to define and use cutoffs by unchecking the standard cutoff box and typing into a dialog box for entry. To guarantee compliance with ISO specifications for surface 2D profilometry, Bruker offers both 45° cone angle tips and 60° cone angle tips for use on the DektakXT platform.

This information has been sourced, reviewed and adapted from materials provided by Bruker Nano Surfaces.

For more information on this source, please visit Bruker Nano Surfaces.