✕

The answer to whether a part meets specification is often a “yes” or “no” on a printed report, a “good” part versus “bad.” However, these numerical values—the information passing from quality to engineering and manufacturing—is often not enough information to solve quality issues or improve processes.

In recent years, more measurement information is becoming available to design and production personnel. For example, the rapid adoption of 3D scanning has enabled engineers to explore data point clouds at their desks. This technology lets design teams visualize shapes and errors of a “digital twin,” and to explore and share that information. The ability to interact with the data offers vastly more information than a static report of numerical measurement results.

When it comes to surface texture however, most decisions are still made based on the handful of texture parameters measured in the lab. Those parameters, unfortunately, offer only a very limited view of how the surface will actually function.

As performance requirements become more demanding in areas such as sealing, wear resistance, friction, noise, vibration, adhesion, etc., the “shape” of surface texture has become far more important than the numerical roughness values. Just as point cloud analysis enables engineers to explore and visualize “macro” part data, the ability to explore and interact with “micro” surface texture data can provide a wealth of information about surfaces and how they perform.

Numerical data only tells part of the story

Over the past several decades, surface roughness measurement and analysis have been greatly developed and refined. Despite that development, however, most roughness specifications are still based on a handful of basic parameters such as Average Roughness (Ra, or Sa). These parameters can be helpful for spotting changes in parts or processes, but they give only the most general information about surfaces.

Figure 1 shows three very different surfaces that have nearly identical Ra values, yet, we can see that they are considerably different and would likely perform differently as well. Ra cannot distinguish peaks from valleys, gives no information about the spacing or position of peaks or valleys, etc. All these characteristics greatly impact how a surface will function.

Figure 1. These surfaces have very different properties but almost identical average roughness (Ra) values. Numerical results are often insufficient to describe surface texture. Courtesy Digital Metrology Solutions

Seeing the data is more powerful than reading the numbers

For most of us, music notes on a page aren’t a “song” until we hear them played. The same is true for surface texture: parameters are just numbers until we see the surface and understand what they indicate.

The most powerful surface analysis tools are the human eye and the human brain. Thousands of times per day we look at surfaces and intuitively determine if they are rough, smooth, safe to walk on, dangerous to touch, nice to look at, etc.

Figure 2 shows presentations of surface texture data that are progressively richer in information. A numerical parameter value (top) provides a summary of the texture. A 2D profile begins to show the shape of the surface, such as peak heights and spacings. A 3D image is even more informative, showing directionality and revealing “areal” features such as pits, porosity, etc. Some software even enables 3D printing of texture data, a powerful tool for exploring and sharing surface data.

Figure 2. Presentations of surface texture with progressively greater levels of information.

|

Courtesy Digital Metrology Solutions

“Offline” data analysis

Seeing data (perhaps in a printed graph) adds to our understanding of a surface. The ability to interact with the data creates the opportunity to study and learn from it.

In most companies, the personnel responsible for designing and manufacturing surfaces are not necessarily those with access to the raw data. Traditionally, most analysis software was native to (and tied to) a particular measurement instrument and therefore only available in the lab. If a designer is analyzing data at that instrument, the instrument cannot serve its primary purpose of making measurements.

With newer software tools, lab personnel can measure and store texture data so that designers and engineers can explore and analyze it at their desks. This “offline” approach allows the quality lab to focus on its mission and expertise of producing accurate data, while design and manufacturing teams can leverage that data to better understand their products and processes.

Figure 3. A static printout (left) gives some indication of surface quality. Offline data (right) allows designers to explore and learn from measurements to improve parts and processes. Courtesy Digital Metrology Solutions

Offline software can typically analyze data from a variety of measuring systems. Thus, engineers can explore data from a variety of sources using a common platform.

Software can guide texture analysis

Surface texture analysis software is typically geared toward quality personnel who are expert in texture measurement. The dozens of parameters that impact measurements present a steep learning curve for users who have a more tangential understanding of texture.

To accommodate a wider range of users, more software tools are incorporating wizards and other organizational tools that guide a user through the analysis process (Figure 4). These tools promote interaction and separate the main data analysis functions from more obscure instrument settings. These tools can also help a non-expert user to understand how each analysis step impacts the data.

Figure 4. Software tools can guide surface texture analysis to help a non-expert achieve meaningful data. Courtesy Digital Metrology Solutions

Sharing data

Visualizations of surface texture can aid with process improvement, both across departments and along the supply change (Figure 5). The ability for management and production personnel to see surface data makes it easier to understand how aspects of texture affect performance.

Software can also offer a common platform for sharing point clouds throughout the supply chain. Sharing data can help solve discrepancies and disputes with suppliers and customers, reducing rework and rejection of parts.

Figure 5. Texture visualization software helps experts and non-experts to share texture information and collaborate on process improvement. Courtesy Chris Potter and COMP Cams.

“Some assembly required”

Offline software offers a wealth of information for manufacturing and design teams. However, it does not replace training in order to reach fundamentally sound conclusions about the data. Quality experts must be involved in verifying any conclusions drawn, and in implementing any process changes based on those conclusions.

That said, when more personnel understand why changes are implemented, the chances for lasting improvement is greatly increased.

Conclusion

The ability to analyze and interpret surface texture data offline has opened vast possibilities for part design, process improvement, and collaboration. By providing intuitive tools for exploring surface texture, offline software provides a common platform for analyzing surface data. Designers can explore surface data directly to improve surface function, while quality and process engineers can dig deeper to discover the root causes of variation and to better control manufacturing for reliable component performance.