Predicting sealing quality by filtering surface texture data

The shape and texture of a surface is critically important to its functionality. This is particularly true in sealing applications. Today, there is no shortage of instruments for measuring surface texture. Yet, despite the great amount of data that can be gleaned from surface texture measurements, most engineers continue to specify only basic height-based parameters, such as average roughness or total flatness. (See Figure 1.)

The challenge, however, is that most surface functionalities (sealing, for example) are not measurable purely in units of height. Local curvatures and the connectedness of surface features can provide much more valuable insight.

In recent years, morphological filters have been introduced in surface texture literature and standards. The morphological opening and closing filters provide a powerful means of simulating surfaces in contact. Furthermore, they are well-suited to predicting how a relatively soft surface (e.g., a gasket or seal) will interact with a relatively rigid surface. This article discusses how sealing performance can be predicted with this analysis via modern surface texture analysis software.

Can average roughness distinguish a “good” sealing surface?

As previously mentioned, the most commonly specified surface finish parameters in sealing applications are based on heights measured along a profile. These common parameters include average roughness (Ra) and average peak-to-valley height (Rz). In some cases, a waviness height (Wt) may also be specified. These parameters are well-known and readily measurable with most surface profiling instruments.

Unfortunately, these statistics do not necessarily indicate how the surface will perform. Figure 2 shows two very different surfaces with virtually identical Ra, Rz and Wt values. If these parameters were specified on a print, the two surfaces would be virtually indistinguishable and would pass similar quality control testing. Yet, none of these parameters would reveal that the repeating milling pattern in the second surface could create leak paths between mating surfaces.

New surface texture filters provide insight into function

In recent years, new parameters have been developed that closely correlate with particular functionalities. Engineers can use these functional parameters to compare design options during development and then specify these parameters on prints to improve component performance and longevity.

When the closing filter is applied to the first surface, it closely follows the waviness profile. Although the waviness is quite large in terms of total height, there are no abrupt changes that would cause significant leak areas.

However, when the same closing filter is applied to the second surface, large, periodic voids are shown that cannot be sealed by the virtual gasket. These voids are of concern.

While the profile graphs clearly show the difference in sealing, a numerical value to indicate the amount of leakage could be desirable to specify and control these critical surfaces. The functional parameter Wvoid (void area per unit length) was developed to satisfy that need. Wvoid, which is shown via the OmniSurf software in Figure 4, can be tracked as a measure of sealing quality. The parameter is normalized per unit length, making it independent of the evaluation length and therefore more repeatable and stable. 

This analysis can be extended into three dimensions as well. Figure 5 shows the application of a closing filter to a measured surface using interactive filtering tools in OmniSurf3D software. In Figure 5, the closing filter has been applied across the dataset (on left) to create the closing surface, shown in transparent blue. On the right, the gaps between the closing surface and measured surface are shown as a void surface, which represents the void areas where leakage may occur. As with the 2D analysis, changing the radius of the closing filter can be related to changes in conformability and sealing forces.

Using Pit and Porosity analysis to trace leak paths

The porosity features shown in red are considered closed pores, meaning that they are completely enclosed by material at this given cutting plane. Blue features are open pores, meaning they open onto the edge of the data set. The green regions may be the most important. These are edge-connected regions, representing potential leak paths that are open along either the X or Y axis of the data set. 

At any cross-sectional height level, the open and closed pores can be counted. Or, the pit density, the number of pits per square centimeter, can be calculated. Pit density can give a better indication of the surface in general, irrespective of the size of the measurement area. Analysis tools like this make it easy to visualize what is happening and provide traceable numbers.

Conclusion

Surface texture filters and functional parameters make it possible to quickly and easily measure the properties of a surface that are most specifically tied to a desired function. 

A morphological closing filter can provide insight as to how well a surface will seal and provide insight about other functionalities. These filters can be used to explore changes to materials and methods that may address the root causes of sealing issues, as well as provide parameters that can be specified on prints to guide production and track surface quality.