Rotary dynamic sealing applications in the electronics manufacturing, automotive, aerospace and other industries depend upon the fit between the shaft and the seals to avoid leakage and premature wear that can reduce performance and service life. Most applications require precision measurements of the surface texture of the shaft, as well as the presence of spiral grooves on the shaft, known as “lead angle,” that can result in leakage.
Surface roughness is traditionally measured with a stylus profilometer, also called a “contact profilometer,” while shaft lead angle is measured by winding a string around the shaft and observing its movement as the shaft rotates. White-light interferometry is a relatively new proven and quantifiable approach that utilizes the properties of light to achieve topographical measurements of surfaces with feature heights from nanometers to several millimeters. This article will explore shaft surface measurement challenges, describe the various measurement options in detail and discuss the pros and cons of each alternative.
Shaft measurement challenges
Leakage in rotary dynamic sealing applications is prevented by preloading the lip of the elastomeric seal, making its internal diameter slightly smaller than the shaft diameter. The surface finish of the shaft in the area where it contacts the lip is critical to avoid fluid leakage and minimize friction. Excessive roughness on the shaft can cause rapid wear on the seal, resulting in leakage. On the other hand, if the shaft is too smooth, the seal will not bed correctly, which will also lead to leakage.
The ideal shaft surface should generate an amount of wear on the lip of the seal that will allow a small quantity of liquid to enter the shaft-seal interface. At this point, the lip seal begins to run on a fluid film and lip wear should end. Sealing is provided by the surface tension of the hydrodynamic oil film between the seal flattened area and the shaft. Ideally, the thickness of this film should be between 1 and 3 microns. The meniscus acts as an interface between the outside air and the fluid. Any break in the meniscus will result in leakage. This can occur if the shaft contains scratches along the seal path. Having a surface with the right amount of roughness is important to avoiding leakage in dynamic sealing.
Even if the surface texture of the shaft meets specifications, the presence of a lead angle can generate leakage. The introduction of a lead angle on a shaft is inherent to all shaft manufacturing operations. Fine helical shaped grooves are inadvertently machined into the shaft due to the feed rate of the cutting tool and the part orientation during turning or grinding. This helical pattern leads to a wicking out of lubricant or ingestion of contaminants through the shaft-seal interface. A left hand lead moves the oil in the direction of the source, causing the seal to dry out. A right hand lead pulls lubricant away from the seal, causing the seal to leak. The International Organization for Standardization (ISO) standard 6194 states that the portion of the shaft that contacts the seal should be free of machining leads. The Rubber Manufacturers Association (RMA) OS-1-21 rev. 2004 specifies a lead angle of zero degrees +/-0.05o for shaft lead. Both of these standards also list parameters for surface roughness that are shown in the table below.
Alternative measurement methods
Surface texture is traditionally measured by a profilometer with a stylus that moves across the sample or part surface. The profile is then filtered so that the areas of the profile above and below a line called “the mean line” are equal. The measurement Ra is the average of the absolute value of the deviations from the mean line over the evaluation length. Rz is the average value of the greatest peak to valley distances in five consecutive sample lengths taken over the assessment length of the filtered profile. Rpm is the average peak to mean height of the average value of the five highest peaks above the median in five consecutive sample lengths, taken over the assessment length of the filtered profile.
The standard approach for measuring lead angle involves mounting the shaft in a holding chuck and lightly coating it with silicone oil. A length of specialized quilting thread is draped over the shaft with a 30 gram plumb weight suspended from one end in such a way that the thread makes an arc of contact with the shaft, ranging between 220o and 240o. The shaft is rotated at 60 rpm in both directions and the string is observed using an optical eyepiece, with a Vernier scale or precision calipers measuring the movement of the thread. If the thread translates in either direction, it signifies the presence of a lead angle. The lead angle (A) can be determined by formula: Arc Tan A = Thread advance per shaft revolution/shaft circumference.
A newer quantifiable and repeatable method involves the use of white light interferometry, also known as optical profiling, to measure surface texture and lead angle very accurately. In an optical profiler, light approaching the sample is split and directed partly at the sample and partly at a high-quality reference surface. The light reflected from these two surfaces is then recombined. Where the sample is near focus, the light interacts to form a pattern of bright and dark lines that track the surface shape. The microscope is scanned vertically with respect to the surface so that each point of the test surface passes through focus. The location of the maximum contrast in the bright and dark lines indicates the best focus position for each pixel, and a full 3D surface map is generated.
The first generation of optical profilometers was primarily designed to perform dimensional measurements and characterize surface texture. In addition to the traditional 2-D surface measurements, such as Ra, Rz and Rpm, the optical profiler captures 3-D topographical information that makes it possible to provide great definition of the texture by 3-D surface parameters Sa, Sz and Srpm. These 3D measurements represent a complete surface area rather than a single line trace, so they are more representative of the critical sealing area shaft and generally are insensitive to alignment.
The latest generation of optical profilometers go one step further by calculating the lead angle based on the surface profile. Calculation of the lead angle to an accuracy of +/-0.05o requires compensation for any off-axis variation associated with the mounting of the part. The operator loads the shaft in a chuck and specifies the locations to measure. At each location the system measures a best fit to true cylinder and lead angle with reference to a charge coupled device (CCD) camera. This is accomplished by a best fit nominal arc calculation to independently determine the orientation of the shaft.
Fig 1. Measurement of glass substrate, challenging for most stylus profilers. Perfect for interferometric optical profiler.]
The system also simultaneously conducts a Fourier transform of the data to determine the angular power spectral density of the angle of the marks on the shaft, again with reference to a CCD camera. These two calculations are subtracted from each other to eliminate any off-axis variation associated with mounting the shaft. The result is a lead angle and surface measurement value at every point specified by the operator.
Tradeoffs between competing approaches
Most companies producing precision equipment that utilize dynamic sealing are currently using the conventional methods of stylus profiler and string winding to measure surface texture and lead angle respectively to avoid leakage and contamination and ensure seal life. The cost of a stylus profilometer ranges from $25,000 to $200,000 and the cost of a string assembly is in the neighborhood of $20,000. These methods have the advantage of being familiar to operators and engineers and also represent the lowest cost alternatives.
Optical profilometers with lead angle measurement capability are comparable to high-end stylus profilometer price points. However, in comparison to stylus profilometers, optical profilometers offer the additional advantage of being able to simultaneously measure both lead angle and surface roughness at a considerably higher level of accuracy. Some authors have reported that shafts with a lead of less than 0.05o degrees often have a dead band using string measurement, which leads to a broad designation of no lead on shafts with lead angles ranging from 0.05o RH to 0.05o LH. Another concern is that since the string must translate a significant distance to achieve good accuracy, local measurements in the critical seal contact areas are often not possible.
Fig 1. Bruker-Nano NPLEX-LA measuring shaft lead angle.]
On the other hand, tests of the accuracy of optical profilometers in measuring plunge ground shafts have shown that when parts with 0.4 o RH, 0.4 o LH and 0.05 o RH lead were measured three times on three different systems, the maximum deviation of results between systems was 0.08 o on the 0.4 o RH part and 0.04 o for the other two parts. The one-sigma standard deviation of results between the systems ranged from 0.02o to 0.04 o inch. These results demonstrate the superior accuracy of the optical profilometer compared to conventional measurement methods. These accuracy improvements should save time and money by reducing field failures and warranty costs.
This article has reviewed the two leading approaches for measuring surface texture and lead angle on shafts used in rotary dynamic sealing systems. In summary the traditional methods of using a stylus profilometer and string winding offer the lowest cost while the newer method involving the use of optical profilometers provides superior measurement accuracy.
About the Author
Javier Vera is a Product Marketing Manager at Bruker, Nano Division and has held various positions as product manager, metrology manager, and application engineer. Currently he has responsibilities as marketing manager for the Stylus and Optical Metrology Unit. Javier has more than two decades of experience in the field of industrial and optical metrology, during which he has written several articles and white papers on the subject. Javier can be reached at: Javier.Vera@bruker-nano.com