When designing a gland in which an O-ring or elastomeric seal is the desired sealing component, there are several aspects that need to be considered. A perfectly designed seal with the right material, ideal compression, gland fill and stretch can have inadequate sealing capability if the surface finish of the hardware is neglected. This blog discusses the ideal surface finish requirements for both the application and testing of seals.
Consider these photos of metal surfaces. At first glance, all three may appear to be identical, but looking closely, the main difference is surface finish. Figure 1 illustrates the appearance of surface finish as it will be discussed.
Figure 1: Left to Right: 500x magnification of 16µin RMS, 32µin RMS, 63µin RMS
Surface finish, as pertinent to seal design, is the measurement of the roughness of the two hardware faces compressing the O-ring or seal. That is why it is sometimes referred to as “surface roughness” as well as surface finish. Maintaining the proper surface finish of these two surfaces is essential to obtaining a good seal. Table 1 outlines the basic guidance suggested in the ORD 5700 O-Ring Handbook.
|Liquid||32µin RMS||16µin RMS|
|Gas or Vacuum||16µin RMS||16µin RMS|
The requirements for sealing gas and vacuum are more restrictive than a liquid due to gas’s ability to find a passage through very minute pathways on a hardware surface. Some estimates are that the viscosity of air is 53x to 55x less than that of water, which equates to about 53x more volume of air passing through the hardware indentation than water would.
For static seals, Parker recommends using a surface roughness value not to exceed 32 µin (32µin RMS) when the seal involves liquid and a maximum of 16µin RMS when the seal involves gas.
If a surface is too rough against a static seal, the O-ring may have difficulty conforming to surface imperfections causing leakage. The Durometer of material can play a role in overcoming surface finish. The softer the material, the more it will fill in the peaks and valleys of the sealing surface, however, this may be at the detriment of other sealing properties, such as contact pressure, compression set resistance, extrusion resistance, or durability.
For static sealing, consideration of the method used to produce the surface finish can certainly play a role and potentially offer an improved sealing margin. Methods such as lathe or some other machining technique that produces tool mark parallel to the groove can be sealed most effectively and in certain situations may seal at roughness values greater than recommended. Other methods, such as end milling or routing, produce tool marks perpendicular to the groove and can be too deep for the O-ring to make full contact which could result in a leak path. In this situation, the recommended roughness values should not be exceeded.
For dynamic seals, the shaft or bore should have a surface finish between 8µin and 16µin RMS. This range of peaks and valleys on the hardware serves the purpose of holding the lubricant against the O-ring and ultimately minimize friction and wear damage.
Surface finishes above 20µin will cause abrasion on the O-ring surface, and no amount of lube will prevent the O-ring from wearing. Surfaces that are better than 10µin will result in the lubricant being wiped away, which thereby increases friction and accelerates wear over the life of the seal.
|Figure 2 Illustration of mismatch on a molded component.|
Questions come from customers with respect to hardware mismatch, surface porosity, and air or helium testing. Each of these questions often simplifies down to the same guidance which has been outlined above. Figure 2 illustrates the mismatch that can be present on a molded housing. While there is not a hard and fast rule for overcoming mismatch, application experience has found success with limiting the step to a maximum of .003”. Like mismatch, surface porosity is another application-specific hardware challenge that can be difficult for a seal to overcome. A general rule of thumb is for the maximum porosity size to be less than half of the contact width of the compressed seal in the least material condition. Lastly, a seal designed to contain fluid that is failing an air leak test often has the root cause of leakage due to surface roughness. This comes back to the reality that air and gas are a much more difficult sealing medium than a liquid and a smoother surface finish can often improve this condition. In some instances, mismatch and surface porosity can be overcome with a custom-designed seal, but it will not be possible for a custom-designed fluid seal to pass a gaseous leak test when the surface finish is the cause of leakage.
If you have additional questions about surface roughness, please visit our website and chat with our support team or reach out to one of our application engineers at OESmailbox@parker.com.
Dorothy Kern, applications engineering lead, Parker O-Ring & Engineered Seals Division
Matt Frye, product design engineer, Parker O-Ring & Engineered Seals Division