At the heart of every spring or air-powered dart blaster, you rely on a few tiny pieces of rubber to keep air precisely where you want it. These seals can be tricky for even the most seasoned designers, and there's much confusion surrounding proper sizing and groove design. Much of the information here is based on specifications from the Parker O-Ring Handbook, and I've included some hobby-specific information as well. 

What's an O-ring?

An o-ring, as the name implies, is an o-shaped ring of rubber, plastic, or metal, typically with a circular cross section. They're used to seal joints and moving parts against liquids and gases. Variants include square, round, x (or quad), and double-x profiles. For hobby purposes, we'll focus on round and x profiles in 70A durometer nitrile rubber, as these are the most common. 

A note on sizing:

SAE and metric o-rings are sized differently! Metric o-rings are true to size. That is, their OD, ID, and cross-section are the same as their nominal spec. They are typically referred to by ID x cross-section - for example, the metric o-rings used on our fixed pusher are listed by the manufacturer as 7.5 x 3.0mm. Why would you expect anything else, you ask? 

Well, SAE sizes according to the AS 568 standard work a little differently. They use a "dash code," written as "Dash xxx" or "-xxx," where xxx is the 3 digit identifier. The first number is the cross section. Within this hobby, the most common cross sections you'll run into are 0 (1/16" nominal*, .070" actual), 1 (3/32" nominal, .103" actual), and 2 (1/8" nominal, .139" actual). The next two numbers denote the nominal ID according to a chart. These sizes are typically available in 1/16" ID increments, with some exceptions on the extremely small and large ends of the range.

*Sizes -001 through -003 have different cross sections

As it turns out, the AS 568 standard makes a lot more sense than you might initially think. The nominal dimensions refer to the appropriate gland size. A gland is the feature that holds an o-ring in place. Glands can be broadly classified as male, female, or face, and dimensions vary between static, dynamic, and floating dynamic applications. I won't cover face seals here, as they aren't commonly used in the hobby, but the same basic concepts apply. 

A complete seal can only occur under mechanical deformation. That is, o-rings require a certain amount of "squish" (that's a technical term) in order to fill the microscopic grooves in a sealing surface and contain a fluid. The amount of compression required varies based on the specific application. This is where the often-confusing AS 568 standard comes into play - the nominal sizes of each o-ring refer to the gland size they're intended for, rather than the relaxed dimensions of the seal itself. This article focuses on SAE o- and x- rings because they're much easier to design around. Metric o-rings require more effort to calculate a properly-sized gland.

Static Seals

Static o-ring seals are very simple to design, since none of the parts need to move once installed. A simple recipe for static gland design is as follows:

1. Determine gland type - male or female. Male glands are intended to seal externally, as on the nose of a pusher. Female glands are intended to seal internally, as on the barrel socket of a Lynx. 
2. Determine nominal o-ring size required. Size the OD or ID of your o-ring according to the mating part it needs to seal with. In the case of a pusher, size your o-ring to the nominal bore of the chamber. In the case of a barrel socket, size it according to the OD of your barrel.
3. Select a cross-section. For static seals, this is mostly driven by space requirements. In general, large cross-sections allow for more forgiving part tolerances because they have a wider range of deformation. 
4. Calculate your gland dimensions. Refer to an AS 568 chart to calculate the appropriate groove width and depth - a simplified chart is shown in Table 1 below. Ignore references to "backup rings," as they are only necessary in high pressure (>1500psi) applications. Subtract 2x the groove depth from the ID of a hole to determine the diameter of a male gland and add 2x the groove depth to the OD of a shaft to determine the diameter of a female gland.

Table 1 - Static O-ring Groove Dimensions

That's a lot of compression! Static glands squish the heck out of an o-ring, and typically have a cross-sectional compression between 17 and 30%! That can greatly increase installation forces. Make sure to apply a 10-20 degree chamfer before the gland in order to aid installation. Male glands will stretch the o-ring somewhat, but this should not be relied on to change the effective diameter of the seal. While there is some wiggle room within a given o-ring size that will allow it to accommodate small deviations from nominal, a gland should not stretch an o-ring more than 5% for optimal reliability and longevity. 

It is acceptable to reduce compression in order to reduce installation force when necessary; testing will determine the minimum compression required for adequate performance in a specific application.

O-ring glands should be straight-walled or slightly tapered - typically no more than 5 degrees from perpendicular. The gland should ideally have a smooth surface finish - if machined, a 16-32 microinch finish is preferred. This is obviously not possible when using 3D printed parts. A softer seal material can be selected in extremely adverse conditions, but it seems that printed sealing surfaces work reasonably well as-is for operating conditions of most dart blasters.

Glands should not have a full root radius because the o-ring needs room to expand. This is referred to as "gland fill," and is the percentage of gland area occupied by the o-ring when compressed. Obviously, you cannot exceed 100% gland fill, but it's important to leave some room in there to accommodate variations. 75% fill is optimal, but anywhere between 60 and 85% will typically work well. The gland dimensions in Table 1 above already factor in gland fill, but you'll have to calculate it yourself if you have special requirements.

Dynamic Seals

Dynamic, or "Industrial Reciprocating" seals are very similar to the static seals described above, but they involve moving parts. You'll notice that the glands are a bit deeper to reduce compression, and therefore sliding friction. This type of seal should be used on relatively slow-moving parts where absolute speed is not a functional requirement. Generally, male glands (piston seals) are preferred, though female glands (rod seals) are not uncommon. Glands are designed with the same process outlined above for static seals, but using the dimensions in Table 2 below. 

Table 2 - Dynamic O-ring Groove Dimensions

An example of a dynamic o-ring application is the ram base o-ring in a Caliburn or Talon Claw.  This seal has a male gland cut into the ram base and seals in a nominally sized plunger tube. Female glands, or rod seals, are much less common in this hobby by comparison, but exist in some blasters like Taffy's Skewer and the Zinc by 118 Design to seal the plunger head on a concentric guide shaft. Often, female rod seals are a greater source of parasitic friction than male piston seals, and may function better using a floating gland design as described in the next section. 

Floating Seals

Floating seals are a bit of a special case, and you'll have to deviate a bit from the standard in order to achieve best results. The Parker Handbook includes a section on floating seals and a lookup table based on bore ID with gland dimensions for each specific o-ring size. However, in-house testing has demonstrated that the o-ring sizes specified are too aggressive to work well on a plunger head, which inherently requires very low friction for best results. Often, an imperfect seal will yield better performance over a tighter, perfect seal. The Parker Handbook's sizes are based on exact compression calculations and prioritize a perfect seal over absolute minimum friction.

The table in the Parker Handbook is not entirely useless here, though. It includes an important detail often overlooked by designers: floating o-rings are extremely sensitive to compression, and generally require a thicker section than would be necessary in a standard industrial static or dynamic seal, where constant compression is created between the gland and mating part. Instead, these floating seals rely on the compression created between the nominal bore and actual OD of the o-ring in order to produce the minimum amount of deformation required to seal.

Generally, my strategy is to select an o-ring with a nominal OD the same as the ID of the bore I'm installing it in. This ensures that the actual OD of the seal will be slightly larger than the ID of the bore. Note that I'm specifically referring to a male gland (piston style) seal here. Contrary to the Parker Handbook, floating rod seals are possible. However, they exhibit poor performance by comparison and may not yield a 100% seal using my guidelines.

I've condensed the section thickness guidelines from the Parker handbook into Table 3 below. The values are approximate, and if there's any doubt, it's typically better to jump to the next section thickness where the design allows.

Table 3 - Floating O-ring Cross Section Selection by Bore Diameter

Note that one of the most common plunger o-ring sizes, -123, falls outside of the optimal ID range for a floating seal with a -1xx cross section. I believe that this is the primary source of general frustration with floating seals. There is simply not enough compression with a 3/32" nominal cross section o-ring in a 1 3/8" nominal bore to form a reliable seal. This issue is compounded by variations in plunger tube dimensions. Typical 1 1/2" x 1 3/8" polycarbonate tubing has an ID tolerance of +/-.015". That's not so great.

It's much worse when you consider that the actual OD of a -123 o-ring is 1.380" +/-.003". That means that in the absolute worst case scenario, you could end up with a 1.377" o-ring in a 1.390" ID tube! Realistically, that's pretty unlikely, but a batch of polycarbonate tubing that was pushing the upper tolerance limits wreaked havoc on dozens of Lynxes produced in early 2021. Users were having to print larger plunger heads which stretched the o-ring - already a no-no that I covered earlier - in order to force a low-quality, high-friction seal with the plunger tube. Frequently, I see modders wrapping Teflon or electrical tape around their plunger heads increase o-ring compression in what was supposed to be a floating application. This is band-aid solution at best, and one that robs you of power by increasing plunger friction. 

The proper solution here is to use a thicker o-ring. A -216 o-ring has the same OD as a -123, but has a 1/8" nominal section thickness. That translates to an actual OD of 1.387" +/-.004". When you compare that to the tolerance band of polycarbonate tubing, it's immediately apparent that you'll have better results. 

Hey, what about gland width and depth?

Don't let the ID of the floating seal get stretched by the gland. That's it. You want the seal to float, as in the name, so it should be loose in the gland. I typically allow about .025" of diametrical clearance past the actual ID of the o-ring. In the case of a -216, the actual ID is 1.109", so I'd select a gland diameter around 1.075-1.085". You don't want the groove so deep that the o-ring can just fall right off, and you also don't want it so shallow (relative to the OD) that it can pop out in use. Make sure that the walls of the gland contact at least half of the section thickness of the o-ring, or it will have a tendency to fall out of the gland during use. Gland width can be the same as or up to about .01" wider than the equivalent industrial reciprocating gland for a given section width.

X-Rings

X-rings serve two main purposes: they increase the number of lines of contact in a seal, and decrease compression required to create a seal equivalent to an o-ring. 

Increasing the contact area in a seal will improve its overall reliability, especially when located close to the edge of a feature. Since they create twice as many contact patches as a round profile o-ring, they have redundancy in the presence of foreign contaminants, where one edge of the seal can act as a wiper to keep dirt and dust away from lubricants and critical surfaces. They also produce higher quality contact with sealing surfaces, since the parting line and any flashing from molding can be located in the valleys of the X profile, away from critical edges. X-rings help improve lubricant retention as well, and have a tendency to trap oil in their peripheral grooves to ensure constant lubrication of sealing surfaces for dynamic applications. 

As a result of these differences, x-rings need slightly different groove geometry to realize their maximum potential in dynamic applications. Fortunately for us, they're almost always used as floating seals in dart blasters and are sized the same way as their round section cousins, so floating seal grooves will work just fine with o- and x-rings. X-rings will tolerate variations in bore diameter better than an equivalent o-ring, and a -123 x-ring can help make up for a slightly oversized plunger tube that might not achieve maximum performance with a standard o-ring due to blow-by.

Static and industrial reciprocating x-ring groove design is outside the scope of this article, and in most cases, users looking to improve performance of an existing, leaky o-ring in such applications will be perfectly satisfied by directly swapping in an x-ring, provided that the existing groove is properly designed with straight walls. Benefits from a tailored x-ring groove are only realized in extremely demanding applications requiring perfect sealing at high pressures while maintaining relatively low dynamic friction and stick-slip. For spring-powered blaster applications, a floating x-ring will vastly exceed the performance of a compressed, industrial reciprocating seal on a plunger head. 

Conclusion

Seals can be daunting for novice and experienced designers alike. Hopefully, this article helped improve your understanding of o-ring seals and gland design. Let me know if you have any questions or concerns in the comments, as well as if there are any other topics you'd like to see covered here in the future!