O-rings are the epitome of elegant engineering: The ring itself costs only a few cents, and the groove it goes in is simple and easy to manufacture. But despite this simplicity, the resulting seal is able to reliably hold many thousands of psi of pressure. O-rings are definitely a machine design component you’ll want to be familiar with, and in this video, we’re going to tell you all about how to design seals with them.
If you are looking for more details, kindly visit peek o ring seals.
An O-ring forms a seal when it is squeezed between two adjacent surfaces. As the ring is squeezed, a contact stress between the O-ring and the surfaces emerges. If the fluid pressure is lower than the contact stress, then the seal prevents the fluid from escaping. In general, as the fluid pressure increases, the O-ring is compressed even tighter into the groove, further increasing the contact pressure, and hence helping the O-ring to seal even better. This positive feedback loop of increasing pressure leading to increased sealing is called “self-energizing”.
Seals can be generally classified as either static or dynamic. This hydraulic cylinder has examples of both. The seals between the cylinder and the end cap, the gland and the end cap, and the piston and the rod are static seals, since these components don’t move relative to one another after the cylinder is assembled. The seal between the piston and the cylinder, and the rod and the end cap are dynamic seals, since these components slide when the cylinder is actuated.
The most common type of seal is a radial seal, which can be designed in one of two ways. If the groove is on the ID of the housing, this is called a rod seal. If the groove is on the OD of the shaft, then this is a piston seal. If you have a choice between a rod and piston seal, it’s better to go with a piston seal, because the grooves are much easier to machine and inspect.
The biggest weakness of radial O-ring seals is that the clearance between components creates a path for the O-ring to extrude due to the pressure acting on it. Components called backup rings can help alleviate this. Backup rings are designed to spring out of the gland and block the extrusion gap. Where an O-ring alone could withstand perhaps only 2000 psi, a backup ring can help it hold 5000 or more.
Backup rings are very cheap, and effective. They are made of a plastic like PEEK or Teflon, and they usually have a scarf cut to help you install them in the gland. Technically speaking, if you only had pressure in one direction, you could get by using only one backup ring. However, it’s very easy to put the ring in on the wrong side, so as a design-for-assembly precaution, if designing with backup rings, you should always design for two.
Another configuration is a face seal, which you might use when trying to seal an enclosure. These are really a type of gasket, and they require a clamping force, usually provided by fasteners, to compress the O-ring. Face seals are actually really tricky to get right, because squeezing the O-ring requires a great deal of pressure. This first lid design is far too thin, and in the middle, there is virtually no squeeze on the O-ring, and hence no sealing. We can fix this design by adding a lip around the perimeter.
A third type of design is a boss seal. You pretty much only see them on hydraulic fittings, but they have a lot of advantages for other applications. In this configuration, the O-ring sits in a triangular space that is usually made with a special form tool. Boss seals are really easy to manufacture, since there aren’t any undercuts, and if the gland gets damaged, they can be reworked by just machining the profile slightly deeper.
Regardless of the gland design you select, you’ll likely need to choose an off-the-shelf O-ring from a catalog. O-rings are available in standardized sizes. The most common standard is AS568, and each size is assigned a “dash” number.
O-rings conforming to these sizes are available in many different materials. The primary considerations for selecting a material is the working fluid and temperature. Design tables, like this one in the Parker O-ring Handbook, are the easiest way to select a material that is appropriate for your application.
The same material is often available in a range of different hardnesses. The hardness is typically expressed as the Durometer Hardness. 70 is a fairly typical hardness and is good for most uses. 55 Durometer is much softer, and is a good choice for pressures below maybe 1000 psi because it is easier to install and less sensitive to surface finish.
90 Durometer is extremely hard, and consequently, is more resistant to extrusion. For higher pressures, exceeding maybe 6000 psi, you’ll definitely want to consider using 90 durometer.
However, the better high-pressure performance comes at a price. 90 durometer O-rings can be really difficult to install, particularly in small sizes. A good tip is to drop them in hot water for a few minutes to let them soften. They will be a bit easier to install.
Another important installation tip is to apply a high-quality O-ring grease before assembling the parts. In addition to lubricating the rubber and helping it slide in easier, most O-ring grease is designed to cause the O-ring to swell slightly, helping increase the squeeze after installation.
When you assemble the components, you need to squeeze the O-ring quite a bit to create a seal. It helps to have a shallow entry angle of about 15 degrees. This surface should be totally smooth and free of burrs so that the ring isn’t inadvertently cut.
Surface finish on the components is extremely important. As a general rule, the side of the gland, and the bore or rod, should have a 32 rms surface finish for static seals. The walls of the gland can be slightly rougher, at 64. This is where the piston seal really shines, since it’s typically more difficult to verify the dimensions and finish on a rod seal gland.
In a dynamic application, O-rings can work, but there are much better options to consider. This is called a T-seal and they are specifically designed for dynamic applications. They are packaged with two backup rings as a unit, and their primary advantage is that they have a wide, flat bottom to keep them from rolling around in the groove. They don’t cost much more than O-rings, and in our experience, are very reliable in dynamic applications.
Up to this point, we haven’t mentioned where the dimensions for the components come from. There are three seal parameters that will define the dimensions: squeeze, stretch, and percent gland fill.
Squeeze is how much you radially compress the O-ring when it’s installed in the gland. In general, 18-25 percent squeeze is appropriate for most static seals, but as high as 30 percent is sometimes used, particularly for cold-service applications.
Stretch corresponds to how much the O-ring is tangentially stretched AFTER it is installed in the groove. It doesn’t directly have to do with how much you stretch the O-ring during installation, though with small sizes of rings, the installation stretch creates other problems. Stretch should be below 5 percent, because high values of stretch cause the cross section to become smaller, decreasing the squeeze.
Volumetric gland fill is the final parameter. The O-ring and backup rings are incompressible, and if you don’t have enough space, you won’t be able to assemble the components because there will be nowhere for the O-ring to go. You also have to watch out for thermal expansion, because the O-ring will get bigger with temperature, and it can actually crush the metal components and yield them if you don’t have enough room. A good guideline is to keep the gland fill below 85%.
While you could use these equations to calculate the gland dimensions yourself, there is a source for pre-calculated values called the Parker O-Ring Handbook.
Unless you have a very unusual or demanding application, the guidelines we’ve given you, and the tables published by Parker should allow you to confidently design reliable O-Ring connections. If you found this video helpful, we’d really appreciate it if you shared it. We have a long list of engineering topics that we’ll be covering in future videos, so be sure to subscribe so you don’t miss these. If you have feedback, a question, or an idea for a future video, be sure to leave a comment!
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS
FIRST NAME
*
LAST NAME
*
*
MESSAGE
*
ADDITIONAL DETAILS
Thanks. We have received your request and will respond promptly.
Log In
Are you an
Engineering professional?
Join Eng-Tips Forums!
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.
Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Eng-Tips Posting Policies
O-Ring Groove Design
thread404-422673 Forum Search FAQs Links MVPsForum
Search
FAQs
Links
Are you interested in learning more about peek valve seats? Contact us today to secure an expert consultation!
MVPs
(Petroleum)
(OP)
21 Mar 17 05:03Hi All,
Does anyone have any references I can look up/purchase with regards to design of both static and dynamic o-rings grooves for high pressure (10,000psi) gas and liquid applications?
I've used the Parker o-ring handbook successfully for other lower pressure (approx 3,600psi) designs.
I'm assuming I just need to tighten up clearances on mating parts, but would like to see some other references for this higher pressure type work, unless the standard Parker grooves/clearances are ok for these pressures.
Thanks!
(Mechanical)
21 Mar 17 11:37You also have to consider using harder seal materials and anti-extrusion components. Use joint designs that create metal-to-metal closure. The grooves can be per the Parker Handbook.
Ted
(Aerospace)
21 Mar 17 13:47I would like to add to hydtools that metal-to-metal is for face seal only to my best experience.
For piston grove seal you will need more than just harder seal and anti extrusion component. Usually such details are not available freely. These are company's (or experts) propriety knowledge.
1. The type of anti extrusion design and material is important
2. Temperature range
3. How long it has to seal and how much leak is allowed over time (If it seals a volume).
4. Housing material and coating
5. Type of gas, sealing Helium is totally different than sealing Air, Nitrogen, Argon, etc.,
(Mechanical)
21 Mar 17 16:40I'd be contacting the 0-ring manufacturer's engineering department.
I think they'll ask you all of Israelkk's excellent questions, and a bunch more.
(Petroleum)
(OP)
22 Mar 17 09:52Thanks for the feedback.
We intend to use 90duro o-rings in an AED compound. Will also be using PTFE or possibly PEEK backup rings.
Will discuss with o-ring vendors hopefully they will have a HP design guide they can share.
If not we will use std Parker groove dims and err on the tighter end of tolerances to give less room for extrusion.
Always open to more feedback!
Thanks.
(Mechanical)
24 Mar 17 03:47Tuckabag,
I routinely spec out o ring grooves for static and dynamic applications. Also use piston seals (polypaks). What exactly is your application? Static or dynamic? Face or radial/axial? Loaded on both sides of seal or just one? Internal pressure? Acidic? Gas or liquid? For a start I would say to just tighten up the tolerances as per parker and give them a call to have them look over your design. They want to help.
Also, have you considered a metal to metal seal? It may be your best bet. See Parkers Metal to Metal Seal Design Guide for the rundown. You should have no prob sealing past 10k in either a face or axial application.
(Petroleum)
(OP)
24 Mar 17 04:05I have both dynamic and static applications for this job.
Basically I have a cylinder with end caps that seal on the ID of the cylinder tube via o-rings. These seals have atmospheric one side, up to 10,000psi the other.
There is also a dynamic piston. I would be looking to use crown seals in lieu of o-rings on this item, however the groove dimensions are the same. This can have up to max dP of 10,000psi, or can be 0psi dP depending on the operation of the unit.
I did find a nice table that showed recommended diametrical clearances based on pressure and o-ring duro. It recommended to reduce clearance down to approx 0.05mm for 10,000psi and 90duro o-rings.
(Mechanical)
24 Mar 17 04:44Interesting you found a chart. Was it from a supplier? Only thing that I've come across is from Parker's fluid power book.See Table 2-4 on page 2-5.All of my seal designs have been done by playing with the squeeze and extrusion gaps. Typically roll with aprox .0025'' (.05mm = .002'') radial extrusion gap and get decent seals upto around 5ksi. Liquid, internal pressure, dynamic piston seals. Mind you they are in a terrible environment and I can tolerate some leak/extrusion. How critical are leaks for your system?
(Petroleum)
(OP)
24 Mar 17 11:59When I'm back in the office I'll post them. Can't remember where I found them.
Leaks are super critical in this application. Typical use is on complex hydrocarbon streams for composition analysis, so we cannot afford to lose any light ends through leak paths.
(Aerospace)
24 Mar 17 15:10If "Leaks are super critical in this application" then using O-rings is not the solution to my best judgment. There is also a gas permeation through the rubber seal. You should may go with all welded sealed metal bellows designed for 10000psi and the movement and forces and life cycles.
(Mechanical)
24 Mar 17 17:12If leaks are a no go up to 10ksi you're sort of forced into using a metal-metal seal. Israelkk is right about the gas permeation issue with the o ring elastomers (or piston seals). Is the piston functionality absolutely mandatory? Guessing you're using the internal piston to force the fluids out for sampling like in a syringe? Could you maybe route any fluids that blowby the piston seal into a secondary expansion vessel and then bleed them off into your sampling chamber to make sure you capture everything?
(Mechanical)
24 Mar 17 20:47I would recommend a self energizing or pressure energized seal. They will have the pressure rating you need.However, because of the contamination issues metal seals like Swagelok VCR fittings might work better for the static seals.Someone like HIP might already have something in their catalog that will do what you need.
(Specifier/Regulator)
24 Mar 17 23:00There's also Bal Seal ( www.balseal.com ).
(Petroleum)
(OP)
27 Mar 17 02:11Thanks for the feedback so far.Metal/metal seals are not really a viable option in this application unfortunately. We need to maintain the piston function, basically we pre-load one side of the piston at or above line pressure with inert gas, then slowly allow the other side to fill from process (at line pressure) without flashing off any light ends by bleeding off the pre-charge pressure.We have used these seals in the past on lower pressure application (up to 5,000psi) with great success. They are very resiliant and provide great sealability.Have also used Bal-seals in the past, though they have not been as successful as crown seals.We also need to disassemble for cleaning/servicing, so o-ring seals for the end caps is really ideal for this service as well. We've never had any real issues with gas permeation before, apart from some RGD issues on high pressure/temp streams with significant CO2 levels. Even then the static o-rings were not too bad, it was the dynamic piston seals and valve stem seals that copped a hiding.VCR/bellows type seals unfortunately aren't going to work for us in this style of application.Also @MechyMarco please see attached the chart I found online re clearance gaps.
(Mechanical)
27 Mar 17 11:11You have not said what your volume is. It is probably small enough this is not an issue. However, you should double check that the increased pressure does not make it regulated pressure vessel.
Sounds like a clever design.
(Mechanical)
27 Mar 17 16:35Ahh right on Tuckabag. It's so frustrating knowing that Parker, Hitech, EPM and all the other seal manufacturers have loads of practical test data that we don't have access to. Makes sense that they don't publicly share their knowledge as that's sort of what distinguishes one from another though ya? It's funny you attached that sheet from Hitech as I was going to ask if that's what you were using. Sounds like you've got the design essentially nailed down. All you can really do is tighten up the squeeze and extrusion gap until you get an acceptable seal. Assuming friction isn't an issue of course.
Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.
Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.
Click Here to join Eng-Tips and talk with other members! Already a Member? Login
If you want to learn more, please visit our website peek seal.