What's new
What's new

Machining O-ring grooves


Jul 20, 2003
Paradise, Ca
I need to make some parts for a friend. Material is 6061 T6 round bar. He needs an O-ring to seal on a 1.000" tube.

I've never dealt with O-rings, so I don't even want to bore the bar out until I get some help on this. Are there any inserts that are designed for O-ring grooves, or do I need to grind my own tooling? What's the best way to figure out how deep to make the groove? I'd like to make it to where the O-ring is somewhat captive, so when the 1" tube slides in, it won't try to roll the O-ring out of its groove. (with some oil on the O-ring, ofcourse)

Clear as mud? I think I could explain better if someone will ask the right questions!

Thanks guys!
hello Mike,

I have a parker oring book. The info I need to determine your cavity parameters are:

What oring? Typically they are sold as a 3 digit size, at nominal section thickness. What cross section thickness oring?

Also OD/ID is generally sold as referenced to an aerospace spec. When you see OD/ID listed as nominal fractions - that is very much approxiamate. See mcmaster carr for some additional info on this.

I am not totally clear on your arrangement. Is the oring on the outside of the one inch bar? Or on an ID that you will bore into the 1" bar-then what is the ID of your bore?

If you are not sure about any of this describe your arrangemrnt:

oring on the piston in a bore? What od/id is piston/bore?


Is the oring in the bore? What od/id is piston/bore

give me what you know and I'll size the rest.

[This message has been edited by morsetaper2 (edited 12-15-2003).]
Hey, thanks for the reply. I'll see my friend tonight, and will get all the exact dimensions I need and let you know... I'll clarify the application then too.
Mike R:

Pull up: www.usseal.com Click on 'O-ring American......', then 'O-ring gland design', and 'American O-ring size'.

Knows all---------shows all.

I'm kinda pressed for time right now, but I'll check your link out later.. thanks Jose!

To make things alittle clearer... my friend wants me to build a sheet metal (aluminum) valve cover for his four banger. It's a hemi head, and has tubes that go down through the top of the valve cover and seal on the head, between the spark plug and head. What I'm working on is how to seal the top of the tubes to the top of the valve cover. I want to have an O-ring in a female part (made from round bar), that the tubes pass through. The female part would weld into the top of the valve cover, on a jig.

Now that I've had a chance to measure everything myself, I have another question: How much tolerance can I get with this setup? The tubes he's using average between 1.230" - 1.251", with an average TIR of .015". I'm hoping, with the right design, that O-rings are still a feasable option, because that would make the cleanest looking seal I can think of!

MorseTaper, without hardly ANY research, it looks like I'm in the market for a 216 or 217 size. Or maybe 316/317??

Thanks again!

[This message has been edited by MikeR (edited 12-15-2003).]
The easiest way to cut internal o-ring groves is with a double-ended boring bar, such as the one shown at the top of page 2338 in McMaster-Carr's catalog.

It's easy to make your own if you want to use a small round bit. Just cross-drill a piece of round bar for the bit, and drill/tap into the end of the bar for a set-screw to hold it. Grind the bit square on the end with a slight relief angle on the sides and a small radius on the corners.

The usseal link looks like a great reference.


[This message has been edited by winchman (edited 12-16-2003).]
Parker has a downloadable version of their O-ring design handbook but it's a PITA to find. Apple Rubber has one that is easier to download. A Google search should turn up one or the other.

I've machined internal and external O-ring grooves and ground a standard HSS lathe bit for both applications. It's a pretty simple job - must be if I was successful.


With o-rings, application determines a lot of things. Sealing low pressure pneumatics, static seals (like your valve covers), and high pressure hydraulics are three different animals. With your valve covers, you want a gland that makes contact with the ID, OD, and face. That's all. No complex formulas, no critical groove dimensions. Your only concern might be the environment. I don't have my Parker book here, but you might want to check that the standard Buna-N material will live at the temps that the engine operates and are compatable with oils involved. You may want to step up to the viton material. All these are discussed in the Parker handbook. Try your local bearing supply houses or fluid power distributor to get a copy. When I was designing pneumatics, it was my bible!

Generally you want a 15% squeeze on an o-ring.
If you are sealing on a 1.000" ID your seal ring groove diameter for varying oring sizes is as follows:

000 series (.070 cross section)
seal ring grrove would be .882" dia. x .09 wide

100 series (.103 c.s.)
seal ring groove would be .825" dia. x .125 wide

200 series (.139 c.s.)
seal ring groove would be .763" dia. x .156 wide

300 series (.210 c.s.)
seal ring groove would be .634" dia. x .234 wide

The widths are not very important, but they must be greater than the cross section of the seal. Best of luck
Just wondering how you are going to get the spark plug tube spacing measurment, and what kind of tolerance they have between each other, as well as their diameter. In other words, if you make this cover to spec, is it going to be a PIA to line up and install? Most OEM valve covers for this application use a very wide and flexible o'ring type seal, about 3/8 inch in (flat) thickness. In fact why not machine your valve cover to accept the OEM seals? Just thinking...

[This message has been edited by Hanz (edited 12-16-2003).]
Hanz (Franz from the Hobart board??), I do have some concerns with this, but there's only one way to find out! I build all my valve covers on jigs which I make to duplicate the head dimensions, NOT the stock valve cover. From what I've seen, factory tolerances on cylinder heads seem to be pretty good, since valve train geometry is so important.

My friend has ordered new spark plug tubes, so it'll be interesting to see if they are as out of round as the old ones. Hopefully not!

Doug, thanks for those numbers! I'm thinking about going with the widest O-ring (cross section) I can find, so I can have as much possible squeeze to make up for any out of round in the tubes. (400 series) I'm also trying to have the deepest groove possible, so the O-ring doesn't try and roll out of its home as the spark plug tubes are installed. Would having a .270" wide groove for a .275" wide O-ring be detremental?

Bottom line.... Am I asking too much from an O-ring??

Winchman, The reason I asked if there are any inserts for this application is because, once I prototype the first ones, I'm going to have my friend make more on his CNC lathe. I'm probably wrong, but I have the impression that HSS bits won't be mountable on a machine like that.

Ok, I'm sure I forgot a bunch of stuff, but I'll bet you guys can get me straitened out! I sure appreciate all of your help guys!!!!!!
I think you are going to have to reduce the TIR of the tubes to get a good seal. Or else your gland width will vary too much with respect to the inconsistent tube diam. You can't lathe turn it to match the TIR to maintain consistent gland.

For instance, the x-sectional tol on a .139 x-sec oring is +/-.004. And the corresponding groove depth specification is .111-.113 according to Parker book. Your TIR blows that away.

Get the TIR under control, download the parker book. Honestly oring sizing isn't as clear cut as it might seem.

You have stated that you will have:

1. Male tube OD (need to improve TIR here-and get final OD dim). You are using a female oring gland

2. Choose most appropriate oring sz to closest match of tube OD (by oring's ID, then x-sec).

Then, you sz the following based on 1&2 (above) using parker chart A5-1:

3. groove diam (female gland)

4. throat diam (female gland)

5. groove width

or start with throat diam and determine other parameters.


[This message has been edited by morsetaper2 (edited 12-17-2003).]
Morse, I know you're not trying to, but I think you're telling me it won't work. You see, the spark plug tubes are very thin, and deep drawn steel. There's no way I could turn them to true them up, nor could I (or my friend w/ the CNC lathe) make new ones with out being extremely expensive. I was just hoping that the application would be docile enough to not have to have everything textbook perfect. Your thoughts?

Owen also mentioned the Parker book, but you said to download it... do you have a link for it?
Hello Mike,

If they are thin tubes I guess you have no choice. I have always had the luxury of specing gland dimensions by the "book". I honestly don't know how well your arrangement will seal given the nonconcentric nature of the gland vs tube. But if this is a race car type of application maybe it will be good enough. Just don't oil the track. You won't make friends that way.

Maybe someone else knows or has experience sealing with less than ideal parameters. These are from the Parker book, .


pgs 4-1 through 4-13 are your pages to look for! I need to get to sleep. Hope this helps.



[This message has been edited by morsetaper2 (edited 12-17-2003).]
As far as o-ring tools,we use two brands
Horn,which have a round flag type insert
that screws on from the front of the bar
or a Circle bar which has a rectangular
insert with dual flags so that you can
use either side.they are held by a side
screw.Inserts for the Horn tool come in
different widths.when we use them we plunge
to depth then cut backwards if nessessary
to bring in size.If they are going to be
done on CNC machines have them programmed
with slight edge breaks at start and finish of cut,so the sharp edge doesn't cut them
when being installed.
Mike, you need to use a bigger groove width than the o-ring cross section. The o-ring HAS to expand somewhere. Hence you normally want a wider groove.
Could you use a flange, or lipped o-ring? You know...the type that has a lip coming off at an angle...I don't know the true name for them... but somebody will know what I mean. It might get around the tir problem tyou are going to have.
Still wondering why you don't machine to accept factory OEM seals, which were designed for this purpose, and would be available at any local parts house should they need to be replaced on the weekend or at the track...? Just out of curiosity, what head are these for? Hanz