Dovetail versus half dovetail gland

[scott82crg]

Hi guys,

I have a large mold that I'm designing. An o-ring gland will be cut into an end plate in a static face seal orientation. Pressure is low (vacuum to 45 psi), but I am heating the mold to 350F. Prior molds have had the o-ring fall out during mold manipulation. I'd like to try a dove-tail or half dove-tail orientation, but every source (including the Parker handbook) says "Only use if absolutely necessary." Is this primarily due to cost? Is there a performance advantage to full versus half?

I'm sure I'm missing some critical detail, but what are your first thoughts?

Thanks,
Scott

 

[tbuelna]

scott82crg-

Machining an o-ring dovetail face groove is an expensive operation, primarily due to the very small diameter dovetail milling cutter that must be used for the undercut.

Your operating temps are high, but have your considered bonding the seal in place?

Good luck.
Terry

 

[MintJulep]

The use of some sort of sticky goop to hold o-rings in place during assembly is common.

Dow-Corning 111 valve lubricant is good for this if your process is tolerant of silicone.

 

[unclesyd]

I concur with the above post in that dovetails are the last resort. If not done correctly they give the O-ring too much room to squirm.

At your temperature you are somewhat limited to materials you may want to try a type of X-ring with a tight groove

http://www.alliedmetrics.com/seals/x-ring-square-ring.php

We use a lot of 3M 77 adhesive spray for holding O-rings and gaskets in place. && does not lubricate, For real tough conditions we use DuPont Krytox Grease to hold and lubricate the O-rings.

 

[racookpe1978]

Starting/stopping the dovetail groove is a bit tricky as well: If a dovetail "starts" at a prependicular surface (like the edge of a plate) then the machiniing tool enterss the metal from the side.

But when the dovetail is cutting below a flat surface (like your gasket sealing surface) then the tool has to cut "down" into the surface before it can begin cutting sideways to make the groove. That entry point is difficult to seal. A rectangular O-ring groove has the same size cut at the top as well as the bottom.

 

[Tmoose]

is the o-ring groove in the end plate a circle

http://img.tootoo.com/mytootoo/upload/48/481397/product/481397_31abb02c06bf28647afee20486107520.jpg

Or some other shape?

http://t1.gstatic.com/images?q=tbn:ANd9GcQDGccYbnv7wdbk1kQjm-Xw6BsG1yzE9Za7CxUXpM_cdMSVvCN9&t=1

In either case, would an o-ring with a slightly undersized length hug the groove better?

 

[scott82crg]

Thanks, everyone. Sorry, I'm a little slow on my response.

This is a very low production, very high value part. We can't tolerate any risk of part failure, so the o-ring is replaced each casting. We've used 77 spray adhesive, but it was a mess and only work partially. Silicone of any type is not tolerated with this operation.

We have a smaller scale mold that we are going to run trials on. I'm leaning towards trying a 2% undersized x-ring prior to jumping into the dovetails.

Our machinist is recommending a 1/2 cord rather than the 3/16 we are currently using. The face seal is 12' diameter and the longitudinal seal is ~20' long. Thoughts here??

 

[imagitec]

I've used dovetail grooves in high end aerospace systems & semiconductor manufacturing equipment with great success. The expense of a dovetail groove is meaningless in these situations, as I suspect it may be in yours. The aerospace applications typically have a 66 ft submersion requirement, and the seminconductor equipment seals against high vacuum, tested with a helium-based leak detector.

Production hates any kind of goop (which isn't allowed in your application in any case) and loves dovetail grooves.

A rectangular groove is machined first, then a plunge cut is made with the dovetail tool & the dovetail feature is chased in the existing rectangular groove. The plunge cut leaves a circular opening at one spot equal to the maximum diameter of the tool. We haven't had any leak issues there; the seal is still primarily made on the faces.

As with any o-ring seal, especially one used in production, be sure the surface finish requirements are met & then protect those surfaces. If possible, make an adapter tool that will permit the seal to be tested when an expensive production run isn't on the line. Make it part of your preventive maintenance progream.

Rob Campbell

Rob Campbell, PE

 

[unclesyd]

You also may want to contact Helicoflex about using their Metal-OFlex sealing concept for your sealing problem. We have used this type sealing on some very critical sealing applications.
The last time I worked with the applications people they were extremely accommodating.

http://www.helicoflex.com/index.jsp

Sorry about the 77 spray. As I post we also use Krytox Grease which is essentially inert, along with Hylomar as a tacking agent for gaskets and O-rings. hylomar is a very thin coating.

 

[btrueblood]

A hard thing to fix with dovetail grooves is the expansion issue. A dovetail that will "hold" the O-ring at room temp. has a geometry that won't allow for thermal expansion of the ring at 350 F. The result will be that the ring will extrude itself under that temperature cycle. It may not be an issue for you, if you truly can stand to throw out an o-ring that has extruded after every thermal cycle. But the extrusion may also be strong enough to bend a light flange and/or stretch bolts if there are such in your design.

 

[49again]

Scott -

I agree with Rob - dovetails work well for us. The biggest downside for us is the cost of the groove, so we've started making our grooves a kind of "hybrid". MOST of the groove is standard rectangular cross-section, with short segments (especially at corners) of a dove-tail cross-section to keep the o-ring retained.

Steve R.

 

[moon161]

They're all over the stuff I draw, capitol vacuum chambers. Good vacuum practice as there is no trapped volume behind the seal when things are seated.

 

[scott82crg]

Thanks for the insight everyone. Have a good weekend.