Leaking Protos Using Established Dimensionals

[gorilla1969]

I'm at a loss with a prototype part I'm developing which should be a relatively simple change. Basically, the proto is an offshoot of an existing part used for two years now with great success.

The part is a rotating knob with various hole sizes drilled along the inner circumference - see the attached graphic. The knob rides on a set of o-rings on each side of the knob which creates a seal for a 20-30 PSI, gas-operated pneumatic device. The holes, depending on the position of the knob, align with the concentrically matching o-rings on each side of the knob and are compressed with a round solid on each side that compresses the assembly with four screws. Grooves are cut around the circumference to allow the screws to pass through the knob and attach to the opposite solid piece and also allow for rotation of the knob. With the prototype knob, we are adding some additional settings, changing the detent grooves (scalloped cuts in the knob) for a softer 'feel', etc.

As part of our functionality tests, we apply pressure in the 20-30 psi range into the unit and attach a pressure gauge to the outlet and check for leaks by shutting of the pressure source and see if the gauge drops in pressure which indicates the gas is escaping. The current knob is a glass-filled nylon part and the new proto is an aluminum piece (which would be clear anodized in it's final form). The o-rings are lubricated and made from 70 duro silicone. The aluminum proto parts leak whereas the nylon does not. I've tried several different lots of the proto parts, checked the dimensionals and surface finish, etc but they all leak at the holes closest to the grooves.

Would a nylon to silicone interface vs the 6061 aluminum be an issue? Do certain elastomers seal better with aluminum?

Any ideas or suggestions?

 

[flexibox]

It would be good if you can provide more detail such as thickness of your part. Also an assembly drawings would help but have you checked the Oring Groove depths? Have you thought of lapping the aluminium part?

 

[gorilla1969]

The mating o-rings are -001 1/2 size (.070 C.S. with .040 I.D.). The mating surface for the knob, where the o-rings make contact, has a thickness of .193" +/-.003. Essentially, the knob rests on the o-rings. For comparative purposes, if the thickness of this area was .200", a solid-on-solid condition would occur. The cavities which house the o-rings have a diameter of .145" X .025" deep. There is little compression but the existing design with the nylon works well assembly and in the field with > 96% total pass rate (inclusive of all failure modes) in assembly. This was an existing design (I wasn't involved with it) from a company we acquired about three years back.

Yes, I have attempted polishing the mating surface on the aluminum prototype to no avail.

Are certain elastimer materials better at sealing on metal than others?

 

[keyepitts]

Have you looked at the aluminum parts with a microscope in the seal contact areas? I've seen pits and inclusions in some aluminum alloy materials that would definitely be a problem if they occurred in an area where a good seal was needed.