O-ring does not remain in its groove.

[RayJohnson2]

Hi all,

I have an annoying sealing problem.
I have a test setup where a robot takes a series of plastic DUT's (Devices Under Test) one by one, and presses them against an air nozzle.
The DUT's are then subjected to an air pressure and a temperature range. The seal between the DUT and the nozzle is made by a small o-ring.
The problem is that, when the test is done and the robot removes the DUT from the air nozzle, often the o-ring sticks to the plastic DUT.
So, when the next DUT is pressed against the air nozzle, there is leakage because the o-ring is missing ...
So I should come up with a solution to keep the o-ring in its groove in the air nozzle.
Oh, and the DUT, while being pressed against the nozzle, is subjected to a temperature range of -50°C to +150°C (-58°F to +302°F).

The o-ring is quite small: inner diameter: 1.78 mm (0.07 inch), cord diameter: 1.02 mm (0.04 inch).
So a dovetail groove cross-section is pretty much impossible to tool.

I could try to glue the o-ring in the groove using silicon rubber glue (the o-ring is silicon rubber). That might work, but it will be a nuisance when the o-ring needs replacing. The remaining glue will be a mess in the groove...

For now, as a workaround, I am using a bigger o-ring, but it is actually too big for the groove so it gets damaged easily by the groove edges.

Any suggestions anyone?

 

[3DDave]

It is small enough one could make it an insert and turn it on a lathe with a belled out end and then rotary burnish it to close the straight groove into a half-tapered groove. If the end is upset too much a cleanup pass could flatten it as required. It would not take much deformation and the insert part can be installed with shaft-locking compound into a hole in the remainder of the test fixture.

 

[hydtools]

Like 3DDave's idea create a half dovetail with a sepatate piece.
Let some air flow to separate the oring from the test piece as the test piece is removed.

Ted

 

[RayJohnson2]

@3DDave.
Good idea. Do you mean something like this?

Making the chamfer on the inside would be best because it does not introduce possible leakage.
Problem is, that it is so tiny. The central hole diameter is only 1 mm (0.039").
Making the chamfer on the outside could be a cause of leakage, though.

Maybe if I would use a conical tool to deform the inner thin-walled edge like this, to create a dovetail on the inner side.

 

[RayJohnson2]

Let some air flow to separate the oring from the test piece as the test piece is removed.

Good idea, Ted.
In fact, I suggested this idea some time ago, but I forgot to follow up on this.
I will introduce the idea again.
It certainly would be the simplest solution.

 

[3DDave]

I really meant on the the OD, hence burnishing rather than jamming a cone into it - the entire cylinder with the groove as one piece rather than adding a tiny part.

O-rings in this instance are sealing on the flat faces, not the outer boundary. The outer part keeps the pressure from causing the o-ring to escape.