low pressure microvalve sealing?

[britishmg]

Does anyone have any experience in designing low pressure (40 psi water, 80 psi air) valves using small o-rings (-001, -002, -003 sizes)?
I am having off and on again leakage issues since you have to use low squeeze to be able to still move the valve (if you use min recommended of at least .007" squeeze the valve will not move since the pressures are so low.

I have had experience that the valve will leak then if I just replace the o-ring that was made from a different batch and cure date (and keep all of the other parts the same), it fixes the leak. At times though in a batch of o-rings I will see a high failure rate because of leaks.

 

[MikeHalloran]

I'll assume that you've already checked >every< tolerance of every mating part and distributed the allowable variations in accord with process capability, and set up a system to assure that the processes do stay in control.

Beyond that, there are two ways to deal with variations in small rubber parts:
1. Pay more for parts with better than commercial tolerances from a supplier who can actually do it.
2. Inspect for leakage immediately after assembly of the fewest possible parts, e.g. with a bubbler at a few inches of water column, and replace the leakers before completing the remainder of the assembly.

Deciding whether to pay, e.g., a nickel for parts with a measurable defect rate or pay a dollar for parts with a miniscule defect rate depends on your labor rate, the rate at which you can replace a leaker, and the proportion of leakers you get. The balance may change dynamically by lot or by season, so I suggest you develop both processes and both supply chains.





Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

In a former life we did a sorting process on O-rings. We used a volumetric (displacement) method with pipettes and a vacuum pump; very slow and laborious. Could've been automated if we had needed more than a few dozens at a time...