PDA

View Full Version : pressure seal with press-fit parts?



darryl
09-27-2011, 01:28 AM
Never thought about this before- if you press fit parts together, say a tube over a stub, what are the chances it will be pressure-tight? Given the typical lathe-turned finish on the stub, under a microscope it's going to look like hills and valleys- same would be true of the tube ID. Press the two together and some of the highs would be mashed over, but it's unlikely that all the gappage would be filled. There might be room for the crew of Fantastic Voyage to maneuver through there. Molecules of gas might be able to make the voyage, but in what numbers, who knows.

Common sense might suggest that a perfect seal would be made, and if you put air pressure to the open end of the tube, no detectable air would pass through.

I'm turning a short stub on the end of some 1/4 inch threaded stainless rod, and pressing on some 3/16 OD stainless tubing. There will be about 1/4 inch of stub inside the tubing, and it looks like a 2 thou interference fit will work without distorting the tubing.

I'm not worried about it leaking, but I thought the question was interesting anyway. Is there a standard procedure that is used in cases like this? In my case I might opt to wet the stub with epoxy before pressing the parts together, just as insurance, but with larger parts you'd have the option of including an O-ring, or depending on materials you might use loctite. Will loctite work with stainless?

In my case, pressures could be as high as 200 psi. I have no doubt that this small diameter tubing can easily handle that- and I know the pressure will work against the 2 thou interference fit by increasing the tubings diameter a tad. I think this will be a small effect though.

Any thoughts?

Cheeseking
09-27-2011, 01:57 AM
We do something similar putting threaded fittings onto tubes. Used to braze the fittings to the tubes but that took 5 minutes each! I developed a method of staking the fittings using a hydraulic press and a carefully devised split ring die. Hard to describe w/o photos but basically compressing the fitting axially while impinging a ring groove into the fitting. It literally hourglasses the tube and mechanically locks it in place. We have tested brass fittings on SS tubes and it holds 60 psi air. In your case though I would be wary of holding 200 psi with only friction of the press fit unless the pressurized media is incompressible. (I have shot my fair share of tubes out of fittings in the pressure testing apparatus using air.) Hydro test is much safer.

lynnl
09-27-2011, 12:49 PM
I'm just guessing here, off the top of my head (top, right rear corner to be exact), but with a 2 thou interference between 1/4" pieces 200 psi would be easily contained.

I'd think that tight a fit in pieces that small would almost have to involve heat and shrinkage.

One of my proudest machining moments was during some evening classes I took early on. The project was one of those straight tap wrenches (like a Starrett), about 10" long OA.
Like this: http://www.amazon.com/L-S-Starrett-STR-91B-Straight-Handle/dp/B002RJ0O7G

On the first try I got a sliding fit for the plunger rod through the handle that would hold a tight suction when the thumb & forefinger were closed over the hole. ...again, this was a sliding fit.

Admittedly, I did do a little sanding on it, but just enough to smooth it.

Probably trivial to many of you, but a DARN BIG source of pride to me at the time. :)

Scottike
09-27-2011, 12:55 PM
I have to agree with Cheeseking, I wouldn't rely on on the press fit alone to hold the stub in place, especially if the joint would be subject to pressure surges from valves opening and closing.
Cheesking's idea of a swaging the tube with a 2 piece shaft collar on the press sounds good, perhaps even turning a shallow grove or two in the stub for the tube to be swaged into would insure a more positive lock.
If your worried about leakage, a THIN coat of soft set pipe joint compound smeared on the ID of tube (Q-tip) will insure a positive seal and provide a little lubricant when you press the parts together.

edit: A simple press fit might hold, but I wouldn't trust it without having a pressure gauge on your press to measure the press pressure while fitting the parts, and even then I think it could be worrisome down the road, always wondering when that bullet might pop out of that tube.

RWO
09-27-2011, 01:29 PM
Loctite works fine on stainless if you apply the specified primer to one surface. A dab of 609 on your stub before assembly would ensure a leak proof joint plus adding significant resistance to pull out.

RWO

vpt
09-27-2011, 02:13 PM
Stubs can't be tig'ed to the tube?

I'd as well use the loctite.

Arcane
09-27-2011, 03:28 PM
A friend of mine has a Rust Check franchise business and I've made a few new tips for his application wands. The tips are bullet shaped with a .014 slot half way through one side at a 45 degree angle with a hole drilled part way up the center from one end and turned down on the outer diameter on that end so they can be pushed into either a 5/16" tube or a 1/4" tube to a distance roughly equal to the tubing diameter.

His air powered pump puts out considerable pressure, several hundred psi, and the only thing he uses to secure the tips into the tubes is Loctite. He initially was quite concerned that he would have to Silver Solder the tips in but decided to give the Loctite a try and see if it would work, which it has.

jrude
09-27-2011, 04:59 PM
Putting some math behind it, if you turn your .250 in threaded rod down to say .200 in, you've got an area of .157 sq inches. Area multiplied by pressure is force so 200 psi * .157 sq in = 31.4 lbs.

I don't know what your press fit load will be, but if it is significantly more than 32 lbs, (4x-5x), I would not worry about it unless you are compressing air since air stores alot of energy.

The industry I used to work in was commercial marine systems. A press fit will usually be water tight but pressure rating depends on the load required to achieve the fit.

Think about freeze plugs in your cars engine. They are usually a lousy stamped part which is tapped into an iron hole or arguable precision with all the force provided by a hammer handle, and they routinely withstand 10-20 psi over a lifespan of greater than 20 years.

vpt
09-27-2011, 08:44 PM
Think about freeze plugs in your cars engine. They are usually a lousy stamped part which is tapped into an iron hole or arguable precision with all the force provided by a hammer handle, and they routinely withstand 10-20 psi over a lifespan of greater than 20 years.



With loctite. ;)

lakeside53
09-27-2011, 08:57 PM
Freeze plugs have been around a lot longer than Loctite!

vpt
09-27-2011, 10:10 PM
Freeze plugs have been around a lot longer than Loctite!



Early days they used paint as a sealant.

Evan
09-27-2011, 10:49 PM
Lead paint to be precise.

darryl
09-27-2011, 11:06 PM
Thanks for the responses. I can't tig the parts because there's no area where it would go without interfering. There will be a washer-like part placed over the stub, then the tube presses on. Almost exactly at that point an O-ring has to seal against the tube.

That bit of math- I figured it would be about 20 odd lbs at most acting to push the stub out, and when I do the math I get about 6 lbs for 200 psi. Maybe my calculations aren't right, but I like the idea of measuring the force to press these parts together. Maybe if I set something up on the bathroom scale and try to push the parts together- if I'm up over 50 lbs before it starts to go I'll call it good, then just do the job on the press.

Seems that using loctite would be a good idea, as it would also act as a lube for assembly. I suppose I should do the pressing quickly so I don't get the parts partway together and find that it's locked up.

I like the idea of swaging or some other deformation that would lock the parts together, though there won't be room for that in these little parts.

But back to the press-fit being air-tight- I suppose when you tighten a fitting on an air line, you're basically holding one surface to another with some pressure, and that 'line of contact' is enough to prevent leakage. I imagine that the contact pressures you get with that method would be in the same ballpark as pressed together parts- the press fit might actually give higher contact pressures.

jrude
09-28-2011, 12:53 PM
That bit of math- I figured it would be about 20 odd lbs at most acting to push the stub out, and when I do the math I get about 6 lbs for 200 psi. Maybe my calculations aren't right, but I like the idea of measuring the force to press these parts together. Maybe if I set something up on the bathroom scale and try to push the parts together- if I'm up over 50 lbs before it starts to go I'll call it good, then just do the job on the press.

You're right. Somehow in my first calculation I forgot to square the diameter when calculating area. :)

Corrected I get 6.3 lbs on an area of .031 square inches at 200 psi.

darryl
09-29-2011, 12:04 AM
Well, I had a revelation today- the stubs I carefully turned last night are slightly tapered. Today I carefully reamed out some collars that are to press over the stub before the tubing goes on. As I'm pressing them on, they start off tight, but become slightly loose just as they reach their seated positions. It went from about 30 lbs or so of force required to near zero. Hmm- that's what happens when the workpiece is so small in diameter that if flexes away from the cutter. The difference is very small, barely noticeable on the dial caliper, but very evident in the actual fit up.

It won't be a problem because the tubing will be a tighter fit and will hold it all together.

In hindsight, I should have used the TP grinder to bring the stubs to final size- they would have had less taper, or none. Of course, there's a trick to that too- small diameter workpieces with a large 'stick out from the chuck' to diameter ratio tend to vibrate when the grinder is working on them. You have to apply resonance damping- some fingertip pressure is usually enough. And then it's easy to burn your finger because the part gets hot very quickly when grinding on it- don't aks me how I know this-

More than one revelation today actually- don't know how I missed this, but it's actually the adjuster stub that will be taking the gas pressure, not the fit between the stub and the tubing. The only worry I would have had was the possible seepage of gas through the press-fit area. And I'm not worried about that. I'll post a few pix of these gizmos when they are functional.

And yes, it's only 6 lbs or so of force against the threads on the adjuster.