Machinist suggestion wanted - lathe/stock sizes not quite there.

[Captain Starfish]

I have a bit of a mismatch between what I have and what I want, and looking for a bit of advice from an experienced fitter/turner on how to address the gap.

What I want:

A cylinder of ali. At the end is a short (3mm? smooth surface, then an internal thread about 1-1.5mm pitch running 2cm into the cylinder. The end of the cylinder makes one wall of an O-ring gland for a face seal, so is also faced smooth.

A plug of ali. External thread to match the cylinder with a less short run of smooth (5mm?) finish, followed by a knurled section 3mm wider. The wall of the knurled section and that 5mm smooth run form the base and 2nd wall of the Oring gland.

Screw the cap in, those smooth sections at the end of the thread form a sliding fit, 3mm O-ring is compressed to about 60% and she's all good.

Needs to be waterproof to 150psi and will take a bit of a beating, glues to be avoided at all costs.

What I have:
Cylinder slightly larger than needed with 3mm thick wall.
Rod slightly larger than required.

So what's the problem?
I don't have enough meat on the cylinder wall to cut the thread and still leave room for the gland wall.

My approach failing better ideas:
Turn a 2cm long cylinder with an OD that forms an interference fit with the end of the main cylinder and has about a 3mm wall thickness, put the internal thread into that, hot/cold/press it into the cylinder so it sits about 3mm inside the end of the cylinder and face the main cylinder. Turn the plug's thread and gland base to fit this combination.

But: is there a better way?

Pictures tonight when I can...

[duffman]

I'm getting flashbacks of that school exercise where 1 person tells the class how to draw something, and everyone then draws something completely different to each other.

Is the ID of the non-threaded portion of the cylinder small enough to allow you to extend the plug, add a groove, and have a radial seal on the other side of the thread?

Or would chamfering the cylinder lead-in (rather than stepping it) help at all? So you then seal on 3 faces rather than 4. (Triangular seal)

Transeals O-ring Handbook is a good reference for o-ring selection and groove design. Suspect you may already be familiar with that stuff though?

Edit: Use a smaller o-ring?

[darkfibre]

The o-ring needs to run on the end face? you can't have it on the internal face (not compressed as much)?

[Captain Starfish]

Thanks, duffman - I was being rough as guts and hadn't actually looked up the numbers. Just as well: a face seal of 3mm O-ring like that needs a minimum 3.9mm gland height which is thicker than the cylinder wall so I'm going to HAVE to bulk it out somehow.

3mm is about as small as I want to go, it's a pretty standard size for diving stuff especially in the 60/70 duro because anything smaller gets extremely unforgiving of any matter caught in the gland and tend to leak.

Face seals are popular rather than the internal ones because they are much less likely to damage the o-ring during clean and lube servicing and they allow the user to visually examine for distortion and impurities whilst in place. A hell of a lot safer this way.

Two in one - top is what I want (now that I know the dimensions I need ), bottom is what I was talking about adjusted for the correct measurements. Or is there an easier way? Must admit it's tempting to just buy some solid rod and bore it out - that way I only need a cap at one end (not two) and I can go the exact size I want. On the other hand, I don't have any boring gear that will run that deep so it means $$$ in tooling which I'd rather not throw at just this job.

[darkfibre]

I have an adjustable boring head and bars, MT3 shank you can borrow if it helps.

[Captain Starfish]

Thanks - but my tailstock is MT-2

Considering this is a one-off and I already have the stock, I'm doing things a little differently than if I was going to crank them out. Production? Rod, boring bar, buy 18mm x 1.5mm and a 20mm x 1mm taps and be done with it.

If I do go ahead with this trick, ID of tube is 26mm, what should the OD of the threaded collar be for a good tight pressure proof fit? 26.2mm? More?

Thanks for help and advice so far guys, this is proving to be quite the learning experience on a machine I've had for years but only ever done very basic stuff on. (yes, I know, this is basic stuff too, but I'm finding it challenging and enjoyable )

[Old frt]

Not entirely sure what you are trying to achieve but standard stainless steel pipe fittings and ptfe may do the job

[Captain Starfish]

I'm trying to achieve a sealed aluminium canister of specific dimensions

Tube with a cap on each end. One of the caps is drilled and tapped with an 18mm x 1.5mm internal thread for a cable gland.

[emrys74]

can you give me exact drawing and dimentions of what you want?

[Old frt]

I'm trying to achieve a sealed aluminium canister of specific dimensions

Tube with a cap on each end. One of the caps is drilled and tapped with an 18mm x 1.5mm internal thread for a cable gland.

I have a M20 x 1,5 tap, dunno if a standard cable gland willl withstand 150 psi though.

[emrys74]

I have a M20 x 1,5 tap, dunno if a standard cable gland willl withstand 150 psi though.

think you can get stainless steel ones they might

[Old frt]

Thinking outside the square, a cheap pipe thread compression fitting with a nylon ferrule may do the same job.

[agrid]

If you get stuck I can ask my brother in the US is a specialist machinist who boasts he can make anything. He has some items on display at the Craftsmanship Museum (The Joe Martin Foundation Museum and Shop Facility) which coincidentally is just around the corner from the famous Carlsbad Motocross track around which I had the pleasure of doing a few laps. I came across this running minature V8.

[Captain Starfish]

Mmmmkay guys, some good ideas there. But...

This is a combination of me trying to replicate existing design, pick up the level of skill on the lathe and use some stock that I've gotten hold of. The query is more about getting ideas of how to cobble this shit together and directions to explore than getting the job done. I'd prefer to do it using the tools and materials I have at hand - if it were about just getting the thing done I would have given it to an engineering shop a month ago, paid the $300 and had it done in a week. The material I have is tube (32mm OD 3mm wall) and 39mm rod plus O rings etc and, of course, the gland. The gland in question is out of an old torch doing the same thing, it's good for the pressures and application.

Trick with fittings (I didn't mention this before) is that either they make a big step out at the head, or they make the throat of the cylinder too small for me to be able to get the batteries in and out for charging and replacing. I'd like to keep it as close to a straight cylinder sans projections as possible.

I'm liking the idea of this threaded insert (plus it gives me a chance to chuck metal shit in the freezer which is always, cough, cool.) But don't know what interference to run or whether the end result will be pressure sealed. The Big Green Book says nothing apart from I'm after a "shrink" interference fit, my hope was that someone would come up with the magical "Oh yeah, 6160 Aluminium you will get a 4/1000 expansion with heat/freeze so go .4% of the diameter difference; scrape a 0.1mm chamfer on the insert leading edge and it will go in sweet and never come loose again", or something like that

@agrid: guys like that should be able to knock this shit up in their sleep and would be laughing at the n00b. I'd love to be able to do stuff that fine but obviously have a way to go yet!

[agrid]

Yep, I wasn't thinking of getting my brother to do your job, just in case you needed some out-of-the-box thinking.

He is on some sort of retainer and when a large plastics company needs to build a mold that no-one else can figure out how to do they call him. He seems to be able to visualize 3D very well. He's made knives, Frizbees, casino chips, a small part for the space shuttle .. even a fucking machine for an adult shop (they supplied the dildo). When I raised my eyebrows he said "Hey, their money is as green as anyone else's".

I sent him the plastic thing on the right and he built a mold and made 10,000 of the things on the left.

[Old frt]

Coefficients of expansion, bung the male bit in the freezer, whack the female bit in the oven, slip em together, stuck like shit to a blanket in no time. There ya go, basic engineering.

[Captain Starfish]

23x10^-6m/m/ºC which, on a 100ºC difference in temp at 26mm means about 0.05mm expansion. Not much

That doesn't help though - that just tells me what temperature difference I need to get the pieces to fit.

What I'm stuck on is what the "same temperature" interference should be to ensure it stays put, if that makes sense. Should the sleeve's OD be 0.01mm bigger than the cylinder's ID? 0.05mm? 0.1mm?

[duffman]

A few suggestions:
-Theres should be a stop of some sort to repertedly set the compression on a face sealing o-ring.
-A good o-ring seal should fully encapsulate the o-ring. As you tighten yours, the oring will extrude out the side. Especially if over tightened.
-Don't leave the o-ring exposed to the elements. It'll collect sand, get eaten by fish etc.

I really don't think a shrink fit is the answer, but this calculator might be of some use to you. For aluminium use:
E - Youngs modulus = 90 GPa
v - poisons ratio = 0.33.
u - co-efficient of friction ~1.0

0.02mm interference would be about right. Subject to length of the insert, and how it's being loaded.
Don't put the insert into the freezer. You'll get condensation. Better to heat the cylinder by an extra ~20C

If it were me, Id take your "desired" drawing, move the o-ring down ~3mm, into a groove in the plug, and extend the non threaded portion of the cylinder ~3mm to the right.
End result being a piston type radial seal.
If I get a chance I'll draw it up at work tomorrow.

[Sean'o]

Why not just put a groove behind the thread on the plug for an o ring to sit in, and machine a 45 degree taper into the end of the cylinder as sealing face. Will allow you to use a reasonably big o ring, depending on groove size, and if you use a 1mm pitch, .5mm depth thread, the taper will give you a little over 2.5mm sealing face on the end of the cylinder. O ring is captive so you dont loose it when you take it apart, and the 45 degree screwing down over it stops it bulging.

[Captain Starfish]

Hmm. Thanks again guys. Now I'm thinking about it, there's a difference between mine and the one I'm copying - mine screws in where the others are a press fit with those luggage style cam clamps. And the twisting, now I think about it, will more than likely pop the O-ring from that seat.

The cylinder is too thin to hold those clamps, so thread it must be (I guess). Original plan is out because of this, so I give up the need to have the o-ring exposed and visible because I don't think it will work.

This means the base cap can be a sliding fit insert with a couple of O-rings in glands and a bolt or dowel (on the "outside" of the o rings) holding it in place. Piece of piss. Main cap can be a 1.5mm thread into the cylinder wall, plug to fit with a couple of o rings on the 'outside' end of the thread.

Can't see the seal once it's formed but no need for shrink fit, the install pressure on the o rings is consistent and I can even set up a pin on the cap (outside the o rings) to make sure it doesn't undo when it shouldn't.

Life just got a whole lot easier. Thank you.