Sealing a metal pin in plastic - insert molding, interference or other? 1

[audiobeer]

Some smart guys in here

Was hoping for some advice - need to seal a stainless 316 pin (for electrical measurement) against 10 bar fluid pressure in plastic injection molded part (food safe e.g. PP).

See rubbish sketch I made with a few ideas. Pros and cons:

Option 1: insert molded straight pin - may leak as plastic shrinks, and insert mold tool more money plus more time (possible danger of damage to mold). But it's quite simple.

Option 2: insert molded with labyrinth features on pin - better as more surface area and longer fluid path so more resistance to leaking. But same issues with mold cost and part cost plus danger of leaking due to shrink still maybe.

Option 3: interference fit barbed pin into molded hole after part is made - simpler mold and no issue with plastic shrinkage just need to make sure pin doesn't come loose. Needs assembly after part is made and requires good tolerance on hole and pin (latter should be ok I guess).

Option 4: any of above with epoxy bead in chamfer or circular pocket - use epoxy that expands a little.

Any thoughts? Or other options?

 

[berkshire]

Audiobeer,
You are going to have a problem with #4 on your list, PP is a low surface energy plastic, that does not readily stick to adhesives, without surface preparation, i.e. plasma etching. 3M does make a cyanoacrylate glue DP8005 that can be used, however I do not know how food safe it is.
It would appear that #3 on your list would be the better option.
B.E.

You are judged not by what you know, but by what you can do.

 

[audiobeer]

Thanks for the response. Yep - we have experimented with this we used Loctite 770 as a primer and then 403 (rather than epoxy) for the fillet - both are approved for potable water.

Why would you go for Option (3) instead of Option (2)? Is Option (3) an accepted thing to do?

Thanks!

 

[MikeHalloran]

What fluid?
How many dishwasher/autoclave cycles/washdown/flushes?

I'd be inclined to use a mechanical joint with o-rings:
inside pin extension exposed to fluid
flange
shallow spline on shank under flange
(or anti-rotation feature like hex flange sunk into molded recess in inner wall)
part wall
molded recess in exterior end of boss
o-ring
flat washer
threaded shank with a drop of RTV for locking
nut
outside pin extension for electrical termination.



Mike Halloran
Pembroke Pines, FL, USA

 

[audiobeer]

I guess you mean something like this:

The main issue is that pin is 1mm diameter. And thats quite a lot of parts if just an interference fit or insert mold will do the same job?

fluid is mainly water but sometimes a slightly alkaline cleaning solution.

Its used in a water dispensing system - so water will be going past it as people pour themselves a drink.

Thanks!

 

[MacGyverS2000]

Thinking outside the box... would a warmed pop rivet style of operation work? Pop rivet is easy to install, and a properly warmed one will give you a partial seal via melted plastic.

Of course, I make no guarantees the plastic you intend to use is a thermoplastic, nor do I guarantee it is possible to find a rivet in stainless.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[audiobeer]

It could work if I'm understanding your suggestion correctly. We will be getting the pin manufactured so can make it however we like.

Has anyone got any experience with interference fitting metal into plastic for a liquid seal?

 

[dgallup]

How much leakage can you stand? How long does it have to work? What is the environment? How many are you making? All these things will affect your choice. If you need a high degree of reliability you need some form of elastomer.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[btrueblood]

I like Mike's idea, but would reverse the position of the o-ring from what you have shown, and put it under the flange on the fluid side. Also, the nut and thread could probably be repaced with a spring washer and e-clip or similar (push nut?), thus obviating any need for anti-rotation keying.

 

[cowski]

10 bar pressure? That's like 90 meter deep water, is your PP housing going to hold up to that pressure?
I recently designed a small plastic housing for a waterproof application, we went with your option #2 (except our pin had a single flange). Note that we were dealing with much lower pressure than you, we started by designing for the IPX7 standard (1 meter deep water, submerged for 30 minutes) but management later decided that even that depth wasn't necessary for our application. However, the molded in pins did pass the 1 meter deep, 30 minute submersion test.

One other option we considered was sonic welding in the pin in a secondary operation. This link shows some threaded inserts meant to be sonic welded into the part post molding, the same could be done with a solid pin. We didn't test this option as the molder we were working with preferred the inserts so I don't know how well it would hold up under fluid pressure. Our "back up" plan was to use an o-ring (something similar to Mike's design), but it didn't come to that.

 

[MikeHalloran]

Thanks, audiobeer, you drew it just as I intended.

My thinking for doing that, or something similar, is that it's probably repairable.
My problem with the molded in or pressed in designs is that when they decide to leak, the 'gozinta' is toast.

If the gozinta is itself easily replaced and not expensive and you can provide distribution channels, then leakage around the pin may not be an issue.
If the gozinta is the major part of the product, then minor leakage becomes a major issue.



Mike Halloran
Pembroke Pines, FL, USA

 

[MacGyverS2000]

Another option I would be interested in seeing someone try...

Option 1 with a flange on the inside, but... you spin it at high speed until you get a friction melt of the plastic near the flange. Stop spinning, then pull the pin/flange outwards slightly. If you do it right, I imagine a lip of molten plastic would form around/over the edge of the flange, holding it in place for a good long time. Don't know how water-resistant the interface would be, though.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[audiobeer]

Thanks for the responses.

dgallup: zero leakage, continuously for years, its in an indoor environment, 10's of thousands. I don't agreed about an elastomer being essential - there are many different ways to seal electrical contacts e.g. for high pressure vessels (1000 + bar) which don't involve elastomers. The challenge here is more about achieving a seal in a high volume product without many parts.

btrueblood: yes I think I would also do it that way around. But for our 1mm pin it wont be possible to use this scheme sadly.

cowski: yep 10 bar. Actually we use PVDF. If you work out the force its equivalent to about 20kg on the widest area of the cavity. The 1mm pin experiences a force of about 80grams over its area. Sonic welding may be a good idea, but it may be easier and neater to do insert molding.

MikeHalloran: The 'gozinta' wont be serviceable as the whole thing gets encapsulated. So replacing internal parts isn't an option.

MacGyverS2000: Suspect that using ultrasonics would be a better approach than this - although I have heard of someone doing something similar before. Possibly a bit hit and miss?

 

[dgallup]

Nothing has zero leakage. Anytime someone suggest that it shows they are a neophyte. Solid steel tanks will still have permeation if your measurement system is accurate enough which it must be to try to get somewhere close to zero. The first thing you need is a real specification before you can design a seal.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[MikeHalloran]

Have you tested to be sure that SS is the right material for your electrode?
If you have to go to Pt, you will want to use much smaller wire, and you will want to recover used/damaged assemblies just for the Pt value.

..... which sort of leads to ...

If you're going to encapsulate the exterior connection, maybe you don't care about a little leakage, you just don't want a breach.
So mold/drill two small holes (maybe 1..2 mm apart) instead of one, and let the inner end of the electrode comprise the short loop of thin wire between the two holes. Bend the wire into a J before insertion, and twist the legs together for retention.





Mike Halloran
Pembroke Pines, FL, USA