NEW WAY TO SEAL PISTON IN CYLINDER

[Brian Rupnow]

A couple of weeks ago, when I picked up all the free gears at the tool repair place, I also picked up 3 burned out portable compressors. I discovered a new/different method of sealing a piston in a cylinder,which I thought would be very applicable to small air engines, possibly even steam engines. I know that you can use a neoprene o-ring, but I have found that o-rings create a very high frictional drag on the cylinder wall, and any advantage gained by using them to seal "compression" is offset by this high drag that they create. It appears that the compressor manufacturer made the piston, seal, and piston cap all the same diameter, then used the flat head capscrew to put just enough pressure on the neoprene seal plate to make it "squish out" a bit and provide an air tight seal with the cylinder wall. It also appears that one might be able to control this amount of "squish out" (for lack of a better technical term), by how much the screw was tightened. Now of course, this would demand some blue loctite on the flat head capscrew to prevent it from working loose, and if using this design, the cylinder bores would have to be marginally deeper to accomodate the extra piston height. I would still put in the 3 small grooves for oil retention, both to lubricate the piston in its bore, and to help seal it. Maybe I'm posting information here that has already been tried and discarded, but I thought it was a very interesting way of providing a piston seal without having to get into the advanced machining skills required to manufacture piston rings.---Brian

 

[bentprop]

I see your point,Brian,but being neoprene as well,would that seal not have the same amount of friction as an o-ring?It probably works fine in a compressor,since the piston is so much bigger.
Still,it's always worth a try.
Hans.

 

[Brian Rupnow]

Hans--The point that I was trying to make is that you would have more control about how much interferance and drag there was by how tight the screw was tightened.---Brian

 

[crankshafter]

A couple of weeks ago, when I picked up all the free gears at the tool repair place, I also picked up 3 burned out portable compressors. I discovered a new/different method of sealing a piston in a cylinder,which I thought would be very applicable to small air engines, possibly even steam engines. I know that you can use a neoprene o-ring, but I have found that o-rings create a very high frictional drag on the cylinder wall, and any advantage gained by using them to seal "compression" is offset by this high drag that they create. It appears that the compressor manufacturer made the piston, seal, and piston cap all the same diameter, then used the flat head capscrew to put just enough pressure on the neoprene seal plate to make it "squish out" a bit and provide an air tight seal with the cylinder wall. It also appears that one might be able to control this amount of "squish out" (for lack of a better technical term), by how much the screw was tightened. Now of course, this would demand some blue loctite on the flat head capscrew to prevent it from working loose, and if using this design, the cylinder bores would have to be marginally deeper to accomodate the extra piston height. I would still put in the 3 small grooves for oil retention, both to lubricate the piston in its bore, and to help seal it. Maybe I'm posting information here that has already been tried and discarded, but I thought it was a very interesting way of providing a piston seal without having to get into the advanced machining skills required to manufacture piston rings.---Brian

Hi Brian.
I will try to explane how I did it.
I used a O-ring(Viton) as piston ring in my Webster 4-cycle. I used an under sized O-ring and stretched it so its barley protrude outside the piston just to eliminate most of the drag.( Brian, and the rest of you I hope you understand what i mean, my English is terrible, I know)
The Webster has some hours running time now and with no problem at all.
Best regards CS

 

[Brian Rupnow]

Crankshafter---I understand very well what you are saying. Your English is as good or better than many people who use English as their native language. I have always found that when cutting o-ring grooves using the information given in the o-ring literature, they fit too tight. I am sure if one has the time to experiment a bit, that they can achieve a fit like you are talking about.---Brian

 

[Maryak]

( Brian, and the rest of you I hope you understand what i mean, my English is terrible, I know)

I bet your English is much better than my Norwegian.

Your explanation was easy to follow.

Best Regards
Bob

 

[shred]

Yeah, O-ring groove specs are for "no-leakage, no-moveage" use cases. Sliding fits are perfectly possible, but need different geometry. The cap-and-sheet method would work with an O-ring, if you wanted adjustability. Cutting sheet into circles can be a big nuisance sometimes too.

 

[kustomkb]

the "Hoglet" IC I am currently working on calls for O-rings instead of piston rings.

Randall Cox the designer says "after trying o-rings I'll never go back to making piston rings again. I don't have the drawings now, but I think He calls for Viton.

Great sketch Brian, The only thing I would add is to make the threaded hole blind.

 

[cfellows]

I used a plain old o-ring from the hardware store for my 1" bore hit n miss gas engine. I made the o-ring groove deep enough that the o-ring barely protrudes above the side of the piston. I get no detectable friction from the o-ring but the engine has excellent compression and fires hard.

Chuck

 

[bentprop]

Sorry,Brian,I should have looked at the drawing more closely. :-[
I see what you mean now.It's an intriguing method of providing the correct seal.
Although the skills required to make this type of seal would seem to be about the same as cutting a groove for an o-ring.Anyway,are you going to try it on a model engine?

 

[Brian Rupnow]

Sorry,Brian,I should have looked at the drawing more closely. :-[
I see what you mean now.It's an intriguing method of providing the correct seal.
Although the skills required to make this type of seal would seem to be about the same as cutting a groove for an o-ring.Anyway,are you going to try it on a model engine?

I hadn't thought of using it. I'm burned out on machining things right now. It just looked like such a novel approach that I wanted to share it---Thats all.

 

[Peter Neill]

Brian, there is another way to do it that’s even easier and less demanding of accurate sizes and fits.

A couple of months ago, you may remember I posted a question about piston/bore clearances, as I needed information for a small R&D project I was looking at for a customer. The responses I got with information were very useful, and I used these to make up some prototypes using ‘O’ rings.

However, like yourself, I found that the drag created – even with just 0.002” compression on the ‘O’ ring – was just too great. I experimented with different clearances on the piston/bore/’O’ ring ; experimented with different materials; and also with different levels of surface finish on the cylinder bore. All still had too much drag using the ‘O’ ring.
Here are some of the prototype parts I made up to test.

Then a little light came on in the brain, and I got some samples of small diameter lip seals used in hydraulic and pneumatic applications. These are also call ‘U’ seals or ‘U’ rings, and are available in various materials and hardness’s. I chose nitrile butadiene rubber (NBR) in a fairly soft grade, around 55 shore D, as I wanted plenty of conformance to the bore, rather than out and out wear resistance at this stage of development. Depending on which direction you fit them, these will expand under pressure to very tightly seal the bore, but then easily compress down to a lighter seal on the return stroke to reduce friction and effort required to move the piston. The more pressure on them, the greater the seal.

I used single acting seals, but you can also get double acting to give a high pressure seal in both directions. Maximum pressure for some of these seals is up to 400 Bar, so more than adequate for model engineering purposes. Here's a sketch I picked off the net to illustrate the seal operation:

The following pictures show the final piston shape arrived at after some experimentation, and the seal as fitted on the piston. The final pistons and cylinder sleeves are both made from gray cast iron, and the piston to bore clearance was set at 0.002”. The seal-to-bore fit at the top (largest diameter) of the seal was 0.020” oversize per side – not a typo btw – and both the sliding fit and the compression it gives are perfect. Piston diameter is around 17mm and overall piston is 9.5mm. The seals are around 2mm in height.

The prototype units are currently under test with some standard units as a benchmark, and the test results should be available from my customer in a couple of weeks. As soon as I have these I will (with their permission) publish them here as a reference, and they will include life cycles, wear rates, and load/demand curves on the 12v electric motors used to power the compressors.

All information is freely provided for anyone to use without paying R&D or royalty costs. ;D ;D

Peter

 

[Lew Hartswick]

Peter, Can you give a link or name of a manufacturer that has some detailed info.
on those on the net? If I get around to building an engine it looks like the way to go.
Thanks.
...lew...

 

[Richard Carlstedt]

One of the reasons for extra drag on O-rings is incorrect groove
configuration. You should NOT make the groove like you would for a ordinary piston ring.
proper O-ring grooves have 10 degree negative sides.
This means the bottom of the groove is "wider" than the top.
I have to admit, that it is pretty hard to make it that way in the home shop. Also the the ring should have on .010" (.25mm) compression total. do this by measuring the cross-section.
If it is .063 (~1.5mm), then your groove bottom should only be .057 from the cylinderwall to produce the .010 fit.

back to the taper walls.
the reason for the reverse taper, is two-fold.
First, the pressure tries to force the ring to extrude itself between the piston and the wall. now I know they look sturdy, but lots of pressure will squeeze the rubber into the few thousandths clearance. This increases drag and wear at the same time. By having the negative angle, increased pressures actually pull the ring back into the groove, lessening the above effect and keeping drag/wear under control..
If you must use a straight groove side wall, keep your clearances down between the piston diameter and the cylinder wall.
Rich

So the important thing fellows is the cross-section diameter !
As a earlier poster mentioned , some rings vary depending on supplier, but measuring the ring ID and OD is not your answer in many cases
Hope this helps

 

[Peter Neill]

Lew, I got these samples from this company in the UK: http://www.claron.co.uk/.

I googled for "hydraulic lip seals", found the company and took it from there. After having a chat with their techical sales guy about what I needed for size and conformity, these were recommended. The part number for them is S00261, but I can't find them on the online catalogue.
They do have a very good downloadable PDF file on seal selection and design though.

I would imagine using a similar search on Google.com instead of the Google.co.uk that I use would produce a source somewhat closer to hand over your side of the pond. However I have got a couple of spares, and if you wanted to use this particular size I'd be happy to pop them in the post to you.

Peter

 

[Peter Neill]

One of the reasons for extra drag on O-rings is incorrect groove
configuration. You should NOT make the groove like you would for a ordinary piston ring.
proper O-ring grooves have 10 degree negative sides.
This means the bottom of the groove is "wider" than the top.
I have to admit, that it is pretty hard to make it that way in the home shop. Also the the ring should have on .010" (.25mm) compression total. do this by measuring the cross-section.
If it is .063 (~1.5mm), then your groove bottom should only be .057 from the cylinderwall to produce the .010 fit.

back to the taper walls.
the reason for the reverse taper, is two-fold.
First, the pressure tries to force the ring to extrude itself between the piston and the wall. now I know they look sturdy, but lots of pressure will squeeze the rubber into the few thousandths clearance. This increases drag and wear at the same time. By having the negative angle, increased pressures actually pull the ring back into the groove, lessening the above effect and keeping drag/wear under control..
If you must use a straight groove side wall, keep your clearances down between the piston diameter and the cylinder wall.
Rich

So the important thing fellows is the cross-section diameter !
As a earlier poster mentioned , some rings vary depending on supplier, but measuring the ring ID and OD is not your answer in many cases
Hope this helps

Rich, you forgot to mention that 'O' ring grooves also need to be wider than the cross-section of the 'O' ring, so that they have room to squish into, as under pressure they deform from a circular section to more of an ovoid shape. Incidentally, on my prototypes the 'O' ring cross-section was 1.5mm, and the groove width was 1.8mm.

Another member here Chip_Monk (David) provided a a useful link to another downloadable (1.4MB) PDF design guide for O rings here: http://www.jameswalker.biz/docs/jwco/o_rings_full_doc.pdf

Peter

 

[Brian Rupnow]

I suppose that it wouldn't be unreasonable to try this method either. By doing it this way, no o-ring groove is necessary in the piston. All of the machining can be done to the cap portion.---Brian

 

[bentprop]

Those seals Peter mentions look very much like the type used in automotive brake cylinders.This would mean there are plenty of sources of supply,and relatively cheap.Although I don't recall ever seeing anything smaller than about 5/8th in diameter.