|
Using O-Rings on
Piston Valves I'm designing a 1.6" scale model of the 8 1/2" Westinghouse Cross Compound Air Compressor. The plan is to use O-Rings on the pistons, in the check valves and as seals around the piston rods. The compressor uses two piston valves. The reversing valve which is driven by the high pressure piston has two rings. The main valve which is operating by steam controlled by the reversing valve has 5 rings. The rings on the pistons valves must slide over the ports (holes) in the side of the valve cylinder. Cast iron rings are normally used on piston valves. O-Ring failures are often traced to rough mating surfaces so the natural conclusion is that O rings would fail quickly if used on a piston valve. However, if the holes were kept small relative to the O-Ring cross section, maybe they would work. For this high temperature application Viton O-Rings would normally be used. Viton is tougher than the Buna-N O-Ring one normally uses so the added toughness might help. I've never seen where anyone has use O-Rings on piston valves and I suspect the manufacturer would discourage their use in this type application. So, to settle my curiosity I decided to test some O-Rings in a situation where the cylinder has holes on the side and the O-Ring must pass over the holes. The timing of this test was related to the failure of the speed reducer on my old band saw. The saw was designed for cutting wood but an optional belt and chain driven speed reducer was added to slow the blade speed so that it could be used to saw metal. The speed reducer was worn out --- the teeth on the final drive sprocket were nearly worn off and the sleeve bearings were oval rather than round. The plan was to remove the speed reducer and make the saw run at high speed only --- great for wood. I would then give the saw to the brother-in-law and buy a new metal cutting band saw. I was able to repair the sprocket enough so that it would drive a light load. So, the O-Ring test had to be run before I gave away the motor with the saw.
The photo above shows the test setup. The output pulley was replaced by a piece of aluminum with a off center 6-32 hole that would hold the screw used as a crank pin. The rod and piston in this case are 1/2" OD cold formed steel. The cylinder is 5/8" OD brass which is clamped in a block of aluminum. The upper end of the aluminum block is held to the speed reducer by a bolt. The block is free to rotate on the bolt as necessary to keep the cylinder, piston and rod in alignment with the crank pin. The stroke is about 3/4" and the drive provides ~80 strokes per minute. This photo shows a close up of the piston, O-Ring and cylinder. These O-Rings are 1/2" OD, 5/16" ID and made of Viton. The holes in the cylinder simulating ports are 1/16" diameter. When the test is running both O-Rings are inside the cylinder for the entire stroke
The test apparatus was run for about 45minutes and then the O-Rings examined --- see photo above. Two different groove depths were used. The left O-Ring had a squeeze of about 0.008" and the one on the right has a squeeze of 0.005" (The standard squeeze is about 0.13" but that will seal many hundreds or possibly thousands of pounds. For my application at 100 psi, 0.005" squeeze is more than adequate.) The white speck on the right O-Ring is a bit of debris related to a filthy workshop and not the test. The O-Ring on the left shows a bit of wear in that the parting ridge is pretty well worn off. Neither ring shows any apparent damage. The piston is cold formed 1/2" rod that was not machined except for the grooves. The rod was polished with fine sandpaper. The pits are apparently from the rolling or forming process. I'd figured there was a good probability that the 1/2" O-Rings would be severely damaged and I'd terminate the test at this point. That didn't happen. So, the next step was to test some 5/16" OD - 3/16" O-Rings which are a better match for the compressor application. The only O-Rings I had in this size are EPDM which are not as tough as the Viton rings I'd probably use.
Photo above shows the new test cylinder for the 5/16" OD O=Ring test. The grooves cut in the outside of the cylinder helped with alignment of the ports (holes) around the cylinder. This time many more holes were drilled. The hole size was 0.055", a bit larger than I'd probably use in a working design.
The photo above shows the piston and O-Rings after about 5,000 strokes. The left O-Ring has a squeeze of 0.010" and the other five have a squeeze of 0.005"
This photo shows a close-up of the left three O-Rings after the 5,000 strokes. Again, no obvious damage.
The next step was to check the O-Rings to see if they seal. A ring was soldered around one of the sets of ports in the cylinder as shown in the photo above. The ring was supplied with 100 psi air and the cylinder was submerged in a bowl of water to check for leaks. No leaks when the piston was aligned such there was an O-Ring on each side of the pressurized ports. The next thought was whether pressure would increase the wear on the O-Rings so the test was run another hour (~ 5.000 strokes) with the one set of ports at 100 psi. The O-Rings were then examined again and there was still no observable wear.
This test was encouraging enough to build the compressor valve using O-Rings. If it turns out that the O-Ring failures are unacceptably high, I can convert it to metal rings by making new valve pistons and of course, making the rings. After running this test Ii was able to examine the parts list of a commercially available steam powered pump and found that O-Rings were used for the main valve in that design which has a number of years of successful operation .
|
. 
