Compression

[Gordon]

I am building an Atkinson Differential engine and I am having a problem getting enough compression to run. I have made several pistons and rings. I have made rings using the Trimble method and been very careful to use the method and hold tolerances. When the pistons move to the right to the firing position I have leakage around the skirt of the LH piston and only low compression. Once it has reached TDC and the piston starts to move to the right it creates a very good vacuum and will hold this for several seconds.

I cannot understand why I am getting so much leakage on the compression stroke and then immediately get a good seal as soon as the piston direction changes. As I said I have made at least four different pistons and at least 8 sets of rings and nothing seems to change.

 

[minh-thanh]

Hi Gordon !

. Once it has reached TDC and the piston starts to move to the right it creates a very good vacuum and will hold this for several seconds.

I cannot understand why I am getting so much leakage on the compression stroke and then immediately get a good seal as soon as the piston direction changes. As I said I have made at least four different pistons and at least 8 sets of rings and nothing seems to change.

I made a " few " cylinders and pistons, I noticed when the cylinder was not "straight" often the cause of the phenomenon you said.

 

[Gordon]

Hi Gordon !

I made a " few " cylinders and pistons, I noticed when the cylinder was not "straight" often the cause of the phenomenon you said.

I honed the cylinder for a long time and any variation in diameter or parallel is not showing up using the measuring equipment that I have which is a snap gauge and micrometer. It seems strange because the piston creates a vacuum with only about 1/4" of travel. I have tried putting pressure on the piston through the spark plug hole and moving the piston up/down/back and forth with no apparent difference. Another person making this same engine is having the same problem and he thinks that the oscillating arms are putting uneven pressure on the piston. I thought that perhaps the ring moving from the top of the groove to the bottom of the groove was permitting a leak but I cannot get a difference in compression with moving the piston around. Between the two of us we have many hours in trying to solve this. I am sure that it is a compression problem because I can get the engine to run for a short time by squirting oil on the piston to create a better seal. The engine is marginal at best but others have gotten it to run. I am not sure how long they have gotten it to run though. A two minute YouTube video is not a true proof of success.

 

[minh-thanh]

I honed the cylinder for a long time and any variation in diameter or parallel is not showing up using the measuring equipment that I have which is a snap gauge and micrometer. It seems strange because the piston creates a vacuum with only about 1/4" of travel. I have tried putting pressure on the piston through the spark plug hole and moving the piston up/down/back and forth with no apparent difference.
I am sure that it is a compression problem because I can get the engine to run for a short time by squirting oil on the piston to create a better seal.

You are sure that it is a compression problem,. So : Check pistons, rings with cylinder , at each of their positions in cylinder with the light !!

The engine is marginal at best but others have gotten it to run. I am not sure how long they have gotten it to run though. A two minute YouTube video is not a true proof of success.

With me, the engine running 1 minute is also successful .

 

[mayhugh1]

I cannot understand why I am getting so much leakage on the compression stroke and then immediately get a good seal as soon as the piston direction changes. As I said I have made at least four different pistons and at least 8 sets of rings and nothing seems to change.

Gordon,
If the piston and cylinder are round, there's a couple other things to look for. The ring groove in the piston could be too wide allowing the ring to rock in the groove. If the ring is allowed to tilt in the groove because of too much clearance, it can make a non-circular contact with the cylinder wall and leak. I like to make this clearance .001" to .0015".
Another possibility is that a rough surface ether on the flat bottom surface of the piston or the ring groove wall itself is too rough and not allowing the two to seal. The ring needs to seal to the piston as well as to the wall of the cylinder. You may have a poor sealing surface on the bottom of the ring but a good sealing surface on the top of the ring creating the difference in sealing depending upon the direction of travel. - Terry

Here is the effect illustrated for a conventional IC engine piston/cylinder combination:

 

[Gordon]

Gordon,
If the piston and cylinder are round, there's a couple other things to look for. The ring groove in the piston could be too wide allowing the ring to rock in the groove. If the ring is allowed to tilt in the groove because of too much clearance, it can make a non-circular contact with the cylinder wall and leak. I like to make this clearance .001" to .0015".
Another possibility is that a rough surface ether on the flat bottom surface of the piston or the ring groove wall itself is too rough and not allowing the two to seal. The ring needs to seal to the piston as well as to the wall of the cylinder. You may have a poor sealing surface on the bottom of the ring but a good sealing surface on the top of the ring creating the difference in sealing depending upon the direction of travel. - Terry

I have about .001 to .0015 side clearance in the groove. In the Gingery book on this engine he ended up running the engine for up to 30 hours using an electric motor to turn it over. I have not run it that long. Maybe I should just hook it to an electric motor and let it turn over for a couple of days. It seems like that is extreme.

 

[dsage]

Gordon:
Are you abandoning your other thread.
https://www.homemodelenginemachinist.com/threads/atkinson-frustrations.30272/
I've been waiting to see if you guys made some progress.
Or perhaps this is a side tour on one of your many issues.
I hope you make it work. I want to build that engine.
Thanks for posting.

 

[Gordon]

Gordon:
Are you abandoning your other thread.
https://www.homemodelenginemachinist.com/threads/atkinson-frustrations.30272/
I've been waiting to see if you guys made some progress.
Or perhaps this is a side tour on one of your many issues.
I hope you make it work. I want to build that engine.
Thanks for posting.

I have not actually abandoned the thread but since low compression is the primary cause of not running I was looking for input which is not specific to the Atkinson. Ray and I have been in regular contact off list and we are both experiencing the same frustrations. This engine is so marginal that any slight problem makes the difference between running or not running. For instance adjusting the carburetor just a couple of thousands goes from running to not running. Periodically both Ray and I will get the engine running for a while but then the engine quits and will not start again until suddenly it will start and run for a minute or two. Ray has had his engine running for over 50 minutes but still cannot get it to run consistently. This is without a doubt the most frustrating engine I have ever worked on. At this point it is a matter of being stubborn and getting this ****** running. It is a very marginal design. I am sure that it would never have become a viable design.

 

[dsage]

I have not actually abandoned the thread but since low compression is the primary cause of not running I was looking for input which is not specific to the Atkinson. Ray and I have been in regular contact off list and we are both experiencing the same frustrations. This engine is so marginal that any slight problem makes the difference between running or not running. For instance adjusting the carburetor just a couple of thousands goes from running to not running. Periodically both Ray and I will get the engine running for a while but then the engine quits and will not start again until suddenly it will start and run for a minute or two. Ray has had his engine running for over 50 minutes but still cannot get it to run consistently. This is without a doubt the most frustrating engine I have ever worked on. At this point it is a matter of being stubborn and getting this ****** running. It is a very marginal design. I am sure that it would never have become a viable design.

Thanks for the update and for being the guinea pig on this one.
I have seen a few running at shows. But I keep forgetting to ask what they did. Maybe they wouldn't say anway.
Keep us posted on both threads.
Thanks

 

[Gordon]

Thanks for the update and for being the guinea pig on this one.
I have seen a few running at shows. But I keep forgetting to ask what they did. Maybe they wouldn't say anway.
Keep us posted on both threads.
Thanks

I have never seen one running, just YouTube videos. I would actually love to talk to someone who actually has gotten one to run. The YouTube videos show them running for a minute or two and I wonder how many of them will only run for short periods.

 

[dsage]

Good question. I saw one a couple of years ago at Cabin Fever. It wasn't running but I assume he had it running - or what's the point of bringing it. I did ask him about some issue I found in the drawings but I'd forgotten exactly what it was. He did confirm something was not right. I guess he fixed "it".

 

[Gordon]

Good question. I saw one a couple of years ago at Cabin Fever. It wasn't running but I assume he had it running - or what's the point of bringing it. I did ask him about some issue I found in the drawings but I'd forgotten exactly what it was. He did confirm something was not right. I guess he fixed "it".

The location of the pin in the oscillating arm where the connecting rod is attached is not shown correctly. I had to compensate by adjusting the location in the connecting rod. It was easier to make a new connecting rod than it was to make a new oscillating arm.

 

[dsage]

The location of the pin in the oscillating arm where the connecting rod is attached is not shown correctly. I had to compensate by adjusting the location in the connecting rod. It was easier to make a new connecting rod than it was to make a new oscillating arm.

Interesting. I wish I'd made better notes of what I found when I drew it in CAD (which is how I discovered the design was a bit sketchy so never proceeded to build it.)
When you get it all working would you mind detailing all that you discovered?
Thanks

 

[Gordon]

Gordon,
If the piston and cylinder are round, there's a couple other things to look for. The ring groove in the piston could be too wide allowing the ring to rock in the groove. If the ring is allowed to tilt in the groove because of too much clearance, it can make a non-circular contact with the cylinder wall and leak. I like to make this clearance .001" to .0015".
Another possibility is that a rough surface ether on the flat bottom surface of the piston or the ring groove wall itself is too rough and not allowing the two to seal. The ring needs to seal to the piston as well as to the wall of the cylinder. You may have a poor sealing surface on the bottom of the ring but a good sealing surface on the top of the ring creating the difference in sealing depending upon the direction of travel. - Terry

Here is the effect illustrated for a conventional IC engine piston/cylinder combination:

View attachment 109341

I think that your sketch shows the problem. The piston is moving to the left and reverses direction to the right just before firing so in order to seal the ring must move to the other side of the ring groove. I believe that it is leaking pressure during that transition. Unfortunately this is not a conventional IC engine. After making several pistons and rings and not noticing a difference in any setup I think that it is something other than poor fit.

 

[Gordon]

Interesting. I wish I'd made better notes of what I found when I drew it in CAD (which is how I discovered the design was a bit sketchy so never proceeded to build it.)
When you get it all working would you mind detailing all that you discovered?
Thanks

Take a look at the other discussion I started.
https://www.homemodelenginemachinist.com/threads/atkinson-differential.27487/

At this point I have a combination design. I am using a variation of another design which is similar but gives a better compression stroke. The compression stroke is the biggest problem with this engine. Compression takes place because the LH piston is moving faster than the RH piston so the LH piston stops just before the RH piston gets to the firing position. That gives a very short compression stroke part of which is a race between the two pistons.

 

[dsage]

I think that your sketch shows the problem. The piston is moving to the left and reverses direction to the right just before firing so in order to seal the ring must move to the other side of the ring groove. I believe that it is leaking pressure during that transition. Unfortunately this is not a conventional IC engine. After making several pistons and rings and not noticing a difference in any setup I think that it is something other than poor fit.

Sure it is (conventional)
Left/ right / up / down. The orientation makes no difference if ring movement and sealing is truly the problem. I suspect you have such low compression you can't afford to lose any. My CAD drawing showed almost none since the two pistons are moving away from each other most of the time. You're going to need at least 3 or 4 to 1 which means if you put a pressure gauge in the sparkplug hole you should measure at least 3 x 15 = 45 pounds. Try it.
Have you tried a leak down test? i.e put air pressure (say 10lbs) into the spark plug hole and listen to where it is leaking? If all is sealing you should hear next to nothing and the piston should be pushed down the cylinder with quite a bit of force. If you can't accomplish that there is something wrong with your rings / piston / fits etc. It's a very easy test if you make a little screw in adapter or use a rubber tip on your air hose. Give it a try. If that passes what kind of pressure do you feel if you put your finger over the spark plug hole? It needs to be a pretty obvious pressure and it should hold it for a bit and fight you turning the crank (or whatever you call the goofy mechanism).

 

[Gordon]

Sure it is (conventional)
Left/ right / up / down. The orientation makes no difference if ring movement and sealing is truly the problem. I suspect you have such low compression you can't afford to lose any. My CAD drawing showed almost none since the two pistons are moving away from each other most of the time. You're going to need at least 3 or 4 to 1 which means if you put a pressure gauge in the sparkplug hole you should measure at least 3 x 15 = 45 pounds. Try it.
Have you tried a leak down test? i.e put air pressure (say 10lbs) into the spark plug hole and listen to where it is leaking? If all is sealing you should hear next to nothing and the piston should be pushed down the cylinder with quite a bit of force. If you can't accomplish that there is something wrong with your rings / piston / fits etc. It's a very easy test if you make a little screw in adapter or use a rubber tip on your air hose. Give it a try. If that passes what kind of pressure do you feel if you put your finger over the spark plug hole? It needs to be a pretty obvious pressure and it should hold it for a bit and fight you turning the crank (or whatever you call the goofy mechanism).

I tried checking it by putting pressure through the spark plug hole but that is hard to check because as soon as the piston starts to return to the right it covers the spark plug hole but by putting oil on the piston skirt it shows bubbles when it reaches TDC and the bubbles stop once the piston travels about 1/4" back to the right which is after the spark plug hole is covered. I can also feel pressure resistance at that point. I think that is why the engine only seems to fire with the timing set to about 15° ATDC.

 

[dsage]

I'm confused by your report. To recap this engine configuration:
The pistons should be closest together when they are both on the left end of their travel (TDC)
AND
the spark plug hole should be accessible to the gap between the pistons at that time.
AND
that is where you should be firing the ignition. Any other arrangement is pointless.
That appears to be when you say you have leakage. That's not good and I guess that's your point of this thread i.e. how do you fix the ring leakage. (see way down below).
On this engine it's pointless to observe leakage at any other point since (as you say) the piston(s) can be in front of the spark plug hole or both down the cylinder.
You should be feeling push back on the flywheel when it's at TDC. If it's at any other time it's probably because the air is escaping past the piston that's in front of the spark plug hole and getting between the two pistons as they travel together to the right.
I guess the bottom line is it's most important to first achieve the mechanical conditions described above and then to fix your leaks.
Sorry I haven't addressed your ring sealing issue. I use the Trible method for making rings exactly as he describes it (no short cuts) and have no issues with rings.
BUT (now the suggestions)
Is it possible your cylinder is out of shape in the area of the spark plug hole (were it was drilled) and so you get a leakage at the worst possible time? Perhaps it needs to be lapped more? Or maybe lapping it with the hole being there the lap has "dug in" going past the hole and messed up the cylinder. Maybe your pistons and rings are fine? You do say they seal at other places in the cylinder. Maybe the cylinder is at fault.
Sorry I don't have more answer. But you have a few more things to think about.

 

[Gordon]

OK Here is my theory. See the attached.

When the two pistons are traveling to the left to TDC the ring in the LH piston is forced by friction to the right side of the ring groove and no gas can get to the bottom of the ring to force it against the cylinder wall or the sealing side of the ring groove. Just spring tension of the ring holds it against the cylinder wall. There is leakage between the ring and the cylinder wall.

Once the piston reaches TDC the crank forces the ring to the left side of the ring groove and allows gas to force the ring out and against the cylinder and back to the left of the ring groove.

Due to the design there is a short compression stroke because the LH piston is just traveling slightly faster than the RH piston and the LH piston reaches TDC shortly before the RH piston reaches TDC. About the best pressure I have been able to measure at TDC is about 12-15 PS I. Once the pistons start moving back to the right I can feel pressure build up but I cannot measure it because at that point the pistons are past the spark plug hole.

Is my theory wrong? I have been wrong many times before.

Gordon

 

[dsage]

OK Here is my theory. See the attached.

When the two pistons are traveling to the left to TDC the ring in the LH piston is forced by friction to the right side of the ring groove and no gas can get to the bottom of the ring to force it against the cylinder wall or the sealing side of the ring groove. Just spring tension of the ring holds it against the cylinder wall. There is leakage between the ring and the cylinder wall.

Once the piston reaches TDC the crank forces the ring to the left side of the ring groove and allows gas to force the ring out and against the cylinder and back to the left of the ring groove.

Due to the design there is a short compression stroke because the LH piston is just traveling slightly faster than the RH piston and the LH piston reaches TDC shortly before the RH piston reaches TDC. About the best pressure I have been able to measure at TDC is about 12-15 PS I. Once the pistons start moving back to the right I can feel pressure build up but I cannot measure it because at that point the pistons are past the spark plug hole.

Is my theory wrong? I have been wrong many times before.

Gordon

This engine might be a perfect candidate for O-rings. Then you won't have to worry about the sealing side of a ring groove or poor ring construction etc. etc. etc. or probably not even a poorly lapped cylinder. None of that with O-rings because they fit nicely in the ring grooves.
How about that idea??
The pistons are basically the same for O-rings. Just a properly sized grooves for the chosen ring. Maybe the piston grooves you have already can be modified for O-rings. That would be my next trial given the state you're in. (Of course you need to check the geometry / timing first.)
In any case, any compression you feel once the pistons are past the spark plug hole is a moot point. It does you no good. Normally in that phase the pressure from ignition would (hopefully) be pushing the pistons to the right. Sort of like how you feel suction on the power stroke of a conventional engine when you turn it by hand because the valves are still closed and there was no previous ignition / expansion to push the pistons down.