Atkinson frustrations

[Gordon]

BTW. I did a theoretical compression calculation from the operating volumes in my CAD drawing and I figure the compression of the engine is only 3.2:1
I might be wrong on that. But if I'm right it's going to be pretty iffy for the engine to run. Also it doesn't leave any room for leaks anywhere.
Let me know if I made a mistake.
You should be able to measure that with a pressure gauge in the spark plug hole while cranking it. Theoretically it should measure about 45lbs. Too low to be reliable IMHO.

Thank you for your input. I found at least two places where the drawings are at least incomplete. On page 81 it shows the connecting rod hole on a 2 7/16 radius line which is pretty much worthless without a location on that radius. I think that I guessed wrong and have been trying to compensate ever since with different connecting rod c/c and different link arm c/c. As I built it the pistons hit each other. Also the intake valve stem interferes with the RH pivot arm and the arm must be reworked to provide clearance.

I redrew the engine also except that I only do 2D which does not show running interference. I may have to bite the bullet and learn 3D but I am comfortable in 2D after using it for about 20 years.

I think that your take on low compression ratio and the engine only runs if everything is exactly right. That would explain the extended run in times most folks experience.

At this point I am torn between just setting it on the shelf at least for a while and keep on trying after spending all this time on it.

Gordon

 

[Gordon]

Latest results. Maybe getting closer. I lowered the fuel tank so that the carburetor has to suck the fuel from the tank. On the drawing it shows the tank up about level with the carburetor so it can actually siphon fuel once it starts flowing. I think that was flooding the chamber. I also made another upper link so that now it is 1 1/2 C/C on the right and 1 7/16 C/C on the left. This gives a little longer exhaust time and a little longer intake time. At this point it fires pretty constantly when it is turning over with the electric motor. One strange thing is that it will fire pretty regularly for a couple of minutes and the kind of fire erratically for a short period of time. If I put my finger over the carburetor intake to choke it it then runs again for a couple of minutes. This indicates that I am not getting proper vaporization in the combustion chamber which is probably due to marginal compression. I will try playing around with stronger and weaker springs.

 

[dsage]

Gordon:
I presume you have a one way clutch on your starter motor or some way of disconnecting the starter once it starts to fire? Once the engine fires you need to get the starter off the engine otherwise the starter will hold it back.
Sounds like your getting close. I did notice on the CAD animation that as-specified the time the intake port is uncovered is only about 20 deg of flywheel rotation (exceedingly short). So your efforts to extend that are probably worthwhile. Part of the problem is the intake hole into the cylinder is pretty small and the one piston covers it up very soon after the other piston starts to move to create suction. I guess your link changes will help with that?? But I guess you'd have to be careful the change does not throw off clearances at the other end??
I wonder if an oval port extending farther down the cylinder would help to extend the intake duration. A big change I know. But port timing is everything in a (sort of) two stroke engine.
The whole engine timing is pretty sketchy.
Keep us posted.
Thanks

 

[dsage]

Gordon:
I found a few minutes to play around with my 3D CAD drawing and I entered your dimensions for the two links and you are correct the right piston moves to give more intake stroke (about 90 deg) BUT I see a severe loss of compression with those dimensions. i.e. the two pistons are considerably farther apart on the compression end. (about double the gap from what I had before).
Now you must take this info with a heavy dose of salt because I could have made a mistake or mis-interpretation of the drawings.
For some reason I had both links at 1.688 C/C even though the drawings showed 1.562 (1-9/16). So I might have fiddled the values when I first drew it and saw problems to give optimal compression. But I will admit these dimensions may be why I now show a very short intake time.
This is only FWIW info. But I think your going to find changing the links might fix one problem and cause another. It only makes sense.
I'm thinking re-positioning / sizing the intake port might be the answer rather than adjusting the motion works. Certainly there is lots of possibility to extend the intake port because there is quite a long intake stroke to take advantage of. BUT there is another problem with the motion of the pistons (as I see it). On the intake motion both pistons are moving at about the same speed (together) from right to left. So in effect there is really very little vacuum created. It's not until the very end when the pistons are both near the left side that the right pistons suddenly moves more than the other causing compression.
It's all very wacky.
I'm sorry I don't have more time to fiddle with it to come up with a solution.

 

[Gordon]

Gordon:
I found a few minutes to play around with my 3D CAD drawing and I entered your dimensions for the two links and you are correct the right piston moves to give more intake stroke (about 90 deg) BUT I see a severe loss of compression with those dimensions. i.e. the two pistons are considerably farther apart on the compression end. (about double the gap from what I had before).
Now you must take this info with a heavy dose of salt because I could have made a mistake or mis-interpretation of the drawings.
For some reason I had both links at 1.688 C/C even though the drawings showed 1.562 (1-9/16). So I might have fiddled the values when I first drew it and saw problems to give optimal compression. But I will admit these dimensions may be why I now show a very short intake time.
This is only FWIW info. But I think your going to find changing the links might fix one problem and cause another. It only makes sense.
I'm thinking re-positioning / sizing the intake port might be the answer rather than adjusting the motion works. Certainly there is lots of possibility to extend the intake port because there is quite a long intake stroke to take advantage of. BUT there is another problem with the motion of the pistons (as I see it). On the intake motion both pistons are moving at about the same speed (together) from right to left. So in effect there is really very little vacuum created. It's not until the very end when the pistons are both near the left side that the right pistons suddenly moves more than the other causing compression.
It's all very wacky.
I'm sorry I don't have more time to fiddle with it to come up with a solution.

This thing is super fussy. It was firing consistently first thing thing this morning and would almost run unassisted. Suddenly it just quit firing . I played around with some adjustments and finally found that the mounting plate which mounts the cylinder to the main body had loosed up and was moving in the holes. The holes are only about 1/32 oversize but it was enough to throw things off. When mounting the cylinder assembly Gingery shows where the pistons should be while looking through the 3/16 dia holes. This is very difficult to achieve. You think that you have it and just tapping the assembly and moving up/down/left/right 1/32 throws everything off. My mounting plate looks like Swiss cheese from all of the times I thought I had it and drilled and tapped the holes only to have it off when I put it back together. I suspect that the folks who have this running just happened to find the sweet spot with everything just right and if they disassembled it and put it back together it would not run. It seems to run best with the ignition timing retarded to a few degrees past TDC.

 

[Ramoye]

This thing is super fussy. It was firing consistently first thing thing this morning and would almost run unassisted. Suddenly it just quit firing . I played around with some adjustments and finally found that the mounting plate which mounts the cylinder to the main body had loosed up and was moving in the holes. The holes are only about 1/32 oversize but it was enough to throw things off. When mounting the cylinder assembly Gingery shows where the pistons should be while looking through the 3/16 dia holes. This is very difficult to achieve. You think that you have it and just tapping the assembly and moving up/down/left/right 1/32 throws everything off. My mounting plate looks like Swiss cheese from all of the times I thought I had it and drilled and tapped the holes only to have it off when I put it back together. I suspect that the folks who have this running just happened to find the sweet spot with everything just right and if they disassembled it and put it back together it would not run. It seems to run best with the ignition timing retarded to a few degrees past TDC.

Hi Gordon,
Glad you are making some progress. What dsage said about the intake hole seems to jive with what the guy on youtube said about slotting the intake 3/16 hole to allow 90 degree rotation of the flywheel. I slotted mine until it was bumping up against the intake valve body mounting hole (about 45 degrees). Saw some improvement when I did this. Height of cylinder seemed best at 9/32 between base and bottom of 2x3 cylinder mounting plate. Have not had a chance to work on my new cylinder in a few days (been working on my house, plus had to order more 1/8 end mills for making the cooling fins). I may look into making an adjustable cylinder (to let it move left or right). Thanks Gordon, and thanks to dsage for looking into this. Maybe we will figure this thing out eventually!

 

[Gordon]

Hi Gordon,
Glad you are making some progress. What dsage said about the intake hole seems to jive with what the guy on youtube said about slotting the intake 3/16 hole to allow 90 degree rotation of the flywheel. I slotted mine until it was bumping up against the intake valve body mounting hole (about 45 degrees). Saw some improvement when I did this. Height of cylinder seemed best at 9/32 between base and bottom of 2x3 cylinder mounting plate. Have not had a chance to work on my new cylinder in a few days (been working on my house, plus had to order more 1/8 end mills for making the cooling fins). I may look into making an adjustable cylinder (to let it move left or right). Thanks Gordon, and thanks to dsage for looking into this. Maybe we will figure this thing out eventually!

I am not doing well today. Performance seems to be worse. I think the position of the cylinder is a big part of the problem. I am thinking about slotting the mounting plate to allow adjustment left and right.

I am not worrying about cooling fins at this point. If I get it running I will go back and put cooling fins on the assembly. I am not sure if they are necessary. I ran it for over an hour yesterday with the electric motor and the engine firing and it did not get excessively warm.

 

[Ramoye]

I am not doing well today. Performance seems to be worse. I think the position of the cylinder is a big part of the problem. I am thinking about slotting the mounting plate to allow adjustment left and right.

I am not worrying about cooling fins at this point. If I get it running I will go back and put cooling fins on the assembly. I am not sure if they are necessary. I ran it for over an hour yesterday with the electric motor and the engine firing and it did not get excessively warm.

I slotted my holes and tried different positions with different links. Thinking of making the cylinder slide on a block by using adjustment bolts. Since I went air cooled, it opens up more possibilities for design changes without worrying about messy water leaks.

 

[Gordon]

The more I look at this engine the more I am convinced that it marginal at best. Obviously some folks have the engine running but there are also a lot of attempts setting on a shelf or under the bench because the builder could not get the exact right combination. I am convinced that the problem is in the compression stroke. As the RH piston is advancing to compress the fuel mix the LH piston is retreating at a slightly faster rate and only for the last few degrees of rotation is the LH piston stopped while the RH piston is still advancing so there is a very short time when the fuel mix is actually being compressed into a combustible mixture. Any slight variation of the ideal state makes ignition marginal at best.

The engine concept is intriguing but it certainly is a reason why it never was a commercial success.

Gordon

 

[Gordon]

I presently am trying to learn Fusion 360 cad to try to visualize the operation and to determine what various modifications do to change the operation. Hopefully I do not just give up before I even make the initial drawings. Lots of things to learn before I can become comfortable with 3D cad. I have discovered that I can import my 2d parts into the program and make 3D parts from that. The manipulation, assembly and animation of the parts is going to be the problem.

Gordon

 

[Cogsy]

Given the amount of trouble this engine has given you I'm definitely keeping it off my 'to build' list. I'm not sure I'd have the patience to go as far as you have but I'm sure you'll work it out eventually. Good luck!

 

[Gordon]

I played around with Fusion 360 yesterday and that is probably going to be as frustrating as trying to get the engine to run. I imported the oscillating arm drawing from my 2D drawing and tried to convert it to 3D. After trying everything I could think of and looking at videos and forums I still have not figured it out. I am not sure that it is worth the frustration to learn the 3D cad. I thought that at least I would be able to take advantage of the work I had already done without starting from scratch. Obviously others have produced 3D drawings so it is possible.

Gordon

 

[bluejets]

This Lars Christensen is really good at describing how to use Fusion 360.



One thing is he originally forgot to show everyone the easy way to get around inside the drawing screen but he rectifys this in a later video.

 

[Gordon]

Looks like it is time to just put this engine on the shelf. Learning Fusion 360 is just another frustration. I do not have enough use for 3D cad to spend the hours learning it. I was hoping that I could learn enough to put together at least a rough assembly so that I could see what happens when I move parts around or make some dimensional changes but that is going to take many hours.

I tried making a temporary eccentric pivot shaft to adjust the location to see if there was a point where the compression stroke could be improved. Not enough difference to affect the operation. No matter what is done the left hand piston and the right hand piston are moving to the left at a similar rate and unless the left hand piston stops before the right hand piston reaches the end of its stroke there is not enough difference in the stroke to actually compress the fuel mix to make it combustible.

I am sure that there is a magic point where the existing design will allow a marginal running condition. I have had the engine firing while driven by an electric motor but the moon moves or the wind shifts or the engine gods frown and suddenly it will not fire at all. Something changes again and it will fire again for a few minutes.

Gordon

 

[johnmcc69]

Hi Gordon, I'm sorry to hear about all the frustrations you've been having in both the construction of the engine & learning 3D CAD.

In my years of experience with 3D CAD (Pro-Engineer), I have never had very good luck in importing 2D sketches into it. I've found it easier to start from scratch, the problem lies within the complexity of the imported 2D geometry. As you have seen, endpoints don't match up & tangent points on arcs are no longer tangent. One of the first things I learned was to keep your sketches simple, leave fillets, arc's, cuts, & holes out of your "first" sketch. Do these AFTER your first extrude/feature. The reason being, it will be SO much easier modifying the additional features in the long run. If I had a nickel for every time a complex sketch failed after modifying it...
When these features are separate features in your "tree", modifying them is as easy as a right click on the model tree.

Your first models don't need to be highly detailed for what you want them to do, leave the cosmetic stuff (rounds, fillets, etc.) Off. Make an assembly with them & see how they go together. When you get that far, you can go back at anytime & change & add features to the individual parts that make up that assembly. It's actually pretty cool.

I hope some of this helps & encourages you,

John

 

[Ramoye]

Looks like it is time to just put this engine on the shelf. Learning Fusion 360 is just another frustration. I do not have enough use for 3D cad to spend the hours learning it. I was hoping that I could learn enough to put together at least a rough assembly so that I could see what happens when I move parts around or make some dimensional changes but that is going to take many hours.

I tried making a temporary eccentric pivot shaft to adjust the location to see if there was a point where the compression stroke could be improved. Not enough difference to affect the operation. No matter what is done the left hand piston and the right hand piston are moving to the left at a similar rate and unless the left hand piston stops before the right hand piston reaches the end of its stroke there is not enough difference in the stroke to actually compress the fuel mix to make it combustible.

I am sure that there is a magic point where the existing design will allow a marginal running condition. I have had the engine firing while driven by an electric motor but the moon moves or the wind shifts or the engine gods frown and suddenly it will not fire at all. Something changes again and it will fire again for a few minutes.

Gordon

Hi Gordon,
I know exactly what you have gone through with this engine (very frustrating, and this is my first engine build). I am still trying to finish my new cylinder, but the USPS lost my package of end mills (what a mess working with the USPS). I will keep working to get it running, but thinking the design is very marginal in working. Ray

 

[Motorteb]

Hello Gordon I am not a very active member as far as posting on this forum but I have been following your frustrations with your Atkinson differential engine with some interest as this is the engine I am building at the moment and today was the day that I actually got the engine running for two half hour runs.(a bit touchy very smooth running) more work to be done yet.

I did not use my plans from the Gingery book the plans I got I’m not sure where they originated from as there is no name or information on how to build with the copied pages I have although, I did buy the Gingery book to use as a reference there are considerable differences in the two engines regarding the pivot points as I drew both of these engines in the CAD program to compare the reference points.

So I can give you no advice regarding the Gingery design I have had similar problems to you but all my problems were man-made which is a long story as I converted the plans from Imperial to metric and sometimes I read numbers back to front which resulted in me having to make eccentric’s sleeves to correct the mistakes that I had made and I must say this has worked beautifully.

It would be a shame not to finish your engine it has taken me about 2 ½ years To get to the stage I am at now, not full-time of course only spare time when available.

I will try to put some photos up of my engine I made patterns for most parts and cast them in cast-iron by local foundry.

Keep trying

Leith

 

[Gordon]

Hello Gordon I am not a very active member as far as posting on this forum but I have been following your frustrations with your Atkinson differential engine with some interest as this is the engine I am building at the moment and today was the day that I actually got the engine running for two half hour runs.(a bit touchy very smooth running) more work to be done yet.

I did not use my plans from the Gingery book the plans I got I’m not sure where they originated from as there is no name or information on how to build with the copied pages I have although, I did buy the Gingery book to use as a reference there are considerable differences in the two engines regarding the pivot points as I drew both of these engines in the CAD program to compare the reference points.

So I can give you no advice regarding the Gingery design I have had similar problems to you but all my problems were man-made which is a long story as I converted the plans from Imperial to metric and sometimes I read numbers back to front which resulted in me having to make eccentric’s sleeves to correct the mistakes that I had made and I must say this has worked beautifully.

It would be a shame not to finish your engine it has taken me about 2 ½ years To get to the stage I am at now, not full-time of course only spare time when available.

I will try to put some photos up of my engine I made patterns for most parts and cast them in cast-iron by local foundry.

Keep trying

Leith

I am convinced that the pivot points are the problem. Unfortunately I am not familiar enough with 3D cad to be able to do an animation to actually see what is happening. I spent a lot of time last week trying to learn Fusion 360 so that I could determine what would correct the situation. I even tried making some wooden pieces and tried moving stuff around but that effort did not prove to be successful.

Is there any way you could share the other design with me without violating copyright law? I really do hate to just abandon my work to this point but it is just getting too frustrating. I am starting to do some improvements on the Howell V2 I made last year. I have a tendency to get an engine running and then move on to the next project without doing the final fine tuning.

I will send you a private message with my email address and would appreciate it if you could at least share the pivot points. I can read DXF cad files.

Gordon

 

[Motorteb]

Yes Gordon I will respond to your request just might have to wait a couple of days for me to put that together for you

Leith

 

[Gordon]

Yes Gordon I will respond to your request just might have to wait a couple of days for me to put that together for you

Leith

Thanks. Maybe I can get this figured our yet.