Another Atkinson Differential build

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I'm not enough of an s-spert to see a difference from the Otto cycle -- but thanks for doing the work! Excel will allow you to label your axes, so that militant axis labeling zealots won't be poised to criticize you severely, and only stop at the last moment when they see that you described them in the text.
 
When I was working on my engine, which was quite some time ago, I tried using O rings. As I remember I got the engine to fire pretty consistently while it was being turned by the electric motor but the O rings had too much friction to make the engine run on its own. The O rings did not last very long before they no longer worked. I now believe that was due to to the nipping of the ring as it passed the intake and exhaust holes in the cylinder. It is still my belief that adequate compression is the answer.

With all of this conversation I am coming close to taking the engine back off the shelf and trying again. I keep hoping that someone comes up with some answers. Note: My engine is close to the original design and uses atmospheric valves instead of mechanically actuated valves. It is also larger with 1 1/8" bore as opposed to this one with 3/4" bore.

Gordon
Hi Gordon,
As I am exploring the pressure cycle of the Atkinson Differential engine, I have been re-reading the thread, trying to extricate some extra glimmer of inspiration.... Your comment "the original design uses atmospheric valves instead of mechanically actuated valves" inspires a question or 2:
I imagine - as I have no drawings of engine - that the "chamfered" piston is actually performing the job of the valve timing by opening and closing ports to atmosphere - as appropriate to the pressure within the chamber? So I guess there is
  1. the exhaust port in the cylinder, which when exposed towards the end of the Power piston's stroke, but part-way down the stroke of the "valve" piston, permits the exhaust to vacate to atmosphere by the pressure differential between the cylinder and the atmosphere:
  2. When the cylinder pressure has then dropped to atmospheric, the non-return "automatic" valve closes on the exhaust as the valve piston continues to travel and cause a negative (to atmospheric) pressure in the chamber:
  3. Then when the valve piston traverses the intake port, the negative pressure causes the intake through the inlet "automatic" (non-return) valve.
  4. As the pistons continue, the power piston creates more negative pressure to draw-in more air through the intake, until the valve piston has traversed back up the bore to close the inlet valve.
  5. But as the pistons proceed, there is some small compression while the exhaust port is still exposed, thus allowing some mixture to vacate the cylinder via the exhaust auto-valve, before the valve piston closes the exhaust valve. - Could this be the cause of some "wet fuel" appearing in the exhaust?
  6. Following closure of the exhaust valve by the transition of the valve piston, the ensuing compression can proceed to ignition firing expansion and the whole thing continues to cycle.
Please correct any foibles or errors in my description? I hope to do a similar numerical model to develop the combustion pressure volume cycle of this design after I have a satisfactory version of the Perreault design of mini-A engine with slide valve.
If you are able to determine timings of all the events from your drawings/CAD of your engine, starting at "ignition" and at approx 3 degree intervals, then this will be a great help to me, but I realise there are hours of mindless work to do so, as the combustion chamber volume is a complex solution of off-set phased piston motions that are not necessarily sinusoidal, so if you are unable then don't loose sleep. I'll find another way.
Thanks,
K
 
I'm not enough of an s-spert to see a difference from the Otto cycle -- but thanks for doing the work! Excel will allow you to label your axes, so that militant axis labeling zealots won't be poised to criticize you severely, and only stop at the last moment when they see that you described them in the text.
Thanks Tim,
I'm making a few power-point slides to explain it with annotation of the graph. First I need to double check the picture - and make it a continuous loop at ignition. (End pressure should equal start pressure for continuous running. - This is the first cycle, where it has effectively started at atmospheric pressure, not at the end of the cycle just before ignition. - I said it was a draft "work-in-progress". - more like a "half-gale" than "draught" really!).
K
 
Just so "we" have a starting point, I found this explanation relating to the Otto cycle. I think I'll use this as a guide to preparing my explanation of the Atkinson differential engine. And yes Tim you are right in that there is little difference from the Otto cycle to my picture of the Atkinson differential engine. The differences are in the detail of the amount of expansion, valve timings and a few subtleties that I have yet to study in detail. ( - like the position of the "cross-over" point between the exhaust curve and compression curve - is it relevant?).

FYI:
Actual and Ideal Otto Cycle

Comparison of Actual and Ideal Otto Cycles
Otto engine vs. Otto cycle In this section it is shown an ideal Otto cycle in which there are a lot of assumptions differs from actual Otto cycle. The main differences between the actual and ideal Otto engine appear in the figure. In reality, the ideal cycle does not occur and there are many losses associated with each process. For an actual cycle, the shape of the pV diagram is similar to the ideal, but the area (work) enclosed by the pV diagram is always less than the ideal value. The ideal Otto cycle is based on the following assumptions:
  • Closed cycle. The largest difference between the two diagrams is the simplification of the intake and exhaust strokes in the ideal cycle. In the exhaust stroke, heat Qout is ejected to the environment, in a real engine, the gas leaves the engine and is replaced by a new mixture of air and fuel.
  • Instantaneous heat addition (isochoric heat addition). In real engines the heat addition is not instantaneous, therefore the peak pressure is not at TDC, but just after TDC.
  • No heat transfer (adiabatic)
    • Compression – The gas (fuel-air mixture) is compressed adiabatically from state 1 to state 2. In real engines, there are always some inefficiencies that reduce the thermal efficiency.
    • Expansion. The gas (fuel-air mixture) expands adiabatically from state 3 to state 4.
  • Complete combustion of fuel-air mixture.
  • No pumping work. Pumping work is the difference between the work done during exhaust stroke and the work done during intake stroke. In real cycles, there is a pressure difference between exhaust and inlet pressures.
  • No blowdown loss. Blowdown loss is caused by the early opening of exhaust valves. This results in a loss of work output during expansion stroke.
  • No blow-by loss. The blow-by loss is caused by the leakage of compressed gases through piston
  • rings and other crevices.
  • No frictional losses.
These simplifying assumptions and losses lead to the fact that the enclosed area (work) of the pV diagram for an actual engine is significantly smaller than the size of the area (work) enclosed by the pV diagram of the ideal cycle. In other words, the ideal engine cycle will overestimate the net work and, if the engines run at the same speed, greater power produced by the actual engine by around 20%.

I must credit the authors for their explanation: https://www.nuclear-power.net/nucle...ycle-otto-engine/actual-and-ideal-otto-cycle/

I hope this is of interest? (and relevant?).
K
 
Hi Gordon,
As I am exploring the pressure cycle of the Atkinson Differential engine, I have been re-reading the thread, trying to extricate some extra glimmer of inspiration.... Your comment "the original design uses atmospheric valves instead of mechanically actuated valves" inspires a question or 2:
I imagine - as I have no drawings of engine - that the "chamfered" piston is actually performing the job of the valve timing by opening and closing ports to atmosphere - as appropriate to the pressure within the chamber? So I guess there is
  1. the exhaust port in the cylinder, which when exposed towards the end of the Power piston's stroke, but part-way down the stroke of the "valve" piston, permits the exhaust to vacate to atmosphere by the pressure differential between the cylinder and the atmosphere:
  2. When the cylinder pressure has then dropped to atmospheric, the non-return "automatic" valve closes on the exhaust as the valve piston continues to travel and cause a negative (to atmospheric) pressure in the chamber:
  3. Then when the valve piston traverses the intake port, the negative pressure causes the intake through the inlet "automatic" (non-return) valve.
  4. As the pistons continue, the power piston creates more negative pressure to draw-in more air through the intake, until the valve piston has traversed back up the bore to close the inlet valve.
  5. But as the pistons proceed, there is some small compression while the exhaust port is still exposed, thus allowing some mixture to vacate the cylinder via the exhaust auto-valve, before the valve piston closes the exhaust valve. - Could this be the cause of some "wet fuel" appearing in the exhaust?
  6. Following closure of the exhaust valve by the transition of the valve piston, the ensuing compression can proceed to ignition firing expansion and the whole thing continues to cycle.
Please correct any foibles or errors in my description? I hope to do a similar numerical model to develop the combustion pressure volume cycle of this design after I have a satisfactory version of the Perreault design of mini-A engine with slide valve.
If you are able to determine timings of all the events from your drawings/CAD of your engine, starting at "ignition" and at approx 3 degree intervals, then this will be a great help to me, but I realise there are hours of mindless work to do so, as the combustion chamber volume is a complex solution of off-set phased piston motions that are not necessarily sinusoidal, so if you are unable then don't loose sleep. I'll find another way.
Thanks,
K
That is basically correct. There is no bevel on either piston. If you look at my drawing on post #205 the intake is through the hole (port) on the top and the exhaust is through the hole in the side which is further to the right. The intake valve is held in position with a weak spring which opens when there is negative pressure in the chamber and the exhaust port is covered by the RH piston. The exhaust is held in position with a stronger spring which is opened by positive pressure in the chamber. The intake valve is mounted with the flat face of the valve facing the chamber so that pressure in the RH chamber holds the valve closed and the exhaust valve is mounted with the tapered face of the valve facing the chamber so that pressure in the RH chamber forces it open. The problem is that if there is no ignition in the left chamber there is very little pressure in the right hand chamber and the exhaust valve does not have enough pressure to open.

Since I do not have 3D cad it is not really practical to try to move in 3° increments. In my 2D cad each 3° increment basically becomes a new drawing. It is possible but extremely difficult.

I hope that I did not confuse you even more.

Gordon
 
That is basically correct. There is no bevel on either piston. If you look at my drawing on post #205 the intake is through the hole (port) on the top and the exhaust is through the hole in the side which is further to the right. The intake valve is held in position with a weak spring which opens when there is negative pressure in the chamber and the exhaust port is covered by the RH piston. The exhaust is held in position with a stronger spring which is opened by positive pressure in the chamber. The intake valve is mounted with the flat face of the valve facing the chamber so that pressure in the RH chamber holds the valve closed and the exhaust valve is mounted with the tapered face of the valve facing the chamber so that pressure in the RH chamber forces it open. The problem is that if there is no ignition in the left chamber there is very little pressure in the right hand chamber and the exhaust valve does not have enough pressure to open.

Since I do not have 3D cad it is not really practical to try to move in 3° increments. In my 2D cad each 3° increment basically becomes a new drawing. It is possible but extremely difficult.

I hope that I did not confuse you even more.

Gordon
Hello, all...
I have designed and built the Atkinson Differential engine with valves on my 10" flywheel engine and it uses 2 atmospheric valves on it.... For intake and exhaust.
The Mini A a half scale version of the Atkinson Differential engine that I have designed and built it does not use atmospheric valves. I designed a single slide valve mechanism that opens and closes the ports for proper timing.
Starting from ignition the slide valve is not in play all. Once the expanding gasses have reach the full stroke the exhaust port on the slide valve is opened and stays open until the completion of the exhaust stroke. The exhaust port is then closed and the intake port on the slide valve is opened. The intake port on the slide remains open while the intake stroke is almost completely finished.
Compression starts and the cycle repeats.
That is a very simple explanation of the slide valve timing... There is some advance and lag designed in the slide valve to assure proper timing.
The intake, exhaust and spark plug port holes only provide access to the cylinder chamber if that make sense. They obviously need to be in the correct spots.
The chamfers on the piston provide a lthe head space for ignition as well as provide a smoother transfer of gasses.
One has to think of these chamfers as similar to the otto engine head space.
You also have to consider that the cylinder ports only start to seal when the rings go past them.

I hope this helps a little in understanding how my slide valve timing works on the Mini A engine.

I have put many hours in getting this design concept to work. It does work!

A side note:
The atmospheric valve engines have valve charter as they do not open and close like mechanically operated ones do.

Regards Dave
 
That is basically correct. There is no bevel on either piston. If you look at my drawing on post #205 the intake is through the hole (port) on the top and the exhaust is through the hole in the side which is further to the right. The intake valve is held in position with a weak spring which opens when there is negative pressure in the chamber and the exhaust port is covered by the RH piston. The exhaust is held in position with a stronger spring which is opened by positive pressure in the chamber. The intake valve is mounted with the flat face of the valve facing the chamber so that pressure in the RH chamber holds the valve closed and the exhaust valve is mounted with the tapered face of the valve facing the chamber so that pressure in the RH chamber forces it open. The problem is that if there is no ignition in the left chamber there is very little pressure in the right hand chamber and the exhaust valve does not have enough pressure to open.

Since I do not have 3D cad it is not really practical to try to move in 3° increments. In my 2D cad each 3° increment basically becomes a new drawing. It is possible but extremely difficult.

I hope that I did not confuse you even more.

Gordon
Thanks Gordon.
Don't bother to do incremental measurements. I'll figure it out from other data. Thanks for the clear explanation of the "atmospheric" valves.
I understand and can model that when I get the Mini A done.
I'll catchg up with you when I have some more.
K
 
Hello, all...
I have designed and built the Atkinson Differential engine with valves on my 10" flywheel engine and it uses 2 atmospheric valves on it.... For intake and exhaust.
The Mini A a half scale version of the Atkinson Differential engine that I have designed and built it does not use atmospheric valves. I designed a single slide valve mechanism that opens and closes the ports for proper timing.
Starting from ignition the slide valve is not in play all. Once the expanding gasses have reach the full stroke the exhaust port on the slide valve is opened and stays open until the completion of the exhaust stroke. The exhaust port is then closed and the intake port on the slide valve is opened. The intake port on the slide remains open while the intake stroke is almost completely finished.
Compression starts and the cycle repeats.
That is a very simple explanation of the slide valve timing... There is some advance and lag designed in the slide valve to assure proper timing.
The intake, exhaust and spark plug port holes only provide access to the cylinder chamber if that make sense. They obviously need to be in the correct spots.
The chamfers on the piston provide all the head space for ignition as well as provide a smoother transfer of gasses.
One has to think of these chamfers as similar to the Otto engine head space.
You also have to consider that the cylinder ports only start to seal when the rings go past them.

I hope this helps a little in understanding how my slide valve timing works on the Mini A engine.

I have put many hours in getting this design concept to work. It does work!

A side note:
The atmospheric valve engines have valve charter as they do not open and close like mechanically operated ones do.

Regards Dave
Thanks Dave, That's what I have had fun modelling! - I'm glad to know my understanding matches your explanation.
Incidentally. I think you have done an excellent job adding the slide valve to the Mini A. From looking at my (crude - and slightly erroneous data) on my numerical model I can see exactly why the valve is timed as it is.
I don't know how you determined the timing - but it looks very good to me. I'm just beginning to understand some of the subtleties of this engine, by seeing how the chamber pressure changes.
K
 
Time for confession, I had taken the engine apart to some extent, seemed minor at the time, then I lost this part the 'CAM ARM', yes really, it went into Mr. Murphys storage! My shop is relatively clean, and as can be seen this is not really a tiny part, regardless I'd erroneously made it original out of aluminum, so now I have to wait for the bearings to arrive before I can reassemble and start attempting to get the little guy running. Now that a new piece is made, want to bet the original shows up? At any rate the bass is a much better looking part;).
 

Attachments

  • 20201025_133917.jpg
    20201025_133917.jpg
    109.1 KB · Views: 236
  • 20201026_121814.jpg
    20201026_121814.jpg
    184.9 KB · Views: 237
Time for confession, I had taken the engine apart to some extent, seemed minor at the time, then I lost this part the 'CAM ARM', yes really, it went into Mr. Murphys storage! My shop is relatively clean, and as can be seen this is not really a tiny part, regardless I'd erroneously made it original out of aluminum, so now I have to wait for the bearings to arrive before I can reassemble and start attempting to get the little guy running. Now that a new piece is made, want to bet the original shows up? At any rate the bass is a much better looking part;).

That is a bummer Ken!
Hope you find it soon!
 
That is a bummer Ken!
Hope you find it soon!
As mentioned, a new better looking brass part is finished, just awaiting bearings.

Now working on a pressure vacuum setup, anyone have experience with 'Schrader Valves' as check valves? I am thinking that an engine such as the Atkinson Differential, which produced both pressure and vacuum in one revolution needs to have two check valves so as not to introduce pressure to the vacuum gage (Damage?), or vacuum to the pressure gage (Again, Damage?). If this line of thinking is correct, seems it ought to work on any engine, with the added benefit of being able to observe the holding capacity (time). Somebody must have been down this road before . . . . . comments?
 
As mentioned, a new better looking brass part is finished, just awaiting bearings.

Now working on a pressure vacuum setup, anyone have experience with 'Schrader Valves' as check valves? I am thinking that an engine such as the Atkinson Differential, which produced both pressure and vacuum in one revolution needs to have two check valves so as not to introduce pressure to the vacuum gage (Damage?), or vacuum to the pressure gage (Again, Damage?). If this line of thinking is correct, seems it ought to work on any engine, with the added benefit of being able to observe the holding capacity (time). Somebody must have been down this road before . . . . . comments?
Schadenfreude valves are fine. My bought compression tester uses one, and my home-made tester the same. Vacuum gauges are just back to front pressure gauges, so should be OK as well.
 
Yesterday I took my Mini-A to a mechanic friend's shop for a smoke test, (Not to see if it burns up when powered up 😏 ), he has a unit that introduces smoke under 1/2 PSI, the unit will indicate any leak down of pressure. Fortunately my Power Piston, valving and everything else are tight, unfortunately my Pump Piston leaks so that need a make over:confused:. Anyone ever make their own 'Smoke Tester'? The unit he has cost him about $1,500 several years ago, he informed me it makes the smoke from baby oil, and can detect pin hole leaks.
 
Last edited:
I'd forgotten this trick until you mentioned it. When I was a young lad, the boss used his lungs and a cigarette... but the health risks probably prohibit publication of this as safe practice! But it was amazing to see smoke emerge through a leaking valve on a cylinder head, when he blew in the port. He did get a dirty ombishure on his face! Alternatively, I was taught to fill intake or exhaust ports with paraffin or other spirit and check for dampness showing on the combustion chamber side of the valves. Spirit Doesn't work on pistons n rings.
You can make one cheaply from a bicycle pump, couple of non-return valves, some plastic pipe and a small bunch of cheap jos-sticks from the local oriental supermarket. The bicycle pump sucks-in smoke, then blows it into the cylinder through the spark-plug hole (make a simple adapter to affix the hose).
Or maybe I mis-understand the need for $1500?
Chefs use smokers for introducing smoke to food... could one of those be adapted? Just a smoke chamber bellows and pipe?
K2
 
I made one once with a NiChrome wire heating element submerged in
oil, baby oil? I forget. The hard part was the electrical feedthrough,
I had a panel mount BNC teflon insulated feed-through. The thing was made from an old pint paint can. Worked fine but the food smoker sounds
like a better idea. Hi Ken
 
Found one called Waazus portable smoke maker. Light some sawdust place in intake and turn on smaller blower and it blows smoke out of the pipe - for food or other use.
K
Ken C,
Now that created some thought! Being that the Air Intake is really small, and the Spark Plug hole is 1/4", some external source as well as portability for use on other engines point my old brain toward an aquarium pump hooked up to some device with a canister where a drop of oil can be heated to create the smoke. Interesting how one project leads to another, then to another and so on😏.
 
Instead of heating a drop of oil, I would simply have a jos-stick in a tube - the air drawn-in past the burning jos-stick will keep it burning and create quite a lot of nice smelling smoke... - Or you could buy "smoke matches"- as used by gas engineers to check the draught into domestic flues at gas fire and boiler installations? - or the humble cigarette? - Or maybe they are too expensive now? Or a teaspoon of smouldering sawdust? - It shouldn't be too hard to make a simple "burner and intake for the air-pump" for the sawdust to sit in - ingited and then the air drawn into the pump... a bit of aluminium foils should be useful?
K2
 

Latest posts

Back
Top