Stirlingengineforum.com

I think my magnetic stirling engine would do this if only it worked . It seems like a good idea but I wonder would it add too much friction to the engine ?

One of my ideas that I want to test is to have an array of MC Stirling engines that http://www.alice-dsl.net/bsuhrbier3/MC_easy.html but instead of having them move, they are all connected to one central power piston which drives a generator. You could then use that energy to power electro-magnets to control the displacers of all of the individual units might not be efficient enough to work though.

spadez wrote:Hi,

I have found a video on youtube. It talks about designing a mechanism which creates idealised stirling timing. I was wondering if anyone had an opinion on this. Would it really be better than simply using a sine wave?

Furthermore, if the travel of the piston is known, how easy would it be to work out the required dimensions for this shape?

This is the video in question: VIDEO LINK

Thanks for any help you can give :)

Just some thoughts, without really knowing what the actual mechanical details might be for this engine...

The term "idealized", to me, means without consideration for the realities of mechanical limitations and constraints.

I'm not sure, or at least I'm having some difficulty imagining what type of linkage (if any ?) might be used so as to duplicate such an "ideal" waveform.

Just for example. If the piston is connected to a rotating crankshaft, I would think that the only possible "wave" (i.e. piston motion plotted against time) would be a sine wave. As the piston is connected to the crankshaft and must keep moving, the piston could not simply go to TDC and linger there as in this animation.

Perhaps this would be possible in a "free piston" engine, or possibly something similar to a "Ross Yoke" linkage might better approximate such a near "square(-ish) wave".

I could see some advantage as far as the displacer. Theoretically, the displacer COULD ride on a cam of some sort. I'm not sure about the feasibility of a piston also being able to drive such a cam as seems to be suggested by the animation.

I'm very interested in the timing of a Stirling engine in general.

It seems to me that the conventional wisdom of using a 90 degree advance must have been worked out in the 1800's with a cast iron engine. Heat transfer through cast iron would be relatively slow as compared with an engine having a copper bottom displacer chamber for example. Therefore I think that the material used would require some alteration of the timing. If the displacer were also driven by a cam so as to follow a near square wave, I would think the expansion of the gas would take place much more rapidly so that the 90 degrees would be too much of an advance in timing possibly resulting in the piston being driven backwards.

I see a problem with this animation/wave form. Freezing the image at the following point in the cycle for example:
The gas in the displacer chamber at this point, presumably, would be fully expanded, yet the piston still has some distance to go ,following this wave form, before it can move. This does not seem particularly desirable or "ideal" even if it were somehow possible mechanically.

spadez wrote:Thank you so much for the reply. I am planning on experimenting with this using a cylindrical cam,...

I'm still not entirely clear in regard to how the power from the piston would be transmitted,... assuming that the intent of this would ultimately be to transmit power to a rotating crankshaft.

The displacer riding on a cylindrical cam I can understand as the displacer is relatively light and is not transmitting power.

Like I say, I'm still not sure exactly what you have in mind, but my thoughts are that perhaps the piston and displacer should be dealt with separately (i.e. the piston attached to the crank in the ordinary way and the displacer riding a cam.)

trying to minimize friction. Using this technique, I can create any waveform for the Stirling I like.

In an internal combustion engine the timing is very critical as the compressed fuel/air mixture, when ignited, expands very rapidly - but not quite instantaneously. The idea is to time the ignition so that the maximum expansion takes place as the piston is just beginning its power stroke (about 20 degrees After Top Dead Center). It takes about 30 degrees or so of rotation for the expansion to take place, so in an internal combustion engine the timing would be advanced so that ignition takes place about 10 degrees BEFORE TDC. (in a car engine this changes at increasing engine speed, generally the advance increases with increasing engine speed. In most small IC engines, like a push mower, the advance remains constant)

In a "normal" Stirling engine however, the equivalent of the "ignition" - that is, when the displacer moves the air from the cold side to the hot side of the displacer chamber - takes place relatively slowly, (during the course of 180 degrees rotation of the crankshaft). So the crank travels somewhere around 110 degrees from the time the displacer starts moving air from the cold side to the hot side to the time enough expansion has taken place to deliver "maximum" power to the piston... again, somewhere around 20 degrees after TDC - so a relatively long advance is needed, generally around 90 or so degrees Before TDC.

Putting the displacer on a cam, I believe, could very well "tighten up" the operation of a Stirling making it more comparable with an Internal combustion engine as the displacer could be made to move the air from the cold to the hot side much more quickly since it no longer needs to follow the rotation of the crankshaft.

Trying to have the piston follow the same curve as the displacer however, I think might be a mistake and I don't really see what advantage could possibly be derived from that.

You seem to be absolutely right about the screenshot though. From what I can see, it would almost be better if the flat "high" surface was removed, meaning the power piston was on the top of the downward slope, would this be correct? In which case, this is getting closer to the sine curve again?

Ideally, I think the piston should be some distance down the "slope" or about 20 degrees or so past TDC at the point of maximum expansion. This would be the "sweet spot" so to speak. You don't want the piston to be right at TDC as this would be a "straight arm" situation. (i.e. the piston would be jammed straight into the crank rather than contributing to the angular momentum.)

Again, I think you should have two different curves if this is possible with whatever your set up is. I think it is a great idea, and would be a great advantage to have the displacer follow as sharp a "square wave" as practicable while having the piston follow a regular "sine wave".

In that way the air would be moved sharply from the cold to the hot side of the displacer chamber providing what would amount to a faster "ignition" or more sudden heating of the air or gas. This should provide more concentrated power and torque to the piston. The timing would then be closer to that of an IC engine depending more upon how quickly the heat transfer can take place between whatever material your displacer chamber is made out of and the air - rather than the speed of the displacer following the crankshaft since with a cam, it would not have to follow the crank through 180 degrees to effect the air movement, but could be made to move much more quickly.

So, in this regard, material considerations would then play a greater role in timing. Here are some thermal conductivity ratings of different metals for comparison. (the higher the number the faster the heat flows through the material). This may not exactly reflect metal to air heat exchange, but rather, how fast the heat can get through the wall of the chamber to the air inside. In that sense it should reflect the potential for heat exchange, but this would also depend on surface area and the amount of turbulence.


pyrex glass: 1
stainless steel: 16
carbon steel: 54
cast iron: 55
tin: 67
nickel: 91
brass: 109
aluminum: 250
gold: 310
copper: 401
silver: 429

http://www.engineeringtoolbox.com/therm ... d_429.html

Im really interested in this, and im wondering, if any waveform was possible, would it be possible to improve on the sine wave?

I'm sure it would be - for the displacer. I can't really see how any advantage could be gained by messing with the piston though. How would it transmit power to the crankshaft otherwise ?

Even a "free piston", I think, would follow a sine wave (steady reciprocating motion). What would be the advantage in having the piston pause at the end of each stroke ? (Though I'm still not clear on what your arrangement is. I may be missing something, so don't hold me to that.)

Having the displacer move quickly and then pause could have a decided advantage. faster and more thorough heat exchange. More "dwell" time for the exchange to take place...

I think it would be really nice if the timing of this displacer cam could somehow be adjusted while the engine is running. Similar to how an IC engine timing can be adjusted while the engine is running. This would allow for dynamic testing of the torque and power output at different advance settings without having to build a whole new engine or turn a new cam or crankshaft to test various different settings.

I haven't worked out any purely mechanical means for doing that. My solution was to move the displacer with a solenoid(s) or electromagnet(s) which could then be controlled any way you like - at least for testing.

Hi,

I was wondering if you might have considered using a simple "Laminar Flow" type set of pistons (No Displacers). This would certainly simplify things. Just one wave form for the pistons and you wouldn't have to worry about the displacer timing as there is no displacer in such engines.

I've only recently become more interested in these engines as previously I had assumed that they would be very inefficient. I arrived at this conclusion with limited knowledge of the thermodynamics involved though. I had assumed that without the displacer the air in the cylinder would not have a chance to cool and this would make re-compression difficult.

After seeing some of these engines running recently though, in particular, those of a "free piston" design on You Tube, it appears that this is not a problem. Going by the thermodynamic principles I was discussing on the other thread, the gas, doing work, converts the heat into work and so in effect contracts on its own drawing the piston back in. This seems to be a lost or forgotten bit of information about how these engines operate, but also from what I've been reading about the Laminar Flow - or so-called "Thermo-acoustic" type Stirlings, they are the MOST efficient Stirling engines of all since they do not have the extra weight, drag and friction of the displacer to overcome.

I've also been thinking about a possible way you might be able to get an actual trace of the exact waveform needed for any particular engine, but you would have to build a working engine with one piston.

Set this engine up so that it drives a weighted pendulum or just a flywheel equivalent to the weight of your cylinder divided by the number of pistons the engine will have. That is, if you are going to have an engine with six pistons and the cylinder would weigh about 3 pounds, make the flywheel for the one piston engine 1/2 pound, or whatever extra you think might be needed for any external load.

You would have to guesstimate the approximate speed or RPM of the engine.

Get something like a roll of register tape and set it up like a seismograph or polygraph machine. Attach a marker to the engines connecting rod so that it will trace a line as the engine is running and the paper can be drawn through at the same time.

Something like this:
If you could do something like this, I think that would at least take some of the guesswork out of it.

Once you have your trace, just wrap the paper around your cylinder and make a transfer. The paper trace is likely to be pretty shaky and would vary with the speed at which the paper is drawn through (theoretical RPM) but it should at least provide you with some clue as to the actual wave form that would most closely match how the piston actually wants to move.