Flames and flame suckers and eaters

[madjackghengis]

With several members busy with flame eaters, I thought I'd take some pictures and delve a bit into how they perform, their nature, and the things which seem to be stumbling blocks.  I am hoping this will be considered open for all to post in, I am putting my own thoughts in, but many others have been working on these, and have experience as well.  This is intended to be expansive, for all our edification, and not to be my own personal log.
    So, to start it off, many have asked about flame propagation, and how flames enter the cylinder, I've taken a couple of videos with this focus in mind, and will post them for observation and comment.  These were taken of my "oddball" engine, which has an iron cylinder, an iron piston, three quarter inch bore, inch and a half stroke, with a stainless steel head, ball bearings on the rod end, and on the crank shafts with a similar valve arangement to "the poppin'", and an alcohol burner.



I have the videos on photo bucket, and have pasted them in the log, but for some reason, you have to click on the picture and be taken to photo bucket to watch them.  If someone wants to school me, I'd be happy to learn how to make them appear in the log as actual videos.



these videos are with the engine started up cold, with the first one actually being the second one, so it demonstrates a warmed up engine.  The flame appears to travel at least an inch through the .375 in port, and perhaps all the way to meet the piston, with the cone of the flame pretty close to filling the port during mid-stroke, and returning to about a .750 ball once the valve closes.  It likes to sit slightly to the side, perhaps an eighth of an inch and at ten degrees of angle or so, minimizing blowing out of the flame.  It runs a consistent 450 rpm until it is fully warmed up, and then jumps to six hundred, and runs steady there with the current valve reed, which is .009 in thick, and the thickest valve I've used.
   All the reeds have developed a vacuum formed depression in the reed after a few hours of running, and need cleaning to keep ease of starting due to burnt oil buildup on the reed and the valve face of the head.  The engine has probably ten solid hours of run time behind it, so is thoroughly broken in.  Comments are invited with enthusiasm.  Cheers, Jack

[Bogstandard]

Jack,

Consider this.

I had a discussion with Jan Ridders about this proposition and he discounted it straight away, but it is still nagging in the back of my mind.

Why do they go faster when reaching a higher temperature, in fact some just won't run until a critical temp is reached?

I suggested that it wasn't a cooling effect at all, otherwise they should run from cold, the colder the better.

We all know that what happens is that supposedly the air in the cylinder expands with the heat, and supposedly when the air is cooled, the vacuum is formed and so pulls the piston back. The vacuum is actually caused not only by the hot gases cooling, but also an unknown amount of fuel burning in the cylinder and burning away the oxygen that is sucked in with it, as the oxygen is burnt with the fuel, a vacuum is automatically produced.

I put it to Jan that the engine relies more on the burning oxygen/fuel principle than the cooling down of the hot gases, so in fact should be termed an internal combustion engine rather than just a vacuum engine.

Another point that supports this theory, is where does the pressure come from that can blow out a small flame as the valve opens? There should be no pressure in there, only a vacuum, if anything it should suck the flame into the cylinder rather than blowing it away. Again, my theory is that the imbibed gases are still burning or the gas in the cylinder is stiil hot enough to cause combustion as the valve opens, and the sudden inrush of oxygen causes a sort of secondary burn in the cylinder which then 'pops' out of the inlet hole, then the returning piston starts to suck the flame and unburnt fuel in again.

Opinions anyone?


Bogs

[lordedmond]

Well John

That certainly puts a different slant on things , in 99% of the time I would concur that the flame is indeed incomplete combustion , in that case it would continue to burn ( in a older car petrol engine there is enough petrol in the exhaust to burn) , maybe that why the crud builds up inside the cylinder. so if this is the case we do need some air to get in not a total flame covering the port ?



Stuart

[cfellows]

If there is continued combustion in the cylinder, why wouldn,t there be continued net expansion as occurs in an internal combustion engine?  I would maintain that if the engine speeds up, it's probably just because things are looser and move more freely.  Regarding the exhaust gas question, I would think this is just the residual gas that was sucked in.  At some point during the return stroke the pressure equalizes then exceeds outside pressure as the swept volume continues to decrease.

Chuck

[John Hill]

Another point that supports this theory, is where does the pressure come from that can blow out a small flame as the valve opens? There should be no pressure in there, only a vacuum, if anything it should suck the flame into the cylinder rather than blowing it away.

John, my understanding is that the hot gas in the cylinder cools and reduces in volume creating a lower pressure and atmospheric pressure then pushs the piston in.  If the valve did not open at all (extreme case for demonstration purposes only) the flywheel would carry the piston past the point of equal pressures and so create a high pressure in the cyclinder.  I suggest that if the flame gets blown out in that way it may be a case for opening the valve earlier.

John

[John Hill]

Why do they go faster when reaching a higher temperature, in fact some just won't run until a critical temp is reached?

That does make me   to come up with a possible explanation and the best I can manage is to propose that the cold cylinder cools the gases too fast and the equal pressure point ocurrs long before the valve opens and the high pressure at that point stalls the action.  A fairly weak theory though..

Makes me  though, if the engine had a passive exhaust valve (I think some people put a little ball valve on top of the cylinder?) that might make a cold engine easier to start and would also avoid the puff that blows the flame out..... but it would have to be a very sensitive valve.

[Bogstandard]

Gents,

I wasn't trying to come up with the theory of how its works. I was just raising a few points that required thinking about, not this is, but what ifs.

I am absolutely positive that somewhere, it isn't just the cooling down of the hot gases that causes them to work as they do. The reason I suspect it is to do with unburnt hydrocarbons was when I was doing trials with both gas and spirit burners, the engine didn't like working with the hottest parts of the flame, which is what you would expect, but areas where unburnt fuels would be expected to be present.


John

[NickG]

Hi guys, sorry if this is a bit long winded but here are my thoughts.

I see the first theory John talked about slightly differently and it took me a while to get around to this way of thinking. I'm not sure if it's correct, but it makes more sense to me and it links a bit more to the laws of thermodynamics from what little I can remeber.

The hot gases, as you say are drawn into the cylinder - and the valve is closed. So you have in there very hot gases - and this is the important bit - at atmospheric pressure, occupying a certain volume. When the gases hit the relatively cold cylinder the gas cools and hence the pressure in the cylinder is greatly reduced to below atmospheric pressure. Hence there is an over pressure on the other side of the piston (atmospheric pressure). I believe this is why they are sometimes referred to as atmospheric engines.

One theory about why they work better once up to a certain temperature is what I've heard called the quenching effect. That is, when the engine is cold the hot gasses cool so quickly the under-pressure is created before the valve is shut, so this just causes an influx of air into the cylinder and it ends up back nearer to atmospheric pressure so there is little or no differential.

I concur with Chuck and John Hill here, the main reason for the pressure that blows out when the piston is nearing the top of its stroke, I think, is as follows. Imagine the hot gases have been sucked in, the piston is now at bottom dead centre and is just about to start its return journey. So the gases occupy the swept volume of the cylinder at atmospheric pressure, the valve is shut and the gases have rapidly cooled. There is now a pressure less than atmospheric (Note: it may not be much less than atmospheric, which is why these engines have so little power. If you think about it, if there was a perfect vacuum, the pressure difference would be 1 atmosphere 14.7 psi - which would actually give you quite a lot of power with a 3/4" or so Bore!) so the atmosphere on the other side of the piston starts to push the piston back towards top dead centre. Now the volume inside the cylinder is decreasing rapidly so the pressure is actually rising back towards atmospheric. Once it goes above atmospheric on most flame lickers the valve will be forced open as there is now an over pressure INSIDE the cylinder. That's what the popping noise is and the pressure blowing out over the flame. This is very obvious on 'poppin' and mad jack's engine and it's easy for the engine to do with the reed type valve. I guess an efficient and well designed engine would have the pressure inside and out not equalise until the piston was right near the top of it's stroke. That would take some serious maths to work out though! A lot of engines actually have a secondary spring loaded relief or exhaust valve incorporated into the design to relive that pressure, otherwise with a rigidly held valve the engine would be fighting against that pressure. Incidentally, on Jan Ridders internal valve engine it has a spigot on the piston that pushes the valve open, but in actual fact, when the pressures equalise the pressure forces the valve open the first little bit until the port is cracked open then the spigot makes sure it's fully open ready to draw in the next flame.

What John said about the incomplete combustion I'd never thought about, but I must admit, even before I'd seen Chucks reply I thought exactly the same as him. If combustion was continuing, I think the pressure would be more likely to increase rather than decrease.

They are just my thoughts and from what I've read / learned over the years. As John (Bogs) said though, neither theory fully supports the some of the behaviours we see in these engines. My poppin will start almost instantly but then stop, then it will need to warm up for a few mins to run reliably! (is that just until it gets rid of moisture build up though?) Also, another strange one as Bogs mentioned, the poppin seems to prefer sucking just a bit of the flame in like Jack's seems to - but how could that make it run faster? I think like you, there must be more to it, but if it was hydrocarbons surely that would produce a higher pressure, not use what little oxygen is in there and help produce less pressure? Think it must be bloomin' complex!

By the way madjack, the engine seems to be running superbly now and I think you've optimised the valve set up by moving it upwards leaving the port unobstructed and having the cylinder at a good angle. 

Nick

[madjackghengis]

Hi all, glad to have all the different perspectives added to mine.  In watching them run, I note they all sweat moisture immediately, and postulate since the basic exhaust of an alcohol burner is water vapor and carbon dioxide, the moisture immediately mixes with any oil residue making a sticky emulsion.  When I spin my cold engine over, it spins freely, but four or five cycles of burner input has it sticking, and I can stick a cotton swab in the port and pull out yellow emulsion.  When I wash it out with WD, I immediately get high speed, the feeling of the solvent burning in addition to the alcohol, adding to the expansion during the valve open, and it generally gets hot enough no water condenses on either the valve or the head and cylinder, leaving just the oil residue.  I've seen the additional valve in some, and have seen a ball sitting by its weight to allow the cooled air to exhaust without going through the port, so I may try this with the next one I'm in the middle of, as there is definitely a volume of air left in the cylinder when it has cooled, causing the pop and blowing out flame.  The alcohol and the butane are the fuels of choice because they leave so little partially burned hydrocarbons, as well as burning at a lower temperature.  These engines need a volume of hot air, hot enough for the temperature differential to allow substantial volume change, but the volume of the hot air must essentially fill the cylinder by the time the port closes, and having a hotter flame with less volume does not accomplish this, as tested with several small butane torches, requiring a "soft flame", as John has said, to provide the volume of hot gasses.
    I would also note, what Nick said about the cam closing the valve, and allowing the spring to open it when the vacuum allows, has been a major factor in the better performance of the oddball, which has been my test piece.  I also note that when I take my Duclos flame sucker, and set it up, brass cylinder, bronze piston, bronze valve, with the tip of the flame just a quarter of an inch from the port when cold, and give it a good spin, if I keep that red, meaning fully burned, no unburned fuel left, as in the blue part, so it is almost exclusively pulled into the port, it will come up to about half speed immediately and if I carefully maintain that tip just going in, while the rest of the flame goes across the valve, pulling it further away as it speeds up, I get no condensation, no popping, and it goes from half speed up to full speed in about thirty seconds of running, and will continue at that speed until the graphite pencil rubbings are gone and it sticks, some fifteen minutes or so.
   The Duclos works well this way, but the oddball is arranged such that it doesn't have as much flame for its volume, and I can't get just the top of the flame going in, as I do with the Duclos.  I believe the designer of the "poppin" was right in his putting a thin edge on the port, to minimize temperature change there, as my head is almost a quarter inch thick, and takes some time before it quits sweating moisture.  If I fail with the Duclos, and get too much flame in say twice, it will require being spun over until warm before it will start off, and will pop, showing the valve is sticking due to the moisture wicking between it and the cylinder.  I would suggest that is where your engine is popping or quacking, John, as the cooled air is forced out against valve pressure, causing the valve to act as a reed, even as a block of graphite.  I suspect engines that are closer to "square" are more efficient as well, as long strokes mean lots of contact area between pistons and cylinders.  I've reached the point with the oddball, where the valve timing is controlling the speed, and its opening is controlled by the vacuum holding the reed until the pressure equalizes, and it is almost silent once it is hot and running at about six hundred rpm.  I appreciate all the chiming in, lots of good ideas to consider.  Cheers, Jack

[BillTodd]

Another take on it....

The flame that is sucked into the cylinder is NOT a gas, it is a plasma, a different state of matter, as its temperature drops it condenses rapidly to a gas dramatically reducing in volume . All gases have a poor conductivity so contraction due to cooling is fairly slow - the plasma loses heat by radiation (those ultra violet/blue photons carry a lot of energy away) so it condenses very rapidly.

I suspect the reason why a cold engine is difficult to start is because the plasma condenses too quickly - while the valve is still open - thus not producing the lower pressure required.

Bill

[madjackghengis]

Hi Bill, I hadn't even considered that, but it certainly sounds dead on target.  I've been long considering how my little flamesucker could achieve some 2500 revs just by pressure changes caused by heat conduction.  I had conjectured harmonics, with it acting with air as the fluid, and heat essentially being "conducted" through it to the cylinder, but the plasma is the logical explanation of how. cheers, Jack

[NickG]

Didn't know about that Bill, but it seems to make a lot of sense. Glad you agree about the condensing too quickly as well, as I said, I'd seen that mentioned before somewhere. Interesting stuff.

What I would find interesting would be the relative power outputs of two engines - 1 flame licker and 1 stirling with the same bore and stroke, same flame. I know there would be many many variables but I'm interested to know which has more power. The first relies on cooling that volume of plasma which when is first sucked in is at more or less atmospheric pressure then rapidly cooled. We know the power stroke is pretty short and that effectively the stirling engine has 2 power strokes per revolution instead of 1, but the latter relies on heating and cooling the same, changing volume of air by conduction - can this produce as high an overpressure as the cooling plasma produces an under pressure? I can't see how but then you have 2 per rev instead of just 1. Has anybody ever seen a stirling with pressure gauges on both hot and cold end? Would it even be possible.

Sorry for going off topic Madjack.

Nick

[BillTodd]

It would be an interesting experiment Nick. Both engine types have a very low efficiency, so you may just be measuring which is the lowest

We know the power stroke is pretty short and that effectively the stirling engine has 2 power strokes per revolution instead of 1, but the latter relies on heating and cooling the same, changing volume of air by conduction

The Stirling's conduction is somewhat aided by the stirring action of the displacer and, if the displacer is perforated, it also adds surface area to aid the conduction.

[madjackghengis]

Hi Nick and Bill, I don't consider it off topic, I am simply wanting those of us who are building and experimenting to put our heads together, and get better acquainted with the facts we deal with, so we improve our product.  We are all amateurs to some degree or another, and yet we all have professions which have taught us some serious science and knowledge.  We don't advance our technology by subsidizing it from government or anywhere else, but by using our noodles and trying new things, looking at the results, and seeing if there's some new facts we haven't encountered yet, or still haven't understood.  We, the kind who are on this forum, are the wave of the future technology, and our common interests improve our work.
   The "greens" of the world are dependent on the idea that the government is going to find the right place to put our money to make life better, yet history shows this has never ever happened before, so it won't happen this time either.  We have to make things better, or accept things as they are.  If we were going to do that, we would watch the tube instead of making engines and such.  The internal combustion engine is a direct relative of flame suckers, they gave rise to it, and we are investigating innovation.  Someone like us is going to come up with the next generation prime mover, and we must do what we do, for such a person to come into existence.  I like hanging out with like minded people, and our discussions are our way of gathering as engineers and discovering our own future.  It's nice to have this tool to do it across borders, languages, and cultures.  I don't want to teach everyone, I opened this column to teach a few some things I know, but even more, to elicit information from all of you out there also on this line, and opening my own eyes to things I haven't thought of.  Nick, I think you're idea is a good one, as stirlings are indeed being brought back into real world work, and it's quite possible other abandoned technology simply needs a boost from some of what's been learned since it was abandoned.  I just want to build engines, and maybe come up with something novel, but I would be quite happy if we found a way to make use of old technology.  Nothing in the "green house" of save the planet is new, all of the technology was discovered a hundred years ago.  Solar cells, windmills, electic cars, steam power, fuel cells, all of its old technology, just dusted off.  Might as well have fun while we're working, shouldn't we?  Cheers, Jack

[BillTodd]

So, to start it off, many have asked about flame propagation, and how flames enter the cylinder,

I've just been it the 'meditation' room thinking about this:

Since the plasma is conductive and moving it might be possible to improve the inlet with a ring magnet - thus forming a sort of plasma conduit (yes, I can feel the tips of my ears getting pointy as I type ;)    ) The magnet could also help to seal a steel valve sheet over the hole.

Once inside the cylinder, good highly polish surfaces would help to keep the plasma hot (by reflecting those energetic photons back into the centre) while the valve closes.

Bill

[NickG]

Hear hear Jack, nice speech for all of us like minded people.

Bill, I guess it's easier to control the efficiency of the stirling, wasting as little heat as possible must be the starting point, then it must be largely down to the conduction of the heat - as you said there are ways of improving that, I believe different gases can be used also. The actual concept and the cycle of the stirling must be quite high efficiency though? eg, low temperature differential engines can run on very small differences in temperature - i know they give little power but the fact that they can take such a small amount of energy and turn it into motion is impressive!

Sounds interesting with the magnet but the downside would be the friction it creates for the sliding valve  and you may get friction latching - the valve may stay stuck longer than required as it's opened by a light spring. That could be solved by other methods of driving the valve though.

Nick

[John Hill]

So, to start it off, many have asked about flame propagation, and how flames enter the cylinder,

I've just been it the 'meditation' room thinking about this:

Since the plasma is conductive and moving it might be possible to improve the inlet with a ring magnet - thus forming a sort of plasma conduit (yes, I can feel the tips of my ears getting pointy as I type ;)    ) The magnet could also help to seal a steel valve sheet over the hole.

Once inside the cylinder, good highly polish surfaces would help to keep the plasma hot (by reflecting those energetic photons back into the centre) while the valve closes.

Bill

Bill, you could fit a magnet or two to the flywheel with a coil to generate a pulse at the right time to energise a coil around the inlet port..

[John Hill]

There must be scope for a tuned draught tube.  If there were seperate inlet and exhaust ports a tuned tube on the exhaust would suck the last of the 'cold' gas out and enhance the sucking of hot gas/plasmas in.


I should stop having these dreams and make a basic one first...

[andyf]

Might the explanation for improved running when the engine has warmed up be simply that when the inlet port and valve it are cold, the hot gases (flame + air) are cooled down a bit as they are drawn in, but after a while the flame heats up the valve and port area so this cooling on the way in doesn't occur?

Just a thought from a simple soul....

Andy

[madjackghengis]

Hi all, good to hear all the differing ideas, and perspectives, I will say, Andy, the port area definitely warms up, and that certainly is a big factor, based on observation, and in my last engine, I did taper the port to supply the smallest contact area with the air going in, and believe the statement made by the fellow who designed the "poppin" was dead on with regard to design correctness on that.  I've taken a small butane torch, and aimed it up the port while the alky flame was running the engine, expecting to get more energy out of it, but had little effect at all.  In analysing it, I think I was also pulling in cool air via the Bernoulli effect, and I think John's idea of a tuned tube with a pressure fed burner would be effective in matching the natural pulse effect of the engine with a timed pulse flame, and would probably improve the efficiency of the engine.  I've been considering a two or four cylinder engine with two cranks, and the heads close together, so with their timing 180 out they could use a common flame being sucked back and forth, with a manifold length to provide a timing effect from the flow length and port diameter.  I've also considered a V twin with each cylinder on its own crank pin, and the timing of the two being similar to the previous example, but offset by the cylinder angle, and with a "Y" shaped manifold, with the flame flipping back and forth within the manifold from port to port.
   These flame eaters were the direct precursers for internal combustion engines, and it is so often said the primary reason for the worst of the pollution is the compression allowing the high temperatures, so if we could come close to achieving thermal efficiency with an external burner, we would eliminate the worst of the pollution, and still have the energy.  That remains the Stirling's prime attraction I think.  cheers, Jack