Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Ambivalent? Are you sure?

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Linear Piston Engine

Straight Up and down
  (+7, -5)
(+7, -5)
  [vote for,

"Crankshafts are inefficient devices for efficiently transferring power from the pistons to the driveline, with losses that can approach 36%. At the top of the piston stroke, where gas pressure is highest, force transfer efficiency is at its lowest, though it rises as the piston descends and the connecting rod’s leverage increases. Peak efficiency happens about 40% through the piston stroke, then drops at an exponential rate that mirrors its rise. In addition, the piston doesn’t travel a path that is strictly parallel to the bore, so an angular force equal to the pressure on the top of the piston is transferred to the cylinder wall. This increases friction, wear, and fuel consumption." -from link

Well, that doesn't sound good. It's true that in any conversion from up and down motion to rotational motion there is going to be some loss, but 36% is a little excessive. So then the Revetec engine was made, using a camshaft instead of crankshaft to control the piston. But no matter what, camshafts are still relatively heavy and produce a lot of friction. Thinking about it, the switch from crankshaft to camshaft is the piston equivalent to the advance from pushrod to overhead camshaft. So then from there where do you advance? What is the piston equivalent of linear actuated valves?

In my design, the engine is a 2-stroke (either diesel or scavenged gasoline). There are two pistons 180 degrees apart as in a boxer engine. Only in this design there is a rod that connects them together and does not flex at all. This rod has teeth on either side that mesh with two gears, one above the rod and one below. Now at this point you're probably thinking that there is no way for this engine to work; the gears would constantly be changing direction. But what if you use one gear per direction? Each gear has a freewheel for a different direction so that while one is harnessing the power, the other is freewheeling. The problem that they would both be spinning in opposit directions can easily be solved with more gearing. Combustion is what controls which way the piston is moving. Once the piston gets close enough to the head on one side, the charge is ignited and combusts. Once the opposing piston gets close enough to the head on the other side, the charge is ignited and combusts.

So what does this all mean? Efficiency! The up and down motion is being converted as efficiently as possible, since it is directly up and down always. The only limiting factors of the engine are the freewheels. With durable enough freewheels, the engine could be extremely high reving, since there would be no sideways force on the one rod to make it break, and there are so few moving parts in the engine itself. See illustration for details, any comments?

acurafan07, Oct 06 2007

Crankshafts http://www.autofiel...rticles/050302.html
Also talks about the Revetec Engine [acurafan07, Oct 06 2007]

Illustration http://i210.photobu...07/4-Strogngine.jpg
Edited from a Wikipedia picture of a conventional engine [acurafan07, Oct 06 2007]

Rack and Pinion Piston http://www.freepate...ne.com/4433649.html
You may be interested in this...acurafan [xenzag, Oct 06 2007]

Gear Lash? http://thekneeslide...-diesel-motorcycle/
Hmm... [acurafan07, Oct 06 2007]

The Original http://upload.wikim...4-Stroke-Engine.gif
This is the illustration I edited to show my idea. [acurafan07, Oct 07 2007, last modified Nov 07 2007]

Illustration 2 http://usera.imagec...rtessyAccurafan.JPG
Is this how the engine should go? [quantum_flux, Oct 07 2007]

Gnome Rotary engine(moving Block) http://www.keveney.com/gnome.html
[jhomrighaus, Oct 08 2007]

Linear engine for electricity generation http://www.lceproject.org/en/principle/
[Srimech, Oct 09 2007]

Hybrid Engine Hybrid_20Engine
This HB idea talks about using magnets on the reciprocating pistons. You can quite efficiently directly generate high-frequency Alternating Current. No need for "freewhell" gadgets which have to survive the large impact-loads of gasoline explosions in the cylinders (my BICYCLE freewheel broke under much less stress). [Vernon, Nov 07 2007]

New Rotary Engine Configuration Perfect_20Engine_20...omotive_20X_20PRIZE
rotation without gears [rotary, Apr 24 2008]

geared opposed pistons http://ebtx.com/motor/mech00.htm
// in this design there is a rod that connects them together and does not flex at all. // [rotary, Apr 25 2008]

Engine innovations http://dds78dan78.free.fr/
- compilation of significant inventions regarding engines. [rotary, Apr 25 2008]


       Your gears are going to have a very short life, not just your freewheels. There is a certain amount of play in a geartrain (which you need for the gears to work; without it they will damage one another) and every time the pistons change direction, they will pass through this play (known as backlash) and then crash into the gear again.   

       They will do this with all the force available to them from the gases which will be at or near their maximum pressure. This one impact will also have to apply enough force to reverse the direction of rotation of the gear (granted, not the shaft due to the freewheel mechanism you mentioned) and teeth will break within minutes.   

       What makes it worse is that this impact will always be applied to the same gear tooth.   

       Definitely creative, but [-] because I don't think it to be workable.
david_scothern, Oct 06 2007

       I think it's a nice idea, [+]. Don't know about the gear wear - I can imagine some arrangement whereby each of the gears would rotate continuously in the same direction, being driven for only half the time. You might need some small lateral movement (at right angles to the reciprocating connecting shaft) to engage and disengage the gears, but this would be quite small.   

       How about the connecting shaft rotating about its axis, making one quarter turn for each complete cycle of the piston pair? The shaft would have teeth in two rows along its length and, as it turned, the rows of teeth would engage first with one of the driven gears, and then the other? That way, nothing (except,inevitably, the pistons and the connecting shaft) has any abrupt change of direction?
MaxwellBuchanan, Oct 06 2007

       Given the immense variety of steam driven piston engines, this may exist in that form, but that's NOT to detract from it. I just have a feeling that with so many moving parts that any mechanical output advantage, would be swallowed with increased friction and wear.
xenzag, Oct 06 2007

       A two-stroke, I assume?   

       Why bother with rotary motion - just halfbake an appropriate linear method of transport. Like walking legs.
Ling, Oct 06 2007

       Yup, two stroke. As to the friction, I know that freewheels can be made to have very little friction when not "locked" so that far less energy would be wasted bringing the piston up (because it would be done by the other piston) and no energy would be transfered to the cylinder walls from side loading.   

       [xenzag], thanks for the link. Looks like someone beat me to this, partially. I still don't really know what a "reduced function crankshaft is", or how much sideloading and friction would be cut from the rack and pinion. But I still believe my concept would work better. And even if it is baked in steam form, It'd be neat to see it in ethanol or biodiesel form.   

       [MaxwellBuchanan], not sure I completely understand what you're proposing, but from what I do understand it sounds like it would well and probably reduce the gear wear and overall friction.   

       [david_scothern] I would have thought the exact same thing about the gears, had I not seen this link [link]. If gear lash were a problem in that engine then it wouldn't be able to run at all.
acurafan07, Oct 06 2007

       You are still trying to convert reciprocating motion into rotary. The Wankel engine does that by virtue of its design.
nuclear hobo, Oct 06 2007

       If you connected two of these, one on each side of the earth, by means of a long coupler, you would create perpetual motion and solve the world's energy crisis. Almost certainly.
vincevincevince, Oct 06 2007

       I have thought up an engine design very similar to this one. But I realized that it probably wasn't so great of an idea. So, I kind of think your idea isn't that great, but I'm not going to vote against it.   

       I think that this engine would probably be nearly equal in efficiency to that of a conventional engine in converting reciprocating motion into one direction of mechanical rotation. This is because I think it would actually be changing the direction of a greater amount of mass, and it doesn't have the advantage of rotational momentum that normal engines have. (I know that at least the motion of the piston from the expanding gas also causes the opposite piston to compress it's charge at the same time (which is a good thing.)
BJS, Oct 07 2007

       //[MaxwellBuchanan], not sure I completely understand what you're proposing// Sorry, I didn't explain very clearly.   

       Imagine you have your two pistons linked by a connecting rod; together, they form a sort of a "dumbell" shape, which is one solid item, yes? For the sake of mental imagery, assume that the dumbell is horizontal, and is reciprocating (along its axis) left and right.   

       OK, now imagine that the "handle" of the dumbell (the connecting shaft) has sleeve over it which is free to rotate. Call the rotation of the sleeve the "twist".   

       Now imagine a row of ridges in a line down one side of the sleeve - acting like the rack in a rack-and-pinion gear. Imagine also that there is a round gear- wheel meshing with this rack. Now, as the pistons move back and forth, the gear-wheel will be driven first clockwise, then anticlockwise. This is obviously not good: we want the gearwheel always to turn the same way.   

       So now, imagine that the sleeve twists in synch with the reciprocating motion of the dumbell. When the dumbell is moving leftwards, the sleeve twists so that the "rack" engages with the gear- wheel. When the dumbell is moving rightwards, the sleeve twists such that the rack is disengaged with the gear- wheel. Viola! The gear-wheel is now only driven during the leftward stroke of the dumbell, and is only ever driven in one direction.   

       Now imagine a second gear-wheel on the opposite side of the sleeve to the first one. This is set up so that it engages with the rack during the *rightward* stroke, and disengages during the leftward stroke.   

       Hey presto, you now have two gearwheels, one driven during the left stroke, and one during the right stroke. The two gearwheels are then each coupled to the output shaft. Each gearwheel is always driven in the same direction (during opposite parts of the cycle), and there is no reciprocating motion anywhere except for the piston set-up.   

       Obviously, you could make this more efficient for four-stroke engine, by ensuring that each gearwheel only engages during the power stroke.   

       I don't think it would be difficult to keep everything in synch and meshed, so perhaps gear wear would not be an issue.
MaxwellBuchanan, Oct 07 2007

       Might have tendency to stall with no rotating mass (like flywheel) upon initial loadup- as in race engines with lower mass flywheels[n]
the dog's breakfast, Oct 07 2007

       I have 3 engine ideas which are similar to this (ie, remove the linear to rotary conversion) but I am actually thinking of patenting them (if I ever get round to doing it). Which reminds me, what is the easiest way to register a patent or intellectual property, etc? I'd like to get them done so I can post them.   

       [acurafan07] One my designs is based on a Bourke engine, so you may want to look at that for inspiration to solve your problems, that's what started me off.
marklar, Oct 07 2007

       [BJS], but are you taking into account that the compression of the gasses in the opposing cylinder during one's power stroke would slow the speed and force of the one power stroke just prior to firing off the other side.
acurafan07, Oct 07 2007

       So, from that picture, am I correct in assuming that the cycle of energy transfer goes like this?   

       1) L-explosion -> Top Flywheel & R-compression   

       2) R-explosion -> Bot Flywheel & L-exhaust   

       3) Top Flywheel reverb -> L-intake & R-exhaust   

       4) Bot Flywheel reverb -> R-intake & L-compression   

       Both flywheels would actually be spinning in the same direction if this were the case, this is easy to see with the right hand rule. Also, the the initial position for the cams would have to look different than that, allow me to slightly modify the diagram.
quantum_flux, Oct 07 2007

       Well I was picturing it as a scavenged 2-stroke for simplicity, and feel free to edit.
acurafan07, Oct 07 2007

       Well, in that case, you could actually make do with 1 large intake port in the center part and 1 individual exhaust port at each of the ends.
quantum_flux, Oct 07 2007

       To the illustration, that is indeed how the 4-stroke version of this would work.
acurafan07, Oct 07 2007

       I don't really know how well this engine would work, maybe someone should just build a working model...
BJS, Oct 07 2007

       You could also use cams to initiate the flywheel strokes by pushing the flywheels onto the piston rod every 180 and having it spring back while the other flywheel is pushed. I like it, good one. [+]
quantum_flux, Oct 08 2007

       Thanks, and good idea about the cams. If you had cams push the flywheels on, you wouldn't even need to have them be flywheels. Since each would be lifted for the opposit direction, they could simply be gears.
acurafan07, Oct 08 2007

       Use herringbone gears.
BJS, Oct 08 2007

       I like it
BunsenHoneydew, Oct 08 2007

       Why not use a set of chain driven cogs on one way bearings so that the main shaft of the engine is centered and only has to rotate in one direction. The one way bearing cogs would be pulled from the top on one side and the bottom on the other kind of like the freewheel on a ten speed bicycle.
jhomrighaus, Oct 08 2007

       Why not have all the pistons pointing down and towards the road at the rear of the car at about 45 degrees, then they can punt the car along.
hippo, Oct 08 2007

       Even though I think this engine would self-destruct, I think it should be funded and built. Just to prove the principle. Everyone is quick to say "I really don't think this would work" but in till it could be prototyped it can't be ruled out.   

       Very intresting. [acura] must be an engineer
evilpenguin, Oct 08 2007

       [hippo] you are responsible for my first bakery-related Coke-spilling LOL. I had a mental image of the car moving in a similar manner to Pepe Le Pew (or the dragonfly from Crazy Legs Crane).
marklar, Oct 08 2007

       This idea is better than my approach to the same issue, which would be to have the pistons stationary, while the block jumps up and down.
normzone, Oct 08 2007

       Old rotary airplane engines actually had fixed pistons and rotating cylinders(block). See link for animation of this type of engine.
jhomrighaus, Oct 08 2007

       [Acura] I disagree about the backlash problem, by the way - the gears in the engine on that link don't change direction, so they don't have to move through their backlash. Hence that engine behaves itself. However, the discussion has moved on since I was last here...
david_scothern, Oct 08 2007

       [MaxwellBuchanan], an illustration would help. Your 2nd try was still hard to visualise.
doanviettrung, Oct 09 2007

       accurafan07, by "With durable enough freewheels, the engine could be extremely high reving, since there would be no sideways force on the one rod to make it break, and there are so few moving parts in the engine itself." do you mean that the output shaft rotation speed would be high, because of the gearing design? or do you mean that the pistons would be reciprocating at a high rate?
BJS, Oct 09 2007

       I was thinking the pistons would reciprocate at a high rate.
acurafan07, Oct 09 2007

       I would like to see the pistons connected via stiff rods to another set of pistons that pump water into a pressurized vessel. Then the water could be squirted onto a pelton turbine. An airspace at the top of the vessell would provide the flywheel.
nomocrow, Oct 09 2007

       //[MaxwellBuchanan], an illustration would help.// Well, sadly I am not able to post one. I can probably sketch something if anyone knows how to put it up. But it's probably not going to change lives.
MaxwellBuchanan, Oct 09 2007

       The design makes conversion to rotary motion more difficult, but if you intend to use it to generate electricity instead it could be quite promising. I've linked one such design.
Srimech, Oct 09 2007

       What mechanism prevents the pistons from smacking into the head on the other side from the one that has just fired?
the dog's breakfast, Oct 10 2007

       The air it would suck in and compress.
acurafan07, Oct 10 2007

       No, that would be overcome. There would need to be something else, or it would smash itself to bits.
the dog's breakfast, Oct 10 2007

       Else as rpm increased the stroke would have to be shortened(ignition timing) then you would get inefficiency from lesser swept volume - larger combustion volumes.   

       Edit:(rpm) of output shaft or otherwise - frequency of oscillation.
the dog's breakfast, Oct 10 2007

       [Srimech] Argh, that was one of my other linear engine ideas. Damn other people for their ingenuity.
marklar, Oct 10 2007

       Could you arrange four cylinders in a cross around the cog wheels for a four stroke cycle?
BunsenHoneydew, Oct 10 2007

       The gears simply wouldn't take it at the cycles per second of a normal IC engine. Tooth pitting and bending would be disastrous. However, who says you need to run at those speeds?
RayfordSteele, Oct 10 2007

       By "rpm", do you guys mean "reciprocations per minute"?
BJS, Oct 10 2007

       //However, who says you need to run at those speeds?//   

       So sayeth St. Irling:   

       "And the speed of thy endeavours shall be slow, and the work extracted therefrom shall be made enough by multiplication. And the heat of the sun or burning chafe may wash thy skivvies."
nomocrow, Oct 11 2007

       I just thought about this again: it might make a decent motorbike engine. It wouldn't need cams because of the two stroke design (scavenged or not) and it could be run on diesel very easily. Since there is no set stroke, once the engine has been started (which would actually be very difficult), each stroke will compress the opposing cylinder's charge until it ignites; no matter how far along in the stroke it is. Too bad I can't build some kind of model just to see the kinds of properties it would have.
acurafan07, Nov 06 2007

       I'd like to point out a fundamental issue with the whole principal of 'linear' engines. The ineffeciency of a crankshaft (or any other linear to rotary device) comes from the linear part of the process. Linear motion must reciprocate, which means accelerating masses (pistons etc.) It is this acceleration of masses which is responsible for much of the internal loads generated in an engine. The successor to the reciprocating engine will be a rotary or mostly rotary device.
Twizz, Nov 08 2007

       The linear motion in itself is not inefficient. The energy used to accelerate the mass is only wasted if the deceleration of the mass is not harnessed. This is the case with a crank where at TDC/BDC the crank forces the piston to stop.
marklar, Nov 08 2007

       If the gears freewheel in one direction (presumeably through one-way bearings), then what is to keep the momentum of the piston from carrying it into the cylinder head during the compression stroke? Also, how will the engine be started? Turning the crankshaft manually will not move the pistons because it will be turning in the "freewheeling" direction.
chemphd, Apr 24 2008

       [chemphd], how about not taking much time evaluating glitches but instead look at my rotary engine, then ask me questions as much as you can later. See link.
rotary, Apr 24 2008

       [rotary] I'm not quite sure you know how things work here. See the entire HB does not revolve around your one idea, even if you think it should, and users are free to asks questions about other ideas that do not relate to your rotary engine.   

       [chemphd], it could be started using compressed air and what stops the momentum from traveling into the head is the combustion. Once the compression cycle is far enough alone, it will compress the charge enough for it to combust and that will repeat this process.
acurafan07, Apr 24 2008

       [acurafan] but you need something in case there's too much or too little compression; in my design the pistons are moved farther apart for more compression when you have more fuel, and since it's an electric generator, the generator itself is used to help or hinder the pistons-shaft if they're out of spec.
FlyingToaster, Apr 25 2008

       Yeah I agree with that; this was a pretty primitive HCCI idea. If you think about it there's really no good way to shut it off either.
acurafan07, Apr 25 2008

       [acurafan], your engine was already fully baked by someone unless he is you. See link.   

       If you want to evaluate most of the greatest innovations in engine technology, see the link provided. (Please determine too which is the best, so that you may appreciate them.)
rotary, Apr 25 2008


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle