We've all seen those radial engines on old airplanes, the round ones with several cylinders aranged in a circle around a common crank. The problem with these is that they're about as smooth as a Harley engine and they require large counterweights to ballance them.
I propose keeping the radial layout, but with a different crank design. There will be three cylinders joined at the crank case but with 120 degrees between each other. Instead of a single crank, there will be three of them, also 120 degrees appart. All the pistons will go "down" (towards the center of the engine) at the same time, eliminating the need for balancing weights (much like a boxer engine). Valve timing could be by ring cams off the crankshaft or by belts and dohc. Otherwise, the heads and cylinders would be similar to other V-Twin engines.
This design would work well for a motorcycle because of its small size and triangular shape. A centrifugal supercharger could be easily mounted on one end of the crankshaft, maybe with planetary gears to change its speed. It would suck air from a carbuerator or throttle body in-line with the crankshaft and throw it out and into the cylinder heads.
Since all three cylinders go "down" at the same time, two would always be on the same cycle. To make the firing order smoother, another cylinder bank could be put on the back of the three cylinder, making a "Y-6" engine. The two banks would be 180 degrees out of phase, making it much smoother. The heads would be like an inline two cylinder. This design would be more suitable for a car.
I believe someone on this forum suggested the idea of having four pistons converge on a common head. You could use the crank case of my three cylinder in a similar way, but this would lead to lubrication issues, and the cranks would get in the way of the flame front.
A more feasable plan would be to use the converging pistons and a one-way valve to pump oil; or as some kind of supercharger. This second idea would have the added benefit of evacuating the crankcase fumes.
Those are all the ideas I can think of right now. I have a CAD drawing of the engine somewhere, but no server to put it on. Since the idea is hard to visualize, I would appreciate it greatly if someone drew an illustration for me.-- discontinuuity, Jun 01 2005 and then there was the Ariel Four Square http://www.wemoto.c.../52ariel1000sq4.jpg [normzone, Jun 01 2005] (?) Feuling W3 http://www.randyscy...es.htm#Fueling%20W33 cyl radial motorcycle [CharlieB, Jul 24 2005] X4 Engine http://s84.photobuc...X4_1st_day1_001.flvGo to www.revetec.com for info [revetec, Apr 05 2007] I must be missing something... If they all go "down" at once, what pushes them back "up"?-- justaguy, Jun 01 2005 And they thought the BMW four-square had a strange vibration...-- normzone, Jun 01 2005 Hey justaguy: They use pistons, con-rods, and cranks just like any other engine. Imagine an in-line three cylinder. Now move two cylinders down and apart 120 degrees. The whole engine should be shaped like an upside-down letter Y.-- discontinuuity, Jun 01 2005 I understand what you're describing. What I don't understand is what force moves all 3 pistons back "up" once combustion has occurred and the pistons are forced into the "down" position.
Maybe I'm just unclear on how this normally works in an ICE.-- justaguy, Jun 01 2005 [justaguy], the cycle brings the piston back up into compression. That's no issue... I don't think. The problem is having them all fire at once. Why??? You're just asking for something loud and uncomfortable. having three cylinders focus on a common point will relieve alot of the vibration in the chassis, but not all of it, one slight of a cylinder and your front end will be jumping all over the place. And, the power will still be jumpy and the noise unbearable. And, you're wasting alot of space with this, which makes it a poor canidate for motorcycles.. Perhaps... Oh yeah, I just remembered I know very little about cars. Later!-- daseva, Jun 01 2005 The point is, though, that the idea describes all 3 cylinders firing at once as a central feature. Whys aside, I don't think this would work. The cycle, in this configuration, would be lopsided - I don't think there would be enough inertia in the pistons to return them to the top of the cylinder, and there's nothing else pushing them there.
In other words, try changing the ignition timing of your car to fire all cylinders at once. Let me know what happens.
I take it all back - this has to be how a single-cylinder engine works.-- justaguy, Jun 01 2005 No, only two cylinders would fire at a time. If the bottom two are in the combustion stroke, the top cylinder is in the intake stroke, and vice versa. The three-cylinder would be rough, but no more than a Harley (which fires one cylinder right after the other), and the six-cylinder would be twice as smooth.-- discontinuuity, Jun 01 2005 [no rougher than a Harley] is not exactly a compliment [ducks in anticipation of the response from the Harley crowd]-- normzone, Jun 02 2005 I'm not one of the engineer/mechanic types who engage in long discussions about optimising the venturi effect etc., but this seems like a really nice engine design. It makes sense and feels nice. Have a pastry.-- wagster, Jun 02 2005 What do you mean "gears"? There is a single crankshaft with three cranks.-- discontinuuity, Jun 02 2005 Sorry, this thing would still be quite out of balance. Here's why: if you have 3 pistons going 'down' at the same time, then you need to have a crankshaft with 3 120 degree sections, and hence you no longer have a true radial, you simply have an inline with the cylinders set at funky angles to eachother. This will cause massive fore-aft inbalance problems. Got a cure or a crankshaft design that can accomodate all 3 cylinders?
Another drawback will be the massive flywheel and a crank that will be needed to handle the stresses of fully-reversed torsional loading. It'll run so rough from not being timed sequentially that it'll blow itself apart, I'm afraid.-- RayfordSteele, Jun 03 2005 I still don't get it. You could use fork shaped con-rods like old Harley V-twins, that would eliminate fore-aft balance problems I think. Or you could make it a six-cylinder, that might help. I doubt it would be any harder to balance than a three-rotor Wankel engine; you might need a few counterweights and a flywheel, but no more than any other engine.-- discontinuuity, Jun 03 2005 Hmm... Maybe. If you have a forked rods that stagger eachother, and a 120 degree crank, you're going to need some support bearings in between the crank sections to take the loads. It might package, if a little niggly.-- RayfordSteele, Jun 03 2005 If you do have a CAD model then check if you can export to :obj or vrml and I'll post some renders.-- not_only_but_also, Jun 03 2005 How do I do that? Anyway, it's just a rough sketch, and 2D, so I'm not sure. I would have to make some kind of 3D model first, plus all those files are on another computer.-- discontinuuity, Jun 03 2005 This is not a true radial; it is more like a Ducati V-Twin with an extra cylinder, in that there are three cranks rather than one (radial engines and Harley V-Twins have a single crank).
I know that for several years in the 40s and 50s, Harley engines had forked con-rods on one cylinder; afterwords, they used staggered cranks. Either way would work for this plan as long as it was balanced propperly. With forked con-rods, you would not need any counter-weights, just a flywheel, since each cylinder would be effectively pushing on the same plane.
In Subaru's H6 3 litre engine, four of the cylinders go "down" (towards the crank) at the same time. This would be similar; all three would go "down" at the same time, and two would be performing the same stroke at any given time.
Smoothness could be improved by stacking two engines together to make a 6 cylinder. Three cylinders would be going down, and three up. So at any given moment, there would be two cyninders compressing, one exhausting, two intaking, and one in power stroke. This would be similar to an inline 6 cylinder or boxer engine.-- discontinuuity, Jul 22 2005 Too many parts, too complex, poorly engineered horrifically balanced, impractical power balance, too large a package, difficult oil control, difficult fuel distribution, difficult intake system, difficult exhaust system, unnecessary bulk
what would make this engine better than others? why not use a 3cyl radial?-- shad, Jul 22 2005 I don't see why this would be so badly out of balance: obviously the crank would have to be counterweighted just like in an inline. I also don't understand why you guys are trying to call this thing an inline in a circle, unless you're talking about how the crank would have to be arranged. A conventional radial uses a crown rod (or whatever you might call it) to which all the other rods are attached, and the crown drives the crank.
It definitely would need a big, heavy flywheel to power all the pistons through the other three cycles. There goes your throttle response.
Can you imagine the exhaust note? (Idling)"whomp whomp whomp whomp" (full throttle) "PAPAPAPA!" I don't think Midas or even Borla could fit an exhaust system that could handle these pulses. Maybe a big four inch sewer pipe...-- elhigh, Jul 22 2005 not only do you need to balance out the rotational mass you also need to balance the power delivery, 3 cylinders 4 strokes and two power cycles doesn't balance-- shad, Jul 24 2005 Most of these ideas have been tried before. Search for "W Engine" for one old take on it. To package it like a radial, use a common crank bearing with 3 separate attachment points for the rods.-- moPuddin, Jul 24 2005 every radial configuration has been tested and tried, they just arent as easy to manufacture as conventional engines are which is why inlines and v's are conventional.
again, turbines are the better choice here, much more efficient and less energy lost to friction-- costellogroup, Jun 23 2007 [ Too many parts, too complex, poorly engineered horrifically balanced, impractical power balance, too large a package, difficult oil control, difficult fuel distribution, difficult intake system, difficult exhaust system, unnecessary bulk ]
Yes, but that does make it ideal for the Halfbakery.-- normzone, Jun 24 2007 random, halfbakery