I'm putting this in the RX-8 Discussion Forum because I think it fits here better than the Tech Garage, although it is very "Techy".

I'm sticking my neck out here, and opening myself up to a lot of flaming, but I'd like to try to change the way people compare rotary and cylinder engines. Please do comment though, I'd like your civilized input. Many of you I'm sure will find my points basic and obvious. Also, please point out where I'm wrong.

From ALL of the posts I've read, people describe rotary engines as being smoother and more efficient than cylinder engines because the latter has pistons that move back and forth whereas rotaries have a rotor going around and around and around.

My point is, the rotary engine does move back and forth. Every type of rotational motion inherently has back and forth movement. In fact, this is how pistons are able to create rotational motion in the first place. Going with a cartesian coordinate system, the rotor in a rotary engine is constantly changing direction in both the X and Y directions in such a way to create rotational motion. So, just like how work is required to move the piston back and forth, similar forces are acting on the rotor so that it can move back and forth in both the X and Y direction. In the rotor, it is the sun gear that opposes the forces of the combusting gas that pushes the rotor back and forth, and thus around and around. However, these forces acting on the eccentric shaft are acting in both the X and Y directions. In a cylinder engine, combusting gases are confined to a one dimensional system, moving the piston back and forth, and translating that motion into rotational motion through the connecting rod. So, the rotational motion of a cylinder engine's crankshaft is caused by forces solely in the direction parallel to the motion of the piston. This leads to inefficiencies because the force is not always tangential to the rotating direction of the crankshaft.

Ping pong motion?

No, the piston is not "bouncing" back and forth. This is not a tennis ball being volleyed back and forth. The forces acting on the piston to move it back and forth are very smoothly confined by the connecting rod in simple sinusoidal fashion. At top and bottom dead center the piston is smoothly slowed, and acceleration is smoothly controlled by the rotational motion of the crankshaft.

So why is the rotary engine smoother than the cylinder engine?

In line 4s all act on the crankshaft in the same direction. V6s act on the crankshaft usually about 45 degrees out of phase. We can think of the rotary engine as a cylinder engine with an infinite number of cylinders, spread evenly around all 360 degrees of the rotating crankshaft. Thus, creating smoother rotational motion. Well, this isn't entirely true. Rather, think about the expanding gases of the combustion stroke smoothly acting on the rotor (nearly) orthogonally (tangential to the rotational direction) through 90 degrees of rotational motion. This is partly why rotary engines are smoother. Another big reason is the beauty of fewer moving parts.

-P23

Another observation:
The Renesis engine redlines at 9000 rpm. The S2000 engine redlines at 8000 rpm. However, the rotor of the Renesis engine still only goes around once every 3 rpm. Thus, at 9000 rpm, it is still only spinning at 3000 rpm. In comparison, the S2000 pistons are all travelling back and forth at an amazing 8000 "cycles per minute." To explain this observation, the rotary engine does cycle once per rotation of flywheel, but this is through the translational motion of the rotor, not rotation. So in essence, the rotary engine is a two banger, with two rotors "bouncing around" as some would say, and at the same time rotating once every six "bounces". This is just another example of how similar rotary and cylinder engines really are.

EDIT: Changed "camshafts" to "crankshafts" :p

I think you have it all wrong. If you analyze the movement of the rotors, they have both rotation around their centers as well as orbiting motion around the center of the eccentric shaft, all in the same direction. You can visualize it as an off-center rotation, that's why there's movement in the x and y directions.
Now, pistons move linearly and do change direction everytime the crankshaft rotates. For one revolution of the crank a piston has to move back and forth once, the connecting rod converts this linear movement into rotation of the crank and visceversa. If you think about it, the rod's movement is one of the most complicated inside a piston engine, the crank end orbits around the center of the crankshaft while the piston end just moves back and forth in a linear path. Now imagine multiple cylinders doing the same thing...

Look in here:
http://auto.howstuffworks.com/rotary-engine.htm

Yes I agree, the connecting rod does follow a very complicated path. Would you say that that the crankshaft in a cylinder engine moves back and forth? My point is that anything revolving has back and forth movement. This back and forth movement is demonstrated perfectly by the piston and crankshaft system. This movement of the piston is a direct reflection of the translation in the X direction of a single point on the edge of the rotating crankshaft.

What I'm saying is that the rotational motion of a rotor has these same types of forces. Say you were swinging a weight at the end of a rope through the X-Y plane. The weight is swinging around and around. If you look at movement of the weight in only the X direction, you would see that it is moving from -r to +r (r = radius of circle, length of rope) in a sinusoidal fashion. The forces in the rope, broken down in just the X direction, are thus constantly pulling on the weight to move it back and forth. These forces in the rope, to move the weight back and forth in the X direction, are principally identical as the ones required to cause the back and forth movement of a piston. With the weight and rope system, movements and forces in the X direction are combined with the same back and forth movement in the Y direction, but 90 degrees out of phase, in order to cause rotational motion.

However, the rotor has movement that is much more complicated than that. This is due to the orbital gearing, causing three translational orbits for every one rotation of the rotor.

On page 5 of the howstuffworks.com section on rotary engines linked above, there is a great animation of rotary motion. Now, try and follow the movement of a single apex seal around and around, but only confined in either the X or Y directions. These graphs are what you get, and as you can see, it's much more complicated than that of a piston engine. And yes, there is back and forth motion in both X and Y axes, resulting in rotation as well as translational motion.

The equations I used:
X=1 sin (t) + 6 sin (t/3)
Y=1 cos (t) + 6 cos (t/3)

So, the rotor motion is the addition of two sinusoids.
In principal, the connecting rod movement is also the addition of two sinusoidal movements: the revolutions of the crankshaft, and the back and forth movement at the piston.

Here is a diagram of what you would get if you combined the back and forth movements of both the X and Y axes, which are graphed above. As you can see, the movement of the X and Y axes are quite complicated, with absolute velocity of the apex seal slowing 3Pi/2 and 9Pi/2 time points, at the middle of the long side of the housing.

Would you say that that the crankshaft in a cylinder engine moves back and forth?

No, I wouldn't; just rotation, same as the eccentric shaft.

Nice graphs! how did you come up with those equations? Bringing back some trig and calculus memories!

I agree there is a back-and-forth component to the rotor's motion. You can see this visually in the animations as the "wobble" of its eccentric orbit which is superimposed on the rotational motion. I may be oversimplifying here, but I think of the rotor's orbit as being analogous to the stroke of a piston. The Renesis has an orbital eccentricity of 1.5 cm so the diameter of its orbit is 3 cm. Compare this to the stroke of a piston which is typically 2-3 times this amount and I believe this is where a lot of the extra smoothness comes from.

Having said that, a rotor is probably heavier than a typical piston and con rod so that might make up much of the difference. But in any case a piston engine has far more things moving around in complicated ways which creates more harmonics which usually produces a harsher sound.

Since I haven't done enough analysis on this yet I will have to give just a couple thoughts on the matter. First an interesting note about the rotary is that the rotor does not travel in a pure circular pattern. But the major thing is that in a piston engine every cylinder hits a singularity point twice in one rotation. So this being that you reach a point in the crank shaft in which a change in direction is necessary.

Whereas in a rotary you are traveling in a constrained and slightly elliptical path. understanding the path it follows gives us a really simple jacobian of the equations of motion. Point being is that a the rotary follows a smooth set of geometrical constarints with minimal hard changes in direction. again I could be wrong right now but as a first look this is what becomes evident

hmm this gives me an idea.. maybe I should fire up my Matlab Students edition and derive the equations of motion for the rotary engine.. yes yes really nerdy but it would be fun. Bring out all the advanced dynamics I learned in school!

Originally posted by portero23
Yes I agree, the connecting rod does follow a very complicated path. Would you say that that the crankshaft in a cylinder engine moves back and forth? My point is that anything revolving has back and forth movement. This back and forth movement is demonstrated perfectly by the piston and crankshaft system. This movement of the piston is a direct reflection of the translation in the X direction of a single point on the edge of the rotating crankshaft.

What I'm saying is that the rotational motion of a rotor has these same types of forces. Say you were swinging a weight at the end of a rope through the X-Y plane. The weight is swinging around and around. If you look at movement of the weight in only the X direction, you would see that it is moving from -r to +r (r = radius of circle, length of rope) in a sinusoidal fashion. The forces in the rope, broken down in just the X direction, are thus constantly pulling on the weight to move it back and forth. These forces in the rope, to move the weight back and forth in the X direction, are principally identical as the ones required to cause the back and forth movement of a piston. With the weight and rope system, movements and forces in the X direction are combined with the same back and forth movement in the Y direction, but 90 degrees out of phase, in order to cause rotational motion.

However, the rotor has movement that is much more complicated than that. This is due to the orbital gearing, causing three translational orbits for every one rotation of the rotor.

On page 5 of the howstuffworks.com section on rotary engines linked above, there is a great animation of rotary motion. Now, try and follow the movement of a single apex seal around and around, but only confined in either the X or Y directions. These graphs are what you get, and as you can see, it's much more complicated than that of a piston engine. And yes, there is back and forth motion in both X and Y axes, resulting in rotation as well as translational motion.

The equations I used:
X=1 sin (t) + 6 sin (t/3)
Y=1 cos (t) + 6 cos (t/3)

So, the rotor motion is the addition of two sinusoids.
In principal, the connecting rod movement is also the addition of two sinusoidal movements: the revolutions of the crankshaft, and the back and forth movement at the piston.


nice graphs what did you use? only thing to add is the real importance is not to look at X and Y but look at the change in r (the radius) So polar coordinates would be the correct usage in determining the changes of rotational momentum. If you can determine the path vector of a point on the apex seal then you should be able to determine the equations of motion of the rotary engine much easier than that of a piston engine. (less rigid bodies to account for)

To confuse you all further do a google search (or whatever search engine you use) about quasiturbine engines. Bizzare stuff!

Claims to have all the mechanical benefits of rotary and all the thermal benefits of piston/cylinder.

I'll save you from searching...

Go here and get very confused!! :D

http://deadbeatdad.org/eliptoid/

Be sure to look in every link! There are diagrams and animations of some weird non-wankel rotary engines.

I actually added a little nutmeg to my cream and sugar to give it a rougher edge. The engine is pretty smooth normally but the nutmeg just adds a little something. Come to think of it I add it to my chocolate chip cookies as well and they are a big hit too. Maybe that's why everyone is so interested in what powers my baby under the hood ;) .