Use Mozilla Firefox as your browser. It spell-checks as you type.
And from that link, the wankel engined vehicle category.
The Mazda Cosmo was pretty funky looking in its first and last incarnations.
Well, I always thought the NSU Ro 80 was a classy looking car. A slightly nutty design, though. NSU must have been like the German equivalent of Citroen.
That’s pretty much exactly what Consumer Reports said about the RX-8, too. 
I’ve been seriously thinking about getting one, but now this thread has me concerned about gas and oil consumption … and a possible engine rebuild as early as 60k miles! :eek:
I think those worries are overblown (except for the gas mileage, but you don’t buy a sports car with the expectation of great gas mileage). Personally I’m thrilled about having to get my timing belt (chain?) changed out every 60k miles on my current car. I can’t afford it now, but I’m already on borrowed time and I’m afraid if it breaks I will have a big pile of worthless metal instead of an engine. I would prefer the gradual degradation of the seals on a Wankel engine that allows me to choose when to perform a PM.
It was classy. The space shuttle is an impressive piece of engineering.
>the pistons themselves […] come to a sudden halt, reverse direction, halt again, reverse again.
>the wear and tear on these kind of components is considerably greater than a rotary, electric motor, for example.
>If you don’t believe me, I suggest looking into some basic mechanical engineering concepts.
>Give it up dude.
Thank you for that.
The pistons don’t come to a sudden halt. They move almost sinusoidally.
Speaking of wear and tear, the downfall of the Wankel was wear and tear on the seal, wasn’t it?
I’m familiar with basic mechanical engineering concepts. Less than 24 hours ago I was helping a mechanical engineer do a finite element analysis on stress and strain relationships in a cylinder internal wall.
If Wankel engines work so well, why does this thread start with “What ever became of the Wankel”?
Perhaps your engines work differently from this Otto cycle animation.
If you’re talking about any motion other than straightline due to tolerances, I’m sure that’s the case, but only of interest to someone designing an engine from scratch or a theoretical physicist. From a practical stanpoint, any way you look at the piston action, it’s straight line with a reverse at each end of travel. There’s no way it can reverse without coming to a stop first. Please explain why you think otherwise.
Yup – Isn’t that one reason given here already?
One of the benefits of the tiny relative size of the rotary engine is that it can be mounted much farther back (in a front-engined configuration) than a piston engine- behind the front axles rather than over it.
The 92-98 RX-7 was the best-handling car in its class mostly because it was able to achieve an almost perfect 50/50 front/rear weight distribution. This was particularly impressive given that Mazda was not noted at the time for producing cars which handled well.
The main weaknesses of the Wankel were the seals, which got improved with development but are still very problematic - and IIRC the thermal inefficiency of having a relatively large surface area in the combustion “chamber”
Seriously, you’re going to lose an argument that the Wankel isn’t inherently smoother than a reciprocating piston engine, just look at RPMs
Easy answer - they produce as much power as a piston engine of twice the size, but consume as much fuel as a piston engine three times the size.
I don’t how you can say that a piston inside a straight sided cylinder with only .01mm clearance between the piston and wall moves in anything buy a straight line. :dubious: 
As I mentioned in the RX-8 thread the wankel has good and bad parts.
Good:
Great power
Smooth as can be
Very compact (the motor on our RX-7 race car was about the size of a beer keg)
Bad:
Drinks gas like there is no tomorrow
Very dirty engine (exhaust wise)
Uses oil to keep the rotor tip seals alive
I think the life span issues are pretty much solved by the latest version of the engine, but you won’t see Irv Gordon like reliability out of a rotary.
What killed the rotary was the tip seal issues on the early engines, the fuel consumption, and the exhaust pollution. Rotor seals have pretty much been solved IMHO, but you can’t fix the thirst and inherent pollution issues.
In fairness, I think he means the acceleration is sinusoidal (plot of speed against time), not the path of the piston
However that still involves an abrupt, though not instantaneous, reversal of travel
The inherent problem with the modern rotary engine is the limited parameters for modification compared to a standard piston ICE motor. This makes it harder to modify the engine and still meat emissions standards. This is what killed the rotary Corvette.
In a modern piston motor it is possible to change intake and exhaust timing, intake/exhaust volume, combustion chamber design, compression, ignition timing, and piston diameter to stroke ratio. It is much more tunable and thus can be modified for a variety of applications.
Rotary engines require a very specific chamber design, which limits the parameters listed above. Given the poor fuel economy and low torque characteristics of such a motor it narrows the applications that would benefit from such an engine.
GM was experimenting with a 4.4 liter 2 rotor and a 6.4 liter 4 rotor for the Corvette but could not meet emissions in the early 70’s. I would imagine the fuel economy of a 6.4 liter rotary would have been ugly by any standards.
Where the rotary engine is most useful is in applications where a small, light weight engines are required. The only thing that really sticks out in my mind is an aviation application and you see it in the homebuilt market for 2-seat helicopters and the illusive Moller Skycar.
Nitpick: Porsche’s 968 used a 3.0L inline four, and balancer shafts were developed by Lanchester - Mitsubishi just reintroduced them.
What he is saying is that the position derivatives (velocity and most importantly acceleration) are quasi-sinusoidal functions; that is, that as the piston approaches TDC it is already slowing. Of course, you still have force balance, and because (even in a flat engine) the cylinders aren’t directly in opposition, you still have vibration. You also have this with a rotary engine, although the character of the vibration is different (and can be countered entirely via a counterrotating flywheel or second rotar). In general, a rotary engine is going to have a lower amplitude of vibration in any direction even if the “quantity” of energy lost to vibration is larger.
Stranger
>There’s no way it can reverse without coming to a stop first. Please explain why you think otherwise.
Well, I don’t, of course. I said they move almost sinusoidally. Their acceleration rises smoothly to a maximum value at the same time they come to a stop and start moving the other way. Right around top dead center or bottom dead center, their linear motion is like something thrown straight up into the air, with a fairly constant acceleration, and a velocity that is first getting smaller, then passing through zero, then getting larger in the other direction. This is smooth motion like a sprung mass as a harmonic oscillator.
>I don’t how you can say that a piston inside a straight […] anything buy a straight line.
>In fairness, I think he means the acceleration is sinusoidal (plot of speed against time)…
>What he is saying is that the position derivatives (velocity and most importantly acceleration) are quasi-sinusoidal functions; that is, that as the piston approaches TDC it is already slowing.
Yes, I mean the acceleration is sinusoidal, that is has the functionality KSin(ctime). And the velocity and the linear position along the length of the cylinder are also sinusoidal. All of these are sinusoidal functions with the same frequency and with phases a quarter turn apart. Well, I was careful to say “almost sinusoidal”. In the limit as the connecting rod gets long, it approaches a sinusoidal function, but the angling of the connecting rod in a real engine makes it a bit different, hence Stranger’s “quasi”. However, this is a bigger effect in mid stroke anyway, not at the ends where the piston stops.
>still involves an abrupt, though not instantaneous, reversal of travel
Don’t think I quite follow you here. I think you could say exactly the opposite: the direction of travel changes in an instant, or in an arbitrarily short period of time. But there is nothing abrupt about the motion. The piston’s velocity, very closely around the time it stops, would graph as a straight line crossing zero. The piston’s acceleration over this same period would graph as a straight and horizontal line, unchanging with time. Since it is acceleration that has to do with force here, the fact that it is a constant would mean it isn’t abrupt at all.
>Perhaps your engines work differently from this Otto cycle animation.
They do.
This animation is pretty funny. They didn’t bother to reproduce the piston movement as sinusoidal. They’ve got the piston moving at a constant velocity of alternating sign, and I bet that is because it is very easy to do with whatever computer animation tool they were using. If you watch the connecting rod carefully, you see that it stretches approaching top dead center and then shrinks just afterwards, to let the piston continue to rise at a constant rate and then suddenly fall at the same constant rate while the crankshaft end of the rod is following a circular path at constant speed and therefore moving with zero vertical rate. So, it’s pretty artwork that somebody clearly fooled with at some length, but the motion is obviously wrong.
>They come to a sudden halt…
>If you don’t believe me, I suggest looking into some basic mechanical engineering concepts.
“The velocity of the reciprocating parts is
c = v(sin a + r sin 2a/2l sqrt(1 - (r^2/l^2) sin^2 a))”
-Mechanical Engineer’s Handbook, Lionel S. Marks
c = velocity, a = angle, v = rotary speed, r = crank radius, l = connecting rod length.
Note that this approaches a sinusoidal function in the limit as the connecting rod length l gets much bigger than the crank radius r.
To extend on this, the maximum force imparted by the piston upon the crankshaft (and thus on the bearings and from thus into the motor block finally motor mounts) is somewhere around mid-stroke (for inertial strokes), near full compression (for compression stroke), and at some point soon after ignition (for the power stroke). The positions on the non-intertial strokes vary with RPM which is why Otto-cycle enignes in general are so hard to balance (although two-stroke engines are even harder), as you end up with several different dynamic vibrational modes going on at once. Even true boxer engines (in which directly opposing pistons reach TDC simultaneously) still have secondary dynamics coming from piston lateral loads and reactions throught the crank, even though the piston dynamics are themselves perfectly balanced. Pistons do not “slam” at end of travel–at least, not in a properly running engine they don’t–because if they did they’d be exceeding the triboelastic pressure of the oil lubricating the crank and wearing directly on the engine bearings, and wearing out the relatively low strength bearings on the connecting rods in short order.
I can’t speak with experience to the dynamics of a rotary engine, since my internal combustion engine classwork only extended to reciprocating piston and turbine engines (rotorys generally considered to be an exotic sideline in the study of IC engines) but from my more cursory review of them it appears that a single rotor is unbalanced by its mass by the arm of the eccentricity from shaft center. This could be compensated for in plane by a single compensating countermass placed in direct opposition (though you would then get a secondary torque along the length of the shaft), double compensating half countermasses placed in opposition (providing balanced torque along the axis), or two or more rotors placed so as to oppose off-axis torque of the system. Given how a rotary works it looks like it would be more simple to balance than a recip engine, and that the character of the vibration would itself be rotational rather than linear (i.e. that the resultant force is tangental to the motion of the moving part rather than in-line with it), owing to the fact that only a small portion of the motion of the rotor is radial as opposed to rotational, whereas with a recip engine all of the motion of the piston is radial, which then has to be converted to rotational motion.
So while I agree with Napier regarding the motion that a piston in a recip engine undergoes is close to sinusoidal (not quite because of the varying angle of the conn rod), the rotary engine clearly has less radial motion to balance and convert. However, there are a few of major advantages that have served to make piston engines more popular, to wit:[ul][li]Easier to manufacture: despite the larger number of parts and bearings, every mechanical joint or path except for the crankshaft and camshaft are either straight or cylindrical. The rotary engine, on the other hand, requires an epitrochoidal chamber and a Reuleaux triangle, and requires them to be machined to very high tolerances to achieve good sealing and prevent wear. Although this can be done to requisite dimensional tolerances on modern CNC milling equipment, it formerly required very specialized tooling and measurement to achieve.[/li][li]Seals and sideloads: the pistons on a reciprocating engine see very little sideloads; hence the piston rings, if properly designed, can effectively seal the cylinder and yet last for billions of cycles. The seals on a rotary engine, however, run at ninety degrees to rotation and see wear on every cycle. This requries more lubrication–hence the demand for lubricating oil and inherent dirtyness of exhaust–and means that the seals wear out sooner. Contary to flex727’s experience, I’m personally pretty disappointed in an engine that requires a rebuild every ~60k miles; most modern rubber timing belts are rated at twice that mileage, and timing chains are generally assumed to have infinite life. I’ve had flat-type engines (Subaru) last >200k miles without rebuild or even major mechanical work. [/li][li]Thermodynamics: there is a lot more wall (and time) to radiate heat away in a rotary than in a piston engine, even when you account for the surface area of multiple pistons. Non-adiabatic rotarys have an inherent thermodynamic disadvantage on that basis, which sucks away efficiency. A rotary has high specific output, but the cycle itself is much further away from an ideal Carnot cycle than the Otto or two-stroke. [/li][li]Experience: because of the above issues (especially manufacturing) we’ve had over a century of building recip piston engines, but only about half a century of limited experience with rotary engines. This means that engineers and mechanics are less familiar with the benefits of the rotary engine and have spent less time mitigating its undesireable aspects. Piston engines, on the other hand, have been refined into an art with devoties to every workable configuration, even the sad sack straight-4 (which should be deep-6’ed IMHO).[/ul][/li]
The rotary engine is unlikely to see broad use, but in applications in which it packages well or its high specific output outweighs its disadvantage it can shine.
Stranger