Mazda rx8 wankel efficiency nearly there? 7

[harveygp]

Looking at the fuel consumption of the rx8 in relation to the power output, it seems to be comparable with piston engine sport cars of similar (high) performance.

Doesn't that suggest that say a single rotor version of half the output (say 110hp) but in average use having a wider open throttle and only half the frictional loss would give a very acceptable power unit for a small saloon with good economy too.

The improvement over the previous rx7 engine seems very impressive but has been got through careful attention to detailed understanding. So the basic engine topology has much more to offer that anyone would have guessed a few years ago.

Should the industry in general be taking a new look at the Wankel topology? Maybe it already is?

Any comment from those who know a bit more than me/

 

[ghamish]

I should point out that the rotary in the RX8 is not turbocharged.

 

[BramCohen]

If the big problem with the wankel is surface to volume ratio, couldn't that theory be run in reverse, as an improvement to the piston? I'm a total n00b about this stuff, but I'm guessing that a long thin piston suffers from too much surface area when fully extended, and a short stubby one has too much when fully compressed, so real ones pick a height/width ratio at some compromise value. If a 'piston' could be designed with a compressible chamber which stayed the same shape during its entire cycle, couldn't it in theory have near optimal surface/volume the entire time? Would that significantly improve engine efficiency?

 

[harveygp]

My view is that that the surface to volume thing may be out-of date dogma thats amounts to a lazy way to rubbish the Wankel concept. If you look at the fuel efficiency of the RX8 it is a huge advance on the RX7 and comparable (or better) than piston engine powered cars in the same power range and market segment. More important, the efficiency must surely now be way better than the state of the art with piston engines back at the time(about 40 years ago)folks first put forward the idea that surface to volume represented a fundamental barrier to acceptable efficiency for the Wankel topology. Clearly, with the benefit of hindsigh, that view was wrong then and maybe wrong now. Since that time I would guess at least 100X the resource has into making the piston topology more efficient compared with the resource expended on the Wankel topology.

 

[GraviMan]

Interesting HarveyGP. I had wondered how much truth was in the surface area to volume thing. An interesting point has been raised in a thread on 2-strokes that the power output is limited to head temp in a piston engine anyway. I imagine that the rotary piston width is very well optimised with this in mind.

Since the "big end" is epicycloidal, the seals can run very hot with no real wear. I imagine that cooling circuits are also optimised to flow in opposite direction to rotation, so as to assist heat transfer. As you say Mazda is keeping up the research that everyone else forgot. Pity someone like Ford doesn't give it another look...

Mart

 

[JrGman7765]

New guy here with a question... I have a new RX8. Its got roughly 8800 miles on it. I added a bottle of octane booster to the tank about a week ago. I ran it through hitting a top speed of 135+/- before i ran out of road. After the tank was drained I refuled and started noticing a hrad idle. The exhaust was "puttering" and the car sounded as if it were going to stall. It has been doing this for a week now, even after 2 refuelings and a bottle of injector cleaner. Could the octane booster have caused a problem with my cat or an O2 sensor, even though it claimed the booster was safe on those componants? PLEASE HELP.

 

[TOHCan]

As a former RX-7 owner, I can relate a few real world points. Mine averaged about 25 mpg with a 4 speed at somewhat elevated speeds (5000 rpm is too smooth to intrude as a "speed alarm"). The main failure mode is having the end seal finally wear enough that it flips out and at the least drops the compression between two of the lobes on that rotor. No warning or symptoms.

I always considered my 11A engine to be a 2.2L six, and found the fuel economy and power levels to roughly coincide with piston engines of the same displacement.

Someone else beat me to the comments about "VTEC"-ing a rotary.

 

[magnograil]

I would like to point out the engine is actually a Umblebee (sp?), invented in the UK in the early 1900's. Wankle took out his 'patent' the day Umblebee's expired. It was not until the kinematics was inverted by the head of NSU's design (forgot his name) did Wankle's engine become useable.
As far as aircraft use, check out Midwest engines who are building a version of the Norton motorcycle engine for aircraft. I have a Norton rotary and building a copy of their NRS588 race bike. BSA originally bought the license (which were based on power output) and built the prototypes. Norton inherited it and developed the liquid cooled version.
First difficulty when comparing to recipricating engines is measuring the displacement. The rotor turns 1/3 the crank (eccentric shaft) speed. I believe Mazda uses 1/2 the swept volume of the rotor. That, I believe is the reason for their quoted high power/displacement. They are lighter however without the valve gear. Comparable to a two stroke. This one uses periferal ports and air cooled rotors. I think Mazda were side ports which had more control over the exhaust/intake overlap.
As far as piston surface area, F1 engines are about 2:1 bore stroke. If area was a problem they would not have such large piston areas. Of course, efficiency is not a prime directive of F1.

 

[harveygp]

If I remember correctly, according to the book by Jan Norbye, there was a lot of debate in the early days about how one should compare engine sizes until someone called R.F.Ansdale "proved" at a Society of Automobile Engineers (?) that the Wankel was equivalent to a 2-stroke (!) of the same swept volume. The guy was either loosing it or had some vested interest, in my humble opinion.

I'm firmly with TOHCan in his last posting. To me it's obvious that the mechanical details (shaft speed etc.)are irrelevant and that to make a fair comparison one should look at the number of simultaneous volumes being swept on a four stroke cycle. So TOHCan is absolutely right. A 2 rotor Wankel is should be compared to a six cylinder piston of capacity six times the swept volume associated with each rotor face. That's not how Mazda spec their engine though.

On a similar theme, many journalists rave about how the Mazda "redlines" at 9000 rpm. But that is the shaft speed and equates to the same number of power impulses as a conventional six cylinder engine at 6000rpm - so nothing special really.

It's a shame that these stupidities persist because it makes ready comparison rather difficult.

(BTW, I think the NSU guy was called Walter Froede)

 

[microhenry]

Am I ill informed, or is there no talk about a hybrid wankel?
I would think that the wankel is the ideal engine for an electric hybrid.

I'm told the wankel will never match a piston because the
lobes don't capture enough of the expanding gas.
Add a big turbo to get back the efficiency, and add an electric
motor to eliminate tutbo lag and add low-end torque.

What am I not seeing?

 

[magnograil]

The surface area problem can be corrected with ceramic coating. Same problem occurs with large round or oval pistons, the surface area picks up more heat from the combustion which has to be cooled by oil jets on the lower surface.
Some believe the Wankle is neither two or four stroke. However, it evacuates the chamber through the exhaust port before opening the intake to refill the chamber like a four stroke. A two stroke uses the incoming charge to help blow out the exhaust which a Wankle does not.
The exhaust/intake overlap emmissions can be eliminated using the same technique as Lotus with their new two stroke. Move the port to later in the cycle and use a supercharger to provide the quantity of flow.
From all accounts of the Norton race bike, low end torque was not a problem and had a wider power band than the piston engines. Adding a turbo eliminates it from most racing classes.
Until batteries become more efficient, electric hybrids will be less efficient than mechanical hybrids.
Harveygp - thanks for the name.

 

[patprimmer]

microhenry

What you are not seeing, is that if you do all these things to a piston engine, they also improve, and thereby maintain an advantage, if developed to the same extent with the same technology.

The inherent disadvantage of the rotary engine is it's high surface to volume ratio. This still applies no matter what else you do to improve it.

There is no argument that it has very good power density, which can result in a lighter car which can help up to the point that this becomes offset as you also need to carry more fuel to maintain range. What range you require becomes critical to the comparison.

Ceramic coatings reduce the losses in piston engine and rotary engines, but the rotary has more scope for improvement as it has higher losses. This does not mean it will end up better in this regard, just worse by a smaller margin.

The fuel efficiency advantage is in the reduced inertia in the rotor vs a piston, but a rotor at constant rpm still has some losses as the rotor moves back and forth across the dog boned shape housing as it rotates.

I have no emotional support for either design concept, as I have had a lot of fun driving both rotary and reciprocating piston powered cars and boats.

By the way, suck, squeeze, bang blow adds up to 4 cycles in my book, no matter what mechanical configuration or valving you use to achieve it.

In a 4 stroke, all 4 are reasonably separate processes, and all occur within the combustion and positive displacement area. In a 2 stroke the suck only occurs in the exhaust system, and suck and blow occur virtually simultaneously outside the positive displacement area.

This clearly, at least in my mind makes a rotary a 4 stroke.

The capacity should be calculated by the displacement of air during one complete cycle or 4 strokes of the engine. This means a twin rotor should be treated as a 6 cylinder, and rotor rpm not crank rpm is the real issue in determining engine speed.

Regards
pat pprimmer@acay.com.au
eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[magnograil]

Patprimmer, perhaps it was just the way you stated it, but the rotary does not move back and forth across the housing. The rotor is mounted by a bearing to the eccentric shaft and connected to the housing by a gear. The rotor center turns in a circle and rotates backward from the eccentric shaft. The curve enscribed by the tips is an epicycloid.
The same can be done with a four sided rotor and three lobed housing or any higher number.
I agree that a fair measurement of its capacity is the total swept volume, whether the volume is moving radially or along a single axis and regardless of the number of shaft revolutions it takes to accomplish it.

 

[hydrae]

JrGman
As another former RX-7 owner,
Did you need the octane booster to the the car to that speed
The octane boost would change the flame shape and may have overheated some of the seals.
Did you bypass the rpm interlock? I assume it has one just like the rx7 at 7000 rpm (loud buzzer and ignition cutout)
Though distasteful and annoying it is what forces you to make the engine last longer.
Have you looked at the tip seals? Remove the plugs and bump the starter until you see the lobes, there may be a relubing procedure to get them to reseal again. Or you may have breakdown and rebuild it. The manual direct from mazda for the rx7 was complete including how to replace all those seals (24 components per rotor)

Harveygp, have you driven a RX? they unlike any piston engine, the power is smooth and does not fall off telling you it is time to shift, the engine will continue to produce more power above the redline than spec'd. Why the journalists were raving about the redline was not about the number but about the power and the feel of the power at redline.

I always thought the engine was comparable to a 4 cylinder you get two bangs per revolution of the crank (the crank is spinning at 3 times rotor speed)

Hydrae

 

[harveygp]

Hydrae
I'm very entusiastic about rotaries, so regarding the redline, I'm only saying that if we could compare apples with apples more easily, we would be able to get a better idea of how (say) the RX8 compares with other cars in the class.

For example, currently if a journalist were to compare bhp at 6000rpm of (say) a BMW six cylider and (say) RX8, that would be unfair to the RX8 since it has 1/3 less power impulses per second.

Crank position is a red herring in terms of determining the equivalent number of cylinders, which most of the recent postings seem to concurr should be six, based on the number of phased four stroke cycles simultaneously repeating. This is important, not just academic, since it would clearly be a nonsense to compare a six cylinder engine with a four cylinder engine using the same piston sizes.

We need to compare like with like to get a true idea of whether for example the intrisic porting advantages and zero valve gear losses of the Wankel outweigh (or not) the disadvantages of higher surface to volume ratio and seal frictional losses.

The engine size confusion has probably had the more damaging effect on the fortunes of the Wankel topology, long term, since it resulted in the engine being outlawed from many forms of motor sport. It's clearly ludicrous to say the Mazdas are equivalent to 1.3 litres.

For example, just pasted the following from a review, but what does it mean? Completely hopeless interms of comparison.

>Mazda RX-8 Specifications
>
>Engine
>1.3L displacement gas engine, 210 hp @ 7200 rpm 164 ft->lbs. @ 5000 rpm, premium unleaded fuel

should perhaps read

2.6L (or 3.9L - not sure, anyone know?)gas engine, 210 hp @ 4800 rpm(equivalent) 246 ft-lbs(equivalent) @ 3300 rpm(equivalent).

Now we're getting somewhere, surely?

I had a ride in an RX8 recently and its an ambition to own one. Never driven one, but look forward to it.

 

[microhenry]

patprimmer

I agree that much can be done to improve the piston engine.
The question I'm posing is: on a weight & total engine size
basis is the wankel not inherently and insurmountable
superior to a 4 stroke piston? Likewise isn't the piston
inherently and insurmountable superior to a wankel in terms
of fuel efficiency?

If the above is true, then a hybrid wankel overcomes the
efficiency deficit, whereas a hybrid piston will gain
efficiency at the expense of weight & size.

 

[zzonbi]

For a 360 degrees rotor cycle there are 6 pulses on the Renesis, so it's more like a 12 cylinder, not 6. The whole 1.3l volume is processed, so we have to make the equivalent V12 2.6l. However, it will only have 1/3 of the quoted 10000rpm, hence 3x the torque value. Actually the high shaft revs are an advantage as it allows a smaller transmission to follow.

The big handicap is surface/area. On the bright side simplicity, dimmensions, robustness and fewer losses, but still insufficient to overcome consumption, regardless of the claimed figures. It is still rather unnefficient. A cool concept though, still hugely underdeveloped vs piston one.

PS So the equivalent is a 2.6 V12 with 250HP at 3000rpm?? Looks suspicious? That's because it is, the actual swept volume is 2x654ccm per shaft rev, which makes the equivalent V12 8.4l ! Now the comparison is fair.

 

[facty]

I have to agree with Harveygp - its the comparison of the cycle times that must be considered. The RX8 has 2 rotors and therefore 6 working volumes. One complete 4-stroke cycle takes place in 3 output shaft revolutions as opposed to 2 in the conventional reciprocating engine. So, if the output shaft of a Wankel engine is turning at 9000 rpm then this will equate to 3000 cycles per second. The crankshaft of a reciprocating equivalent would have to be turning at 6000 rpm to produce 3000 cycles per second.

So, it is reasonable to say that the RX8 engine with the following specs (detailed above):
210 hp @ 7200 rpm 164 ft->lbs. @ 5000 rpm, premium unleaded fuel.

could be equivalent to a 6-cylinder reciprocating engine with total volume 3.9L and the following respective specs as also given by Harveygp):
210 hp @ 4800 rpm 246 ft->lbs. @ 3333 rpm, premium unleaded fuel.

In this light it looks rather poor!! Another point worth noting is the Fuel Consumption quoted for the RX* as this also helps give the game away -: 18 to 24 mpg. Its certainly no 1.3L!!

 

[GregLocock]

Simplistically

In one minute the output shaft goes round N times
In one minute it makes X bangs
In one minute it inhales V m^3 of air at 100% VE

Why make it more complex than that?





Cheers

Greg Locock

 

[zzonbi]

It has 6 chambers, but fired twice as frequent. It is extremely well balanced too, just like a V12. So per rotor revolution it fires as a 12 cyl.

The engine wear is more related to rotor speed than shaft speed, 10000rpm is used mostly for bragging.

In my opinion 654ccm is the volume of a single chamber of the 6. By the time this volume is cycled out, so do the other 5 volumes. That would suggest 3.9l. However, an equivalent classic 3.9l cycles its internal components twice in this interval, so from this perspective it would take a 7.8l, working at the rotor speed.

250hp from a 7.8 is way lower than modern 100hp/l standards, but the 1/3 actual rotor speed helps explaining that.


Simplistically

In one minute the output shaft goes round N times
It goes 10000 times.

In one minute it makes X bangs
It bangs 20000 times

In one minute it inhales V m^3 of air at 100% VE
It inhales 654*20000*10^(-6) m^3, or 13 m^3

Why make it more complex than that?
Did it get simpler?

In the end it depends on the equivalence factor used. If it is just output shaft speed, the engine is like a 2.6 at 10000rpm, but considering also wear and smothness it is rather a 7.8

And I am not sure any polygonal shape can be used as rotor. The Wankel rotor is a constant width figure, a Reuleaux triangle.

 

[GregLocock]

So its black box equivalent is a 10000 rpm 4 cylinder with a capacity of 2.6 litres

Any other conclusion is marketing, not engineering.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.